Structural Mechanics Module

User’s Guide
Structural Mechanics Module User’s Guide
© 1998–2017 COMSOL
Protected by U.S. Patents listed on www.comsol.com/patents, and U.S. Patents 7,519,518; 7,596,474;
7,623,991; 8,457,932; 8,954,302; 9,098,106; 9,146,652; 9,323,503; 9,372,673; and 9,454,625. Patents
pending.
This Documentation and the Programs described herein are furnished under the COMSOL Software License
Agreement (www.comsol.com/comsol-license-agreement) and may be used or copied only under the terms
of the license agreement.
COMSOL, the COMSOL logo, COMSOL Multiphysics, Capture the Concept, COMSOL Desktop,
LiveLink, and COMSOL Server are either registered trademarks or trademarks of COMSOL AB. All other
trademarks are the property of their respective owners, and COMSOL AB and its subsidiaries and products
are not affiliated with, endorsed by, sponsored by, or supported by those trademark owners. For a list of such
trademark owners, see www.comsol.com/trademarks.
Version: COMSOL 5.3

Contact Information
Visit the Contact COMSOL page at www.comsol.com/contact to submit general
inquiries, contact Technical Support, or search for an address and phone number. You can
also visit the Worldwide Sales Offices page at www.comsol.com/contact/offices for
address and contact information.

If you need to contact Support, an online request form is located at the COMSOL Access
page at www.comsol.com/support/case. Other useful links include:

• Support Center: www.comsol.com/support

• Product Download: www.comsol.com/product-download
• Product Updates: www.comsol.com/support/updates
• COMSOL Blog: www.comsol.com/blogs
• Discussion Forum: www.comsol.com/community
• Events: www.comsol.com/events
• COMSOL Video Gallery: www.comsol.com/video
• Support Knowledge Base: www.comsol.com/support/knowledgebase

Part number: CM021101

C o n t e n t s

Chapter 1: Introduction

About the Structural Mechanics Module 18

Why Structural Mechanics is Important for Modeling . . . . . . . . . 18
What Problems Can It Solve? . . . . . . . . . . . . . . . . . . 19
The Structural Mechanics Module Physics Interface Guide . . . . . . . 22
Common Physics Interface and Feature Settings and Nodes. . . . . . . 25
Geometry Levels for Study Capabilities . . . . . . . . . . . . . . 25
Where Do I Access the Documentation and Application Libraries? . . . . 27

Overview of the User’s Guide 30

Chapter 2: Structural Mechanics Modeling

Study Types 35
Introduction . . . . . . . . . . . . . . . . . . . . . . . . 35
Stationary Analysis . . . . . . . . . . . . . . . . . . . . . . 35
Eigenfrequency Analysis . . . . . . . . . . . . . . . . . . . . 36
Mode Analysis . . . . . . . . . . . . . . . . . . . . . . . . 39
Time Domain Analysis . . . . . . . . . . . . . . . . . . . . . 40
Frequency Domain Analysis . . . . . . . . . . . . . . . . . . . 41
Modal Superposition . . . . . . . . . . . . . . . . . . . . . 44
Harmonic Perturbation . . . . . . . . . . . . . . . . . . . . 46
Modal Reduced Order . . . . . . . . . . . . . . . . . . . . . 48
Linearized Buckling Analysis . . . . . . . . . . . . . . . . . . . 53
Bolt Pre-Tension Study. . . . . . . . . . . . . . . . . . . . . 55

Selecting the Physics Interface 56

Solid Mechanics . . . . . . . . . . . . . . . . . . . . . . . 56
Shell and Plate . . . . . . . . . . . . . . . . . . . . . . . . 59
Membrane . . . . . . . . . . . . . . . . . . . . . . . . . 59
Beam . . . . . . . . . . . . . . . . . . . . . . . . . . . 60

CONTENTS |3
 Truss . . . . . . . . . . . . . . . . . . . . . . . . . . . 60

Selecting Discretization 61
Shape Function Order . . . . . . . . . . . . . . . . . . . . . 61
Lagrange and Serendipity Shape Functions . . . . . . . . . . . . . 62
Choosing Shape Functions in Multiphysics Models . . . . . . . . . . 62
Implicit Shape Function Orders . . . . . . . . . . . . . . . . . 63

Coupling Different Element Types 65

Introduction to the Element Types . . . . . . . . . . . . . . . . 65
Coupling Techniques . . . . . . . . . . . . . . . . . . . . . 65

Applying Loads 71
Units, Orientation, and Visualization . . . . . . . . . . . . . . . 71
Load Cases . . . . . . . . . . . . . . . . . . . . . . . . . 73
Singular Loads . . . . . . . . . . . . . . . . . . . . . . . . 74
Moments in the Solid Mechanics Interface . . . . . . . . . . . . . 75
Pressure . . . . . . . . . . . . . . . . . . . . . . . . . . 76
Acceleration Loads . . . . . . . . . . . . . . . . . . . . . . 76
Temperature Loads—Thermal Expansion. . . . . . . . . . . . . . 76
Hygroscopic Swelling . . . . . . . . . . . . . . . . . . . . . 77
Total Loads . . . . . . . . . . . . . . . . . . . . . . . . . 77

Defining Constraints 78
Rigid body motion . . . . . . . . . . . . . . . . . . . . . . 78
Orientation . . . . . . . . . . . . . . . . . . . . . . . . . 79
Prescribed Displacements, Velocities, and Accelerations . . . . . . . . 80
Symmetry Constraints . . . . . . . . . . . . . . . . . . . . . 81
Elemental and Nodal Constraints. . . . . . . . . . . . . . . . . 83
Kinematic Constraints . . . . . . . . . . . . . . . . . . . . . 84
Rotational Joints . . . . . . . . . . . . . . . . . . . . . . . 85

Calculating Reaction Forces 86

Using Predefined Variables to Evaluate Reaction Forces . . . . . . . . 86
Using Weak Constraints to Evaluate Reaction Forces . . . . . . . . . 87
Using Surface Traction to Evaluate Reaction Forces . . . . . . . . . . 88
Evaluating Surface Traction Forces on Internal Boundaries . . . . . . . 89

4 | CONTENTS
Introduction to Material Models 90
Material Models for Structural Mechanics . . . . . . . . . . . . . 90
Combination of Material Models . . . . . . . . . . . . . . . . . 91
Entering material data . . . . . . . . . . . . . . . . . . . . . 92
Introduction to Linear Elastic Materials . . . . . . . . . . . . . . 92
Introduction to Viscoelastic Materials . . . . . . . . . . . . . . . 93
Mixed Formulation . . . . . . . . . . . . . . . . . . . . . . 93
About the Material Databases for the Structural Mechanics Module . . . 95

Modeling Piezoelectric Problems 96

About Piezoelectric Materials . . . . . . . . . . . . . . . . . . 96
Piezoelectric Material Orientation . . . . . . . . . . . . . . . . 97
Piezoelectric Losses. . . . . . . . . . . . . . . . . . . . . 105
References for Piezoelectric Damping . . . . . . . . . . . . . . 109
Piezoelectric Coupling . . . . . . . . . . . . . . . . . . . . 110
Create the Piezoelectric Effect Interface and Define Domains . . . . . 110
Complete Settings of Piezoelectric Materials . . . . . . . . . . . 112
Add Damping and Loss . . . . . . . . . . . . . . . . . . . 113
Define Material Properties . . . . . . . . . . . . . . . . . . 113
Coupling Piezoelectric Devices with Acoustics . . . . . . . . . . 114

Modeling Magnetostrictive Materials 116

Magnetostriction Coupling . . . . . . . . . . . . . . . . . . 116
Create the Magnetostriction Interface and Define Domains. . . . . . 116
Linear vs Nonlinear Magnetostriction . . . . . . . . . . . . . . 118

Mechanical Damping and Losses 119

Damping Sources . . . . . . . . . . . . . . . . . . . . . . 119
About Damping . . . . . . . . . . . . . . . . . . . . . . 120
Rayleigh Damping. . . . . . . . . . . . . . . . . . . . . . 122
Loss Factor Damping . . . . . . . . . . . . . . . . . . . . 124
Viscoelastic Damping . . . . . . . . . . . . . . . . . . . . 125
Explicit Damping . . . . . . . . . . . . . . . . . . . . . . 125
Equivalent Viscous Damping. . . . . . . . . . . . . . . . . . 125
Piezoelectric Damping . . . . . . . . . . . . . . . . . . . . 126
Adding Damping in the Modal Solver . . . . . . . . . . . . . . 126

CONTENTS |5
 Modeling Geometric Nonlinearity 127
Geometric Nonlinearity, Frames, and the ALE Method . . . . . . . 127
Geometric Nonlinearity for the Solid Mechanics Interface . . . . . . 133
Geometric Nonlinearity for the Shell, Plate, Membrane, Beam and
Truss Interfaces . . . . . . . . . . . . . . . . . . . . . 134
Solving Geometrically Nonlinear Problems . . . . . . . . . . . . 135
Prestressed Structures. . . . . . . . . . . . . . . . . . . . 135
Geometric Nonlinearity for the Piezoelectric Material. . . . . . . . 137
References . . . . . . . . . . . . . . . . . . . . . . . . 140

Contact Modeling 141

General . . . . . . . . . . . . . . . . . . . . . . . . . 141
Setting up a contact problem . . . . . . . . . . . . . . . . . 141
Contact Pairs . . . . . . . . . . . . . . . . . . . . . . . 144
Meshing for Contact Analysis . . . . . . . . . . . . . . . . . 145
Constraints . . . . . . . . . . . . . . . . . . . . . . . . 146
Settings for Contact Nodes . . . . . . . . . . . . . . . . . . 147
Time-Dependent Contact Analysis . . . . . . . . . . . . . . . 152
Multiphysics Contact . . . . . . . . . . . . . . . . . . . . 152
Fallback Nodes for Contact . . . . . . . . . . . . . . . . . . 153
Solver Settings for Contact Analysis. . . . . . . . . . . . . . . 153
Monitoring the Solution . . . . . . . . . . . . . . . . . . . 155
Dependent Variables in Contact Analysis . . . . . . . . . . . . . 155
Important Contact Variables . . . . . . . . . . . . . . . . . 158

Springs and Dampers 160

Defining Multiphysics Models 162

Thermal-Structural Interaction. . . . . . . . . . . . . . . . . 162
Acoustic-Structure Interaction. . . . . . . . . . . . . . . . . 163
Thermal-Electric-Structural Interaction . . . . . . . . . . . . . 163

Thermally Coupled Problems 165

Temperature Dependent Material Data . . . . . . . . . . . . . 165
Thermal Expansion . . . . . . . . . . . . . . . . . . . . . 166
Constraints and Thermal Expansion. . . . . . . . . . . . . . . 172
Thermoelastic Damping . . . . . . . . . . . . . . . . . . . 173

6 | CONTENTS
 Computing Mass Properties 175

Pre-tensioned Bolts 178

Stress Linearization 184

Solver Settings 190

Symmetric Matrices . . . . . . . . . . . . . . . . . . . . . 190
Selecting Iterative Solvers . . . . . . . . . . . . . . . . . . 191
Solver Settings for Viscoelasticity, Creep, and Viscoplasticity . . . . . 192
Specifying Tolerances and Scaling for the Solution Components . . . . 193
Solver Settings . . . . . . . . . . . . . . . . . . . . . . . 193

Part Libraries 195

Chapter 3: Structural Mechanics Theory

Solid Mechanics Theory 198

Introduction . . . . . . . . . . . . . . . . . . . . . . . 198

Frames and Coordinate Systems 200

Material and Spatial Coordinates . . . . . . . . . . . . . . . . 200
Coordinate Systems. . . . . . . . . . . . . . . . . . . . . 201

Analysis of Deformation 203

Lagrangian Formulation . . . . . . . . . . . . . . . . . . . 203
Deformation Measures . . . . . . . . . . . . . . . . . . . 206
Invariants of Strain . . . . . . . . . . . . . . . . . . . . . 209
Inelastic Strain Contributions . . . . . . . . . . . . . . . . . 212
Axial Symmetry and Deformation . . . . . . . . . . . . . . . 216
Out-of-plane and Circumferential Modes . . . . . . . . . . . . . 219

Stresses 222
Defining Stress. . . . . . . . . . . . . . . . . . . . . . . 222
Invariants of the Stress Tensor . . . . . . . . . . . . . . . . . 223
Plane Strain and Plane Stress Cases . . . . . . . . . . . . . . . 227

CONTENTS |7
 Initial Stresses and Strains . . . . . . . . . . . . . . . . . . 227
External Stress. . . . . . . . . . . . . . . . . . . . . . . 228
Axial Symmetry and Stresses . . . . . . . . . . . . . . . . . 229

Equation of Motion 230

Material Models 233

Linear Elastic Material . . . . . . . . . . . . . . . . . . . . 236
Nonlinear Elastic Material Models . . . . . . . . . . . . . . . 244
Linear Viscoelastic Materials . . . . . . . . . . . . . . . . . 253
Large Strain Viscoelasticity . . . . . . . . . . . . . . . . . . 262
Hyperelastic Material Models . . . . . . . . . . . . . . . . . 263
Elastoplastic Material Models . . . . . . . . . . . . . . . . . 281
Failure Criteria for Concrete, Rocks, and Other Brittle Material . . . . 312
Cam-Clay Material Model . . . . . . . . . . . . . . . . . . 318
Creep and Viscoplasticity. . . . . . . . . . . . . . . . . . . 323
Piezoelectric Material . . . . . . . . . . . . . . . . . . . . 337
Magnetostrictive Material. . . . . . . . . . . . . . . . . . . 341
Rigid Domain Material Model . . . . . . . . . . . . . . . . . 349
Safety Factor Evaluation . . . . . . . . . . . . . . . . . . . 358

Formulation of the Equilibrium Equations 367

Equation Implementation . . . . . . . . . . . . . . . . . . . 367

Damping 372
Rayleigh Damping. . . . . . . . . . . . . . . . . . . . . . 372
Loss Factor Damping . . . . . . . . . . . . . . . . . . . . 373
Viscous Damping . . . . . . . . . . . . . . . . . . . . . . 374

Loads and Boundary Conditions 376

Distributed Loads . . . . . . . . . . . . . . . . . . . . . 376
Rotating Frame and Gravity . . . . . . . . . . . . . . . . . . 377
Spring Foundation and Thin Elastic Layer . . . . . . . . . . . . . 379
Added Mass . . . . . . . . . . . . . . . . . . . . . . . . 383
Rigid Motion Suppression Boundary Condition . . . . . . . . . . 384
Low-Reflecting Boundary Condition . . . . . . . . . . . . . . 386
Cyclic Symmetry and Floquet Periodic Conditions . . . . . . . . . 387
Thermal Expansion of Constraints . . . . . . . . . . . . . . . 389

8 | CONTENTS
 Rigid Connector Theory . . . . . . . . . . . . . . . . . . . 391
Contact Analysis Theory . . . . . . . . . . . . . . . . . . . 392

Stress Linearization 404

Energy Quantities 407

Elastic Energy . . . . . . . . . . . . . . . . . . . . . . . 407
Kinetic Energy . . . . . . . . . . . . . . . . . . . . . . . 409
Dissipated Energy . . . . . . . . . . . . . . . . . . . . . 409
Mechanical Energy Flux . . . . . . . . . . . . . . . . . . . 410
Energy Variables . . . . . . . . . . . . . . . . . . . . . . 410

References 414

Chapter 4: Solid Mechanics

The Solid Mechanics Interface 422

Domain, Boundary, Edge, Point, and Pair Nodes for Solid Mechanics . . 426
Initial Values. . . . . . . . . . . . . . . . . . . . . . . . 430
Change Thickness . . . . . . . . . . . . . . . . . . . . . 430
Linear Elastic Material . . . . . . . . . . . . . . . . . . . . 431
Nonlinear Elastic Material . . . . . . . . . . . . . . . . . . 437
Hyperelastic Material . . . . . . . . . . . . . . . . . . . . 442
Piezoelectric Material . . . . . . . . . . . . . . . . . . . . 445
Magnetostrictive Material. . . . . . . . . . . . . . . . . . . 447
Cam-Clay Material . . . . . . . . . . . . . . . . . . . . . 450
Viscoelasticity . . . . . . . . . . . . . . . . . . . . . . . 451
Plasticity . . . . . . . . . . . . . . . . . . . . . . . . . 454
Creep . . . . . . . . . . . . . . . . . . . . . . . . . . 459
Viscoplasticity . . . . . . . . . . . . . . . . . . . . . . . 462
Porous Plasticity . . . . . . . . . . . . . . . . . . . . . . 467
Soil Plasticity . . . . . . . . . . . . . . . . . . . . . . . 469
Concrete . . . . . . . . . . . . . . . . . . . . . . . . . 471
Rocks . . . . . . . . . . . . . . . . . . . . . . . . . . 472
Thermal Expansion (for materials) . . . . . . . . . . . . . . . 473
Hygroscopic Swelling . . . . . . . . . . . . . . . . . . . . 475

CONTENTS |9
 Initial Stress and Strain. . . . . . . . . . . . . . . . . . . . 477
External Stress. . . . . . . . . . . . . . . . . . . . . . . 478
External Strain . . . . . . . . . . . . . . . . . . . . . . . 480
Safety . . . . . . . . . . . . . . . . . . . . . . . . . . 483
Damping . . . . . . . . . . . . . . . . . . . . . . . . . 487
Mechanical Damping . . . . . . . . . . . . . . . . . . . . 491
Coupling Loss . . . . . . . . . . . . . . . . . . . . . . . 492
Dielectric Loss. . . . . . . . . . . . . . . . . . . . . . . 493
Conduction Loss (Time-Harmonic) . . . . . . . . . . . . . . . 493
External Stress-Strain Relation . . . . . . . . . . . . . . . . . 494
Rigid Domain . . . . . . . . . . . . . . . . . . . . . . . 495
Initial Values (Rigid Domain) . . . . . . . . . . . . . . . . . 498
Fixed Constraint (Rigid Domain) . . . . . . . . . . . . . . . . 499
Prescribed Displacement/Rotation . . . . . . . . . . . . . . . 500
Applied Force (Rigid Domain) . . . . . . . . . . . . . . . . . 502
Location Nodes . . . . . . . . . . . . . . . . . . . . . . 504
Applied Moment (Rigid Domain) . . . . . . . . . . . . . . . . 504
Mass and Moment of Inertia (Rigid Domain) . . . . . . . . . . . 505
Center of Mass Nodes. . . . . . . . . . . . . . . . . . . . 506
Spring Foundation (Rigid Domain) . . . . . . . . . . . . . . . 507
Free. . . . . . . . . . . . . . . . . . . . . . . . . . . 509
Prescribed Displacement . . . . . . . . . . . . . . . . . . . 510
Prescribed Velocity . . . . . . . . . . . . . . . . . . . . . 513
Prescribed Acceleration . . . . . . . . . . . . . . . . . . . 515
Fixed Constraint . . . . . . . . . . . . . . . . . . . . . . 516
Thermal Expansion (for constraints) . . . . . . . . . . . . . . 518
Roller . . . . . . . . . . . . . . . . . . . . . . . . . . 520
Symmetry . . . . . . . . . . . . . . . . . . . . . . . . 521
Antisymmetry . . . . . . . . . . . . . . . . . . . . . . . 522
Rigid Motion Suppression . . . . . . . . . . . . . . . . . . 523
Body Load . . . . . . . . . . . . . . . . . . . . . . . . 525
Gravity . . . . . . . . . . . . . . . . . . . . . . . . . 526
Rotating Frame . . . . . . . . . . . . . . . . . . . . . . 528
Boundary Load . . . . . . . . . . . . . . . . . . . . . . 529
Edge Load . . . . . . . . . . . . . . . . . . . . . . . . 531
Point Load . . . . . . . . . . . . . . . . . . . . . . . . 532
Ring Load . . . . . . . . . . . . . . . . . . . . . . . . 533
Point Load (on Axis) . . . . . . . . . . . . . . . . . . . . 534

10 | C O N T E N T S
 Spring Foundation . . . . . . . . . . . . . . . . . . . . . 535
Thin Elastic Layer. . . . . . . . . . . . . . . . . . . . . . 540
Predeformation . . . . . . . . . . . . . . . . . . . . . . 542
Added Mass . . . . . . . . . . . . . . . . . . . . . . . . 543
Periodic Condition . . . . . . . . . . . . . . . . . . . . . 545
Low-Reflecting Boundary . . . . . . . . . . . . . . . . . . . 548
Thin-Film Damping . . . . . . . . . . . . . . . . . . . . . 548
Rigid Connector . . . . . . . . . . . . . . . . . . . . . . 549
Center of Rotation Nodes . . . . . . . . . . . . . . . . . . 553
Thermal Expansion (Rigid Connector) . . . . . . . . . . . . . . 554
Applied Force (Rigid Connector) . . . . . . . . . . . . . . . . 555
Applied Moment (Rigid Connector) . . . . . . . . . . . . . . . 556
Mass and Moment of Inertia (Rigid Connector) . . . . . . . . . . 557
Spring Foundation (Rigid Connector) . . . . . . . . . . . . . . 558
Attachment . . . . . . . . . . . . . . . . . . . . . . . . 561
Thermal Expansion (Attachment). . . . . . . . . . . . . . . . 562
Beam Connection . . . . . . . . . . . . . . . . . . . . . 563
Shell Connection . . . . . . . . . . . . . . . . . . . . . . 565
Contact . . . . . . . . . . . . . . . . . . . . . . . . . 566
Friction . . . . . . . . . . . . . . . . . . . . . . . . . 572
Adhesion . . . . . . . . . . . . . . . . . . . . . . . . . 574
Bolt Pre-Tension . . . . . . . . . . . . . . . . . . . . . . 576
Bolt Selection . . . . . . . . . . . . . . . . . . . . . . . 578
Phase . . . . . . . . . . . . . . . . . . . . . . . . . . 579
Harmonic Perturbation . . . . . . . . . . . . . . . . . . . 581
Stress Linearization . . . . . . . . . . . . . . . . . . . . . 582
Important Variables in the Solid Mechanics Interface . . . . . . . . 583

Chapter 5: Shell and Plate

Theory for Shell and Plate Interfaces 588

About Shells and Plates . . . . . . . . . . . . . . . . . . . 588
Theory Background for the Shell and Plate Interfaces . . . . . . . . 590
Reference for the Shell Interface . . . . . . . . . . . . . . . . 612

CONTENTS | 11
 The Shell and Plate Interfaces 613
Domain, Boundary, Edge, Point, and Pair Nodes for the Shell and
Plate Interfaces . . . . . . . . . . . . . . . . . . . . . 618
Initial Values. . . . . . . . . . . . . . . . . . . . . . . . 620
Change Thickness . . . . . . . . . . . . . . . . . . . . . 621
Linear Elastic Material . . . . . . . . . . . . . . . . . . . . 623
Shell Local System . . . . . . . . . . . . . . . . . . . . . 624
Thermal Expansion (for materials) . . . . . . . . . . . . . . . 625
Hygroscopic Swelling . . . . . . . . . . . . . . . . . . . . 626
Initial Stress and Strain. . . . . . . . . . . . . . . . . . . . 628
External Stress. . . . . . . . . . . . . . . . . . . . . . . 629
Prescribed Displacement/Rotation . . . . . . . . . . . . . . . 631
Thermal Expansion (for constraints) . . . . . . . . . . . . . . 634
Prescribed Velocity . . . . . . . . . . . . . . . . . . . . . 636
Prescribed Acceleration . . . . . . . . . . . . . . . . . . . 638
Pinned . . . . . . . . . . . . . . . . . . . . . . . . . . 640
No Rotation . . . . . . . . . . . . . . . . . . . . . . . 641
Simply Supported. . . . . . . . . . . . . . . . . . . . . . 642
Symmetry . . . . . . . . . . . . . . . . . . . . . . . . 643
Antisymmetry . . . . . . . . . . . . . . . . . . . . . . . 645
Body Load . . . . . . . . . . . . . . . . . . . . . . . . 647
Face Load . . . . . . . . . . . . . . . . . . . . . . . . 648
Edge Load . . . . . . . . . . . . . . . . . . . . . . . . 649
Point Load . . . . . . . . . . . . . . . . . . . . . . . . 651
Point Mass . . . . . . . . . . . . . . . . . . . . . . . . 652
Point Mass Damping . . . . . . . . . . . . . . . . . . . . 653
Periodic Condition . . . . . . . . . . . . . . . . . . . . . 654
Solid Connection . . . . . . . . . . . . . . . . . . . . . . 658
Beam Connection . . . . . . . . . . . . . . . . . . . . . 659
Rigid Connector . . . . . . . . . . . . . . . . . . . . . . 661
Thermal Expansion (Rigid Connector) . . . . . . . . . . . . . . 664
Attachment . . . . . . . . . . . . . . . . . . . . . . . . 666
Thermal Expansion (Attachment). . . . . . . . . . . . . . . . 666
Phase . . . . . . . . . . . . . . . . . . . . . . . . . . 668
Harmonic Perturbation . . . . . . . . . . . . . . . . . . . 669

12 | C O N T E N T S
Chapter 6: Membrane

Theory for the Membrane Interface 672

About Membranes . . . . . . . . . . . . . . . . . . . . . 672
Theory Background for the Membrane Interface . . . . . . . . . . 673

The Membrane Interface 677

Boundary, Edge, Point, and Pair Nodes for the Membrane Interface. . . 679
Linear Elastic Material . . . . . . . . . . . . . . . . . . . . 681
Initial Stress and Strain. . . . . . . . . . . . . . . . . . . . 684
External Stress. . . . . . . . . . . . . . . . . . . . . . . 686

Chapter 7: Beam

Theory for the Beam Interface 690

About Beams . . . . . . . . . . . . . . . . . . . . . . . 690
In-Plane Beams . . . . . . . . . . . . . . . . . . . . . . 691
3D Beam . . . . . . . . . . . . . . . . . . . . . . . . . 692
Geometric Variables . . . . . . . . . . . . . . . . . . . . 693
Shape Functions . . . . . . . . . . . . . . . . . . . . . . 694
Geometric Nonlinearity . . . . . . . . . . . . . . . . . . . 696
Strain-Displacement/Rotation Relation. . . . . . . . . . . . . . 699
Stress-Strain Relation . . . . . . . . . . . . . . . . . . . . 699
Thermal Strain . . . . . . . . . . . . . . . . . . . . . . . 700
Hygroscopic Swelling . . . . . . . . . . . . . . . . . . . . 700
Initial Load and Strain . . . . . . . . . . . . . . . . . . . . 701
Implementation . . . . . . . . . . . . . . . . . . . . . . 702
Stress Evaluation . . . . . . . . . . . . . . . . . . . . . . 704
Common Cross Sections . . . . . . . . . . . . . . . . . . . 706

The Beam Interface 719

Boundary, Edge, Point, and Pair Nodes for the Beam Interface. . . . . 722
Initial Values. . . . . . . . . . . . . . . . . . . . . . . . 724
Cross Section Data . . . . . . . . . . . . . . . . . . . . . 724
Section Orientation . . . . . . . . . . . . . . . . . . . . . 728

CONTENTS | 13
 Linear Elastic Material . . . . . . . . . . . . . . . . . . . . 731
Thermal Expansion (for materials) . . . . . . . . . . . . . . . 733
Hygroscopic Swelling . . . . . . . . . . . . . . . . . . . . 734
Initial Stress and Strain. . . . . . . . . . . . . . . . . . . . 735
External Stress. . . . . . . . . . . . . . . . . . . . . . . 736
Prescribed Displacement/Rotation . . . . . . . . . . . . . . . 737
Prescribed Velocity . . . . . . . . . . . . . . . . . . . . . 739
Prescribed Acceleration . . . . . . . . . . . . . . . . . . . 740
Pinned . . . . . . . . . . . . . . . . . . . . . . . . . . 742
Thermal Expansion (for constraints) . . . . . . . . . . . . . . 743
No Rotation . . . . . . . . . . . . . . . . . . . . . . . 744
Symmetry . . . . . . . . . . . . . . . . . . . . . . . . 745
Antisymmetry . . . . . . . . . . . . . . . . . . . . . . . 746
Edge Load . . . . . . . . . . . . . . . . . . . . . . . . 747
Point Load . . . . . . . . . . . . . . . . . . . . . . . . 750
Point Mass . . . . . . . . . . . . . . . . . . . . . . . . 751
Point Mass Damping . . . . . . . . . . . . . . . . . . . . 751
Solid Connection . . . . . . . . . . . . . . . . . . . . . . 752
Shell Connection . . . . . . . . . . . . . . . . . . . . . . 753
Rigid Connector . . . . . . . . . . . . . . . . . . . . . . 754
Attachment . . . . . . . . . . . . . . . . . . . . . . . . 757
Phase . . . . . . . . . . . . . . . . . . . . . . . . . . 758
Harmonic Perturbation . . . . . . . . . . . . . . . . . . . 758

Chapter 8: Beam Cross Section

Using the Beam Cross Section Interface 762

About Beams and Cross Section Data . . . . . . . . . . . . . . 762
Using the Beam Cross Section Interface . . . . . . . . . . . . . 763

Theory for the Beam Cross Section Interface 768

Cross Section Properties . . . . . . . . . . . . . . . . . . . 768
Computation of Stresses . . . . . . . . . . . . . . . . . . . 782

The Beam Cross Section Interface 785

Hole . . . . . . . . . . . . . . . . . . . . . . . . . . 786

14 | C O N T E N T S
 Safety . . . . . . . . . . . . . . . . . . . . . . . . . . 787

C h a p t e r 9 : Tr u s s

Modeling with Truss Elements 790

Theory for the Truss Interface 792

Theory Background for the Truss Interface . . . . . . . . . . . . 792

The Truss Interface 800

Boundary, Edge, Point, and Pair Nodes for the Truss Interface . . . . . 801
Cross Section Data . . . . . . . . . . . . . . . . . . . . . 803
Straight Edge Constraint . . . . . . . . . . . . . . . . . . . 804
Linear Elastic Material . . . . . . . . . . . . . . . . . . . . 805
Thermal Expansion (for materials) . . . . . . . . . . . . . . . 806
Hygroscopic Swelling . . . . . . . . . . . . . . . . . . . . 807
Initial Stress and Strain. . . . . . . . . . . . . . . . . . . . 808
External Stress. . . . . . . . . . . . . . . . . . . . . . . 808
Spring-Damper . . . . . . . . . . . . . . . . . . . . . . 810
Pinned . . . . . . . . . . . . . . . . . . . . . . . . . . 811
Thermal Expansion (for constraints) . . . . . . . . . . . . . . 812
Symmetry . . . . . . . . . . . . . . . . . . . . . . . . 813
Antisymmetry . . . . . . . . . . . . . . . . . . . . . . . 814
Edge Load . . . . . . . . . . . . . . . . . . . . . . . . 815
Point Mass . . . . . . . . . . . . . . . . . . . . . . . . 816
Point Mass Damping . . . . . . . . . . . . . . . . . . . . 817
Phase . . . . . . . . . . . . . . . . . . . . . . . . . . 817

Chapter 10: Multiphysics Interfaces and Couplings

The Thermal Stress Interface 820

Thermal Expansion (Multiphysics Coupling) . . . . . . . . . . . . 822

The Fluid-Structure Interaction Interface 825

Domain, Boundary, Edge, Point, and Pair Nodes for the

CONTENTS | 15
 Fluid-Structure Interaction Interface . . . . . . . . . . . . . 828
Initial Values. . . . . . . . . . . . . . . . . . . . . . . . 831
Wall. . . . . . . . . . . . . . . . . . . . . . . . . . . 831
Fluid-Solid Interface Boundary . . . . . . . . . . . . . . . . . 833
Basic Modeling Steps for Fluid-Structure Interaction . . . . . . . . 833

The Fluid-Structure Interaction, Fixed Geometry Interface 836

Fluid-Structure Interaction, Fixed Geometry . . . . . . . . . . . 838

Theory for the Fluid-Structure Interaction Interface 840

The Joule Heating and Thermal Expansion Interface 842

The Piezoelectric Devices Interface 846

Piezoelectric Effect . . . . . . . . . . . . . . . . . . . . . 849
Coupling Piezoelectric Devices with Acoustics . . . . . . . . . . 850

The Magnetostriction Interface 852

Magnetostriction . . . . . . . . . . . . . . . . . . . . . . 854

Hygroscopic Swelling Coupling 856

Hygroscopic Swelling . . . . . . . . . . . . . . . . . . . . 856

Couplings Between Structural Mechanics Interfaces 858

Solid-Shell Connection. . . . . . . . . . . . . . . . . . . . 858
Solid-Beam Connection . . . . . . . . . . . . . . . . . . . 860
Shell-Beam Connection . . . . . . . . . . . . . . . . . . . 861

Chapter 11: Glossary

Glossary of Terms 866

16 | C O N T E N T S
 1

Introduction

This guide describes the Structural Mechanics Module, an optional add-on

package that extends the COMSOL Multiphysics® modeling environment with
customized physics interfaces that solve problems in the fields of structural and
solid mechanics, including special physics interface for modeling of shells,
membranes, beams, plates, and trusses.

This chapter introduces you to the capabilities of this module and includes a
summary of the physics interfaces as well as information about where you can find
additional documentation and model examples. The last section is a brief overview
with links to each chapter in this guide.

• About the Structural Mechanics Module

• Overview of the User’s Guide

17
 About the Structural Mechanics
Module
In this section:

• Why Structural Mechanics is Important for Modeling

• What Problems Can It Solve?
• The Structural Mechanics Module Physics Interface Guide
• Common Physics Interface and Feature Settings and Nodes
• Geometry Levels for Study Capabilities
• Where Do I Access the Documentation and Application Libraries?

The Physics Interfaces and Building a COMSOL Multiphysics Model in

the COMSOL Multiphysics Reference Manual

Why Structural Mechanics is Important for Modeling

The Structural Mechanics Module solves problems in the fields of structural and solid
mechanics, adding special physics interfaces for modeling shells and beams, for
example.

The physics interfaces in this module are fully multiphysics enabled, making it possible
to couple them to any other physics interfaces in COMSOL Multiphysics or the other
modules. Available physics interfaces include:

• Solid mechanics for 2D plane stress and plane strain, axial symmetry, and 3D solids
• Beams in 2D and 3D, Euler and Timoshenko theory
• Truss and cable elements
• Shells and plates, Mindlin theory
• Membranes

The module’s study capabilities include static, eigenfrequency, time dependent

(transient), frequency response, buckling, and parametric studies.

18 | CHAPTER 1: INTRODUCTION
There are several material models:

• Linear Elastic Materials can be isotropic, orthotropic, or fully anisotropic, and you
can use local coordinate systems to specify material properties.
• Linear Viscoelastic Materials
• Piezoelectric Material
• Magnetostrictive Materials are available when used together with the AC/DC
module.
• Material models for hyperelasticity, metal plasticity, creep, viscoplasticity, nonlinear
elasticity, soil plasticity, concrete, rocks, and clay are available with the optional
Nonlinear Structural Materials Module and Geomechanics Module.

Large deformations as well as contact and friction, can also be modeled.

Coupling structural analysis with thermal analysis is one example of multiphysics easily
implemented with the module, which provides predefined multiphysics couplings for
thermal stress and other types of multiphysics. Piezoelectric materials, coupling the
electric field and strain in both directions are fully supported inside the module
through special multiphysics couplings solving for both the electric potential and
displacements. Structural mechanics couplings are common in simulations done with
COMSOL Multiphysics and occur in interaction with, for example, fluid flow
(fluid-structure interaction, FSI), chemical reactions, acoustics, electric fields,
magnetic fields, and optical wave propagation.

What Problems Can It Solve?

The Structural Mechanics Module contains a set of physics interfaces adapted to a
broad category of structural-mechanics analysis. The module serves as an excellent tool
for the professional engineer, researcher, and teacher. In education, the benefit of the
short learning curve is especially useful because educators need not spend excessive
time learning the software and can instead focus on the physics and the modeling
process.

The module is a collection of physics interfaces for COMSOL Multiphysics that

handles static, eigenfrequency, transient, frequency response, parametric, transient
thermal stress, and other analyses for applications in structural mechanics, solid
mechanics, and piezoelectricity.

STATIC ANALYSIS
In a static analysis the load and constraints are fixed in time.

ABOUT THE STRUCTURAL MECHANICS MODULE | 19

 EIGENFREQUENCY ANALYSIS
An eigenfrequency analysis finds the damped or undamped eigenfrequencies and mode
shapes of a structure. Sometimes referred to as the free vibration of a structure.
Pre-stress effects and damping can be taken into account.

TR A N S I E N T A N A L Y S I S
A transient analysis finds the transient response for a time-dependent model, taking
into account mass, mass moment of inertia. The transient analysis can be either direct,
or using a modal solution.

FREQUENCY RESPONSE ANALYSIS

A frequency response analysis finds the steady-state response to harmonic loads. The
frequency-response analysis can be either direct, or using a modal solution. Effects of
pre-stress can be included.

LINEAR BUCKLING STUDY

A linear buckling analysis uses the stiffness coming from stresses and material to
define an eigenvalue problem where the eigenvalue is a load factor that, when
multiplied with the actual load, gives the critical load in a linear context.

PARAMETRIC ANALYSIS
A parametric analysis finds the solution dependence due to the variation of a specific
parameter, which could be, for instance, a material property or the position of a load.

THERMAL STRESS
In a transient thermal stress study, the program neglects mass effects, assuming that the
time scale in the structural mechanics problem is much smaller than the time scale in
the thermal problem.

LARGE DEFORMATIONS
You can also enable geometric nonlinearity for all structural mechanics interfaces. The
engineering strain is then replaced with the Green-Lagrange strain and the stress with
the second Piola-Kirchhoff stress. To solve the problem, the program uses a total
Lagrangian formulation.

ELASTOPLASTIC MATERIALS
An elastoplastic analysis involves a nonlinear material with or without hardening.
Several isotropic and kinematic hardening models are available.

The material models allow large strains.

20 | CHAPTER 1: INTRODUCTION
The elastoplastic material models are available in the Solid Mechanics, Membrane, and
Truss interfaces.

CREEP AND VISCOPLASTIC MATERIALS

A number of different material models for creep and viscoplasticity are available. In
these materials the rate of elastic strain depends on the stress.

HYPERELASTIC MATERIALS
In hyperelastic materials the stresses are computed from a strain energy density
function. They are often used to model rubber and biological tissue, but are also used
in acoustic elasticity. Many different models are available.

The hyperelastic materials are available in the Solid Mechanics and Membrane
interfaces.

NONLINEAR ELASTIC MATERIALS

The nonlinear elastic materials are intended for materials that exhibit a nonlinear
behavior already at small strains. Some brittle material as well as soils show this
behavior.

VISCOELASTIC MATERIALS
Viscoelastic materials have a time-dependent response, even if the loading is constant.
The Viscoelasticity materials are available in the Solid Mechanics and Membrane
interfaces.

• Hyperelastic, elastoplastic, creep, viscoplastic and nonlinear elastic

material models are available with the Nonlinear Structural Materials
Module.
• Additional functionality and material models for geomechanics and soil
mechanics—nonlinear elasticity, soil plasticity, concrete, rock, and clay
material models—is available with the Geomechanics Module.

CONTACT MODELING
You can model contact between parts of a structure. The Solid Mechanics interface
supports contact with or without friction. The two contact algorithms available are
based on the augmented Lagrangian and penalty methods. The contact models can
be augmented with adhesion and decohesion.

ABOUT THE STRUCTURAL MECHANICS MODULE | 21

 The Structural Mechanics Module Physics Interface Guide

The Poroelasticity interface requires, and couple with, the Structural

Mechanics Module and is discussed in the Subsurface Flow Module
User’s Guide.

At any time, a new model can be created or physics interfaces added. Right-click the
Root (top) node and select Add Component or right-click a Component node and select
Add Physics.

Depending on the physics interface, specify parameters defining a problem on points,

edges (3D), boundaries, and domains. It is possible to specify loads and constraints on
all available geometry levels, but material properties can only be specified for the
domains, except for shells, membranes, beams, and trusses, where they are defined on
the boundary or edge level.

In the COMSOL Multiphysics Reference Manual:

• Studies and Solvers

• The Physics Interfaces
• For a list of all the core physics interfaces included with a COMSOL
Multiphysics license, see Physics Interface Guide.

PHYSICS INTERFACE ICON TAG SPACE AVAILABLE PRESET STUDY

DIMENSION TYPE

Fluid Flow

Fluid-Structure Interaction fsi 3D, 2D, 2D stationary; stationary,

axisymmetric one-way coupled; time
dependent; time
dependent, one-way
coupled
Fluid-Structure — 3D, 2D, 2D stationary; time-dependent
Interaction, Fixed axisymmetric
Geometry2

22 | CHAPTER 1: INTRODUCTION
PHYSICS INTERFACE ICON TAG SPACE AVAILABLE PRESET STUDY
DIMENSION TYPE

Structural Mechanics

Solid Mechanics1 solid 3D, 2D, 2D stationary; eigenfrequency;

axisymmetric prestressed analysis,
eigenfrequency; mode
analysis; time dependent;
time-dependent modal;
frequency domain;
frequency-domain modal;
prestressed analysis,
frequency domain; modal
reduced order model;
linear buckling; bolt
pre-tension
Thermal Stress2 — 3D, 2D, 2D stationary; time dependent
axisymmetric
Joule Heating and Thermal — 3D, 2D, 2D stationary; time dependent
Expansion2 axisymmetric
Shell shell 3D stationary; eigenfrequency;
prestressed analysis,
eigenfrequency; time
dependent; time-dependent
modal; frequency domain;
frequency-domain modal;
prestressed analysis,
frequency domain; modal
reduced order model;
linear buckling
Plate plate 2D stationary; eigenfrequency;
prestressed analysis,
eigenfrequency; time
dependent; time-dependent
modal; frequency domain;
frequency-domain modal;
prestressed analysis,
frequency domain; modal
reduced order model;
linear buckling

ABOUT THE STRUCTURAL MECHANICS MODULE | 23

 PHYSICS INTERFACE ICON TAG SPACE AVAILABLE PRESET STUDY
DIMENSION TYPE

Beam beam 3D, 2D stationary; eigenfrequency;

frequency domain;
frequency-domain modal;
time dependent;
time-dependent modal;
modal reduced order
model; linear buckling
Beam Cross Section bcs 2D stationary

Truss truss 3D, 2D stationary; eigenfrequency;

prestressed analysis,
eigenfrequency; time
dependent; time-dependent
modal; frequency domain;
frequency-domain modal;
prestressed analysis,
frequency domain; modal
reduced order model;
linear buckling
Membrane mbrn 3D, 2D stationary; eigenfrequency;
axisymmetric prestressed analysis,
eigenfrequency; time
dependent; time-dependent
modal; frequency domain;
frequency-domain modal;
prestressed analysis,
frequency domain
Piezoelectric Devices2 — 3D, 2D, 2D stationary; eigenfrequency;
axisymmetric time dependent;
time-dependent modal;
frequency domain;
frequency-domain modal;
small-signal analysis,
frequency domain;
prestressed analysis,
eigenfrequency;
prestressed analysis,
frequency domain; modal
reduced order model

24 | CHAPTER 1: INTRODUCTION
PHYSICS INTERFACE ICON TAG SPACE AVAILABLE PRESET STUDY
DIMENSION TYPE

Magnetostriction2 — 3D, 2D, 2D stationary; eigenfrequency;

axisymmetric time dependent; frequency
domain; small-signal
analysis, frequency domain;
prestressed analysis,
eigenfrequency;
prestressed analysis,
frequency domain
1 This physics interface is included with the core COMSOL package but has added
functionality for this module.
2
This physics interface is a predefined multiphysics coupling that automatically adds all the
physics interfaces and coupling features required.

Common Physics Interface and Feature Settings and Nodes

There are several common settings and sections available for the physics interfaces and
feature nodes. Some of these sections also have similar settings or are implemented in
the same way no matter the physics interface or feature being used. There are also some
physics feature nodes that display in COMSOL Multiphysics.

In each module’s documentation, only unique or extra information is included;

standard information and procedures are centralized in the COMSOL Multiphysics
Reference Manual.

In the COMSOL Multiphysics Reference Manual see Table 2-3 for

links to common sections and Table 2-4 to common feature nodes.
You can also search for information: press F1 to open the Help
window or Ctrl+F1 to open the Documentation window.

Geometry Levels for Study Capabilities

The column for the dependent variables shows the field variables that formulate the
underlying equations. Depending on the engineering assumptions and the geometry
dimension, these variables include a subset of the displacement field u, v, and w in the
global coordinate system, pressure, and temperature. The Shell and Plate interfaces use
as dependent variables the variables ax, ay, and az, which are the displacements of the

ABOUT THE STRUCTURAL MECHANICS MODULE | 25

 shell normals in the global x, y, and z directions, respectively. Such variables can be
expressed in terms of customary rotations φx, φy, and φz about the global axes.

For each physics interface, the table indicates dependent variables and the geometry
levels (where data such as material properties, loads, and constraints are specified).
Edges exist only in 3D geometries. Studies are listed in a separate table in The
Structural Mechanics Module Physics Interface Guide section.

Studies and Solvers in the COMSOL Multiphysics Reference Manual

PHYSICS GEOMETRY LEVEL

INTERFACE

NAME DEPENDENT POINTS EDGES BOUNDARIES DOMAINS

VARIABLES

STRUCTURAL MECHANICS

Solid Mechanics solid u, (p) √ √ √ √

Shell shell u, a √ √ √
Plate (3 DOF) plate w, ax, ay √ √ √
Plate (6 DOF) plate u, a √ √ √
Beam beam u, θ √ √
Truss truss u √ √
Membrane mbrn u √ √ √
Thermal Stress n/a* u, (p), T √ √ √ √
Joule Heating and n/a* u, (p), T, √ √ √ √
Thermal V
Expansion
Piezoelectric n/a* u,V √ √ √ √
Devices
Fluid Flow
Fluid-Structure fsi usolid, √ √ √ √
Interaction ufluid, p
Fluid-Structure n/a* usolid, √ √ √ √
Interaction, ufluid, p
Fixed Geometry
* These are multiphysics couplings that do not have names.

26 | CHAPTER 1: INTRODUCTION
Where Do I Access the Documentation and Application Libraries?
A number of internet resources have more information about COMSOL, including
licensing and technical information. The electronic documentation, topic-based (or
context-based) help, and the application libraries are all accessed through the
COMSOL Desktop.

If you are reading the documentation as a PDF file on your computer,

the blue links do not work to open an application or content
referenced in a different guide. However, if you are using the Help
system in COMSOL Multiphysics, these links work to open other
modules (as long as you have a license), application examples, and
documentation sets.

THE DOCUMENTATION AND ONLINE HELP

The COMSOL Multiphysics Reference Manual describes the core physics interfaces
and functionality included with the COMSOL Multiphysics license. This book also has
instructions about how to use COMSOL Multiphysics and how to access the
electronic Documentation and Help content.

Opening Topic-Based Help

The Help window is useful as it is connected to many of the features on the GUI. To
learn more about a node in the Model Builder, or a window on the Desktop, click to
highlight a node or window, then press F1 to open the Help window, which then
displays information about that feature (or click a node in the Model Builder followed
by the Help button ( ). This is called topic-based (or context) help.

To open the Help window:

• In the Model Builder, Application Builder, or Physics Builder click a node or

window and then press F1.
• On any toolbar (for example, Home, Definitions, or Geometry), hover the
mouse over a button (for example, Add Physics or Build All) and then
press F1.
• From the File menu, click Help ( ).
• In the upper-right corner of the COMSOL Desktop, click the Help ( )
button.

ABOUT THE STRUCTURAL MECHANICS MODULE | 27

 To open the Help window:

• In the Model Builder or Physics Builder click a node or window and then
press F1.
• On the main toolbar, click the Help ( ) button.
• From the main menu, select Help>Help.

Opening the Documentation Window

To open the Documentation window:

• Press Ctrl+F1.
• From the File menu select Help>Documentation ( ).

To open the Documentation window:

• Press Ctrl+F1.
• On the main toolbar, click the Documentation ( ) button.
• From the main menu, select Help>Documentation.

THE APPLICATION LIBRARIES WINDOW

Each application includes documentation with the theoretical background and
step-by-step instructions to create a model application. The applications are available
in COMSOL as MPH-files that you can open for further investigation. You can use the
step-by-step instructions and the actual applications as a template for your own
modeling and applications. In most models, SI units are used to describe the relevant
properties, parameters, and dimensions in most examples, but other unit systems are
available.

Once the Application Libraries window is opened, you can search by name or browse
under a module folder name. Click to view a summary of the application and its
properties, including options to open it or a PDF document.

The Application Libraries Window in the COMSOL Multiphysics

Reference Manual.

28 | CHAPTER 1: INTRODUCTION
Opening the Application Libraries Window
To open the Application Libraries window ( ):

• From the Home toolbar, Windows menu, click ( ) Applications

Libraries.
• From the File menu select Application Libraries.

To include the latest versions of model examples, from the File>Help

menu, select ( ) Update COMSOL Application Library.

Select Application Libraries from the main File> or Windows> menus.

To include the latest versions of model examples, from the Help menu
select ( ) Update COMSOL Application Library.

CONTACTING COMSOL BY EMAIL

For general product information, contact COMSOL at [email protected].

To receive technical support from COMSOL for the COMSOL products, please
contact your local COMSOL representative or send your questions to
[email protected]. An automatic notification and a case number are sent to you by
email.

COMSOL ONLINE RESOURCES

COMSOL website www.comsol.com

Contact COMSOL www.comsol.com/contact
Support Center www.comsol.com/support
Product Download www.comsol.com/product-download
Product Updates www.comsol.com/support/updates
COMSOL Blog www.comsol.com/blogs
Discussion Forum www.comsol.com/community
Events www.comsol.com/events
COMSOL Video Gallery www.comsol.com/video
Support Knowledge Base www.comsol.com/support/knowledgebase

ABOUT THE STRUCTURAL MECHANICS MODULE | 29

 Overview of the User’s Guide
The Structural Mechanics Module User’s Guide gets you started with modeling using
COMSOL Multiphysics. The information in this guide is specific to this module.
Instructions how to use COMSOL in general are included with the COMSOL
Multiphysics Reference Manual.

As detailed in the section Where Do I Access the Documentation and

Application Libraries? this information can also be searched from the Help
menu in COMSOL Multiphysics.

TA B L E O F C O N T E N T S , G L O S S A R Y, A N D I N D E X
To help you navigate through this guide, see the Contents, Glossary, and Index.

MODELING WITH THE STRUCTURAL MECHANICS MODULE

The Structural Mechanics Modeling chapter gives you an insight on how to approach
the modeling of various structural mechanics problems.

STRUCTURAL MECHANICS THEORY

The Structural Mechanics Theory chapter introduces the general theory on which the
physics interfaces in the Structural Mechanics Module are based.

THE SOLID MECHANICS INTERFACE

The Solid Mechanics chapter describes The Solid Mechanics Interface, which is used
to model 3D solids, plane strain and plane stress 2D models, and axisymmetric models.

THE SHELL AND PLATE INTERFACES

The Shell and Plate chapter describes The Shell and Plate Interfaces, which are used to
model thin 3D structures (shell) and out-of-plane loaded plates (plate). The
underlying theory is described in Theory for Shell and Plate Interfaces.

THE BEAM INTERFACE

The Beam chapter describes The Beam Interface, which contains Euler
(Euler-Bernoulli) and Timoshenko beams for modeling slender 3D and 2D structures.
Typical examples are frameworks and latticeworks. The underlying theory for the
physics interface is described in Theory for the Beam Interface.

30 | CHAPTER 1: INTRODUCTION
THE BEAM CROSS SECTION INTERFACE
The Beam Cross Section chapter describes The Beam Cross Section Interface, which
is used for computing cross section properties for beams. It can also be used for a
detailed evaluation of stresses in a beam when the section forces to which it is subjected
are known. The first section discusses Using the Beam Cross Section Interface, and the
underlying theory is described in Theory for the Beam Cross Section Interface.

T H E TR U S S I N T E R F A C E
The Truss chapter describes The Truss Interface, which models slender 3D and 2D
structures with components capable to withstand axial forces only. Typical applications
are latticeworks, but it can also be used for modeling cables. In the section Modeling
with Truss Elements, the you will find a discussion how to set up models using this
interface. The underlying theory for the physics interface is described in Theory for the
Truss Interface.

THE MEMBRANE INTERFACE

The Membrane chapter describes The Membrane Interface, which can be used for
prestressed membranes, cladding on solids, and balloons, for example. The underlying
theory for the physics interface is also included in Theory for the Membrane Interface.

THE MULTIPHYSICS INTERFACES

The Multiphysics Interfaces and Couplings chapter describes these physics interfaces
found under the Structural Mechanics branch when adding a physics interface:

• The Thermal Stress Interface combines a Solid Mechanics interface with a Heat
Transfer interface. The coupling appears on the domain level, where the
temperature from the Heat Transfer interface acts as a thermal load for the Solid
Mechanics interface, causing thermal expansion.
• The Joule Heating and Thermal Expansion Interface combines solid mechanics
using a thermal linear elastic material with an electromagnetic Joule heating model.
This is a multiphysics combination of solid mechanics, electric currents, and heat
transfer for modeling of, for example, thermoelectromechanical (TEM)
applications.
• The Piezoelectric Devices Interface combines a Solid Mechanics with an
Electrostatics interface. Piezoelectric materials in 3D, 2D plane strain and plane
stress, and axial symmetry can be modeled.

O V E R V I E W O F T H E U S E R ’S G U I D E | 31
 • The Magnetostriction Interface combines a Solid Mechanics with a Magnetic Fields
interface. Using this interface you can solve problems in the magnetostrictive field
with linear as well as nonlinear material models.
• The Fluid-Structure Interaction Interface, found under the Fluid Flow branch, is
described in this chapter. The physics interface combines fluid flow with solid
mechanics to capture the interaction between the fluid and the solid structure.
• The Fluid-Structure Interaction, Fixed Geometry Interface, found under the Fluid
Flow branch, is described in this chapter. The interface can be used to model
phenomena where a fluid and a deformable solid structure affect each other.

32 | CHAPTER 1: INTRODUCTION
 2

Structural Mechanics Modeling

The goal of this chapter is to give you an insight on how to approach the modeling
of various structural mechanics problems.

Some physics interfaces and features discussed in this chapter are only available with
certain products. For a detailed overview of the functionality available in each
product, visit http://www.comsol.com/products/specifications/

33
 In this chapter:

• Study Types • Modeling Geometric Nonlinearity

• Selecting the Physics Interface • Contact Modeling
• Selecting Discretization • Springs and Dampers
• Coupling Different Element Types • Defining Multiphysics Models
• Applying Loads • Thermally Coupled Problems
• Defining Constraints • Computing Mass Properties
• Calculating Reaction Forces • Pre-tensioned Bolts
• Introduction to Material Models • Stress Linearization
• Modeling Piezoelectric Problems • Solver Settings
• Mechanical Damping and Losses • Part Libraries

34 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

Study Types
Introduction
In this section, you will find information about when and how to apply the study types
which are available for structural mechanics problems:

• Stationary Analysis
• Eigenfrequency Analysis
• Mode Analysis
• Time Domain Analysis
• Frequency Domain Analysis
• Modal Superposition
• Harmonic Perturbation
• Modal Reduced Order
• Linearized Buckling Analysis
• Bolt Pre-Tension Study

For general information about study types and solvers, see Studies and
Solvers in the COMSOL Multiphysics Reference Manual

Stationary Analysis
You can consider a structural mechanics problem as stationary if the following two
criteria are fulfilled:

• The loads vary so slowly that inertial forces are negligible. Problems of this type are
referred to as quasi-static.
• There are no explicit time dependencies in the material model. Viscoelasticity and
creep have such time dependences.

To perform this type of analysis, you use a Stationary study step.

S T U D Y TY P E S | 35
 In many cases, there is a variation in the load, even though the solution for each value
of the load can be considered as stationary. There are three conceptually different
cases:

• The load values are independent; it is just a number of different load cases you want
to compute. The load case handling functionality described in Load Cases is well
suited for this purpose.
• You want to study a nonlinear problem where the solution is path dependent, or
where the load must be increased in small increments in order to obtain a converged
solution. In this case you should use the parametric continuation solver. Create a
parameter under Global Definitions>Parameters, which you use to control the
variation of the load. Then select Auxiliary sweep under Study Extensions in the
settings for the Stationary solver. In the table for the auxiliary sweep parameters, add
the load controlling parameter, and define its range of variation.
• In a multiphysics problem, another physical quantity might be truly
time-dependent, but on a time scale that is “slow” from the structural mechanics
point of view. This is usually the case with, for example, problems coupled to heat
transfer or diffusion. If the problem also is one-way coupled in the sense that the
structural deformations do not affect the other physics, it will be unnecessarily
expensive to solve also the structural problem in the time domain, irrespective of
whether it is linear or nonlinear. In this situation, you should first solve the other
physics in a time-dependent study, and then the structural mechanics problem in a
subsequent stationary study step using the time t as the parameter in the auxiliary
sweep.

CONSTRAINTS
A stationary problems is solvable only if the structure is sufficiently constrained. There
must not be any possible rigid body modes; thus no stress-free deformation states are
allowed.

For a more detailed discussion about sufficient constraints, see Rigid body
motion.

Eigenfrequency Analysis
An eigenfrequency study solves for the eigenfrequencies (natural frequencies) and the
shapes of the corresponding eigenmodes.

36 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

When performing an eigenfrequency analysis, you can specify whether to look at the
mathematically more fundamental eigenvalue, λ, or the eigenfrequency, f, which is
more commonly used in a structural mechanics context. The relation between the two
is

λ
f = – ---------
2πi

where i is the imaginary unit.

Because only the shape and not the size of the modes (eigenvectors) have physical
significance, the computed modes can be scaled arbitrarily. You can select the method
for scaling in the Eigenvalue Solver node of the solver sequence. If output of mass
participation factors is required, then Scaling of eigenvectors must be set to Mass
matrix. This means that the eigenmodes U are orthogonalized with respect to the mass
matrix M so that

T
U i MU i = 1 (2-1)

This is a common choice for the scaling of eigenvectors within the structural mechanics
field.

The mass (or ‘modal’) participation factor for mode i in direction j, rij, is defined as

T
r ij = U i Md j

Here, dj is a vector containing unity displacement in all degrees of freedom

representing translation in direction j. The mass participation factor gives an indication
of to which extent a certain mode might respond to an excitation in that direction.

The mass participation factors have the important property that when their squares for
a certain direction are summed over all modes, this sum approaches the total mass of
the model:

 rij =
2
m tot
i=1

If you use a modal superposition method to solve a forced response problem, then in
practice you do not solve for all possible modes but just a limited number. This
property if the mass participation factors can then be used for investigating how well a
certain number of selected modes represent the total mass of the system.

S T U D Y TY P E S | 37
 The mass participation factors are available as a global variables, and these can be
shown in a table using a Global Evaluation node under Derived Values in the Results
branch, for example. The participation factor variables are available as predefined
variables in the Solver submenu.

For an example showing how to compute modal mass, see In-Plane

Framework with Discrete Mass and Mass Moment of Inertia:
Application Library path Structural_Mechanics_Module/
Verification_Examples/inplane_framework_freq.

For an example showing an eigenfrequency computation in a model

having a rigid body mode, see Eigenfrequency Analysis of a Free
Cylinder: Application Library path Structural_Mechanics_Module/
Verification_Examples/free_cylinder.

In the COMSOL Multiphysics Reference Manual:

• Eigenvalue Solver
• Studies and Solvers
• Derived Values and Tables

It is possible to compute eigenfrequencies for structures which are not fully

constrained; this is sometimes referred to as free-free modes. For each possible rigid
body mode, there is one eigenvalue which in theory is zero. The number of possible
rigid body modes for different geometrical dimensions is shown in the table below.
TABLE 2-1: NUMBER OF POSSIBLE RIGID BODY MODES

DIMENSION NUMBER OF RIGID BODY MODES

3D 6 (3 translations + 3 rotations)
2D axisymmetric 1 (Z-direction translation)
2D (solid, beam, truss) 3 (2 translations + 1 rotation)
2D (plate) 3 (1 translation + 2 rotations)

In a piezoelectric model, one more zero eigenfrequency could appear if you have not
set a reference value for the electric potential.

In practice, the natural frequencies of the rigid body modes are not computed as
exactly zero, but can appear as small numbers which can even be negative or complex.
If rigid body modes are present in the model, then it is important to use a nonzero

38 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

value in the Search for eigenfrequencies around text field in the settings for the
Eigenfrequency study step. The value should reflect the order of magnitude of the first
important nonzero eigenfrequency.

DAMPING
If any type of damping is included in the model, an eigenfrequency solution
automatically returns the damped eigenvalues. The eigenfrequencies and, in general,
also the mode shapes are complex in this case. A complex-valued eigenfrequency can
be interpreted so that the real part represents the actual frequency, and the imaginary
part represents the damping. The ratio between the imaginary and real parts of the
eigenfrequency is the relative damping of the corresponding eigenmode,

imag ( ω i )
ζ i = -------------------------
real ( ω i )

In a complex mode shape there are phase shifts between different parts of the
structure, so that not all points reach the maximum at the same time under free
vibration.

Some damping types will still give real valued eigenmodes, this is the case for Rayleigh
damping and loss factor damping.

PRESTRESSED ANALYSIS
In a loaded structure, the natural frequencies may be shifted due to stress stiffening.
With the Prestressed Analysis, Eigenfrequency study type you can compute
eigenfrequencies taking this effect into account.

Prestressed Structures

Mode Analysis
The Mode Analysis study type( ) is available with the Solid Mechanics interface in
2D plane strain.

Elastic waves can propagate over large distances in structures like rails and pipes, with
a generic name referred to as waveguides. After some distance of propagation in a
waveguide of uniform cross section, such guided waves can be described as a sum of
just a few discrete propagating modes, each with its own shape and phase speed. The

S T U D Y TY P E S | 39
 equation governing these modes can be obtained as a spatial Fourier transform of the
linearized time-harmonic equation in the waveguide axial z direction or by inserting
the assumption that the mode is harmonic in space,

u = ue – ikz z

and eliminating all out-of-plane z dependence.

Similar to the full time-harmonic equation, the transformed equation can be solved at
a given frequency with a nonzero excitation for most axial wave numbers kz. But at
certain discrete values the equation breaks down. These values are the propagation
constants or wave numbers of the propagating or evanescent waveguide modes. The
eigenvalue solver can solve for these propagation constants together with the
corresponding mode shapes.

The propagating wave number is a function of the frequency. The relation

between the two is commonly referred to as a dispersion curve.

The most common use for the Mode Analysis is to define sources for a subsequent
time-harmonic simulation. If there is a component with one or more waveguide
connections, its behavior can be described by simulating its response to the discrete set
of propagating modes on the waveguide opening cross sections.

• Out-of-plane and Circumferential Modes in the Structural Mechanics

Theory Chapter
• Studies and Solvers and Mode Analysis in the COMSOL Multiphysics
Reference Manual

Time Domain Analysis

There are two classes of problems where a stationary solution cannot be used:

• When the inertial forces no longer are negligible, the full problem as given by
Newton’s first law must be solved.
• When there are time dependencies in the material model, as for creep or
viscoelasticity.

40 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

The most general way of doing that is to use a Time Dependent study. In this type of
analysis, you can incorporate any type of nonlinearity, and there are no limitations on
the time dependence of the loads.

A time domain solution can be preceded by a stationary study, if for example prestress
effects are needed.

For a linear problem including inertia, using the modal superposition method is often
much more efficient than using the standard direct method.

SOLVER SELECTION
The two classes of dynamic problems presented above have quite different properties.
The inertial forces in the full structural dynamics problem contain second-order time
derivatives of the displacements, whereas creep and viscoelasticity only have first-order
derivatives. The physical and numerical properties of these equations differ
significantly.

There are two general solvers for time-dependent problems in COMSOL

Multiphysics.

• The Generalized alpha method, which is recommended for structural dynamics

problems. This is the default solver if Structural Transient Behavior is set to Include
inertial terms in the physics interface settings.
• The BDF method, which is recommended for the first-order problems. This is the
default solver if Structural Transient Behavior is set to Quasi-static in the physics
interface settings.

In the COMSOL Multiphysics Reference Manual:

• Time-Dependent Solver
• Studies and Solvers

Frequency Domain Analysis

In a frequency domain analysis, you study the response to a harmonic steady state
excitation for certain frequencies. Such a steady state can prevail once all transient
effects have been damped out.

The response must be linear, so that the single frequency harmonic excitation gives a
pure harmonic response with the same frequency. The model may however contain
nonlinearities; the harmonic response is computed around a linearization point. In

S T U D Y TY P E S | 41
 such a case, the frequency domain analysis can be considered as a very small
perturbation around the linearization point.

Harmonic Perturbation

All loads and responses are in general complex valued quantities. If not all loads have
the same phase, you can prescribe the phase of a certain in two ways:

• Add a Phase subnode to the load, in which you give the phase angle.
• Enter the load as a complex value, for example as
100[N]*(1+0.5*i)/sqrt(1.25).

Most results of a frequency domain analysis are complex. In results evaluation, the real
value of any result quantity will be shown. Assuming that you want to display for
example the displacement in the x-direction, u, you have following options:

• Plot u or real(u). This gives the displacement at zero phase angle.

• Plot imag(u). This gives the displacement at a phase angle of 90 degrees.
• Plot abs(u). This gives the amplitude of the displacement.
• Plot arg(u). This gives the phase angle of the displacement.

The reference phase, with respect to which the results above are reported.

42 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

Result quantities that are nonlinear in terms of the displacements, such as principal
stresses, should be interpreted with great care. They will in general not be harmonic,
so the amplitude and phase information is not reliable.

Some extra variables for postprocessing are created in a frequency-domain

analysis. As an example, in the Solid Mechanics interface the following
variables are defined:

• solid.disp—Norm of displacement (at current phase angle)

• solid.vel—Norm of velocity (at current phase angle)
• solid.acc—Norm of acceleration (at current phase angle)
• solid.disp_rms—RMS displacement over a cycle
• solid.vel_rms—RMS velocity over a cycle
• solid.acc_rms—RMS acceleration over a cycle
• solid.uAmpX—Amplitude of displacement in X direction
• solid.uAmp_tX—Amplitude of velocity in X direction
• solid.uAmp_ttX—Amplitude of acceleration in X direction
• solid.uPhaseX—Phase of X displacement, in radians
• solid.uPhase_tX—Phase of X velocity, in radians
• colid.uPhase_ttX—Phase of X displacement, in radians

The components in other coordinate directions are obtained by replacing

X by another coordinate name.

PRESTRESSED ANALYSIS
The shift in the natural frequencies in a prestressed structure may have a significant
effect on the frequency response. This is particularly important when the frequencies
of the load are close to any of the natural frequencies of the structure. The Prestressed
Analysis, Frequency Domain study type is designed to take such effects into account.

The prestress loading can include a contact analysis, in which case the subsequent
frequency domain analysis provide as linearization around the current contact state.

• Prestressed Structures
• Harmonic Perturbation

S T U D Y TY P E S | 43
 OBTAINING A TIME HISTORY
Sometimes you want to study the time history over a period for the results of a
frequency domain analysis. You can do that by adding a Frequency to Time FFT study
step. The frequency response results are then viewed as terms in a Fourier series, which
can be transformed to time domain. It is possible combine the results for several
frequencies into a single time history, under the assumption that they are all multiples
of the same fundamental frequency.

For an examples showing how to obtain a time history from frequency

domain results, see

• Viscoelastic Structural Damper: Application Library path

Structural_Mechanics_Module/Dynamics_and_Vibration/
viscoelastic_damper_frequency.
• Vibration Analysis of a Deep Beam: Application Library path
Structural_Mechanics_Module/Verification_Examples/
vibrating_deep_beam.

Modal Superposition
Analyzing forced dynamic response for large models can be very time-consuming. You
can often improve the performance dramatically by using the modal superposition
technique. The following requirements must be met for a modal solution to be
possible:

• The analysis is linear. It is possible, however, that the structure has been subjected
to a preceding nonlinear history. The modal response can then be a linear
perturbation around that state.
• There are no nonzero prescribed displacements.
• The important frequency content of the load is limited to a range that is small when
compared to all the eigenfrequencies of the model, so that its response can be
approximated with a small number of eigenmodes. In practice, this excludes wave
and shock type problems.
• If the modal solution is performed in the time domain, all loads must have the same
dependency on the time.

When using the Structural Mechanics Module, there are two predefined study types
for modal superposition: Time-Dependent Modal and Frequency-Domain Modal.

44 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

Both these study types consist of two study steps: One step for computing the
eigenfrequencies and one step for the modal response.

In practice, you have often computed the eigenfrequencies already, and then want to
use them in a modal superposition. In this case, start by adding an empty study, and
then add a Time-Dependent Modal or Frequency-Domain Modal study step to it. After
having added the study step this way, you must point the modal solver to the solution
containing the eigenfrequencies and eigenmodes. You do this by first selecting Show
default solver at the study level, and then selecting the eigenfrequency solution to be
used in the Eigenpairs section of the generated modal solver.

In a modal superposition, the deformation of the structure is represented by a linear

combination of its eigenmodes. The amplitudes of these modes are the degrees of
freedom of the reduced problem. You must select which eigenmodes to include in the
analysis. This choice is usually based on a comparison between the eigenfrequencies of
the structure and the frequency content of the load. As a rule of thumb, select
eigenmodes up to approximately twice the highest frequency of the excitation.

In the modal superposition formulation in COMSOL, the full model is projected onto
the subspace spanned by the eigenmodes. A problem having the number of degrees
equal to the number of included modes is then solved. This means that there are no
restrictions on the type of damping that can be used in a modal superposition analysis,
as it would have been the case if the modal equations were assumed to be totally
decoupled.

For many common cases, the modal superposition analysis is not sensitive to whether
the eigenmodes were computed using damping or not. The reason is that the
eigenmodes of problems with Rayleigh damping and loss factor damping can be shown
to be identical to those of the undamped problem, so that the projection to the
subspace spanned by the eigenmodes is the same in both cases. For more general
damping, it is however recommended that you suppress all contributions to the
damping during the eigenfrequency step, and thus base the modal superposition on
the solution to the undamped eigenfrequency problem.

FREQUENCY DOMAIN ANALYSIS

All loads are assumed to have a harmonic variation. This is a perturbation type analysis,
so only loads having the Harmonic perturbation property selected are then included in
the analysis.

S T U D Y TY P E S | 45
 TIME-DEPENDENT ANALYSIS
Only the factor of the load which is independent of time should be specified in the load
features. The dependency on time is specified as Load factor under the Advanced section
of the modal solver. This factor is then applied to all loads.

• Modal Solver and Studies and Solvers in the COMSOL Multiphysics

Reference Manual
• Mechanical Damping and Losses

For an example showing how to perform modal superposition in time and

frequency domain, see Various Analyses of an Elbow Bracket:
Application Library path Structural_Mechanics_Module/Tutorials/
elbow_bracket.

Harmonic Perturbation
Analyses in the frequency domain assume that the problem your study is linear, at least
with respect to the response to the harmonic excitation. There may be other
nonlinearities, such that the structure has responded nonlinearly to a previous loading.
This loading could, for example, have caused a large rotations or prestress of a rubber
membrane.

The concept of harmonic perturbation is in COMSOL Multiphysics used for

distinguishing the linear harmonic analysis from a possible prestress analysis. The most
important implication is that if a load has the Harmonic Perturbation selection, it is
applied only in a study that is of the perturbation type. A load without this selection is,
on the other hand, ignored in such a study. In this way two sets of loads can be
distinguished from each other. Technically speaking, the effect of marking a load as
Harmonic Perturbation is that the linper() operator is applied to the value of the load.

The default settings for the different structural mechanics study types in the frequency
domain are summarized in Table 2-2.
TABLE 2-2: DEFAULT PERTURBATION SETTINGS FOR STRUCTURAL MECHANICS STUDY TYPES

STUDY TYPE STUDY STEP PERTURBATION

Frequency Domain Frequency Domain No

46 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

TABLE 2-2: DEFAULT PERTURBATION SETTINGS FOR STRUCTURAL MECHANICS STUDY TYPES

STUDY TYPE STUDY STEP PERTURBATION

Prestressed Analysis, Frequency Stationary No

Domain Frequency-Domain, Perturbation Yes
Frequency-Domain, Modal Eigenfrequency N/A
Frequency-Domain, Modal Yes
Note the following:

• With the default settings you cannot use the same set of loads for a Frequency Domain
and a Frequency-Domain, Modal study because only the latter responds to
perturbation loads.
• You can change the behavior of a Frequency Domain study to be of the perturbation
type by modifying the solver sequence. In the General section of the settings for the
Stationary Solver, change Linearity to Linear perturbation.
• A solver that does not have Linearity set to either Linear perturbation or Linear may
respond to nonlinear effects. There are multiphysics problems where this is wanted
because there may be a nonlinearity in another physics, even though the harmonic
solution within structural mechanics is linear. But if there are nonlinearities within
the structural mechanics parts of the model, you must be careful with the solver
settings.

In a Frequency-Domain, Perturbation study step, that is when Linearity is set

to Linear perturbation, geometric nonlinearity will be incorporated in the
sense that there is a split between the material and spatial frames. This
makes it possible to take into account for example stiffness from follower
loads, and to use a contact solution as linearization point.

This frame split was introduced in version 5.3. As an effect, models

created in an earlier version, in which some expressions have a frame
dependency may produce results that differ from before. Examples of
such cases are:

• Pressure loads
• Loads defined in coordinate systems with deformation dependent axis
orientation
• User-defined expressions containing spatial (‘lowercase’) coordinates

S T U D Y TY P E S | 47
 In the COMSOL Multiphysics Reference Manual:

• Frequency-Domain Perturbation Study Step

• Harmonic Perturbation — Exclusive and Contributing Nodes
• Built-In Operators (linper() operator)

For most load types, the use of Harmonic Perturbation is straightforward, but some
cases need a more detailed discussion:

• A Rigid Connector can be assigned an Harmonic Perturbation subnode in which you

can prescribe harmonic perturbation values to constrained degrees of freedom. If
you have added Applied Force or Applied Moment nodes under a Rigid Connector, you
can independently assign Harmonic Perturbation to these nodes, so that the loads are
considered as being of the perturbation type.
• Even though initial stresses and strains are not usually considered as loads, you can
assign Harmonic Perturbation also to the Initial Stress and Strain nodes.

For an example showing how to use harmonic perturbation, see

Bracket—Frequency-Response Analysis: Application Library path
Structural_Mechanics_Module/Tutorials/bracket_frequency.

Modal Reduced Order

Reduced-order modeling seeks to reduce the number of degrees of freedom in a
physical model, whilst still retaining the essential physics. For a lightly damped
resonant system driven at one of its resonant frequencies, it is reasonable to consider
only the contributions to the system of a small number (m) of modes within the signal
bandwidth. In some cases a single mode is sufficient. A system with n degrees of
freedom has mass, stiffness, and damping matrices of size n-by-n. A reduced-order
representation of the system considering m modes has size m-by-m. The reduction in
complexity of the system, and the computational speed up is therefore significant when
m « n. This section describes the theory of the reduced-order system and gives
guidelines on how to obtain reduced-order models from a COMSOL model.

This can be employed in two different ways: Either you can use the built-in modal
solvers for the time or frequency domain, or you can export the small equivalent system
and analyze it outside COMSOL, for example, as a component in a larger system
simulation.

48 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

THE MODAL COORDINATE SYSTEM
Consider a mechanical system, with n degrees of freedom, described by an equation of
the form

·· ·
Mu + Du + Ku = F (2-2)

where u is the displacement vector (size: n-by-1), K is the stiffness matrix (size:
n-by-n), D is the damping matrix (size: n-by-n), and M is the mass matrix (size:
n-by-n). In the frequency domain the problem takes the form

2
– ω M u 0 + iωDu 0 + Ku 0 = F

where u = u0eiωt.

Initially consider the system in the absence of damping and forces. The undamped
system has n eigenvalues ωi, which satisfy the equation

ˆ = ω 2 Mu
Ku ˆ (2-3)
i i i

These eigenvectors can be shown to be orthogonal with respect to both M and K:

ˆ T Mu
u ˆ = 0 i ≠ j , ωi ≠ ωj (2-4)
j i

ˆ T Ku
u ˆ = 0 i ≠ j , ωi ≠ ωj (2-5)
j i

Next the following n-by-n matrix is constructed, with columns taken from the n
eigenvectors:

ˆ ,u
U = u ˆ …u
ˆ
1 2 n

Then consider the following matrix:

ˆ T Mu
u ˆ u ˆT ˆ
1 1 1 Mu 2
ˆ T Mu
u ˆ u ˆT ˆ
2 1 2 Mu 2
T
U MU =
ˆ T Mu
u ˆ ˆT ˆ
n–1 n – 1 u n – 1 Mu n
ˆ T Mu
u ˆ ˆ T Mu
u ˆ
n n–1 n n

S T U D Y TY P E S | 49
 From Equation 2-4 it is clear that this is a diagonal matrix. Similarly from Equation 2-5
it is clear that UTKU is also diagonal.

From the properties of the eigenvectors it is possible to expand any function in terms
of the eigenvectors. Thus the displacement u can be written as:

 ai ui
u = ˆ

i=1

This equation can also be expressed in the form:

u = Ua (2-6)

where a is a column vector containing the coefficients ai as rows. In general a is

time-dependent.

Now consider the original equation: Equation 2-2. First substitute for u using
Equation 2-6. Then transform the equation to the modal coordinate system by
premultiplying by UT. This gives:

T ·· T · T T
U MUa + U DUa + U KUa = U F (2-7)

It has already been established that the matrices UTMU and UTKU are diagonal and
frequently a damping model is chosen that results in a diagonal damping matrix. For
example, in Rayleigh damping D = αM + βK, where α and β are constants. For a
general damping, the transformed damping matrix is however not diagonal. As an
alternative, a damping ratio, ζi, can be assigned to each mode.

EIGENVALUE SCALING
The precise form of Equation 2-7 is determined by the normalization adopted for the
eigenfunctions. In structural applications the eigenfunctions are often normalized such
that UTMU = I. This is referred to as mass matrix scaling in the eigenvalue solver. In
this case Equation 2-3 gives

ˆ T Ku
u ˆ T Mu
ˆ = ω 2u ˆ = ω2
i i i i i i

so that

T 2
U KU = diag ( ω i )

50 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

where diag(ωi2) is the diagonal matrix with diagonal elements ωi2. Similarly, if
damping ratios for each mode are defined, the damping matrix can be expressed in the
form

T
U DU = diag ( 2ζ i ω i )

Thus if mass matrix scaling is used Equation 2-7 takes the form

·· · 2 T
a + diag ( 2ζ i ω i )a + diag ( ω i )a = U F (2-8)

It is also possible to scale the eigenvectors so that the point of maximum displacement
has unit displacement. This is referred to as max scaling in the eigenvalue solver. For
an individual mode this scaling has a simple physical interpretation—the
corresponding component of a, ai, is the amplitude of the i:th mode, measured at the
point of maximum displacement, when the mode is driven by the force F. In this case
Equation 2-7 takes the form

·· · T
diag ( m eff , i )a + diag ( c eff , i )a + diag ( k eff , i )a = U F (2-9)

Here meff,i is the effective mass of the i:th mode, ceff,i = 2meff,iξiωi is the effective
damping parameter for the mode, and keff,i is the effective spring constant. Each
element of the vector UTF gives the force component that acts on each of the
respective modes.

When using max scaling, it is the largest value of a degree of freedom

which is scaled to 1. The total displacement in that node will thus be
between 1 and 3 .

If degrees of freedom other than displacements are active in the

eigenfrequency problem, the maximum value may occur in another type
of degree of freedom such as electric potential or pressure. Consequently,
the peak displacement in that mode can then be less than 1.

REDUCED-ORDER MODELS
The preceding discussion did not consider how to reduce the number of degrees of
freedom in the system. For systems in which the vector UTF has only a few significant
components (for example, components i = 1, …, m where m « n) the following
approximation can be made:

S T U D Y TY P E S | 51
 m

 ai ui
u≈ ˆ

i=1

The expression for u in matrix becomes:

u = U'a'

where U' is now an m-by-n and a' is a vector with m components. The equation
system in modal coordinates now takes the form

T ·· T · T T
U' MU'a' + U' DU'a' + U' KU'a' = U' F (2-10)

The matrices U'TMU', U'TDU', and U'TKU' now have dimensions m-by-m. Similarly
the vector U'TF has m components. This results in a significant reduction in the
system complexity.

REDUCED-ORDER MODELS WITH PHYSICAL DAMPING

If physically relevant damping is present in the system, the above theory must be
modified as the damping matrix is no longer diagonal in the modal coordinate system.
COMSOL can still handle this case as the modal solver does not make the assumption
that any of the matrices are diagonal. In this case the eigenvalues become complex and
the eigenvectors split into right and left eigenvectors. The right eigenvectors Ur are
solutions of the equation:

2 ˆ + iω Du
–ωi M u ˆ + Kuˆ
r ,i i r ,i r ,i = F

As in the previous section, for a reduced set of modes, it is assumed that:

u = U' r a'

where U'r is the n-by-m matrix containing the right eigenvectors chosen for the modal
analysis. Once again a' is a vector with m components. The system in modal
coordinates takes the form

T ·· T · T T
U' l MU' r a' + U' l DU' r a' + U' l KU' r a' = U' l F

where U'l is the n-by-m matrix containing the left eigenvectors chosen for the modal
analysis.

52 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

The matrices U'lTMU'r, U'lTDU'r, and U'lTKU'r are no longer necessarily diagonal.
The modal solver accepts any linearly independent set of vectors to project the solution
vector and equations onto and constructs the reduced-order system accordingly.

CONSTRUCTING REDUCED-ORDER MODELS WITH COMSOL

To obtain the data necessary to construct reduced-order models from a COMSOL
simulation the Modal Reduced Order Model study step is available. Add this study step
after an existing Eigenvalue study step, by right-clicking the Study 1 node and selecting
Study Steps>Time Dependent>Modal Reduced Order Model. Then solve the model. After
the model has solved, right-click the Results>Derived Values node and select System
Matrices. In the output section choose the Matrix to display in the list. The mass matrix
corresponds to the matrix U'lTMU'r the stiffness matrix corresponds to U'lTKU'r, and
the damping matrix corresponds to U'lTDU'r. The vector U'lTF is available as the load
vector. The Modal Reduced Order Model exports these matrices in a format that
respects the normalization of the preceding Eigenvalue study. To change this select the
Study 1>Solver Configurations>Solver 1>Eigenvalue Solver 1 node and change the Scaling
of Eigenvectors setting under the Output section. Use the Max setting if an equivalent
Mass-Spring-Damper system is required, in which case the modal amplitude
corresponds to the maximum displacement of the mode.

Linearized Buckling Analysis

A linearized buckling analysis can be used for estimating the critical load at which a
structure becomes unstable. This is a predefined study type that consists of two study
steps: An initial step in which a unit load is applied to the structure, and a second step
in which an eigenvalue problem is solved for the critical buckling load.

The idea behind this type of analysis can be described in the following way:

Consider the equation system to be solved for a stationary load f,

Ku = ( K L + K NL )u = f

Here the total stiffness matrix, K, has been split into a linear part, KL, and a nonlinear
contribution, KNL.

In a first order approximation, KNL is proportional to the stress in the structure and
thus to the external load. So if the linear problem is solved first for an arbitrary initial
load level f0,

KL u0 = f0

S T U D Y TY P E S | 53
 then the nonlinear problem can be approximated as

( K L + λK NL ( u 0 ) )u = λf 0

where λ is called the load multiplier.

An instability is reached when this system of equations becomes singular so that the
displacements tend to infinity. The value of the load at which this instability occurs can
be determined by, in a second study step, solving an eigenvalue problem for the load
multiplier λ.

( K L + λK NL ( u 0 ) )u = 0

COMSOL reports a critical load factor, which is the value of λ at which the structure
becomes unstable. The corresponding deformation is the shape of the structure in its
buckled state.

The level of the initial load used is immaterial since a linear problem is solved. If the
initial load actually was larger than the buckling load, then the critical value of λ is
smaller than 1.

You should not select geometric nonlinearity in a linearized buckling

analysis. Since the nonlinear effect of the stress is already taken into
account in the formulation, an explicit use of geometric nonlinearity
would make the computed buckling load dependent on the load level
used in the pre-load study step.

Be aware that for some structures, the true buckling load can be
significantly smaller that what is computed using a linearized analysis.
This phenomenon is sometimes called imperfection sensitivity. Small
deviations from the theoretical geometrical shape can then have a large
impact on the actual buckling load. This is especially important for curved
shells.

For a structure that exhibits axial symmetry in the geometry, constraints,

and loads, the critical buckling mode shape can still be non-axisymmetric.
A full 3D model should always be used when computing buckling loads.

54 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

 Studies and Solvers and Linear Buckling in the COMSOL Multiphysics
Reference Manual

• Bracket — Linear Buckling Analysis: Application Library path

Structural_Mechanics_Module/Tutorials/bracket_linear_buckling
• Linear Buckling Analysis of a Truss Tower: Application Library path
Structural_Mechanics_Module/Buckling/truss_tower_buckling

Bolt Pre-Tension Study

The Bolt Pre-Tension study step is a special case of a Stationary study step, where the
special degrees of freedoms used for modeling prestressed bolts are solved for. In all
other study types, these degrees of freedom are inactive. Typically, you include a Bolt
Pre-Tension study step as the first step in a study in order to simulate the state after the
assembly of a bolted joint. You can the add any other types of study steps for
computing the effects of the service loads.

Pre-tensioned Bolts

Prestressed Bolts in a Tube Connection: Application Library path

Structural_Mechanics_Module/Contact_and_Friction/tube_connection

S T U D Y TY P E S | 55
 Selecting the Physics Interface
The structural mechanics products contain a number of physics interfaces for a wide
range of applications. This section contains some guidelines for how to select an
appropriate physics interface for your analysis.

The basic physics interfaces for structural mechanics are:

• Solid Mechanics—General modeling of objects with extension in all directions

• Shell and Plate—For objects which are thin in one direction, but have significant
bending stiffness
• Membrane—For objects which are thin in one direction, and have negligible
bending stiffness
• Beam—For objects where two directions have significantly smaller dimensions than
the third; significant bending stiffness
• Truss—For objects where two directions have significantly smaller dimensions than
the third; only axial forces can be transmitted

For a detailed overview of the functionality available in each product, visit

http://www.comsol.com/products/specifications/

Solid Mechanics
The Solid Mechanics interface offers the most general modeling of structural
mechanics problems and is formulated based on general principles of continuum
mechanics. It is the interface which contains the largest number of material models,
and the most advanced boundary conditions. It is also the only physics interface that
supports contact analysis.

The drawback with using solid elements is that the models can become
computationally expensive, especially in 3D. For structures which are thin or slender,
you should consider using one of the specialized physics interfaces.

56 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

3D SOLID GEOMETRY
The degrees of freedom (dependent variables) in 3D are the global displacements u, v,
and w in the global x, y, and z directions, respectively.

Figure 2-1: Loads and constraints applied to a 3D solid using the Solid Mechanics
interface.

2D GEOMETRY

Plane Stress
The plane stress variant of the 2D physics interface is useful for analyzing thin in-plane
loaded plates. For a state of plane stress, the out-of-plane components of the stress
tensor are zero.

Figure 2-2: Plane stress models plates where the loads are only in the plane; it does not
include any out-of-plane stress components.

The 2D physics interface for plane stress allows loads in the x and y directions, and it
assumes that these are constant throughout the material’s thickness, which can vary
with x and y. The plane stress condition prevails in a thin flat plate in the xy-plane
loaded only in its own plane and without any z direction restraint.

SELECTING THE PHYSICS INTERFACE | 57

 Plane Strain
The plane strain variant of the 2D physics interface that assumes that all out-of-plane
strain components of the total strain εz, εyz, and εxz are zero.

Figure 2-3: A geometry suitable for plane strain analysis.

Loads in the x and y directions are allowed. The loads are assumed to be constant
throughout the thickness of the material, but the thickness can vary with x and y. The
plane strain condition prevails in geometries, whose extent is large in the z direction
compared to in the x and y directions, or when the z displacement is in some way
restricted. One example is a long tunnel along the z-axis where it is sufficient to study
a unit-depth slice in the xy-plane.

AXISYMMETRIC GEOMETRY
The axisymmetric variant of the Solid Mechanics interface uses cylindrical coordinates
r, ϕ (phi), and z. Loads are independent of ϕ, and the axisymmetric variant of the
physics interface allows loads only in the r and z directions.

The 2D axisymmetric geometry is viewed as the intersection between the original

axially symmetric 3D solid and the half plane ϕ = 0, r ≥ 0. Therefore the geometry is
drawn only in the half plane r ≥ 0 and recover the original 3D solid by rotating the 2D
geometry about the z-axis.

Figure 2-4: Rotating a 2D geometry to recover a 3D solid.

58 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

Shell and Plate
The Shell interface is useful when the object is thin in one direction. Structures built
from welded or bolted flat plates is an archetypal shell structure, and so are pressure
vessels.

The Plate interface is a specialization of the Shell interface, used for 2D modeling in
the XY-plane. A plate model has its main action in bending out of the plane, but it can
also treat in-plane forces. If the loads act only in the plane, using Solid Mechanics with
the Plane Stress option is a better choice.

Shells are modeled on boundaries, and the transverse direction is represented only by
the mathematical model. The degrees of freedom consist of displacements and
rotations at the modeled boundary. This results in an assumption where the in-plane
stresses and strains vary linearly through the thickness, and the stress in the thickness
direction is zero. The thickness of a shell does not have to be constant, although this
is by far the most common case.

The Shell and Plate interfaces can be used both for ‘thin’ and ‘thick’ shells. Shear
deformations are taken into account; this is usually called Mindlin theory. The material
model is linear elastic.

When modeling with shells, it important to keep track of ‘top’ and ‘bottom’ side when
applying loads and interpreting the results.

The in-plane stiffness of a shell is proportional to the thickness h, while the bending
stiffness is proportional to h3. The difference in stiffness along different directions can
thus become very large. When an object is very thin, a shell model may be numerically
ill-posed. It is then better to use the Membrane interface.

Membrane
The Membrane interface can be used for very thin objects, like cloth, where only
in-plane forces are important. Membranes can be considered as plane stress elements,
but with an arbitrary, possibly curved configuration is space.

In most applications, a membrane must be pre-tensioned in order to have a stable

configuration, so it will almost invariably be used in a geometrically nonlinear analysis.
The only exception is when it is used as a ‘cladding’ on top of a solid, since it will then
be stabilized by the solid.

In the Membrane interface a large number of different material models can be used.

SELECTING THE PHYSICS INTERFACE | 59

 Beam
A beam is an abstract model where only the extension in the axial direction is modeled
explicitly on an edge. The cross section is specified in terms of properties such as area
and moments of inertia.

The exact stress distribution in the beam is not explicitly modeled. It is actually not
even fully determined by the cross-sectional properties. Instead, six (in 3D) resultant
section forces are used: axial force, shear forces in two perpendicular directions, two
bending moments, and one twisting moment.

Two formulations are available in the Beam interface:

• The classical Euler-Bernoulli beam theory, which is applicable for slender beams.
• Timoshenko theory, where shear deformations are taken into account. This allows
the Beam interface to be applied to rather thick beams.

Truss
The Truss interface has four distinct purposes:

• Modeling of trusses, consisting of straight bars carrying only axial forces

• Modeling of cables and wires
• As reinforcements, used in conjunction with other physics interfaces
• For modeling discrete springs and dampers, often added to other physics interfaces

For a truss model, only one geometrical property is needed, the cross section area. The
material model can be linear elastic or elastoplastic. There is also a special material
model for creating spring/damper data.

The truss element has no stiffness in the directions perpendicular to its extension. For
trusses, this is usually not a problem since they are designed such that each member is
stabilized by its neighbors. For cables however, a prestress is necessary to keep them
stable.

60 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

Selecting Discretization
Shape Function Order
In structural mechanics analysis, the focus is often on the stresses and strains rather
than on the displacements. Since the strains are derivatives of the displacement field,
the accuracy of the strains will be one order less than the accuracy of the displacements.
For this reason, second-order shape functions are used as default in most of the
structural mechanics interfaces. Often this gives the best trade-off between model size
and accuracy.

It is well known that using first-order shape functions in solid mechanics will give an
overly stiff solution, unless a very fine mesh is used. This is especially noticeable for
triangular and tetrahedral elements.

If the purpose of the analysis is only to compute stiffness, rather than stresses, the use
of linear shape functions can still be justified. This is the default choice in the
Multibody Dynamics interface.

If the solution contains discontinuities, like when some type of front is moving
through the material, first-order elements and a fine mesh is often a good choice, since
the advantage of the higher-order elements lies in their ability to represent smooth
gradients.

TR U S S E L E M E N T S
In the Truss interface the default is to use first-order shape functions, since the
elements are mainly used in a context where the axial force in each element is constant.

When truss elements share an edge with other structural elements, you should choose
the same discretization in both interfaces, usually quadratic.

BEAM ELEMENTS
The beam elements have only one set of shape functions, which cannot be changed.
The axial displacement and the twist are represented by first-order shape functions,
while the bending is represented by cubic Hermitian shape functions. This element can
then represent a constant axial force, a constant twisting torque, a linear bending
moment, and a constant shear force. This is the exact solution for a beam having no
distributed loads.

SELECTING DISCRETIZATION | 61
 A consequence of this formulation is that it may not possible to obtain a perfectly
conforming approximation if a beam shares an edge with elements from another
physics interface.

Lagrange and Serendipity Shape Functions

In the Solid Mechanics and Membrane interfaces, you can choose between two
families of shape functions: Lagrange and serendipity. The default is to use serendipity
shape functions.

The serendipity elements have the advantage of generating significantly fewer degrees
of freedom for structured meshes. The accuracy is in most cases almost as good as for
the Lagrange elements. The Lagrange elements are however less sensitive to strong
mesh distortions.

The serendipity shape functions differs from the Lagrange shape functions only for the
following element shapes:

• 2D: Quadrilateral elements of discretization order higher than 1

• 3D: Hexahedral, prism, and pyramid elements of discretization order higher than 1

In the COMSOL Multiphysics Reference Manual:

• The Lagrange Element

• The Nodal Serendipity Element

When coupling to other structural mechanics physics interfaces, the same type of shape
functions should be used in both interfaces to ensure conformity in displacement shape
functions. Since there is no difference between the two families of shape functions in
1D, this is not an issue when connecting edges.

Choosing Shape Functions in Multiphysics Models

In problems where several physics fields participate, the accuracy can sometimes be
improved by considering how the different fields interact. In structural mechanics, it
is common that other physics fields directly affect the inelastic strains. This is the case
in, for example, thermal expansion and hygroscopic swelling.

62 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

In thermal expansion, the elastic strain used in most constitutive relations is the
difference between the total strain, which is computed from derivatives of the
displacement field, and the thermal strain:

ε el = ε tot – ε th = ε tot – α ( T – Tref )

Since the thermal strain is directly proportional to the temperature, a consistent

approximation would be to use one order lower discretization order for the
temperature than for the displacements. These kinds of considerations are
automatically made when you add the built-in multiphysics interfaces, such as The
Thermal Stress Interface.

Another type of coupling appears on the boundary between two domains having
different physics, as in fluid-structure interaction and acoustic-structure interaction.
When, for example, Thermoviscous Acoustics is coupled to Solid Mechanics, then the
time derivative of the displacement in the solid is set equal to the velocity in the
acoustic medium on the shared boundary. In this case, it makes sense to have the same
shape function order for these two fields.

Implicit Shape Function Orders

Some solid mechanics formulations contain other degrees of freedom in addition to
the displacements. The shape functions are then selected internally based on your
choice of displacement discretization.

MIXED FORMULATION
When the Nearly incompressible material check box is selected for a certain material, the
pressure (mean stress) is added as an extra degree of freedom to form what is called a
mixed formulation.

The shape function used for the pressure is continuous Lagrangian, having an order
that is one below what is used for the displacement shape functions. The pressure
degree of freedom is, however, discontinuous over domain boundaries in order to
allow for changes in material properties.

VISCOELASTIC STRESS AND STRAIN

When using the Viscoelastic material, auxiliary degrees of freedom are added either for
the viscoelastic strains or the viscoelastic stresses, depending on whether a linear or a
nonlinear formulation is used. These degrees of freedom are local to the element, and
you can select either the discontinuous Lagrange or Gauss point data type.

SELECTING DISCRETIZATION | 63
 The discontinuous Lagrange shape functions will have an order that is one below what
is used for the displacement shape functions.

If Gauss point data is used, the same integration points as used for the numerical
integration of the stiffness matrix are used. This order depends on the selected
displacement discretization order.

INELASTIC STRAINS
For material models like plasticity and creep, the inelastic strains are formally degrees
of freedom. They will be allocated at the same integration points as used for the
numerical integration of the stiffness matrix. This order depends on the selected
displacement discretization order.

64 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

Coupling Different Element Types
In this section:

• Introduction to the Element Types

• Coupling Techniques
• Solver Settings

Introduction to the Element Types

In some engineering structures, the optimal idealization is a mixture of different
element types. Some examples are:

• Structures that are thin in large regions but more three-dimensional at certain
locations. A mixture of solids and shells can then significantly reduce the model size.
• Plates or shells having beams as stiffeners.
• Truss elements acting as reinforcement bars in a concrete structure.
• A thin layer of one material on top of another material. In this case, an idealization
with shells or membranes covering the boundary of a solid can be useful.

When several physics interfaces are added in COMSOL, the default is always that each
physics interface has its individual degrees of freedom. In structural mechanics the first
physics interface has the displacement variables (u, v, w), then the second physics
interface has (u2, v2, w2), and so on. This means that the physics interfaces initially are
independent even when defined on the same geometrical part. To get the intended
interaction requires that a coupling is established between the physics interfaces.

Various methods to couple different element types are discussed in this section.

Coupling Techniques
The following basic techniques to connect physics interfaces with displacement
degrees of freedom is discussed in this section:

• Renaming Degrees of Freedom

• Using Customized Coupling Features
• Using General Coupling Operators

C O U P L I N G D I F F E R E N T E L E M E N T TY P E S | 65
 RENAMING DEGREES OF FREEDOM
The easiest coupling method is to rename the displacement degrees of freedom so that
these are the same for all physics interfaces. This is sufficient, for example, when using
membranes as cladding on a solid boundary or truss elements as reinforcement bars in
a solid.

In the Beam, Shell, and Plate interfaces, the deformation is described also by rotational
degrees of freedom. In the general case, these degrees of freedom interact with the
translational degrees of freedom in a connection.

In some special cases—for example, when a thin shell acts as cladding on a solid—it is
sufficient to make the degree of freedom names for the displacements common; the
rotational degrees of freedom are not important. If, however, a shell edge is connected
to a solid, it acts as a ‘hinge’, which in most cases is not the intended behavior. You
then need to use the more sophisticated techniques described next.

The default shape functions in the Solid Mechanics interface are of the
serendipity type, whereas in the Shell interface Lagrange shape functions
are used. If you are placing a shell element on the boundary of a solid
element, you must select Lagrange shape functions also in the Solid
Mechanics interface so that the two physics interfaces share the same node
points.

The shape functions used in the Beam interface have special properties,
and a beam cannot have the same degrees of freedom as another physics
interface if the same edge or boundary are shared.

Also, the representation of rotations differs between the Shell and Plate
interfaces (displacement of normal) and the Beam interface (rotation
angle). It is therefore not possible to use common degree of freedom
names for the rotational degrees of freedom.

USING CUSTOMIZED COUPLING FEATURES

There are a number of built-in couplings, by which you can add connections that are
difficult to set up manually:

• Shell Edge to Solid Boundary (3D)

• Shell Boundary to Solid Boundary (3D)
• Beam Point to Solid Boundary (2D)

66 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

• Beam Edge to Solid Boundary (2D)
• Beam Edge to Shell Edge (3D)
• Beam Point to Shell Boundary (3D)
• Beam Point to Shell Edge (3D)

Shell Edge to Solid Boundary (3D)

A shell can be coupled to a solid by adding a Solid-Shell Connection multiphysics
coupling. In the settings, set Connection type to Solid boundaries to shell edges. This
situation typically occurs when you want to make a transition from a thin region to one
which is thicker. Usually, shell assumptions should be valid on both sides of the
transition.

Shell Boundary to Solid Boundary (3D)

A shell can also be coupled to a solid by adding a Solid-Shell Connection multiphysics
coupling with Connection type set to Solid and shell shared boundaries or Solid and shell
parallel boundaries. This connection is used to add a shell on top of a solid as a
‘cladding’. It is possible to include an offset distance. The boundaries may be
coincident or parallel.

Beam Point to Solid Boundary (2D)

A beam can be coupled to a solid by adding a Solid-Beam Connection multiphysics
coupling. In the settings, set Connection type to Solid boundaries to beam points. This
coupling is intended for modeling a transition from a beam to a solid where beam
assumptions are valid on both sides of the connection.

Beam Edge to Solid Boundary (2D)

A beam can also be coupled to a solid by adding a Solid-Beam Connection
multiphysics coupling with Connection type set to Solid and beam parallel boundaries.
This connection is used for adding a beam on top of a solid as a ‘cladding’. The
boundaries are assumed to be parallel.

Beam Edge to Shell Edge (3D)

A beam can be coupled to a shell by adding a Shell-Beam Connection multiphysics
coupling with Connection type set to either Shell and beam shared boundaries or
Shell and beam parallel boundaries. This connection is used for adding beams as
stiffeners to shells. The edges may be coincident or parallel. It is possible to prescribe
that the beam has an offset from the shell when a coincident edge is used.

C O U P L I N G D I F F E R E N T E L E M E N T TY P E S | 67
 Beam Point to Shell Boundary (3D)
A beam can be coupled to a shell by adding a Shell-Beam Connection multiphysics
coupling with Connection type set to Shell boundaries to beam points. This connection
is used for modeling a beam with one end ‘welded’ to the face of the shell. You can
specify the size of the are around the beam end that is connected in several ways.

Beam Point to Shell Edge (3D)

A beam can be coupled to a shell by adding a Shell-Beam Connection multiphysics
coupling with Connection type set to Shell edges to beam points. This connection is
used for modeling a beam with one end ‘welded’ to the edge of the shell. You can
specify how large portion of the edge that is connected to the beam end in several ways.

The underlying theory and more details can be found in Connection Between Shells
and Solids and Connection Between Shells and Beams.

• Examples of all types of couplings between shells and beams are shown
in Connecting Shells and Beams: Application Library path
Structural_Mechanics_Module/Tutorials/shell_beam_connection
• An example of couplings between shells and solids is shown in
Connecting Shells and Solids: Application Library path
Structural_Mechanics_Module/Tutorials/shell_solid_connection

USING GENERAL COUPLING OPERATORS

The most general method of connecting parts modeled with different physics
interfaces is by using a General Extrusion operator. In this case the parts need not even
be in contact, so the connection is an abstraction.

An example could be a shell stiffened by beams. In practice, you would probably use
the built-in coupling described in Beam Edge to Shell Edge (3D) for this case, but the
examples displays the principles.

In structure like this, the beam is usually placed at one side of the shell, so that the
centerline of the beam and the midsurface of the shell do not coincide. This difference

68 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

must be taken into account, so the edges representing the beam are geometrically
disconnected from the midsurface of the shell.

Beam centerline

Mathematically, the connection between the beam and the shell can be expressed as


 u beam = u shell + φ × ( X beam – X shell )

 φ beam = φ shell


or equivalently as


 u beam = u shell + ( ( X beam – X shell ) ⋅ n )a

 φ beam = φ shell


Here φ is the rotation vector, which contains the rotational degrees of freedom in the
Beam interface. The rotation vector is also available as a variable in the Shell interface,
where it is derived from the rotational degrees of freedom a. The shell normal is
denoted by n.

To create the coupling:

1 Add a General Extrusion node under Definitions. Select the line on the shell
midsurface as source. Enter data in the Destination Map.

C O U P L I N G D I F F E R E N T E L E M E N T TY P E S | 69
 2 Add a Prescribed Displacement/Rotation node in the Beam interface and select the
corresponding edge.
3 Enter data for the prescribed displacements and rotations, for example
genext1(u)+genext1(shell.thy)*zdist, where zdist is some expression
defining the distance from the beam axis to the shell midsurface.

Because a shell does not have a valid rotation degree of freedom around
its normal, the rotation of the beam should not be connected in that
direction.

In the COMSOL Multiphysics Reference Manual:

• Component Couplings and Coupling Operators and General

Extrusion
• About Component Couplings

70 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

Apply i ng L o a ds
An important aspect of structural analysis is the formulation of the forces applied to
the modeled structure. The freedom is available to use custom expressions, predefined
or user-defined coordinate systems, and even variables from other modeling physics
interfaces.

Loads can be applied in the structural mechanics interfaces on the body, face, edge, or
point levels. Add The Solid Mechanics Interface ( ) to the Model Builder, then from
the Physics toolbar, Domains, Boundaries, Edges, and Points menus, click to select Body
Load, Face Load, Edge Load, or Point Load. This guide includes a detailed description of
the functionality for each physics interface.

In this section:

• Units, Orientation, and • Pressure

Visualization • Acceleration Loads
• Load Cases • Temperature Loads—Thermal
• Singular Loads Expansion
• Moments in the Solid Mechanics • Hygroscopic Swelling
Interface • Total Loads

Units, Orientation, and Visualization

USING UNITS
Enter loads in any unit, independently of the base SI unit system in the model, because
COMSOL automatically converts any unit to the base SI unit system. To use the
feature for automatic unit conversion, enter the unit in square brackets, for example,
100[lbf/in^2].

PREDEFINED AND CUSTOM COORDINATE SYSTEMS

In this module, different predefined coordinate systems are available when loads are
specified. There is always the global coordinate system. Depending on the
dimensionality of the part being worked with, there can also be predefined coordinate
systems such as and the local tangent and normal coordinate system for boundaries.

APPLYING LOADS | 71
 Custom coordinate systems are also available and are useful, for example, to specify a
load in any direction without splitting it into components. From the Definitions
toolbar, select a Coordinate System ( ) from the menu.

Some coordinate systems can have solution dependent axis directions. If

you use such a system for defining a load, the directions of the load follow
the moving coordinate axis directions if the Include geometric nonlinearity
check box is selected under the Study settings section of the current study
step.

VISUALIZATION
If you have switched on the physics symbols (see Displaying Physics Symbols in the
Graphics Window — An Example in the COMSOL Multiphysics Reference Manual),
then an applied load is indicated by a symbol together with a coordinate system
indicator displaying the definition directions for the load. The actual direction or
magnitude of the load you enter is not, however, reflected by the symbol. As a load in
COMSOL can be a function of parameters, variables, the solution, or results from
other physics interfaces, it is not possible to display it with only the information
available in the individual load feature.

You can always display the loads actually used after the analysis, since they are available
among the result quantities. Sometimes, especially if you have entered complicated
load expressions in a large model, it is important to inspect the load distribution before
you run the analysis. You can then do like this:

1 On the Study toolbar, click Get Initial Value. This operation is fast when compared
to actually solving the problem.
2 From the Results toolbar, add a 3D Plot Group or 2D Plot Group.
3 Add a suitable plot type to the plot group, for example an Arrow Surface plot.
4 Select a result from the Load group in the results for the physics interface.

In the COMSOL Multiphysics Reference Manual:

• Physics Symbols
• Using Units
• Coordinate Systems

72 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

 For an example showing how to examine the load distribution, see
Bracket—Static Analysis: Application Library path
Structural_Mechanics_Module/Tutorials/bracket_static. This is also the first
example used in the Introduction to the Structural Mechanics Module.

Load Cases
For a Stationary study, you can define load cases and constraint cases. A load or
constraint can be assigned to a load or constraint group, and then used conditionally.

For most load types, the load case acts as a simple multiplier, but some cases need a
more detailed discussion:

• A Prescribed Displacement or Prescribed Displacement/Rotation node can be assigned

both a constraint group and a load group. You can use the constraint group to
switch off the whole constraint. The load group acts as a multiplier to non-zero
prescribed values of displacement and rotations.
• When a load case multiplier is used for Thermal Expansion, the multiplier is applied
not to the actual temperature, but to the difference between the temperature and
the strain free reference temperature. The temperature difference, and thus the
thermal strain, is proportional to the load case multiplier.
• Since Thermal Expansion nodes are exclusive (only the last one given gives a
contribution for a certain domain), you cannot switch between different Thermal
Expansion nodes only by assigning them to different load cases.
• A Spring Foundation or Thin Elastic Layer node can be assigned a constraint group,
which you can use to switch it on and off. If there is also a Predeformation subnode,
then you can assign a load group to that subnode. The prescribed predeformation
is then multiplied by the load case multiplier. Predeformation nodes are exclusive,
you cannot switch between them by assigning them to different load cases.
• A Rigid Connector can be assigned both a constraint group and a load group. You can
use the constraint group to switch off the prescribed displacements and rotations.
The load group acts as a multiplier to non-zero prescribed values of displacement
and rotations.
• If you have added Applied Force or Applied Moment nodes under a Rigid Connector,
you can assign individual load groups to these nodes

APPLYING LOADS | 73
 • If you have added Applied Force or Applied Moment nodes under a Rigid Domain, you
can assign individual load groups to these nodes.

For an example about how to set up expressions for controlling position

and distribution of loads using load cases, see Pratt Truss Bridge:
Application Library path Structural_Mechanics_Module/Civil_Engineering/
pratt_truss_bridge.

Singular Loads
In reality, loads always act on a finite area. However, in a model a load is sometimes
defined on a point or an edge, which leads to a singularity. The reason for this is that
points and lines have no area, so the stress becomes infinite. Because of the stress
singularity, there are high stress values in the area surrounding the applied load. The
size of this area and the magnitude of the stresses depend on both the mesh and the
material properties. The stress distribution at locations far from these singularities is
unaffected according a to a well-known principle in solid mechanics, the St. Venant’s
principle. It states that for an elastic body, statically equivalent systems of forces
produce the same stresses in the body, except in the immediate region where the loads
are applied.

Figure 2-5 shows a plate with a hole in plane stress loaded with a distributed load and
a point load of the same magnitude. The mesh consists of triangular elements with
quadratic shape functions. The high stress around the point load is dissipated within
the length of a few elements for both mesh cases. The stresses in the middle of the plate
and around the hole are in agreement for the distributed load and the point load. The
problem is that due to the high stress around the singular load it is easy to overlook
the high stress region around the hole. When the point load is applied, the range must
be manually set for the stress plot to get the same visual feedback of the high stress
region around the hole in the two cases. This is because the default plot settings
automatically set the range based on the extreme values of the expression that is
plotted.

Despite these findings it is good modeling practice to avoid singular loads because it is
difficult to estimate the size of the singular region. In the Structural Mechanics

74 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

Module it is possible to define loads on all boundary types. However, avoid singular
loads altogether with elastoplastic or creep materials.

The Plasticity and Creep nodes are available as a subnode to Linear Elastic
Material nodes with the Nonlinear Structural Materials Module or the
Geomechanics Module.

normal mesh size

finer mesh size

Figure 2-5: A plate with a hole subject to a distributed load (left) and a point load (right).

Moments in the Solid Mechanics Interface

The Solid Mechanics interface, as opposed to the Beam, Plate, and Shell interfaces,
does not have rotational degrees of freedom. This makes the direct specification of
moment loads somewhat more complicated. To specify moments, attach a rigid
connector to the loaded area. The rigid connector has rotational degrees of freedom,
and it is possible to apply moments directly.

APPLYING LOADS | 75
 Pressure
A pressure is a load acting toward the normal of a face of the structure. If there are
large deformations in the model and the Include geometric nonlinearity check box is
selected under the Study settings section of the current study step, the pressure acts as
a follower load. The pressure is then defined with respect to the geometry and, as the
geometry deforms locally, the orientation of the load changes. The size of the loaded
area can also change as an effect of straining.

Acceleration Loads
Acceleration loads can be found, for example, in the structural mechanics analysis of
an airplane seat. Acceleration or deceleration of the aircraft produces a force that an
accurate simulation must include. Because expressions can be used when specifying
loads, it is easy to model acceleration loads.

For modeling rotating parts under static conditions, use centrifugal acceleration loads.
The body load in the radial direction is

K r = ρω 2 r (2-11)

where ρ is the density of the material, ω is the angular frequency, and r is the radial
distance from the axis of rotation. A cylindrical coordinate system is often useful here.

Temperature Loads—Thermal Expansion

When performing thermal expansion analysis, temperature loads are specified by
entering a temperature and a reference temperature in a thermal expansion subnode
which is available from the context menu (right-click the parent node, a Linear Elastic
Material node, for example) or from the Physics toolbar, Attributes menu. Enter a
constant temperature or an analytic expression that can depend on the coordinates or
dependent variables. For beams, plates, and shells it is also possible to specify bending
temperature loads. More details are available in the descriptions for each physics
interface.

When a separate physics interface is used to model heat transfer in the material, the
entry for the temperature is the dependent variable for the temperature from that

76 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

physics interface, typically T. In most cases, possible temperature variables from other
physics interfaces can be directly selected from a list.

• For more information about how to couple heat transfer analysis with
structural mechanics analysis, see Thermal-Structural Interaction. This
module also includes The Thermal Stress Interface.
• For a detailed discussion about thermal effects in structural mechanics
models, see Thermally Coupled Problems.

Hygroscopic Swelling
Some materials have the capability to absorb significant amounts of moisture through
diffusion processes. Changes in the moisture content may then cause volume changes.

To include the effects of hygroscopic swelling, the Hygroscopic Swelling subnode is

available from the context menu (right-click the parent node, Linear Elastic Material
node, for example) or from the Physics toolbar, Attributes menu. Enter a constant
concentration or an analytic expression that can depend on the coordinates or
dependent variables. For beams, plates, and shells it is also possible to specify bending
swelling loads caused by concentration gradient in the transverse direction. More
details are available in the descriptions for each physics interface.

When a separate physics interface is used to model the moisture diffusion in the
material, the entry for the concentration is the dependent variable for the
concentration from that physics interface, typically c. In most cases, possible
concentration variables from other physics interfaces can be directly selected from a
list.

The diffusion of the moisture into the material also adds to the mass density. You can
choose to automatically include this effect in a dynamic analysis, and also in mass
proportional loads, such as gravity and rotating frame loads.

Total Loads
You can specify a load either as a distributed load per unit length, area, or volume, or
as a total force to be distributed on a boundary. In the case of a total load, the applied
distributed load is the given load divided by the area (or length, or volume) on which
its acts.

APPLYING LOADS | 77
 Defining Constraints
Defining the proper constraints for structural mechanics models is just as important as
defining the loads as together they make up the model boundary conditions. This
module has many useful predefined physics features to define the constraints or to
create user-defined expressions that define constraints.

In this section:

• Rigid body motion

• Orientation
• Prescribed Displacements, Velocities, and Accelerations
• Symmetry Constraints
• Kinematic Constraints
• Rotational Joints

Rigid body motion

In most cases, a structure must have a set of constraints which is sufficient to suppress
any rigid body motions. A stationary problems is solvable only if the structure is
sufficiently constrained. There must not be any possible rigid body modes; thus no
stress-free deformation states are allowed. In a dynamic analysis, rigid body motions
are admissible. The inertial forces will then balance the external forces.

The number of possible rigid body modes for different geometrical dimensions is
shown in the table below.
TABLE 2-3: NUMBER OF POSSIBLE RIGID BODY MODES

DIMENSION NUMBER OF RIGID BODY MODES

3D 6 (3 translations + 3 rotations)
2D axisymmetric 1 (Z-direction translation)
2D (solid, beam, truss) 3 (2 translations + 1 rotation)
2D (plate) 3 (1 translation + 2 rotations)

If the model is underconstrained, you may encounter the following problems:

• The solver reports that the stiffness matrix is singular.

78 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

• The solver reports that the stiffness matrix is ill-conditioned. Theoretically, the
matrix is singular for a structure with rigid body modes, but because of the
round-off errors during the solution this is not exactly determined.
• A nonlinear analysis fails to converge.
• An iterative linear equation solver fails to converge.
• You get a solution with an extremely large displacement, orders of magnitude larger
than what is expected.

For a single body, it is seldom difficult to see whether it is fully constrained or not, but
for a more complex assembly, including several physics interfaces, or advanced
couplings and boundary conditions, it may not be trivial. If you suspect that rigid body
modes is a problem in your model, you can run an eigenfrequency analysis, and check
for modes with zero eigenfrequency as described in Eigenfrequency Analysis.

If there are no constraints which are dictated by the physical boundary conditions of
the structure, you can use the Rigid Motion Suppression boundary condition to
automatically remove the rigid body motions. When you do this, the assumption is that
the external loads are in equilibrium. If not, reaction forces and stress concentrations
will appear at seemingly arbitrary points where the automatic constraints were placed.

As an alternative to applying constraints, you can also add elastic supports through a
Spring Foundation node to suppress rigid body motion.

Rigid Motion Suppression Boundary Condition in the Structural

Mechanics Theory chapter.

Orientation
You can specify constraints in global as well as in any previously defined local
coordinate system.

Coordinate Systems in the COMSOL Multiphysics Reference Manual

DEFINING CONSTRAINTS | 79
 Prescribed Displacements, Velocities, and Accelerations
The most fundamental constraint is the prescribed displacement, where the individual
components of displacement or rotation can be prescribed to zero or non-zero values
for points, edges, boundaries, or domains.

For dynamic analysis, you can also directly prescribe the velocity or acceleration. The
conditions for prescribing displacements, velocities, or accelerations are mutually
exclusive for the same geometrical object since they prescribe the same degree of
freedom.

In frequency domain, a prescribed velocity vp or prescribed acceleration ap can be

directly interpreted as a prescribed displacement up:

vp
u p = ------
iω

–ap
u p = --------
2
-
ω

where ω is the angular frequency.

In the case of a time-dependent analysis, the prescribed displacement is obtained as


u p ( t ) = u 0 ( t 0 ) + v p ( τ ) dτ
t0

or

t τ
 
u p ( t ) = u 0 ( t 0 ) + v 0 ( t 0 ) + a p ( ζ ) dζ dτ
 

 
t0  t0 

where u0 and v0 are is given by the initial conditions. It is not possible to set explicit
initial conditions, but if initial values are taken from a previous study, they will be
respected. In order to compute the integrals, up is introduced as a separate degree of
freedom which is solved for by adding an extra ODE.

As prescribing the velocity or acceleration in time domain comes with an extra cost,
you should always consider using a prescribed displacement instead. As long as the

80 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

time history of the velocity or acceleration is a known a priori and does not depend on
the solution itself, this is possible.

• When the velocity or acceleration has a simple time dependence, you can integrate
it analytically one or two times to obtain the displacement, and directly prescribe the
displacement instead.
• When you have complicated known velocity or acceleration histories, for example
from measurements, you can use the integrate() operator. In this case, you enter
the prescribed displacement as integrate(my_data(tau),tau,0,t). Here
my_data is the measured data as function of time, and tau is a dummy integration
variable
In a stationary analysis, the prescribed velocity and acceleration nodes can have two
different behaviors. As a default, they are ignored, but you can also select that the
degrees of freedom having a prescribed velocity or acceleration in a dynamic analysis
should be constrained to zero in a static analysis.

When a local coordinate system is used for prescribing a prescribed velocity or

acceleration, the axis directions must be fixed in space. As an example, you cannot use
a Boundary System rotating with the deformation.

Symmetry Constraints
In many cases symmetry of the geometry and loads can be used to your advantage in
modeling. Symmetries can often greatly reduce the size of a model and hence reduce
the memory requirements and solution time. When a structure exhibits axial
symmetry, use the axisymmetric physics interfaces. A solid that is generated by rotating
a planar shape about an axis is said to have axial symmetry. In order to make use of
the axisymmetric physics interfaces, all loads and constraints must also be the same
around the circumference.

For other types of symmetry, use the predefined symmetry and antisymmetry
constraints. This means that no expressions need to be entered—instead just add the
type of constraint to apply to the model.

Physics Interface Axial Symmetry Node in the COMSOL Multiphysics

Reference Manual

DEFINING CONSTRAINTS | 81
 If the geometry exhibits two symmetry planes (Figure 2-6), model a quarter of the
geometry by using the Symmetry node for the two selected surfaces.

Symmetry planes Apply symmetry constraints

Figure 2-6: If the geometry exhibits two symmetry planes, model a quarter of the geometry
by using the Symmetry feature for the two selected surfaces.

Both geometric symmetry and loads are important when selecting the
correct constraints for a model.

In an eigenfrequency or buckling analysis, the eigenmodes might be

non-symmetric even if the structure is symmetric.

Figure 2-7 shows symmetric and antisymmetric loading of a symmetric geometry.

When modeling half of the geometry, the correct constraint for the face at the middle
of the object would be Antisymmetry in the case of antisymmetric loading and
Symmetry in the case of symmetric loading of the object.

Symmetry plane Antisymmetry plane

Figure 2-7: Symmetry plane (left) and antisymmetry plane (right).

82 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

Elemental and Nodal Constraints
For most constraints, you can select between using elemental and nodal constraints.
To do this, select Advanced Physics Options, so that the Constraint Settings section is
displayed.

When using nodal constraints, one constraint is generated for each node within the
selection a certain constraint feature. With elemental constraints, the number of
constraints added at a node equals the number of elements connected to that node.
This means that if some values used in the constraints differ between the elements,
then different constraints will be generated by the elemental method, whereas with the
nodal method an average is computed at the node before adding the constraint.

When several constraints are present at a node, the internal constraint elimination
algorithm is responsible for reducing them to a minimum unique set. Using elemental
constraints will clearly put an extra burden on this algorithm, so whenever possible you
should use nodal constraints.

The two different options exist, since under some circumstances the actual constraints
can differ between the two methods. Consider for example a symmetry constraint,
where the displacement in the direction normal to the boundary is constrained by the
equation

u⋅n = 0

where n is the unit normal vector.

If there are several intersecting symmetry planes, like in Figure 2-6, using nodal
constraints could cause a problem:

• If both boundaries are selected in the same Symmetry node, then only a single
constraint is applied for each node along the common edge, while you actually want
constraints along the normals of both planes. The normal used would be pointing
somewhere between the two planes, since a nodal constraint uses averaging of the
values from the adjacent elements.
• If two Symmetry nodes are used, so that the selection in any one of them only
contains boundaries without a normal direction discontinuity, the intended
constraints are added. On the common edge, there will be two contributions, one
from each node, and each using the normal direction of its boundary. If you want
to use nodal constraints, you must set up your model in this way if the constraints
are orientation dependent.

DEFINING CONSTRAINTS | 83
 Elemental constraints, on the other hand, can cause problems if the constraints added
by adjacent elements are not exactly the same. This could for example happen if the
normal orientation differs between neighboring elements. In such a case, a boundary
could behave as if it were fixed when a Symmetry, Antisymmetry, or Roller constraint is
applied. Such a situation could occur when the component consists of an imported
mesh, so that no underlying geometry exists.

The default type of the constraint, nodal or elemental, differs between different
constraint features. A nodal formulation is the default whenever it is considered safe,
like for a Fixed Constraint. Whenever the constraint can have a dependency on the
surface orientation, the default value is elemental.

See also Constraint Settings in the COMSOL Multiphysics Reference

Manual.

Kinematic Constraints
Kinematic constraints are equations that control the motion of solids, faces, edges, or
points. Add a Prescribed Displacement constraint to enter expressions for constraints.
You can define the equations using predefined coordinate systems as well as custom
coordinate systems. Special constraints, for instance to keep an edge of body straight
or to make a boundary rotate, require such constraint equations.

For an example showing how to force a boundary to remain plane, but

still allow it to translate in its normal direction, see Thermo-Mechanical
Analysis of a Surface-Mounted Resistor: Application Library path
Structural_Mechanics_Module/Thermal-Structure_Interaction/
surface_resistor.

In the 3D and 2D Solid Mechanics interfaces and in the Shell interface

there is a special constraint called a Rigid Connector. A rigid connector is
applied to one or more boundaries or edges and force them to behave as
connected to a common rigid body. The rigid connector can be given
prescribed displacements and rotations and thus simplifies the realization
of some constraints.

84 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

Rotational Joints
Joints between elements in The Truss Interface are automatically rotational joints
because the truss elements have no rotational degrees of freedom. For beams, however,
the rotational degrees of freedom are by default coupled between elements. To create
a rotational joint between two beam elements, add one additional Beam interface to a
geometry. Make sure that it is only active for the edge that includes the point where
the joint is positioned and that no other physics interface is active here. Couple the
translational degrees of freedom and leave the rotational degrees of freedom
uncoupled at the joint.

DEFINING CONSTRAINTS | 85
 Calculating Reaction Forces
There are different ways to evaluate reaction forces and these are discussed in this
section.

• Using Predefined Variables to Evaluate Reaction Forces

• Using Weak Constraints to Evaluate Reaction Forces
• Using Surface Traction to Evaluate Reaction Forces
• Using Surface Traction to Evaluate Reaction Forces

The following sections describe the merits and costs of these methods.

Using Predefined Variables to Evaluate Reaction Forces

The results analysis capabilities include easy access to the reaction forces and moments.
They are available as predefined variables. The reaction force variables are available
only at the nodes, and not as a continuous field, so they are not suitable for graphic
presentation.

To compute the sum of the reaction forces over a region, use

Volume Integration, Surface Integration, or Line Integration under
Results>Derived Values. The integration method discovers that the
reaction forces are discrete values and applies a summation instead of an
integration.

If you create an integration operator under

Component>Definitions>Component Couplings>Integration to sum reaction
forces, you must explicitly set Method to Summation over nodes.

Reaction forces are computed as the sum of the nodal values over the selected volume,
face, or edge. Reaction moments are calculated as the sum of the moment from the
reaction forces with respect to a reference point, and any explicit reaction moments (if
there are rotational degrees of freedom).

Specify the default coordinates of the Reference Point for Moment Computation at the
top level of the Settings window for the physics interface. After editing the reference
point coordinates, you need to right-click the Study node and select Update Solution
for the change to take effect on the reaction moment calculation. During

86 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

postprocessing, you can modify the coordinates of the reference point in the
Parameters section of a result feature..

• Reaction forces are not available for eigenfrequency analysis or when

weak constraints are used.
• Reaction force variables are computed where there are constraints, that
is Dirichlet conditions like Fixed Constraint, or Prescribed Displacement.
From the physical point of view, there can be other sources of reaction
forces, such as the elastic and viscous forces from a Spring Foundation.
Such forces are not included in the reaction force variables.
• If reaction forces are summed independently for two adjacent
boundaries, the total sum is not the same as if the reaction forces were
summed for both boundaries in one operation. The values of the nodes
at the common edge always contain contributions from the elements
at both sides of the edge.

Derived Values and Tables in the COMSOL Multiphysics Reference

Manual

Using Weak Constraints to Evaluate Reaction Forces

Select the Use weak constraints check box to get accurate distributed reactions. Extra
variables that correspond to the reaction traction distribution are automatically added
to the solution components.

With weak constraints activated, COMSOL Multiphysics adds the reaction forces to
the solution components. The variables are denoted X_lm, where X is the name of the
constrained degree of freedom (as, for example, u_lm and v_lm). The extension lm

CALCULATING REACTION FORCES | 87

 stands for Lagrange multipliers. It is only possible to evaluate reaction forces on
constrained boundaries in the directions of the constraints.

To compute the total reaction force on a boundary, integrate one of the

variables X_lm using Volume Integration, Surface Integration, or Line
Integration under Derived Values.

If the constraint is defined in a local coordinate system, the degrees of

freedom for the weak constraint variables are defined along the directions
of that system.

Since the reaction force variables are added to the solution components, the number
of DOFs for the model increases slightly, depending on the mesh size for the
boundaries in question. Boundaries that are adjacent to each other must have the same
constraint settings. The reason for this is that adjacent boundaries share a common
node.

Using weak constraints affects the structure of the equation system to be solved, and
is not suitable for all types of equation solvers.

In the COMSOL Multiphysics Reference Manual:

• Derived Values and Tables

• Symmetric and Nonsymmetric Constraints

Using Surface Traction to Evaluate Reaction Forces

As an alternative method, you can obtain values of the reaction forces on constrained
boundaries by using boundary integration of the relevant components of the surface
traction vector.

For 2D components, multiply the surface traction by the cross section

thickness before integrating to calculate the total reaction force.

Two different types of surface traction results can be computed in COMSOL

Multiphysics:

The first type, contained in the variables interface.Tax, is computed from the
stresses. It is always available. Since the surface traction vector is based on computed

88 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

stress results, this method is less accurate for computing reactions than the other
methods.

The second type, contained in the variables interface.Tracx, is computed using a

method similar to the weak constraints, but without introducing the Lagrange
multipliers as extra degrees of freedom. The accuracy is high, but there is an extra
computational cost. These traction variables are computed only if the check box
Compute boundary fluxes in the Discretization section is selected for the Solid Mechanics
interface.

In case of geometric nonlinearity, the two types of traction variables are

interpreted differently. The interface.Tax variables are based on
Cauchy stress, and contains a force per current area. If you integrate them
you must use the spatial frame. The interface.Tracx variables are based
on First Piola-Kirchhoff stresses and contains a force per undeformed
area. An integration must then be done on the material frame.

Evaluating Surface Traction Forces on Internal Boundaries

As opposed to the other methods for reaction force computation, the boundary flux
based tractions are computed not only on external boundaries, but also on internal
boundaries. On internal boundaries, there are then two traction fields: One acting
from each of the domains sharing the boundary. These internal traction fields are
contained in the variables interface.iTracux and interface.iTracdx. The letters
u and d in the variable names indicate the up and down side of the boundary
respectively. If you need the value of the total force acting on an internal section
through your model, these variable can be integrated. The interface.iTracux
andinterface.iTracdx variables are only available if the Compute boundary fluxes
check box is selected in the Discretization section of the physics interface, and there are
internal boundaries in your model.

Computing Accurate Fluxes in the COMSOL Multiphysics Reference

Manual

CALCULATING REACTION FORCES | 89

 Introduction to Material Models
In this section:

• Material Models for Structural Mechanics

• Entering material data
• Introduction to Linear Elastic Materials
• Introduction to Viscoelastic Materials
• Mixed Formulation
• About the Material Databases for the Structural Mechanics Module

Material Models for Structural Mechanics

The Structural Mechanics Module without any add-on modules provides the Linear
Elastic material with Viscoelasticity and Thermal Expansion modeling capabilities. It
also provides access to Piezoelectric materials modeling, which is described in detail in
the section Modeling Piezoelectric Problems.

If you have the optional products Nonlinear Structural Materials Module or

Geomechanics Module, many other classes of nonlinear materials are also available.
These models can be modified and extended, and custom material models can be
defined.

You can also add a material model which you have coded yourself and made available
as a binary library file using an External Stress-Strain Relation.

In the COMSOL Multiphysics Reference Manual:

• Working with External Materials

• External Material

90 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

In Table 2-4 you can find an overview of the families of materials, and their
applicability in the various structural mechanics interfaces.
TABLE 2-4: MATERIAL MODELS IN THE BASIC PHYSICS INTERFACES

MATERIAL SOLID SHELL/ MEMBRANE BEAM TRUSS

MODEL MECHANICS PLATE

Linear Elastic Material X X X X X

Nonlinear Elastic X - X - -
Material
Hyperelastic Material X - X - -
Piezoelectric Material X - - - -
Magnetostrictive X - - - -
Material
Viscoelasticity X - X - -
Plasticity X - X - X
Soil Plasticity X - - - -
Creep X - X - -
Concrete X - - - -
Rocks X - - - -
Cam-Clay Material X - - - -
External Stress-Strain X - - - -
Relation
Rigid Domain X - - - -
Spring-Damper - - - - X

Many of the material models can be augmented by effects like thermal expansion,
hygroscopic swelling, initial stresses and strains, and external stress.

Combination of Material Models

It is possible to combine many of the effects in an additive manner. The models based
on elasticity all have the same structure where

1 An elastic strain is computed by removing all inelastic strains (for example, plastic or
thermal strains) from the total strain.
2 An “elastic stress” is computed from the elastic strains.

INTRODUCTION TO MATERIAL MODELS | 91

 3 Any additional stresses (for example viscous stresses, or initial stresses) are added to
form the total stress.

This concept will give you a great freedom in combining effects. Some such useful
combinations are

• Plasticity and thermal expansion

• Plasticity and creep
• Two different creep models
• Thermal expansion and hygroscopic swelling
• Viscoelasticity and creep

Entering material data

For most material data, you have the option to choose between From material and
User defined. The preferred way of supplying the material data, is through the Materials
node. If you are using data from the Material Library, this is the only option, but also
when supplying your own data this will improve clarity of model.

Constitutive matrices, such as the elasticity matrix for an anisotropic material, are in
many cases per definition symmetric. Only the upper diagonal of the matrix given as
input is used for forming the matrix used, so you need not enter the lower diagonal
part.

Introduction to Linear Elastic Materials

Linear elasticity forms the basis for the majority of structural mechanics simulations.

For isotropic linear elasticity, two parameters are enough to describe the material
behavior. The number of parameters increases to (at most) 21 for the fully anisotropic
case in 3D. When setting up a model, make sure that the material parameters are
defined in agreement with the type of relationship used. If necessary, transform the
material data before entering it in the physics interface. For example, for orthotropic
materials calculate the Poisson’s ratio νxy by

92 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

 Ex
ν xy = ν yx ------
Ey

In the theory section Linear Elastic Material

Introduction to Viscoelastic Materials

The generalized Maxwell, standard linear solid (SLS) and Kelvin-Voigt viscoelastic
materials are available. All the models are linear, and the corresponding materials can
be described as consisting of one or more branches with a spring and a dashpot acting
in parallel to a linear elastic material. For each viscoelastic branch, the shear modulus
and the relaxation time (or viscosity) are entered.

• Linear Viscoelastic Materials

• Viscoelasticity

Mixed Formulation
Nearly incompressible materials can cause numerical problems if only displacements
are used in the interpolating functions. Small errors in the evaluation of volumetric
strain, due to the finite resolution of the discrete model, are exaggerated by the high
bulk modulus. This leads to an unstable representation of stresses, and in general to
underestimation of the deformation because spurious volumetric stresses might
balance also applied shear and bending loads.

When the Nearly incompressible material check box is selected in the Settings window
for the material, the negative volumetric stress pw is treated as an additional dependent
variable. The resulting mixed formulation is also known as u-p formulation. This
formulation removes the effect of volumetric strain from the original stress tensor and
replaces it with an interpolated pressure, pw. A separate equation constrains the
interpolated pressure to make it equal (in a finite-element sense) to the original
pressure calculated from the strains.

INTRODUCTION TO MATERIAL MODELS | 93

 Select this setting when the material data is close to incompressibility. For an isotropic
material, this happens when Poisson’s ratio approaches 0.5.

The mixed formulation is useful not only for linear elastic materials but
also for elastoplastic materials, hyperelastic materials, and viscoelastic
materials. The Hyperelastic Material and Plasticity nodes are available with
the Nonlinear Structural Materials Module. The Plasticity node is also
available with the Geomechanics Module.

The order of the shape function for the auxiliary pressure variable should
be one order less than that for the displacements. Thus, it is not
recommended to use linear elements for the displacement variables in the
domains, where the mixed formulation is turned on. Also note that some
iterative solvers do not work well together with mixed formulation
because the stiffness matrix becomes indefinite.

For an isotropic linear elastic material, the second Piola-Kirchhoff stress tensor S,
computed directly from the strains, is replaced by a modified version:

s˜ = s + ( p – p w )I

where I is the unit tensor and the pressure p is calculated from the stress tensor

1
p = – --- trace ( s )
3

The auxiliary dependent variable pw is set equal to p using the equation

pw p
------- – --- = 0 (2-12)
κ κ

where κ is the bulk modulus.

The modified stress tensor s˜ is used then in calculations of the energy variation.

For orthotropic and anisotropic materials, the auxiliary pressure equation is scaled to
make the stiffness matrix symmetric. Note, however, that the stiffness matrix in this
formulation is not positive definite and even contains a zero block on the diagonal in

94 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

the incompressible limit. This limits the possible choices of direct and iterative linear
solver.

In case of linear elastic materials without geometric nonlinearity (and also

for hyperelastic materials), the stress tensor s in the above equations is
replaced by the 2nd Piola-Kirchhoff stress tensor S, and the pressure p
with:

1
p p = – --- trace ( S )
3

About the Material Databases for the Structural Mechanics Module

The Structural Mechanics Module includes these material databases: Liquids and
Gases, with temperature-dependent fluid dynamic and thermal properties, MEMS, an
extended solid materials library with metals, semiconductors, insulators, and polymers
common in MEMS devices, and a Piezoelectric database with over 20 common
piezoelectric materials. The materials include temperature-dependent fluid dynamic
and thermal properties.

In the COMSOL Multiphysics Reference Manual:

• MEMS Materials Database

• Piezoelectric Materials Database
• Liquids and Gases Materials Database
• Materials

For an example of the MEMS materials database and Piezoelectric

materials database, see Piezoelectric Shear-Actuated Beam: Application
Library path Structural_Mechanics_Module/Piezoelectric_Effects/
shear_bender.

INTRODUCTION TO MATERIAL MODELS | 95

 Modeling Piezoelectric Problems
In this section:

• About Piezoelectric Materials

• Piezoelectric Coupling
• Create the Piezoelectric Effect Interface and Define Domains
• Complete Settings of Piezoelectric Materials
• Add Damping and Loss
• Define Material Properties

• Multiphysics Modeling Approaches in the COMSOL Multiphysics

Reference Manual.
• Piezoelectric Coupling
• Piezoelectric Material in the theory section

About Piezoelectric Materials

Piezoelectric materials become electrically polarized when strained. From a
microscopic perspective, the displacement of atoms within the unit cell (when the solid
is deformed) results into electric dipoles within the medium. In certain crystal
structures, this combines to give an average macroscopic dipole moment or electric
polarization. This effect, known as the direct piezoelectric effect, is always
accompanied by the converse piezoelectric effect, in which the solid becomes strained
when placed in an electric field.

Within a piezoelectric, there is a coupling between the strain and the electric field,
which is determined by the constitutive relation:

T
S = sE T + d E
(2-13)
D = dT + ε T E

Here, the naming convention used in piezoelectricity theory is assumed: S is the strain,
T is the stress, E is the electric field, and D is the electric displacement field. The
material parameters sE, d, and εT, correspond to the material compliance, the coupling
properties and the permittivity. These quantities are tensors of rank 4, 3, and 2

96 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

respectively. The tensors however are highly symmetric for physical reasons, and they
can be represented as matrices within an abbreviated subscript notation, which is
usually more convenient. In the Piezoelectric Devices interface, the Voigt notation is
used, which is standard in the literature for piezoelectricity but which differs from the
defaults in the Solid Mechanics interface.

The Piezoelectric Material uses the structural mechanics nomenclature.

The strain is named ε (instead of S) and the stresses are denoted by either
σ or S (instead of T). This makes the names consistent with those used in
the other structural mechanics interfaces.

Equation 2-13 will, using the notation from structural mechanics, then read

T
ε = sEσ + d E
(2-14)
D = dσ + ε 0 ε rT E

Equation 2-13 (or Equation 2-14) is known as the strain-charge form of the
constitutive relations. The equation can be re-arranged into the stress-charge form,
which relates the material stresses to the electric field:

T
σ = cE ε – e E
(2-15)
D = eε + ε 0 ε rS E

The material properties, cE, e, and εS are related to sE, d, and εT. It is possible to use
either form of the constitutive relations. In addition to Equation 2-13 or
Equation 2-15, the equations of solid mechanics and electrostatics must also be solved
within the material.

• Piezoelectric Coupling
• Modeling Piezoelectric Problems
• Piezoelectric Material in the theory section

Piezoelectric Material Orientation

The orientation of a piezoelectric crystal cut is frequently defined by the system
introduced by the IRE standard of 1949 (Ref. 8). This standard has undergone a
number of subsequent revisions, with the final revision being the IEEE standard of

MODELING PIEZOELECTRIC PROBLEMS | 97

 1987 (Ref. 9). Unfortunately the 1987 standard contained a number of serious errors
and the IEEE subsequently withdrew it. COMSOL therefore adopts the preceding
1978 standard (Ref. 10), which is similar to the 1987 standard, for material property
definitions. Most of the material properties in the material library are based on the
values given in the book by Auld (Ref. 11), which uses the 1978 IEEE conventions.
This is consistent with general practice except in the specific case of quartz, where it is
more common to use the 1949 IRE standard to define the material properties.
COMSOL therefore provides additional sets of material properties consistent with this
standard for the case of quartz. Note that the material properties for quartz are based
on Ref. 12, which uses the 1949 IRE standard (the properties are appropriately
modified according to the different standards).

The stiffness, compliance, coupling, and dielectric material property matrices are
defined with the crystal axes aligned with the local coordinate axes. Note that the signs
of several matrix components differ between the 1949 and the 1978 standards (see
Table 2-5). In the absence of a user-defined coordinate system, the local system
corresponds to the global X, Y, and Z coordinate axes. When an alternative coordinate
system is selected this system defines the orientation of the crystal axes. This is the
mechanism used in COMSOL to define a particular crystal cut, and typically it is
necessary to calculate the appropriate Euler angles for the cut (given the thickness
orientation for the wafer). All piezoelectric material properties are defined using the
Voigt form of the abbreviated subscript notation, which is universally employed in the
literature (this differs from the standard notation used for the Solid Mechanics
interface material properties). The material properties are defined in the material
frame, so that if the solid rotates during deformation the material properties rotate
with the solid. See Modeling Geometric Nonlinearity.

Crystal cuts are usually defined by a mechanism introduced by the IEEE/IRE

standards. Both standards use a notation that defines the orientation of a virtual slice
(the plate) through the crystal. The crystal axes are denoted X, Y, and Z and the plate,
which is usually rectangular, is defined as having sides l, w, and t (length, width, and
thickness). Initially the plate is aligned with respect to the crystal axes and then up to
three rotations are defined, using a right-handed convention about axes embedded
along the l, w, and t sides of the plate. Taking AT cut quartz as an example, the IEEE
1978 standard defines the cut as: (YXl) −35.25°. The first two letters in the bracketed
expression always refer to the initial orientation of the thickness and the length of the
plate. Subsequent bracketed letters then define up to three rotational axes, which move
with the plate as it is rotated. Angles of rotation about these axes are specified after the
bracketed expression in the order of the letters, using a right-handed convention. For

98 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

AT cut quartz only one rotation, about the l axis, is required. This is illustrated in
Figure 2-9. Note that within the 1949 IRE Standard AT cut quartz is denoted as:
(YXl) +35.25°, since the X-axis is rotated by 180° in this convention and positive
angles therefore correspond to the opposite direction of rotation (see Figure 2-8).
Table 2-6 summarizes the differences between the standards for different crystal cuts.

When defining the material properties of Quartz, the orientation of the

X, Y, and Z axes with respect to the crystal differs between the 1987 IEEE
standard and the 1949 IRE standard. Figure 2-8 shows the alignment of
the axis for the case of right-handed quartz. A consequence of this is that
both the material property matrices and the crystal cuts differ between the
two standards. Table 2-5 summarizes the signs for the important matrix
elements under the two conventions. Table 2-6 shows the different
definitions of the crystal cuts under the two conventions.

TABLE 2-5: SIGNS FOR THE MATERIAL PROPERTIES OF QUARTZ, WITHIN THE TWO STANDARDS COMMONLY
EMPLOYED

IRE 1949 STANDARD IEEE 1978 STANDARD

MATERIAL RIGHT HANDED LEFT HANDED RIGHT HANDED LEFT HANDED

PROPERTY QUARTZ QUARTZ QUARTZ QUARTZ

s14 + + - -
c14 - - + +
d11 - + + -
d14 - + - +
e11 - + + -
e14 + - + -

MODELING PIEZOELECTRIC PROBLEMS | 99

 TABLE 2-6: CRYSTAL CUT DEFINITIONS FOR QUARTZ CUTS WITHIN THE TWO STANDARDS COMMONLY
EMPLOYED AND THE CORRESPONDING EULER ANGLES FOR DIFFERENT ORIENTATIONS OF THE CRYSTAL
THICKNESS

STANDARD REPRESENTATION AT CUT BT CUT

IRE 1949 Standard (YXl) +35.25° (YXl) −49°

Y-thickness Euler (ZXZ: 0°,−35.25°,0°) (ZXZ: 0°,49°,0°)
angles
Z-thickness Euler (ZXZ: 0°,−125.25°,0°) (ZXZ: 0°, −41°,0°)
angles
IEEE 1978 Standard (YXl) −35.25° (YXl) +49°
Y-thickness Euler (ZXZ: 0°, 35.25°,0°) (ZXZ: 0°, −49°,0°)
angles
Z-thickness Euler (ZXZ: 0°, −54.75°,0°) (ZXZ: 0°, −139°,0°)
angles

When defining the material orientation, it is necessary to consider the orientation of

the plate with respect to the global coordinate system in addition to the orientation of
the plate with respect to the crystallographic axes. Consider the example of AT cut
quartz in Figure 2-9. The definition of the appropriate local coordinate system
depends on the desired final orientation of the plate in the global coordinate system.
One way to set up the plate is to orientate its normal parallel to the Y axis in the global
coordinate system. Figure 2-10 shows how to define the local coordinate system in this
case. Figure 2-11 shows how to define the local system such that the plate has its
normal parallel to the global Z axis. In both cases it is critical to keep track of the
orientation of the local system with respect to the global system, which is defined
depending on the desired orientation of the plate in the model.

There are also a number of methods to define the local coordinate system with respect
to the global system. Usually, it is most convenient to define the local coordinates with
a Rotated System node, which defines three Euler angles according to the ZXZ
convention (rotation about Z, then X, then Z again). Note that these Euler angles
define the local (crystal) axes with respect to the global axes—this is distinct from the
approach of defining the cut (global) axes with respect to the crystal (local) axes.

100 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

Figure 2-8: Crystallographic axes defined for right-handed quartz in COMSOL and the
1978 IEEE standard (color). The 1949 standard axes are shown for comparison (gray).

Figure 2-8 is reproduced with permission from: IEEE Std 176-1987 -

IEEE Standard on Piezoelectricity, reprinted with permission from
IEEE, 3 Park Avenue, New York, NY 10016-5997 USA, copyright 1987,
by IEEE. This figure must not be reprinted or further distributed without
prior written permission from the IEEE.

MODELING PIEZOELECTRIC PROBLEMS | 101

 Figure 2-9: Definition of the AT cut of quartz within the IEEE 1978 standard. The AT
cut is defined as: (YXl) −35.25°. The first two bracketed letters specify the initial
orientation of the plate, with the thickness direction, t, along the crystal Y axis and the
length direction, l, along the X axis. Then up to three rotations about axes that move with
the plate are specified by the corresponding bracketed letters and the subsequent angles. In
this case only one rotation is required about the l axis, of −35.25° (in a right-handed
sense).

102 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

Figure 2-10: Defining an AT cut crystal plate within COMSOL, with normal in the
global Y-direction. Within the 1978 IEEE standard the AT cut is defined as (YXl)
-35.25°. Start with the plate normal or thickness in the Ycr direction (a) and rotate the
plate −35.25° about the l axis (b). The global coordinate system rotates with the plate.
Finally rotate the entire system so that the global coordinate system is orientated as it
appears in COMSOL (c). The local coordinate system should be defined with the Euler
angles (ZXZ - 0, 35.25°, 0).(d) shows a coordinate system for this system in COMSOL.

MODELING PIEZOELECTRIC PROBLEMS | 103

 Figure 2-11: Defining an AT cut crystal plate within COMSOL, with normal in the
global Z-direction. Within the 1978 IEEE standard the AT cut is defined as (YXl)
−35.25°. Begin with the plate normal in the Zcr-direction, so the crystal and global systems
are coincident. Rotate the plate so that its thickness points in the Ycr-direction (the
starting point for the IEEE definition), the global system rotates with the plate (b). Rotate
the plate −35.25° about the l axis (d). Finally rotate the entire system so that the global
coordinate system is orientated as it appears in COMSOL (d). The local coordinate system
should be defined with the Euler angles (ZXZ: 0, -54.75°, 0). (e) shows a coordinate system
for this system in COMSOL.

104 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

Piezoelectric Losses
Losses in piezoelectric materials can be generated both mechanically and electrically.

In the frequency domain, these can be represented by introducing complex material

properties in the elasticity and permittivity matrices, respectively. Taking the
mechanical case as an example, this introduces a phase lag between the stress and the
strain, which corresponds to a Hysteretic Loss. These losses can be added to the
Piezoelectric Material by three subnodes: Mechanical Damping, Coupling Loss, and
Dielectric Loss. The losses typically defined as loss factors (see below).

The hysteretic electrical losses are usually used to represent high frequency electrical
losses that occur as a result of friction impeding the rotation of the microscopic dipoles
that produce the material permittivity.

Low frequency losses, corresponding to a finite material conductivity, can be added to

the model through an Electrical Conductivity (Time Harmonic) subnode. This feature
operates only in the frequency domain.

In the time domain, the losses can be added by using the Rayleigh Damping option in
the Mechanical Damping and Coupling Loss subnodes, and by using the Dielectric
Dispersion option in the Dielectric Loss subnodes. These types of damping are also
available in the frequency domain.

Rayleigh Damping

HYSTERETIC LOSS
In the frequency domain, the dissipative behavior of the material can be modeled using
complex-valued material properties, irrespective of the loss mechanism. Such hysteretic
losses can be applied to model both electrical and mechanical losses. For the case of
piezoelectric materials, this means that the constitutive equations are written as
follows.

For the stress-charge formulation

T
σ = c˜ E ε – e˜ E
D = e˜ ε + ε̃ E
S

and for the strain-charge formulation

MODELING PIEZOELECTRIC PROBLEMS | 105

 T
ε = s˜ E σ + d˜ E
D = d˜ σ + ε̃ E T

where c˜ E , d˜ , and ε are complex-valued matrices, where the imaginary part defines the
dissipative function of the material.

Both the real and complex parts of the material data must be defined so as to respect
the symmetry properties of the material being modeled and with restrictions imposed
by the laws of physics.

A key requirement is that the dissipation density is positive; that is, there
is no power gain from the passive material. This requirement sets rules for
the relative magnitudes for all material parameters. This is important
when defining the coupling losses.

In COMSOL, you can enter the complex-valued data directly or by means of loss
˜
factors. When loss factors are used, the complex data X is represented as pairs of a
real-valued parameter

˜
X = real ( X )

and a loss factor

˜ ˜
η X = imag ( X ) ⁄ real ( X )

the ratio of the imaginary and real part, and the complex data is then:

˜
X = X ( 1 ± jη X )

where the sign depends on the material property used. The loss factors are specific to
the material property, and thus they are named according to the property they refer to,
for example, ηcE. For a structural material without coupling, simply use ηs, the
structural loss factor.

The loss factors are defined so that a positive loss factor value corresponds to a positive
loss. The complex-valued data is then based on sign rules.

By default, there is no damping until at least one of the damping and losses related
subfeatures is added.

106 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

For the Piezoelectric Material node, the following equations apply via the
corresponding three subnodes:

Mechanical Damping
m, n m, n m, n
c˜ E = c E ( 1 + jη cE )
m, n m, n m, n
s˜ E = s E ( 1 – jη sE )

where m and n refer to components of each matrix.

Coupling Loss
m, n m, n m, n
e˜ =e ( 1 + jη e )
m, n m, n m, n
d˜ =d ( 1 + jη d )

Dielectric Loss
m, n m, n m, n
ε̃ rS = ε rS ( 1 – jη εS )
m, n m, n m, n
ε̃ rT = ε rT ( 1 – jη εT )

Note that the multiplication is applied component-wise.

In practice, it is often difficult to find complex-valued data for each of the matrix
elements in the literature. The loss factors can also be entered as scalar isotropic factors
independently of the material and the other coefficients.

For more information about hysteretic losses, see Ref. 1 to Ref. 4.

ELECTRICAL CONDUCTIVITY (TIME HARMONIC)

For frequency domain analyses, the electrical conductivity of the piezoelectric material
(see Ref. 2, Ref. 5, and Ref. 6) can be defined. Depending on the formulation of the
electrical equation, the electrical conductivity appears in the equation as an effective
electric displacement

σe E
∇ ⋅  D – j ----------- = 0
 ω 

where σe is the material electrical conductivity, and E is the electric field. Note that the
displacement current variables themselves do not contain any conductivity effects.

MODELING PIEZOELECTRIC PROBLEMS | 107

 Both a dielectric loss factor and the electrical conductivity can be defined at the same
time. In such case, ensure that the loss factor refers to the alternating current loss
tangent, which dominates at high frequencies, where the effect of ohmic conductivity
vanishes (Ref. 7).

The use of electrical conductivity in a damped eigenfrequency analysis leads to a

nonlinear eigenvalue problem, which must be solved iteratively. To compute the
correct eigenfrequency, run the eigenvalue solver once for a single mode. Then, set the
computed solution to be the linearization point for the eigenvalue solver, which is
defined in the Settings window for the Eigenvalue Solver node. Re-run the eigenvalue
solver repeatedly until the solution no longer changes. This process must be repeated
for each mode separately.

DIELECTRIC DISPERSION
When the Dielectric Loss subnode is used with the Dielectric Dispersion option, the
following equations are solved in the time domain:

∂D
∇ ⋅  D + τd + ε 0 Δε rS E = 0
 ∂t 
D = eε + ε 0 ε rS E

where two material parameters can be specified, the relaxation time τd, and the relative
permittivity increment ΔεrS. The latter can be either a matrix or a scalar quantity. This
model is a one-term version of the more general Debye dispersion model, Ref. 13. In
the frequency domain, the time derivative is replaced by the factor jω, and the above
equation can be rewritten as

 Δε rS  τ d ωε 0 Δε rS
∇ ⋅ eε + ε 0  ε rS + --------------------------2- E – j ---------------------------2- E = 0
 1 + ( τd ω )  1 + ( τd ω )

which shows how the dispersion parameters contribute into the polarization and losses.
Thus, the effective permittivity varies from εrS + ΔεrS down to ε0εrS as the excitation
frequency increases from zero. The damping effect vanishes for both large and small
frequencies, and it reaches the maximum for ω = 1/τd.

In the COMSOL Multiphysics Reference Manual:

• Selecting a Stationary, Time-Dependent, or Eigenvalue Solver

• Eigenvalue Solver

108 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

References for Piezoelectric Damping
1. R. Holland and E.P. EerNisse, Design of Resonant Piezoelectric Devices, Research
Monograph No. 56, The M.I.T. Press, 1969.

2. T. Ikeda, Fundamentals of Piezoelectricity, Oxford University Press, 1990.

3. A.V. Mezheritsky, “Elastic, Dielectric, and Piezoelectric Losses in Piezoceramics:

How it Works all Together,” IEEE Transactions on Ultrasonics, Ferroelectrics, and
Frequency Control, vol. 51, no. 6, 2004.

4. K. Uchino and S. Hirose, “Loss Mechanisms in Piezoelectrics: How to Measure

Different Losses Separately,” IEEE Transactions on Ultrasonics, Ferroelectrics, and
Frequency Control, vol. 48, no. 1, pp. 307–321, 2001.

5. P.C.Y. Lee, N.H. Liu, and A. Ballato, “Thickness Vibrations of a Piezoelectric Plate
With Dissipation,” IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency
Control, vol. 51, no. 1, 2004.

6. P.C.Y. Lee and N.H. Liu, “Plane Harmonic Waves in an Infinite Piezoelectric Plate
With Dissipation,” Frequency Control Symposium and PDA Exhibition, pp. 162–
169, IEEE International, 2002.

7. C.A. Balanis, “Electrical Properties of Matter,” Advanced Engineering

Electromagnetics, John Wiley & Sons, chapter 2, 1989.

8. “Standards on Piezoelectric Crystals, 1949”, Proceedings of the I. R. E.,vol. 37,

no.12, pp. 1378–1395, 1949.

9. IEEE Standard on Piezoelectricity, ANSI/IEEE Standard 176-1987, 1987.

10. IEEE Standard on Piezoelectricity, ANSI/IEEE Standard 176-1978, 1978.

11. B.A. Auld, Acoustic Fields and Waves in Solids, Krieger Publishing, 1990.

12. R. Bechmann “Elastic and Piezoelectric Constants of Alpha-Quartz,” Physical

Review B, vol. 110 no. 5, pp. 1060-1061, 1958.

13. N.S. Stoykov, T.A. Kuiken, M.M. Lowery, and A. Taflove, “Finite-element
time-domain algorithms for modeling linear Debye and Lorentz dielectric dispersions
at low frequencies,” IEEE Transactions on Biomedical Engineering, vol. 50, no. 9,
pp. 1100-1107, 2003.

MODELING PIEZOELECTRIC PROBLEMS | 109

 Piezoelectric Coupling
The piezoelectric effect is an interaction between the mechanical and electrical physics,
where a stress applied on a piezoelectric material generates a voltage (direct effect) or
a voltage applied on it generates the deformation of the material (inverse effect). In
COMSOL Multiphysics, the Piezoelectric Devices interface is constituted of one Solid
Mechanics and one Electrostatics interface, which are coupled together by a Piezoelectric
Effect multiphysics feature. Hence a piezoelectric problem contains solid and
electrostatic domains, with at least one domain shared by the two physics interfaces
and with the piezoelectric coupling defined on it.

Create the Piezoelectric Effect Interface and Define Domains

A piezoelectric problem can be set up in different ways:

• By selecting Piezoelectric Devices from the Model Wizard,

• By choosing Piezoelectric Devices from the Add Physics menu when working in an
existing model, or
• By adding the corresponding features to create the coupling manually.

In the first two cases, by default all the domains in the model are assumed to be
piezoelectric materials.

When setting up the problem manually (that is, by adding single physics interfaces one
at a time) both Solid Mechanics and Electrostatics interfaces should be added. Then, you
have to specify which domains are in each physics, and which domains are to be
modeled as piezoelectric materials.

1 On the Solid Mechanics interface Settings window, locate the Domain Selection
section. Select the domains which undergo structural deformation, including the
piezoelectric material domains.
2 Go to the Solid Mechanics>Piezoelectric Material node (if it is not yet available, add
it). Select the domains where the piezoelectric effect applies. Non-piezoelectric
domains can be modeled using the other available material models.

110 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

3 Go to the Electrostatics node and under Domain Selection select the domains the
electrostatics equations must be solved. These domains include all the piezoelectric
domains, and any other insulating domains.

Since metals usually have a conductivity several orders of magnitude

higher than other materials, it is normal not to solve the electrostatics
equations in their corresponding domains, but instead the surfaces of the
material are represented as isopotentials with an appropriate boundary
condition (usually a terminal or floating potential).

4 Go to the Electrostatics>Charge Conservation, Piezoelectric node (if it is not yet

available, add it). Select the domains where the piezoelectric effect has to be
modeled.
5 A Multiphysics>Piezoelectric Effect node is already present if the coupling was added
using either the Model Wizard or Add Physics window. If the model is set up manually
(that is, single physics interfaces are added), right-click the Multiphysics node to add
a Piezoelectric Effect coupling.
6 Confirm that all the domains where the piezoelectric material is present are selected.

Only domains that have both Charge Conservation, Piezoelectric selected in

the Electrostatics interface and Piezoelectric Material selected in the Solid
Mechanics interface are selected. The selection of this feature cannot be
edited. If several Solid Mechanics or Electrostatics interfaces are present,
select the correct ones.

7 Confirm that remaining domains are well assigned:

- Solid and electrostatic domains: In these domains, the electrostatics and
structural problems are solved independently without any piezoelectric coupling.
This is the case for insulators.
- Solid-only domains: In these domains, only mechanical phenomena are modeled,
and the electrostatics phenomena are neglected. Metals are typically modeled in
this way, because their conductivity is so high that their surfaces can be treated as
isopotential surfaces within the electrostatics problem. In some cases, insulators
are modeled using these settings, for example, when there is no potential applied

MODELING PIEZOELECTRIC PROBLEMS | 111

 across the domain, and correspondingly solving the electrostatics equations
would produce a constant potential and waste computing resources.
- Electrostatics-only domains: Those domains are non-solid, this is typically the
case of air, in which the electrostatics is solved but not mechanics.
- Non-solid and non-electrostatics domains: In those domains another physics is
solved. A typical example is an acoustic domain in the case of piezoacoustic
modeling.

Complete Settings of Piezoelectric Materials

Go to the Solid Mechanics>Piezoelectric Material node. On the Settings window
complete these settings:

• Coordinate System Selection section: The material is poled in the x3 direction of the
coordinate system (x1,x2,x3) specified in this section. By default, it is set to the global
coordinate system. If the piezoelectric material is poled along an other direction,
you need to define a coordinate system so that its third direction is aligned with the
poling direction. Then, assign it as the coordinate system which orients the material
in the Coordinate System Selection section.

If a given piezoelectric material is present with several orientations (such

as stacked piezoelectric disks) you need to define several Piezoelectric
Material nodes and to assign a different coordinate system for each of
them.

• Piezoelectric Material Properties: Select whether the constitutive relation of

piezoelectric material is in Stress-charge or in Strain-charge form. This choice
defines the type of material properties that will be used.
• Geometric Nonlinearity: select the check box to force strains to be linear.
• Energy Dissipation: Select the check box to enable the calculation of the dissipated
energy.

112 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

Add Damping and Loss
In the physics toolbar you can add attributes to the Piezoelectric Material node,
especially the following damping and loss contributions:

• Mechanical Damping: Specify the domains of application, then choose if you want to
define a loss factor for cE, a loss factor for sE (in Strain-charge form), an isotropic
loss factor, or a Rayleigh damping.
• Dielectric Loss: Specify the domains of application, then choose if you want to define
a loss factor for εS, a loss factor for εT (in Strain-charge form), or dispersion.
• Coupling Loss: Specify the domains of application, then choose if you want to define
a loss factor for e, a loss factor for d (in Strain-charge form), or Rayleigh damping.
• Conduction Loss (Time-Harmonic): Specify the domains of application, then choose
how you want to define the Electrical conductivity.

Define Material Properties

Define material properties for the piezoelectric materials. The material library contains
several common piezoelectric materials under the piezoelectric section. If you want to
define your own piezoelectric material, you need to specify its properties by hand. The
required properties depend on whether the constitutive relations are in Stress-charge
or Strain-charge form, and which damping and loss attributes are created. Defining all
piezoelectric settings before materials preselects the required properties and makes the
completion easier:

MECHANICAL PROPERTIES:
3
• Density rho (SI unit: kg/m )
• Elasticity matrix cE (SI unit: Pa) in Stress-charge form.
• Compliance Matrix sE (SI unit: 1/Pa) in Strain-charge form.

ELECTROSTATIC PROPERTIES:
• Relative Permittivity epsilonrS (dimensionless) in Stress-charge form.
• Relative Permittivity epsilonrT (dimensionless) in Strain-charge form.

COUPLING PROPERTIES
• Coupling matrix eES (SI unit: C/m2) in Stress-Charge form.
• Coupling matrix dET (SI unit: C/N) in Strain-charge form.

MODELING PIEZOELECTRIC PROBLEMS | 113

 DAMPING AND LOSS PROPERTIES
• Loss factor for elasticity matrix cE, eta_cE, or loss factor for elasticity matrix sE,
eta_sE: required when Mechanical Damping is present. The latter is valid only in
Strain-Charge form.
• Loss factor for electrical permittivity matrix εrS, eta_epsilonS, or Loss factor for
electrical permittivity matrix εrT, eta_epsilonT: required when Dielectric Loss is
present. The latter is valid only in Strain-Charge form.
• Loss factor for coupling matrix e, eta_eES, or loss factor for coupling matrix d,
eta_dET: required when Coupling Loss is present. The latter is valid only in
Strain-Charge form.
• Electrical conductivity sigma (SI unit: S/m): required when Conduction Loss is
present.

Working with Materials and Piezoelectric Materials Database in the

COMSOL Multiphysics Reference Manual.

Coupling Piezoelectric Devices with Acoustics

Using piezoelectric materials for an acoustic application is common, such as in sonars,
microphones, sensors, and so forth. This is why coupling piezoelectric devices with
acoustic domains is of particular interest for these applications.

Compared to a single piezoelectric model, you need to add a pressure acoustics

interface, for example, Pressure Acoustics, Frequency Domain or Pressure Acoustics,
Transient (depending on which study type you want to use) plus an Acoustic-Structure
Boundary coupling under the Multiphysics node. You can also directly create the nodes
that are needed for coupling by adding an Acoustic-Piezoelectric Interaction interface
from the Model Wizard or Add Physics windows. If solid and acoustic domains are
correctly defined, then the right coupling boundaries are automatically selected. Then
specify domains of application for each physics.

• Select solid domains and Piezoelectric Material domains in Solid Mechanics.

• Select electrostatic domains and Charge Conservation, Piezoelectric domains in
Electrostatics.
• Select acoustic domains in a Pressure Acoustics node.
• Under the Multiphysics branch, confirm that selections for Piezoelectric Effect and
Acoustic-Structure Boundary are the right ones. If several Pressure Acoustics, Solid

114 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

 Mechanics, or Electrostatics interfaces are present, select the right ones that need to
be coupled in the multiphysics interfaces.
• Continue the modeling process by entering the settings for each physics interface
and feature and define materials.

MODELING PIEZOELECTRIC PROBLEMS | 115

 Modeling Magnetostrictive Materials
Magnetostriction Coupling
Magnetostriction is an interaction between the mechanical and magnetic physics,
where a magnetic field applied on magnetostrictive material generates the deformation
of the material (direct effect), and a stress applied on it changes the material
magnetization (inverse effect). In COMSOL Multiphysics, the Magnetostriction
interface is constituted of one Solid Mechanics and one Magnetic Fields interface, which
are coupled together via a Magnetostriction multiphysics feature. Hence a
magnetostriction problem contains solid and magnetic domains, with at least one
domain shared by the two physics interfaces and with the magnetostriction coupling
defined on it.

Create the Magnetostriction Interface and Define Domains

A magnetostriction problem can be set up in different ways:

• By selecting Magnetostriction from the Model Wizard,

• By choosing Magnetostriction from the Add Physics menu when working in an
existing model, or
• By adding the corresponding features to create the coupling manually.

In the first two cases, by default all the domains in the model are assumed to be
magnetostrictive materials.

When setting up the problem manually (that is, by adding single physics interfaces one
at a time) both Solid Mechanics and Magnetic Fields interfaces should be added. Then,
you have to specify which domains are in each physics, and which domains are to be
modeled as magnetostrictive materials.

1 On the Solid Mechanics interface Settings window, locate the Domain Selection
section. Select the domains which undergo structural deformation, including the
magnetostrictive material domains.
2 Go to the Solid Mechanics>Magnetostrictive Material node (if it is not yet available,
add it). Select the domains where the magnetostriction effect applies.
Nonmagnetostrictive domains can be modeled using the other available material
models.

116 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

3 Go to the Magnetic Fields node and under Domain Selection select the domains the
magnetics equations must be solved. These domains include all the magnetostrictive
domains, and any other magnetic domains.
4 Go to the Magnetic Fields>Ampere’s Law, Magnetostrictive node (if it is not yet
available, add it). Select the domains where the magnetostriction effect has to be
modeled.
5 A Multiphysics>Magnetostriction node is already present if the coupling was added
using either the Model Wizard or Add Physics window. If the model is set up manually
(that is, single physics interfaces are added), right-click the Multiphysics node to add
a Magnetostriction coupling.
6 Confirm that all the domains where the magnetostrictive material is present are
selected.

Only domains that have both Ampere’s Law, Magnetostrictive selected in

the Magnetic Fields interface and Magnetostrictive Material selected in the
Solid Mechanics interface are active in the selection for Magnetostriction
coupling. The selection of this multiphysics feature cannot be edited. If
several Solid Mechanics or Magnetic Fields interfaces are present, select the
correct ones.

7 Confirm that remaining domains are well assigned:

- Solid and magnetic domains: In these domains, the magnetics and structural
problems are solved independently without any magnetostriction coupling. This
represents deformable magnetic materials for which the magnetostriction
coupling is negligible.
- Solid-only domains: In these domains, only mechanical phenomena are modeled,
and the magnetics phenomena are neglected. This represents deformable solid
nonmagnetic materials.
- Magnetics-only domains: Those domains are non-solid, this is typically the case
of air, in which the magnetics is solved but not mechanics.
- Non-solid and non-magnetics domains: In those domains another physics is
solved. Typical examples are acoustic domains and fluid flow domains, which
might be present in various multiphysics models.

MODELING MAGNETOSTRICTIVE MATERIALS | 117

 Linear vs Nonlinear Magnetostriction
The magnetostrictive strain has a nonlinear dependence on the magnetic field and the
mechanical stress in the material. However, the effect can be modeled using linear
coupled constitutive equations if the response of the material consists of small
deviations around an operating point (bias point).

For linear magnetostriction model, it is possible to express the relation between the
stress S, strain ε, magnetic field H, and magnetic flux density B in either a
stress-magnetization form or strain-magnetization form. In COMSOL
Multiphysics, both constitutive forms can be used; simply select one, and the software
will make all necessary transformations if needed. For details, see the corresponding
the theory section Linear Magnetostriction Model. You find all the necessary material
data inputs within the Magnetostrictive Material selected in the Solid Mechanics interface.

The nonlinear model of magnetostrictive strain can be used for the whole range from
full demagnetization to saturation magnetization. For more details, see the
corresponding the theory section Nonlinear Magnetostriction.

In case of nonlinear magnetostriction, the magnetization model can be selected. The

following options are available: Langevin function, Hyperbolic tangent, User defined,
and Linear. The last option will make it possible to find an explicit expression for the
magnetization. Note however that such model does not have a proper saturation
behavior, and thus it should be used only in the operating range far from saturation.
For ll other choices, the magnetization vector components will be treated as extra
dependent variables.

118 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

Mechanical Damping and Losses
Damping and losses are important for determining the response in Time Dependent
and Frequency Domain studies, and sometimes also when computing
eigenfrequencies. This section describes how to model damping and loss using
different damping models. In this section:

In this section:

• About Damping
• Rayleigh Damping
• Loss Factor Damping
• Viscoelastic Damping
• Explicit Damping
• Equivalent Viscous Damping
• Piezoelectric Damping
• Adding Damping in the Modal Solver

Damping Sources
There are many sources of damping in a system. Some some them are:

• Dissipation in the material. This dissipation can be for example be proportional to

the amplitude of the strain rate (viscous damping) or to the amplitude of the strain
itself (hysteretic damping).
• Thermoelastic damping, which is a thermodynamic effect which is related not only
to the state in a point, but also to the gradients of the temperature field.
• Damping caused by the surrounding medium, often air or water.
• Friction between joined parts.
• Components intended to supply damping like a dashpot in a car suspension.

It is often difficult to separate and quantify these effects, so damping modeling is one
of the biggest challenges in structural dynamics.

MECHANICAL DAMPING AND LOSSES | 119

 About Damping
Phenomenological damping models are typically invoked to model the intrinsic
frictional damping present in most materials (material damping). These models are
easiest to understand in the context of a system with a single degree of freedom. The
following equation of motion describes the dynamics of such a system with viscous
damping:

2
d u du
m ---------2- + c ------- + ku = f ( t ) (2-16)
dt dt

In this equation u is the displacement of the degree of freedom, m is its mass, c is the
damping parameter, and k is the stiffness of the system. The time (t) dependent forcing
term is f(t). This equation is often written in the form:

2
d u- du 2 f( t)
---------
2
+ 2ζω 0 ------- + ω 0 u = --------- (2-17)
dt dt m

where ξ = c/(2mω0) and ω02 = k/m. In this case ζ is the damping ratio (ζ = 1 for
critical damping) and ω0 is the undamped resonant frequency of the system. In the
literature it is more common to give values of ζ than c. ζ can also be readily related to
many of the various measures of damping employed in different disciplines. These are
summarized in Table 2-7.
TABLE 2-7: RELATIONSHIPS BETWEEN MEASURES OF DAMPING

DAMPING DEFINITION RELATION TO

PARAMETER DAMPING RATIO

Damping ratio ζ = c ⁄ c critical –

Logarithmic u ( t0 ) δ d ≈ 2πζ
δ d = ln  ----------------------
decrement  u ( t 0 + τ )
(ζ « 1)
where t0 is a reference time and τ is the period
of vibration for a decaying, unforced degree of
freedom.

120 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

TABLE 2-7: RELATIONSHIPS BETWEEN MEASURES OF DAMPING

DAMPING DEFINITION RELATION TO

PARAMETER DAMPING RATIO

Quality factor Q = ω ⁄ Δω Q ≈ 1 ⁄ ( 2ζ )

where Δω is the bandwidth of the amplitude (ζ « 1)

resonance measured at 1 ⁄ 2 of its peak.
Loss factor 1 Qh At the resonant
η = ------  -------- frequency:
2π W h
η ≈ 2ζ
where Qh is the energy lost per cycle and Wh is
the maximum potential energy stored in the (ζ « 1)
cycle. The variables Qh and Wh are available as
solid.Qh and solid.Wh.

In the frequency domain, the time dependence of the force and the displacement can
be represented by introducing a complex force term and assuming a similar time
dependence for the displacement. The equations

jωt jωt
f ( t ) = Re { Fe } and u ( t ) = Re { Ue }

are written where ω is the angular frequency and the amplitude terms U and F can in
general be complex (the arguments provide information on the relative phase of
signals). Usually the real part is taken as implicit and is subsequently dropped.
Equation 2-16 takes the following form in the frequency domain:

2
– ω mU + jωcU + kU = F (2-18)

where the time dependence has canceled out on both sides. Alternatively this equation
can be written as:

2 2 F
– ω U + 2jωω 0 ξU + ω 0 U = ----- (2-19)
m

There are three basic damping models available in the structural mechanics interfaces
for explicit modeling of material damping—Rayleigh damping, viscous damping, and
loss factor models based on introducing complex quantities into the equation system.
There are also other phenomena which contribute to the damping. Some material
models, such as viscoelasticity and plasticity are inherently dissipative. It is also possible
to model damping in spring conditions.

MECHANICAL DAMPING AND LOSSES | 121

 Rayleigh Damping
A common method of modeling damping is Rayleigh damping, where two damping
coefficients are specified. This type of damping is not directly related to any physical
process, but must be seen as a way to take the total damping of a structure into
account.

Rayleigh damping introduces damping in a form based on Equation 2-16. This means
that the method can be applied generally in either the time or frequency domain. The
parameter c in Equation 2-16 is defined as a fraction of the mass and the stiffness using
two parameters, αdM and βdK, such that

c = α dM m + β dK k (2-20)

Substituting this relationship into Equation 2-16 and rearranging into the form of
Equation 2-17 gives:

2
d u- 2 du 2 f(t)
--------- + ( α dM + β dK ω 0 ) ------- + ω 0 u = ---------
dt
2 dt m

When there are many degrees of freedom m, k, and c become matrices and the
technique can be generalized.

Rayleigh damping can therefore be identified as equivalent to the damping ratio at

resonance of:

1 α dM
ζ = ---  ----------- + β dK ω 0 (2-21)
2 ω0

Note that Equation 2-21 holds separately for each vibrational mode in the system at
its resonant frequency. In the frequency domain it is possible to use frequency
dependent values of αdM and βdK. For example setting αdM = 0 and βdK = 2ζ/ω0
produces a equivalent viscous damping model at the resonant frequency.

While Rayleigh damping is numerically convenient, the model does not agree with
experimental results for the frequency dependence of material damping over an
extended range of frequencies. This is because the material damping forces behave
more like frictional forces (which are frequency independent) than viscous damping
forces (which increase linearly with frequency as implied by Equation 2-18). In the
frequency domain it is possible to introduce loss factor damping, which has the desired
property of frequency independence.

122 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

A complication with the Rayleigh damping model is to obtain good values for the
damping parameters αdM and βdK. A more physical damping measure is the relative
damping, the ratio between actual and critical damping, often expressed as a
percentage of the critical damping. Commonly used values of relative damping can be
found in the literature.

Using Equation 2-21, this relationship at two frequencies, f1 and f2, with different
relative damping, ζ1 and ζ2, results in an equation system that can be solved for αdM
and βdK:

1 -
----------- πf
4πf 1 1 α dM ζ1
=
1 - β dK ζ2
----------- πf
4πf 2 2

The damping ratios are then

ζ1 f2 – ζ2 f1
α dM = 4πf 1 f 2 ---------------------------
2 2
f2 – f1
ζ2 f2 – ζ1 f1
β dK = ---------------------------
2 2
π ( f2 – f1 )

Using the same relative damping, ζ1 = ζ2, does not result in a constant damping factor
inside the interval f1 < f < f2. It can be shown that the damping factor is lower inside
the interval, as Figure 2-12 shows.
Relative damping

Rayleigh damping

Specified damping

f1 f2
f

Figure 2-12: An example of Rayleigh damping.

MECHANICAL DAMPING AND LOSSES | 123

 Since the coefficients αdM and βdK. should not be negative, the damping ratios are
constrained by the respective frequencies as

f1 ζ2 f2
---- ≤ ----- ≤ ----
f2 ζ1 f1

For many applications it is sufficient to leave αdM as zero and to define damping only
using the βdK coefficient. Then according to Equation 2-21, a damping which
increases linearly with frequency is obtained. If the damping ratio ξ(f0) or loss factor
η(f0) is known at a given frequency f0, the appropriate value for βdK is:

β dK = ζ ⁄ ( πf 0 ) = η ⁄ ( 2πf 0 )

Loss Factor Damping

Loss factor damping (sometimes referred to as material damping, structural damping,
or hysteretic damping) can be applied in the frequency domain.

The loss factor is a measure of the inherent damping in a material when it is

dynamically loaded. It is typically defined as the ratio of energy dissipated in unit
volume per radian of oscillation to the maximum strain energy per unit volume.

In COMSOL Multiphysics the loss information appears as a multiplier to the elastic

constitutive matrix Dc

c
D = ( 1 + jη s )D

For a nonlinear elastic material, this applies to the tangential stiffness.

The use of loss factor damping traditionally refers to a scalar-valued loss factor ηs. But
there is no reason that ηs must be scalar. Because the loss factor is a value deduced from
true complex-valued material data, it can be represented by a matrix of the same
dimensions as the anisotropic stiffness matrix. Especially for orthotropic materials,
there should be a set of loss factors of all normal and shear elasticity modulus
components, and COMSOL allows all these options, so a more general expression is.

c
D mn = ( 1 + jη s, mn )D mn

For hyperelastic materials the loss information appears as a multiplier in strain energy
density, and thus in the second Piola-Kirchhoff stress, S:

124 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

 ∂W s
S = ( 1 + jη s ) ----------
∂E

Loss factor damping is available for frequency response analysis and damped
eigenfrequency analysis in all interfaces.

Viscoelastic Damping
In some cases damping is included implicitly in the material model. This is the case for
a viscoelastic material, where damping operates on the shear components of stress and
strain.

When viscoelasticity is modeled in the frequency domain, it will act as a loss factor
damping. The complex modulus G*(ω) is the frequency-domain representation of the
stress relaxation function of viscoelastic material. It is defined as

G∗ = G′ + jG″ = ( 1 + jη s )G′

where G' is the storage modulus, G'' is the loss modulus, and their ratio ηs = G''/G' is
the loss factor. The term G' defines the amount of stored energy for the applied strain,
whereas G'' defines the amount of energy dissipated as heat; G', G'', and ηs can all be
frequency dependent.

Explicit Damping
It is possible to define damping by modeling the dissipative behavior of the material
using complex-valued material properties. In COMSOL Multiphysics, you can enter
the complex-valued data directly, using i or sqrt(-1) for the imaginary unit.

Equivalent Viscous Damping

Although equivalent viscous damping is independent of frequency, it is only possible
to use it in a frequency response analysis. Equivalent viscous damping also uses a loss
factor as the damping parameter, and can be implemented using the Rayleigh damping
feature, by setting the stiffness damping parameter βdK, to the loss factor, η, divided
by the excitation frequency:

η η
β dK = --------- = ----
2πf ω

The mass damping factor, αdM, should be set to zero.

MECHANICAL DAMPING AND LOSSES | 125

 Piezoelectric Damping
Piezoelectric losses are more complex and include coupling and electrical losses in
addition to the material terms. For damping in piezoelectric materials, see Piezoelectric
Losses.

For piezoelectric materials, βdK is only used as a multiplier of the structural

contribution to the stiffness matrix when building-up the damping matrix as given by
Equation 2-20. In the frequency domain studies, you can use the coupling and
dielectric loss factors equal to βdKω to effectively achieve the Rayleigh damping
involving the whole stiffness matrix.

Adding Damping in the Modal Solver

In COMSOL it is possible to solve a problem for a set of modes in the absence of
damping, and then to use those solutions as a modal basis to solve a problem in the
time (using a time domain modal study) or frequency domain (using a frequency
domain modal study). In both of these cases it is possible to manually assign a damping
ratio to the computed modes in the time or frequency domain study. To do that, right
click on the study and choose Show Default Solver, then expand the solver sequence
until the Modal Solver node is visible. In the settings window for that node, add
damping ratios for each of the modes.

For more details, see the section Modal Solver in the COMSOL
Multiphysics Reference Manual.

126 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

Modeling Geometric Nonlinearity
This section discusses how to model problems where displacements or strains are of a
size where the deformation of the structure has to be taken into account when
formulating the equations. Examples of the type of problems where this feature is
useful include:

• Thin structures, where the deflection is of the same order of magnitude as the
thickness.
• Where the structure exhibits large rotations. A rigid body rotation of only a few
degrees causes significant strains and stresses in a material where a linear strain
representation is used.
• Where the strains are larger than a few percent.
• Contact problems.
• Where a prestress must be taken into account for computing the dynamic response
of a structure.
• Buckling problems.
• Where a deformed mesh is used.
• Fluid-structure interaction problems.

• Contact Modeling
• The Fluid-Structure Interaction Interface
• The more formal theory is described in Analysis of Deformation

Geometric Nonlinearity, Frames, and the ALE Method

Consider the bending of a beam in the general case of a large deformation (see
Figure 2-13). In this case the deformation of the beam introduces an effect known as
geometric nonlinearity into the equations of solid mechanics.

Figure 2-13 shows that as the beam deforms, the shape, orientation, and position of
the element at its tip changes significantly. Each edge of the infinitesimal cube
undergoes both a change in length and a rotation that depends on position.
Additionally the three edges of the cube are no longer orthogonal in the deformed

MODELING GEOMETRIC NONLINEARITY | 127

 configuration (although typically for practical strains the effect of the
non-orthogonality can be neglected in comparison to the rotation).

From a simulation perspective it is possible to solve the equations of solid mechanics

on either a fixed domain (this is often called a Total Lagrangian formulation), or on
a domain that changes continuously with the deformation. The latter approach is often
called an Updated Lagrangian formulation. These two approaches also stand in
contrast to the Eulerian formulation which is often used for fluid mechanics. In an
Eulerian formulation the flow through a domain fixed in space is considered, while in
the Lagrangian formulation a fixed volume of material is considered.

In COMSOL Multiphysics, the concepts of a material frame and a

spatial frame are used. Equations formulated in the material frame will
give a Total Lagrangian formulation, while equations formulated in the
spatial frame will give an Eulerian formulation.

See Frames and Coordinate Systems in the theory chapter for more
details.

Solid mechanics in COMSOL Multiphysics is formulated on the material frame. This

is achieved by defining a displacement field for every point in the solid, usually with
components u, v, and w. At a given coordinate (X, Y, Z) in the reference configuration
(on the left of Figure 2-13) the value of u describes the displacement of the point
relative to its original position. The displacement is considered as a function of the
material coordinates (X, Y, Z), but it is not explicitly a function of the spatial
coordinates (x, y, z). The spatial coordinates is the current location in space of a point
in the deformed solid. As a consequence, it is only possible to compute derivatives with
respect to the material coordinates.

Taking derivatives of the displacement with respect to X, Y, and Z enables the

definition of a strain tensor. There are possible representations of the deformation. Any
reasonable representation must however be able to represent a rigid rotation of an
unstrained body without producing any strain. The engineering strain fails here, thus
it cannot be used for general geometrically nonlinear cases. One common choice for
representing large strains is the Green-Lagrange strain. It contains derivatives of the
displacements with respect to the original configuration. The values therefore
represent strains in material directions. This allows a physical interpretation, but it
must be realized that even for a uniaxial case, the Green-Lagrange strain is strongly
nonlinear with respect to the displacement. If an object is stretched to twice its original

128 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

length, the Green-Lagrange strain is 1.5 in the stretching direction. If the object is
compressed to half its length, the strain would read -0.375.

An even more fundamental quantity is the deformation gradient, which contains the
derivatives of the deformed coordinates with respect to the original coordinates:

∂x
F = ------- = ∇u + I
∂X

The deformation gradient contains all information about the local deformation in the
solid, and can be used to form many other strain quantities. As an example, the
Green-Lagrange strain is

1 T
ε = --- ( F F – I )
2

An element at point (X, Y, Z) specified in the material frame moves with a single piece
of material throughout a solid mechanics simulation. It is often convenient to define
material properties in the material frame as these properties move and rotate naturally
together with the volume element at the point at which they are defined as the
simulation progresses. In Figure 2-13 this point is illustrated by the small cube
highlighted at the end of the beam, which is stretched, translated, and rotated as the
beam deforms. The three mutually perpendicular faces of the cube in the Lagrange

MODELING GEOMETRIC NONLINEARITY | 129

 frame are no longer perpendicular in the deformed (spatial) frame. The deformed
frame Coordinates in this frame are denoted (x, y, z) in COMSOL.

Figure 2-13: An example of the deformation of a beam showing the undeformed state
(left) and the deformed state (right) of the beam itself with an element near its tip
highlighted (top), of the element (center) and of lines parallel to the x-axis in the
undeformed state (bottom).

It is important to note that, as the solid deforms, the Lagrangian frame

becomes a non-orthogonal curvilinear coordinate system (see the lower
part of Figure 2-13 to see the deformation of the X-axis). Particular care
is therefore required when defining physics in this coordinate system.

STRESS MEASURES
For example, in the spatial frame it is easy to define forces per unit area (known as
tractions) that act within the solid, and to define a stress tensor that represents all of
these forces that act on a volume element. Such forces could be physically measured,
for example using an implanted piezoresistor. The stress tensor in the spatial frame is
called the Cauchy or true stress tensor (in COMSOL this is referred to as the spatial
stress tensor). To construct the stress tensor in the Lagrangian frame a tensor
transformation must be performed on the Cauchy stress. This produces the second
Piola-Kirchhoff (or material) stress, which can be used with the material strain to solve

130 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

the solid mechanics problem in the (fixed) material frame. This is how the Solid
Mechanics interface works when geometric nonlinearities are enabled.

For the Cauchy stress tensor, both the force components and the normal to the area
on which the force is acting have fixed directions in space. This means that if a stressed
body is subjected to a pure rotation, the actual values of the stress components will
change. What was originally a uniaxial stress state might be transformed into a full
tensor with both normal and shear stress components. In many cases, this is neither
what you want to use nor what you would expect.

Consider for example an orthotropic material with fibers having a certain orientation.
It is much more plausible that you want to see the stress in the fiber direction, even if
the component is rotated. The Second Piola-Kirchhoff stress has this property as it is
defined along the material directions. In the figure below, an originally straight
cantilever beam has been subjected to bending by a pure moment at the tip. The
xx-component of the Cauchy stress and Second Piola-Kirchhoff stress are shown. Since
the stress is physically directed along the beam, the xx-component of the Cauchy stress
(which is related to the global x-direction) decreases with the deflection. The Second
Piola-Kirchhoff stress however, has the same through-thickness distribution all along
the beam, even in the deformed configuration.

Figure 2-14: xx-components of the Cauchy stress tensor (top) and Second-Piola-Kirchoff
stress tensor (below) for an initially straight beam

MODELING GEOMETRIC NONLINEARITY | 131

 Another stress measure available in COMSOL is the First Piola-Kirchhoff stress. It is a
multiaxial generalization of the nominal (or engineering) stress. The stress is defined
as the force in the current configuration acting on the original area. The First
Piola-Kirchhoff stress is an unsymmetric tensor, and is for that reason less attractive to
work with. Sometimes you may also encounter the Kirchhoff stress, although it is not
used in COMSOL. The Kirchhoff stress is just the Cauchy stress scaled by the volume
change. It has little physical significance, but can be convenient in some mathematical
and numerical operations.

Unfortunately, even without a rotation, the actual values of all these stress
representations are not the same. All of them scale differently with respect to local
volume changes and stretches. This is illustrated in the graph below. The
xx-component of four different stress measures are plotted at the fixed end of the beam
from the example above. At this point, the beam axis coincides with the x-axis, so the
directions of all stress tensor components coincide. In the center of the beam, where
strains, and thereby volume changes are small, all values approach each other. For a
case with large rotation but small strains, the different stress representations can be
seen as pure rotations of the same stress tensor.

Figure 2-15: Stress distribution across the beam at the constrained end

132 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

If you want to compute the resulting force or moment on a certain boundary based on
the stresses, there are in practice only two possible choices: Either integrate the Cauchy
stress over the deformed boundary, or integrate the First Piola-Kirchhoff stress over
the same boundary in the undeformed configuration. In COMSOL Multiphysics this
corresponds to selecting either Spatial frame or Material frame in the settings for the
integration operator.

ALE METHOD
In the case of solid mechanics, the material and spatial frames are associated directly
with the Lagrangian and Eulerian frames, respectively. In a more general case (for
example, when tracking the deformation of a fluid, such as a volume of air surrounding
a moving structure) tying the Lagrangian frame to the material frame becomes less
desirable. Fluid must be able to flow both into and out of the computational domain,
without taking the mesh with it. The Arbitrary Lagrangian-Eulerian (ALE)
method allows the material frame to be defined with a more general mapping to the
spatial or Eulerian frame. In COMSOL, a separate equation is solved to produce this
mapping—defined by the mesh smoothing method (Laplacian, Winslow, hyperelastic,
or Yeoh) with boundary conditions that determine the limits of deformation (these are
usually determined by the physics of the system, whilst the domain level equations are
typically being defined for numerical convenience). The ALE method offers significant
advantages since the physical equations describing the system can be solved in a
moving domain.

Deformed Geometry and Moving Mesh in the COMSOL Multiphysics

Reference Manual

Geometric Nonlinearity for the Solid Mechanics Interface

For the Solid Mechanics interface, or any multiphysics interface derived from it, you
enable a geometrically nonlinear analysis for a certain study step by selecting the Include
geometric nonlinearity check box in the Study Settings section of a study step.

If any active feature in the model requires the analysis to be geometrically nonlinear,
the Include geometric nonlinearity check box is selected automatically, and it cannot be
cleared. The physics features which force this behavior are:

• Contact, because the deformation between the contacting boundaries must be part
of the solution

MODELING GEOMETRIC NONLINEARITY | 133

 • The Fluid-Structure Interaction interface.
• Hyperelastic materials, which are always formulated for large strains,

Usually you would also want to use geometric nonlinearity when a Moving Mesh
interface is present, but this is not forced by the program.

When you select a geometrically nonlinear study step, the behavior of several features
differ from that in a geometrically linear case:

• There is an important difference between using uppercase (X, Y, Z, R) and

lowercase (x, y, z, r) coordinates in expressions. The lowercase coordinates
represent the deformed position, and this introduces a dependency on the solution.
• Many features, such as coupling operators, can be specified as operating either in the
material (X, Y, Z) or the spatial (x, y, z) frame. This setting does not make a
difference unless a geometrically nonlinear analysis is performed. In most cases you
would want to do the operation in the material frame.
• The strain representation is changed from using engineering strains to
Green-Lagrange strains, unless the Force linear strains check box is selected in the
Geometric Nonlinearity section of a certain material.
• Where Green-Lagrange strains are used, Second Piola-Kirchhoff stresses are also
used. This means that material data must be given in terms of these quantities. This
distinction is important only when the strains actually are large.
• Pressure loads are interpreted as follower loads, so that the direction of the load as
well as the loaded area are deformation dependent.
• Rigid connectors take finite rotations into account.

Studies and Solvers in the COMSOL Multiphysics Reference Manual

Geometric Nonlinearity for the Shell, Plate, Membrane, Beam and

Truss Interfaces
For the Shell, Plate, Membrane, Beam, and Truss interfaces, a geometrically nonlinear
analysis is enabled in the same way described above. For the Membrane interface,
geometric nonlinearity must almost always be used, since it is nonlinear effects which
supply the stiffness in the transverse direction.

134 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

The geometric nonlinearity in the Beam interface is limited to large rotations and
displacements, but small strains.

The effect of using geometric nonlinearity in these interfaces is limited to the stress and
strain representation as the other effects described in Geometric Nonlinearity for the
Solid Mechanics Interface are not relevant.

Solving Geometrically Nonlinear Problems

Depending on the geometrically nonlinear effects that appear in your model, you may
have to use different solution strategies. Some problems in this class are strongly
nonlinear, while others show only a weak deviation from linearity. Some guidelines are:

• If the problem has a path dependent solution, then it must be solved in an

incremental way in order to give a correct solution. Problems including for example
plasticity or friction belong to this class. If you do the analysis in time domain, then
the solution is inherently incremental. If the analysis is stationary, invoke the
parametric continuation solver by adding an Auxiliary Sweep, and ramp up some
loading parameter. In either case, make sure that the step size is not too large.
• Problem that have a unique solution, like an elastic model subjected to large
rotations or strains can be solved in a single static load step without loss of accuracy.
It is however possible that such an approach will not converge, in which case a
parametric continuation solver must be used.
• In problems involving large rotations, the default settings of the nonlinear solver
will sometimes give a too conservative solution strategy. You can often decrease the
solution time significantly by modifying the settings under Method and Termination
in the settings for the Fully Coupled or Segregated Step node in the solver sequence.
Set Nonlinear method to Constant (Newton) and use a rather high Damping factor. In
most cases the value 1 will work.
• If you model a situation which to a large extent is a rigid rotation, it is often
necessary to use tighter tolerances than usual in order to avoid spurious stresses.
Since the strains are computed from the differences of the displacements in an
element, even a small relative error in the displacements (which are large) can cause
significant strains. This will be visible in a case where the actual stresses are small.

Prestressed Structures
You can analyze eigenfrequency or frequency domain problems where the dynamic
properties of the structure are affected by a preload, such as a tensioned string.

MODELING GEOMETRIC NONLINEARITY | 135

 Usually, a study of a prestressed problem includes using study steps. The first step is a
Stationary step in which the static preload is applied. The effects of the preload can be
computed with or without taking geometric nonlinearity into account. In the second
study step, where the you compute the eigenfrequency or the frequency response, it is
necessary to take geometric nonlinearity into account. Even if the displacements and
strains are small, this is what gives the prestress contribution to the equations.

The same principles apply also to a linear buckling analysis, except that both study steps
should be geometrically linear. The nonlinear contribution is included in the
formulation of the buckling eigenvalue itself.

FOLLOWER LOADS
Loads which change orientation with deformation, such as a pressure, actually
contribute not only to the load, but also to the stiffness. This is a physical effect, and
not just a numerical artifact. Whether such loads are included or not in an
Eigenfrequency study step will affect the computed eigenfrequencies. If you for some
reason do not want this effect, you must suppress the load in the Physics and Variables
section of the Eigenfrequency node.

If you use a local coordinate system for describing a load, you must in case
of geometric nonlinearity pay attention to whether that coordinate system
has constant axis orientations or not. As an example, the default boundary
system has Frame set to Deformed Configuration, so that a load represented
in that system will behave as a follower load. Change to Reference
Configuration if the load should act in fixed directions.

CONTACT AND FREQUENCY DOMAIN ANALYSIS

If a contact is included in the prestress load case, you can perform a subsequent
frequency domain analysis in which then the linearization will be made around the
computed contact state. It is however not possible perform an eigenvalue analysis
based on a contact solution.

136 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

INELASTIC STRESSES AND STRAINS
When inelastic stresses and strains are part of the problem description, you need to
make some extra considerations. Such contributions are formally part of the
constitutive model, but generate load vector contributions.

There are three Preset study types which can be used to set up these two
study steps: Prestressed Analysis, Eigenfrequency; Prestressed Analysis,
Frequency Domain; and Linear Buckling.

If you want to explicitly prescribe the stress field for a prestressed analysis
rather than solving for it, you should not use the two study step
procedure.

In such a case, prescribe the stress field using an Initial Stress and Strain or
External Stress node. Then add a separate Eigenfrequency study and select
Include Geometric Nonlinearity in the settings for the study step.

Prestressed Analysis, Eigenfrequency, Prestressed Analysis, Frequency

Domain, and Linear Buckling in the COMSOL Multiphysics Reference
Manual

• For an example of a general prestressed eigenfrequency analysis see

Bracket—Eigenfrequency Analysis: Application Library path
Structural_Mechanics_Module/Tutorials/bracket_eigenfrequency.
• For an example of an analysis where the stress state is explicitly
prescribed, see Vibrating Membrane: Application Library path
Structural_Mechanics_Module/Verification_Examples/
vibrating_membrane.

Geometric Nonlinearity for the Piezoelectric Material

PIEZOELECTRIC MATERIALS WITH LARGE DEFORMATIONS

The linear piezoelectric equations as presented in About Piezoelectric Materials with
engineering strains are valid if the model undergoes only relatively small deformations.
As soon as the model contains larger displacements or rotations, these equations

MODELING GEOMETRIC NONLINEARITY | 137

 produce spurious strains that result in an incorrect solution. To overcome this
problem, so-called large deformation piezoelectrical equations are required.

The Piezoelectric Material implements the large deformation piezoelectrical equations

according to Yang (Ref. 1). Key items of this formulation are:

• The strains are calculated as the Green-Lagrange strains, εij:

1  ∂u i ∂u j ∂u k ∂u k
ε ij = ---  -------- + -------- +
2  ∂X j ∂X i  --------
- ---------
∂X i ∂X j 
(2-22)
k

Green-Lagrange strains are defined with reference to an undeformed geometry.

Hence, they represent a Lagrangian description. In a small-strain, large rotational
analysis, the Green-Lagrange strain corresponds to the engineering strain in
directions that follow the deformed body.
• Electrical field variables are calculated in the material directions, and the electric
displacement relation is replaced by an expression that produce electric polarization
in the material orientation of the solid.
• In the variational formulation, the electrical energy is split into two parts: The
polarization energy within the solid and the electric energy of free space occupied
by the deformed solid.

The first two items above result in another set of constitutive equations for large
deformation piezoelectricity:

T
S = cE ε – e E m
P m = eε + ε 0 ( ε rS – I )E m

where S is the second Piola-Kirchhoff stress; ε is the Green-Lagrange strain, Em and

Pm are the electric field and electric polarization in the material orientation; I is the
identity matrix; and cE, e, and εrS are the piezoelectric material constants. The
expression within parentheses equals the dielectric susceptibility of the solid:

χ = ε rS – I

The electric displacement field in the material orientation results from the following
relation

–1
D m = P m + ε 0 JC E m

where C is the right Cauchy-Green tensor

138 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

 T
C = F F

Fields in the global orientation result from the following transformation rules:

–T
E = F Em
–1
P = J FP m
(2-23)
–1
D = J FD m
–1
ρv = ρV J

where F is the deformation gradient; J is the determinant of F; and ρv and ρV are the
volume charge density in spatial and material coordinates respectively. The
deformation gradient is defined as the gradient of the present position of a material
point x = X + u:

∂x
F = -------
∂X

Finally, one can rewrite the constitutive equations as

T
S = cE ε – e Em
–1
D m = eε + ε 0 ε rS E m + ε 0 ( JC – I )E m

DECOUPLED MATERIALS WITH LARGE DEFORMATIONS

The large deformation formulation described in the previous section applies directly to
non-piezoelectric materials if the coupling term is set to zero: e = 0. In that case, the
structural part corresponds to the large deformation formulation described for the
solid mechanics interfaces.

The electrical part separates into two different cases: For solid domains, the electric
energy consists of polarization energy within the solid and the electric energy of free
space occupied by the deformed solid—the same as for the piezoelectric materials. For
non-solid domains this separation does not occur, and the electric displacement in
these domains directly results from the electric field—the electric displacement
relation:

MODELING GEOMETRIC NONLINEARITY | 139

 D = ε0 εr E

The Electrostatics Interface in the COMSOL Multiphysics Reference

Manual

LARGE DEFORMATION AND DEFORMED MESH

The Piezoelectric Devices Interface can be coupled with the Moving Mesh (ALE)
interface in a way so that the electrical degrees of freedom are solved in an ALE frame.
This feature is intended to be used in applications where a model contains non-solid
domains, such as modeling of electrostatically actuated structures. This functionality is
not required for modeling of piezoelectric or other solid materials.

The use of ALE has impacts on the formulation of the electrical large deformation
equations. The first impact is that with ALE, the gradient of electric potential directly
results in the electric field in the global orientation, and the material electric field
results after transformation.

The most visible impact is on the boundary conditions. With ALE any surface charge
density or electric displacement is defined per the present deformed boundary area,
whereas for the case without ALE they are defined per the undeformed reference area.

Deformed Geometry and Moving Mesh in the COMSOL Multiphysics

Reference Manual

References
1. J. Yang, An Introduction to the Theory of Piezoelectricity, Springer Science and
Business Media, N.Y., 2005.

2. A.F. Bower, Applied Mechanics of Solids, CRC Press, Boca Raton, FL (http://
www.solidmechanics.org), 2010.

140 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

C o nt a c t M o de lin g
General
Situations where objects come into contact with each other occur frequently in
mechanical simulations. Setting up and solving contact problems is sometimes a
challenging task, and this section contains information about important aspects of
creating models involving contact.

In a contact analysis, you solve for the contact state and the contact forces. If friction
is ignored, the state only consists of being in contact or not, and the force variable is
the contact pressure in the normal direction. With friction included, there are two
possible states for the relative tangential displacement when in contact: sticking and
sliding. The tangential force vector is added as force variable.

The contact analysis functionality in COMSOL Multiphysics also includes the

possibility to prescribe adhesion and decohesion between the contacting objects.

In this section:

• Setting up a contact problem • Multiphysics Contact

• Contact Pairs • Fallback Nodes for Contact
• Meshing for Contact Analysis • Solver Settings for Contact Analysis
• Constraints • Monitoring the Solution
• Settings for Contact Nodes • Dependent Variables in Contact
• Time-Dependent Contact Analysis Analysis
• Important Contact Variables

Setting up a contact problem

Mechanical contact can be modeled in the Solid Mechanics and Multibody Dynamics
interfaces. To model a mechanical contact problem, you must do the following:

• Add Contact Pair nodes under Definitions. A contact pair consists of two sets of
boundaries, which are called the source and destination boundaries.

CONTACT MODELING | 141

 • Add Contact nodes in the physics interface. In the Contact node, you select the
contact pairs to be used, and provide the settings for the physical and numerical
properties of the contact model.
• In the finalization step of the geometry sequence, you should normally have Action
set to Form an assembly. If Form a union is used, then the contacting boundaries must
be geometrically separated in the initial configuration.
• Select a suitable study type. You can analyze contact using a Stationary or Time
Dependent study step.

• Documentation of the Contact, Friction and Adhesion nodes.

• Contact Analysis Theory.
• Identity and Contact Pairs in the COMSOL Multiphysics Reference
Manual.
• Time-Dependent Contact Analysis.

In a multiphysics analysis, a contact problem can also incorporate for example changes
in the heat flux or electric current through the contacting boundaries. You will then
also need to add corresponding features in the other participating interfaces, like a
Thermal Contact node in the Heat Transfer in Solids interface. The contact state and
contact pressure used in other physics interfaces is always supplied by the structural
mechanics interface.

The fact that you add a Contact node to your model will make all study steps
geometrically nonlinear.

INCLUDING FRICTION
In real life, there is always some friction between contacting objects, but this is often
ignored. There are several reasons to do this simplification:

• In many cases, only the normal forces are significant for the general force
distribution in the structure, while the frictional forces only cause a minor local
effect.
• The values of the friction coefficients are difficult to obtain, and unless the structure
is assembled under well controlled conditions, the magnitude of the friction can
vary a lot.
• Adding friction to a contact problem will often increase the computation time
significantly, or even cause convergence problems.

142 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

There are a number of situations when friction modeling cannot be avoided. Some of
them are:

• When a significant portion of the load is carried by shear stresses in the contact
boundaries.
• When a tangential force is necessary to maintain stability and avoid rigid body
motions. In many cases, it is however possible to replace the friction by a suitable
constraint instead, as long as there are no resultant forces being resisted by such a
constraint.
• When the frictional dissipation is an important component of a dynamics problem,
or when it is needed as a heat source in a thermal analysis.

ADHESION AND DECOHESION

You can also specify that the contacting boundaries stick to each other so that they will
not separate or slide. The onset of adhesion, when the boundaries become
permanently attached to each other, can be based on several criteria:

• When a certain contact pressure is exceeded.

• When the gap distance between the contact boundaries is smaller than a certain
value.
• When a user specified logical expression is fulfilled. This can for example be used if
an adhesive cures when a certain temperature is exceeded.
• From the start of the analysis. This case is particularly interesting if you are
interested in modeling the tearing of a thin glue layer by decohesion.

If adhesion is active between the contact boundaries, it is possible to break the bond
by adding a decohesion rule. You can choose between several different decohesion
laws.

SELECTING THE CONTACT ALGORITHM

In COMSOL Multiphysics, there are two possible methods for solving contact
problems: the augmented Lagrangian method and the penalty method.

The default augmented Lagrangian method provides better accuracy, but at a higher
computational cost. It requires additional degrees of freedom, and is less stable from
the convergence point of view. This method ensures that there will not be any
penetration between the contacting objects in a well converged solution. The contact
pressure and friction forces are added as degrees of freedom. The contact pair is
asymmetric (that is, it is a source/destination pair). The destination contact domain
is constrained not to penetrate the source domain, but not vice versa. The contact

CONTACT MODELING | 143

 condition is evaluated in the integration points on the destination boundary. It is thus
possible for a node to have a small penetration into the source boundary, even in a
converged state.

In comparison, the penalty method is rather simple and robust. Roughly speaking, it
is based on inserting a stiff spring, active only in compression, between the contacting
boundaries. In addition to the robustness, it has the advantage that no special solver is
required, which makes it easier to set up multiphysics problems and time-dependent
studies.

The penalty method must be used when modeling adhesion.

You can select the method used for computing the normal direction pressure and the
friction forces independently.

Contact Pairs
To decide which boundaries to assign as source and destination in a contact pair
consider the following guidelines:

• Make sure that the source boundary stiffness in the normal direction is higher than
the destination boundary stiffness. This is especially important if the difference in
stiffness is quite large (for example, over ten times larger).
• If one of the boundaries belongs to a part that is rigid, either since it is a rigid
domain, or because of the constraints applied, it should be selected as the source
boundary.
• When the contacting parts have approximately the same stiffness, consider the
geometry of the boundaries instead. Make a concave boundary the source and a
convex boundary the destination rather than the opposite.

For efficiency, include only those boundaries that can actually come in contact in the
destination selection. For the source, it is often a bit more efficient to make it so large
that every destination point ‘sees’ a corresponding source point. The source point is
obtained by following the normal to the destination until it reaches the source
boundary.

FIXED GEOMETRY CONTACT

In some situations, the relative sliding between the contacting boundaries is small. This
is often the case for shrink fit simulations or when mounting a component using
prestressed bolts. The sliding can be considered as small if it is significantly less than
the length of an element edge.

144 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

In this case, it is possible to simplify the problem by selecting the Mapping Method to
be Initial Configuration in the Contact Pair node. With this setting, a certain point on
the destination boundary will see the same point on the source boundary during the
entire simulation. This will make the contact analysis to run faster and convergence to
be more stable.

The analysis is still geometrically nonlinear when using this option, and the contact
region can still have arbitrarily large displacements and rotations.

AUTOMATIC GENERATION OF CONTACT PAIRS

Contact pairs can be automatically generated during the finalization of the geometry
sequence. When Action is set to Form an assembly, you can select Create pairs, and use
Contact pair as Pair type. Boundaries which are in geometrical contact with each other
will then be placed in contact pairs. When you add Contact nodes in the physics
interface, you select which of these suggested pairs to actually use for the contact
analysis.

The automatic pair generation will not know which side to use as source
or destination. Based on the rules above, you may need to switch them
using the Swap Source and Destination button in the Source Boundaries
section of the pair settings.

Meshing for Contact Analysis

Once the source and destination boundaries are chosen, mesh the destination finer
than the source. Do not make the destination mesh just barely finer than the source
because this can cause nonphysical oscillations in the contact pressure. Make the
destination at least two times finer than the source. The reason is that the algorithm is
asymmetric; the points on the destination side connects to the source side, and not vice
versa. So with a coarse mesh on the destination side, a large portion of an element (or
even a whole element) on the source side could be without connection to the
destination.

It is always important that the geometry is well resolved, so that a curved contact
boundary actually is seen as curved rather than ‘faceted’. The density of the mesh often
needs to be finer than what would be needed to resolve stresses on a similar boundary
without the contact conditions. If the normals to the contact boundaries change much
from one element to the next, there is a risk that the contact analysis does not
converge.

CONTACT MODELING | 145

 If the source boundary is rigid, there are no requirements on the mesh size of the
destination boundary. In this case, you may use a significantly finer mesh on the source
boundary than on the destination boundary. This is sometimes necessary in order to
resolve the geometry well.

Constraints
Make sure that the bodies are sufficiently constrained. If the bodies are not in contact
in the initial configuration, and there are no constraints on them, there will be possible
rigid body displacements. This causes the solver to fail and must be avoided.

For a more detailed discussion about sufficient constraints, see

Constraints under Stationary Analysis.

Sometimes, as when simulating mounting processes, the structure is fully constrained

only once the contact is fully established. There are some strategies for how you can
deal with this problem.

• Create the geometry or set initial values for the displacement variables so that there
is a small penetration in the initial configuration.
• Use boundary conditions giving a prescribed displacement rather than a prescribed
force. When possible, this is usually the best solution to the stability problem. Note
that you can always obtain the force actually applied from the reaction forces.
• Add a temporary weak spring during the beginning of the simulation. Assuming
that a parameter p, ranging from 0 to 1, is used for applying the external load, you
can introduce a stabilizing spring with stiffness kx in the x direction as
– ( p ⋅ 10 )
k x = k ( 1 – p )2

and similarly in any other direction that needs to be restrained. It is not important
whether the spring is applied to domains, boundaries, or edges, but it should not create
local forces. The value for the stiffness k should be chosen so that the force generated
by the spring balances the external load at a sufficiently small displacement. A too weak
spring will give a too large initial overclosure of the contact boundaries. A too stiff
spring might influence the solution too much.

146 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

Settings for Contact Nodes

PENA LT Y FACT OR
An important parameter in the settings for the Contact node is the penalty factor. It is
used for both the augmented Lagrangian method and the penalty method, but the
interpretation is different:

• In the augmented Lagrangian method, the penalty factor is a numerical parameter,

which affects the convergence properties of the algorithm. A higher value gives a
faster convergence, but at the price of less stability.
• In the penalty method, the penalty factor has a straightforward interpretation as the
stiffness of a distributed spring connecting the two contacting boundaries. A higher
value will decrease the unphysical penetration, but can be detrimental to the
numerical conditioning of the stiffness matrix.
• When modeling adhesion, the penalty factor is the stiffness in the normal direction
of the adhesive layer.

In the augmented Lagrangian method, the value of the penalty factor

does not affect the accuracy of the final solution, like it does in the penalty
method.

When running into convergence problems, check the penalty settings. If the iteration
process fails in some of the first iterations, lower the penalty parameters. If the model
seems to converge but very slowly, consider increasing the maximum value of the
penalty parameters.

Increasing the penalty factor can lead to an ill-conditioned Jacobian matrix and
convergence problems in the Newton iterations. This is often seen by noting that the
damping factor becomes less than 1 for many Newton iterations or that the structure
“jumps” into an unphysical state. If this occurs, decrease the penalty factors.

The default values for the penalty factors is based on an “equivalent” Young’s modulus
(Eequiv) of the material on the destination side. For linear elastic isotropic materials
Eequiv is equal to the actual Young’s modulus. For other types of materials Eequiv is
defined by an estimate of a similar stiffness at zero strain. For materials that are user
defined or in other senses nonstandard (for example, anisotropic with large differences

CONTACT MODELING | 147

 in stiffness in different directions), Eequiv might need to be replaced with another
estimate.

For nonlinear materials in general, and for elastoplastic materials in

particular, there can be a significant change in stiffness during the solution
process. Choose the source and destination boundaries accordingly. You
may even have to adjust the penalty factor as the solution progresses when
such materials form a contact boundary.

Penalty Factor Relaxation

When using the augmented Lagrangian method, having a well tuned penalty factor is
an important factor for the performance of the contact iterations.

The default value is selected as a compromise between speed and stability, but with
more weight on stability. The strategy is to for each new step (parametric step or time
step) start with a softened penalty factor, which is then increased over the first four
iterations. The purpose is to stabilize the problem if there are large overclosures in the
first iterations. This is called relaxation.

In a situation where the contact is well established, using relaxation will however cost
extra iterations, and it may even lead to a loss of convergence.

The penalty factor can be tuned in several ways. You have three basic choices, ranging
from simple to advanced:

• With a Preset penalty factor, you can choose having it tuned for Stability or Speed.
With Stability, relaxation is used in every step. With Speed, a constant penalty factor
is used all the time, and the value used is also higher than the final value obtained
when using Stability.
• With Manual tuning, you can make adjustments to the magnitude of the penalty
factor, and to the relaxation strategy.
• With User defined, you can enter any expression for the penalty factor.

For details about these settings, see the documentation of the settings for
the Contact node.

Some hints for selecting the penalty factor:

• Use relaxation only when large changes in the contact state is expected.

148 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

• If an analysis takes a long time, check the convergence graphs. If the contact
variables show a steady, but slow, convergence it may help to increase the penalty
factor. Increase by a factor of 2–5.
• If a model fails to converge and the contacting parts appear to have moved far from
each other, it is probable that the penalty factor is too high. You can then either
decrease the total stiffness or add more relaxation.

TR I G G E R C U T B A C K
If, during the iterations, a contact problem comes into a state where it is far from the
converged solution, it is unlikely that the solution will ever converge. In such a case,
much computing time can be spent before the maximum number of iterations is
reached, and the solver makes an attempt with a smaller time or parameter step. To
speed up this process, you can select the Trigger cutback check box. You can then enter
a logical expression which, when fulfilled, will force the solver to immediately abandon
the iterations and try a smaller step. Such an expression can, for example, be a
maximum displacement (like solid.disp > 5[mm]), based on what is physically
realizable for the structure. The expression is evaluated in all points on the boundary,
but you can also use a global value, like an average or a maximum.

OFFSET AND ADJUSTMENT

It is possible to assign an offset to both the source and destination boundaries. When
an offset is given, the boundary used in the computations is not the geometrical
boundary, but a virtual boundary displaced by the offset value. You can use this option
for several purposes:

• When analyzing problems with for example shrink fit, nominal dimensions can be
used for the geometry, and the overclosure in included in the gap by using the
offset.
• When there is a small clearance between two boundaries, a negative offset can be
used instead of changing the geometry.
• If geometries are such that a large overclosure exists in the initial configuration, the
contact algorithms may not converge. You can then add a negative offset, which is
slowly removed by letting it depend on time or on the parameter in the parametric
continuation solver.

When the source or destination boundaries are curved, the discretization introduced
by the meshing may lead to small variations in the computed distance between the
source and destination boundaries, even though the geometrical shapes of the two
objects are ideal. When modeling for example a shrinkage fit, this effect can cause

CONTACT MODELING | 149

 significant fluctuations in the contact pressure. If you select Force zero initial gap, the
computed distance from destination to source will be adjusted by the initial distance.
This adjustment can be combined with an offset. The offset is applied to the adjusted
value.

The mesh as such is not adjusted, it is only the distance computation which is affected.
This type of adjustment is most useful when the sliding is small, so that the gap distance
is always computed between the same points on source and destination.

INITIAL VALUE
In the augmented Lagrangian method, where the contact pressure is a dependent
variable, it can be given an initial value. In force-controlled contact problems where no
other stiffness prohibits the deformation except the contact, the initial contact pressure
is crucial for convergence. If it is too low the parts might pass through each other in
the first iteration. If it is too high they never come into contact.

DISCRETIZATION
When using the augmented Lagrangian method it is possible to change the order of
the shape functions used for the contact pressure and friction force degrees of freedom.
The default is linear shape functions, and this matches the quadratic shape functions
used as default for the displacement degrees of freedom in the Solid Mechanics
interface. The only situation when you should consider changing the discretization for
the contact variables is if you use cubic or higher shape functions for the displacements.

Selecting any other discretization else than Linear requires that the solver sequence is
modified manually, since the default lumped solver is no longer optimal for the contact
pressure update. An ordinary segregated step should then be used.

ADHESION
When using the penalty method, you can specify that the boundaries in the contact pair
should stick to each other after coming into contact.

The adhesive layer is conceptually without thickness, but by specifying on offset in the
Contact node, you can to some extent include the dimensions of the adhesive layer.

The adhesive layer always has a finite stiffness. For a glue layer, this represents the true
stiffness. For a more conceptual joining of two boundaries, this stiffness should be
considered in the same way as the penalty stiffness in the contact formulation. The
stiffness can differ between tension and compression: In compression the stiffness is
always taken as the penalty stiffness, whereas you can change the tensile stiffness.

150 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

Decohesion
Two boundaries joined by adhesion can separate if you select any of the decohesion
laws. You must then specify the maximum strength and the total energy release during
decohesion for pure tension and pure shear. In most cases, the decohesion process is a
mixture between tension and shear, so you also have to specify the interaction rule.

As long as the stress is below the maximum strength, the adhesive layer retains it
original stiffness. Once the displacement in the layer is so large that the maximum
strength is exceeded, the material is considered as damaged, and the stiffness is
decreased. The same damage factor is applied to both the tensile and the shear stiffness,
while the compressive stiffness is unchanged.

Decohesion is an inherently unstable process. The elastic energy in the strained

adhesive layer will be released during decohesion. Numerically, the decreasing
stress-strain curve manifests itself as a local negative stiffness. Such a problem may be
possible to solve, if the surrounding material can absorb the released energy. The
numerical stability is closely coupled to the physical stability of the structure.

• If the structure is in load controlled situation, there is no possibility to continue the

analysis when the peak external load has been exceeded. Physically, this corresponds
to a sudden collapse of the component.
• If the structure is loaded by prescribed displacements, it is usually possible to
continue the analysis further. It may however happen that the stored elastic energy
in the structure is large enough to force a complete breakdown of the adhesive layer
once a certain external displacement is reached.
• If possible, use prescribed displacements in a decohesion analysis, and evaluate the
applied load from the reaction forces.
• Sometimes it is not possible to use prescribed displacements, for example if the load
is distributed. You can then add a Global Equation to control the loading rate by
some other quantity that increases monotonically. This is the same technique that is
used for post-buckling problems.
• You may have to change the settings in the Method and Termination section of the
settings for the Fully Coupled or Segregated nodes in the solver sequence, for example
by allowing a larger number of iterations.

CONTACT MODELING | 151

 • The true unstable failure is a dynamic event. This can be analyzed using a time
dependent solver, but the computational cost can be high. The inertial forces will
then balance the released energy.

• Documentation of the Contact, Friction and Adhesion nodes.

• Adhesion and Decohesion in the Structural Mechanics Theory chapter.

For an example showing decohesion analysis, including how to use a

global equation to control an unstable problem, see Mixed-Mode
Debonding of a Laminated Composite: Application Library path
Structural_Mechanics_Module/Contact_and_Friction/
cohesive_zone_debonding

Time-Dependent Contact Analysis

The contact formulation is strictly valid only for stationary problems. It is still possible
to use contact modeling in a time-dependent analysis, as long as inertial effects are not
important in the contact region or you use a sufficiently short time step. In practice,
this means that you can solve both quasi-static problems and truly dynamic problems,
as long as situations with impact are avoided. When in doubt, try to do an a posteriori
check of conservation of momentum and energy to ensure that the solution is
acceptable.

In time-dependent contact models, the penalty method is usually the better choice.

Multiphysics Contact
Multiphysics contact problems often exhibit a high degree of nonlinearity, which leads
to convergence problems for the nonlinear solver. As ab example, consider heat
transfer through the contact area, where initially only a small spot is in contact. The
solution for the temperature field is then extremely sensitive to the size of the contact
area. If at the same time, the solid deforms due to thermal expansion, there may be
large changes in the contact area between each iteration,

It is important to resolve the size of the contact area accurately, that is, to
use a very fine mesh in the contact area when modeling fully coupled
multiphysics problems.

152 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

If the contact area is larger, a fine mesh is not required because then the temperature
solution is not that sensitive to the size of the contact area. If possible, start with an
initial configuration where the contact area is not very small.

You can use the contact variables (gap and contact pressure) in expressions for
quantities in other physics interfaces. As an example, a thermal resistance in the contact
region can depend on the contact pressure.

In many cases, the penalty method is preferred in multiphysics contact problems

because of its better stability and less restrictive requirements on solver selections. If
the contact conditions depend strongly on the contact pressure, use the augmented
Lagrangian method because if its higher accuracy.

Fallback Nodes for Contact

When a Contact node has been added, it always contains one subnode, Free. This is the
default fallback condition on the part of the boundary where contact is not established.
By right-clicking the Contact node, you can replace it by another fallback condition.
The most common case is when there is a pressure load acting on the part of the
boundary that is currently not in contact. In this case you would add a Boundary Load
fallback node.

• If you add an ordinary Boundary Load node after a Contact node and use
the same boundary selection, the load will be applied also to the part
of the boundary that is in contact. In most cases, this is not a physical
behavior.
• If a Boundary Load node appears before a Contact node in the model
tree, the load on the common selection will be ignored, irrespective of
whether the boundary actually is in contact or not.

Solver Settings for Contact Analysis

The following solver settings can help to successfully solve contact models:

• In a contact analysis, you almost invariably use an incremental approach. It is

possible to solve a problem without friction in a single stationary load step, but such
an approach will often fail to converge. In a stationary analysis, you should then use
the parametric continuation solver, and gradually increase the load or displacement.

CONTACT MODELING | 153

 Enable it by selecting Auxiliary sweep under Study Extensions in the settings for the
Stationary solver.
• Use a direct solver instead of an iterative solver as linear equation solver if the
problem size allows it. Direct solvers are less sensitive and can provide better
convergence.
• As a default, the double dogleg nonlinear solver is selected when a stationary study
is generated and Contact nodes are present in the model. For the majority of contact
problems this solver has more stable convergence properties than the Newton
solver, which is the generally used default solver for other problems. Using the same
settings, the Double dogleg solver tends to be somewhat slower than the Newton
solver on problems where both solvers converge. It is however often possible to take
larger load steps when using the double dogleg solver. For some problems, the
Newton solver can still be the better choice, so if you experience problems using the
default settings, try to switch solver.
• It is important to scale the contact degrees of freedom manually. The convergence
check relies on the scaling of the degrees of freedom, but since contact pressures and
friction forces often are zero over parts of the simulation, you should not rely on
automatic scaling. When the solver sequence is first created, both contact pressure
and friction forces are given a manual scaling which is relevant for typical
metal-to-metal contact. You should in most cases change this to values appropriate
for your application. The variable scaling is accessed under Dependent Variables in the
solver sequence. Set the scale for each variable to a value which is representative for
the expected result. Too large values may give a too early convergence, while too
small values may lead to an excessive number of iterations.
• For some contact problems, it is necessary to let the parametric solver use a more
defensive strategy when going to the next parameter step. This can be controlled by
setting the value of Predictor in the Parametric feature to Constant.
• When using the augmented Lagrangian method, at least one lumped step will be
generated in the segregated solver for each Contact node. The number of lumped
steps will be increased even more if you select the Group contact variables in solver
per pair check box in the Advanced section of the settings for the Contact node. This
split of variables into different lumped steps does not influence the solution as such;
you can equally well group the contact variables in a single lumped step. Each
lumped step will however generate an individual curve in the convergence plot,
making it easier to pinpoint the source of convergence problems.
• If the model includes friction, try solving the problem without friction first. When
the model seems to work without friction, friction can be added.

154 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

• Always solve contact problems which contain friction or decohesion incrementally,
using a parametric or time-dependent solver. The evolution of the friction forces is
history dependent. For contact problems without friction an incremental strategy is
not necessary but often a good choice.
• For models that include decohesion, see also suggestions under Decohesion.

Monitoring the Solution

It is often useful to monitor the solution during a contact analysis. This can be done
in different ways.

Using the Results while solving functionality in the study step is a good practice. You
can either use a stress plot, or a plot of the contact pressure. In most cases, the scale of
a deformed plot should be set to 1 when monitoring contact problems. Note that if
you select Results while solving in the Segregated feature, the plot is updated after each
iteration, thus allowing you to monitor the convergence in detail.

For each contact pair, two global variables that can be used in probe plots are available.
These are the maximum contact pressure (solid.Tnmax_pair) and the minimum gap
distance (solid.gapmin_pair).

Looking at the convergence plot will give valuable information about the convergence
properties. There will, as a default, be one graph per Contact node in the Model Tree,
which will help you pinpoint the source of a convergence problem. You can increase
the granularity further, by selecting the Group contact variables in solver per pair check
box in the Advanced section of the settings for the Contact node.

You can also select to include information about the contact state in the solver log. To
do that, select the Add contact status to solver log check box in the Advanced section of
the settings for the Contact node. For each contact pair, messages like

69 points of 120 are now in contact.

33 points started to stick. 72 points are now sticking.

12 points started to slide. 47 points are now sliding.

will be generated for each time or parameter step. Only changes are reported.

Dependent Variables in Contact Analysis

The Contact and Friction nodes will generate a number of degrees of freedom which
depend on the settings and study type. You will see these degrees of freedom appear

CONTACT MODELING | 155

 under Dependent Variables in the solver sequence. There are two types of extra variables
created:

• Variables changed until convergence is reached during the iterations. These variables
appear in the Lumped Step nodes in the Segregated solver or in the Fully Coupled
node.
• Variables used to store the state, once the iterations have converged for a certain
time or parameter step. These variables are appear in the Previous Solution node. In
a stationary solver, the path to this node is Stationary Solver>Parametric>
Previous Solution. In a time-dependent solver it appears directly under
Time-Dependent Solver.

If you change settings in the Contact or Friction nodes after the solver
sequence has been generated, dependent variables may be added or
removed. The second case is never a problem, but when new dependent
variables are created, they are not automatically added to the groups in the
segregated solver. You may then encounter the error message
“Segregated solver steps do not involve all components.” You
can then either regenerate the solver sequence, or manually insert the
missing variables into the Lumped Step and Previous Solution nodes.

In Table 2-8 the dependent variables which can be created by the Contact or Friction
nodes are summarized. To shorten the variable names, the full scope has been
removed. As an example, the contact pressure variable for pair p1 in component comp1,
generated in the Solid Mechanics interface solid, will have the full name
comp1.solid.Tn_p1. In the table, it is shown as Tn.

TABLE 2-8: DEPENDENT VARIABLES OCCURRING IN CONTACT ANALYSIS

VARIABLE NAME DESCRIPTION EXPLANATION CREATED SOLVER NODE

Tn Contact pressure The contact Contact: Lumped

pressure in the Augmented Step
normal direction Lagrangian used
Tt Friction force The friction force Friction: Lumped
vector Augmented Step
Lagrangian used

156 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

TABLE 2-8: DEPENDENT VARIABLES OCCURRING IN CONTACT ANALYSIS

VARIABLE NAME DESCRIPTION EXPLANATION CREATED SOLVER NODE

cm_old Previous mapped The location on Friction node Previous

source the source where present Solution
coordinates this point was
located at last
converged
solution
sliptot_old Previous Total slip in this Friction: Store Previous
accumulated slip point at last accumulated slip Solution
converged selected
solution
contact_old Contact variable Non-zero if the Friction node Previous
in previous step point was in present Solution
contact at last
converged
solution within
the friction
detection
tolerance
isContact_old Contact status Non-zero if the Contact: Add Previous
variable in point was in contact status to Solution
previous step contact at last solver log
converged selected
solution
isSliding_old Sliding friction Non-zero if the Contact: Add Previous
status variable point was in a contact status to Solution
previous step sliding state at last solver log
converged selected. Friction
solution node present.
isSticking_old Sticking friction Non-zero if the Contact: Add Previous
status variable point was in a contact status to Solution
previous step sticking state at solver log
last converged selected. Friction
solution node present.
Wcnt Contact energy Elastic energy Contact: Store Segregated
density stored in normal energy variables step
contact by the selected.
penalty factor.

CONTACT MODELING | 157

 TABLE 2-8: DEPENDENT VARIABLES OCCURRING IN CONTACT ANALYSIS

VARIABLE NAME DESCRIPTION EXPLANATION CREATED SOLVER NODE

Wfric Frictional Energy dissipated Contact: Store Segregated

dissipation energy by friction energy variables step
density selected. Friction
node present.
cma_old Previous adhesive Position on the Adhesion node Previous
mapped source source where this present Solution
coordinates point was located
when adhesion
was detected.
u_max_old Maximum Maximum norm Adhesion node Previous
adhesion of adhesive layer present Solution
displacement at displacement
previous step reached until
previous step.
Also used as
indication of
adhesion
activation.

Important Contact Variables

In n this section you will find a summary of variables created in the contact feature
which can be useful in postprocessing.I
TABLE 2-9: IMPORTANT CONTACT VARIABLES

VARIABLE DESCRIPTION DEFINED IN COMMENTS

solid.gap_p1 Gap distance including Contact

offsets, contact pair
p1
solid.gapmin_p1 Minimum gap Contact
distance, contact pair
p1
solid.cnt1.Tn Contact pressure Contact
solid.Tnmax_p1 Maximum contact Contact
pressure, contact pair
p1
solid.cnt1. Friction force norm Friction
fric1.Ttnorm

158 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

TABLE 2-9: IMPORTANT CONTACT VARIABLES

VARIABLE DESCRIPTION DEFINED IN COMMENTS

solid.cnt1. Friction force, x Friction

fric1.Ttx component
solid.Wcnt_tot Total contact energy Contact Integration over
all contact
boundaries
solid.Wfric_tot Total frictional Friction ntegration over all
dissipation energy contact
boundaries

CONTACT MODELING | 159

 Springs and Dampers
The Spring Foundation and Thin Elastic Layer physics nodes supply elastic and
damping boundary conditions for domains, boundaries, edges, and points.

The features are completely analogous, with the difference that a Spring Foundation
connects the structural part on which it is acting to a fixed “ground,” while the Thin
Elastic Layer acts between two parts, either on an internal boundary or on a pair.

A Spring Foundation is most commonly used for simulating boundary conditions with
a certain flexibility, such as the soil surrounding a construction. Another important use
is for stabilizing parts that would otherwise have a rigid-body singularity. This is a
common problem in contact modeling before an assembly has actually settled. In this
case a Spring Foundation acting on the entire domain is useful because it avoids the
introduction of local forces.

A Thin Elastic Layer used as a pair condition can simulate thin layers with material
properties that differ significantly from the surrounding domains. Common
applications are gaskets and adhesives.

When a Thin Elastic Layer is applied on an interior boundary, it usually models a local
flexibility, such as a fracture zone in a geological model.

The following types of data are defined by these nodes:

• Spring Data
• Loss Factor Damping
• Viscous Damping

SPRING DATA
The elastic properties can be defined either by a spring stiffness or by a force as function
of displacement. The force as a function of displacement can be more convenient for
nonlinear springs. Each spring node has three displacement variables defined, which
can be used to describe the deformation dependency. These variables are named
<interface_name>.uspring1_<tag>, <interface_name>.uspring2_<tag>, and
<interface_name>.uspring3_<tag> for the three directions given by the local
coordinate system. In the variable names, <tag> represents the tag of the feature
defining the variable. The tag could, for example, be spf1 or tel1 for a Spring
Foundation or a Thin Elastic Layer, respectively. These variables measure the relative
extension of the spring after subtraction of any predeformation.

160 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

In addition to explicitly supplying a spring stiffness, you can choose to enter elastic
material data and the layer thickness. The spring stiffness is then computed internally,
based on an assumption of plane strain conditions.

LOSS FACTOR DAMPING

The loss factor damping adds a loss factor to the spring data above, so that the total
force exerted by the spring with loss is

f sl = ( 1 + iη )f s

where fs is the elastic spring force, and η is the loss factor.

Loss factor damping is only applicable for eigenfrequency and frequency-domain

analysis. In time-dependent analysis the loss factor is ignored.

VISCOUS DAMPING
It is also possible to add viscous damping to the Spring Foundation and Thin Elastic
Layer features. The viscous damping adds a force proportional to the velocity (or in
the case of Thin Elastic Layer: the relative velocity between the two boundaries). The
viscosity constant of the feature can be made dependent on the velocity by using the
variables named <interface_name>.vdamper1_<tag>,
<interface_name>.vdamper2_<tag>, and <interface_name>.vdamper3_<tag>,
which contain the velocities in the three local directions.

Bracket—Spring Foundation Analysis: Application Library path

Structural_Mechanics_Module/Tutorials/bracket_spring

Spring Foundation and Thin Elastic Layer

SPRINGS AND DAMPERS | 161

 Defining Multiphysics Models
The following modeling tips are about how to define multiphysics models. A good
place to start reading is in Building a COMSOL Multiphysics Model in the COMSOL
Multiphysics Reference Manual.

In this section:

• Thermal-Structural Interaction
• Acoustic-Structure Interaction
• Thermal-Electric-Structural Interaction

Thermal-Structural Interaction
The Thermal Stress Interface included with this module has a predefined one-way
coupling for thermal-structure interaction (thermal stress), which combines a Solid
Mechanics interface with a Heat Transfer interface from the Heat Transfer Module or
COMSOL Multiphysics.

By default, COMSOL Multiphysics takes advantage of the one-way coupling and

solves the problem sequentially using the segregated solver. The solution for the
temperature is separated from the stress-strain analysis which then uses the computed
temperature field from the heat transfer equation.

Using a single iteration in the segregated solver does not produce a

correct result if there are thermal properties that depend on the
displacements. Examples are when a heat source causes mechanical losses
(damping) in the material or when thermal contact is present.

162 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

Acoustic-Structure Interaction
When the Structural Mechanics Module is used together with an acoustics physics
interface from the Acoustics Module, it is possible to model a wide range of
acoustic-structure interaction problems..

There are several physics interfaces available that are documented and
described in the Acoustic-Structure Interaction Interfaces chapter in
Acoustic Module User’s Guide

• The Acoustic-Solid Interaction, Frequency Domain Interface

• The Acoustic-Solid Interaction, Transient Interface
• The Acoustic-Piezoelectric Interaction, Frequency Domain Interface
• The Acoustic-Piezoelectric Interaction, Transient Interface
• The Solid Mechanics (Elastic Waves) Interface
• The Poroelastic Waves Interface
• The Acoustic-Solid-Poroelastic Waves Interaction Interface
• The Acoustic-Poroelastic Waves Interaction Interface
• The Acoustic-Shell Interaction, Frequency Domain Interface
• The Acoustic-Shell Interaction, Transient Interface

Thermal-Electric-Structural Interaction
The Joule Heating and Thermal Expansion Interface enables
thermal-electric-structural interaction. This is a combination of three physics
interfaces: Solid Mechanics, Heat Transfer in Solids, and Electric Currents.

The thermal-electric coupling is bidirectional, with joule heating and temperature

dependent electrical properties, while the temperature coupling to the Solid
Mechanics interface is unidirectional.

By default, COMSOL Multiphysics takes advantage of the one-way coupling and

solves the problem sequentially using the segregated solver. Temperature and electric
potential are solved using a coupled approach and then the stress-strain analysis uses
the computed temperature field from the heat transfer equation.

DEFINING MULTIPHYSICS MODELS | 163

 Using a single iteration does not produce a correct result if there are
thermal properties or electrical that depend on the displacements, making
the thermal-structure part into a two-way coupling.

164 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

Thermally Coupled Problems
A wide class of structural mechanics problems are related to effects of variations in
temperature. In this section various such effects are discussed.

Temperatures can either be computed using another physics interface, usually Heat
Transfer in Solids, or directly be prescribed in the input for the various physics nodes.

For information about the predefined coupling between the Solid

Mechanics and Heat Transfer in Solids interfaces, see The Thermal Stress
Interface.

In this section:

• Temperature Dependent Material Data

• Thermal Expansion
• Constraints and Thermal Expansion
• Thermoelastic Damping

Temperature Dependent Material Data

Many material properties, such as Young’s modulus, coefficient of thermal expansion,
and yield stress, can have a significant dependence on temperature. In many cases,
materials supplied in the material libraries and databases have such dependencies
incorporated.

Materials in the COMSOL Multiphysics Reference Manual

If a material property under the Materials branch has a temperature dependence, you
have to input the temperature to be used in the Model Inputs section in the Settings
window for the node in the physics interface that references the property. It is possible
that not all aspects of a material are defined in the same node in the Model Builder tree.
For example, if a problem is run with thermal expansion and plasticity, then:

• Young’s modulus, Poisson’s ratio, and density are given in the Linear Elastic
Material node.

THERMALLY COUPLED PROBLEMS | 165

 • Yield stress and hardening function are given in the Plasticity node.
• Coefficient of thermal expansion is given in the Thermal Expansion (for materials)
subnode.

The Plasticity node is available as a subnode to the Linear Elastic Material

node when you have either the Nonlinear Structural Materials Module or
the Geomechanics Module.

For each of these nodes there is a Model Inputs section on the Settings window. Some
of these sections may be empty, specifically if none of the properties given in that node
has a temperature dependence. In general you have to supply the temperature in all the
Model Inputs sections. This can be done either by explicitly giving a temperature or by
selecting a temperature variable from another physics interface.

If there is a Temperature Coupling node under the Multiphysics branch,

then you cannot change the temperature input under Model Inputs. The
temperature is automatically taken from the connected Heat Transfer in
Solids interface.

DENSITY
All Structural Mechanics interfaces are formulated on the material frame. This means
that the equations of motion are formulated for a certain volume in its initial
configuration.

The preservation of mass requires that the mass density is constant. In a

structural mechanics problem this means you must not use a temperature
dependent density. All effects of volume change with temperature are
incorporated through the thermal expansion effects.

Thermal Expansion
As the temperature changes, most materials react by a change of volume. For a
constrained structure, the stresses that evolve even with moderate temperature changes
can be considerable. The volume change can be is represented a thermal strain εth,
which produces stress-free deformations. For a linear elastic material, the constitutive
law is

166 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

 σ = C : ( ε – ε th )

In the computations, the thermal expansion appears as a load, even though it formally
is a part of the constitutive relations.

• You can include thermal expansion in a model either by adding a

Thermal Expansion (for materials) subnode to the chosen material, or
by using the Thermal Expansion (Multiphysics Coupling) and
Temperature Coupling (described in the COMSOL Multiphysics
Reference Manual).
• When the temperature distribution is computed by another physics
interface (often Heat Transfer in Solids), it is a good practice to use a
discretization which in the heat transfer interface is one order lower
than what is used in the structural mechanics interface. The thermal
strains are proportional to the temperature, while the strains in the
structural problem are computed as derivatives of the displacements, so
the approximations will then be consistent. This is automatically taken
care of when you use the The Thermal Stress Interface or the The Joule
Heating and Thermal Expansion Interface.

TE M P E R A T U R E D E P E N D E N C E O F T H E T H E R M A L E X P A N S I O N
COEFFICIENT
When performing an analysis over a larger range of temperatures, you often need to
take into account the temperature dependence in the values of the thermal expansion
coefficient itself.

As long as you are using materials from the COMSOL Material Library, everything is
handled internally. When you want to enter data from your own measurements or from
the literature, you do however need to be aware of some details in the definitions used.

Tangent or Secant Data

Thermal expansion coefficients can appear in two forms: tangent and secant.

The tangent form states that the increment in length is

dL
-------- = α t ( T )dT (2-24)
L

where αt is the tangential thermal expansion coefficient. This form, which is the
thermodynamic definition, is conceptually simple, because αt is uniquely defined at

THERMALLY COUPLED PROBLEMS | 167

 each temperature. It is however less convenient to use in practice, since an integration
is required for determining the actual change in length for a finite temperature
difference.

The secant formulation, which is the default in COMSOL Multiphysics, is often used
in engineering:

ΔL
-------- = α ( T )ΔT
L0

In the secant formulation, the actual values of α will however depend on the choice of
reference temperature, Tref, at which the material has the reference length L0:

ΔL
-------- = α ( T,T ref ) ( T – T ref )
L0

Converting from Tangent Form to Secant Form

Equation 2-24 can be integrated, giving

ln  ------ =
L
 L 0  α t ( τ )dτ = I ( T,T ref ) (2-25)
T ref

Thus,

ΔL L I ( T,T ref )
-------- = ------ –1 = e –1
L0 L0

giving the secant thermal expansion coefficient as

I ( T,T ref )
e –1
α ( T,T ref ) = ---------------------------------
( T – T ref )

For most materials and temperature ranges I ( T,T ref ) « 1 , which makes it possible to
approximate with the simpler expression

I ( T,T ref )
α ( T,T ref ) = ------------------------- (2-26)
( T – T ref )

168 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

If you have access to tangent data, you can choose between two different methods for
using them in COMSOL Multiphysics:

• In most of the physics interfaces, you can enter tangent data directly by selecting
Tangent coefficient of thermal expansion in the settings for Thermal Expansion. When
using this option, a numerical integration of Equation 2-25 will be performed each
time the thermal strain is used. This will have a negative impact on the performance,
when compared to using a secant coefficient of thermal expansion.
• Pre-compute the expression in Equation 2-24 externally for the intended range of
temperatures. This can for example be done in a spreadsheet program. Enter the
computed result as a function, which is then used as any other secant temperature
dependent thermal expansion coefficient.

Thermal Expansion Coefficient Dependence on Reference Temperature

Let αm(T) be the temperature-dependent function that represents the measured values
of the secant thermal expansion coefficient. The change in length of a sample at a given
temperature T with respect to the sample’s original length at a temperature Tm is called
dilation.

Note that by definition, the dilation at T=Tm is zero, so Tm denotes the strain-free
state of the material as far as the measured values of αm(T) is concerned. Denote the
length of the sample at a temperature T as L(T) and the strain-free length as
L0 = L(Tm). The dilation can be then expressed as L ( T ) – L ( T m ) . Using the
definition of the secant coefficient of thermal expansion, L(T) can be written as:

L ( T ) = [ 1 + α m ( T ) ( T – T m ) ]L ( T m ) (2-27)

When using the measured data, it is possible that the strain-free state occurs at a
temperature Tref which differs from Tm. The dilation at any temperature T would then
be defined as L ( T ) – L ( T ref ) , where L(Tref) can be written as.

L ( T ref ) = [ 1 + α m ( T ref ) ( T ref – T m ) ]L ( T m ) (2-28)

As a result of this shift in the strain-free temperature, it is necessary to redefine the

thermal expansion coefficient so that L(T) and L(Tref) can be related using
Equation 2-27 but with Tm replaced by Tref.

L ( T ) = [ 1 + α r ( T ) ( T – ( T ref ) ) ]L ( ( T ref ) ) (2-29)

Here αr(T) is the redefined thermal expansion coefficient, based on Tref. It can be
derived from the relations above. Using Equation 2-27 and Equation 2-29 there are
two ways of writing the current length L(T), so that

THERMALLY COUPLED PROBLEMS | 169

 [ 1 + α r ( T ) ( T – ( T ref ) ) ]L ( ( T ref ) ) = [ 1 + α m ( T ) ( T – T m ) ]L ( T m ) (2-30)

Equation 2-28 makes it is possible to eliminate L(Tref) and L(Tm) from

Equation 2-30, and after some algebra αr(T) can then be written as

α m ( T ) – α m ( T ref )
α m ( T ) + ( T ref – T m ) -----------------------------------------------
T – T ref
α r ( T ) = ------------------------------------------------------------------------------------------------------ (2-31)
1 + α m ( T ref ) ( T ref – T m )

Representation in COMSOL
Most materials listed in COMSOL’s material libraries and databases contain a function
for the measured temperature-dependent thermal expansion coefficient curve. You can
find this from the Materials branch, as shown in Figure 2-16. The Piecewise function
named alpha_solid_1 is the measured thermal expansion coefficient αm(T).

Using Functions in Materials in the COMSOL Multiphysics Reference

Manual

The Material Contents section in Figure 2-16 shows the material property alpha,
which is the redefined thermal expansion coefficient αr(T). The complete expression
for alpha is as follows:

(alpha_solid_1(T[1/K])[1/K]+(Tempref-293[K])*
if(abs(T-Tempref)>1e-3,(alpha_solid_1(T[1/K])[1/K]
-alpha_solid_1(Tempref[1/K])[1/K])/(T-Tempref),
d(alpha_solid_1(T[1/K]),T)[1/K]))/
(1+alpha_solid_1(Tempref[1/K])[1/K]*(Tempref-293[K]))

This is essentially Equation 2-31, but with a small modification to avoid problems if
T=Tref.

170 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

 αm(T)

αr(T)

Figure 2-16: An example in COMSOL Multiphysics showing the Materials branch and
where to find the temperature-dependent thermal expansion coefficient.

In the definition of alpha (to be more specific: <material_tag>.def.alpha) in

COMSOL, Tm is set as 293 K and Tref is obtained from the variable Tempref which
typically fetches its value from the physics interface.

Using Your Own Material Data

If you use our own material data in COMSOL Multiphysics (via an interpolation or
any other function), you can still copy-paste the built-in expression for alpha into your
New Material. You just need to:

• Replace the function name alpha_solid_1(T[1/K])[1/K], with the function

name that you have assigned to the temperature-dependent measured thermal
expansion coefficient and use the correct temperature units. This is particularly

THERMALLY COUPLED PROBLEMS | 171

 important if the measured data used the Fahrenheit scale (denoted by [degF] in
COMSOL).
• Ensure that the value of Tm is changed from 293 K to the actual value of
temperature that was used as the strain-free temperature to compute αm(T) in the
function that you created.

User-Defined Materials and Libraries in the COMSOL Multiphysics

Reference Manual

Constraints and Thermal Expansion

When a structure is undergoes thermal expansion, the fact that the motion is restricted
by constraints will in general cause stresses. There are three types of such effects:

• The global expansion is restricted by constraints at different locations, so that

internal forces are introduced throughout the structure.
• At a boundary which is constrained, local stresses can appear if the boundary is not
free to expand in the tangential direction.
• Internally, the same type of local constraint effects will be caused by rigid objects,
such as Rigid Domains.

In many cases, not only the structure which actually is modeled deforms due to the
changes in temperature, but also the surroundings (which are approximated by
constraints) will deform. You can take this effect into account by adding a Thermal
Expansion subnode to the constraints. The constraints will then provide an extra
displacement based on a given temperature field. For thermal strains which have a
simple variation in space (for example linear temperature variations), it is possible to
completely offset the constraint stresses using this method. For more general cases, the
stresses caused by the constraint can be significantly reduced.

The thermal expansions of the constraints are independent of that of the material in
the adjacent domain, so that the surrounding structure can be made from another
material, or have a different temperature distribution.

172 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

You can however also inherit temperature and thermal expansion coefficient from the
domain being constrained. This is useful for the common case that the temperature
and materials are the same over the modeled structure body and its surroundings.

The spatial variation of the temperature and coefficient of thermal

expansion must be explicit functions of the material frame coordinates. It
is not possible to use a computed temperature distribution for the thermal
expansion of the constraints.

• Thermal Expansion (for constraints)

• Thermal Expansion of Constraints in the theory section.

For an example showing how to relive the stress at constraints in a heated

structure, see Thermal Expansion in a MEMS Device: Application
Library path Structural_Mechanics_Module/Thermal-Structure Interaction/
thermal_expansion.

Thermoelastic Damping
In most engineering problems, the coupling between temperatures and structural
problems can be considered as unidirectional. Only the thermal expansion is taken into
account.

The opposite effect, where changes in stress cause heat generation may be important
in small structures vibrating at high frequencies. The Thermoelasticity interface,
available with the MEMS module, is designed for analyzing such problems.

It is also possible to take this effect into account by adding the Thermoelastic Damping
node to the Heat Transfer in Solids interface. When you add a Thermal Expansion node
to a material in the Solid Mechanics interface, the heat source term is computed and
made available to the Heat Transfer in Solids interface.

When you add a Thermal Expansion node under the Multiphysics branch, it is possible
to select whether the thermoelastic damping effect should be taken into account or
not. The heat source contribution is then included automatically without adding any
data in the heat transfer interface.

THERMALLY COUPLED PROBLEMS | 173

 In either case, the heat source term is only present when Structural Transient Behavior
is set to Include inertial terms.

See also

• Entropy and Thermoelasticity

• Thermal Expansion (for materials)
• Thermal Expansion (Multiphysics Coupling)

174 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

Computing Mass Properties
In structural mechanics analysis, especially when modeling dynamic problems, the
mass properties of a structure or its part can be an important aspect of the design. To
compute such mass properties, you can use Mass Properties node, which can be added
under Component>Definitions. There, you can select the geometry domains to be
included into the computations, and select physics interfaces that will define the mass
properties. You can add and configure several mass property contributions if needed.

COMPUTING MASS PROPERTIES | 175

 Volume, mass, center of mass, and moments of inertia will be computed. They will
become available as predefined variables which you can find in the equation view under
the corresponding Mass Properties node.

Besides for postprocessing purposes, you can also use these variables in any
user-defined expressions, user inputs, and in optimization criteria.

Structural mechanics interfaces contribute to the mass properties in several ways:

• All material models, including Rigid Domain, define mass density contributions.
• Added Mass nodes for all geometric entity levels can also contribute with added mass
density. It is possible to suppress the contribution from an Added Mass node by using
the Exclude contribution check box in the Frame Acceleration Forces section.
• Point Mass nodes will contribute both with mass and with the specified mass moment
pf inertia. It is possible to suppress the contribution from an Point Mass node by
using the Exclude contribution check box in the Frame Acceleration Forces section.
• The Hygroscopic Swelling nodes, which can be added to most material models for all
structural elements, can use the moisture content as an extra mass density
contribution.

176 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

• Rigid Domain and Rigid Connector can contribute directly to the total mass properties
via their Mass and Moment of Inertia subnodes

• Structural elements like beams and shells take their true geometrical dimensions into
account when contributing to the mass properties. As an example, a beam
contributes to the rotational inertial around its axis, even though the geometrical
model is only an edge. The beam cross section properties are used to compute the
data.

The mass properties can be computed on both initial geometry (material frame) and
deformed geometry (spatial frame). The results may differ considerably in case of large
deformations. To compute the results in the undeformed geometry, you do not have
to perform the whole analysis, it is sufficient to chose Get Initial Values under the Study
node. To obtain the mass properties in the deformed configuration, you need to the
full analysis, so that the displacement results are available.

In the COMSOL Multiphysics Reference Manual:

• Mass Properties
• Studies and Solvers
• Derived Values and Tables

COMPUTING MASS PROPERTIES | 177

 Pre-tensioned Bolts
The information about pre-tensioned bolts is applicable if your license
includes the Structural Mechanics Module.

Bolted joints are common in mechanical and civil engineering structures. If you are
interested in analyzing the details of a bolted joint, the pre-stress in the bolt must be
taken into account in order to correctly capture the behavior under service loads. The
Bolt Pre-Tension functionality in COMSOL Multiphysics designed to simplify such
analyses.

During mounting, a bolt is tightened to a certain pre-stress. The mounting of the bolt
is, in general, accompanied by deformations of the surrounding structure. In the
subsequent service, the force in the bolt can then change due to external loads.

MODELING THE BOLTS

You must use a specific modeling technique in order to use a bolt in a prestress analysis.

1 Model each bolt using solid elements in 3D. Usually it is most efficient to add the
predefined bolt geometries from the Part Libraries.
2 Make sure that there is at least one internal boundary perpendicular to the bolt axis
somewhere in the shank. In the following, this boundary is referred to as the slit
boundary (Figure 2-17). The slit boundary can be composed of several boundaries
in the geometry.
3 If you are using bolts from the Part Libraries, a slit boundary is predefined, and has
the selection name Pre-tension cut. In order to make this boundary selection
visible from the physics interface, select its Keep check box in the Boundary Selections
section of the settings for the part instance (Figure 2-18).
4 If needed, add contact conditions between the bolt head and the component, and
between different components clamped by the bolt.
5 Add a Bolt Pre-Tension node, in which the pre-tension force or stress is prescribed
for a set of bolts with the same data.

178 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

6 For each bolt having the identical pre-tension stress, add a Bolt Selection subnode
where the slit boundary is selected.

When a bolt is located in a symmetry plane (so that only half the bolt is
modeled), and Automatic symmetry detection is selected in the Bolt
Selection node, the given pre-tension force is interpreted as the force in
the whole bolt, not as the force in the modeled half. This makes it possible
to use the same Bolt Pre-Tension node for a set of similar bolts where some
of them are located in symmetry planes.

Figure 2-17: Example of a bolted joint.

PRE-TENSIONED BOLTS | 179

 Figure 2-18: Getting access to the slit boundary selection for a bolt from the Part Libraries.

Each bolt defined in the Bolt Selection node has a single degree of freedom called
predeformation, d. At the slit boundary, the two sides of the bolt are disconnected so
that the displacements over it can be discontinuous. The discontinuity is represented
by:

u u = u d – dn

Here the subscript u denotes the upside of the slit boundary, and d denotes the
downside. n is the normal pointing out from the downside. The sign has been selected
so that d gets a positive value when the bolt force is tensile.

The axial force in the bolt is thus caused by a small overlap between the two sides of
the slit boundary. It is computed as the reaction force belonging to the degree of
freedom d.

SETTING UP THE STUDY STEPS

In an analysis of prestressed bolts, you have to use two or more separate study steps.
They can be part of a single study, or be placed in different studies. The first study step,
in which the bolt pre-stress is prescribed, simulates the mounting process. If you would
only use a force to load the bolt, for example as an initial stress, the resulting stress in
the bolt would be less than the intended, due to the compression of the material

180 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

around the bolt. The prestress step ensures that the bolts have the intended prestress,
irrespective of the flexibility of the surrounding structure and their interaction.

In the subsequent studies the bolt force is allowed to change, while keeping the
extension of the bolt, as caused by the first study, fixed. The procedure to do this is as
follows:

1 Run the study step for the mounting simulation. The predefined study type Bolt
Pre-Tension is designed for this.
2 Add one or more studies or study steps to analyze the effects of the service loads.
3 Since the pre-tension degrees of freedom are not solved for in the service load study
steps, they must obtain their values from the pre-tension study step. If the study
steps are sequential within the same study, this action is not needed, since the default
then is to inherit values from the previous study step. For other cases, go to the
Values of Dependent Variables section of the study step, set Values of variables not
solved for to User controlled, and then select the pre-tension study step.

PRE-TENSIONED BOLTS | 181

 It is important to make sure that you only solve for the bolt
predeformation degrees of freedom in the pre-tension study step, and not
when analyzing the service loads.

If you use the Bolt Pre-Tension study type for the pre-tension study step,
and any other study type to analyze the service loads, the solvers are
automatically set up to handle this. The Bolt Pre-Tension study type is
actually a special case of a Stationary study step, with the sole purpose of
activating the predeformation degrees of freedom. These degrees of
freedom are by default not solved for in any other study type.

If you however set up your studies manually, the information below is

useful. Also, in versions prior to 5.3, this automatic mechanism was not
available, so in older models the studies were always set up manually.

You enable or disable the solution of individual degrees of freedom under

the Dependent Variables node for a certain study step in the solver
sequence. If required, begin by clicking Show Default Solver in the study
node or in the Solver Configurations node of the study. Then move to the
Dependent Variables node, and in the General section, set Defined by study
step to User defined.

You can now go to the node for each predeformation degree of freedom
below Dependent Variables and adjust the state of the Solve for this state
check box.

For more information, see also Dependent Variables and Studies and
Solvers in the COMSOL Multiphysics Reference Manual.

RESULTS
The results in a bolt do not belong to any part of the geometry, but are global
variables. To access the result from a certain bolt, a full scope of the type
<interface>.<Bolt Pre-Tension tag>.<Bolt Selection tag>.<variable>

182 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

must be used. An example could be solid.pblt1.sblt1.F_bolt. The bolt results
are summarized in the table below.
TABLE 2-10: BOLT VARIABLES

VARIABLE DESCRIPTION

d_pre Predeformation
F_bolt Axial force in the bolt
F_shear Shear force in the bolt

If you place a bolt in a symmetry plane, that only half of the bolt is modeled, this will
automatically be detected. The results are reported for the whole bolt, not for the
symmetric half.

Studies and Solvers in the COMSOL Multiphysics Reference Manual

Prestressed Bolts in a Tube Connection: Application Library path

Structural_Mechanics_Module/Contact_and_Friction/tube_connection

PRE-TENSIONED BOLTS | 183

 Stress Linearization
The information about stress linearization is applicable if your license
includes the Structural Mechanics Module.

Stress linearization is a procedure in which the stress distribution along a line through
the thickness in a solid is approximated with an equivalent linear stress distribution,
similar to what would be the result of an analysis using shell theory. The line is
commonly referred to a stress classification line, SCL. This type of evaluation is
common is the analysis of pressure vessels. It is also useful for the dimensioning of
reinforcements for concrete structures, and for fatigue analysis of welds.

To perform a stress linearization, you add one Stress Linearization node for each SCL.

ORIENTATION OF STRESS COMPONENTS

The stresses along the SCL are represented in a local orthonormal coordinate system,
x1-x2-x3. The x1 direction is oriented along the SCL, so it is mainly the stresses in the
second and third directions which are of interest.

• In 3D, you must specify the x2 direction, and thus implicitly the x3 direction. You
specify the orientation either by selecting a point in the x1-x2 plane, or by defining
an orientation vector in an approximate x2 direction. In either case, the actual x2
direction is chosen so that it is perpendicular to the SCL, and lies in the plane you
have specified. The x3 orientation is then taken as perpendicular to x1 and x2. As
long as you are only interested in a stress intensity, the choice of orientation is
arbitrary.
• In 2D, the x3 direction is the out-of-plane direction, and the x2 direction is
perpendicular to the SCL in the X-Y plane.
• In 2D axial symmetry, the x3 direction is the azimuthal direction, and the x2
direction is perpendicular to the SCL in the R-Z plane.

CREATING THE STRESS CLASSIFICATION LINE

The most straightforward way to create an SCL, is to include it in the geometry, and
then select it in a Stress Linearization node.

184 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

Figure 2-19: Four stress classification lines in a transition region at a pressure vessel
nozzle.

There are however some situations where the direct approach is less convenient:

• When the introduction of the SCL in the geometry must be done after the analysis,
since the locations of critical regions were not obvious when setting up the initial
analysis. It is of course possible to add new edges and re-run the analysis, but this
may not be a good solutions if the analysis time is long.
• When the introduction of the edges for the SCL makes the meshing more difficult.
It may for example not any longer be possible to use swept meshes, or the mesh
quality is reduced in critical regions.

In this case, a possible solution is to operate on a copy of the geometry in another

component. It is then possible to use submodeling, that is to study only a local region,
with its boundaries have displacements controlled by the solution from the larger
model.

Below is an outline of the steps you need to take for this approach.

1 In the original component (assume that its tag is comp1), add a General Extrusion
operator. Set Source frame to Material. You can name the operator, but in the

STRESS LINEARIZATION | 185

 following description, the default name genext1 is assumed. This operator will be
used for mapping results to the second component.
2 Add a new component of the same dimension as the one in which you performed
the original analysis.
3 Add the geometry, using for example Insert Sequence or Import. to the new
component (comp2).
4 Add the lines you are going to use for stress linearization to the imported geometry.
5 Add a Solid Mechanics interface, and select only the domains which contain the new
lines. You can even change the geometry so that you cut out parts of the domains.
6 Add a Prescribed Displacement node having domain selection. Select All domains.
Prescribe the displacement in all directions to be the same as in the original model
with expressions like comp1.genext1(u) for Prescribed in x direction.
7 If the original analysis contains inelastic strains, such as thermal expansions, these
must also be included. You can do this either by adding a Thermal Expansion node
with appropriate settings, or by explicitly importing the inelastic strains using an
External Strain node. In the latter case, you would use expressions like
comp1.genext1(solid.eth11) for the strain components.

8 Add Stress Linearization nodes for the new linearization lines.

9 Add the materials that were used in comp1. The most efficient approach is to add
them under Global Definitions, and link to the same material definitions from both
components.
10 Create a mesh for the domains in comp2 which you are solving for. It is only the
mesh close to the new edges to which you need to pay any attention.
11 Add a new stationary study, in which you solve only for the new Solid Mechanics
interface in comp2.
12 In the settings for the new study set Values of variables not solved for to point to the
solution from which you want to pick the results in comp1. You can also add an
Auxiliary sweep, if the original analysis contains results for several parameters or time
steps.
13 Run the new study.

STUDIES AND SOLUTIONS

Stress linearization is a pure postprocessing operation. The a Stress Linearization node
will only create a number of variables, which can be evaluated under Results. It is thus
possible to add such nodes after the main analysis has been performed. In order to

186 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

make the new variables available for postprocessing, you must then run an Update
Solution.

RESULTS
When you have included one or more Stress Linearization nodes in a model, a number
of data sets and an extra default plot are generated.

One edge data set is created for each SCL. These data sets are named Linearization
Line <n>, where n is an integer number.

Figure 2-20: Generated data sets in a model with four SCL.

The default plot contains graphs for the 22 component of the actual stress, the
membrane stress, and the linearized stress. The first Linearization Line data set is
selected. By changing edge date set in the plot group, you can easily move between the
different stress classification lines.

Figure 2-21: Default plot along a stress classification line.

STRESS LINEARIZATION | 187

 VAR IA BL E N A MES
Each Stress Linearization node adds a number of variables. Many of these variables exist
with two different scopes, physics scope and feature scope. The physics scope is the
name of the physics interface, having the default value ‘solid’. The feature scope
contains also the tag of the Stress Linearization node, for example ‘sl1’.

As an example, the variable solid.Sm22 and the variable solid.sl1.Sm22 have the
same value. The variables with physics scope makes it more convenient to create
expressions using postprocessing. You could for example make a line plot of
solid.sb22, and get all edges having a stress linearization colored by their individual
results.
TABLE 2-11: VARIABLE FOR STRESS LINEARIZATION

VARIABLE DESCRIPTION VARIABLE IN COMMENT SCOPE

THEORY SECTION

Sllij Stress tensor in σ ij ij = 11, 12, 13, 22, Physics,

local coordinate 23, 33 Feature
system
Smij Membrane stress σ m, ij ij = 11, 12, 13, 22, Physics,
23, 33 Feature
Sbmaxij Maximum bending σ b ( max ), ij ij = 11, 12, 13, 22, Feature
stress 23, 33
Sbij Bending Stress σ b, ij ij = 11, 12, 13, 22, Physics,
23, 33 Feature
Smbij Membrane + σ mb, ij ij = 11, 12, 13, 22, Physics,
bending stress 23, 33 Feature
Spsij Peak stress, start σ p ( start ), ij ij = 11, 12, 13, 22, Feature
point 23, 33
Speij Peak stress, start σ p ( end ), ij ij = 11, 12, 13, 22, Feature
point 23, 33
SIm Stress intensity, σ int Physics,
membrane Feature
SImbs Stress intensity, σ int Feature
membrane +
bending, start
point
SIbme Stress intensity, σ int Feature
membrane +
bending, end
point

188 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

TABLE 2-11: VARIABLE FOR STRESS LINEARIZATION

VARIABLE DESCRIPTION VARIABLE IN COMMENT SCOPE

THEORY SECTION

SImb Stress intensity, max(SImbs, SImbe) Physics,

membrane + Feature
bending,
Nij Local in-plane N ij ij = 22, 23, 33 Feature
force
Mij Local bending M ij ij = 22, 23, 33 Feature
moment
Qi Local out-of-plane i = 2, 3 Feature
shear force

lengthtot Length of SCL L Feature

arclength Coordinate along X1 Feature
SCL

Stress Linearization in the Structural Mechanics Theory Chapter.

STRESS LINEARIZATION | 189

 Solver Settings
COMSOL Multiphysics includes many solvers and solver settings. To make it easier to
use a suitable solver and its associated solver parameters, the physics interfaces have
different default settings based on the study type and features used. In some situations,
the default settings may need to be changed. This section helps you to select a solver
and its settings to solve structural mechanics and related multiphysics problems.

In this section:

• Symmetric Matrices
• Selecting Iterative Solvers
• Specifying Tolerances and Scaling for the Solution Components

Studies and Solvers in the COMSOL Multiphysics Reference Manual

If you make changes to the physics, this will not be reflected in the solver
settings unless you regenerate the solver sequence.

Symmetric Matrices
Use the Matrix symmetry list (see the General section on the Settings window for
Advanced for a solver node such as Stationary Solver). There you can explicitly state
whether the assembled matrices (stiffness matrix, mass matrix) resulting from the
compiled equations are symmetric or not.

Normally the matrices from a single-physics structural mechanics problem are

symmetric, but there are exceptions, including the following cases:

• Multiphysics models solving for several physics simultaneously, for example, heat
transfer and structural mechanics. Solving for several Structural Mechanics
interfaces, such as shells combined with beams, does not create unsymmetrical
matrices.

190 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

• Contact with friction.
• Elastoplastic analysis

Advanced and Stationary Solver in the COMSOL Multiphysics Reference

Manual

One of the benefits of using the symmetric solvers is that they use less memory and are
faster. The default option is Automatic, which means the solver automatically detects if
the system is symmetric or not. Some solvers do not support symmetric matrices and
always solve the full system regardless of symmetry.

Selecting the Symmetric option for a model with unsymmetric matrices

may lead to incorrect results. For a nonlinear problem with only a weak
unsymmetry, it may still be useful, since the faster solution of the
symmetric problem may offset the lower convergence rate. This is for
example the case for contact problems with a low coefficient of friction.

Complex matrices can be unsymmetric, symmetric, or Hermitian. Hermitian matrices

do not appear in structural mechanics problems.

Selecting the Hermitian option for a model with complex-valued

symmetric matrices produces incorrect results.

Selecting Iterative Solvers

The default solver for structural mechanics is the MUMPS direct solver in both 2D and
3D. For large 3D problems (several hundred thousands or millions of degrees of
freedom) it is beneficial to use iterative solvers when possible to save time and memory.
The drawback is that they are more sensitive and might not converge if the mesh
quality is low. The iterative solvers also have more options than the direct solvers.

For Stationary, Time Dependent studies and Frequency Domain studies in 3D, a
GMRES iterative solver is preconfigured and available as an alternative solver
suggestion for solid mechanics models.

If the model is set up using quadratic (default) or higher order elements for the
displacement dependent variable, this GMRES solver will use geometric multigrid
(GMG) as the preconditioner.

SOLVER SETTINGS | 191

 For slender geometries, changing to SOR Line as presmoother and postsmoother can
give better results compared to SOR that is the default for GMG preconditioner.

For models using linear elements for the displacement dependent variable, the
preconditioner will be changed to smoothed aggregated algebraic multigrid (AMG).
This is to avoid remeshing when creating the discretization on the coarse level. Note
that you need to manually regenerate the solver sequence after you change the element
order if you want to make use of such change in the predefined iterative solver
configuration.

• For eigenfrequency/eigenvalue studies, use the default direct solver (MUMPS).

Specifying a shift frequency greater than the lowest eigenfrequency results

in indefinite matrices.

Studies and Solvers in the COMSOL Multiphysics Reference Manual

Solver Settings for Viscoelasticity, Creep, and Viscoplasticity

Viscoelastivity, creep, and viscoplasticity are time-dependent phenomena. The time
scale is however often such that inertial effects can be ignored. When that is the case,
you can modify the solver settings to improve the performance of the time-dependent
analysis.

PHYSICS INTERFACE SETTINGS

In the Model Builder, click the Solid Mechanics node. In the Settings window, under
Structural Transient Behavior, select Quasi-static to treat the elastic behavior as
quasi-static (with no mass effects; that is, no second-order time derivatives for the
displacement variables). Selecting this option gives a more efficient solution for
problems where the variation in time is slow when compared to the natural frequencies
of the system since no mass matrix will be created.

SOLVER SETTINGS
When Quasi-static is selected on the physics interface Settings window, the automatic
solver suggestion changes the method for the Time Stepping from Generalized alpha to
BDF.

192 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

For a Fully Coupled node (or Segregated node for multiphysics problems), the default
Nonlinear method under Method and Termination is Automatic (Newton). To get a faster
computation time when the effective strain rate is low or moderate, select Constant
(Newton) as the Nonlinear method instead.

In the COMSOL Multiphysics Reference Manual:

• Studies and Solvers

• Fully Coupled and Segregated

Specifying Tolerances and Scaling for the Solution Components

The absolute-tolerance parameters used for time-dependent studies are problem
specific. By default, the absolute tolerance is applied to scaled variables, with the
default value being 0.001 for all solution components.

The default scaling for the displacement components is based on the size of the
geometry in the model, and certain reasonable scales are used for the pressure and
contact force variables, if any. You are encouraged to change these scales as soon as
better values are known or can be guessed or estimated from the applied forces, yield
stress, reaction forces, maximum von Mises stress, or similar. The same suggestion
applies to the displacement scale, which can be estimated easily if the problem is
displacement controlled. This approach can significantly improve the robustness of the
solution. The variable scaling is accessed under Dependent Variables in the solver
sequence. The scales need to be entered using the main unit system in the model.

Solver Settings
In many situations, the default in COMSOL when having several physics interfaces is
to generate a solver sequence with a segregated solver. When several structural
mechanics interfaces are present, it can happen that the degrees of freedom are placed
in different segregated steps by the default solver generation. It is, however, not
possible to solve a model where the structural mechanics degrees of freedom are placed
in different segregated groups, so in this case you must modify the solver settings.

• If the model only consists of structural mechanics interfaces, the best option is
usually to replace the segregated solver with a fully coupled solver.
• If the model contains contact conditions where the augmented Lagrangian method
is used, then a segregated solver must be used. Place all structural mechanics degrees

SOLVER SETTINGS | 193

 of freedom except the contact variables in one segregated step. The contact variables
should remain in the lumped step.
• If there are other types of physics interfaces being solved in the same study, then the
segregated solver should usually be kept. Make sure that all structural mechanics
degrees of freedom except contact variables are solved in one segregated step.

In the COMSOL Multiphysics Reference Manual:

• Solution Operation Nodes and Solvers

• About the Stationary Solver

194 | CHAPTER 2: STRUCTURAL MECHANICS MODELING

Part Libraries
The Part Libraries can be used to store and access a set of standard geometric entities.
It is possible to create custom user-defined geometry libraries or to use built-in
collections of geometric entities that are available with many add-on modules for
COMSOL Multiphysics.

A dedicated part library for the Structural Mechanics Module is included. The library
consists of two main parts:

• Bolts, nuts, and washers. These are parts intended for quick modeling of fasteners.
The bolts geometries are prepared for use with the Pre-tensioned Bolts
functionality.
• Beam cross sections. These are 2D geometries, primarily intended for use in The
Beam Cross Section Interface. The library includes geometries for all standard beam
cross sections according to European and US standards.

All entities in the built-in the Part Libraries are fully parameterized, making them easy
to use as parts in large-scale industrial models.

Part Libraries in the COMSOL Multiphysics Reference Manual

PART LIBRARIES | 195

196 | CHAPTER 2: STRUCTURAL MECHANICS MODELING
 3

Structural Mechanics Theory

This chapter contains the theory behind the implementation of the structural
mechanics functionality in COMSOL Multiphysics. It is assumed that the reader
has a basic knowledge of solid mechanics.

In this chapter:

• Solid Mechanics Theory

• Frames and Coordinate Systems
• Analysis of Deformation
• Stresses
• Material Models
• Formulation of the Equilibrium Equations
• Damping
• Loads and Boundary Conditions
• Stress Linearization
• Energy Quantities

197
 Solid Mechanics Theory
Introduction
In the following, the theory for the Solid Mechanics interface is described. To a large
extent, this theory covers other structural mechanics physics interfaces, such as Shell
and Beam, which are included with the Structural Mechanics Module. For these other
interfaces, only the details which are specific to a certain interface are described its
documentation.

• Theory for Shell and Plate Interfaces

• Theory for the Membrane Interface
• Theory for the Beam Interface
• Theory for the Beam Cross Section Interface
• Theory for the Truss Interface

TE N S O R N O T A T I O N
Some of the theory is developed using tensor notation. In most cases, explicit index
notation is avoided. This means that the order of a tensor usually must be understood
from the context. As an example, Hooke’s law for linear elasticity is usually written like

σ = C :ε

Here, the stress tensor σ and the strain tensor ε are second-order tensors, while the
constitutive tensor C is a fourth-order tensor. The ‘:’ symbol means a contraction over
two indices. In a notation where the indices are shown, the same equation would read

σ ij = C ijkl ε kl

where the Einstein summation convention has been used as a shorthand for

3 3

σ ij =   Cijkl εkl
k = 1l = 1

In a few cases, non-orthonormal coordinate systems must be considered. It is then

necessary to keep track of the covariance and contravariance properties of tensors. In
such a case, Hooke’s law is written

198 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

 ij ijkl
σ = C ε kl

The stress and constitutive tensors have contravariant components, while the strain
tensor has covariant components.

SOLID MECHANICS THEORY | 199

 Frames and Coordinate Systems
In this section:

• Material and Spatial Coordinates

• Coordinate Systems

Material and Spatial Coordinates

The Solid Mechanics interface, through its equations, describes the motion and
deformation of solid objects in a 2- or 3-dimensional space. In COMSOL Multiphysics
terminology, this physical space is known as the spatial frame, and positions in the
physical space are identified by lowercase spatial coordinate variables x, y, and z (or r,
ϕ , and z in axisymmetric components).

Continuum mechanics theory also makes use of a second set of coordinates, known as
material (or reference) coordinates. These are normally denoted by uppercase
variables X, Y, and Z (or R, Φ, and Z) and are used to label material particles. Any
material particle is uniquely identified by its position in some given initial or reference
configuration. As long as the solid stays in this configuration, material and spatial
coordinates of every particle coincide and displacements are zero by definition.

When the solid object deforms due to external or internal forces and constraints, each
material particle keeps its material coordinates X (bold font is used to denote
coordinate vectors), while its spatial coordinates change with time and applied forces
such that it follows a path

x = x ( X, t ) = X + u ( X, t ) (3-1)

in space. Because the material coordinates are constant, the current spatial position is
uniquely determined by the displacement vector u, pointing from the reference
position to the current position. The global Cartesian components of this displacement
vector in the spatial frame, by default called u, v, and w, are the primary dependent
variables in the Solid Mechanics interface.

By default, the Solid Mechanics interface uses the calculated displacement and
Equation 3-1 to define the difference between spatial coordinates x and material
coordinates X. This means the material coordinates relate to the original geometry,
while the spatial coordinates are solution dependent.

200 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

Material coordinate variables X, Y, and Z must be used in coordinate-dependent
expressions that refer to positions in the original geometry, for example, for material
properties that are supposed to follow the material during deformation. On the other
hand, quantities that have a coordinate dependence in physical space — for example, a
spatially varying electromagnetic field acting as a force on the solid — must be
described using spatial coordinate variables x, y, and z.

In a geometrically linear analysis, no difference is made between the two

coordinate systems. For this case, the material and spatial coordinates
coincide. This may seem inconsistent with equation Equation 3-1 but
ensures linearity for problems that are expected to be linear. It is then, for
example, equivalent to choose a coordinate system related to the material
frame or one related to the spatial frame. In a geometrically nonlinear
analysis, however, any use of a spatial coordinate in an expression will
introduce a nonlinear contribution because it will be deformation
dependent.

Coordinate Systems
Force vectors, stress and strain tensors, as well as various material tensors are
represented by their components in a specified coordinate system. By default, material
properties use the canonical system in the material frame. This is the system whose
basis vectors coincide with the X, Y, and Z axes. When the solid deforms, these vectors
rotate with the material.

Loads and constraints, on the other hand, are applied in spatial directions, by default
in the canonical spatial coordinate system. This system has basis vectors in the x, y, and
z directions, which are forever fixed in space. Both the material and spatial default
coordinate system are referred to as the global coordinate system in the physics
interface.

FRAMES AND COORDINATE SYSTEMS | 201

 Vector and tensor quantities defined in the global coordinate system on either frame
use the frame’s coordinate variable names as indices in the tensor component variable
names.

For example, solid.SXY is the material frame XY-plane shear stress, also
known as a second Piola-Kirchhoff stress, while solid.sxy is the
corresponding spatial frame stress, or Cauchy stress. There are also a few
mixed tensors, most notably the deformation gradient solid.FdxY,
which has one spatial and one material index because it is used in
converting quantities between the material and spatial frames.

It is possible to define any number of user coordinate systems on the material and
spatial frames. Most types of coordinate systems are specified only as a rotation of the
basis with respect to the canonical basis in the underlying frame. This means that they
can be used both in contexts requiring a material system and in contexts requiring a
spatial one. A coordinate system defined on the spatial frame will in general introduce
nonlinearities in the problem, since its directions are deformation dependent in case of
a geometrically nonlinear analysis.

The coordinate system can be selected separately for each added material model, load,
and constraint. This is convenient when for example, an anisotropic material with
different orientation in different domains is required. The currently selected
coordinate system is called the local coordinate system.

Coordinate systems used for describing a material must be defined on the material
frame.

202 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

Analysis of Deformation
The analysis of deformation aims at deriving descriptions of the local deformation in a
material suitable for use in a constitutive relation. Often, but not always, this amounts
to deriving a strain tensor.

This section starts by a general description of finite deformation in solids. At the end,
the specialization to engineering strains used in geometrically linear analysis, is made.

In this section:

• Lagrangian Formulation
• Deformation Measures
• Invariants of Strain
• Inelastic Strain Contributions
• Axial Symmetry and Deformation
• Out-of-plane and Circumferential Modes

Lagrangian Formulation
The formulation used for structural analysis in COMSOL Multiphysics for both small
and finite deformations is a total Lagrangian formulation. This means that the
computed stress and deformation state is always referred to the material configuration
rather than to current position in space.

Likewise, material properties are always given for material particles and with tensor
components referring to a coordinate system based on the material frame. This has the
obvious advantage that spatially varying material properties can be evaluated just once
for the initial material configuration, and they do not change as the solid deforms and
rotates.

Consider a certain physical particle, initially located at the coordinate X. During

deformation, this particle follows a path

x = x ( X, t )

Here, x is the spatial coordinate and X is the material coordinate.

For simplicity, assume that undeformed and deformed positions are measured in the
same coordinate system. Using the displacement u it is then possible to write

ANALYSIS OF DEFORMATION | 203

 x = X + u ( X, t )

The displacement is considered as a function of the material coordinates (X, Y, Z), but
it is not explicitly a function of the spatial coordinates (x, y, z). It is thus only possible
to compute derivatives with respect to the material coordinates.

In the following, the gradient operator is assumed to be a gradient with respect to the
material coordinates, unless something else is explicitly stated.

∇ = ∇X = ∂ ∂ ∂
∂X ∂Y ∂Z

The gradient of the displacement, which occurs frequently in the following theory, is
always computed with respect to material coordinates. In 3D:

∂u ∂u ∂u
∂X ∂Y ∂Z
∇u = ∂v ∂v ∂v
∂X ∂Y ∂Z
∂w ∂w ∂w
∂X ∂Y ∂Z

The deformation gradient tensor F shows how an infinitesimal line element, dX, is
mapped to the corresponding deformed line element dx by

204 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

 ∂x
dx = ------- dX = F dX
∂X

The deformation gradient F contains the complete information about the local
straining and rotation of the material. It is a two-point tensor (or a double vector),
which transforms as a vector with respect to each of its indices. It involves both the
reference and present configurations.

In terms of the displacement gradient, F can be written as

∂x
F = ------- = ∇u + I
∂X

The deformation of the material (stretching) will in general cause changes in the
material density. The ratio between current and initial volume (or mass density) is
given by

dV ρ0
---------- = ------ = det ( F ) = J
dV 0 ρ

Here, ρ0 is the initial density and ρ is the current density after deformation. The
determinant of the deformation gradient tensor F is related to volumetric changes with
respect to the initial state. A pure rigid body displacement implies J = 1. Also, an
incompressible material is represented by J = 1. These are called isochoric processes.

The determinant of the deformation gradient tensor is always positive (since a negative
mass density is unphysical). The relation ρ = ρ0/J implies that for J < 1 there is
compression, and for J > 1 there is expansion. Since J > 0, the deformation gradient
F is invertible.

In the material formulations used within the structural mechanics interfaces, the mass
density should in general be constant because the equations are formulated for fixed
material particles. You should thus not use temperature-dependent material data for
the mass density. The changes in volume caused by temperature changes are
incorporated using the coefficient of thermal expansion when you add Thermal
Expansion (for materials) to the material model.

The variable solid.rho represents a “reference” or “initial” density ρ0,

and not the “current” density ρ. If you are interested in finding the
density of the deformed material (the density in the spatial frame), you
can compute it using the expression solid.rho/solid.J.

ANALYSIS OF DEFORMATION | 205

 Deformation Measures
Since the deformation tensor F is a two-point tensor, it combines both spatial and
material frames. It is not symmetric. Applying a singular value decomposition on the
deformation gradient tensor gives an insight into how much stretch and rotation a unit
volume of material has been subjected to. The right polar decomposition is defined as

F = RU

where R is a proper orthogonal tensor ( det ( R ) = 1 , and R – 1 = R T ) and U is the

right stretch tensor given in the material frame. The rotation tensor R describes the
rigid rotation, and all information about the deformation of the material is contained
in the symmetric tensor U.

• The internal variables for the deformation gradient tensor with respect
to global material coordinates are named solid.FdxX, solid.FdxY,
and so on.
• The internal variables for the deformation gradient tensor with respect
to local material coordinates are named solid.Fdx1, solid.Fdx2,
and so on.
• The rotation tensor components are named solid.RotxX,
solid.RotxY, and so on.
• The right stretch tensor components are named solid.UstchXX,
solid.UstchXY, and so on.

An upper case index refers to the material frame, and a lower case index
refers to the spatial frame.

The stretch tensor contains physically important information about the deformation
state. The eigenvalues of the U tensor are the principal stretches, λ1, λ2, and λ3. The
stretch of a line element with initial length L0 and current length L is

L
λ = ------ = 1 + ε eng
L0

where εeng is the engineering strain. The three principal stretches act along three
orthogonal directions. In the coordinate system defined by these principal directions,
the U tensor will be diagonal:

206 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

 λ1 0 0
U = 0 λ2 0
0 0 λ3

The right Cauchy-Green deformation tensor C defined by

T
C = F F = U2

It is a symmetric and positive definite tensor, which accounts for the strain but not for
the rotation. The eigenvalues of the C tensor are the squared principal stretches, thus
providing a more efficient way to compute the principal stretches than by using the
stretch tensor U directly.

The Green-Lagrange strain tensor is a symmetric tensor defined as

1 1 T
ε = --- ( C – I ) = --- ( F F – I )
2 2

Since C is independent of rigid body rotations, this applies also to the Green-Lagrange
strain tensor.

Using the displacement components and Cartesian coordinates, the Green-Lagrange

strain tensor can be written on component form as

1 ∂u i ∂u j ∂u k ∂u k
ε ij = ---  -------- + -------- + --------- ⋅ --------- .
2  ∂X j ∂X i ∂X i ∂X j  (3-2)

The rotation independence of the Green-Lagrange strain tensor, together

with the fact that it for small strain approaches the engineering strain
tensor explains why it is a common choice in constitutive models for small
strain- finite rotation. As an opposite, a pure rigid rotation causes strains
when engineering strains are used.

The Green-Lagrange is the natural strain representation in a Lagrangian

description. Since it is a tensor in the material frame, its values should be
interpreted in along the undeformed axis orientations.

ANALYSIS OF DEFORMATION | 207

 • The internal variables for the Green-Lagrange strains are named
solid.eX, solid.eXY, and so on.

• The internal variables for the Green-Lagrange strain tensor in local

coordinates are named solid.el11, solid.el12, and so on.
• In a geometrically linear analysis, the strain variables
solid.eX,solid.el11, and so on, will instead represent engineering
strain.
• The right Cauchy-Green deformation tensor in local coordinate system
are named solid.Cl11, solid.Cl12, and so on.

Some textbooks prefer to use the left Cauchy-Green deformation tensor

B = FFT, which is also symmetric and positive definite but it is defined in
the spatial frame.

ENGINEERING STRAIN
Under the assumption of small displacements and rotations, the normal strain
components and the shear strain components are related to the deformation as follows:

∂u γ xy 1 ∂u ∂v
εx = ε xy = ------- = ---  + 
∂x 2 2  ∂ y ∂ x
∂v γ yz 1 ∂v ∂w
εy = ε yz = ------- = ---  +  (3-3)
∂y 2 2 ∂z ∂y 
∂w γ xz 1 ∂u ∂w
εz = ε xz = ------- = ---  +  .
∂z 2 2 ∂z ∂x 

In COMSOL Multiphysics, the tensor form of strain representation (εxy, εyz, εxz) is
used.

In the documentation, the symbol ε is used to denote strain in general. In

a geometrically nonlinear analysis, the strain should be interpreted as a
Green-Lagrange strain. In a geometrically linear analysis, the engineering
strain is used.

The symmetric strain tensor ε consists of both normal and shear strain components:

208 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

 ε x ε xy ε xz
ε = ε xy ε y ε yz
ε xz ε yz ε z

The strain-displacement relationships for the axial symmetry case for small
displacements are

∂u u ∂w ∂u ∂w
εr = , ε ϕ = --- , εz = , and γ rz = +
∂r r ∂z ∂z ∂r

A general description of the axially symmetric case is given in Axial Symmetry and
Deformation.

Invariants of Strain

PRINCIPAL STRAINS
The principal strains are the eigenvalues of the strain tensor (ε), computed from the
eigenvalue equation

( ε – ε p I )v p = 0

The three principal strains are sorted so that

ε p1 ≥ ε p2 ≥ ε p3

This sorting is true also for the 2D cases. The corresponding vectors in the principal
directions, vpi, are orthonormal.

• The internal variables for the principal strains are named solid.ep1,
solid.ep2, and solid.ep3 respectively.

• The internal variables for the components of the directions of the first
principal strains are named solid.ep1X, solid.ep1Y, and
solid.ep1Z. The direction vectors for the other two principal strains
are named analogously.

In a geometrically nonlinear analysis, the Green-Lagrange strain tensor is

used. The orientations of the principal directions will thus be with respect
to the material directions. If you plot the principal strains as arrows, you
should thus use an undeformed plot.

ANALYSIS OF DEFORMATION | 209

 PRINCIPAL STRETCHES
The principal stretches are the eigenvalues of the stretch tensor U, and are also sorted
by size:

λ p1 ≥ λ p2 ≥ λ p3

The internal variables for the principal stretches are named

solid.stchp1, solid.stchp2, and solid.stchp3 respectively. The
elastic principal stretches are named solid.stchelp1,
solid.stchelp2, and solid.stchelp3 respectively.

The different invariants of the strain tensor form are useful for constitutive modeling
and result interpretation. The three fundamental invariants for any tensor are

I 1 ( ε ) = trace ( ε )
1 2 2
I 2 ( ε ) = --- ( trace ( ε ) – trace ( ε ) )
2
I 3 ( ε ) = det ( ε )

The invariants of the strain deviator tensor is also useful.

1
dev ( ε ) = ε – --- trace ( ε )I
3
J 1 ( ε ) = trace ( dev ( ε ) ) = 0
1 1 2
J 2 ( ε ) = --- dev ( ε ):dev ( ε ) = --- I 1 ( ε ) – I 2 ( ε )
2 3
2 3 1
J 3 ( ε ) = det ( dev ( ε ) ) = ------ I 1 ( ε ) – --- I 1 ( ε )I 2 ( ε ) + I 3 ( ε )
27 3

As defined above J2 ≥ 0. I1 represents the relative change in volume for infinitesimal

strains and J2 represents the magnitude of shear strain.

In tensor component notation, the invariants can be written as

210 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

 trace ( ε ) = ε kk
ε kk
dev ( ε ) ij = ε ij – -------- δ ij
3
1
I 2 ( ε ) = --- ( ε ii ε jj – ε ij ε ji )
2
1
J 2 ( ε ) = --- dev ( ε ) ij dev ( ε ) ji
2

The volumetric strain is defined as

ε vol = trace ( ε ) .

The internal variable for the volumetric strain is solid.evol.

In terms of the principal strains, the stress invariants can be written as

I 1 ( ε ) = ε p1 + ε p2 + ε p3
I 2 ( ε ) = ε p1 ε p2 + ε p2 ε p3 + ε p1 ε p3
I 3 ( ε ) = ε p1 ε p2 ε p3

The principal stresses are the roots of the characteristic equation (Cayley–Hamilton
theorem)

3 2
εp – I1 εp + I2 εp – I3 = 0

STRAIN RATE AND SPIN

The spatial velocity gradient is defined in components as

∂ v
L kl = ( x, t )
∂ xl k

where v k ( x, t ) is the spatial velocity field. It can be shown that L can be computed in
terms of the deformation gradient as

dF – 1
L = F
dt

where the material time derivative is used.

ANALYSIS OF DEFORMATION | 211

 The velocity gradient can be decomposed into symmetric and skew-symmetric parts

L = Ld + Lw

where

1 T
L d = --- ( L + L )
2

is called the rate of strain tensor, and

1 T
L w = --- ( L – L )
2

is called the spin tensor. Both tensors are defined on the spatial frame.

It can be shown that the material time derivative of the Green-Lagrange strain tensor
can be related to the rate of strain tensor as

dε
------ = F T L d F
dt

The spin tensor Lw(x,t) accounts for an instantaneous local rigid-body rotation about
an axis passing through the point x.

Components of both Ld and Lw are available as results and analysis

variables under the Solid Mechanics interface.

• The components of the rate of strain tensor are named solid.Ldx,

solid.Ldxy, and so on.

• The components of the spin tensor are named solid.Lwx,

solid.Lwxy, and so on.

Inelastic Strain Contributions

Many of the material models in COMSOL Multiphysics will compute a stress based on
an elastic strain. The elastic strain tensor is obtained after removing any inelastic
deformation contribution from the total deformation from the displacements. There
are several possible inelastic strain contributions:

• Initial strain, ε0
• External strain, εex

212 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

• Thermal strain, εth
• Hygroscopic strain, εhs
• Plastic strain, εpl
• Creep strain, εcr
• Viscoplastic strain, εvp

ADDITIVE DECOMPOSITION
In a geometrically linear analysis, the elastic strain is computed by a straightforward
subtraction of the inelastic strain:

ε e = ε – ε inel

where

ε inel = ε 0 + ε ex + ε th + ε hs + ε pl + ε cr + ε vp

Additive decomposition of strains can also be used in a geometrically nonlinear

analysis. In this case, it can however only be justified as long as the strains are small. In
the case of large deformations, the different strain contributions may not even be
commutative.

You can choose to use additive decomposition also for geometric nonlinearity by
selecting the Additive strain decomposition check box in the settings for Linear Elastic
Material or Nonlinear Elastic Material.

MULTIPLICATIVE DECOMPOSITION
In the finite deformation case, the inelastic strain is instead removed using a
multiplicative decomposition of the deformation gradient tensor. The elastic
deformation gradient tensor is the basis for all strain energy formulations in
hyperelastic materials, and also for the elastic strain in linear and nonlinear elasticity. It
is derived by removing the inelastic deformation from the total deformation gradient
tensor.

The total deformation gradient tensor is defined as the result of two successive
operations, an inelastic deformation followed by an elastic deformation:

ANALYSIS OF DEFORMATION | 213

 F = F el F inel (3-4)

The order is important here, multiplication from the left makes the elastic
deformation act on the inelastically deformed state.

Since a deformation gradient tensor describes a mapping from one frame

to another, there are actually three frames involved in this operation. The
F tensor is defined by the displacements as usual and describes the
mapping from the material frame to the spatial frame. The Finel tensor,
however, describes a mapping from the material frame to an intermediate
frame, and the Fel tensor describes a mapping from the intermediate
frame to the spatial frame.

When the inelastic deformation gradient tensor is known, the elastic deformation
gradient tensor is computed as

–1
F el = FF inel (3-5)

so the inelastic deformations are removed from the total deformation gradient tensor.
The elastic right Cauchy-Green deformation tensor is then computed from Fel.

T
C el = F el F el

and the elastic Green-Lagrange strain tensor is computed as:

1
ε el = --- ( C el – I )
2

The inelastic deformation tensor Finel is derived from inelastic processes, such as
thermal expansion or plasticity. When there are several inelastic contributions, they are
applied sequentially to obtain the total inelastic deformation tensor Finel.

F inel = F 1 F 2 F 3 F 4 ...

214 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

where Fi is the inelastic strain contribution from subnode i under a Linear Elastic
Material, Nonlinear Elastic Material, or Hyperelastic Material.

• The order is important when deformations are finite. The

contributions are applied in the same order as the subnodes appear in
the model tree. If a Thermal Expansion node appears before a Plasticity
node, then the physical process can be viewed as a thermal expansion
followed by a plastic deformation.
• When a certain inelastic strain contribution is small, the order is not
significant.
• If the inelastic strain is a pure isotropic volume change, as is often the
case for thermal expansion and hygroscopic swelling, the order is not
significant.

The internal variables for the elastic right Cauchy-Green deformation

tensor in the local coordinate system are named solid.Cel11,
solid.Cel12, and so on; and for the elastic Green-Lagrange tensor in
local coordinates solid.eel11, solid.eel12, and so on.

The elastic, inelastic, and total volume ratios are related as

det ( F ) = det ( F el )det ( F inel ) or J = J el J inel

The internal variables for the elastic, inelastic, and total volume ratio are
named solid.Jel, solid.Ji, and solid.J.

Large strain plasticity

In case of large strain plasticity, the plastic strains are primarily not represented as
strains, but as the plastic deformation gradient tensor, Fpl.

The plastic Green-Lagrange strain tensor is computed from the plastic deformation
gradient tensor as

1 T
ε pl = --- ( F pl F pl – I )
2

As opposed to the small strain formulation, the total, plastic, and elastic
Green-Lagrange strain tensors are related as

ANALYSIS OF DEFORMATION | 215

 –T –1
ε el = F pl ( ε – ε pl )F pl

Axial Symmetry and Deformation

The axially symmetric geometry uses a cylindrical coordinate system. Such a coordinate
system is orthogonal but curvilinear, and you can choose between a covariant basis
e1, e2, e3 and a contravariant basis e1, e2, e3 when formulating the theory.

The metric tensor is

1 0 0
[ g ij ] = 0 r 2 0
0 0 1

in the coordinate system given by e1, e2, e3, and

1 0 0
g
ij = 0 r –2 0
0 0 1

in e1, e2, e3.

The metric tensor plays the role of a unit tensor for a curvilinear coordinate system.

For any vector or tensor A, the metric tensor can be used for conversion between
covariant, contravariant, and mixed components:

 ( Aim g
j mj
Ai = )
m

ni mj
 ( Anm g
ij
A = g )
m, n

In both covariant and contravariant basis, the base vector in the azimuthal direction
has a non-unit length. To cope with this issue, the so-called physical basis vectors of
unit length are introduced. These are

1 1 2 3
e r = e 1 = e , e ϕ = --- e 2 = re , e z = e 3 = e
r

The corresponding components for any vector or tensor are called physical.

216 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

For any tensor, the physical components are defined as

phys ij
A ij = g ii g jj A

where no summation is done over repeated indices.

MIXED COMPONENTS AND PRINCIPAL INVARIANTS

The mixed tensor components are given by

 (g
i im
Aj = A mj )
m

The principal invariants are

1
 Ai
i i
I 1 ( A ) = trace [ A i ] = = A 11 + A 22 ----2- + A 33
i
r

1 j
 Aj Ai
2 i
I 2 ( A ) = ---  ( I 1 ( A ) ) –
2
i, j

i
I 3 ( A ) = det [ A i ]

DISPLACEMENTS AND AXIAL SYMMETRY ASSUMPTIONS

The axial symmetry implementation in COMSOL Multiphysics assumes independence
of the angle, and also that the azimuthal component of the displacement is identically
zero. The physical components of the radial and axial displacement, u and w, are used
as dependent variables for the axially symmetric geometry.

STRAINS
The right Cauchy-Green deformation tensor is defined as

T T
C = ( ( ∇u ) + ∇u + ( ∇u ) ∇u ) + g

and the Green-Lagrange strain tensor is

1
ε = --- ( C – g )
2

Under the axial symmetry assumptions, the covariant components of C are

ANALYSIS OF DEFORMATION | 217

 ∂u ∂u 2 ∂w 2
C 11 = 2 +   +  +1
∂r  ∂r   ∂r 

C 12 = C 23 = 0

∂u ∂w ∂u ∂-----
u- ∂------
w ∂w
C 13 = + + + - -------
∂z ∂r ∂ r ∂z ∂r ∂z

2 2
C 22 = 2ru + [ ( u ) ] + r

∂v ∂u ∂v
C 23 = r ------ + – v ------ + u ------
∂z ∂z ∂z

∂w ∂u 2 ∂w 2
C 33 = 2 ------- +  ------ +  ------- + 1
∂z  ∂z   ∂z 

For geometrically linear analysis, drop the nonlinear terms inside square brackets in the
above expressions.

The physical components of ε are

1
ε r = ε 11 = --- ( C 11 – 1 )
2

ε rϕ = ε ϕz = 0

1
ε rz = ε 13 = --- C 13
2

ε 22 1 2
ε ϕ = -------
2
- = --------- ( C 22 – r )
2
r 2r

1
ε z = ε 33 = --- ( C 33 – 1 )
2

The volumetric strain is

ε vol = I 1 ( ε ) = ε r + ε ϕ + ε z

218 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

Out-of-plane and Circumferential Modes

OUT-OF-PLANE WAVES
If a 2D plane strain model represents a cross-section of the structure that has a
significant uniform extension in the out-of-plane Z-direction, the following 3D
solution can be sought in form of the amplitude expansion:

u ( X, Y, Z, t ) = u 0 ( X, Y ) + u 1 ( X, Y ) exp ( iωt – ik Z Z )

The first term represents a static in-plane prestress solution:

u 0 ( X, Y )
u 0 = v ( X, Y )
0
0

This can be obtained by a standard static analysis using 2D geometry for the cross
section with the corresponding boundary conditions.

The second part of the solution presents a time-harmonic linear perturbation with an
amplitude that can be a function of the in-plane coordinates. Such a perturbation can
be seen as an out-of -plane wave, with a small amplitude that propagates in the
Z -direction, and has a wave length L and phase velocity c:

k Z = 2π ⁄ L
c = ω ⁄ kZ
2πi
u 1 ( X, Y ) exp --------- ( ct – Z )
L

Note that in contrast to the prestress, the perturbation amplitude can have non-zero
values in all three displacement components:

u 1 ( X, Y )
u1 = v 1 ( X, Y )
w 1 ( X, Y )

There are two alternative approaches. The wave length, and thus the wave number kZ,
can be considered as a parameter. Then, ω can be computed by an eigenfrequency
analysis for the 2D cross-section with all three displacement components taken as
dependent variables. As a result, one obtains

ANALYSIS OF DEFORMATION | 219

 ω = 2πf
f = f ( L, u 0 )
c ( λ, u 0 ) = Lf ( L, u 0 )

Thus, the wave speed for the out-of-plane wave is computed as a function of the wave
length and possible prestress in the material. The dependence of the wave speed on the
wave length is often called dispersion.

Alternatively, the frequency f (and thus ω) can be taken as a parameter. Then, the
solution can be computed via eigenvalue analysis with respect to the wave number kZ
using the 2D cross section geometry. Hence,

k Z = 2π ⁄ L
L = L ( f, u 0 )
c ( f, u 0 ) = fL ( f, u 0 )

which determine the wave length and speed for the wave that can propagate
out-of-plane for a given frequency under given in-plane prestress condition. Such
interpretation of the perturbation solution is sometimes called a signaling problem.

CIRCUMFERENTIAL MODES
A standard 2D axially symmetric representation of the structure geometry assumes
zero twist displacement component together with independence of the circumferential
position φ. The following 3D solution form represent an extension of these
assumptions:

u ( R, φ, Z, t ) = u 0 ( R, Z ) + u 1 ( R, Z ) exp ( iωt – imφ )

where m a circumferential mode number that can only have integer values to obey the
axially symmetric nature of the corresponding 3D problem. Thus,

u ( R, φ + 2π, Z, t ) = u ( R, φ, Z, t )

The circumferential wave number can be introduced as k m = m ⁄ R .

The static prestress solution u 0 has zero twist component

u 0 ( R, Z )
u0 = 0
w 0 ( R, Z )

220 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

and it can be obtained using a standard static analysis in 2D axially symmetric
geometry.

The displacement vector u 1 can have non-zero values in all three components, which
are functions of the radial and axial positions. For a given circumferential mode
number m, it can be found using an eigenfrequency analysis in a 2D axially symmetric
geometry. Hence,

ω = 2πf
f = f ( m, u 0 )

and the perturbation solution becomes

u 1 ( R, Z ) exp [ 2πif ( m, u 0 )t – imφ ]

This represents eigenmodes in the corresponding 3D structure, which can be

computed assuming certain constraints on the axis and possible static prestress without
twist and independent off the position along the axis.

ANALYSIS OF DEFORMATION | 221

 Stresses
In this section:

• Defining Stress
• Invariants of the Stress Tensor
• Plane Strain and Plane Stress Cases
• Initial Stresses and Strains
• Axial Symmetry and Stresses

Defining Stress
This section summarizes the definition of different stress measures, stress invariants,
and other important definitions.

Three different stress measures are used in COMSOL Multiphysics:

• Cauchy stress σ defined as force/deformed area in fixed spatial directions not

following the body. This is a symmetric tensor.
• First Piola-Kirchhoff stress P. The forces in the spatial directions are related to the
area in the original (material) frame. This is an unsymmetric two-point tensor.
• Second Piola-Kirchhoff stress S. Both force and area are represented in the material
configuration. For small strains the values are the same as Cauchy stress tensor but
the directions are rotating with the body. This is a symmetric tensor.

The stresses relate to each other as

–1
S = F P
–1 T –1 T
σ = J PF = J FSF

In a geometrically linear analysis, the distinction between the stress measures disappear,
and they all converge to the same values.

In the documentation, the symbol σ is used to denote not only Cauchy

stress, but stress in general. The symbols P and S are used whenever it is
necessary to make a distinction. In geometrically nonlinear analysis, the
stress should in general be interpreted as second Piola-Kirchhoff stress.

222 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

 • The internal variables for the Cauchy stresses are named solid.sx,
solid.sxy, and so on.

• The internal variables for the first Piola-Kirchhoff stresses are named
solid.PxX, solid.PxY, and so on.
• The internal variables for the second Piola-Kirchhoff stresses are
named solid.SX, solid.SXY, and so on.

SIGN CONVENTIONS
A positive normal stress in COMSOL Multiphysics acts in tension. This is the most
widely used definition in general physics and engineering.

Within the field of geomechanics it is however common to let compressive stresses be

positive, since compression is almost always dominant in that field of science. When
working with the material models intended for soils and rocks, you must be aware of
that the “positive in tension” convention is used also there in order to maintain
consistency within the software.

Specifically, the ordering of Principal Stresses is such that σ p1 ≥ σ p2 ≥ σ p3 (including

signs). In geotechnical applications the dominant principal stresses will usually be
compressive, so the third principal stress will the be the one which you may consider
as “largest”.

The convention used in Ref. 1 refers to the hydrostatic pressure (trace of the stress
Cauchy tensor) with a positive sign. The use of the first invariant of Cauchy stress
tensor I1(σ) is preferred in this document where there is a risk of confusion. The
pressure in COMSOL Multiphysics is always defined as positive under compression, or
equivalently, it has the opposite sign of the Cauchy stress tensor’s trace.

Invariants of the Stress Tensor

The different invariants of the stress tensor form an important basis for constitutive
models and also for interpretation of stress results. The three fundamental invariants
for any tensor are

I 1 ( σ ) = trace ( σ )
1 2 2
I 2 ( σ ) = --- ( trace ( σ ) – trace ( σ ) ) (3-6)
2
I 3 ( σ ) = det ( σ )

STRESSES | 223
 In many cases, the invariants of the deviatoric stress tensor are also useful.

1
dev ( σ ) = σ – --- trace ( σ )I
3
J 1 ( σ ) = trace ( dev ( σ ) ) = 0
1 1 2
(3-7)
J 2 ( σ ) = --- dev ( σ ):dev ( σ ) = --- I 1 ( σ ) – I 2 ( σ )
2 3
2 3 1
J 3 ( σ ) = det ( dev ( σ ) ) = ------ I 1 ( σ ) – --- I 1 ( σ )I 2 ( σ ) + I 3 ( σ )
27 3

As defined above J2 ≥ 0. In many material models, the most relevant invariants are I1,
J2, and J3. I1 represents the effect of mean stress, J2 represents the magnitude of shear
stress, and J3 contains information about the direction of the shear stress.

In tensor component notation, the invariants can be written as

trace ( σ ) = σ kk
σ kk
dev ( σ ) ij = σ ij – --------- δ ij
3
1
I 2 ( σ ) = --- ( σ ii σ jj – σ ij σ ji )
2
1
J 2 ( σ ) = --- dev ( σ ) ij dev ( σ ) ji
2

The pressure is defined as

– trace ( σ ) –I1 ( σ )
p = ------------------------ = -----------------
3 3

and is thus positive in compression.

• The internal variables for the invariants I1, I2, and I3 are named
solid.I1s, solid.I2s, and solid.I3s, respectively.

• The internal variables for the invariants J2 and J3 are named

solid.II2s and solid.II3s, respectively.
• The internal variable for the pressure is solid.pm.

PRINCIPAL STRESSES
The principal stresses are the eigenvalues of the stress tensor, computed from the
eigenvalue equation.

224 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

 ( σ – σ p I )v p = 0

The three principal stresses are ordered so that

σ p1 ≥ σ p2 ≥ σ p3

This ordering is true also for the 2D cases. The corresponding principal directions vpi
are orthonormal.

• The internal variables for the principal stresses are named solid.sp1,
solid.sp2, and solid.sp3 respectively.
• The internal variables for the components of the directions of the first
principal stress are named solid.sp1x, solid.sp1y, and
solid.sp1z. The direction vectors for the other two principal stresses
are named analogously.

In terms of the principal stresses, the stress invariants can be written as

I 1 ( σ ) = σ p1 + σ p2 + σ p3
I 2 ( σ ) = σ p1 σ p2 + σ p2 σ p3 + σ p1 σ p3
I 3 ( σ ) = σ p1 σ p2 σ p3

The principal stresses are the roots of the characteristic equation (Cayley–Hamilton
theorem)

3 2
σp – I1 σp + I2 σp – I3 = 0

OTHER STRESS INVARIANTS

It is possible to define other invariants in terms of the primary invariants. One common
auxiliary invariant is the Lode angle θ.

3 3 J3
cos 3θ = ----------- ⋅ -----------
- (3-8)
2 J3 ⁄ 2
2

The Lode angle is bounded to 0 ≤ θ ≤ π/3 when the principal stresses are sorted as
σp1 ≥ σp2 ≥ σp3 (Ref. 1).

STRESSES | 225
 Following this convention, θ = 0 corresponds to the tensile meridian, and θ = π/3
corresponds to the compressive meridian. The Lode angle is part of a cylindrical
coordinate system (the Haigh–Westergaard coordinates) with height (hydrostatic axis)

ξ = I 1 / 3 and radius r = 2J 2 .

The Lode angle is undefined at the hydrostatic axis, where all three
principal stresses are equal (σp1 = σp2 = σp3 = I1/3) and J2 = 0. To avoid
division by zero, the Lode angle is actually computed from the inverse
tangent function atan2, instead of the inverse cosine, as stated in
Equation 3-8.

The Lode angle and the effective (von Mises) stress can be called in
user-defined yield criteria by referencing the variables solid.thetaL and
solid.mises, where solid is the name of the physics interface node.

The octahedral plane (also called π-plane) is defined perpendicular to the hydrostatic
axis in the Haigh–Westergaard coordinate system. The stress normal to this plane is
σoct = I1/3, and the shear stress on that plane is defined by

τ oct = 2/3J 2

The functions described in Equation 3-6 and Equation 3-7 enter into expressions that
define various kind of yield and failure surfaces. A yield surface is a surface in the 3D
space of principal stresses that circumscribe an elastic state of stress.

The principal stresses (σp1, σp2, and σp3) can, when sorted as σp1 ≥ σp2 ≥ σp3, be
written by using the invariants I1 and J2 and the Lode angle (Ref. 1):

1 4J
σ p1 = --- I 1 + ---------2- cos θ
3 3
4J
σ p2 = --- I 1 + ---------2- cos  θ – ------
1 2π
3 3  3
4J
σ p3 = --- I 1 + ---------2- cos  θ + ------
1 2π
3 3  3

226 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

Plane Strain and Plane Stress Cases
For two-dimensional problems, there are two possible approximations: plane strain
and plane stress. The selection is made in the settings for the Solid Mechanics node.

In the plane stress formulation in COMSOL Multiphysics, the plane stress conditions

σ zz = σ xz = σ yz = 0 (3-9)

are not enforced through a modified constitutive relation, as is common in many

textbooks. Instead, an extra set of degrees of freedom for the out-of-plane strains are
introduced, and Equation 3-9 is enforced by solving for the strains.

For a general anisotropic linear elastic material in case of plane stress, COMSOL
Multiphysics solves three equations. For isotropy and orthotropy, only one extra
degree of freedom is needed since all out-of-plane shear components of both stress are
zero.

• For isotropic and orthotropic materials, the extra degree of freedom is

∂w
named wZ, and represents ------- .
∂Z
• For anisotropic materials, two more degrees of freedom area added, uZ
∂u ∂v
and vZ. They represent ------ and ------ .
∂Z ∂Z

Initial Stresses and Strains

Initial stresses and strains refer to a stress and strain state that would exist even without
the external loads. Initial stresses and strains are not initial values in the mathematical
sense. They apply all through the solution, and may even vary with time or solution
parameters. They should rather be considered as an offset to the stress and strain state
in the constitutive relation.

The initial strain is subtracted from the total strain, before the constitutive law is
applied for computing the stresses. The initial stress is added to the stress computed by
using the constitutive law. As an example, linear elasticity including both in initial strain
ε0 and an initial stress s0 can be written as

s = s0 + C : ( ε – ε0 )

It can also be noted that the effect of the initial strain is analogous to that of a for
example a thermal strain.

STRESSES | 227
 A common case is when you have results from another analysis or another physics
interface, which you want to incorporate as initial stresses or strains. You should then
use either the strain or the stress, but not both.

The Initial Stress and Strain node can be added to Linear Elastic Material, Nonlinear
Elastic Material Models, Piezoelectric Material, or Cam-Clay Material Model.

Both the initial stress and strains are tensor variables defined via components in the
local coordinate system for each domain.

In case of nearly incompressible material (mixed formulation), the components of the

total initial stress (that is, without volumetric-deviatoric split) are still input. The initial
pressure in the equation for the pressure help variable pw is computed as

1
p 0 = – --- I 1 ( s 0 )
3

In the case of geometric nonlinearity, the initial stress represents the second
Piola-Kirchhoff stress, not the Cauchy stress. The initial strain is interpreted as a
Green-Lagrange strain.

OTHER POSSIBLE USES OF INITIAL STRAINS AND STRESSES

Many inelastic effects in solids mechanics (for example creep, plasticity, damping,
viscoelasticity, poroelasticity, and so on) are additive contributions to either the total
strain or total stress. Then the initial value input fields can be used for coupling the
elastic equations (solid mechanics) to the constitutive equations (usually General
Form PDEs) modeling such extra effects. When adding stress contributions, you may
however find it more convenient to use the External Stress concept.

External Stress
The external stress is a stress contribution which has a source other than the
constitutive relation. It is similar to the initial stress described in the previous section,
and the two features can to a large extent be used interchangeably.

There is however one large difference: using the External Stress, you can also prescribe
a stress contribution that only acts as a load on the structure but is not added into the
stress tensor. The typical case is when there is a pore pressure in a porous material, a
common case in geotechnical engineering. The stress carried by the solid material,
excluding the pore pressure, is often called the effective stress. So the load from the
pore pressure helps to balance the external loads, while not contributing to the stress
tensor of the solid. The contribution to the virtual work of the external stress is then

228 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

 δW = V –δε : σext dv
The External Stress subnode can be added to Linear Elastic Material, Nonlinear Elastic
Material Models, Piezoelectric Material, or Cam-Clay Material Model.

Axial Symmetry and Stresses

The physical stress components are defined on the global coordinate system:

11
σr = σ

2 22
σϕ = r σ

33
σz = σ

13
σ rz = σ

The first invariant of the stress tensor is

 (σ
ij
I1 ( σ ) = g ij ) = σ r + σ ϕ + σ z
i, j

STRESSES | 229
 Equation of Motion
The first Piola-Kirchhoff stress P is calculated from the second Piola-Kirchhoff stress
as P = FS. The first Piola-Kirchhoff stress relates forces in the present configuration
with areas in the reference configuration, and it is sometimes called the nominal stress.

Using the first Piola-Kirchhoff stress tensor, the equation of motion can be written in
the following form:

2
∂ u
ρ0 2
= FV – ∇X ⋅ P (3-10)
∂t

where the density corresponds to the material density in the initial undeformed state,
the volume force vector FV has components in the actual configuration but given with
respect to the undeformed volume, and the tensor divergence operator is computed
with respect to the coordinates on the material frame. Equation 3-10 is the strong
form that corresponds to the weak form equations within the Solid Mechanics
interface (and many related multiphysics interfaces) in COMSOL Multiphysics. Using
vector and tensor components, the equation can be written as

2
∂ ux ∂P xX ∂P xY ∂P xZ
ρ0 = F Vx –  + +
2 ∂X ∂Y ∂Z 
∂t
2
∂ uy ∂P yX ∂P yY ∂P yZ
ρ0 = F Vy –  + +
∂t
2 ∂X ∂Y ∂Z 
2
∂ uz ∂P zX ∂P zY ∂P zZ
ρ0 = F Vz –  + +
∂t
2 ∂X ∂Y ∂Z 

The components of the first Piola-Kirchhoff stress tensor are non symmetric in the
general case, thus

P iJ ≠ P Ij

because the component indexes correspond to different frames. Such tensors are called
two-point tensors.

230 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

The boundary load vector FA in case of geometric nonlinearity can be related to the
first Piola-Kirchhoff stress tensor via the following formula:

FA = P ⋅ n0

where the normal n0 corresponds to the undeformed surface element. Such a force
vector is often referred to as the nominal traction. In components, it can be written as

F Ax = P xX n X + P xY n Y + P xZ n Z
F Ay = P yX n X + P yY n Y + P yZ n Z
F Az = P zX n X + P zY n Y + P zZ n Z

The Cauchy stress, s, can be calculated as

–1 T –1 T
s = J PF = J FSF

The Cauchy stress is a true stress that relates forces in the present configuration (spatial
frame) to areas in the present configuration, and it is a symmetric tensor.
Equation 3-10 can be rewritten in terms of the Cauchy stress as

2
∂ u
ρ 2
= fV – ∇x ⋅ s
∂t

where the density corresponds to the density in the actual deformed state, the volume
force vector fV has components in the actual configuration (spatial frame) given with
respect to the deformed volume, and the divergence operator is computed with respect
to the spatial coordinates.

The pressure is computed as

1
p = – --- trace ( s )
3

which corresponds to the volumetric part of the Cauchy stress. The deviatoric part is
defined as

s d = s + pI

The second invariant of the deviatoric stress

1
J 2 ( s ) = --- s d :s d
2

EQUATION OF MOTION | 231

 is used for the computation of von Mises (effective) stress

s mises = 3J 2 ( s )

232 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

Material Models
There are many material models available for structural analysis in COMSOL
Multiphysics. In this section, theory for all material models is presented. The material
models available, and also some detailed aspects of them, depend on which licenses you
have. The material models, grouped by families, are as follows:

• Linear Elastic Material

• Nonlinear Elastic Material Models
- Ramberg-Osgood
- Power Law
- Bilinear Elastic
- Uniaxial Data
- Hyperbolic Law
- Hardin-Drnevich
- Duncan-Chang
- Duncan-Selig
- User Defined

• Linear Viscoelastic Materials

- Generalized Maxwell Model
- Standard Linear Solid Model
- Kelvin-Voigt Model

MATERIAL MODELS | 233

 • Hyperelastic Material Models
- Neo-Hookean
- St Venant-Kirchhoff
- Mooney-Rivlin, Two Parameters
- Mooney-Rivlin, Five Parameters
- Mooney-Rivlin, Nine Parameters
- Yeoh
- Ogden
- Storakers
- Varga
- Arruda-Boyce
- Gent
- Blatz-Ko
- Gao
- Murnaghan
- User Defined

• Elastoplastic Material Models

- von Mises Criterion
- Tresca Criterion
- Shima-Oyane Criterion
- Gurson Criterion
- Gurson-Tvergaard-Needleman Criterion
- Fleck-Kuhn-McMeeking Criterion
- FKM-GTN Criterion
- Mohr-Coulomb Criterion
- Drucker-Prager Criterion
- Matsuoka-Nakai Criterion
- Lade-Duncan Criterion
- Hill Orthotropic Plasticity

234 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

• Failure Criteria for Concrete, Rocks, and Other Brittle Material
- Bresler-Pister Criterion
- Willam-Warnke Criterion
- Ottosen Criterion
- Original Hoek-Brown Criterion
- Generalized Hoek-Brown Criterion

• Cam-Clay Material Model

• Creep and Viscoplasticity
- Norton Law (Power law)
- Norton-Bailey Law
- Garofalo Law (Hyperbolic Sine Law)
- Navarro-Herring Creep (Diffusional Creep)
- Coble Creep (Diffusional Creep)
- Weertman Creep (Dislocation Creep)
- Anand Viscoplastic Model
- Chaboche Viscoplastic Model
- Perzyna Viscoplastic Model
- Creep Potential
- Volumetric Creep
- Deviatoric Creep
- User-Defined Creep
• Piezoelectric Material
• Magnetostrictive Material
• Rigid Domain Material Model
• Safety Factor Evaluation

You can also add a material model which you have coded yourself and made available
as a binary library file using an External Stress-Strain Relation.

In the COMSOL Multiphysics Reference Manual:

• Working with External Materials

• External Material

MATERIAL MODELS | 235

 Linear Elastic Material
For a linear elastic material, Hooke’s law relates the stress tensor to the elastic strain
tensor:

σ = σ ex + C : ε el = σ ex + C : ( ε – ε inel ) (3-11)

where C is the 4th order elasticity tensor, “:” stands for the double-dot tensor product
(or double contraction). The elastic strain εel is the difference between the total strain
ε and all inelastic strains εinel. There may also be an extra stress contribution σex with
contributions from initial stresses and viscoelastic stresses. In case of geometric
nonlinearity, the second Piola-Kirchhoff stress tensor and the Green-Lagrange strain
tensor are used.

The elastic strain energy density is

1 1
W s = --- ε el : ( C : ε el + 2σ 0 ) = --- ε el : ( σ + σ 0 ) (3-12)
2 2

This expression assumes that the initial stress contribution is constant during the
straining of the material.

TE N S O R V S . M A T R I X F O R M U L A T I O N S
Because of the symmetry, the strain tensor can be written as the following matrix:

ε x ε xy ε xz
ε xy ε y ε yz
ε xz ε yz ε z

A similar representation applies to the stress tensor:

σ x σ xy σ xz
σ xy σ y σ yz
σ xz σ yz σ z

Due to the symmetry, the elasticity tensor can be completely represented by a

symmetric 6-by-6 matrix as:

236 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

 D 11 D 12 D 13 D 14 D 15 D 16 1111 1122 1133 1112 1123 1113
C C C C C C
D 12 D 22 D 23 D 24 D 25 D 26 1122 2222 2233 2212 2223 2213
C C C C C C
D 13 D 23 D 33 D 34 D 35 D 36 1133 2233 3333 3312 3323 3313
D = = C C C C C C
D 14 D 24 D 34 D 44 D 45 D 46 C
1112
C
2212
C
3312
C
1212
C
1223
C
1213

D 15 D 25 D 35 D 45 D 55 D 56 C
1123
C
2223
C
3323
C
1223
C
2323
C
2313

D 16 D 26 D 36 D 46 D 56 D 66 1113 2213 3313 1213 2313 1313

C C C C C C

which is the elasticity matrix.

ISOTROPIC MATERIAL AND ELASTIC MODULI

In this case, the elasticity matrix becomes

1–ν ν ν 0 0 0
ν 1–ν ν 0 0 0
ν ν 1–ν 0 0 0
E 1 – 2ν-
D = --------------------------------------- 0 0 0 --------------- 0 0
( 1 + ν ) ( 1 – 2ν ) 2
1 – 2ν-
0 0 0 0 --------------- 0
2
1 – 2ν-
0 0 0 0 0 ---------------
2

Different pairs of elastic moduli can be used, and as long as two moduli are defined,
the others can be computed according to Table 3-1.
TABLE 3-1: EXPRESSIONS FOR THE ELASTIC MODULI.

DESCRIPTION VARIABLE D(E,ν) D(E,G) D(K,G) D(λ,μ)

Young’s E= E E 9KG - 3λ + 2μ
------------------ μ --------------------
modulus 3K + G λ+μ
Poisson’s ν= ν E- λ -
---  1 – ------------------
3G -
------- –1 1 --------------------
ratio 2G 2 3K + G 2(λ + μ)

Bulk K= E - EG - K 2μ
----------------------- -------------------------- λ + -------
modulus 3 ( 1 – 2ν ) 3 ( 3G – E ) 3
Shear G= E - G G μ
--------------------
modulus 2(1 + ν)

MATERIAL MODELS | 237

 TABLE 3-1: EXPRESSIONS FOR THE ELASTIC MODULI.

DESCRIPTION VARIABLE D(E,ν) D(E,G) D(K,G) D(λ,μ)

Lamé λ= Eν G ( E – 2G )- 2G λ
--------------------------------------
- --------------------------- K – --------
parameter ( 1 + ν ) ( 1 – 2ν ) 3G – E 3
λ
Lamé μ= E - G G μ
--------------------
parameter 2(1 + ν)
μ
Pressure- cp = E(1 – ν) - G ( 4G – E )- K + 4G ⁄ 3- λ + 2μ
wave speed ----------------------------------------- --------------------------- ------------------------- ----------------
ρ ( 1 + ν ) ( 1 – 2ν ) ρ ( 3G – E ) ρ ρ
Shear-wave cs = E - G⁄ρ G⁄ρ μ⁄ρ
speed ------------------------
2ρ ( 1 + ν )
According to Table 3-1, the elasticity matrix D for isotropic materials is written in
terms of Lamé parameters λ and μ,

λ + 2μ λ λ 0 0 0
λ λ + 2μ λ 0 0 0
D = λ λ λ + 2μ 0 0 0
0 0 0 μ 0 0
0 0 0 0 μ 0
0 0 0 0 0 μ

or in terms of the bulk modulus K and shear modulus G:

4G 2G 2G
K + -------- K – -------- K – -------- 0 0 0
3 3 3
2G 4G 2G
K – -------- K + -------- K – -------- 0 0 0
3 3 3
D = 2G 2G 4G
K – -------- K – -------- K + -------- 0 0 0
3 3 3
0 0 0 G 0 0
0 0 0 0 G 0
0 0 0 0 0 G

ORTHOTROPIC AND ANISOTROPIC MATERIALS

There are two different ways to represent orthotropic or anisotropic data. The
Standard (XX, YY, ZZ, XY, YZ, XZ) material data ordering converts the indices as:

238 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

 11 1 x
22 2 y
33 ↔ 3 ↔ z
12, 21 4 xy
23, 32 5 yz
13, 31 6 xz

thus, Hooke’s law is presented in the form involving the elasticity matrix D and the
following vectors:

σx σx  εx εx 
 
σy σy  εy εy 
 
σz σz  εz εz 
= + D – 
σ xy σ xy  2ε xy 2ε xy 
 
σ yz σ yz  2ε yz 2ε yz 
 
σ xz σ xz  2ε xz 2ε xz 
ex inel

COMSOL Multiphysics uses the complete tensor representation internally to perform

the coordinate system transformations correctly.

Beside the Standard (XX, YY, ZZ, XY, YZ, XZ) Material data ordering, the elasticity
coefficients can be entered following the Voigt notation. In the Voigt (XX, YY, ZZ, YZ,
XZ, XY) Material data ordering, the sorting of indices is:

11 1 x
22 2 y
33 ↔ 3 ↔ z
23, 32 4 yz
13, 31 5 xz
12, 21 6 xy

The last three rows and columns in the elasticity matrix D are thus swapped.

Orthotropic Material
The elasticity matrix for orthotropic material in the Standard (XX, YY, ZZ, XY, YZ, XZ)
Material data ordering has the following structure:

MATERIAL MODELS | 239

 D 11 D 12 D 13 0 0 0
D 12 D 22 D 23 0 0 0
D 13 D 23 D 33 0 0 0
D =
0 0 0 D 44 0 0
0 0 0 0 D 55 0
0 0 0 0 0 D 66

where the components are as follows:

2 2
E x ( E z ν yz – E y ) E x E y ( E z ν yz ν xz + E y ν xy )
D 11 = ---------------------------------------- , D 12 = – -----------------------------------------------------------------
D denom D denom

2 2
E x E y E y ( ν xy ν yz + ν xz ) E y ( E z ν xz – E x )
D 13 = – ---------------------------------------------------------- , D 22 = ----------------------------------------
D denom D denom

2
E y E z ( E y ν xy ν xz + E x ν yz ) E y E z ( E y ν xy – E x )
D 23 = – ----------------------------------------------------------------- , D 33 = -----------------------------------------------
D denom D denom

D 44 = G xy , D 55 = G yz , and D 66 = G xz

where

2 2 2 2
D denom = E y E z ν xz – E x E y + 2ν xy ν yz ν xz E y E z + E x E z ν yz + E y ν xy

The values of Ex, Ey, Ez, νxy, νyz, νxz, Gxy, Gyz, and Gxz are supplied in designated
fields in the physics interface. COMSOL Multiphysics deduces the remaining
components—νyx, νzx, and νzy—using the fact that the matrices D and D−1 are
symmetric. The compliance matrix has the following form:

240 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

 1 ν yx ν zx
------ – -------- – -------- 0 0 0
Ex Ey Ez
ν xy 1 ν zy
– -------- ------ – -------- 0 0 0
Ex E y Ez
ν xz ν yz 1
–1
– -------- – -------- ------ 0 0 0
D = Ex Ey Ez
1
0 0 0 --------- 0 0
G xy
1
0 0 0 0 --------- 0
G yz
1
0 0 0 0 0 ---------
G xz

The values of νxy and νyx are different for an orthotropic material. For a
certain set of given material data, you must make sure that the definition
of the indices is consistent with the definition used in COMSOL
Multiphysics.

The elasticity matrix in the Voigt (XX, YY, ZZ, YZ, XZ, XY) Material data ordering changes
the sorting of the last three elements in the elasticity matrix:

D 44 = G yz , D 55 = G xz , and D 66 = G xy

Anisotropic Material
In the general case of fully anisotropic material, you provide explicitly all 21
components of the symmetric elasticity matrix D, in either Standard (XX, YY, ZZ, XY, YZ,
XZ) or Voigt (XX, YY, ZZ, YZ, XZ, XY) Material data ordering.

AXIAL SYMMETRY
For the linear elastic material, the stress components in coordinate system are

ij ij ijkl
σ = σ ex + C ( ε kl – ε inel, kl )

For anisotropic and orthotropic materials, the 4th-order elasticity tensor is defined
from D matrix according to:

MATERIAL MODELS | 241

 σr σr  εr εr 
 
σϕ σϕ  εϕ εϕ 
= + D – 
σz σz  εz εz 
 
σ rz σ rz
ex
 2ε rz 2ε rz
inel


The user input D matrix always contains the physical components of the elasticity
tensor

phys
C ijkl

and the corresponding tensor components are computed internally according to:

phys
ijkl C ijkl
C = -----------------------------------------------
g ii g jj g kk g ll

For an isotropic material:

ijkl ij kl ik jl il jk
C = λg g + μ(g g + g g )

where λ and μ are the first and second Lamé elastic parameters and g is the metric
tensor.

For a hyperelastic material, the second Piola-Kirchhoff stress tensor is computed as

∂W s
S = 2
∂C

which is computed as the contravariant components of the stress in the local

coordinate system:

ij ∂W s
S = 2
∂ C ij

The energy variation is computed as

S
ij
S : test ( ε ) = test ( ε ij )
i, j

which can be also written as

242 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

 S r test ( ε r ) + S ϕ test ( ε ϕ ) + S z test ( ε z ) + 2S rz test ( ε rz )

ENTROPY AND THERMOELASTICITY

The free energy for the linear thermoelastic material can be written as

F = ρf 0 ( T ) + W s ( ε, T )

where the strain energy density Ws(ε, T) is given by Equation 3-12. Hence, the stress
can be found as

∂F ∂W
σ =   =   = C : ( ε – ε inel )
∂ε T ∂ε T

and the entropy per unit volume can be calculated as

∂F 
–  = ρC p log ( T ⁄ T 0 ) + S elast
∂ T ε

where T0 is a reference temperature, the volumetric heat capacity ρCp can be assumed
to be independent of the temperature (Dulong-Petit law), and the elastic entropy is

S elast = α : σ

where α is the thermal expansion coefficient tensor. For an isotropic material, it

simplifies into

S elast = α ( σ x + σ y + σ z )

The heat balance equation can be written as

∂T ∂
ρC p + T S elast = ∇ ⋅ ( k ∇T ) + Q h
∂t ∂t

where k is the thermal conductivity matrix, and the heat source caused by the
dissipation is

·
Qh = τ : ε

·
where ε is the strain-rate tensor and the tensor τ represents all possible inelastic stresses
(for example, a viscous stress).

Using the tensor components, the heat balance can be rewritten as:

MATERIAL MODELS | 243

 ∂T ∂
ρC p
∂t
+  Tαmn ∂ t σmn = ∇ ⋅ ( k ∇T ) + Q h (3-13)
m, n

In many cases, the second term can be neglected in the left-hand side of Equation 3-13
because all Tαmn are small. The resulting approximation is often called uncoupled
thermoelasticity.

Nonlinear Elastic Material Models

As opposed to hyperelastic materials, where the stress-strain relationship becomes
significantly nonlinear at moderate to large strains, nonlinear elastic materials present
nonlinear stress-strain relationships even at infinitesimal strains.

Here, nonlinear effects on the strain tensor are not as relevant as the nonlinearity of
the elastic properties. Important materials of this class are Ramberg-Osgood for
modeling metal and other ductile materials, and the Duncan-Chang soil model.

The nonlinear elastic materials as such do not include strain-rate nor stress-rate in the
constitutive equations. It is however possible to add linear viscoelasticity to these
materials.

For a nonlinear material to be “energetically sound” it should be possible to take any

path in stress-strain space and return to the undeformed state without producing or
dissipating any net energy. A requirement is then that the bulk modulus depends only
on the volumetric strain, and the shear modulus depends only on the shear strains.

The splitting into volumetric and deviatoric components of the stress tensor helps
ensuring the “path independent” restriction for isotropic nonlinear elastic materials.

For isotropic linear elastic materials, the stress tensor follows Hooke’s law:

σ = σ ex + C : ε el = σ ex + C : ( ε – ε inel )

For a more detailed discussion, see Equation 3-11.

It is possible to split the stress and elastic strain tensors into the deviatoric and
volumetric contributions

1
σ = dev ( σ ) + --- trace ( σ )I
3

and

244 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

 1
ε el = dev ( ε el ) + --- trace ( ε el )I
3

Assuming only elastic stresses in linear isotropic elastic medium, Hooke’s law simplifies
to

1
σ = dev ( σ ) + --- trace ( σ )I = 2Gdev ( ε el ) + Ktrace ( ε el )I
3

where K is the bulk modulus and G is the shear modulus. By using the convention that
the pressure is the mean stress defined as positive in compression,

1
p = – --- trace ( σ )
3

The volumetric strain (positive in tension) is

ε el,vol = trace ( ε el )

The linear relation between pressure and volumetric elastic strain is thus

p = – Kε el,vol

The deviatoric stress and deviatoric elastic strain tensors are related by the shear
modulus

dev ( σ ) = 2Gdev ( ε el )

By using the contraction of the deviatoric stress and strain tensors, we can alternatively
relate the invariants of these tensors

1 2
J 2 ( σ ) = --- dev ( σ ):dev ( σ ) = ( 2G ) J 2 ( ε el )
2

For a body subject to pure torsion on the plane 12, the stress tensor components are
zero except the shear stress σ12 = σ21 = τ, and also the elastic strain tensor has zero
components beside the shear strains on that plane ε12 = ε21 = γel/2.

We can then write

1 2
J 2 ( σ ) = --- dev ( σ ):dev ( σ ) = τ
2

and

MATERIAL MODELS | 245

 1 1 2
J 2 ( ε el ) = --- dev ( ε el ):dev ( ε el ) = --- γ el
2 4

The shear stress on the plane τ = J 2 ( σ ) is then related to the elastic shear strain
γ el = 2ε 12 = 2 J 2 ( ε el ) by the shear modulus

τ = Gγ el

Nonlinear Moduli
For nonlinear elastic materials, there is a nonlinear relation between shear stress and
shear strain and/or a nonlinear relation between pressure and volumetric strain.

For the purpose of this discussion, τ = J 2 ( σ ) and γ el = 2 J 2 ( ε el ) are used

alternatively as variables.

In the most general case:

p = p ( ε el,vol ) and τ = τ ( γel )

Tangent and secant moduli

The tangent shear modulus Gt(γel) and the secant shear modulus Gs(γel) in the most
general case depend non-linearly on the shear strain, and are defined as

∂τ τ
G t = --------- and G s = ------
∂γ el γ el

The tangent bulk modulus Kt(εel,vol) and the secant bulk modulus Ks(εel,vol) depend
on the elastic volumetric strain, and are defined as

∂p p
K t = – ---------------- and K s = – -------------
∂ε el,vol ε el,vol

For linear elastic materials, it is clear that Gt = Gs = G and Kt = Ks = K, but this is not
the case for nonlinear elastic materials.

At zero strain, the secant and shear moduli are equal to each other Gs(0) = Gt (0) and
Ks(0) = Kt (0).

The nonlinear elastic materials described in the next sections are represented by
introducing nonlinear secant shear and/or bulk moduli.

Geometric nonlinearity
The nonlinear elastic material models are primarily intended for small strain analysis.
When used in a geometrically nonlinear study step, the strains will be interpreted as

246 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

Green-Lagrange strains and the stresses will be interpreted as second Piola-Kirchhoff
stresses. This is relevant for a situation with large rotations but small strains. If the
strains become larger than a few percent, then you must be careful when interpreting
input parameters and results since the strain and stress tensors also have a nonlinear
dependence on the displacements.

RAMBERG-OSGOOD
Ramberg-Osgood material model (Ref. 1) is a non-linear elastic material commonly
used to model plastic deformation in metals, but it also often used in soil engineering.
As it is an elastic model, it can only represent plasticity during pure on-loading
conditions.

For uniaxial extension, the stress-strain curve is defined by the expression

σ σ n
ε = ---- + ε ref  ---------
E σ ref

Here, E means the initial Young’s modulus, and εref is the strain at a reference stress
σref. The parameter n is the stress exponent. It is common to use εref = 0.002, so σref
is the stress at 0.2% strain, typically denoted by the symbol σ0.2. This parameter has
several names depending on the literature: 0.2% offset yield strength, 0.2% proof stress,
0.2% proof strength, or 0.2% yield stress. Typical values for stainless steel are E = 200
GPa, σ0.2 = 600 GPa, and n = 4.8.

The linear strain is given by

σ
ε el = ----
E

and the nonlinear strain by

σ n
ε nl = ε ref  ---------
 σ ref

The total strain is the sum of linear an nonlinear strains

σ σ n
ε = ε el + ε nl = ---- + ε ref  ---------
E σ ref

In order to avoid a circular dependence of internal variables, the nonlinear strain εnl is
defined with an auxiliary degree of freedom, so the stress reads σ = E(ε − εnl).

MATERIAL MODELS | 247

 Ramberg-Osgood material in soil engineering
In soil engineering, it is common to write Ramberg-Osgood material with the
stress-strain expression

σ σ ref σ n
ε = ---- + α ---------  --------- (3-14)
E E  σ ref

so at the reference stress σref, the strain is ε = (1 + α)σref/E. It is common to use α = 3/

7, so σref represents the stress level at which the secant Young’s modulus has been
reduced to 70% its initial value: E0.7 = E/(1 + α) = 0.7E. At this reference stress the
strain is ε = σref/E0.7.

POWER LAW
For this type of material the shear stress is related to the shear strain by the strain
exponent n and a reference shear strain γref (Ref. 2)

γ n
τ = G 0 γ ref  -------
γ ref

The secant shear modulus is given by the power law relation

γ n–1
G s = G 0  -------
 γ ref

The strain exponent controls the nonlinear deformation:

• For n > 1 the material behaves as a dilatant (shear-thickening) solid

• For n = 1 the material is linear elastic
• For 0 < n< 1 the material behaves as pseudoplastic (shear-thinning) solid
• For n = 0 the material is perfectly plastic

BILINEAR ELASTIC
The most commonly mentioned model of “bilinear elastic” material is defined with
two different bulk moduli for either tension and compression. Commonly, brittle
materials like graphite and ceramics exhibit this behavior. The secant bulk modulus
reads:

K = K t for εel,vol > 0

and

248 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

 K = K c for εel,vol < 0

where εel,vol is the volumetric strain, Kc is the bulk modulus for compression, and Kt
the bulk modulus for tension.

UNIAXIAL DATA
Many nonlinear stress-strain curves are measured in a tensile test, for which a nonlinear
curve of force vs displacement is obtained.

If only the uniaxial behavior is measured, the measurements do not fully define the
material behavior. An extra assumption is needed. This material model allows you to
assume either a constant Poisson’s ratio, or a constant bulk modulus.

For the uniaxial tensile test, the principal stresses are σ1 = σmises, σ2 = σ3 = 0. The
principal (axial) strain is positive, and the other two strains (transverse) are negative and
related by the Poisson’s ratio ε2 = ε3 = −νε1.

For uniaxial compression, the axial strain is negative, and the other two strains
(transverse) are positive.

The uniaxial test defines then the relation between the axial stress and axial strain as

σ ax = E s ε ax

where Es is the secant Young’s modulus, and the axial stress σax is considered as a
function of axial strain εax. Thus

σ ax
E s = ---------
ε ax

At zero strain, the secant Young’s modulus is defined as

dσ ax
E s = -------------
dε ax
ε ax = 0

Assuming a constant Poisson’s ratio, the secant shear modulus is then defined as

Es
G s = ---------------------
2(1 + ν)

and the secant bulk modulus as

MATERIAL MODELS | 249

 Es
K s = ------------------------
3 ( 1 – 2ν )

When using the constant bulk modulus assumption, the secant shear modulus is
instead defined as

3KE s
G s = --------------------
9K – E s

HYPERBOLIC LAW
An hyperbolic relation between shear stress and shear strain is obtained by setting the
secant shear modulus

1
G s = G -------------------------n-
γ
1 +  -------
 γ ref

where the strain exponent n and a reference shear strain γref control the shape of the
hyperbola.

For hyperbolic material models, the maximum shear modulus is occurs at zero shear
strain, so practitioners might call G the “maximum shear modulus” and use the
notation Gmax. Sometimes it is also called “small strain shear modulus”.

HARDIN-DRNEVICH
The Hardin-Drnevich model (Ref. 3) is an hyperbolic soil model (with n = 1) defined
by two input parameters: the initial shear modulus G and a reference shear strain γref

1
G s = G -----------------
γ
1 + -------
γ ref

This nonlinear soil model is commonly used for modeling soil dynamics in earthquake
engineering problems.

Since τ = Gsγ, the shear stress is bounded by τmax = Gγref as the shear strain increases.

The hyperbolic Hardin-Drnevich model is normally used for quantifying stiffness

reduction curves in soils. Commonly, the reference shear strain γref is replaced by the
reference shear strain at which the secant shear modulus has been decreased to 70% its
initial value. Calling this shear strain value γ0.7, the reference strain is written as

250 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

 7
γ ref = --- γ 0,7
3

and the secant shear modulus as

1
G s = G -----------------------
3 γ
1 + --- ---------
7 γ 0,7

so that when γ = γ0.7 the secant shear modulus is Gs = 0.7G.

DUNCAN-CHANG
The original model was originated by Kondner to fit triaxial test data for undrained
soils. Duncan and Chang (Ref. 4) and other coworkers (Ref. 5) developed this
hyperbolic model to its current state. The material model is written in terms of the axial
and radial stresses σ1 and σ3 and the axial strain ε, and it describes the stress-strain
curve by fitting the hyperbola

ε
σ 1 – σ 3 = ----------------
a + bε

here a and b are material parameters obtained by curve fitting data from the triaxial
test. The parameter a is related to the initial Young's modulus E

1
a = ----
E

and the parameter b defines the asymptote of the hyperbola, which is related to the
ultimate value of σ1 − σ3 denoted qult

1
--- = ( σ 1 – σ 3 ) = q ult
b ult

The ultimate value qult is related to the strength of the soil.

For the triaxial test, the axial strain ε is related to the shear strain γ by the Poisson’s
ratio as

3
ε = --------------------- γ
2(1 + ν)

and the axial and radial stresses are related to the shear stress as σ 1 – σ 3 = 3τ .

It is possible then to write the relation between shear stress and shear strain as

MATERIAL MODELS | 251

 1 -
-------------------- γ
2(1 + ν)
τ = ---------------------------------------------
1- -------- 3 -
1 --------------------
--- + γ
E q ult 2 ( 1 + ν )

Since the initial shear modulus is related to the initial Young’s modulus as
G = E ⁄ 2 ( 1 + ν ) , we can alternatively write this stress-strain relation as

γ
τ = ---------------------
1- 3γ
--- + ----------
G q ult

which is an hyperbolic law with a secant shear modulus of

G
G s = -----------------------------
G
1 + -------- 3γ
q ult

DUNCAN-SELIG
The Duncan-Selig model is a combination of the Duncan’s hyperbolic material model
(Ref. 4, Ref. 5) and Selig’s model to describe nonlinear bulk modulus behavior. Selig
(Ref. 6) further developed the model of Duncan et al. in order to include a nonlinear
volumetric response in soils.

The model defines the nonlinear volumetric response for the pressure as

ε el,vol
p = – K -----------------------
ε el,vol
1 – -------------
ε ult

where εel,vol is the volumetric strain, and εult is the asymptote of the hyperbola, the
maximum value for the volumetric compression. Note that K represents the bulk
modulus at zero strain.

The secant (nonlinear) bulk modulus is defined for this material model as

1
K s = K -----------------------
ε el,vol
1 – -------------
ε ult

252 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

USER DEFINED
The user defined option allows you to write explicitly how the pressure depends on the
elastic volumetric strain. This could be an analytic function or data interpolated from
a table.

The elastic volumetric strain ε el,vol can be called in user-defined

expressions by referencing the variables solid.eelvol, where solid is
the Name of the physics interface node.

Linear Viscoelastic Materials

Viscoelastic materials have a time-dependent response even if the loading is constant
in time. Many polymers and biological tissues exhibit this behavior. Linear
viscoelasticity is a commonly used approximation where the stress depends linearly on
the strain and its time derivatives (strain rate). Also, linear viscoelasticity deals with the
additive decomposition of stresses and strains. It is usually assumed that the viscous
part of the deformation is incompressible so that the volumetric deformation is purely
elastic.

GENERALIZED MAXWELL MODEL

For isotropic linear elastic materials in the absence of inelastic stresses, Hooke’s law in
Equation 3-11 reduces to

σ = C : εel

where the elastic strain tensor εel = ε − εinel represents the total strain minus initial and
inelastic strains, such as thermal strains.

The stress tensor can be decomposed into a pressure and a deviatoric stress:

σ = – pI + σ d

The pressure, mean stress, or volumetric stress, is given with a positive sign in
compression

1
p = – --- trace ( σ )
3

and the deviatoric stress is computed from the total stress minus the volumetric
contribution

MATERIAL MODELS | 253

 σ d = dev ( σ ) = σ + pI

The elastic strain tensor εel can in the same way be decomposed into volumetric and
deviatoric components

1
ε el = --- ε el,vol I + ε d
3

with the volumetric elastic strain given by

ε el,vol = trace ( ε el )

and the deviatoric contribution by

ε d = dev ( ε el )

For isotropic linear elastic materials, the pressure is then related to the volumetric
elastic deformation by the bulk modulus K

1
p = – --- trace ( σ ) = – Ktrace ( ε el ) = – Kε el,vol
3

and the deviatoric stress tensor is linearly related to the deviatoric elastic strain tensor
by the shear modulus G

σ d = dev ( σ ) = 2Gdev ( ε el ) = 2Gε d

The total stress in Equation 3-11 is then

σ = Kε el,vol I + 2Gε d

In case of geometric nonlinearity, σ represents the second Piola-Kirchhoff stress tensor

and εel the elastic Green-Lagrange strain tensor.

For viscoelastic materials, the deviatoric stress σd is not linearly related to the deviatoric
strain εd but it also depends on the strain history. It is normally defined by the
hereditary integral:

t
∂ε d

σ d = 2 Γ ( t – t' ) -------- dt'
∂t'
0

254 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

The function Γ(t) is called the relaxation shear modulus function (or just relaxation
function) and it can be found by measuring the stress evolution in time when the
material is held at a constant strain.

The relaxation function is often approximated by a Prony series:

G m exp  – -------
t
Γ(t) = G +   τ m
m=1

A physical interpretation of this approach, often called the generalized Maxwell

model, is shown in Figure 3-1
σd

G1 G2 Gm qm
ε G ..... σm
η1 η2 ηm γm

τ1 τ2 τm
σd

Figure 3-1: Generalized Maxwell model.

Hence, G is the stiffness of the main elastic branch, Gm represents the stiffness of the
spring in branch m, and τm is the relaxation time constant of the spring-dashpot pair
in branch m.

The auxiliary strain variable qm is introduced to represent the extension of the

corresponding abstract spring, and the auxiliary variables γm = ε − qm represent the
extensions in the dashpots.

The shear modulus of the elastic branch G is normally called the long-term shear
modulus, or steady-state stiffness, and is often denoted with the symbol G ∞ . The
instantaneous shear modulus G0 is defined as the sum of the stiffness of all the branches

G0 = G +  Gm
m=1

This is the stiffness when the external load is applied much faster than the shortest
relaxation time of any viscous branch.

MATERIAL MODELS | 255

 The relaxations time τm is normally measured in the frequency domain, so the viscosity
of the dashpot is not a physical quantity but instead it is derived from stiffness and
relaxation time measurements. The viscosity of each branch can be expressed in terms
of the shear modulus and relaxation time as

ηm = Gm τm

The stress per branch can be written either in terms of the strain in the spring qm or
the strain in the dashpot γm

·
σ m = 2G m q m = 2η m γ m = 2G m ( ε – γ m )

The sum of the stresses in the viscoelastic branches is then computed from

N N

σq =  σm =  2G m ( ε – γ m )
m=1 m=1

The total stress in Hooke’s law (Equation 3-11) is then augmented by the viscoelastic
stress σq

σ = σ 0 + C : ε el + σ q (3-15)

Computing the Stress on Each Branch

The auxiliary variable γm is a symmetric strain tensor, which has as many components
as the number of strain components of the problem class. Since the stress per branch
is written as

·
σ m = 2G m q m = 2η m γ m = 2G m ( ε – γ m )

the auxiliary variables γm can be computed by solving the ODE

·
ηm γm = Gm ( ε – γm ) (3-16)

The relation between viscosity and relaxation time is

ηm = Gm τm

so that Equation 3-16 can equivalently be written as

·
τm γm + γm = ε (3-17)

256 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

The viscoelastic strain variables γm are treated as additional degrees of freedom. The
shape functions are chosen to be one order lower than those used for the displacements
because these variables add to the strains and stresses computed from displacement
derivatives. The viscoelastic strain variables do not require continuity so discontinuous
shape functions are used.

The viscoelastic strain variables γm are called solid.lemm1.vis1.ev,

where solid is the Name of the physics interface node, and lemm1 is the
name of the elastic material node.

Energy Dissipation
The dissipated energy density rate (SI unit: W/m3) in each branch m is

· ·
W m = σ m :γ m

The rate of total dissipated energy density in the Generalized Maxwell material is then

N
· ·
Wv =  Wm
m=1
·
In order to compute the dissipated energy density, the variable W v is integrated over
time. For frequency domain studies, the dissipation of viscous forces averaged over a
time period 2π/ω is computed from the shear loss modulus G’’ as

W v = ωG''ε˜d : conj ( ε˜d )

MATERIAL MODELS | 257

 STANDARD LINEAR SOLID MODEL
The standard linear solid model, also called SLS model, Zener model, or
three-parameter model, is a simplification of the generalized Maxwell model with only
one spring-dashpot branch:
σd

G1 q1
ε G σq
η1 γ1

τ1
σd

Figure 3-2: Standard linear solid (SLS) model.

The stress in the single branch is computed as

·
σ q = 2G 1 q 1 = 2η 1 γ 1 = 2G 1 ( ε – γ 1 )

where the relaxation time is related to the stiffness and relaxation time as η1 = τ1G1.

The auxiliary strain tensor γ1 is computed after solving by the ODE

·
τ1 γ1 + γ1 = ε

and the dissipated energy density rate of the single branch is calculated from

· ·
W v = σ q :γ 1

258 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

K E L V I N - VO I G T M O D E L
The Kelvin–Voigt viscoelastic model is represented by a spring connected in parallel
with a damper:
σd

ε G η σq

τ
σd

Figure 3-3: Kelvin-Voigt model.

The stress tensor in the viscous branch is computed from the elastic strain rate

· ·
σ q = 2ηε = 2Gτε (3-18)

so there is no need to add the extra DOFs to compute the auxiliary strain tensor γ.

The relaxation time relates the viscosity and shear modulus by η = τG.

The dissipated energy density rate of the Kelvin-Voigt model is then computed from
its rate

· ·
W v = σ q :ε

TE M P E R A T U R E E F F E C T S
For many polymers, the viscoelastic properties have a strong dependence on the
temperature. A common assumption is that the material is thermorheologically simple
(TRS). In a material of this class, a change in the temperature can be transformed
directly into a change in the time scale. The reduced time is defined as

t
dt'
tr =  α------------------------
T ( T ( t' ) )
0

where αT(T) is a temperature-dependent shift function.

The implication is that the problem can be solved using the original material data,
provided that the time is transformed into the reduced time.

MATERIAL MODELS | 259

 Think of the shift function αT(T) as a multiplier to the viscosity in the dashpot in the
Generalized Maxwell model. This shifts the relaxation time, so Equation 3-17 for a
TRS material is modified to

·
α T ( T )τ m γ m + γ m = ε

For the SLS model, the shift applies to a single branch

·
α T ( T )τ1 γ 1 + γ 1 = ε

and for the Kelvin-Voigt model, it applies to the viscosity in the damper.

Williams-Landel-Ferry Shift
One commonly used shift function is defined by the WLF (Williams-Landel-Ferry)
equation:

–C1 ( T – T0 )
log ( α T ) = -----------------------------------
C2 + ( T – T0 )

where a base-10 logarithm is assumed. This shift is only valid over a certain range of
temperature, typically around the glass transition temperature.

The first step to compute the shift factor αT consists of building a master curve based
on experimental data. To do this, the curves of the viscoelastic properties (shear
modulus, Young’s modulus, and so forth.) versus time or frequency are measured at a
reference temperature T0. Then, the same properties are measured at different
temperatures.

The shift value of each curve with respect to the master curve obtained at the
temperature T0 defines the shift factor αT(T). The constants C1 and C2 are material
dependent and are calculated after plotting log(αT) versus T − T0.

αT(T0) = 1 so that T0 is the temperature at which the master curve is

given. If the temperature drops below T0 − C2, the WLF equation is no
longer valid.

Since the master curve is measured at an arbitrary reference temperature T0, the shift
factor αT(T) can be derived with respect to any temperature, and it is commonly taken
as the shift with respect to the glass transition temperature. The values C1 = 17.4 and
C2 = 51.6 K are reasonable approximations for many polymers at this reference
temperature.

260 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

Arrhenius Shift
Below the Vicat softening temperature, the shift factor in polymers is normally
assumed to follow an Arrhenius law. In this case, the shift factor is given by the
equation

log ( α T ) = ----  ---- – ------

Q 1 1
R  T T 0

here, a base-e logarithm is assumed, Q is the activation energy (SI unit: : J/mol), and
R is the universal gas constant.

STATIONARY ANALYSIS
For stationary analysis it is possible to select either the long-term stiffness, in which
case the stiffness of the viscoelastic branches is neglected, or the instantaneous stiffness,
in which case the contribution from all branches is used.

The instantaneous shear modulus G0 is defined as the sum of the stiffness of all the
branches

G0 = G +  Gm
m=1

FREQUENCY DOMAIN ANALYSIS AND DAMPING

For frequency domain analysis, the frequency decomposition is performed as

jωt
σ d = real ( s˜ d e )

jωt
ε d = real ( ε˜d e )

Equation 3-15 and Equation 3-17 are then simplified to

σ̃ d = 2 ( G' + jG'' )ε˜d

where the shear storage modulus G’ and the shear loss modulus G” are defined for the
generalized Maxwell model as

N 2 N
( ωτ m ) ωτ m
G' = G +  Gm ----------------------------2 and G'' =
1 + ( ωτ m )
 G m ----------------------------2
1 + ( ωτ m )
m=1 m=1

MATERIAL MODELS | 261

 and for the SLS model as

2
( ωτ 1 ) ωτ 1
G' = G + G 1 --------------------------2- and G'' = G 1 --------------------------2-
1 + ( ωτ 1 ) 1 + ( ωτ 1 )

and for the Kelvin-Voigt model as

G' = G and G'' = ηω = Gτω

The internal work of viscous forces averaged over a time period 2π/ω is computed as

Q h = ωG''ε˜d : conj ( ε˜d )

See also the description of Viscoelasticity in the Solid Mechanics interface

documentation.

Large Strain Viscoelasticity

The implementation for large strain viscoelasticity follows the derivation by Holzapfel
(Ref. 1).

The generalized Maxwell model is based on the splitting of the strain energy density
into volumetric, isochoric and the contribution from the viscoelastic branches

W s = W iso + W vol +  ψm
m=1

The strain energy in the main hyperelastic branch is normally denoted with the
superscript ∞ to denote the long-term equilibrium (as t → ∞ ).

The second Piola-Kirchhoff stress is computed from

N
∂W s
S = 2
∂C
= S iso + S vol +  Qm
m=1

where the auxiliary second Piola-Kirchhoff stress tensors Qm are defined as

∂ψ m
Qm = 2
∂C

262 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

The time evolution of the auxiliary stress tensor Qm in each viscoelastic branch is given
by the rate

· 1 ·
Q m + ------- Q m = S iso,m
τm

here, Siso,m is the isochoric second Piola-Kirchhoff stress tensor in the branch m.
These tensors are derived from the strain energy density in the main hyperelastic
branch and the energy factors βm as

∂W iso, m ∂W iso
S iso, m = 2 = 2β m = β m S iso
∂C ∂C

so the time evolution of the auxiliary stress tensor Qm is given by

· 1 ·
Q m + ------- Q m = β m S iso
τm

This equation is not well suited for modeling prestressed bodies. Applying the change
of variables

q m = β m S iso – Q m

the time evolution of the auxiliary stress tensor qm reads

·
τ m q m + q m = β m S iso

TE M P E R A T U R E E F F E C T S
The same options for defining Temperature Effects as described for Linear Viscoelastic
Materials are available for large strain viscoelasticity.

Hyperelastic Material Models

A hyperelastic material is defined by its elastic strain energy density Ws, which is a
function of the elastic strain state. It is often referred to as the energy density. The
hyperelastic formulation normally gives a nonlinear relation between stress and strain,
as opposed to Hooke’s law in linear elasticity.

Most of the time, the right Cauchy-Green deformation tensor C is used to describe
the current state of strain (although one could use the left Cauchy-Green tensor B, the
deformation gradient tensor F, and so forth), so the strain energy density is written as
Ws(C).

MATERIAL MODELS | 263

 For isotropic hyperelastic materials, any state of strain can be described in terms of
three independent variables—common choices are the invariants of the right
Cauchy-Green tensor C, the invariants of the Green-Lagrange strain tensor, or the
principal stretches.

Once the strain energy density is defined, the second Piola-Kirchhoff stress in the local
coordinate system is computed as

∂W s
S = 2
∂C

In the general case, the expression for the energy Ws is symbolically evaluated down to
the components of C using the invariants definitions prior to the calculations of the
components of the second Piola-Kirchhoff stress tensor. The differentiation is
performed in components on the local coordinate system.

In Equation View, the definition of the stress components are shown as

solid.Sl11 = 2*d(solid.Ws,solid.Cl11),
solid.Sl12 = d(solid.Ws,solid.Cl12) etc.
The factor 2 in front of the differentiation operator for the shear stresses
is omitted, since the symmetry in the Cauchy-Green tensor will cause two
equal contributions.

Modeling Geometric Nonlinearity

THERMAL EXPANSION
If thermal expansion is present, a stress-free volume change occurs. This is a pure
volumetric change, so the multiplicative decomposition of the deformation gradient
tensor in Equation 3-4 implies

det ( F ) J
J el = --------------------- = --------
det ( F th ) J th

Here, the thermal volume ratio, Jth, depends on the thermal stretch λth, which for
linear thermal expansion in isotropic materials can be written in terms of the isotropic
coefficient of thermal expansion, αiso, and the absolute change in temperature

J th = λ th
3 and λ
th = 1 + α iso ( T – T ref )

264 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

Here, the term αiso(T−Tref) is the thermal strain. The isotropic thermal gradient is
therefore a diagonal tensor defined as

F th = λ th I

When the coefficient of thermal expansion α is anisotropic, the thermal strain is

computed from

ε th = α ( T – Tref )

and the anisotropic thermal gradient is defined as

F th = I + ε th

The internal variables for the thermal stretch and the thermal volume
ratio are named solid.stchth and solid.Jth.

HYGROSCOPIC SWELLING
Hygroscopic swelling is an internal strain caused by changes in moisture content. This
strain depends linearly on the moisture content

ε hs = β h ( c mo – c mo,ref )

where βh is the coefficient of hygroscopic swelling, cmo is the moisture concentration,

and cmo,ref is the strain-free reference concentration. The coefficient of hygroscopic
swelling can represent isotropic or anisotropic swelling. The anisotropic hygroscopic
gradient is defined as

F hs = I + ε hs

HYPERELASTICITY WITH PLASTICITY

It is possible to combine the hyperelastic material models with plasticity. Since these
models are primarily used for large strain applications, only the large strain plasticity
formulation is available. The decomposition between elastic and plastic deformation is
made using a multiplicative decomposition of the deformation gradient tensor,

–1
F el = FF pl

MATERIAL MODELS | 265

 Here, the plastic deformation tensor Fpl depends on the plastic flow rule, yield
function, and plastic potential.

• Multiplicative Decomposition
• Plastic Flow for Large Strains

ISOCHORIC ELASTIC DEFORMATION

For some classes of hyperelastic materials it is convenient to split the strain energy
density into volumetric (also called dilatational) and isochoric (also called
distortional or volume-preserving) contributions. The elastic deformation tensor is
then multiplicatively decomposed into the volumetric and isochoric components

F el = F el,vol F el

with Fel,vol as the volumetric elastic deformation (a diagonal tensor) and F el the
isochoric elastic deformation gradient. Isochoric deformation means that the volume
ratio is kept constant during deformation, so the isochoric elastic deformation is
computed by scaling it by the elastic volume ratio. The elastic volume ratio is defined
by

J el = det ( F el ) = det ( F el,vol )

and the volumetric deformation as

1 ⁄ 3I
F el,vol = J el

By using Jel it is possible to define the isochoric-elastic deformation gradient

– 1/3
F el = J el F el

the isochoric-elastic right Cauchy-Green tensor

T – 2/3
C el = F el F el = J el C el

and the isochoric-elastic Green-Lagrange strain tensor

1
ε el = --- ( C el – I )
2

This scaling changes the eigenvalues of the tensor, but not its principal directions, so
the original and isochoric tensors remain coaxial to each other.

266 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

Some authors call F el and C el the modified tensors. Note that

det ( F el ) = det ( C el ) = 1

The internal variables for the isochoric-elastic Cauchy-Green deformation

tensor in local coordinate system are named solid.CIel11,
solid.CIel12, and so on.

The other two invariants normally used together with Jel are the first and second
invariant of the isochoric-elastic right Cauchy-Green deformation tensor

I 1 ( C el ) and I 2 ( C el )

In these equations:

– 2/3
I 1 ( C el ) = trace ( C el ) = J el I 1 ( C el )

1 2 2 – 4/3
I 2 ( C el ) = --- ( I 1 ( C el ) – trace ( C el ) ) = J el I 2 ( C el )
2

Since I 3 ( C el ) = det ( C el ) = 1 , the third invariant is never explicitly used.

The internal variables for the invariants Jel, I 1 ( C el ) , and I 2 ( C el ) are

named solid.Jel, solid.I1CIel, and solid.I2CIel.

The invariants of the isochoric (modified) elastic Green-Lagrange strain tensor are
related to the invariants of the isochoric-elastic right Cauchy-Green deformation
tensor

1
I 1 ( ε el ) = trace ( ε el ) = --- ( I 1 ( C el ) – 3 )
2

1 2 1
I 2 ( ε el ) = --- ( I 1 ( ε el ) – trace ( ε el 2 ) ) = --- ( I 2 ( C el ) – 2I 1 ( C el ) + 3 )
2 4

MATERIAL MODELS | 267

 1
I 3 ( ε el ) = det ( ε el ) = --- ( I 1 ( C el ) – I 2 ( C el ) ) s
8

The internal variables for the invariants of the isochoric elastic

Green-Lagrange strain tensor are named solid.I1eIel, solid.I2eIel,
and solid.I3eIel.

NEARLY INCOMPRESSIBLE HYPERELASTIC MATERIALS

If the Nearly incompressible material check box is selected for the Hyperelastic Material
node, the total elastic energy function is split into two parts as:

W s = W iso + W vol

where Wiso is the isochoric strain energy density and Wvol is the volumetric strain
energy density.

The isochoric strain energy density can be entered as an expression involving the
following:

• Components of the isochoric elastic right Cauchy-Green tensor C el in the local

coordinate system.
• Principal invariants of the isochoric elastic right Cauchy-Green tensor C el .
• Principal invariants of the isochoric elastic Green-Lagrange strain ε el .

The volumetric strain energy density, Wvol, can only be defined as an expression of the
elastic volumetric deformation. The most commonly used form is:

1 2
W vol ( J el ) = --- κ ( J el – 1 ) (3-19)
2

where κ is the initial bulk modulus. From here, the volumetric stress (pressure) is
calculated as

∂W vol
p m = – --------------
∂J

When the expression in Equation 3-19 is used, the pressure becomes linearly related
to the volume change:

pm = –κ ( J – 1 )

The second Piola-Kirchhoff stress is then given by

268 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

 –1 ∂W iso
S = – p m JC + 2 -------------- (3-20)
∂C

and the Cauchy stress tensor by

∂W iso
σ = J – 1 FSF T = – p m I + 2J – 1 F -------------- F T
∂C

THE LOCKING PROBLEM

A numerical scheme is said to exhibit locking if the accuracy of the approximation
deteriorates as a parameter tends to a limiting value (Ref. 11). Finite elements in solid
mechanics are said to “lock” when exhibiting an unphysical response to deformation
(Ref. 12). Locking can occur for many different reasons. For linear elastic materials,
this typically happens as Poisson’s ratio tends to 0.5, or the bulk modulus is much
larger than the shear modulus. Numerical errors arise because the shape functions are
unable to properly describe the volume preserving deformation.

To avoid the locking problem in computations, the mixed formulation replaces pm in

Equation 3-20 with a corresponding interpolated pressure help variable pw, which
adds extra degrees of freedom to the ones defined by the displacement vector u.

The general procedure is the same as when the Nearly incompressible material check
box is selected for the Linear Elastic Materials node.

THEORY FOR THE PREDEFINED HYPERELASTIC MATERIAL MODELS

Different hyperelastic material models are constructed by specifying different elastic
strain energy expressions. This module has several predefined material models and also
has the option to enter user-defined expressions for the strain energy density.

NEO-HOOKEAN
The strain energy density for the compressible version of the Neo-Hookean material is
written in terms of the elastic volume ratio Jel and the first invariant of the elastic right
Cauchy-Green deformation tensor I1(Cel) (Ref. 10).

1 1 2
W s = --- μ ( I 1 – 3 ) – μ ln ( J el ) + --- λ [ ln ( J el ) ]
2 2

Here, λ and μ are the Lamé coefficients.

The nearly incompressible version uses the isochoric invariant I 1 ( C el ) and the initial
bulk modulus κ

MATERIAL MODELS | 269

 1 1 2
W s = --- μ ( I 1 – 3 ) + --- κ ( J el – 1 )
2 2

See also the description of the Neo-Hookean material model in the Solid
Mechanics interface documentation.

S T VE N A N T - K I R C H H O F F
One of the simplest hyperelastic material models is the St Venant-Kirchhoff material,
which is an extension of a linear elastic material into the hyperelastic regime.

The elastic strain energy density is written with two parameters (the two Lamé
coefficients) and two invariants of the elastic Green-Lagrange strain tensor, I1(εel) and
I2(εel)

1
W s = --- ( λ + 2μ )I 12 – 2μI 2
2

Here, λ and μ are the Lamé parameters.

The nearly incompressible version uses the isochoric invariants I 1 ( ε el ) and I 2 ( ε el ) ,

and the initial bulk modulus κ is calculated from the Lamé parameters κ = λ+2μ/3

1 1 2
W s = --- ( λ + 2μ )I 1 2 – 2μI 2 + --- κ ( J el – 1 )
2 2

See also the description of the St Venant-Kirchhoff material model in the

Solid Mechanics interface documentation.

M O O N E Y - R I V L I N , TW O P A R A M E T E R S
Only a nearly incompressible version is available, and the elastic strain energy density
is written in terms of the two isochoric invariants of the elastic right Cauchy-Green
deformation tensors I 1 ( C el ) and I 2 ( C el ) , and the elastic volume ratio Jel

1 2
W s = C 10 ( I 1 – 3 ) + C 01 ( I 2 – 3 ) + --- κ ( J el – 1 )
2

270 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

The material parameters C10 and C01 are related to the Lamé parameter
μ = 2(C10+C01).

See also the description of the Mooney-Rivlin, Two Parameters material

model in the Solid Mechanics interface documentation.

MOONEY-RIVLIN, FIVE PARAMETERS

Rivlin and Saunders (Ref. 2) proposed a phenomenological model for small
deformations in rubber-based materials on a polynomial expansion of the first two
invariants of the elastic right Cauchy-Green deformation, so the strain energy density
is written as an infinite series

∞ ∞

Ws =   Cmn ( I1 – 3 )m ( I2 – 3 ) n
m = 0n = 0

with C00 = 0. This material model is sometimes also called polynomial hyperelastic
material.

In the first-order approximation, the material model recovers the Mooney-Rivlin strain
energy density

W s = C 10 ( I 1 – 3 ) + C 01 ( I 2 – 3 )

while the second-order approximation incorporates second-order terms

W s = C 10 ( I 1 – 3 ) + C 01 ( I 2 – 3 ) + C 20 ( I 1 – 3 ) 2 + C 02 ( I 2 – 3 ) 2 + C 11 ( I 1 – 3 ) ( I 2 – 3 )

The nearly incompressible version uses the isochoric invariants of the elastic right
Cauchy-Green deformation tensors

I 1 ( C el ) and I 2 ( C el )

and it adds a contribution due to the elastic volume ratio. The strain energy density is
then computed from

2 2
1
  Cmn ( I1 – 3 )
m 2
Ws = ( I 2 – 3 ) n + --- κ ( J el – 1 )
2
m = 0n = 0

MATERIAL MODELS | 271

 Here, κ is the initial bulk modulus.

See also the description of the Mooney-Rivlin, Five Parameters material

model in the Solid Mechanics interface documentation.

MOONEY-RIVLIN, NINE PARAMETERS

The Mooney-Rivlin, nine parameters material model is an extension of the polynomial
expression to third order terms and the strain energy density is written as

3 3
1
  Cmn ( I1 – 3 )
m 2
Ws = ( I 2 – 3 ) n + --- κ ( J el – 1 )
2
m = 0n = 0

See also the description of the Mooney-Rivlin, Nine Parameters material

model in the Solid Mechanics interface documentation.

YE O H
Yeoh proposed (Ref. 1) a phenomenological model in order to fit experimental data of
filled rubbers, where Mooney-Rivlin and Neo-Hookean models were to simple to
describe the stiffening effect in the large strain regime. The strain energy was fitted to
experimental data by means of three parameters, and the first invariant of the elastic
right Cauchy-Green deformation tensors I1(Cel)

Ws = c1 ( I1 – 3 ) + c2 ( I1 – 3 ) 2 + c3 ( I1 – 3 ) 3

The shear modulus depends on the deformation, and it is calculated as

 ∂W s ∂W s
μ = 2 +  = 2c 1 + 4c 2 ( I 1 – 3 ) + 6c 3 ( I 1 – 3 ) 2
 ∂ I1 ∂ I2 

This imposes a restriction on the coefficients c 1, c 2, c 3 , since μ > 0 .

The nearly incompressible version uses the isochoric invariant of the elastic right
Cauchy-Green deformation tensor I 1 ( C el ) , and it adds a contribution from the elastic
volume ratio

272 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

 1 2
W s = c 1 ( I 1 – 3 ) + c 2 ( I 1 – 3 ) 2 + c 3 ( I 1 – 3 ) 3 + --- κ ( J el – 1 )
2

See also the description of the Yeoh material model in the Solid
Mechanics interface documentation.

OGDEN
The Neo-Hookean material model usually fits well to experimental data at moderate
strains but fails to model hyperelastic deformations at high strains. In order to model
rubber-like materials at high strains, Ogden adapted (Ref. 1) the energy of a
Neo-Hookean material to

N
μp αp αp αp
Ws =  -----
- ( λ + λ el2 + λ el3 – 3 )
α p el1
p=1

Here αp and μp are material parameters, and λel1, λel2, and λel3 are the principal elastic
stretches such as Jel = λel1λel2λel3.

The Ogden model is empirical, in the sense that it does not relate the material
parameters αp and μp to physical phenomena. The parameters αp and μp are obtained
by curve-fitting measured data, which can be difficult for N > 2. The most common
implementation of Ogden material is with N = 2, so four parameters are needed.

The nearly incompressible version uses the isochoric elastic stretches

1⁄3
λ eli = λ eli ⁄ J el

and the initial bulk modulus κ

N
μp αp αp αp 1
 -----
2
Ws = - ( λ + λ el2 + λ el3 – 3 ) + --- κ ( J el – 1 )
α p el1 2
p=1

The isochoric elastic stretches define a volume preserving deformation, since

λ el1 λ el2 λ el3 = λ el1 λ el2 λ el3 ⁄ J el = 1

The initial shear modulus is

MATERIAL MODELS | 273

 N
1
μ = ---
2  αk μk
k=1

Sometimes a slightly different definition of the strain energy function for

the Ogden material is used. If you have material data for an Ogden
material given, be careful that the definitions used are the same. If not,
you will need to rescale the μk coefficients.

See also the description of the Ogden material model in the Solid
Mechanics interface documentation.

STORAKERS
The Storakers material (Ref. 12 and Ref. 15) is used to model highly compressible
foams. The strain energy density is written in a similar fashion as in Ogden material:

N
2μ k –αk βk
-  λ + λ el2 + λ el3 – 3 + ----- ( J el – 1 )
αk αk αk 1
Ws =  --------
α
2  el1 βk
k=1 k

The initial shear and bulk moduli can be computed from the parameters μk and βk as

N N

 2μk  βk + --3-
1
μ =  μk and κ =
k=1 k=1

for constant parameters βk = β, the initial bulk modulus becomes κ = 2μ(β + 1/3), so a
stable material requires μ > 0 and β > −1/3. In this case, the Poisson's ratio is given by
ν = β/(2β+2/3), which means that for a Poisson’s ratio larger than −1, β > −2/9 is
needed.

See also the description of the Storakers material model in the Solid
Mechanics interface documentation.

VARGA
The Varga material model (Ref. 1) describes the strain energy in terms of the elastic
stretches as

274 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

 W s = c 1 ( λ el1 + λ el2 + λ el3 – 3 ) + c 2 ( λ el1 λ el2 + λ el2 λ el3 + λ el1 λ el3 – 3 )

The nearly incompressible version uses the isochoric elastic stretches defined as

1⁄3
λ eli = λ eli ⁄ J el

and the initial bulk modulus κ

1 2
W s = c 1 ( λ el1 + λ el2 + λ el3 – 3 ) + c 2 ( λ el1 λ el2 + λ el2 λ el3 + λ el1 λ el3 – 3 ) + --- κ ( J el – 1 )
2

The simplest Varga model is obtained by setting c1 = μ and c2 = 0:

1 2
W s = μ ( λ el1 + λ el2 + λ el3 – 3 ) + --- κ ( J el – 1 )
2

See also the description of the Varga material model in the Solid
Mechanics interface documentation.

ARRUDA-BOYCE
The other hyperelastic materials described are phenomenological models in the sense
that they do not relate the different material parameters (normally obtained by
curve-fitting experimental data) to physical phenomena.

Arruda and Boyce (Ref. 3) derived a material model based on Langevin statistics of
polymer chains. The strain energy density is defined by

 cp ( I1 – 3
p p
Ws = μ0 )
p=1

Here, μ0 is the initial macroscopic shear modulus, I1(Cel) is the first invariant of the
elastic right Cauchy-Green deformation tensor, and the coefficients cp are obtained by
series expansion of the inverse Langevin function.

Arruda and Boyce truncated the series and used only the first five terms as listed in
Table 3-2:
TABLE 3-2: FIRST FIVE COEFFICIENTS OF ARRUDA-BOYCE MATERIAL MODEL

C1 C2 C3 C4 C5

1/2 1/20 N 11/1050 N2 19/7000 N3 519/673750 N4

MATERIAL MODELS | 275

 The strain energy density is then defined as

 cp ( I1 – 3
p p
Ws = μ0 )
p=1

Other authors (Ref. 1) use only the first three coefficients of the series. The number of
segments in the polymeric chain is specified by the parameter N so the material model
is described by only two parameters, μ0 and N.

This material model is sometimes also called the eight-chain model since it was derived
for N = 8.

The nearly incompressible version uses the isochoric invariant I 1 ( C el ) and the initial
bulk modulus κ

5
1
 cp ( I1 – 3
p p 2
Ws = μ0 ) + --- κ ( J el – 1 )
2
p=1

See also the description of the Arruda-Boyce material model in the Solid
Mechanics interface documentation.

GENT
Many hyperelastic material models are difficult to fit to experimental data. Gent
material (Ref. 13 and Ref. 14) is a simple phenomenological constitutive model based
on only two parameters, μ and jm, which defines the strain energy density as:

μ  I 1 – 3
W s = – --- j m log  1 – --------------
2  jm 

Here, μ is the shear modulus and jm is a limiting value for I1− 3, which takes care of
the limiting polymeric chain extensibility of the material.

Since the strain energy density does not depend on the second invariant I2, the Gent
model is often classified as a generalized Neo-Hookean material. The strain energy
density tends to be the one of incompressible Neo-Hookean material as j m → ∞ .

The nearly incompressible formulation uses the isochoric invariants I 1 ( C el ) and the
initial bulk modulus κ:

276 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

 μ  I 1 – 3 1 2
W s = – --- j m log  1 – -------------- + --- κ ( J el – 1 )
2  jm  2

Gent material is the simplest model of the limiting chain extensibility family.

See also the description of the Gent material model in the Solid
Mechanics interface documentation.

BLATZ-KO
The Blatz-Ko material model was developed for foamed elastomers and polyurethane
rubbers, and it is valid for compressible isotropic hyperelastic materials (Ref. 1).

The elastic strain energy density is written with three parameters and the three
invariants of the elastic right Cauchy-Green deformation tensor, I1(Cel), I2(Cel), and
I3(Cel)

μ μ I2
W s = φ ---  ( I 1 – 3 ) + --- ( I 3 – 1 ) + ( 1 – φ ) ---   ----- – 3 + --- ( I 3 – 1 )
1 –β 1 β
2 β 2  I3  β

Here, φ is an interpolation parameter bounded to 0 < φ < 1, µ is the shear modulus, and
β is an expression of Poisson’s ratio.

When the parameter β → ∞ , the strain energy simplifies to a similar form of the
Mooney-Rivlin material model

μ μ I2
W s = φ --- ( I 1 – 3 ) + ( 1 – φ ) ---  ----- – 3
2 2  I3 

In the special case of φ = 1, the strain energy reduces to a similar form of the
Neo-Hookean model

μ μ – 2β
W s = --- ( I 1 – 3 ) + ------ ( J – 1)
2 2β

See also the description of the Blatz-Ko material model in the Solid
Mechanics interface documentation.

MATERIAL MODELS | 277

 GAO
Gao proposed (Ref. 16) a simple hyperelastic material where the strain energy density
is defined by two parameters, a and n, and two invariants of the elastic right
Cauchy-Green deformation tensors Cel:

n n
Ws = a ( I 1 + I –1 )

Here, the invariant I-1(Cel) is calculated as:

–1 I 2 ( C el )
I –1 = trace ( C el ) = ------------------
I 3 ( C el )

Gao proposed that the material is unconditionally stable when the parameters are
bounded to 3 > n > 1 and a > 0, and related these parameters under small strain to the
Young’s modulus and Poisson’s ratio by:

n 2
3 n 8a n–1
E = -------------------- and ν = -----------------
2n + 1 2n + 1

Since n = (1+ν)/(1−2ν) and it is bounded to 1 < n < 3, this material model is stable for
materials with an initial Poisson’s ratio in the range of 0 < ν < 2/7.

See also the description of the Gao material model in the Solid Mechanics
interface documentation.

MURNAGHAN
The Murnaghan potential is used in nonlinear acoustoelasticity. Most conveniently it
is expressed in terms of the three invariants of the elastic Green-Lagrange strain tensor,
I1(εel), I2(εel), and I3(εel)

1 1
W s = --- ( λ + 2μ )I 12 – 2μI 2 + --- ( l + 2m )I 13 – 2mI 1 I 2 + nI 3
2 3

Here, l, m, and n are the Murnaghan third-order elastic moduli, which can be found
experimentally for many commonly encountered materials such as steel and aluminum,
and λ and μ are the Lamé parameters.

See also the description of the Murnaghan material model in the Solid
Mechanics interface documentation.

278 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

USER DEFINED
When a material model is user-defined, an expression for the elastic strain energy Ws is
entered, which can include any expressions involving the following:

• Components of Cel, the elastic right Cauchy-Green deformation tensor in the local
material coordinate system.
• Principal invariants of Cel
I 1 ( C el ) = trace ( C el )

1 2 2
I 2 ( C el ) = --- ( I 1 ( C el ) – trace ( C el ) )
2

I 3 ( C el ) = det ( C el )

The internal variables for these invariants are named solid.I1Cel,

solid.I2Cel, and solid.I3Cel.

• Components of the elastic Green-Lagrange strain tensor εel in the local coordinate
system.
• Principal elastic stretches λel1, λel2, and λel3, which are the square-root of the
eigenvalues of the elastic right Cauchy-Green deformation tensor Cel.

The internal variables for the principal elastic stretches are named
solid.stchelp1, solid.stchelp2, and solid.stchelp3.

• Invariants of the elastic Green-Lagrange strain tensor. Since

1
ε el = --- ( C el – I )
2

the invariants of εel are written in terms of the invariants of Cel:

1
I 1 ( ε el ) = trace ( ε el ) = --- ( I 1 ( C el ) – 3 )
2

1 2 2 1
I 2 ( ε el ) = --- ( I 1 ( ε el ) – trace ( ε el ) ) = --- ( I 2 ( C el ) – 2I 1 ( C el ) + 3 )
2 4

MATERIAL MODELS | 279

 1
I 3 ( ε el ) = det ( ε el ) = --- ( I 3 ( C el ) – I 2 ( C el ) + I 1 ( C el ) – 1 )
8

The internal variables for these invariants are named solid.I1eel,

solid.I2eel, and solid.I3eel.

• When the Nearly incompressible material check box is selected for the Hyperelastic
Material node, the elastic strain energy is decoupled into the volumetric and
isochoric components.:
The volumetric strain energy Wvol, which can be an expression involving the elastic
volume ratio
J el = det ( F el )

The isochoric strain energy, Wiso, as an expression involving the invariants of the
isochoric elastic right Cauchy-Green tensor

I 1 ( C el ) and I 2 ( C el )

or the invariants of the isochoric elastic Green-Lagrange strain

I 1 ( ε el ) , I 2 ( ε el ) , and I 3 ( ε el )

The internal variables for Jel, I 1 ( C el ) , and I 2 ( C el ) are named

solid.Jel, solid.I1CIel, and solid.I2CIel.

The internal variables for I 1 ( ε el ) , I 2 ( ε el ) , and I 3 ( ε el ) are named

solid.I1eIel, solid.I2eIel, and solid.I3eIel.

The strain energy density must not contain any other expressions
involving displacement or their derivatives. This excludes components of
the displacement gradient ∇u and deformation gradient F = ∇u + I
tensors, their transpose, inversions, as well as the global material system
components of C and ε. If they occur, such variables are treated as
constants during symbolic differentiations.

See also the description of the User defined material model in the Solid
Mechanics interface documentation.

280 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

Elastoplastic Material Models
In this section:

• Introduction to Small and Large Plastic Strains

• Plastic Flow for Small Strains
• Isotropic Plasticity
• Yield Function
• Hill Orthotropic Plasticity
• Isotropic Hardening
• Plastic Flow for Large Strains
• Numerical Solution of the Elastoplastic Conditions

INTRODUCTION
Many materials have a distinct elastic regime, in which the deformations are
recoverable and path independent. When the stresses exceed a certain level, the yield
limit, permanent plastic strains will appear.

The elastic part of the constitutive relation can be described by either a Linear Elastic
Material, Nonlinear Elastic Material Models, or by Hyperelastic Material Models.

Elastoplastic material models are common, both when modeling metals and soils.

In geotechnical applications it is common to define compressive stresses as having

positive signs. In COMSOL Multiphysics, the convention is however to always use
positive signs for tensile stresses.

See also Sign Conventions.

DEFINING THE YIELD SURFACE

A yield criterion serves to define the stress condition under which plastic deformation
occurs. Stress paths inside the yield surface result in purely recoverable deformations
(elastic behavior), while paths intersecting the yield surface produces both recoverable
and permanent deformations (plastic strains).

In general, the yield surface can be described as

F = f ( σ ) – fc = 0

MATERIAL MODELS | 281

 where fc can be a constant value (for perfectly plastic materials), or a variable for
strain-hardening materials. The yield surface F is a surface in the space of principal
stresses, in which the elastic regime (F ≤ 0) is enclosed.

For brittle materials, the yield surface represents a failure surface, which is a stress
level at which the material collapses instead of deforms plastically.

Some authors define the yield criterion as f (σ) = fc, while the yield
surface is an isosurface in the space of principal stresses F = 0, which can
2 2
be chosen for numerical purposes as F = f ( σ ) – f c = 0 .

ISOTROPIC PLASTICITY
For isotropic plasticity, the plastic potential Qp is written in terms of at most three
invariants of Cauchy’s stress tensor

Q p ( σ ) = Q p ( I 1 ( σ ), J 2 ( σ ), J 3 ( σ ) )

where the invariants of the stress tensor are

I 1 ( σ ) = trace ( σ )
1
J 2 ( σ ) = --- dev ( σ ):dev ( σ )
2
J 3 ( σ ) = det ( dev ( σ ) )

·
so that the increment of the plastic strain tensor ε p can be decomposed into

∂Q p ∂Q p ∂I 1 ∂Q p ∂J 2 ∂Q p ∂J 3
ε p = λ ---------- = λ  ---------- -------- + ---------- --------- + ---------- ---------
·
∂σ  ∂I 1 ∂σ ∂J 2 ∂σ ∂J 3 ∂σ 

·
The increment in the plastic strain tensor ε p includes in a general case both deviatoric
·
and volumetric parts. The tensor ε p is symmetric given the following properties

∂I 1
-------- = I
∂σ
∂J 2 (3-21)
--------- = dev ( σ )
∂σ
∂J 3
--------- = dev ( σ )dev ( σ ) – 2
--- J 2 I
∂σ 3

282 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

A common measure of inelastic deformation is the effective plastic strain rate, which
is defined as

· 2
ε pe = --- ε· p :ε· p (3-22)
3

The trace of the incremental plastic strain tensor, which is called the volumetric plastic
·
strain rate ε pvol , is only a result of dependence of the plastic potential on the first
invariant I1(σ), since ∂J2/∂σ and ∂J3/∂σ are deviatoric tensors

∂Q p ∂Q p
ε pvol = trace ( ε p ) = λ trace  ---------- = 3λ ----------
· ·
∂σ ∂I 1

For metal plasticity under the von Mises or Tresca criteria, the volumetric plastic strain
rate is always zero because the plastic potential is independent of the invariant I1(σ).
This is known as J2 plasticity.

Incompressible plastic deformation is experimentally observed in metals, but it not the

case for most materials used in geotechnical applications. For instance, a nonzero
volumetric plastic strain is explicitly used in the Cam-Clay material.

• The effective plastic strain and the volumetric plastic strain are available
in the variables solid.epe and solid.epvol.
• In a time dependent analysis, rates of plastic strains can be computed
with expressions like d(solid.epe,TIME).

YIELD FUNCTION
When an associated flow rule is applied, the yield function must be smooth, that is,
continuously differentiable with respect to the stress. In COMSOL Multiphysics, the
following form is used:

F y = φ ( σ ) – σ ys

where σys is the yield stress. The scalar function φ(σ) is called effective stress. The
default form of the effective stress is the von Mises stress, which is often used in metal
plasticity:

3
σ mises = 3J 2 ( σ ) = --- dev ( σ ):dev ( σ )
2

MATERIAL MODELS | 283

 Other expressions can be defined, such as Tresca stress, Hill orthotropic plasticity, or
a user-defined expression.

The Tresca effective stress is calculated from the difference between the largest and the
smallest principal stress

σ tresca = σ p 1 – σ p 3

A user-defined yield function can by expressed in terms of invariants of the stress tensor
such as the pressure (volumetric stress)

1
p = – --- I 1 ( σ )
3

the effective (von Mises) stress σmises, or other invariants, principal stresses, or stress
tensor components.

von Mises Criterion

The von Mises criterion suggests that the yielding of the material begins when the
second deviatoric stress invariant J2 reaches a critical value. This criterion can be
written in terms of the elements of Cauchy’s stress tensor (Ref. 1)

1 2 2 2 2 2 2
J 2 = --- ( ( σ 11 – σ 22 ) + ( σ 22 – σ 33 ) + ( σ 33 – σ 11 ) ) + σ 12 + σ 23 + σ 13 = k 2
6

or equivalently J2 = k .

The von Mises criterion is implemented as

F = 3J 2 – σ ys = 0

where σys is the yield stress level (yield stress in uniaxial tension).

The effective or von Mises stress ( σ mises = 3J 2 ) is available in the

variable solid.mises, where solid is the name of the physics interface
node.

Tresca Criterion
The Tresca yield surface is normally expressed in terms of the principal stress
components

1
--- max ( σ 1 – σ 2 , σ 1 – σ 3 , σ 2 – σ 3 ) = k
2

284 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

The Tresca criterion is a hexagonal prism with its axis equally inclined to the three
principal stress axes. When the principal stresses fulfill σ1 ≥ σ2 ≥ σ3, this criterion is
written as

1
--- ( σ 1 – σ 3 ) = k
2

By using the representation of principal stresses in term of the invariants J2 and the
Lode angle 0 ≤ θ ≤ π/3, this criterion can alternatively be written as

4J π
--- ---------2-  cos ( θ ) – cos  θ + ------  = J 2 sin  θ + --- = k
1 2π
2 3   3   3

or equivalently

π
J 2 cos  θ – --- = k
 6

The maximum shear stress is reached at the meridians (θ = 0 or θ = π/3). The Tresca
criterion can be circumscribed by setting the Lode angle θ = 0, or equivalently, by a
von Mises criterion

3J 2 = 2k

The minimum shear is reached at θ = π/6, so the Tresca criterion can be inscribed by
setting a von Mises criterion

J2 = k

When dealing with soils, the parameter k is also called undrained shear strength.

Tresca criterion can be used with either an associated or non-associated flow rule, in
which case von Mises stress is applied in order to get better numerical performance.

The Tresca effective stress, σtresca = σ1 − σ3, is implemented in the

variable solid.tresca, where solid is the name of the physics interface
node.

MATERIAL MODELS | 285

 θ

Figure 3-4: The upper and lower limits of the Tresca criterion.

Figure 3-5: Classical yield criteria for metals. Tresca criterion (left) and von Mises
criterion (right).

The von Mises and Tresca criteria are independent of the first stress invariant I1 and
are mainly used for the analysis of plastic deformation in metals and ductile materials,
though some researchers also use these criteria for describing fully saturated cohesive
soils under undrained conditions. The von Mises and Tresca criteria belongs to what
researchers call volume preserving or J2 plasticity, as the plastic flow is independent on
the mean pressure.

286 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

PO RO US PLAST ICITY
The description of plastic deformation in soils, porous metals and aggregates has a
main difference with respect to traditional metal plasticity; as the yield function and
plastic potential are not only defined in terms of the deviatoric stress tensor component
(or the deviatoric stress invariant J2), but they also include dependencies on the
hydrostatic pressure.

Shima-Oyane Criterion
Shima and Oyane (Ref. 14) proposed a plastic yield surface for modeling the
compaction of porous metallic structures fabricated by sintering. This criterion can
been applied for powder compaction at both low and high temperatures. The yield
function and associated plastic potential is defined by an ellipsoid in the stress space.
The plastic potential Qp is written in terms of both von Mises effective stress and mean
pressure, and it also considers isotropic hardening due to changes in porosity. The
plastic potential is defined by

σe 2 γ pm
2
Q p ( σ ) =  ------ + α ( 1 – ρ rel )  ------- – ρ rel
m
 σ 0  σ0 

here, σe is the effective stress, σ0 is the yield stress, pm is the pressure, and ρrel is the
relative density, which is related to the porosity φ by ρrel = 1 − φ. The material
parameters α, γ, and m are obtained from curve fitting experimental data. Typical
material parameter values for copper powder are α = 6.2, γ = 1.03, and m = 5.

Gurson Criterion
Gurson criterion (Ref. 15) consists in a pressure dependent yield function to describe
the constitutive response of porous metals, this yield function is derived from the
analytical expression of an isolated void immersed in a continuum medium. The void
volume fraction, or porosity φ is chosen as main variable. The yield function and
associated plastic potential is not an ellipse in the stress space, as in Shima-Oyane
Criterion, but it is defined in terms of the hyperbolic cosine function. The plastic
potential for Gurson criterion reads

σe 2 3p m
Q p ( σ ) =  ------ + 2φ cosh  ----------- – ( 1 + φ )
2
σ0 2σ 0

here, σe is the effective stress, σ0 is the initial yield stress, pm is the pressure, and φ is
the porosity.

MATERIAL MODELS | 287

 Gurson-Tvergaard-Needleman Criterion
Tvergaard and Needleman modified Gurson Criterion for porous plasticity to include
parameters to better fit experimental data (Ref. 16-17). The resulting criterion is called
in the literature Gurson-Tvergaard-Needleman (GTN) criterion. The plastic potential
for GTN criterion reads

σe 2  3q 2 p m
Q p ( σ ) =  ------ + 2q 1 φ e cosh  ----------------- – ( 1 + q 3 φ e )
2
σ0  2σ 0 

here, σe is the effective stress, σ0 is the initial yield stress, pm is the pressure, and φe is
the effective void volume fraction (effective porosity). Typical correction parameter
values are q1 = 1.5, q2 = 1.03, and q3 = q12.

The effective void value fraction (or effective porosity) φe used in the plastic potential
is a function of the current porosity φ and other material parameters:

φ for φ < φ c
φe = φm – φc
φ c + ------------------ ( φ – φ c ) for φ c < φ < φ f
φf – φc

here φc is the critical void volume fraction (critical porosity) at which void coalescence
begins, and φf is the void volume fraction at failure. When the porosity increases up to
value of failure, the effective porosity takes a maximum value of φm; at this high
porosity, the porous material loses the capacity to carry stresses. This maximum
porosity value is derived from other parameters

q 1 + q 12 – q 3
φ m = -----------------------------------
q3

since typically the correction parameters are related as q3 = q12, then φm = 1/q1.

Fleck-Kuhn-McMeeking Criterion
The Fleck-Kuhn-McMeeking criterion (Ref. 19), also called FKM criterion, was
developed to model the plastic yielding of metal aggregates of high porosity. The yield
function and associated plastic potential is derived from expressions for randomly
distributed particles. The criterion is considered relevant for aggregates with porosity
between 10% and 35%. The plastic potential for FKM criterion reads

5 σe 2 2 5 pm 2
Q p =  ------ ----- + --- +  --- ------- – 1
 18 p f 3  3 pf 

288 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

here, σe is the effective stress and pm is the pressure.The flow strength of the material
under hydrostatic loading, pf, is computed from

2 φm – φ
p f = 2.97 ( 1 – φ ) ---------------- σ 0
φm

here, σ0 is the initial yield stress, and φ is the void volume fraction (porosity). The
maximum void volume fraction φm typically takes the value of 36%, the limit of dense
random packing of sintered powder.

FKM-GTN Criterion
The FKM-GTN criterion is a combination of the Fleck-Kuhn-McMeeking Criterion
and Gurson-Tvergaard-Needleman Criterion, intended to cover a wider range of
porosities (Ref. 20-21). For low void volume fractions (porosity lower than 10%), the
GTN model is used and for void volume fractions higher than 25%, the FKM criterion
is used. In the transition zone, a linear combination of both criteria is used.

SOIL PLASTICITY

Mohr-Coulomb Criterion
The Mohr-Coulomb criterion is the most popular criterion in soil mechanics. It was
developed by Coulomb before the Tresca and von Mises criteria for metals, and it was
the first criterion to account for the hydrostatic pressure. The criterion states that
failure occurs when the shear stress and the normal stress acting on any element in the
material satisfy the equation

τ + σ tan φ – c = 0

here, τ is the shear stress, c the cohesion, and φ denotes the angle of internal friction.

With the help of Mohr’s circle, this criterion can be written as

1 1
--- ( σ 1 – σ 3 ) + --- ( σ 1 + σ 3 ) sin φ – c cos φ = 0
2 2

The Mohr-Coulomb criterion defines an irregular hexagonal pyramid in the space of

principal stresses, which generates singularities in the derivatives of the yield function.

MATERIAL MODELS | 289

 ia n
rid
me
an
ile
i
rid
ns

me
Te

s ive
res
omp
C

Figure 3-6: The Mohr-Coulomb criterion. The cone opens toward the compressive axis.

The Mohr-Coulomb criterion can be written in terms of the invariants I1 and J2 and
the Lode angle 0 ≤ θ ≤ π/3 (Ref. 1, Ref. 9) when the principal stresses are sorted as
σ1 ≥ σ2 ≥ σ3. The yield function then reads

J
F y = --- I 1 sin φ + -----2-  ( 1 + sin φ ) cos θ – ( 1 – sin φ ) cos  θ + ------  – c cos φ = 0
1 2π
3 3 3

The tensile meridian is defined when θ = 0 and the compressive meridian when θ = π/3.

Rearranging terms, the Mohr-Coulomb criterion reads

Fy = J 2 m ( θ ) + αI 1 – k = 0

where

m ( θ ) = cos ( θ – π ⁄ 6 ) – 1 ⁄ 3 sin φ sin ( θ – π ⁄ 6 ) , α = sin φ ⁄ 3 , and k = c cos φ

In the special case of frictionless material, ( φ = 0 , α = 0, k = c), the Mohr-Coulomb

criterion reduces to a Tresca’s maximum shear stress criterion, (σ1 − σ3) = 2k or
equivalently

290 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

 π
Fy = J 2 cos  θ – --- – k = 0
6

See also the description of the Mohr-Coulomb material model in the

Solid Mechanics interface documentation.

Drucker-Prager Criterion
The Mohr-Coulomb criterion causes numerical difficulties when treating the plastic
flow at the corners of the yield surface. The Drucker-Prager model neglects the
influence of the invariant J3 (introduced by the Lode angle) on the cross-sectional
shape of the yield surface. It can be considered as the first attempt to approximate the
Mohr-Coulomb criterion by a smooth function based on the invariants I1 and J2
together with two material constants (which can be related to Mohr-Coulomb’s
coefficients)

Fy = J 2 + αI 1 – k = 0

This is sometimes also called the extended von Mises criterion, since it is equivalent to
the von Mises criterion for metals when setting α = 0.

Figure 3-7: The Drucker-Prager criterion. The cone opens toward the compressive axis.

The coefficients in the Drucker-Prager model can be matched to the coefficients in the
Mohr-Coulomb criterion by

2 sin φ 2 3c cos φ
α = ------- ⋅ -------------------------- and k = ---------------------------
3 ( 3 ± sin φ ) ( 3 ± sin φ )

MATERIAL MODELS | 291

 The symbol ± is related to either matching the tensile meridian (positive sign) or the
compressive meridian (negative sign) of Mohr-Coulomb’s pyramid.

The matching at the tensile meridian (θ = 0) comes from setting

m ( 0 ) = ( 3 + sin φ ) ⁄ ( 2 3 )

in the Mohr-Coulomb criterion, and the matching at the compressive meridian (θ = π/

3) from setting

m ( π ⁄ 3 ) = ( 3 – sin φ ) ⁄ ( 2 3 )

C
o

Figure 3-8: The Drucker-Prager criterion showing the tensile and compressive meridians
(inner and outer circles), and the Lode angle compared to the cross section of
Mohr-Coulomb criterion in the π-plane.

In the special case of frictionless material, ( φ = 0 , α = 0, k = 2c ⁄ 3 ), the Drucker

-Prager criterion reduces to the von Mises criterion

J 2 = 2c ⁄ 3

When matching Drucker-Prager criterion to Mohr-Coulomb criterion in 2D

plane-strain applications, the parameters are

tan φ 3c
α = ------------------------------------ and k = ------------------------------------
2 2
9 + 12 tan φ 9 + 12 tan φ

292 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

and when matching both criteria in 2D plane-stress applications, the matching
parameters are:

1 2
α = ------- ⋅ sin φ and k = ------- ⋅ c cos φ
3 3

Dilatation Angle
The Mohr-Coulomb yield criterion is sometimes used with a non-associated plastic
potential. This plastic potential could be either a Drucker-Prager criterion, or the same
Mohr-Coulomb yield function but with a different slope with respect to the
hydrostatic axis, in which case the angle of internal friction is replaced by the dilatation
angle, which is normally smaller (Ref. 5).

Also, when using a Drucker-Prager criterion matched to a Mohr-Coulomb criterion,

the plastic potential could also be non-associated, in which case the difference between
the dilatation angle and the angle of internal friction would result in a yield surface and
plastic potential portrayed by two cones with different angles with respect to the
hydrostatic angle.

Elliptic Cap
The Mohr-Coulomb and Drucker-Prager criteria portray a conic yield surface which
opens in the hydrostatic axis direction. Normally, these soil models are not accurate
above a given limit pressure because real-life materials cannot bear infinite loads and
still behave elastically. A simple way to overcome this problem is to add an elliptical
end-cap on the compressive side to these models.

The elliptic cap is an elliptic yield surface of semi-axes as shown in Figure 3-9. The
initial pressure pa (SI units: Pa) denotes the pressure at which the elastic range
circumscribed by either a Mohr-Coulomb pyramid or a Drucker-Prager cone is not
valid any longer, so a cap surface is added. The limit pressure pb gives the curvature of

MATERIAL MODELS | 293

 the ellipse, and denotes the maximum admissible hydrostatic pressure for which the
material starts deforming plastically. Pressures higher than pb are not allowed

pa pb p

Figure 3-9: Elliptic cap model in Haigh–Westergaard coordinate system.

Note that the sign convention for the pressure is taken from the Structural Mechanics
Module: positive sign under compression, so pa and pb are positive parameters.
Figure 3-9 shows the cap in terms of the variables p and q

q = 3J 2 ( σ ) and p = – I 1 ( σ ) ⁄ 3

In terms of these variables, the equation for the elliptic cap reads

p – pa  2  q  2
 -----------------
- + ----- = 1
 p b – p a  q a

the point (pa, qa) in the Haigh–Westergaard coordinate system is where the elliptic cap
intersects either the Mohr-Coulomb or the Drucker-Prager cone.

Elliptic Cap With Hardening

It is also possible to add isotropic hardening to the cap surface. In this case, the center
of the ellipse is shifted as the volumetric plastic strain increases, also, the size of the
ellipse’s semi-axes grow as hardening evolves. The intersection of the elliptic cap with
the pressure axis is given by

ε pvol
p b = p b0 – K iso log  1 + --------------------
 ε pvol,max

294 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

here, pb0 is the initial value for the limit pressure pb, Kiso is the isotropic hardening
modulus, εpvol the volumetric plastic strain, and εpvol,max the maximum volumetric
plastic strain. Instead of providing the value for the initial pressure pa (SI units: Pa),
the ellipse’s aspect ratio R is entered.

Note that the volumetric plastic strain εpvol is negative in compression, so the limit
pressure pb is increased from pb0 as hardening evolves.

See also the description of the Drucker-Prager material model in the Solid
Mechanics interface documentation.

Matsuoka-Nakai Criterion
Matsuoka and Nakai (Ref. 3) discovered that the sliding of soil particles occurs in the
plane in which the ratio of shear stress to normal stress has its maximum value, which
they called the mobilized plane. They defined the yield surface as

F y = ( 9 + 9μ 2 )I 3 – I 1 I 2 = 0

where the parameter μ = (τ/σn)STP equals the maximum ratio between shear stress
and normal stress in the spatially mobilized plane (STP-plane), and the invariants are
applied over the effective stress tensor (this is the Cauchy stress tensor minus the fluid
pore pressure).

The Matsuoka-Nakai criterion circumscribes the Mohr-Coulomb criterion in dry soils,

when

2 2
μ = ----------- tan φ
3

MATERIAL MODELS | 295

 and φ denotes the angle of internal friction in Mohr-Coulomb criterion.

Figure 3-10: The Matsuoka-Nakai criterion and Mohr-Coulomb criterion in the

principal stress space.

See also the description of the Matsuoka-Nakai material model in the

Solid Mechanics interface documentation.

Lade-Duncan Criterion
The Lade-Duncan criterion was originally developed to model a large volume of
laboratory sample test data of cohesionless soils. This criterion is defined as

3
F y = kI 3 – I 1 = 0

where I1 and I3 are the first and third stress invariants respectively, and k is a parameter
related to the direction of the plastic strain increment in the triaxial plane. The
parameter k can vary from 27 for hydrostatic stress conditions (σ1 = σ2 = σ3), up to a
critical value kc at failure. In terms of the invariants I1, J2,and J3, this criterion can be
written as

F y = J 3 – --- I 1 J 2 +  ------ – --- I 1 = 0

1 1 1 3
3  27 k

The Lade-Duncan criterion can be fitted to the compressive meridian of the

Mohr-Coulomb surface by choosing

k = ( 3 – sin φ ) 3 / ( cos φ 2 ⋅ ( 1 – sin φ ) )

296 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

with φ as the angle of internal friction in Mohr-Coulomb criterion

Figure 3-11: Comparing the Mohr-Coulomb, Matsuoka-Nakai, and Lade-Duncan

criteria when matching the tensile meridian.

The Lade-Duncan criterion does not match the Mohr-Coulomb criterion

(nor the Matsuoka-Nakai criterion) at the tensile meridian.

See also the description of the Lade-Duncan material model in the Solid
Mechanics interface documentation.

Tension Cut-Off
It appears that the Mohr-Coulomb and Drucker-Prager criteria predict tensile
strengths larger than the experimental measurements on soil samples. This discordance
can be mended by the introduction of the Rankine or tension cut-off criterion.

The Rankine criterion states that a material stops deforming elastically when the
biggest principal stress σ1 reaches a maximum tensile stress, also called tension cut-off
limit σt.

In terms of the principal stress, Rankine criterion reads

F = σ1 – σt = 0

For soils and clays, the maximum tensile stress can be estimated from the material
parameters, such as the cohesion c and the friction angle φ . For instance, the tip of the
cone in Mohr-Coulomb criterion is reached when

cos φ
σ 1 = c ------------
sin φ

MATERIAL MODELS | 297

 therefore, the tension cut-off should be chosen such as

cos φ
σ t < c ------------
sin φ

The Mohr-Coulomb criterion together with a tension cut-off is sometimes called

modified Mohr-Coulomb criterion (Ref. 19).

Tension cut-off is also available with Concrete material models.

HILL ORTHOTROPIC PLASTICITY

Hill (Ref. 12, Ref. 13) proposed a quadratic yield function (and associated plastic
potential) in a local coordinate system given by the principal axes of orthotropy ai

Q p = F ( σ 22 – σ 33 ) 2 + G ( σ 33 – σ 11 ) 2 + H ( σ 11 – σ 22 ) 2 +
(3-23)
2 2 2
2Lσ 23 + 2Mσ 31 + 2Nσ 12 – 1

The six parameters F, G, H, L, M, and N are related to the state of anisotropy. As with
isotropic plasticity, the elastic region Qp < 0 is bounded by the yield surface Qp = 0.

Hill demonstrated that this type of anisotropic plasticity is volume preserving, this is,
given the associated flow rule

· ∂Q p
ε p = λ ----------
∂σ

the trace of the plastic strain rate tensor is zero, which follows from the expressions for
·
the diagonal elements of ε p

· ∂Q p
ε p11 = λ ------------ = 2λ ( – G ( σ 33 – σ 11 ) + H ( σ 11 – σ 22 ) )
∂σ 11
· ∂Q p
ε p22 = λ ------------ = 2λ ( F ( σ 22 – σ 33 ) – H ( σ 11 – σ 22 ) )
∂σ 22
· ∂Q p
ε p33 = λ ------------ = 2λ ( – F ( σ 22 – σ 33 ) + G ( σ 33 – σ 11 ) )
∂σ 33

so the plastic volumetric strain rate is zero

298 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

 · · · · ·
ε pvol = trace ( ε p ) = ε p11 + ε p22 + ε p33 = 0

Hill plasticity is an extension of J2 (von Mises) plasticity, in the sense that

it is volume preserving. Due to this assumption, six parameters are needed
to define orthotropic plasticity, as opposed to orthotropic elasticity, where
nine elastic coefficients are needed.

Expressions for the Coefficients F, G, H, L, M, N

Hill noticed that the parameters L, M, and N are related to the yield stress in shear with
respect to the axes of orthotropy ai, thus they are positive parameters

1 1 1
L = ----------------
2
, M = ----------------
2
, N = ----------------
2
2σ ys23 2σ ys31 2σ ys12

Here, σysij represents the yield stress in shear on the plane ij.

The material parameters σys1, σys2, and σys3 represent the tensile yield stress in the
direction, a1, a2, and a3, and they are related to Hill’s parameters F, G, and H as

1-
--------- = G+H
2
σ ys1
1-
--------- = H+F
2
σ ys2
1-
---------
2
= F+G
σ ys3

or equivalently

1 1 - ---------
1
2F = ---------
2
- + ---------
2
– 2-
σ ys2 σ ys3 σ ys1
1
1 - --------- 1-
2G = ---------
2
+ 2 - – ---------
2
σ ys3 σ ys1 σ ys2
1 1 - ---------
1
2H = ---------
2
- + ---------
2
– 2-
σ ys1 σ ys2 σ ys3

MATERIAL MODELS | 299

 Note that at most one of the three coefficients F, G, and H can be negative.

In case of hardening, these coefficients (either Hill’s coefficients or the

shear and tensile yield stresses) are renamed with the “initial” prefix.

In order to define a yield function and plastic potential suitable for isotropic or
kinematic hardening, the average initial yield stress σys0 is calculated from the Hill’s
parameters F, G, and H (this is equivalent to the initial yield stress σys0 in von Mises
plasticity)

1- 2 1  1 - --------- 1 1 -
----------
2
= --- ( F + G + H ) = ---  --------- + 2 - + ---------  (3-24)
σ ys0 3 3  σ ys1 σ ys2 σ 2ys3
2

Defining Hill’s effective stress as (Ref. 13)

2 2
σ hill = σ ys0 ( F ( σ 22 – σ 33 ) 2 + G ( σ 33 – σ 11 ) 2 + H ( σ 11 – σ 22 ) 2
2 2 2
+ 2 ( Lσ 23 + Mσ 31 + Nσ 12 ) )

makes it possible to write the plastic potential in a similar way to von Mises plasticity.

Isotropic hardening is then applied on the average yield stress variable σys0, by using
the plastic potential

Q p = σ hill – σ ys

Here, the average yield stress

σ ys = σ ys0 + σ h ( ε pe )

now depends on the initial yield stress σys0, the hardening function σh, and the
effective plastic strain εpe.

ISOTROPIC HARDENING
Plasticity implements seven different kinds of isotropic hardening models for
elastoplastic materials:

• Perfect plasticity (no isotropic hardening)

• Linear isotropic hardening
• Ludwik

300 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

• Swift
• Voce
• Hockett-Sherby
• User defined

Perfect (or Ideal) Plasticity

For perfect elastoplastic materials, the yield surface is fixed in the space of principal
stresses, and therefore, plastic deformations occur only when the stress path moves on
the yield surface (the regime inside the yield surface is elastic, and stress paths beyond
the yield surface are not allowed).

In this case the plasticity algorithm solves either the associated or non-associated flow
rule for the plastic potential Qp

· ∂Q p
ε p = λ ----------
∂σ

with the yield function

F y = φ ( σ ) – σ ys0

In the settings for plasticity you specify the effective stress φ(σ) for the yield function
from von Mises stress, Tresca stress, Hill effective stress, or a user-defined expression;
and σys0 is the initial yield stress that defines the onset of plastic deformation.

When Large plastic strain is selected as the plasticity model for the
Plasticity node, either the associate or non-associated flow rule is applied
as written in Equation 3-30.

Linear Isotropic Hardening

In this case the plasticity algorithm solves either the associated or non-associated flow
rule for the plastic potential Qp

· ∂Q p
ε p = λ ----------
∂σ

with the yield function

F y = φ ( σ ) – σ ys ( ε pe )

where the yield stress σys(εpe) now depends on the effective plastic strain εpe.

MATERIAL MODELS | 301

 The yield stress σys(εpe) is then a function of the effective plastic strain and the initial
yield stress σys0

σ ys ( ε pe ) = σ ys0 + E iso ε pe

here, the isotropic hardening modulus Eiso is calculated from

1- 1 1
---------- = -------------- – ----
E iso E Tiso E

For linear isotropic hardening, the isotropic tangent modulus ETiso is defined as
(stress increment / total strain increment). A value for ETiso is entered in the isotropic
tangent modulus section for the Plasticity node. The Young’s modulus E is taken from
the parent material (Linear Elastic, Nonlinear Elastic or Hyperelastic material model).
For orthotropic and anisotropic elastic materials, E represents an effective Young’s
modulus.

Ludwik
In Ludwik model for nonlinear isotropic hardening, the yield stress σys(εpe) is defined
by a nonlinear function of the effective plastic strain. Ludwik equation (also called
Ludwik-Hollomon equation) for isotropic hardening is given by the power-law

n
σ ys ( ε pe ) = σ ys0 + kε pe

here, k is the strength coefficient and n is the hardening exponent. Setting n = 1 would
result in linear isotropic hardening.

Swift
For non-saturating materials, the Swift power-law equation relates the initial yield
stress σys0 and the isotropic hardening σh, to the effective plastic strain as

n
σ ys ( ε pe ) = σ ys0 + σ h ( ε pe ) = k ( ε 0 + ε pe )

here, k is the strength coefficient, n is the hardening exponent and ε0 is a reference

strain. Noting that at zero plastic strain the initial yield stress is related to the strength
coefficient and hardening exponent as

n
σ ys ( 0 ) = σ ys0 = kε 0

the yield stress is written as

302 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

 ε pe n
σ ys ( ε pe ) = σ ys0  1 + --------
 ε0 

Voce
Voce rule for nonlinear isotropic hardening is intended for materials that exhibit a
saturating evolution of hardening. The isotropic hardening σh is exponentially related
to the effective plastic strain as

– βε pe
σ h ( ε pe ) = σ sat  1 – e 
 

The yield stress σys(εpe) is then defined as

– βε pe
σ ys ( ε pe ) = σ ys0 + σ h ( ε pe ) = σ ys0 + σ sat  1 – e 


The value of the saturation exponent parameter β determines the saturation rate of
the hysteresis loop for cyclic loading. The saturation flow stress σsat characterizes the
maximum distance by which the yield surface can expand in the stress space. For values
εpe >> 1/β, the yield stress saturates to

σ ys → σ ys0 + σ sat

Hockett-Sherby
Hockett-Sherby rule for nonlinear isotropic hardening is also intended for materials
which yield stress saturates as effective plastic strain increases. It is similar to Voce rule,
but it includes an exponential dependency of the form

– m ε pe
n

σ ys ( ε pe ) = σ ys0 + ( σ ∞ – σ ys0 )  1 – e 


where σ∝ is the steady-state flow stress, m the saturation coefficient and n the saturation
exponent. For values mεpen >> 1 the yield stress saturates to

σ ys → σ ∞

User defined
The yield stress versus the effective plastic strain can be specified with the help of a
hardening curve that could also depend on other variables, such as temperature.

In this case, define the (usually nonlinear) hardening function σh(εpe) such that the
yields stress reads

MATERIAL MODELS | 303

 σ ys ( ε pe ) = σ ys0 + σ h ( ε pe )

The internal variable for the effective plastic strain is named solid.epe.
The effective plastic strain evaluated at Gauss points is named
solid.epeGp.

When Large plastic strain is selected as the plasticity model for the
Plasticity node, either the associate or non-associated flow rule is applied
as written in Equation 3-30.

KINEMATIC HARDENING
There are few options for computing either linear or nonlinear kinematic hardening
for plasticity:

• No kinematic hardening (default)

• Linear kinematic hardening
• Armstrong-Frederik
• Chaboche

For any of the kinematic hardening models, the algorithm solves either the associated
or non-associated flow rule for the plastic potential Qp

· ∂Q p
ε p = λ ----------
∂σ

with the yield function defined as

F y = φ ( σ – σ back ) – σ ys

Here, σys is the yield stress (which may include a linear or nonlinear isotropic
hardening model), and the effective stress φ(σ) is either the von Mises, Tresca, or Hill
stress; or a user-defined expression. The stress tensor used in the yield function is
shifted by what is usually called the back stress, σback.

Linear Kinematic Hardening

The back stress is generally not only a function of the current plastic strain but also of
its history. In the case of linear kinematic hardening, the back stress σback is a linear
function of the plastic strain tensor εp, this is also known as Prager’s hardening rule.

304 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

The implementation of linear kinematic hardening assumes a linear evolution of the
back stress tensor with respect to the plastic strain tensor:

2
σ back = --- C k ε p
3

where the kinematic hardening modulus Ck is calculated from

1- 1 1
------ = ------- – ----
Ck Ek E

The value for Ek is entered in the kinematic tangent modulus section and the Young’s
modulus E is taken from the linear or nonlinear elastic material model. For orthotropic
and anisotropic elastic materials, E represents an averaged Young’s modulus. Note that
some authors define the kinematic hardening modulus as Hk = 2/3Ck.

When kinematic hardening is added, both the plastic potential and the
yield surface are calculated with effective invariants, that is, the invariants
of the tensor defined by the difference between the stress tensor minus
the back-stress, σ eff = σ – σ back . The invariant of effective deviatoric
tensor is named solid.II2sEff, which is used when a von Mises, Tresca
or Hill orthotropic plasticity is computed together with kinematic
hardening.

Armstrong-Frederik Hardening Model

Armstrong and Frederick (Ref. 7) added memory to Prager’s linear kinematic
hardening model. This nonlinear kinematic hardening model allows to capture the
Bauschinger effect and nonlinear behavior by non symmetrical tension-compression
loading.

The nonlinear evolution of the back stress σback is governed by the rate

· 2 · ·
σ back = --- C k ε p – γ k ε pe σ back
3

here, Ck is the kinematic hardening modulus, γk is a kinematic hardening parameter,

and εpe the effective plastic strain. Setting γk = 0 recovers Prager’s rule for linear
kinematic hardening.

To solve this rate, internal degrees of freedom are added to account for the back stress
components. In order to have the same units as used for the plastic strain, the
algorithm solves for the back strain εback as proposed in (Ref. 8), which is related to

MATERIAL MODELS | 305

 the back stress as

2
σ back = --- C k ε back
3

The nonlinear evolution for the back strain reads

· · ·
ε back = ε p – γ k ε pe ε back

Chaboche Hardening Model

Chaboche (Ref. 8) proposed a nonlinear kinematic hardening model based on the
superposition of N back stress tensors

σ back =  σback,i
i

each of these back stress tensors σback,i follows a nonlinear Frederick-Armstrong

kinematic hardening rule

· 2 · ·
σ back,i = --- C i ε p – γ i ε pe σ back,i
3

Practitioners would normally select γk = 0 for one of the back stress equations, thus
recovering Prager’s linear rule for that branch

2
σ back,0 = --- C 0 ε p
3

The back stress tensor σback is then defined by the superposition of N back stress
tensors

N
2
σ back = --- C 0 ε p +
3  σback,i
i=1

As done for Armstrong-Frederik kinematic hardening, the algorithm solves for the
back strain tensors εback,i instead of the back stress tensors. The change of variables is

2
σ back,i = --- C i ε back,i
3

and the nonlinear evolution for the back strain tensors reads

306 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

 · · ·
ε back,i = ε p – γ i ε pe ε back,i

INTRODUCTION TO SMALL AND LARGE PLASTIC STRAINS

There are two implementations of plasticity available in COMSOL Multiphysics. One
is based on the additive decomposition of strains, which is the most suitable approach
in the case of small strains, and the other one is based on the multiplicative
decomposition of the deformation gradient, which is more suitable when large plastic
strains occurs. The additive and multiplicative decomposition of strains is described in
Inelastic Strain Contributions.

When small plastic strain is selected as the plasticity model, an additive

decomposition is used. If the elastic or plastic strains are large, the additive
decomposition might produce incorrect results. As an example, the volume
preservation, which is an important assumption in metal plasticity, will no longer be
respected. The additive decomposition of strains is however widely used both for metal
and soil plasticity.

When Small plastic strain is selected as the plasticity model for the
Plasticity node, and the Include geometric nonlinearity check box is
selected on the study Settings window, a Cauchy stress tensor is used to
evaluate the yield function and plastic potential. The components of this
stress tensor are oriented along the material directions, so it can be viewed
as a scaled second Piola-Kirchhoff stress tensor. The additive
decomposition of strains is understood as the summation of
Green-Lagrange strains.

When large plastic strain is selected as the plasticity model, the total deformation
gradient tensor is multiplicatively decomposed into an elastic deformation gradient
and a plastic deformation gradient.

PLASTIC FLOW FOR SMALL STRAINS

The flow rule defines the relationship between the increment of the plastic strain
·
tensor ε p and the current state of stress, σ, for a yielded material subject to further
loading. When Small plastic strain is selected as the plasticity model for the Plasticity
node, the direction of the plastic strain increment is defined by

· ∂Q p
ε p = λ ----------
∂σ

MATERIAL MODELS | 307

 Here, λ is a positive multiplier (also called the consistency parameter or plastic
multiplier), which depends on the current state of stress and the load history, and Qp
is the plastic potential.

·
The “dot” (for ε p ) means the rate at which the plastic strain tensor
changes with respect to ∂Qp/∂σ. It does not represent a true time
derivative. Some authors call this formulation rate independent
plasticity.

·
The direction of the plastic strain increment ε p is perpendicular to the surface (in the
hyperspace spanned by the stress tensor components) defined by the plastic potential
Qp.

The plastic multiplier λ is determined by the complementarity or Kuhn-Tucker

conditions

λ ≥ 0 , F y ≤ 0 and λF y = 0 (3-25)

where Fy is the yield function. The yield surface encloses the elastic region defined by
Fy < 0. Plastic flow occurs when Fy = 0.

If the plastic potential and the yield surface coincide with each other (Qp = Fy), the
flow rule is called associated, and the rate in Equation 3-26 is solved together with the
conditions in Equation 3-25.

· ∂F y
ε p = λ --------- (3-26)
∂σ

For a non-associated flow rule, the yield function does not coincide with the plastic
potential, and together with the conditions in Equation 3-25, the rate in
Equation 3-27 is solved for the plastic potential Qp (often, a smoothed version of Fy).

· ∂Q p
ε p = λ ---------- (3-27)
∂σ
·
The evolution of the plastic strain tensor ε p (with either Equation 3-26 or
Equation 3-27, plus the conditions in Equation 3-25) is implemented at Gauss points
in the plastic element elplastic.

308 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

PLASTIC FLOW FOR LARGE STRAINS
When Large plastic strain is selected as the plasticity model for the Plasticity node, a
multiplicative decomposition of deformation (Ref. 9, Ref. 10, and Ref. 11) is used,
and the associated plastic flow rule can be written as the Lie derivative of the elastic
left Cauchy-Green deformation tensor Bel:

1 ∂Φ
– --- L ( B el ) = λ ------- B el (3-28)
2 ∂τ

The plastic multiplier λ and the yield function Φ (written in terms of the Kirchhoff
stress tensor τ) satisfy the Kuhn-Tucker condition, as done for infinitesimal strain
plasticity

λ ≥ 0 , Φ ≤ 0 and λΦ = 0

The yield function Φ in Ref. 9 and Ref. 10 was written in terms of

Kirchhoff stress τ and not Cauchy stress σ because the authors defined the
plastic dissipation with the conjugate energy pair τ and d, where d is the
rate of strain tensor.

The Lie derivative of Bel is then written in terms of the plastic right Cauchy-Green rate

·
L ( B el ) = FC p– 1 F T (3-29)

By using Equation 3-28 and Equation 3-29, the either associated or non-associated
plastic flow rule for large strains is written as (Ref. 10)

1 · ∂Q p
– --- FC p– 1 F T = λ ---------- B el (3-30)
2 ∂τ

together with the Kuhn-Tucker conditions for the plastic multiplier λ and the yield
function Fy

λ ≥ 0 , F y ≤ 0 and λF y = 0 (3-31)

For the associated flow rule, the plastic potential and the yield surface coincide with
each other (Qp = Fy), and for the non-associated case, the yield function does not
coincide with the plastic potential.

In COMSOL Multiphysics, the elastic left Cauchy-Green tensor is written in terms of

the deformation gradient and the right Cauchy-Green tensor, so Bel = FCp−1FT. The
flow rule then reads

MATERIAL MODELS | 309

 · ∂Q p
C p– 1 = – 2λ F – 1 ---------- FC p– 1 (3-32)
∂τ

The plastic flow rule is then solved at Gauss points in the plastic element elplastic
for the inverse of the plastic deformation gradient Fp−1, so that the variables in
Equation 3-32 are replaced by

· · ·
C p– 1 = F p– 1 F p– T , C p–1 = F p– 1 F p–T + F p–1 F p– T and B el = FF p– 1 F p–T F T

After integrating the flow rule in Equation 3-32, the plastic Green-Lagrange strain
tensor is computed from the plastic deformation tensor

1 T
ε p = --- ( F p F p – I )
2

and the elastic Green-Lagrange strain tensor is computed from the elastic deformation
gradient tensor Fel = FFp−1

1 T
ε el = --- ( F el F el – I )
2

When Large plastic strain is selected as the plasticity model for the
Plasticity node, the effective plastic strain variable is computed as the true
effective plastic strain (also called Hencky or logarithmic plastic strain).

When either Large plastic strain or Small plastic strain is selected as the
plasticity model for the Plasticity node, the out-of-plane shear strain
components are not computed in 2D, neither for plane stress nor plane
strain assumption.

NUMERICAL SOLUTION OF THE ELASTOPLASTIC CONDITIONS

A backward Euler discretization of the pseudo-time derivative is used in the plastic flow
rule. For small plastic strains, this gives

∂Q p
ε p – ε p, old = Λ ----------
∂σ

where “old” denotes the previous time step and Λ = λΔt, where Δt is the pseudo-time
step length.

310 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

For large plastic strains, Equation 3-32 is numerically solved with the so-called
exponential mapping technique

∂Q p
C p– 1 = F –1 exp  – 2Λ ---------- FC p,
–1
∂σ old

where C p–1 = F p– 1 F p–T and C p,

–1 –1 –T
old = F p, old F p, old .

For each Gauss point, the plastic state variables (εp or Fp−1, depending on whether
small strain or large strain plasticity is selected) and the plastic multiplier, Λ, are
computed by solving either of the above time-discretized flow rules together with the
complementarity conditions

Λ ≥ 0, F y ≤ 0, ΛF y = 0

This is done as follows (Ref. 9):

1 Elastic-predictor: Try the elastic solution εp = εp,old (or F p– 1 = F p,

–1
old ) and Λ = 0.
If this satisfies Fy ≤ 0 it is done.
2 Plastic-corrector: If the elastic solution does not work (this is Fy > 0), solve the
nonlinear system consisting of the flow rule and the equation Fy = 0 using a damped
Newton method..

The numerical tolerance to fulfill the condition Fy = 0 is given in SI units

of Pascals, and it depends on the initial yield stress (in case of plasticity and
porous plasticity) or it is defined in terms of other material parameter (for
soil plasticity). This numerical tolerance is 0.1% the value defined in the
variable item.tol., where item is the name of the node.

ENERGY DISSIPATION
Since plasticity is an inelastic process, the dissipated energy density can be calculated
by integrating the pseudo-rate given by

· · ∂Q p
W p = σ:ε p = λσ: ----------
∂σ

As plasticity is rate independent, the plastic dissipation density Wp is obtained after

integrating an extra variable in the plastic flow rule.

MATERIAL MODELS | 311

 The total energy dissipated by plasticity in a given volume can be calculated by a
volume integration of the plastic dissipation density Wp.

When the Calculate dissipated energy check box is selected, the plastic
dissipation density is available under the variable solid.Wp and the total
plastic dissipation under the variable solid.Wp_tot.

Failure Criteria for Concrete, Rocks, and Other Brittle Material

In this section:

• Bresler-Pister Criterion
• Willam-Warnke Criterion
• Ottosen Criterion
• Original Hoek-Brown Criterion
• Generalized Hoek-Brown Criterion

BRESLER-PISTER CRITERION
The Bresler-Pister criterion (Ref. 2) was originally devised to predict the strength of
concrete under multiaxial stresses. This failure criterion is an extension of the
Drucker-Prager criterion to brittle materials and can be expressed in terms of the stress
invariants as

2
Fy = J2 + k1 I1 + k2 I1 + k3 = 0 (3-33)

here, k1, k2, and k3 are material parameters.

This criterion can also be written (Ref. 17) in term of the uniaxial compressive strength
fc and the octahedral normal and shear stresses

σ oct = I 1 /3 and τ oct = 2J 2 /3

so Equation 3-33 simplifies to

τ oct σ oct σ oct

2
F y = --------- – a + b ---------- – c ---------
- = 0
fc fc f c2

312 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

Here, the parameters a, b, and c are obtained from the uniaxial compression, uniaxial
tension, and biaxial compression tests, respectively. The octahedral normal stress σoct
is considered positive when tensile, and fc is taken positive.

WIL LA M -WA R N KE C R IT ERI O N

The Willam-Warnke criterion (Ref. 10) is used to predict failure in concrete and other
cohesive-frictional materials such as rock, soil, and ceramics. Just as the Bresler-Pister
criterion, it depends only on three parameters. It was developed to describe initial
concrete failure under triaxial conditions. The failure surface is convex, continuously
differentiable, and is fitted to test data in the low compression range. The material is
considered perfect elastoplastic (no hardening).

The original “three-parameter” Willam-Warnke failure criterion was defined as

3 τ--------
--- - + r ( θ )   --- – ---- σ oct – 1 = 0
oct 1 1
Fy = (3-34)
5 fc ft fb

where fc is the uniaxial compressive strength, ft is uniaxial tensile strength, and fb is

obtained from the biaxial compression test. All parameters are positive. The octahedral
normal and shear stresses are defined as usual

σ oct = I 1 /3 , and τ oct = 2J 2 /3

This can be written as

--- r ( θ )  ------ I 1 – f c = 0

1
Fy = J2 + 5
2 3χ

with

χ = fb ft / ( fb fc – ft fc )

The function r(θ) describes the segment of an ellipse on the octahedral plane when
0 ≤ θ ≤ π ⁄ 3 . By using the Lode angle θ, the dimensionless function r(θ) is defined as

2
2r c ( r c2 – r t2 ) cos θ + r c ( 2r t – r c ) 4 ( r c2 – r t2 ) cos θ + 5r t2 – 4 r c r t
r ( θ ) = --------------------------------------------------------------------------------------------------------------------------------------------------------------
2
-
4 ( r c2 – r t2 ) cos θ + ( r c – 2r t ) 2

Here, the tensile and compressive meridian rt and rc are defined in terms of the positive
parameters fc, fb, and ft:

MATERIAL MODELS | 313

 fb ft
rt = 6
--- ----------------------------
5 2f b f c + f t f c

fb ft
rc = 6
--- -------------------------------------------
5 3f b f t + f b f c – f t f c

The function r(θ) can be interpreted as the friction angle which depends on the Lode
angle (Ref. 10).

π/6

rc
r θ
rt

Figure 3-12: The deviatoric section of Willam-Warnke failure criterion.

OTTOSEN CRITERION
The Ottosen criterion is a four-parameter failure criterion proposed for short-time
loading of concrete. It corresponds to a smooth convex failure surface with curved
meridians, which is open in the negative (compressive) direction of the hydrostatic
axis. The curve in the pi-plane changes from almost triangular to a more circular shape
with increasing hydrostatic pressure. The criterion is in agreement with experimental
results over a wide range of stress states, including both triaxial tests along the tensile
and the compressive meridian and biaxial tests (Ref. 18).

The Ottosen criterion is commonly written as (Ref. 17, Ref. 18):

a
F y = ---- J 2 + λ ( θ ) J 2 + bI 1 – f c = 0
fc

314 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

In this formulation, the parameters a>0, b>0, k1>0, and k2>0 are dimensionless, and
fc>0 is the uniaxial compressive strength of concrete (with a positive sign). The
function λ ( θ ) > 0 (dimensionless) is defined as


 k cos  1 --- acos ( k 2 cos ( 3θ ) ) J3 > 0
 1 3 

λ(θ) = 

 π 1 
 k 1 cos  --3- – --3- acos ( – k 2 cos ( 3θ ) ) J 3 < 0,


The parameter k1 is called the size factor. The parameter k2 (also called shape factor)
is positive and bounded to 0 ≤ k2 ≤ 1(Ref. 17, Ref. 18).

Typical values for these parameters are obtained by curve-fitting the uniaxial
compressive strength fc, uniaxial tensile strength ft, and from the biaxial and triaxial
data (for instance, a typical biaxial compressive strength of concrete is 16% higher than
the uniaxial compressive strength). The parameters fc, fb, and ft are positive.
TABLE 3-3: TYPICAL PARAMETER VALUES FOR OTTOSEN FAILURE CRITERION ( Ref. 18).
ft/ fc a b k1 k2 lt lc
0.08 1.8076 4.0962 14.4863 0.9914 14.4725 7.7834
0.10 1.2759 3.1962 11.7365 0.9801 11.7109 6.5315
0.12 0.9218 2.5969 9.9110 0.9647 9.8720 5.6979

The compressive and tensile meridians (as defined in the Willam-Warnke criterion) are

1 1
r c = ----- = -------------------
λ c λ ( π/3 )

1 1
r t = ----- = -----------
λt λ ( 0 )

For concrete, the ratio

λ c /λ t = r t /r c

MATERIAL MODELS | 315

 normally lies between 0.54~0.58.

The Ottosen criterion is equivalent to Drucker-Prager when a = 0 and

λ = constant.

ORIGINAL HOEK-BROWN CRITERION

The Hoek-Brown criterion is an empirical type of model which is commonly used
when dealing with rock masses of variable quality. The Hoek-Brown criterion is widely
used within civil engineering and is popular because the material parameters can be
estimated based on simple field observations together with knowledge of the uniaxial
compressive strength of the intact rock material. The Hoek–Brown criterion is one of
the few nonlinear criteria widely accepted and used by engineers to estimate the yield
and failure of rock masses. The original Hoek-Brown failure criterion states (Ref. 5)

σ 1 = σ 3 + mσ c σ 3 + sσ c2

where σ1 ≥ σ2 ≥ σ3 ≥ 0 are the principal stresses at failure (as defined in geotechnical

engineering; that is, an absolute value), σc is the uniaxial compressive strength of the
intact rock (positive parameter), and m and s are positive material parameters.

If the expression is converted into to the sign convention for principal stresses in the
Structural Mechanics Module, it becomes

σ 1 = σ 3 + – mσ c σ 1 + sσ c2

with σc, m, and s positive material parameters. (In this case, note that σ1 < sσc/m).

As developed originally, there is no relation between the parameters m and s and the
physical characteristics of a rock mass measured in laboratory tests. However, for intact
rock, s = 1 and m = mi, which is measured in a triaxial test.

For jointed rock masses, 0 ≤ s < 1 and m < mi. The parameter m usually lies in the
range 5 < m < 30 (Ref. 7).
TABLE 3-4: CHARACTERISTIC VALUES FOR DIFFERENT ROCK TYPES

m ROCK TYPE

5 Carbonate rocks, dolomite, limestone

10 Consolidated rocks, mudstone, shale
15 Sandstone

316 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

TABLE 3-4: CHARACTERISTIC VALUES FOR DIFFERENT ROCK TYPES

m ROCK TYPE

20 Fine-grained rocks
25 Coarse grained rocks
The Hoek-Brown criterion can be written in terms of the invariants I1 and J2 and the
Lode angle 0 ≤ θ ≤ π/3, so

π σ1
F y = 2 J 2 sin  θ + --- – σ c s – m ------ = 0
3 σc

GENERALIZED HOEK-BROWN CRITERION

The generalized Hoek-Brown criterion was developed in order to fit the Geological
Strength Index (GSI) classification of isotropic rock masses (Ref. 6). A new
relationship between GSI, m, s and the newly introduced parameter a was developed,
to give a smoother transition between very poor quality rock masses (GSI < 25) and
stronger rocks

σ1 a
σ 1 – σ 3 = σ c  s – m b ------
 σc

In terms of the invariants J2 and the Lode angle 0 ≤ θ ≤ π ⁄ 3 this equals

π σ1 a
F y = 2 J 2 sin  θ + --- – σ c  s – m b ------ = 0
 3   σc

where σ1 ≥ σ2 ≥ σ3 are the principal stresses (using the Structural Mechanics Module
conventions) of the effective stress tensor (this is, the stress tensor minus the fluid pore
pressure).

The positive parameter mb is a reduced value of the material constant mi:

m b = m i exp  --------------------------
GSI – 100
 28 – 14D 

s and a are positive parameters for the rock mass given by the following relationships:

s = exp  --------------------------
GSI – 100
 9 – 3D 

a = --- + ---  exp  ------------- – exp  --------- 

1 1 – GSI – 20
2 6 15 3

MATERIAL MODELS | 317

 The disturbance factor D was introduced to account for the effects of stress relaxation
and blast damage, and it varies from 0 for undisturbed in-situ rock masses to 1 for very
damaged rock masses.
TABLE 3-5: DISTURBANCE FACTOR IN ROCK MASSES

D DESCRIPTION OF ROCK MASS

0 Undisturbed rock mass

0~0.5 Poor quality rock mass
0.8 Damaged rock mass
1.0 Severely damaged rock mass

Cam-Clay Material Model

ABOUT THE CAM-CLAY MATERIAL MODEL

The Cam-clay material model was developed at the University of Cambridge in the
1970s, and since then it has experienced different modifications. The modified
Cam-clay model is the most commonly used due to the smooth yield surface, and it is
the one implemented in the Geomechanics Module.

The Cam-clay model is a so-called critical state model, where the loading and
unloading of the material follows different trajectories in stress space. The model also
features hardening and softening of clays. Different formulations can be found in
textbooks about these models (see Ref. 13, Ref. 14, and Ref. 15).

The yield function is written in terms of the variables

q = 3J 2 ( σ ) and p = – I 1 ( σ ) ⁄ 3

Following the Structural Mechanics Module sign convention:

F y = q 2 + M 2 ( p – p c )p = 0

This is an ellipse in p-q plane, with a cross section independent of Lode angle and
smooth for differentiation. Note that p, q, and pc are always positive variables.

The material parameter M > 0 defines the slope of a line in the p-q space called critical
state line, and it can be related to the angle of internal friction φ in the
Mohr-Coulomb criterion

6 sin ( φ )
M = --------------------------
3 – sin ( φ )

318 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

 q

pc/2

Figure 3-13: Modified Cam-clay surface in the pq-plane. The ellipse circumscribes a
nonlinear elastic region.

In the Cam-clay model, hardening is controlled by the consolidation pressure pc,

which depends exponentially on the volumetric plastic strain εpl,vol.

– B pl ε pl, vol
p c = p c0 e (3-35)

The volumetric plastic strain is available in the variable solid.epvol and

the consolidation pressure in the variable solid.Pc.

Here, the parameter pc0 is the initial consolidation pressure, and the exponent Bpl is a
parameter which depends on the initial void ratio e0, the swelling index κ, and the
compression index λ:

1 + e0
B pl = ---------------
λ–κ

The initial void ratio, the compression index, and the swelling index are all positive
parameters and must fulfill

0 < κ < λ , so B pl > 0

MATERIAL MODELS | 319

 The void ratio e is the ratio between pore volume and solid volume. It can
be written in terms of the porosity φ as e = φ /(1− φ ).

In this formulation, the compression index λ is the slope of the virgin isotropic
consolidation line, and κ is the slope of the rebound-reloading line (also called
loading-reloading line) in the e versus ln(p) plane.

e
N
λ
κ

ln(prefN) ln(pC0) ln(p)

Figure 3-14: Slopes of the virgin isotropic consolidation line and rebound-reloading line
in the e vs. ln(p) plane. The reference void ratio N is measured at the reference pressure
prefN.

If an Initial Stress and Strain feature subnode is added to the Cam-clay

material, the initial consolidation pressure pc0 must be made equal or
bigger than one third of minus the trace of the initial stress tensor,
otherwise the initial stress state is outside the Cam-clay ellipse.

VO L U M E T R I C E L A S T I C B E H AV I O R
The stress-strain relation beyond the elastic range is of great importance in soil
mechanics. For additive decomposition of strains, Cauchy’s stress tensor is written as

σ – σ 0 = C: ( ε – ε inel )

Here, σ is the Cauchy stress tensor, ε is the total strain tensor, εinel is the inelastic strain
tensor, σ0 is the initial stress tensor, and C is the fourth-order elasticity tensor.

320 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

For a linear elastic material, the trace of the Cauchy’s stress tensor is linearly related to
the volumetric elastic strain (the trace of the elastic strain tensor) by the elastic bulk
modulus

p = – I 1 ( σ ) ⁄ 3 = p 0 – Kε el,vol

here p0 = −trace(σ0)/3 is the trace of the initial stress tensor σ0, and K is the bulk
modulus, a constant parameter independent of the stress or strain.

The elastic volumetric strain is available in the variable solid.eelvol.

The modified Cam-clay model introduces a nonlinear relation between stress and
volumetric elastic strain

– B el ε el, vol 1 + e0
p – p0 = –K0 e with B el = ---------------
κ

and K0 a reference bulk modulus. This formulation gives a tangent bulk modulus
KT = −Bel(p−p0). The reference bulk modulus is calculated from the initial
consolidation pressure pc0, and the void ratio at reference pressure N.

HARDENING AND SOFTENING

The yield surface for the modified Cam-clay model reads

F y = q 2 + M 2 ( p – p c )p = 0

The associated flow rule (Qp = Fy) and the yield surface written in terms of these two
invariants, Fy(I1, J2), gives a rate equation for the plastic strain tensor calculated from
the derivatives of Fy with respect to the stress tensor σ

∂Q p ∂Q p ∂I 1 ∂Q p ∂J 2
ε p = λ p ---------- = λ p  ---------- -------- + ---------- ---------
·
∂σ  ∂I 1 ∂σ ∂J 2 ∂σ 

Here, λp means the plastic multiplier, see Plastic Flow for Small Strains
and Isotropic Hardening.

·
The plastic strain rate tensor ε p includes both deviatoric and isotropic parts. Note that

MATERIAL MODELS | 321

 ∂I 1 ⁄ ∂σ = I and ∂J 2 ⁄ ∂σ = dev ( σ )

These relations can be used for writing the plastic flow as

1 ∂Q p ∂Q p 3
ε p = λ p  – --- ---------- I + ---------- ------- dev ( σ ) = λ p  – --- M 2 ( 2p – p c ) I + 3dev ( σ )
· 1
 3 ∂p ∂q 2q   3 

·
The trace of the plastic strain rate tensor (the volumetric plastic strain rate ε pvol )
then reads

· ·
ε pvol = trace ( ε p ) = λ p ( – M 2 ( 2p – p c ) )

This relation explains the reason why there is isotropic hardening for p > pc/2 and
isotropic softening for p < pc/2. So the volumetric plastic strain can either increase or
decrease.

In the Cam-clay model, the hardening is controlled by the consolidation pressure

variable p c ( ε pvol ) as a function of volumetric plastic strain, as written in
Equation 3-35. Hardening introduces changes in the shape of the Cam-clay ellipse,
since its major semi-axis depends on the value of pc.

INCLUDIN G PO RE FLUID PRESSURE

When a pore fluid pressure pf is added to the Cam-clay material, the yield surface is
shifted on the p axis

Fy = q 2 + M 2 ( p – pf – pc ) ( p – pf ) = 0

The quantity p − pf is normally regarded as the effective pressure, or effective stress,

which should not be confused with von Mises stress.

See also the description of the Cam-Clay Material material model in the
Solid Mechanics interface documentation.

322 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

Creep and Viscoplasticity

ABOUT CREEP

In the literature, the terms viscoplasticity and creep are often used
interchangeably to refer to the class of problems related to
rate-dependent plasticity.

Creep is an inelastic time-dependent deformation that occurs when a material is

subjected to stress (typically much less than the yield stress) at sufficiently high
temperatures.

The creep strain rate, in a general case, depends on stress, temperature, and time,
usually in a nonlinear manner:

·
ε cr = F cr ( σ, T, t )

It is often possible to separate these effects as shown in this equation:

F cr ( σ, t, T ) = f 1 ( σ )f 2 ( T )f 3 ( t )

Experimental data shows three types of behavior for the creep strain rate at constant
stress as function of time. Researchers normally subdivide the creep curve into three
regimes, based on the fact that many different materials show similar responses:

• In the initial primary creep regime (also called transient creep) the creep strain rate
decreases with time to a minimum steady-state value.
• In the secondary creep regime the creep strain rate is almost constant. This is also
called steady-state creep.
• In the tertiary creep regime the creep strain increases with time until a failure
occurs.

When this distinction is assumed, the total creep rate can be additively split into
primary, secondary, and tertiary creep rates

·
ε cr = F cr1 + F cr2 + F cr3

In most cases, Fcr1 and Fcr3 depend on stress, temperature and time, while secondary
creep, Fcr2, depends only on stress level and temperature. Normally, secondary creep

MATERIAL MODELS | 323

 is the dominant process. Tertiary creep is seldom important because it only accounts
for a small fraction of the total lifetime of a structure.

εcr

σ1

p
σ2

σ1 > σ2

Figure 3-15: Uniaxial creep as a function of logarithmic time.

Creep and Viscoplastic Material Models

Despite the fact that the creep response of a given material is related to its atomic
structure, a macroscopic (continuum mechanics) description is normally appropriate
for modeling scientific and engineering problems.

In COMSOL Multiphysics, there are several creep models. These models can be split
into two main groups. One set of models are more general, and you will have to
express the creep rate yourself, based on other variables as a stress tensor and
temperature. These models are:

• Creep Potential
• Volumetric Creep
• Deviatoric Creep
• User-Defined Creep

In addition to the basic models for creep described above, there are also predefined
material models for creep in metals and crystalline solids:

• Norton Law (Power law)

• Norton-Bailey Law

324 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

• Garofalo Law (Hyperbolic Sine Law)
• Navarro-Herring Creep (Diffusional Creep)
• Coble Creep (Diffusional Creep)
• Weertman Creep (Dislocation Creep)
• Anand Viscoplastic Model
• Chaboche Viscoplastic Model
• Perzyna Viscoplastic Model

All creep models are contributing subnodes to the a basic material model like Linear
Elastic Material and they can be combined with any other subnodes, such as Plasticity
or Thermal Expansion to create more advanced models. They can also be combined with
each other to model several creep mechanisms acting at the same time.

CREEP POTENTIAL
Some authors use a creep potential to describe the secondary creep rate, so that the
creep rate is written in a way similar to the flow rule for plasticity:

· ∂Q cr
ε cr = η ----------- and η > 0
∂σ

Here, Qcr is a user-defined creep potential, which is normally written in terms of

invariants of the stress tensor.

Volumetric creep is obtained when the creep potential depends only on the first
invariant of Cauchy stress tensor, I1(σ), since

∂Q cr ∂Q cr
----------- = ----------- I
∂σ ∂I 1

This is equivalent to that the creep potential would depend on the pressure p = −I1/3.

When the creep potential depends only on the second deviatoric invariant of Cauchy
stress tensor, J2(σ), the deviatoric creep model is obtained since

∂Q cr ∂Q cr
----------- = ----------- dev ( σ )
∂σ ∂J 2

This is equivalent to that the creep potential would depend on the effective stress
σe = √3J2.

MATERIAL MODELS | 325

 When (in SI units) the creep potential, Qcr, is given in units of Pa, the rate multiplier
η is given in units of 1/s.

VO L U M E T R I C C R E E P
The creep strain rate is calculated by solving the rate equation

· 1
ε cr = --- F cr I
3

so that the creep rate tensor is a diagonal tensor. The trace of the creep rate tensor, the
volumetric creep strain rate, equals the user input Fcr

·
trace ( ε cr ) = F cr (3-36)

The creep rate, Fcr, usually depends on the first invariant of Cauchy stress I1(σ) or the
pressure p = −I1/3, in addition to the temperature and other material parameters.

Volumetric creep is not generally used to model creep in metals, but it is commonly
used to model creep in soils or other geological materials.

DEVIATORIC CREEP
The creep strain rate is calculated by solving the rate equation

·
ε cr = F cr n D

Here, nD is a deviatoric tensor coaxial to the stress tensor.

The creep rate, Fcr, normally depends on the second deviatoric invariant of the stress
J2(σ) or the effective or von Mises (effective) stress σe, in addition to the temperature
and other material parameters.

The deviatoric tensor nD is defined as

3 dev ( σ )
n D = --- ----------------- (3-37)
2 σe

The resulting creep strain rate tensor is also deviatoric, since trace (nD) = 0

·
trace ( ε cr ) = F cr trace ( n D ) = 0

Given the property

3
n D :n D = ---
2

326 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

the effective creep strain rate equals the absolute value of the user input Fcr

· 2
ε cr,e = --- ε· cr :ε· cr = F cr
3

Deviatoric creep is very popular to model creep in metals and alloys. For example,
Norton’s law is a deviatoric creep model.

The effective creep strain and the effective creep strain rate are available in
the variables solid.ece and solid.ecet.

USER-DEFINED CREEP
The creep strain rate is calculated by solving the rate equation

·
ε cr = F cr

where Fcr is a user-defined symmetric tensor field.

NO R T ON LAW (POWER L AW)

The most common model for secondary creep is the Norton equation where the creep
strain rate is proportional to a power of the effective stress, σe:

· n
ε cr ∝ σ e

This is normally true at intermediate to high stress levels and at absolute temperatures
of T/Tm > 0.5, where Tm is the melting temperature (that is, the temperature in the
solid is at least as high as half the melting temperature Tm). An “Arrhenius type”
temperature dependency can also be included. It is defined by

· – Q ⁄ RT
ε cr ∝ e

where Q is the activation energy (SI unit: J/mol), R is the gas constant, and T is the
absolute temperature (SI unit: K).

Norton creep is a deviatoric temperature-dependent creep model, with a creep rate

equation written as

Q
σ e n – --------
-
ε cr = A  --------- e
· RT D
n (3-38)
σ ref

MATERIAL MODELS | 327

 Here, A is the creep rate coefficient (SI unit: 1/s), n is the stress exponent
(dimensionless), σref a reference stress level (SI unit: Pa), and nD is a deviatoric tensor
coaxial to the stress tensor as defined in Equation 3-37.

• For a discussion about how to convert common creep data equations

into the form used in COMSOL Multiphysics, see Converting
Between Different Creep Data Representations
• See also the description of the Norton material model in the Solid
Mechanics interface documentation.

NOR TON-BAILEY LAW

A common model for modeling primary and secondary creep together is the so-called
Norton-Bailey (or Bailey-Norton) model. Here, the creep strain is proportional to a
power of time and to a power of the effective stress

n
ε cr ∝ σ e t m

which for the creep strain rate becomes a time hardening formulation of Norton’s law.
Differentiating with respect to time will give the rate form.

· n
ε cr ∝ σ e mt m – 1

Norton-Bailey creep is a deviatoric temperature-dependent creep model, furbished

with either a time-hardening or a strain-hardening primary creep model. The creep
rate equation for the time-hardening model used in COMSOL Multiphysics is written
as

t + t shift m – 1
ε cr = F cr m  -------------------
·
nD (3-39)
 t ref 

where nD is a deviatoric tensor coaxial to the stress tensor as defined in Equation 3-37,
and Fcr is expressed as in the Norton model:

Q
σ e n – --------
-
F cr = A  --------- e
RT
(3-40)
 σ ref

Here, A is the creep rate coefficient (SI unit: 1/s), n is the stress exponent
(dimensionless), σref is a reference stress level (SI unit: Pa), tref and tshift are the

328 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

reference and shift times (SI unit: s), and m is the time-hardening exponent
(dimensionless).

The strain-hardening variant of this creep law is implemented as

m–1
--------------
·  ε cr,e + ε shift m
ε cr = F cr m  --------------------------- nD (3-41)
 t ref F cr 

where εcr,e is the effective creep strain, and εshift is the effective creep strain shift.

The time and frequency shifts in Equation 3-39 and Equation 3-41 serve two
purposes:

• They can be used to initialize a study where some hardening has already taken place.
• The strain rate expressions actually predicts an infinite creep rate at t = 0, unless a
shift is used. This singularity is weak in the sense that the time integral is well
defined, but it will cause problems for the numerical solution. You can then add a
small shift to overcome this problem.

• For a discussion about how to convert common creep data equations

into the form used in COMSOL Multiphysics, see Converting
Between Different Creep Data Representations
• See also the description of the Norton-Bailey material model in the
Solid Mechanics interface documentation.

GAROFALO LAW (HYPERBOLIC SINE LAW)

At very high stress levels, the creep rate is proportional to the exponential of the
effective stress

ασ e
·
ε cr ∝ e

Garofalo showed (Ref. 8, Ref. 9) that the power-law and exponential creep are limiting
cases for the general empirical expression

· n
ε cr ∝ ( sinh ( ασ e ) )

This equation reduces to a power-law (Norton law) for ασe < 0.8 and approaches
exponential creep for ασe > 1.2, where 1/α is a reference effective stress level.

MATERIAL MODELS | 329

 Garofalo creep is also a deviatoric creep model with a creep rate proportional to the
hyperbolic sine function. It can also be augmented by an “Arrhenius type” temperature
dependency such that

· – Q ⁄ RT
ε cr ∝ e

where Q is the activation energy (SI unit: J/mol), R is the gas constant, and T is the
absolute temperature (SI unit: K). The complete creep rate equation as used in
COMSOL Multiphysics then reads

Q
σ e n – -------- -
ε cr = A  sinh  ---------  e
· RT D
n
σ ref

where, A is the creep rate (SI unit: 1/s), n is the stress exponent (dimensionless), and
σref a reference effective stress level (SI unit: Pa). nD is a deviatoric tensor coaxial to
the stress tensor as defined in Equation 3-37.

• For a discussion about how to convert common creep data equations

into the form used in COMSOL Multiphysics, see Converting
Between Different Creep Data Representations
• See also the description of the Garofalo (hyperbolic sine) material
model in the Solid Mechanics interface documentation.

NAVARRO-HERRING CREEP (DIFFUSIONAL CREEP)

At low stress levels and high temperatures, Navarro and Herring (Ref. 6, Ref. 7)
independently derived an expression for the creep rate as a function of atomic diffusion

· 7D v b 3
ε cr = -----------------2- σ e n D
k B Td

Here, d is the grain diameter, Dv is the volume diffusivity through the grain interior,
b is Burgers vector, kB is the Boltzmann’s constant, and T is the absolute temperature.
nD is a deviatoric tensor coaxial to the stress tensor as defined in Equation 3-37.

See also the description of the Navarro-Herring material model in the

Solid Mechanics interface documentation.

330 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

COBLE CREEP (DIFFUSIONAL CREEP)
Coble creep (Ref. 6, Ref. 7) is closely related to Navarro-Herring creep but takes into
account the ionic diffusivity along grain boundaries Dgb

· 50D gb b 4
ε cr = ----------------------
- σ nD
k B Td 3 e

Coble creep is more sensitive to grain size than Navarro-Herring creep.

See also the description of the Coble material model in the physics
interface documentation.

WE E R T MA N C RE EP ( D I S L OC AT I ON C REE P )
At intermediate to high stress levels and temperatures T/Tm > 0.5, the creep
mechanism is assumed to be diffusion-controlled movements of dislocations in the
crystal lattices (Ref. 7)

σe n
ε cr = ----------- σ ref  --------- n D
· Db
kB T  σ ref

where and nD is a deviatoric tensor coaxial to the stress tensor as defined in

Equation 3-37. Generally, the stress exponent n takes values between 3 and 5.

A general relation between creep rate and several material parameters is the
Mukherjee-Bird-Dorn equation (Ref. 6)

DGb σ e n b p – Q ⁄ RT
ε cr ∝ -------------  ------  --- e
·
k B T  G   d

Here, T is the temperature, d is the grain size, b is the Burgers vector, D is the self
diffusion coefficient, G is the shear modulus, and e−Q/RT is an “Arrhenius” type of
temperature dependency.

For high temperatures, Mukherjee-Bird-Dorn equation describes Weertman creep

when setting p = 0. Setting n = 0 and p = 2 describes Nabarro-Herring, and setting

MATERIAL MODELS | 331

 n = 0 and p = 3 describes Coble creep. Harper-Dorn creep is obtained by setting n = 1
and p = 0.

See also the description of the Weertman material model in the Solid
Mechanics interface documentation.

ANAND VISCOPLASTIC MODEL

The Anand viscoplasticity (Ref. 9) is a deviatoric creep model suitable for large,
isotropic, viscoplastic deformations in combination with small elastic deformations.

The viscoplastic strain rate equation reads

·
ε cr = F cr n D

where nD is a deviatoric tensor coaxial to the stress tensor as defined in Equation 3-37,
and the creep rate is calculated from

1-
----
σe m
sinh  ξ -----
– Q ⁄ RT
F cr = Ae
 sa 

Here, A is the creep rate coefficient (SI unit: s−1), Q is the activation energy
(SI unit: J/mol), m is the stress sensitivity, ξ is the multiplier of stress, R is the gas
constant, and T is the absolute temperature (SI unit: K).

The internal variable, sa, is called deformation resistance (SI unit: Pa) and is calculated
from the rate equation

sa a – 1 sa 
·
s a = h 0 1 – ----*-  1 – ----*- F cr
sa  s a

with the initial condition sa(0) = sinit. Here, h0 is the hardening constant
(SI unit: Pa), and a is the hardening sensitivity.

The variable sa* is the saturation value of the deformation resistance sa, which is
calculated from the expression

F cr – Q ⁄ RT n
s a* = s 0  ------- e
A 

332 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

where s0 is the deformation resistance saturation coefficient (SI unit: Pa), and n is
the deformation resistance sensitivity.

CHABOCHE VISCOPLASTIC MODEL

The viscoplastic strain rate tensor is given by

· Fy n
ε vp = A  --------- n D
σ ref

Here, A is the viscoplastic rate coefficient (SI unit: 1/s), n is the stress exponent
(dimensionless), σref a reference stress level (SI unit: Pa), and nD is a deviatoric tensor
coaxial to the stress tensor. The Macaulay brackets are applied on the yield function,
which is defined as done for plasticity

F y = φ ( σ ) – σ ys

The effective stress φ(σ) is either the von Mises, Tresca, or Hill stress; or a user-defined
expression, and σys is the yield stress (which may include a linear or nonlinear Isotropic
Hardening model). The stress tensor used in the effective stress φ(σ) is shifted by what
is usually called the back stress, σback when Kinematic Hardening is included.

The deviatoric tensor nD is computed from the plastic potential Qp

∂Q p
n D = ----------
∂σ

When von Misses effective stress is used, the associated flow rule reads Qp = Fy, and
the deviatoric tensor nD is defined as done for deviatoric creep

3 dev ( σ )
n D = --- ----------------- (3-42)
2 σ mises

Given the property

3
n D :n D = ---
2

the effective viscoplastic strain rate is equivalent to

· 2 Fy n
ε vpe = --- ε· vp :ε· vp = A  --------
-
3 σ ref

MATERIAL MODELS | 333

 PERZYNA VISCOPLASTIC MODEL
Perzyna viscoplastic model is similar to the Chaboche model, with the exception than
the stress exponent is set equal to one. The viscoplastic strain rate tensor is then given
by

· A
ε vp = ---------  F y n D
σ ref

Some authors denote the viscosity as the quotient η = σref/A.

See also the description of Viscoplasticity in the physics interface

documentation.

ENERGY DISSIPATION
Since creep and viscoplasticity are inelastic processes, the dissipated energy density can
be calculated by integrating the creep dissipation rate density (SI unit: W/m3) given
by

· ·
W cdr = σ:ε cr

In case many creep sub-nodes are added, the creep dissipation rate density is calculated
·
from the cumulative creep strain rate tensor ε cr .

The total energy dissipated in a given volume can be calculated by a volume integration
of the dissipated creep energy density Wc (SI unit: J/m3).

When the Calculate dissipated energy check box is selected, the dissipation
rate density due to creep is available under the variable solid.Wcdr and
the dissipation rate density due to viscoplasticity is available under the
variable solid.Wvpdr. The dissipated energy density due to creep is
available under the variable solid.Wc and due to viscoplasticity under the
variable solid.Wvp. Here solid denotes the name of the physics
interface node.

CONVERTING BETWEEN DIFFERENT CREEP DATA REPRESENTATIONS

The equation forms described in for the different creep models above differ from the
forms most commonly found in the literature. The difference lies in the introduction
of normalizing reference values such as the reference stress σref and reference time tref.

334 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

These values are in a sense superfluous, and can in principle be chosen arbitrarily. The
choice of reference values will however affect the numerical values to be entered for
the material data. This system has two advantages

• It is possible to use the built-in unit management system in COMSOL Multiphysics.

• You do not need to do any difficult unit conversions when creep data are available
in units other that the model units. Since many creep models contain the stress or
time raised to a non-integer power, such conversions are error prone.

Norton law
Material data for a Norton law is often available as the parameters AN and n in the
equation

· n
ε cr = A N σ e

The coefficient AN has a physical dimension which depends on the value of n and the
unit have an implicit dependence on the stress and time units. Converting the data to
the form used in COMSOL Multiphysics (Equation 3-38) requires the introduction
of the reference stress σref. It is here convenient to use the implicit stress unit for which
AN is given as reference stress. The creep rate coefficient A will then have the same
numerical value as AN, and you do not need to do any conversions.

The physical dimension of A is however (time)-1, whereas the physical dimension of

AN is (stress)-n(time)-1.

Another popular way of representing creep data is to supply the stress giving a certain
creep rate. As an example, σc7 is the stress at which the creep strain rate is 10-7/h. Data
on this form is also easy to enter: You set the reference stress to the value of σc7, and
enter the creep rate coefficient as 1e-7[1/h].

Example

Assume that a carbon steel has the following two equivalent descriptions of its creep
properties at a certain temperature:

• σc7 = 70 MPa, and stress exponent n = 4.5.

• AN = 4.98·10-16 with respect to units MPa and hours, and stress exponent n = 4.5.

In the first case, enter:

• σref as 70[MPa]

MATERIAL MODELS | 335

 • n as 4.5
• A as 1e-7[1/h].

In the second case, enter:

• σref as 1[MPa]
• n as 4.5
• A as 4.98e-16[1/h]

These two sets of data describe the same material.

Norton-Bailey law
Material data for a Norton-Bailey law usually is usually written in terms of the creep
strain, rather than the creep strain rate, so that the form of the constitutive relation is

n m
ε cr = A NB σ e t

In this case, the coefficient ANB has implicit dimension and units which depend on the
values of n and m, and on the stress and time units.

Converting the data to the form used in COMSOL Multiphysics (Equation 3-39 and
Equation 3-40) requires the introduction of an both an arbitrary reference stress σref
and an arbitrary reference time tref. If you use the implicit units for which ANB is given
as the reference values, then the constant A will have the same numerical value as ANB.

Garofalo Law
Since the stress inside in the Garofalo law appears as an argument to a sinh() function,
it must necessarily be nondimensionalized. Most commonly this is however written as

· n
ε cr ∝ ( sinh ( ασ e ) )

Comparing with the expression in COMSOL Multiphysics,

σe n
ε cr ∝  sinh  --------- 
·
σ ref

it is evident that the reference stress should be chosen as

1
σ ref = ---
α

In this case, there is no arbitrariness in the choice of σref, since α is an actual material
parameter.

336 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

Piezoelectric Material

THE PIEZOELECTRIC EFFECT

The piezoelectric effect manifests itself as a transfer of electric to mechanical energy
and vice versa. It is present in many crystalline materials, while some materials such as
quartz, Rochelle salt, and lead titanate zirconate ceramics display the phenomenon
strongly enough for it to be of practical use.

The direct piezoelectric effect consists of an electric polarization in a fixed direction

when the piezoelectric crystal is deformed. The polarization is proportional to the
deformation and causes an electric potential difference over the crystal.

The inverse piezoelectric effect, on the other hand, constitutes the opposite of the
direct effect. This means that an applied potential difference induces a deformation of
the crystal.

PIEZOELECTRIC CONSTITUTIVE RELATIONS

It is possible to express the relation between the stress, strain, electric field, and electric
displacement field in either a stress-charge form or strain-charge form:

Stress-Charge
T
T = cE S – e E
D = eS + ε S E

Strain-Charge
T
S = sE T + d E
D = dT + ε T E

In the above relations, the naming convention used in piezoelectricity theory is

assumed, so that the structural strain is denoted by S, and the stress is denoted by T.
Thus, the naming convention differs in piezoelectricity theory compared to structural
mechanics theory.

The Piezoelectric Material uses the structural mechanics nomenclature.

The strain is named ε (instead of S) and the stresses are denoted by either
σ or S (instead of T). This makes the names consistent with those used in
the other structural mechanics interfaces.

MATERIAL MODELS | 337

 The constitutive relation using COMSOL Multiphysics symbols for the different
constitutive forms are thus:

STRESS-CHARGE
T
σ = cE ε – e E
(3-43)
D = eε + ε 0 ε rS E

The Stress-Charge form is always used in the variational formulation (weak equation
form) which COMSOL Multiphysics uses for discretization and computation.

STRAIN-CHARGE
T
ε = sEσ + d E
D = dσ + ε 0 ε rT E

Most material data appears in the strain-charge form, and it can be easily transformed
into the stress-charge form. In COMSOL Multiphysics both constitutive forms can be
used; simply select one, and the software makes any necessary transformations. The
following equations transform strain-charge material data to stress-charge data:

–1
cE = sE
–1
e = d sE
–1 T
ε S = ε 0 ε rS = ε 0 ε rT – d s E d

You find all the necessary material data inputs within the Piezoelectric Material feature
under the Solid Mechanics interface, which are added automatically when you add a
predefined Piezoelectric Devices multiphysics interface. Such node can be also added
manually under any Solid Mechanics interface similar to all other material model
features. The piezoelectric material uses the Voigt notation for the anisotropic material
data, as customary in this field. More details about the data ordering can be found in
Orthotropic and Anisotropic Materials section.

GOVER NING EQUATIONS

The equations of Piezoelectricity combine the momentum equation Equation 3-51
with the charge conservation equation of Electrostatics,

∇ ⋅ D = ρV (3-44)

338 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

where the ρ V is the electric charge concentration. The electric field is computed from
the electric potential V as

E = – ∇V

In both Equation 3-44 and Equation 3-51, the constitutive relations Equation 3-43
are used, which makes the resulting system of equations closed. The dependent
variables are the structural displacement vector u and the electric potential V.

WAVE P ROP A GAT IO N

In case of geometric linearity, the governing equations are linear and have the
following time-harmonic wave solutions:

u=u ˆ exp [ ik ( r + ct ) ]
ˆ
V = V exp [ ik ( r + ct ) ]

where k is the wave number vector that determines the direction of the wave
propagation, and c is the phase velocity (or wave speed).

The expressions for the wave speed can be computed analytically for waves of different
types, polarizations and directions of propagations. For example, the pressure wave
propagating in the X axis direction is a particular solution, for which

u = uˆ exp [ ik X ( X + c X t ) ]
v=0
w=0
ˆ
V = V exp [ ik X ( X + c X t ) ]

The corresponding pressure wave speed is given by

2 1⁄2
1 e X1
c pX = --- c E, 11 + -----------------------
ρ ε 0 ε rS, XX

The shear wave propagation in the X axis direction and with XY plane polarization is
a solution such that

MATERIAL MODELS | 339

 u=0
v = vˆ exp [ ik X ( X + c X t ) ]
w=0
ˆ
V = V exp [ ik X ( X + c X t ) ]

and the corresponding wave speed is computed as

2 1⁄2
1 e X6
c sXY = --- c E, 66 + -----------------------
ρ ε 0 ε rS, XX

COMSOL Multiphysics provides predefined variables for the waves speeds for waves
of different types and polarizations propagating in the X, Y and Z directions.

PIEZOELECTRIC DISSIPATION
In order to define dissipation in the piezoelectric material for a time-harmonic analysis,
all material properties in the constitutive relations can be complex-valued matrices
where the imaginary part defines the dissipative function of the material.

Complex-valued data can be defined directly in the fields for the material properties,
or a real-valued material X and a set of loss factors ηX can be defined, which together
form the complex-valued material data

˜
X = X ( 1 ± jη X )

See also the same references for an explanation of the sign convention.

It is also possible to define the electrical conductivity of the piezoelectric material, σ.

Electrical conductivity appears as an additional term in the variational formulation
(weak equation form). The conductivity does not change during transformation
between the formulations.

The energy dissipation modeling is also available in time domain. The options are:
dielectric dispersion for the electrical part, and Rayleigh damping for the mechanical
and coupling parts of the problem. The total dissipated energy can be computed as a
function of time.

Piezoelectric Losses

340 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

INITIAL STRESS, STRAIN, AND ELECTRIC DISPLACEMENT
Using the functionality available under the Piezoelectric Material feature and Solid
Mechanics interface, one can define initial stress (S0), initial strain (ε0), and remanent
electric displacement (Dr) for models. In the constitutive relation for piezoelectric
material these additions appear in the stress-charge formulation:

T
σ = cE ( ε – ε0 ) – e E + σ0
D = e ( ε – ε 0 ) + ε 0, vac ε rS E + D r

When solving the model, the program does not interpret these fields as a constant
initial state, but they operate as additional fields that are continuously evaluated. Thus
use these initial field to add, for example, thermal expansion or pyroelectric effects to
models.

Magnetostrictive Material

MAGNETOSTRICTION
Magnetostriction describes the change in dimensions of a material due to a change in
its magnetization. This phenomenon is a manifestation of magnetoelastic coupling,
which is exhibited by all magnetic materials to some extent. The effects related to
magnetoelastic coupling are described by various names. The Joule effect describes the
change in length due to a change in the magnetization state of the material. This
magnetostrictive effect is used in transducers for applications in sonars, acoustic
devices, active vibration control, position control, and fuel injection systems.

The inverse effect accounts for the change in magnetization due to mechanical stress
in the material. This effect is also known Villari effect. This effect is mostly useful in
sensors.

Magnetostriction has a quantum-mechanical origin. The magneto-mechanical

coupling takes place at the atomic level due to spin-orbit coupling. From a system level,
the material can be assumed to consist of a number of tiny ellipsoidal magnets which
rotate due to the torque produced by the externally applied magnetic field. The

MATERIAL MODELS | 341

 rotation of these elemental magnets produces a dimensional change leading to free
strain in the material.

LINEAR MAGNETOSTRICTION MODEL

The magnetostriction has a nonlinear dependence on the magnetic field and the
mechanical stress in the material. However, the effect can be modeled using linear
coupled constitutive equations if the response of the material consists of small
deviations around an operating point (bias point).

It is possible to express the relation between the stress S, strain ε, magnetic field H,
and magnetic flux density B in either a stress-magnetization form or
strain-magnetization form:

Stress-Magnetization
T
S = c H ε – e HS H
B = e HS ε el + μ 0 μ rS H

Strain-Magnetization
T
ε = s H S + d HT H
B = d HT S + μ 0 μ rT H

where μ 0 is the magnetic permeability of free space, c H and s H are respectively the
stiffness and compliance matrices measured at constant magnetic field, and μ rS and
μ rT are the relative magnetic permeabilities measured at constant strain and constant
stress, respectively. The matrices d HT and e HS are called piezomagnetic coupling
matrices.

342 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

In COMSOL Multiphysics, both constitutive forms can be used; simply select one, and
the software makes all necessary transformations. The following equations transform
strain-magnetization material data to stress-magnetization data:

–1
cH = sH
–1
e HS = d HT s H
1 –1 T
μ rS = μ rT – ------ d HT s E d HT
μ0

You find all the necessary material data inputs within the Magnetostrictive Material
node under the Solid Mechanics interface, which are added automatically when you
add a predefined Magnetostriction multiphysics interface. Such a node can be also
added manually under any Solid Mechanics interface similar to all other material model
features. The Magnetostrictive Material uses Voigt notation for the anisotropic
material data. More details about the data ordering can be found in Orthotropic and
Anisotropic Materials section.

For a crystalline material with tetragonal symmetry, the strain-magnetization form of

the constitutive relations is the following:

1
ε 11 s 11 s 12 s 13 0 0 0 S 11 0 0 – --- d
2 33
ε 22 s 12 s 11 s 13 0 0 0 S 22 1
0 0 – --- d H1
ε 33 s 13 s 13 s 33 0 0 0 S 33 2 33
= + 0 0 d 33 H 2
2ε 23 0 0 0 s 44 0 0 S 23
0 d 15 0 H3
2ε 13 0 0 0 0 s 44 0 S 13
d 15 0 0
2ε 12 0 0 0 0 0 2 ( s 11 – s 12 ) S 12
0 0 0

S 11

B1 0 0 0 0 d 15 0 S 22 μ 11 0 0 H 1
B2 = 0 0 0 d 15 0 0 S 33 + μ 0 μ
0 11 0 H2
1 1 S 23
B3 – --- d – --- d d 33 0 0 0 0 0 μ 33 H 3
2 33 2 33 S 13
S 12

The following material data corresponds to Terfenol-D at 100 kA/m bias and 30 MPa
prestress (Ref. 5):

MATERIAL MODELS | 343

 4.4 – 1.1 – 1.65 0 0 0
– 1.1 4.4 – 1.65 0 0 0
⋅ – 1.65 – 1.65 3.8 0 0 0 [ 1 ⁄ Pa ]
– 11
sH = 10
0 0 0 24 0 0
0 0 0 0 24 0
0 0 0 0 0 11

0 0 0 0 16.5 0 8.1 0 0
–9
d HT = 10 ⋅ 0 0 0 16.5 0 0 [ m ⁄ A ] μ rT = 0 8.1 0
– 4.3 – 4.3 8.6 0 0 0 0 0 3

NONLINEAR MAGNETOSTRICTION
A commonly accepted micromagnetic description of the magnetostriction is as follows
(Ref. 2):

All domains have magnetization of the same magnitude M = M s , but the

magnetization can have different orientations characterized by the corresponding
direction vector m = M ⁄ M s for each domain. The applied magnetic field changes
the domain orientation.

In this section, the term domain refers to a small part of magnetic

material. This is typical for micromagnetics literature, and it should not
be mistaken with the concept of domain as part of the model geometry,
the latter is often used in COMSOL Multiphysics documentation.

For a single crystal with cubic symmetry, the magnetostrictive strain tensor can be
written as the following quadratic form:

ε me = --- λ 100  m ⊗ m – --- I + ( λ 111 – λ 100 )

3 1
2  3   mi mj ( ei ⊗ ej ) (3-45)
i≠j

where ( m ⊗ m ) ij = m i m j . Note that the magnetostrictive strain is represented by a

deviatoric tensor, i.e. tr ( ε me ) = 0 . This is because the deformation is related to the
magnetic domain rotation, and such process should not change the material volume.

The strain in any direction given by the directional cosines β i can be written as

344 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

 Δl
λ = ------ =
l  βi βj εij
i, j

The notation lambda is used for strain in this section, which is typical for
micromagnetics literature. This should not be mistaken with the same
notation used for stretch in other parts of the theory in COMSOL
Multiphysics documentation.

Using Equation 3-45, one gets

λ = --- λ  m 1 β 1 + m 2 β 2 + m 3 β 3 – --- +
3 2 2 2 2 2 2 1
2 100  3 (3-46)
3λ 111 ( m 1 m 2 β 1 β 2 + m 2 m 3 β 2 β 3 + m 1 m 3 β 1 β 3 )

When both magnetization and measurement direction are parallel to the same crystal
direction [100], one has m 1 = β 1 = 1 and all other components are zero, so that

λ = λ || = λ 100

In a similar way for the [111] direction, one has for all components
m i = β i = 1 ⁄ 3 and λ = λ 111 .

If the strain is measured in [100] direction, while all the magnetization vectors are
aligned perpendicular to it, one has only the following two nonzero components:
m 2 = β 1 = 1 and consequently:

1
λ = λ ⊥ = – --- λ
2 100

In many applications, such alignment of the domains is achieved by applying a

compressive prestress. Thus, the maximum usable magnetostriction is achieved via a
90 degree rotation of the domains

3
λ || – λ ⊥ = --- λ
2 100

For an isotropic material, λ 100 = λ 111 = λ s , and Equation 3-45 becomes

MATERIAL MODELS | 345

 ε me = --- λ s  m ⊗ m – --- I
3 1
2  3 

For a polycrystalline material without preferred orientation, the following

approximation can be used (Ref. 1):

2 3
λ s = --- λ + --- λ
5 100 5 111

In COMSOL Multiphysics, this description of the magnetostriction is modeled using

the following equation for the magnetostrictive strain:

3 1
ε me = --- -------2- λ 100 dev ( M ⊗ M ) + ( λ 111 – λ 100 )
2M  Mi Mj ( ei ⊗ ej ) (3-47)
s i≠j

The strain field is deviatoric, and Equation 3-47 exhibits the same properties as
Equation 3-45 at saturation, i.e. when M → M s . Equation 3-46 is replaced by

3 λ 100
- M1  β1 – 1
--- + M 2  β 2 – 1
--- + M 3  β 3 – 1
--- +
2 2 2 2 2 2
λ = --- ----------
2 M2  3  3  3
s
λ 111
- ( M1 M2 β β2 + M2 M3 β β3 + M1 M β1 β3 )
3 ----------
2 1 2 3
Ms

Note that the strain vanishes when M → 0 , which makes the model applicable in
the whole range from full demagnetization to saturation.

For isotropic materials, the magnetostrictive strain is modeled as the following

quadratic isotropic form of the magnetization field (Ref. 3):

3 λs
ε me = --- -------2- dev ( M ⊗ M ) (3-48)
2M
s

The stress in the magnetostrictive material is modeled as

S = c H [ ε – ε me ( M ) ]

For isotropic materials, the stiffness matrix c H can be represented in terms of two
parameters, for example, using the Young’s modulus and Poisson’s ration. Cubic
materials possess only three independent components: c 11 , c 12 and c 44 .

346 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

Using Equation 3-48, one can derived a linear response around a given bias state
characterized by a pre-magnetization vector M 0 . Thus,

M = M 0 + M 1 = [ M 0, 1, M 0, 2, M 0, 3 ] + M 1

where M 1 is a perturbation, and

2M 0, 1 – M 0, 2 – M 0, 3
– M 0, 1 2M 0, 2 – M 0, 3
λ s – M 0, 1 – M 0, 2 2M 0, 3
ε me = -------2- M1
Ms 0 3M 0, 3 3M 0, 2
3M 0, 3 0 3M 0, 1
3M 0, 2 3M 0, 1 0

If one assumes a unidirectional state, for example

M 0 = [ 0, 0, M 0 ]

it will further simplify into

0 0 –1
0 0 –1
λs M0 0 0 2 M
ε me = -------------
2
- 1
Ms 0 3 0
3 0 0
0 0 0

MAGNETIZATION
Nonlinear magnetization in the magnetostrictive material is found from the following
nonlinear implicit relation (Ref. 4):

H eff
M = M s L ( H eff ) --------------
H eff

where L is the Langevin function

3χ m H eff Ms
L = coth  -------------------------- – --------------------------
 Ms  3χ m H eff

with χ m being the magnetic susceptibility in the initial linear region.

MATERIAL MODELS | 347

 Other possible choices of the L function are a hyperbolic tangent

χ m H eff
L = tanh  ----------------------
Ms

and a linear function

χ m H eff
L = ----------------------
Ms

The latter option will make it possible to find an explicit expression for the
magnetization. However, such model does not have a proper saturation behavior, and
thus it should be used only in the operating range far from saturation. Both the
Langevin function and hyperbolic tangent models requires the magnetization vector
components to be treated as extra dependent variables.

For cubic crystals, the effective field in the material is given by

3
μ0 Ms

H eff = H + --------------2- λ 100 S ed + ( λ 111 – λ 100 ) ( S ed ) ij ( e i ⊗ e j ) M (3-49)
i≠j

where H is the applied magnetic field. The second term in Equation 3-49 represents
the mechanical stress contribution to the effective magnetic field, and thus to the
material magnetization, which is called the Villari effect. The deviatoric stress tensor
is related to the strain as

S ed = dev ( c H ε )

For isotropic materials, the effective magnetic field expression simplifies into

3λ s
H eff = H + --------------2- S ed M
μ0 Ms

In addition, the magnetization and magnetic field are related to each other and to the
magnetic flux density (also called the B-field) by

B = μ0 ( H + M )

COMSOL Multiphysics solves for the magnetic vector potential A whose curl yields
the vector B-field. The H-field is then obtained as a function of the B-field and
magnetization.

348 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

Rigid Domain Material Model
A Rigid Domain, or a rigid body, is an idealization of a body in which the deformation
is neglected. In other words, the distance between any two given points of a rigid body
remains constant in time regardless of any external forces acting on it. An object can
be assumed to be perfectly rigid if its flexibility can be neglected in comparison with
other flexibilities in the system, and when there is no need to compute the stress in the
object.

Rigid Domain is available as a material model both in the Solid

Mechanics, Shell, Beam, and Multibody Dynamics interfaces. This theory
section applies to all interfaces. There are some minor differences between
the versions of the Rigid Domain which will be described as they appear.

A Rigid Domain can consist of a selection of several geometrical domains.

These domains will act as a single rigid object, irrespective of whether
they are geometrically connected or not.

Adjacent geometrical objects selected in different Rigid Domains nodes

are independent, and can even penetrate each other.

The rigid domain is a material model, which is mutually exclusive to all other material
models. The only material property needed is the mass density.

RIGID DOMAIN KINEMATICS

When a body is rigid, it is sufficient to describe the motion of at least three
non-collinear particles. It is then possible to reconstruct the motion of all other
particles in the body. Usually a mathematically more convenient, but equivalent,
approach is used. The motion of the whole body is represented by:

• The linear motion of the body. The motion of one of the particles of the body,
chosen as a reference point (often coinciding with the center of mass).
• The angular motion (also known as orientation or attitude) of the body.

The degrees of freedom needed to represent the linear and angular motion are known
as rigid body translation and rigid body rotation degrees of freedom.

In 2D, this is represented by two in-plane translations and the rotation around the
z-axis.

MATERIAL MODELS | 349

 In 3D the situation is more complex. Six degrees of freedom are necessary. They are
usually selected as three translations and three parameters for the rotation. For finite
rotations any choice of three rotation parameters is however singular at some specific
set of angles. For this reason, a four-parameter quaternion representation is used for
the rotations in COMSOL Multiphysics. Thus, each rigid domain in 3D actually has
seven degrees of freedom: three for the translation, and four for the rotation. The
quaternion parameters are called a, b, c, and d, respectively. These four parameters are
not independent, so an extra equation stating that

2 2 2 2
a +b +c +d = 1

is added.

The connection between the quaternion parameters and the rotation matrix R is:

2 2 2 2
a +b –c –d 2bc – 2 ad 2ac + 2bd
R = 2ad + 2bc
2 2
a –b +c –d
2 2
2cd – 2ab
2 2 2 2
2bd – 2 ac 2ab + 2cd a –b –c +d

For the geometrically linear case, the quaternion constraint and the rotation matrix
definition are reduced to:

1 – 2d 2c
R = 2d 1 – 2b
– 2c 2b 1
a=1

In 2D, the rotational degree of freedom is the angle of rotation about the z-axis φ ,
and its relation with the rotation matrix R is:

cos φ – sin φ 0
R = sin φ cos φ 0
0 0 1

For the geometrically linear case, the 2D rotation matrix is reduced to:

1 –φ 0
R = φ 1 0
0 0 1

350 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

Under translation and rotation of a rigid domain, the complete expression for the
displacement of any point on the rigid body is given by:

ud = u + ( R – I ) ⋅ ( X – XM )

where X are the material coordinates of any point in the rigid domain, X M is the center
of mass of the rigid domain, u is the translation vector at the center of mass, and I is
the identity matrix.

The rigid body displacement at the center of mass (u) are degrees of freedom. Thus
the rigid body translational velocity and acceleration can be evaluated by directly taking
the time derivatives of u. In the time domain it can be expressed as:

∂u
u= ∂u
· ·· ·
u =
∂t ∂t

In the frequency domain, they can be expressed in terms of frequency (ω):

· ·· · 2
u = ( iω )u u = ( iω )u = – ω u

The same is true for the rotation in 2D since the rigid body rotation φ is the degree
of freedom. The rigid body angular velocity and acceleration can be evaluated by
directly taking the time derivatives of φ .

In 3D, the situation is different and the total rotation of the rigid domain can be
presented as a function of quaternion:

b
2 acos ( a )
Θ = ----------------------------------- c
2 2 2
b +c +d d

The parameter a can be considered as measuring the rotation, while b, c, and d can be
interpreted as the orientation of the rotation vector. For small rotations, this relation
simplifies to:

b
Θ = 2 c
d

The angular velocity of the rigid domain is computed as:

MATERIAL MODELS | 351

 Q2 a
·
q = b
·
Θ = Q3 Q = 2(q ⊗ q)
c
Q4
d

Here q is the conjugate of q, and the symbol ⊗ denotes quaternion

multiplication.

The angular acceleration of the rigid domain can be evaluated by taking the time
derivative of the angular velocity.

RIGID DOMAIN DYNAMICS

The governing equation for a rigid domain can be written as a balance between the
inertial (internal) forces and applied external forces. A rigid domain has only one
internal force, the inertial force. This means that only the mass density of a domain is
required to define the rigid domain material model.

The inertial forces and inertial moments about the center of mass are:

·· ··
F = mu M = IΘ
·· ··
where u and Θ are the linear and angular accelerations of a rigid domain.

The inertial properties mass (m) and moment of inertia tensor (I) of a rigid domain
are computed as:

m = ρ dV 
ρX dV
X M = -------------------
m


T T
I = ( ( ( X – X M ) ⋅ ( X – X M ) )E 3 – ( X – X M ) ⋅ ( X – X M ) )ρ dV

where E 3 and X M are the identity matrix and the center of mass of a rigid domain,
respectively. The special case for the Shell interface is described in Rigid Domain for
Shells.

In 2D, the expressions for inertial forces, inertial moments, and moment of inertia
reduce to:

·· ··
F = mu M = Izφ

352 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

 Iz =  ( ( X – XM ) ⋅ ( X – XM ) )ρh dA
where the volume integration has been replaced by an area integration multiplied by
the out-of-plane thickness h.

The equations of motion for the rigid domain are:

··
mu +  FI =  Fext
and

 MI =  Mext
T ·· · × ( RIR T Θ
·)+
RIR Θ +Θ

Here, the subscripts ‘I’ and ‘ext’ denotes inertial and external forces respectively, and
R is the current rotation matrix. The inertial forces are contributions from Mass and
Moment of Inertia nodes.

In 2D, the moment equations are simplified to the scalar equation

 MI =  Mext
··
Iz ϕ +

IN IT IA L VA LUE
As a Rigid Domain is a separate material model, it overrides the default Linear Elastic
Material model and its default Initial Values node. The initial values are given in a
separate Initial Values subnode for each Rigid Domain.

In the Multibody Dynamics interface version of the Rigid Domain, it is

also possible to get initial values for all domains from the interface level
Initial Values section. This is the default option. The Initial Values subnode
is only present under Rigid domain if Locally defined has been selected.

If many rigid domains are present in a system and the have the same initial
values, then it is often better to define initial values at the interface level
once and to reference it in all features.

The initial values for the rigid body translation, rigid body rotation, and the first time
derivatives can be prescribed about any point—a center of rotation—in a selected
coordinate system. The center of rotation can be defined using

• The center of mass of the rigid domain

MATERIAL MODELS | 353

 • Global coordinates of the center of rotation
• The centroid of a set of selected entities (boundaries, edges, or points)
∂u
Given the initial values of translation (u), rotation ( φ ), translational velocity ( ) and
∂t
angular velocity ( ω ) about a center of rotation ( X c ), the rigid body displacement and
quaternion degrees of freedom are initialized as:

u0 = u + ur
u r = ( – r + Ω̂ ( Ω̂ ⋅ r ) ) ( 1 – cos φ ) + ( Ω̂ × r ) sin φ
r = XM – Xc

 ∂u  = ( ω × ( r + u ) ) + ∂u
∂t  0 r ∂t

φ
cos ---
a = 2
b φ
0 sin --- Ω̂
2

∂a
∂t 1 
= ---  0 ⊗ a 
2 ω b 0
∂b
∂t 0

In 2D, the expressions for the initial values reduce to:

u0 = u + ur
u r = – r ( 1 – cos φ ) + ( e z × r ) sin φ
r = XM – Xc

 ∂u  = ( e × ( r + u ) ) ∂φ + ∂u
∂t  0 z r ∂t ∂t
φ0 = φ

 ∂φ = ∂φ
∂t 0 ∂t

The variable ur is the translation at the center of mass due to a rotation around the
center of rotation, and is thus zero when the two points coincide. In the case that you
are entering the data using a separate center of rotation, you must pay special attention
to how the initial displacement and velocity are composed if initial rotations and
rotational velocities are present.

354 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

Figure 3-16: Initial displacement of a rigid body

MASS AND MOMENT OF INERTIA

Sometimes a rigid domain needs the added effect of an associated abstract rigid object,
which is physically not modeled and where the inertial properties are known. You can
model this using Mass and Moment of Inertia, where the inertial properties of this
abstract domain (center of mass, mass, and moment of inertia tensor) can be directly
entered.

The formulation for an abstract rigid object is similar to the physical rigid domain with
these exceptions:

• The inertial properties of the abstract rigid object are input by the user instead of
being computed from a physical domain.
• No extra degrees of freedom are created. The inertial forces generated by this
feature will be computed based on the distance from the center of gravity of the rigid
domain to which it belongs, and the values of the degrees of freedom there.

The inertial force contributions are

2
F I = m d ( u + ( R – E 3 )X mc )
2
dt

where Xmc is the vector from the center of mass of the rigid domain (XM) to the center
of mass of this contribution (Xm),

X mc = X m – X M

The inertial moment contributions are

MATERIAL MODELS | 355

 T ·· · × ( RIR T Θ
·)+X ×F
M I = RIR Θ +Θ mc I

In 2D, there is only a scalar moment contribution:

··
M I = I z ϕ + ( X mc × F I ) ⋅ e z

CONSTRAINTS
The constraints for a rigid domain are different in nature than those applied to flexible
domain. In a flexible domain, a constraint can be applied at various entity levels:
domains, boundaries, edges, or points. Since the degrees of freedom of the rigid
domain are global and present only at the center of mass, boundary conditions are used
to constrain these global degrees of freedom, which is why a global selection is needed.

The constraints used for a flexible domain, for example Fixed Constraint,
Prescribed Displacement, Rigid Connector, or Attachment, are not applicable
to a rigid domain.

In a rigid domain the Prescribed Displacement/Rotation or Fixed Constraint

subnode is used instead to constrain its degrees of freedom.

The Prescribed Displacement/Rotation node can be used to:

• Prescribe the displacement components in arbitrary directions at a given point.

• Constrain rotations in arbitrary directions.
• Prescribe a non-zero rotation around an arbitrary axis.

The displacement and rotation can be prescribed in a selected coordinate system about
an arbitrary center of rotation. The center of rotation can be defined using

• The center of mass of the rigid domain.

• Global coordinates of the center of rotation.
• The centroid of a set of selected entities (boundaries, edges, or points).

The displacement at the center of rotation is computed as:

uc = u + ( R – I ) ⋅ ( Xc – XM ) (3-50)

The components of this displacement vector are prescribed individually in the selected
coordinate system. Through Equation 3-50, a constraint on a translation will impose

356 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

a relation between translational and rotational degrees of freedom if the center of
rotation differs from the center of mass.

To prescribe the rotation in 3D, the imaginary part of the quaternion is prescribed as:

φ0
b = Ω̂ sin -----
2

where Ω̂ and φ 0 are the axis of rotation and angle of rotation respectively.

In 2D, the out-of-plane rotation angle is directly constrained to the prescribed value
of the rotation.

LOADS
The loads available for a flexible domain can also be used for a rigid domain. In
addition to these boundary conditions, a rigid domain also has global subnodes for
applying forces and moments. If you use Applied Force, a force and its location can be
prescribed in a selected coordinate system. A force implicitly also contributes to the
moment unless it is applied at the center of mass of a rigid domain. If an Applied
Moment node is used, a moment can be prescribed in a selected coordinate system.

CONNECTING TO OTHER BODIES

When a rigid domain and a flexible domain share a boundary (Shell: edge, Beam:
point), the connection is automatic. All displacements on the flexible domain are
controlled by the degrees of freedom of the rigid domain, so that

u flex = ( R – I ) ⋅ ( X – X M ) + u

MATERIAL MODELS | 357

 where X is a coordinate on the boundary. If rotational degrees of freedom are present,
which is the case in the Shell and Beam interfaces, the rotations are set equal to those
of the rigid domain.

In the Multibody Dynamics interface, a rigid domain can be also be

connected to another rigid or flexible domain using joints. A rigid
component can be directly selected in the joints to establish a connection.
This stands in contrast to a flexible component, which needs an
attachment.

A Rigid Domain which contains a selection of several geometrical

domains will act as a single rigid object.

Adjacent domains selected in different Rigid Domains nodes are

independent objects.

Safety Factor Evaluation

There are many theories available in the literature for predicting material failure, these
can predict, for instance whether a ductile material will yield or not, or if a brittle
material will crack under a given set of loads.

Tsai and Wu (Ref. 13, Ref. 22) proposed a stress-dependent criterion intended at
modeling failure in composites. Under Tsai-Wu criterion, failure occurs when a given
quadratic function of stress is grater than zero. The failure criterion is given by

g ( σ ) = σ: ( Fσ ) + f:σ – 1

here, σ is the stress tensor, F a fourth rank tensor (SI unit: 1/Pa^2) and f is a second
rank tensor (SI unit: 1/Pa). For Tsai-Wu criterion, failure occurs when g(σ) ≥ 0.

Due to the symmetry of these tensors, the fourth rank tensor can be represented by a
symmetric 6-by-6 matrix, and the second rank tensor by a 6-by-1 vector (see Voigt
order in the section Tensor vs. Matrix Formulations).

Certain constraints ensure that the failure surface g(σ) = 0 forms a closed ellipsoid in
the stress space. Also, thermodynamic considerations restrict the value of some
components of the fourth rank tensor to be positive only. These restrictions are
summarized as (no summation of the indices)

2
F ii > 0 and F ii F jj ≥ F ij

358 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

The failure index is computed from the failure criterion as

fi = g ( σ ) + 1

so failure is predicted for a failure index greater than one, fi ≥ 1.

The damage index is given by a boolean expression based on the failure criterion

1 g( σ) ≥ 0
di =
0 otherwise

here di = 1 means damage, and di = 0 represents a healthy material.

The safety factor, also called reserve factor or strength ratio, is computed by scaling the
stress tensor such as the failure criterion is equal to zero

g ( sf σ ) = 0

For a quadratic failure criterion, as the Tsai-Wu criterion, this means solving a
quadratic equation for the safety factor variable sf

2
s f ( σ: ( Fσ ) ) + s f ( f:σ ) – 1 = 0

the safety factor is then obtained from the smallest positive root.

For a isotropic criteria, such as von Mises criterion, g(σ) = σmises/σts−1, and the safety
factor is given by sf = σts/σmises.

The margin of safety (Ref. 23) is then computed from the safety factor

ms = s f – 1

Use the Safety subnode to set up variables which can be used to check the risk of failure
according to various criteria. It can be used in combination with Linear Elastic Material
or Nonlinear Elastic Material Models.

Following Tsai-Wu formalism, different orthotropic criteria can be defined by setting

appropriate values for the coefficients in F and f tensors.

AN IS OT RO PIC T SAI - WU CR ITE RI O N

For this criterion, enter twenty one coefficients to define the 6-by-6 matrix F, and six
coefficients to define the vector f. The failure criterion is evaluated from the expression

MATERIAL MODELS | 359

 T T
σ 11 F 11 F 12 F 13 F 14 F 15 F 16 σ 11 f1 σ 11
σ 22 F 12 F 22 F 23 F 24 F 25 F 26 σ 22 f2 σ 22
σ 33 F 13 F 23 F 33 F 34 F 35 F 36 σ 33 f3 σ 33
g( σ) = + –1
σ 23 F 14 F 24 F 34 F 44 F 45 F 46 σ 23 f4 σ 23
σ 13 F 15 F 25 F 35 F 45 F 55 F 56 σ 13 f5 σ 13
σ 12 F 16 F 26 F 36 F 46 F 56 F 66 σ 12 f6 σ 12

here, σij are the stress tensor components given in the local coordinate system of the
parent node.

OR THOTROPIC TSAI-WU CRITERION

For this criterion, enter nine coefficients corresponding to the tensile strengths σtsi,
compressive strengths σcsi, and shear strengths σssij given in the local coordinate
system of the parent node. Tsai-Wu coefficients are then computed from

1 1 1
F 11 = -------------------- , F 22 = -------------------- , F 33 = -------------------- ,
σ cs1 σ ts1 σ cs2 σ ts2 σ cs3 σ ts3

1 - 1 - 1 -
F 44 = ----------- , F 55 = ----------- , F 66 = ----------- ,
σ ss23
2 σ ss13
2 σ ss12
2

1 1 1
F 12 = – --- F 11 F 22 , F 13 = – --- F 11 F 33 , F 23 = – --- F 22 F 33
2 2 2

1 1 1 1 1 1
f 1 = --------- – ---------- , f 2 = --------- – ---------- , f 3 = --------- – ----------
σ ts1 σ cs1 σ ts2 σ cs2 σ ts3 σ cs3

all the other coefficients in F and f tensors are set to zero.

ORTHOTROPIC TSAI-HILL CRITERION

For this criterion, enter six coefficients corresponding to the tensile strengths σtsi and
shear strengths σssij given in the local coordinate system of the parent node. Tsai-Wu
coefficients are then computed from

1- 1- 1-
F 11 = --------- , F 22 = --------- , F 33 = --------- ,
σ ts1
2 σ ts2
2 σ ts3
2

1 - 1 - 1 -
F 44 = ----------- , F 55 = ----------- , F 66 = ----------- ,
σ ss23
2 σ ss13
2 σ ss12
2

360 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

 1 1
F 12 = – --- ( F 11 + F 22 – F 33 ) , F 13 = – --- ( F 11 + F 33 – F 22 ) ,
2 2
1
F 23 = – --- ( F 22 + F 33 – F 11 )
2

all the other coefficients in F and f tensors are set to zero. See also Hill Orthotropic
Plasticity.

ORTHOTROPIC HOFFMAN CRITERION

For this criterion, enter nine coefficients corresponding to the tensile strengths σtsi,
compressive strengths σcsi, and shear strengths σssij given in the local coordinate
system of the parent node. Tsai-Wu coefficients are then computed from

1 1 1
F 11 = -------------------- , F 22 = -------------------- , F 33 = -------------------- ,
σ cs1 σ ts1 σ cs2 σ ts2 σ cs3 σ ts3

1 - 1 - 1 -
F 44 = ----------- , F 55 = ----------- , F 66 = ----------- ,
σ ss23
2 σ ss13
2 σ ss12
2

1 1
F 12 = – --- ( F 11 + F 22 – F 33 ) , F 13 = – --- ( F 11 + F 33 – F 22 ) ,
2 2
1
F 23 = – --- ( F 22 + F 33 – F 11 )
2

1 1 1 1 1 1
f 1 = --------- – ---------- , f 2 = --------- – ---------- , f 3 = --------- – ----------
σ ts1 σ cs1 σ ts2 σ cs2 σ ts3 σ cs3

all the other coefficients in F and f tensors are set to zero.

ORTHOTROPIC JENKINS CRITERION

Enter nine coefficients corresponding to the tensile strengths σtsi, compressive
strengths σcsi, and shear strengths σssij given in the local coordinate system of the
parent node. The failure criterion is then computed from

σ 1 σ 2 σ 3 σ 23 σ 13 σ 12
g ( σ ) = max  --------, --------, --------, ------------, ------------, ------------ – 1
 σ s1 σ s2 σ s3 σ ss23 σ ss13 σ ss12

here, εsi is either the tensile strength or the compressive strength depending whether
the stress in the i-direction, σi, is positive or negative. The absolute value of the shear
stress σij in the ij-plane is compared to the corresponding shear strength σssij.

MATERIAL MODELS | 361

 OR T HOTROPIC WA DDOUPS CRITER ION
Waddoups criterion is similar to Jenkins criterion, but the failure criterion is given in
terms of strains, not strengths. For this criterion, enter nine coefficients corresponding
to the ultimate tensile strains εtsi, ultimate compressive strains εcsi, and ultimate shear
strains γssij given in the local coordinate system of the parent node. The failure
criterion is then computed from

ε 1 ε 2 ε 3 γ 23 γ 13 γ 12
g ( ε ) = max  -------, -------, -------, -----------, -----------, ----------- – 1
 ε s1 ε s2 ε s3 γ ss23 γ ss13 γ ss12

here, εsi is either the ultimate tensile strain or the ultimate compressive strain
depending whether the strain in the i-direction, εi, is positive or negative. The absolute
value of the shear strain γij in the ij-plane is compared to the corresponding ultimate
shear strain γssij.

MODIFIED TSAI-HILL CRITERION

This criterion is derived from Tsai-Wu theory for two-dimensional plane stress
problems (Ref. 23). It is available in 2D for the Pate interface and the Solid Mechanics
interface in plane stress, and for the Shell interface in 3D. Enter the coefficients
corresponding to the tensile strengths σtsi, compressive strengths σcsi, and shear
strengths σssij given in the local coordinate system of the parent node. The failure
criterion is then computed from the in plane stresses

g ( σ ) = F 11 σ 11
2 + F σ 2 + 2F σ σ + F σ 2 – 1
22 22 12 11 22 66 12

Tsai-Wu coefficients are then computed from

1- 1-
F 11 = --------- for σ 11 > 0 , or F 11 = --------- for σ 11 < 0
σ ts1
2 σ cs1
2

1- 1-
F 22 = --------- for σ 22 > 0 , or F 22 = --------- for σ 22 < 0
σ ts2
2 σ cs2
2

1 1- 1 1-
F 12 = – --- --------- for σ 11 σ 22 > 0 or F 12 = – --- --------- for σ 11 σ 22 < 0
2 σ ts1 2 2 σ cs1 2

1 -
F 66 = -----------
σ ss12
2

all the other coefficients in F and f tensors are set to zero.

362 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

AZZI-TSAI-HILL CRITERION
This criterion is derived from Tsai-Wu theory for two-dimensional plane stress
problems. It is available in 2D for the Pate interface and the Solid Mechanics interface
in plane stress, and for the Shell interface in 3D. Enter the coefficients corresponding
to the tensile strengths σtsi, compressive strengths σcsi, and shear strengths σssij given
in the local coordinate system of the parent node. The failure criterion is then
computed from the in plane stresses

g ( σ ) = F 11 σ 11
2 + F σ 2 + 2F σ σ + F σ 2 – 1
22 22 12 11 22 66 12

Tsai-Wu coefficients are then computed from

1- 1-
F 11 = --------- for σ 11 > 0 , or F 11 = --------- for σ 11 < 0
σ ts1
2 σ cs1
2

1- 1-
F 22 = --------- for σ 22 > 0 , or F 22 = --------- for σ 22 < 0
σ ts2
2 σ cs2
2

1 1- 1 1-
F 12 = – --- --------- for σ 11 > 0 or F 12 = – --- --------- for σ 11 < 0
2 σ ts1 2 2 σ cs1 2

1 -
F 66 = -----------
σ ss12
2

all the other coefficients in F and f tensors are set to zero.

NORRIS CRITERION
This criterion is derived from Tsai-Wu theory for two-dimensional plane stress
problems. It is available in 2D for the Pate interface and the Solid Mechanics interface
in plane stress, and for the Shell interface in 3D. Enter the coefficients corresponding
to the tensile strengths σtsi, compressive strengths σcsi, and shear strengths σssij given
in the local coordinate system of the parent node. The failure criterion is then
computed from the in plane stresses

g ( σ ) = F 11 σ 11
2 + F σ 2 + 2F σ σ + F σ 2 – 1
22 22 12 11 22 66 12

Tsai-Wu coefficients are then computed from

1- 1-
F 11 = --------- for σ 11 > 0 , or F 11 = --------- for σ 11 < 0
σ ts1
2 σ cs1
2

MATERIAL MODELS | 363

 1- 1-
F 22 = --------- for σ 22 > 0 , or F 22 = --------- for σ 22 < 0
σ ts2
2 σ cs2
2

1 -
F 66 = -----------
σ ss12
2

1
F 12 = – --- F 11 F 22
2

all the other coefficients in F and f tensors are set to zero.

ISOTRO PIC VO N MISES CRIT ERION

Von Mises criterion is one of the simplest isotropic criteria to predict yielding in metals
and other ductile materials. The failure criterion is computed from the isotropic tensile
strength σts

σ mises
g ( σ ) = -------------- – 1
σ ts

The effective von Mises stress σmises is defined from the deviatoric stress tensor, see
the section about plasticity and von Mises Criterion. For ductile materials the tensile
strength corresponds to the yield stress, while for brittle materials it corresponds to the
failure strength.

I S O T RO P I C TR E S C A C R I T E R I O N
Tresca criterion is similar to von Mises criterion. The failure criterion is computed from
the isotropic tensile strength σts

σ tresca
g ( σ ) = --------------- – 1
σ ts

here, Tresca effective stress is defined in terms of principal stresses, σtresca = σ1 − σ3,
see Tresca Criterion. For ductile materials the tensile strength corresponds to the yield
stress, while for brittle materials it corresponds to the failure strength.

ISOTROPIC RANKINE CRITERION

St. Venant criterion is similar to Tresca criterion, as the failure criterion is given in terms
of principal stresses. For this criterion, enter the tensile strength σts, and the
compressive strength σcs. The failure criterion is then computed from

364 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

 σ p1 σ p2 σ p3
g ( σ ) = max  ---------, ---------, --------- – 1
 σs σs σs 

here, σs is either the tensile strength or the compressive strength depending whether
the principal stress, σpi, is positive or negative. For ductile materials the tensile strength
corresponds to the yield stress, while for brittle materials it corresponds to the failure
strength.

I S O T R O P I C S T . VE N A N T C R I T E R I O N
St. Venant criterion is similar to Waddoups criterion, as the failure criterion is given in
terms of strains, not strengths. For this criterion, enter the ultimate tensile strains εts,
and the ultimate compressive strains εcs. The failure criterion is then computed from

ε p1 ε p2 ε p3
g ( ε ) = max  --------, --------, -------- – 1
εs εs εs

here, εs is either the ultimate tensile strain or the ultimate compressive strain
depending whether the principal strain, εpi, is positive or negative. For ductile materials
the ultimate tensile strain corresponds to the strain at yielding, while for brittle
materials it corresponds to the strain at failure.

ISOTROPIC MOHR-COULOMB CRITERION

Mohr-Coulomb criterion is similar to Tresca criterion, as the failure criterion is given
in terms of principal stresses, see Mohr-Coulomb Criterion for soil plasticity. For this
criterion, enter the cohesion c, and the angle of internal friction φ. The failure criterion
is then computed from

1 σ p1 – σ p3 1 σ p1 + σ p3
g ( σ ) = ---  ------------------------ + ---  ------------------------- sin ( φ ) – 1

2 c cos ( φ )  2  c cos ( φ ) 

here, the principal stresses are sorted as σp1≥σp2≥σp3.

The cohesion and the angle of internal friction are related to the tensile and
compressive strengths by the expressions

σ cs1 – σ ts1 σ cs1 σ ts1

sin ( φ ) = -------------------------- and c cos ( φ ) = --------------------------
σ cs1 + σ ts1 σ cs1 + σ ts1

ISOTROPIC DRUCKER-PRAGER CRITERION

Drucker-Prager criterion approximates the Mohr-Coulomb criterion by a smooth
function (a cone in the stress space). The failure criterion is computed from the stress

MATERIAL MODELS | 365

 invariants I1 and J2, and two material parameters, α and k, see Drucker-Prager
Criterion for soil plasticity

J2
g ( σ ) = ------------------ – 1
k – αI 1

The material parameters α and k are related to the cohesion c and angle of internal
friction φ in the Mohr-Coulomb criterion by the expressions

2 sin ( φ ) 2 3c cos ( φ )
α = ------- ⋅ -------------------------- and k = -------------------------------
3 3 ± sin ( φ ) 3 ± sin ( φ )

The symbol ± is related to either matching the tensile meridian (positive sign) or
matching the compressive meridian (negative sign) of Mohr-Coulomb’s pyramid.
Also, the cohesion and the angle of internal friction are related to the tensile and
compressive strengths, see Isotropic Mohr-Coulomb criterion.

USER DEFINED
The user defined option allows you to write explicitly how the failure criterion and the
safety factor depend on stress and/or strain. These could be analytic functions of stress
or strain tensor components, principal stresses, principal strains, stress or strain
invariants, or data interpolated from tables.

You can add any number of Safety nodes to a single material model. The contents of
this feature will not affect the analysis results as such, as this feature does not account
for post-failure analysis. You can add Safety nodes after having performed an analysis
and just do an Update Solution in order to access to the new variables for result
evaluation.

366 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

Formulation of the Equilibrium
Equations
In this section:

• Equation Implementation

Equation Implementation
The equilibrium equations for solid mechanics are given by Newton’s second law. It is
usually written using a spatial formulation in terms of the Cauchy stress tensor σ:

2
∂ u
ρ 2
= ∇ xσ + f V
∂t

Here fV is a body force per unit deformed volume, and ρ is the current mass density.
For the material frame formulation used in COMSOL Multiphysics, it is more
appropriate to use a Lagrangian version if the equation:

2
∂ u
ρ0 2
= ∇X P + F V
∂t

Now that the first Piola-Kirchhoff stress tensor, P, is used. FV is a body force with
components in the current configuration but given with respect to the undeformed
volume, and ρ0 is the initial mass density. Note the gradient operators are not the same:
in the first case the gradient is taken with respect to the spatial coordinates, and in the
second case with respect to the material coordinates. Using the more common second
Piola-Kirchhoff stress tensor, S, the same equation reads

2
∂ u
ρ0 2
= ∇X( FS ) + F V (3-51)
∂t

where F is the deformation gradient. The COMSOL Multiphysics implementation of

the equations in the Solid Mechanics interface is however not based on the equation
of motion directly, but rather on the principle of virtual work.

The principle of virtual work states that the sum of the internal virtual work and the
external virtual work are equal. The internal virtual work is the work done by the

FORMULATION OF THE EQUILIBRIUM EQUATIONS | 367

 current stress state on a kinematically admissible variation in strains. The external
virtual work is the work done by all forces (acting on domains, boundaries, edges, or
points) when multiplied with the variation in displacements corresponding to the
variation in strains. The virtual displacements δu are in the finite element formulation
represented by the test() operator in COMSOL Multiphysics. For a stationary case,
the virtual work δW is written as

δW = V ( –δε : σ + δu ⋅ FV )dv +
 ( δu ⋅ FS ) ds +  ( δu ⋅ FL ) dl +  ( δu ⋅ Fp ) ]
S L p

The strains are computed from the gradients of the displacements, and the stresses are
given by the constitutive relation.

In a dynamic analysis, the inertial forces are included in the volume forces, according
to d’Alembert’s principle.

δW =  ( –δε : σ + δu ⋅ FV – ρδu ⋅ utt )dv +

V
(3-52)
 ( δu ⋅ FS ) ds +  ( δu ⋅ FL ) dl +  ( δu ⋅ Fp )
S L p

Since the equations are formulated on the material frame, all integrals are taken over
the undeformed geometry. The stress and strain contributions must be interpreted
differently depending on whether the formulation is geometrically nonlinear or not.

• If the study step is geometrically linear, the strain ε is the engineering strain. The
stress σ could in principle be any of the stress measures, as they all converge to the
same engineering stress in this case.
• If the study step is geometrically nonlinear, the strain ε is the Green-Lagrange strain
and the stress σ is the second Piola-Kirchhoff stress.
The Solid Mechanics interface supports Stationary (static), Eigenfrequency, Time
Dependent (transient), Frequency Domain, and Modal solver study types as well as
linear buckling.

368 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

FREQUENCY-DOMAIN STUDIES
In the frequency domain, the frequency response is studied when applying harmonic
loads. Harmonic loads are specified using two components:

• The amplitude value in direction m, Fm

• The corresponding phase, φ m

To derive the equations for the linear response from harmonic excitation loads

F m, freq = F m ( f ) cos ( ωt + φ m )

F x, freq
F freq = F y, freq
F z, freq

Assume a harmonic response with the same angular frequency as the excitation load

u = u amp cos ( ωt + φ u )

u
u= v
w

The relationship can also be described using complex notation with

jφ u jωt jωt jφ
u = Re ( u amp e e ) = Re ( u˜ e ) where u˜ = u amp e u

˜ e jωt )
u = Re ( u

and

jφ m jωt jωt
F m, freq = Re ( F m ( ω )e e ) = Re ( F˜m e )

F˜x
˜
F = F˜
y

F˜z

The primary results, such as displacements, velocities, accelerations, and stress and
strain components are all complex valued.

FORMULATION OF THE EQUILIBRIUM EQUATIONS | 369

 EIGENFREQUENCY STUDIES
The eigenfrequency equations are derived by assuming a harmonic displacement field,
similar as for the frequency response formulation. The difference is that this study type
uses a new variable jω explicitly expressed in the eigenvalue jω = −λ. The
eigenfrequency f is then derived from jω as

λ
f = – ---------
2πj

Damped eigenfrequencies can also be studied, so λ is not necessarily a purely imaginary

number. Any damping included in the problem will automatically cause the
eigenfrequencies to become complex valued.

In addition to the eigenfrequency, the quality factor, Q, and decay factor, δ, for the
model can be examined:

Im ( λ )
Q = -------------------
2Re ( λ )

δ = Re ( λ )

LINEAR BUCKLING
The linear buckling analysis consists of two steps. First a stationary problem is solved
using a unit load of arbitrary size. The critical load is then obtained by solving an
eigenvalue problem, where the eigenvalue λ is the multiplier to the original load that
would cause buckling.

For a general introduction, see Linearized Buckling Analysis

The formulation in terms of virtual work is

δW = V ( –δε : σ –λδ ( εGL – ε ) : σ1 )dv = 0 (3-53)

Here ε us the engineering strain, εGL is the Green-Lagrange strain and σ1 is the stress
caused by the unit load. In terms of stiffness matrices, this corresponds to

( K L + λK NL ( u 0 ) )u = 0

370 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

where KL is the linear stiffness matrix, and KNL is the nonlinear contribution to the
full stiffness matrix. The symbolic linearization point u0 is the displacement vector
caused by the unit load.

Strictly speaking, this formulation assumes that geometric nonlinearity is not used in
the eigenvalue step. The Green-Lagrange tensor is inserted explicitly in the second
term of Equation 3-53, while the first term uses the linear (engineering) strain tensor.

If, however, geometric nonlinearity is selected, then Equation 3-53 is replaced by

δW = V ( –δεGL : σ –( λ – 1 )δ ( εGL – ε ) : σ1 )dv = 0

By using the term (λ-1), the effect of using the Green-Lagrange strain tensor in the
first term is to a large extent removed. Unless the unit load is significantly larger than
the buckling load, the result will be the same as the intended, even if geometric
nonlinearity was inadvertently selected in the eigenvalue study step.

FORMULATION OF THE EQUILIBRIUM EQUATIONS | 371

 Damping
For dynamic problems, the damping of the structure is usually an important property.

In this section:

• Rayleigh Damping
• Loss Factor Damping
• Viscous Damping

Rayleigh Damping
Rayleigh damping is described by two coefficients: the mass damping coefficient αdM
and the stiffness damping coefficient βdK. Rayleigh damping will give the following
contribution to the virtual work

∂P ∂u
δW =   –δ ( ∇u ) : βdK ------
∂t dM
– α ρδu ⋅ -------  dv
∂t 
V

Here P is the first Piola-Kirchhoff stress tensor.

Since Rayleigh damping is added directly to the virtual work equation, it does not
affect the constitutive relation. As a consequence, the stresses and strains will for a
linear elastic material still be in phase. This stands in contrast to the other damping
models.

372 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

Rayleigh damping can be used both in the time domain and in the frequency domain.

Rayleigh damping is not directly related to any physical property.

Historically, it was introduced since it was numerically attractive to have
a damping matrix which was a linear combination of the mass and stiffness
matrices

C = αM + βK

This operation is usually implied to be done at the global assembled

matrix level. Such an interpretation is however only meaningful for pure
structural mechanics problems, but not in a general multiphysics context.
For this reason, Rayleigh damping in COMSOL Multiphysics is a material
property, rather than a global property of the system of equations. If you
enter the same Rayleigh damping parameters for all materials, and solve a
pure structural mechanics problem, then the classical definition will be
retrieved.

Loss Factor Damping

Loss factor damping is only applicable in frequency domain. When using loss factor
damping, a complex constitutive matrix is used. With an isotropic loss factor η s , this
means that

c
D = ( 1 + jη s )D

where D is the constitutive matrix computed from the material data, and Dc is the
complex constitutive matrix used when computing the stresses. For a linear elastic
material, this would be equivalent to multiplying Young’s modulus by the factor
( 1 + jη s ) . For a nonlinear elastic material, this applies to the tangential stiffness.

It is also possible to give individual loss factors for each entry in the constitutive matrix,
so that

c
D mn = ( 1 + jη s, mn )D mn

In the case of an orthotropic material, yet another option is available, where each
individual component of Young’s modulus and shear modulus can be given an
individual loss coefficient:

DAMPING | 373
 c
E m = ( 1 + jη E, m )E m
c
G m = ( 1 + jη G, m )G m
m = 1, 2, 3

The complex moduli are then used to form the constitutive matrix.

For hyperelastic materials, the loss information appears as a multiplier in the strain
energy density, and thus in the second Piola-Kirchhoff stress:

∂W s
S = ( 1 + jη s ) ----------
∂E

For loss factor damping, the following definition is used for the elastic part of the
entropy:

S elast = α : ( s – jη s ( C : ε ) )

note that S elast here denotes the entropy contribution and not any stress.

This is because the entropy is a function of state and thus independent of the strain
rate, while the damping represents the rate-dependent effects in the material (for
example, viscous or viscoelastic effects). The internal work of such inelastic forces
averaged over the time period 2π/ω can be computed as:

1
Q h = --- ωη s Real ( ε : Conj ( C : ε ) )
2

Qh can be used as a heat source for modeling of the heat generation in vibrating
structures, when coupled with the frequency-domain analysis for the stresses and
strains.

Viscous Damping
Viscous damping can be added to the material models. It will cause an extra stress S q
proportional to the rate of elastic strain in the material,

· ·
S q = η b ε el, vol + η v ε el, dev

where η b and η v are the bulk and shear viscosity coefficients, respectively.

Viscous damping can be used in both frequency and time domain analyses.

374 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

In case of geometric nonlinearity, the viscous stress is treated as being a Cauchy stress
acting in the actual configuration (spatial frame). The resulting contribution to the
second Piola-Kirchhoff stress is calculated as

· –1
S q =  η b – --- η v J el C el + η v C el ε el C el
2 –T · –1
 3 
–1
where C el is the inverse of the elastic Cauchy-Green tensor, and J el is the elastic
volume ratio.

DAMPING | 375
 Loads and Boundary Conditions
In this section:

• Distributed Loads
• Rotating Frame and Gravity
• Spring Foundation and Thin Elastic Layer
• Added Mass
• Rigid Motion Suppression Boundary Condition
• Low-Reflecting Boundary Condition
• Cyclic Symmetry and Floquet Periodic Conditions
• Thermal Expansion of Constraints
• Rigid Connector Theory
• Contact Analysis Theory
• Energy Quantities

Distributed Loads
The direction of an explicitly applied distributed load must be given with reference to
a local or global coordinate system in the spatial frame, but its magnitude must be with
reference to the undeformed reference (or material) area. That is, the relation between
the true force f acting on the current area da and the specified distributed load F
acting on the material area dA is f da= FdA. When the solid is subjected to an external
pressure, p, the true force on a surface element acts with magnitude p in the current
area da in the normal direction n:

f = pnda

Therefore, the pressure load type specifies the distributed load as

da
F = pn --------
dA

where both the normal n and area element da are functions of the current
displacement field.

Another view of how to interpret the load, is to express it in the first Piola-Kirchhoff
stress tensor P via the following formula:

376 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

 F = P ⋅ n0

where the normal n0 corresponds to the undeformed surface element. Such a force
vector is often referred to as the nominal traction.

Plane Stress
In a plane stress condition the out-of-plane deformation causes the thickness to
change, and this area effect is included explicitly. The equation transforms to

∂w
F = pn --------  1 + 
da
dA  ∂z 

Axial Symmetry
To account for the radial deformation changing the circumference and therefore the
area element, the distributed load is applied as

da ( R + u )
F = pn -------- -------------------
dA R

Rotating Frame and Gravity

You can add Rotating Frame and Gravity nodes to create the loads caused by gravity or
accelerated frames. This gives load contributions from all nodes in the physics interface
which have a density or mass, such as Linear Elastic Material, Rigid Domain, Added
Mass, or Point Mass.

Only features which have a geometrical selection contribute to the mass

forces. The Mass and Moment of Inertia nodes are global features and will
not get any contribution from Rotating Frame and Gravity nodes.

In the following, the mass density ρ should be considered as generalized. It can

represent mass per unit volume, mass per unit area, mass per unit length, or even mass,
depending on the dimensionality of the object giving the contribution.

ROTATING FRAME
Centrifugal, Coriolis and Euler forces are fictitious forces that need to be introduced
in a rotating frame of reference, since it is not an inertial system. They can be added as
loads.

LOADS AND BOUNDARY CONDITIONS | 377

 Alternatively, the total acceleration in the rotating frame can be augmented to include
the frame acceleration effects:

··
a tot = u + a f

a f = a cen + a cor + a eul

GRAVITY
The gravity acts in a fixed spatial direction eg. The intensity is

g = ρge g

where g is the acceleration of gravity. The action of gravity can also be presented as a
linearly accelerated frame of reference. Thus, it can be accounted for as a contribution
into the total acceleration via the frame acceleration term given by:

af = –g

CENTRIFUGAL FORCE
A centrifugal force acts radially outwards from the axis of rotation defined by the axial
direction vector eax. The rotation is represented by the angular velocity vector:

Ω = Ωeax

where Ω is the angular velocity. In vector form, the acceleration contribution and the
loads are:

a cen = Ω × ( Ω × r p )

F cen = – ρa cen

where rp is the rotation position vector that contains the coordinates with respect to
any point on the axis of rotation. The point is given by its radius vector in the global
coordinate system rbp.

SPIN-SOFTENING EFFECT
The structural displacement can be accounted for when computing the rotation
position, so that

r p = X + u – r bp

378 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

This results in a contribution from the extra acceleration terms caused by the
deformation into the system’s stiffness matrix. The effect is often called spin-softening.

CORIOLIS FORCE
For a Coriolis force to appear, the object studied must have a velocity relative to the
rotating frame. The acceleration contribution and the load are:

∂u
a cor = 2Ω ×
∂t

F cor = – ρa cor

This gives a damping contribution since it is proportional to the velocity.

EULER FORCE
The Euler force occurs when the rate of rotation is not constant in time. The force acts
in the plane of rotation perpendicular to the centrifugal force. The acceleration
contribution and the load are:

∂Ω
a eul = × rp
∂t

F eul = – ρa eul

Spring Foundation and Thin Elastic Layer

In this section, the equations for the spring type physics nodes are developed using
boundaries, but the generalizations to geometrical objects of other dimensions are
obvious. Also, for cases where rotational springs are present, the relations between
moments and rotations are analogous to the relations between forces and
displacements described below.

SPRING FOUNDATION
A spring gives a force that depends on the displacement and acts in the opposite
direction. In the case of a force that is proportional to the displacement, this is called
Hooke’s law. In a suitable coordinate system, a spring condition can be represented as

fs = –K ⋅ ( u – u0 )

LOADS AND BOUNDARY CONDITIONS | 379

 where fs is a force/unit area, u is the displacement deforming the spring, and K is a
stiffness matrix. u0 is an optional deformation offset, which describes the stress-free
state of the spring.

If the spring stiffness is not constant, then it is in general easier to directly describe the
force as a function of the displacement, so that

fs = f ( u – u0 )

In the same way, a viscous damping can be described as a force proportional to the
velocity

· ·
fv = –D ( u – u0 )

where D is a matrix representing the viscosity.

Structural (“loss factor”) damping is only relevant for frequency domain analysis and
is defined as

f l = – iηK ⋅ ( u – u 0 )

where η is the loss factor and i is the imaginary unit. It is also possible to give individual
loss factors for each component in the stiffness matrix K.

If the elastic part of the spring definition is given as a force versus displacement
relation, the stiffness K is taken as the stiffness at the linearization point at which the
frequency response analysis is performed. Since the loss factor force is proportional to
the elastic force, the equation can be written as

f l = iηf s

The contribution to the virtual work is

δW =  ( fs + fl + fv ) ⋅ δu dA
A

T H I N E L A S T I C L A Y E R B E T W E E N TW O P A R T S
A spring or damper can also act between two boundaries of an identity pair. The spring
force then depends on the difference in displacement between the two boundaries.

f sD = – f sS = – K ( u D – u S – u 0 )

380 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

The uppercase indices refer to “source” and “destination”. When a force versus
displacement description is used,

f sD = – f sS = f ( u – u 0 )
u = uD – uS

The viscous and structural damping forces have analogous properties,

· · ·
f vD = – f vS = – D ( u D – u S – u 0 )

f lD = – f lS = – iηK ( u D – u S – u 0 )

or

f lD = iηf sD

The virtual work expression is formulated on the destination side of the pair as

δW =  ( fsD + flD + fvD ) ⋅ δ ( uD – uS ) dAD

AD

Here the displacements from the source side are obtained using the src2dst operator
of the identity pair. If there is a difference in mesh density on the two sides of the pair,
you should select the side with the finer mesh as destination.

THIN ELASTIC LAYER ON INTERIOR BOUNDARIES

On an interior boundary, the Thin Elastic Layer decouples the displacements between
two sides of the boundary. The two boundaries are then connected by elastic and
viscous forces with equal size but opposite directions, proportional to the relative
displacements and velocities.

If an interface which is active on boundaries (Shell or Membrane in 3D

for example) is added on the same interior boundary as a Thin Elastic
Layer, then the virtual slit between the two sides of the boundary may be
closed again. This happens if the domain interface and the boundary
interface share the same displacement degrees of freedom.

The spring force can be written as

f su = – f sd = – K ( u u – u d – u 0 )

LOADS AND BOUNDARY CONDITIONS | 381

 or

f su = – f sd = f ( u – u 0 )
u = uu – ud

The viscous force is

· · ·
f vu = – f vd = – D ( u u – u d – u 0 )

and the structural damping force is

f lu = – f ld = – iηK ( u u – u d – u 0 )

or

f lu = iηf su

The subscripts ‘u’ and ‘d’ denote the “up” and “down” sides of the interior boundary,
respectively.

The virtual work expression is formulated as

δW =  ( fsd + fld + fvd ) ⋅ δ ( uu – ud ) dAu

Au

STIFFNESS FROM MATERIAL DATA

When the stiffness is given in terms of actual material data and layer thickness ds, the
stiffness in the normal direction is computed based on a state of plane strain, so that

E(1 – ν)
k n = -----------------------------------------
ds ( 1 + ν ) ( 1 – ν )

The assumption of plane strain conditions is relevant when the material in the elastic
layer is softer than its surroundings, and this is normally the case.

The shear stiffness is isotropic in the tangential plane, having the value

G E
k t = ------ = ---------------------------
ds 2d s ( 1 + ν )

Since the layer thickness is known in this case, it is also possible to compute a strain in
the elastic layer. The strain tensor has the is stored in a variable with a name like
<interface>.<feature>.etelij, for example solid.tel1.etelxx for the normal

382 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

strain. The two shear strains are stored in the xy and xz components of the tensor. In
3D, the orientation of the two local directions y and z used for the two shear strain
directions is obtained using the following scheme:

1 Choose an auxiliary direction. Unless the normal to the layer is very close to the
global X-direction, use e aux = e X . If the X-direction cannot be used, the
Y-direction is instead used as the auxiliary direction, e aux = e Y .
2 The local y-direction is obtained from the part of the auxiliary direction which is
orthogonal to the normal direction n:
e aux – ( e aux ⋅ n )n
e y = ---------------------------------------------------
e aux – ( e aux ⋅ n )n

3 The local z-direction is orthogonal to the normal and the local y-direction:

ez = n × ey

Springs and Dampers in the Structural Mechanics Modeling chapter.

Added Mass
The Added Mass node can be used for supplying inertia that is not part of the material
itself. Such inertia does not need to be isotropic, in the sense that the inertial effects
are not the same in all directions. This is, for example, the case when a structure
immersed in a fluid vibrates. The fluid is added to the inertia for acceleration in the
direction normal to the boundary, but not tangential to it.

Other uses for added mass are when sheets or strips of a material that is heavy, but
having a comparatively low stiffness, are added to a structure. The data for the base
material can then be kept unaltered, while the added material is represented purely as
added mass.

The value of an added mass can also be negative. You can use such a negative value for
adjusting the mass when a part imported from a CAD system does not get exactly the
correct total mass due to simplifications of the geometry.

Added mass can exist on domains, boundaries, and edges. The inertial forces from
added mass can be written as

LOADS AND BOUNDARY CONDITIONS | 383

 2
∂ u
fm = –M 2
∂t

where M is a diagonal mass distribution matrix. For added mass on a boundary (and
for objects of other dimensions), the contribution to the virtual work is:

δW =  fm ⋅ δu dA
A

Rigid Motion Suppression Boundary Condition

The Rigid Motion Suppression boundary condition is a convenient way to
automatically create a set of constraints which are sufficient to inhibit any rigid body
modes. The constraints are selected so that no reaction forces are introduced as long
as the external loads are in equilibrium.

3 D W I T H TR A N S L A T I O N A L D E G R E E S O F F RE E D O M
In this case, which includes the Solid Mechanics, Multibody Dynamics, Membrane,
and Truss interfaces, six degrees of freedom must be constrained. As it is not possible
to directly constrain rotations, this must be done by a proper selection of locations and
orientations for the constraints.

The following scheme is used:

1 Select three points p1, p2, and p3 that are not located on a common straight line.
2 Compute the unit vectors from p1 to p2 and p3:

X2 – X1
e 12 = -----------------------
X2 – X1

X3 – X1
e 13 = -----------------------
X3 – X1

3 Compute the normal to the plane spanned by these two vectors.

e 12 × e 13
e n1 = --------------------------
e 12 × e 13

4 This normal is perpendicular to the line between p1 and p2. Compute a second
perpendicular direction, orthogonal to en1.

384 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

 e n2 = e 12 × e n1

5 Fix the first point is all three directions, so that there are no rigid body translations.

u p1 = 0

6 Constrain the second point p2 in two directions, so that all possible rotations except
around the line e12 are suppressed:

u p2 ⋅ e n1 = 0
u p2 ⋅ e n2 = 0

7 Constrain the third point p3 so that the remaining rotation is suppressed:

u p3 ⋅ e n1 = 0

The body is now constrained against rigid body rotations, while still free to stretch in
any direction.

P3

en1 e13
P2

P1 e12

en2

Figure 3-17: Selection of constraint orientations: P1 is fixed, P2 is constrained so that it

can only move in the e12 direction, and P3 is constrained in the en1 direction.

LOADS AND BOUNDARY CONDITIONS | 385

 3D WITH ROTATIONAL DEGREES OF FREEDOM
In the Shell interface, the rigid body suppression is implemented using the translational
degrees of freedom in three points, using the same algorithm as above.

In the Beam interface, a single point is constrained in all six degrees of freedom.

2 D W I T H TR A N S L A T I O N A L D E G R E E S O F F RE E D O M
In this case, which includes the Solid Mechanics, Multibody Dynamics, and Truss
interfaces, three degrees of freedom must be constrained. As it is not possible to
directly constrain the rotations, this must be done by a proper selection of locations
and orientations for the constraints.

Two points, p1 and p2 are selected. The first point p1 is constrained in both directions
in order to suppress translational motion. The second point is constrained in the
direction orthogonal to the line joining the two points, so that rotation around the
out-of-plane direction is suppressed:

u p1 = 0

X2 – X1
e 12 = -----------------------
X2 – X1

u p2 ⋅ ( e 12 × e Z ) = 0

2D WITH ROTATIONAL DEGREES OF FREEDOM

In the Plate interface, you can choose to use either three or six degrees of freedom.
When six degrees of freedom are used, the same approach as for 3D solids and shells
is used. In the case of three degrees of freedom, three points which are not located on
a straight line are constrained against out-of-plane translation.

In the Beam interface, a single point is constrained in all three degrees of freedom.

2D AXIALLY SYMMETRIC
For both the Solid Mechanics and Membrane interfaces, a single point is constrained
in the axial (Z) direction.

Low-Reflecting Boundary Condition

The low-reflecting boundary condition is mainly intended for letting waves pass out
from the model domain without reflection in time-dependent analyses. It is also

386 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

available in the frequency domain, but then adding a perfectly matched layer (PML) is
usually a better option.

As a default, the low-reflecting boundary condition takes the material data from the
adjacent domain in an attempt to create a perfect impedance match for both pressure
waves and shear waves, so that

∂u ∂u
σ ⋅ n = – ρc p  ⋅ n n – ρc s  ⋅ t t
∂t  ∂t 

where n and t are the unit normal and tangential vectors at the boundary, respectively,
and cp and cs are the speeds of the pressure and shear waves in the material. This
approach works best when the wave direction in close to the normal at the wall.

In the general case, you can use

∂u
σ ⋅ n = – d i ( ρ, c p, c s )
∂t

where the mechanical impedance di is a diagonal matrix available as the user input, and
by default it is set to

cp + cs
d i = ρ ---------------- I
2

More information about modeling using low-reflecting boundary

conditions can be found in Ref. 1.

Cyclic Symmetry and Floquet Periodic Conditions

These boundary conditions are based on the Floquet theory which can be applied to
the problem of small-amplitude vibrations of spatially periodic structures.

If the problem is to determine the frequency response to a small-amplitude

time-periodic excitation that also possesses spatial periodicity, the theory states that the
solution can be sought in the form of a product of two functions. One follows the
periodicity of the structure, while the other one follows the periodicity of the

LOADS AND BOUNDARY CONDITIONS | 387

 excitation. The problem can be solved on a unit cell of periodicity by applying the
corresponding periodicity conditions to each of the two components in the product.

This section describes the theory for solids in 3D, but is equally applicable
to shells. In the case of shells, the periodicity condition is applied to edges,
and the rotational degrees of freedom (displacements of the normal
vector) are treated in the same way as the translational displacements.

The problem can be modeled using the full solution without applying the above
described multiplicative decomposition. For such a solution, the Floquet periodicity
conditions at the corresponding boundaries of the periodicity cell are expressed as

u destination = exp [ – ik F ( r destination – r source ) ] u source

where u is a vector of dependent variables, r is the position, and the vector kF

represents the spatial periodicity of the excitation.

The cyclic symmetry boundary condition presents a special but important case of
Floquet periodicity, for which the unit periodicity cell is a sector of a structure that
consists of a number of identical sectors. The frequency response problem can then be
solved in one sector of periodicity by applying the periodicity condition. The situation
is often referred to as dynamic cyclic symmetry.

For an eigenfrequency study, all the eigenmodes of the full problem can be found by
performing the analysis on one sector of symmetry only and imposing the cyclic
symmetry of the eigenmodes with an angle of periodicity ϕ = mθ , where the cyclic
symmetry mode number m can vary from 0 to N/2, with N being the total number of
sectors so that θ = 2π/N.

The Floquet periodicity conditions at the sides of the sector of symmetry can be
expressed as

– iϕ T
u destination = e R θ u source

where the u represents the displacement vectors with the components given in the
default Cartesian coordinates. Multiplication by the rotation matrix given by

cos ( θ ) – sin ( θ ) 0
Rθ = sin ( θ ) cos ( θ ) 0
0 0 1

388 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

makes the corresponding displacement components in the cylindrical coordinate
system differ by the factor exp ( – iϕ ) only. For scalar dependent variables, a similar
condition applies, for which the rotation matrix is replaced by a unit matrix.

The angle ϕ represents either the periodicity of the eigenmode for an eigenfrequency
analysis or the periodicity of the excitation signal in case of a frequency-response
analysis. In the latter case, the excitation is typically given as a load vector

F = – F 0 exp [ – im atan ( Y ⁄ X ) ]

when modeled using the Cartesian coordinates. The parameter m is often referred to
as the azimuthal wave number.

Ref. 2 contains more information about cyclic symmetry conditions.

Thermal Expansion of Constraints

Constraints like Fixed Constraint and Prescribed Displacement will in general cause
stresses near the constrained boundaries when the structure undergoes temperature
changes. The same is true also for rigid objects like Rigid Domain, Rigid Connector, and
Attachment. By adding a Thermal Expansion subnode to these features, you can allow
the constrained boundaries to have a thermal expansion displacement.

The thermal strains will in general have a spatial distribution give by

ε ( X ) = α ( X, T ) ( T – T ref )

Note that this is the thermal expansion of the virtual surroundings of the structure
being analyzed, so it is unrelated to the thermal strains of the structure itself.

The strain field must be converted into a displacement field u(X) such that

1 ∂u i ∂u j
ε ij = ---  -------- + --------
2  ∂X j ∂x i 

If the strain field fulfills the general compatibility relations, it is in principle possible to
integrate the above relation. The procedure is outlined in Ref. 3, giving

X ∂ε il ∂ε kl
ui ( X ) = ui ( X0 ) + X  εil + ( Xk – X'k )  ---------
0
- – ----------  dX' l
∂X k ∂X i  

LOADS AND BOUNDARY CONDITIONS | 389

 Summation over double indices is implied. The rigid body rotation term given in
Ref. 3 is omitted, since it cannot be derived from the strain field. The reference point
is chosen so that the displacement (caused by the strain field) is zero, so that the
u i ( x 0 ) term can be omitted. The integral is path independent when the compatibility
is fulfilled. Since the constrained region is a virtual object, the integration path does
not have to be inside a domain. For simplicity, a straight line from X0 to X is used for
the integration. Let p be the vector between the two points,

p = X – X0

The distance along the integration path can then be parameterized by a parameter s
running from 0 to 1as

X' = X 0 + sp

giving

1 ∂ε il ∂ε kl
ui ( X ) = 0  εil + ( 1 – s )pk  ---------
- – ----------  p ds
∂X k ∂X i   l
(3-54)

This integral can be computed using the built-in integrate() operator as long as the
strain field is an explicit function of the material frame coordinates X.

For the physics interfaces which have rotational degrees of freedom (Beam, Shell and
Plate), not only the displacement, but also the rotation of the constraint is needed. For
a given displacement field u(X), the infinitesimal rotation vector Θ is given by

1
Θ = --- ∇×u
2

Applying the rotation operator to Equation 3-54 gives

2
1 ∂ε il ∂ ε il 
1
Θ m ( X ) = --- ε mni
2  0
 ( 3 – 2s ) ---------- + ( 1 – s )p k
 ∂ X n ∂ X k ∂X n
 p l ds


where εmni is the permutation tensor.

Note that the general compatibility requirements will not be fulfilled for arbitrary
expressions for the thermal strain distribution. In such cases, the stresses caused by the

390 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

constraints cannot completely be removed, but they will be significantly decreased.
The results will then depend on the choice of reference point.

Constraints and Thermal Expansion in the Structural Mechanics

Modeling chapter.

Rigid Connector Theory

The rigid connector is a special kinematic constraint, which can be attached to one or
several boundaries, edges or points. The effect is that all connected entities behave as
if they were connected by a common rigid body.

The only degrees of freedom needed to represent this assembly are the ones needed to
represent the movement of a rigid body. In 2D this is simply two in-plane translations,
and the rotation around the z-axis.

In 3D the situation is more complex. Six degrees of freedom, usually selected as three
translations and three parameters for the rotation, are necessary. For finite rotations,
however, any choice of three rotation parameters is singular at some specific set of
angles. For this reason, a four-parameter quaternion representation is used for the
rotations in COMSOL Multiphysics. Thus, each rigid connector in 3D actually has
seven degrees of freedom, three for the translation and four for the rotation. The
quaternion parameters are called a, b, c, and d, respectively. These four parameters are
not independent, so an extra equation stating that the following relation is added:

2 2 2 2
a +b +c +d = 1

The connection between the quaternion parameters and a rotation matrix R is

2 2 2 2
a +b –c –d 2bc – 2 ad 2ac + 2bd
R = 2ad + 2bc
2 2
a –b +c –d
2 2
2cd – 2ab
2 2 2 2
2bd – 2 ac 2ab + 2cd a –b –c +d

Under pure rotation, a vector from the center of rotation (Xc) of the rigid connector
to a point X on the undeformed object is rotated into

x – Xc = R ⋅ ( X – Xc )

LOADS AND BOUNDARY CONDITIONS | 391

 where x is the new position of the point originally at X. The displacement is by
definition

u = x – X = ( R – I ) ⋅ ( X – Xc )

where I is the unit matrix.

When the center of rotation of the rigid connector also has a translation uc, then the
complete expression for the rigid body displacements is

u = ( R – I ) ⋅ ( X – Xc ) + uc

The total rotation of the rigid connector can be also presented as a rotation vector. Its
definition is

b
2 acos ( a )
Θ = ----------------------------------- c
2 2 2
b +c +d d

The parameter a can be considered as measuring the rotation, while b, c, and d can be
interpreted as the orientation of the rotation vector. For small rotations, this relation
simplifies to

b
Θ = 2 c
d

The rotation vector is available as the variables thx_tag, thy_tag, and thz_tag. Here
tag is the tag of the Rigid Connector node in the Model Builder tree.

It is possible to apply forces and moments directly to a rigid connector. A force

implicitly contributes also to the moment if it is not applied at the center of rotation
of the rigid connector. The directions of the forces and moments are fixed in space and
do not follow the rotation of the rigid connector.

Contact Analysis Theory

In COMSOL Multiphysics you can model contact between a group of boundaries in
2D or 3D. There are two algorithms available: an augmented Lagrangian method
and a penalty method.

392 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

In the augmented Lagrangian method, the system of equations is solved in a
segregated way. Augmentation components are introduced for the contact pressure Tn
and the components Tti of the friction traction vector Tt. An additional iteration level
is added where the usual displacement variables are solved separated from the contact
pressure and traction variables. The algorithm repeats this procedure until it fulfills a
convergence criterion.

In the penalty method, no extra degrees of freedom are needed for the contact pressure
or the friction traction vector. These results are just computed from the displacements
and the penalty stiffness. This means that no special solver strategies are necessary.

• Contact Modeling
• Documentation of the Contact, Friction and Adhesion features
• Identity and Contact Pairs in the COMSOL Multiphysics Reference
Manual

In the following equations F is the deformation gradient matrix. When looking at

expressions evaluated on the destination boundaries, the expression map (E) denotes
the value of the expression E evaluated at a corresponding source point, and g is the
current gap distance between the destination and source boundary.

Both the contact map operator map (E) and the gap distance variable are defined by
the contact element elcontact. For each destination point where the operator or gap
is evaluated, a corresponding source point is sought by searching in the direction
normal to the destination boundary.

Source
map(x)

g
x
Destination

Before the boundaries come in contact, the source point found is not necessarily the
point on the source boundary closest to the destination point. However, as the
boundaries approach one another, the source point converges to the closest point as
the gap distance approaches zero.

LOADS AND BOUNDARY CONDITIONS | 393

 It is possible to add an offset value both to the source (osrc), and to the destination
(odst). If an offset is used, the gap will be computed with the geometry treated as larger
(or smaller, in the case of a negative offset) than what is actually modeled. The offset
is given in a direction normal to the boundary. It can vary along the geometry, and
may also vary in time (or by parameter value when using the continuation solver).

The effective gap distance dg used in the analysis is thus

o src
d g = g – o dst – --------------------------------------
– ( n ⋅ map ( n ) )

The correction of the source side offset is necessary since the normal on the source side
is not necessarily pointing towards the current point on the destination. When in
contact, the normals will be exactly opposite.

In some cases, the discretization caused by the meshing of curved boundaries will give
small irregularities in the gap values. You can then choose to adjust the gap by the
computed gap value in the initial configuration, ginit. In this case, the effective gap
distance is

o src
d g = g – g init – o dst – --------------------------------------
– ( n ⋅ map ( n ) )

When including friction in the contact problem, it is important to keep track of not
only the gap in the normal direction, but also the slip between the boundaries in the
tangential direction. The slip s since the last converged step is defined as

s = map ( F ) ( X m – X m, old )

where the index ‘m’ indicates that the coordinates are taken as the mapped coordinates
from the source side,

X m = map ( X )

and Xm,old is the value of Xm in the last converged time or parameter step. The
coordinates are material (undeformed) coordinates. The slip vector is thus
approximated using a backward Euler step. The deformation gradient F contains
information about the local rotation and stretching.

AUGMENTED LAGRANGIAN METHOD

Using the special gap distance variable (solid.gap), the penalized contact pressure
Tnp is defined on the destination boundary as

394 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

  Tn –pn dg if d g ≤ 0

T np =  pn dg (3-55)
– ----------
-
 Tn
 Tn e otherwise

pn is the user-defined normal penalty factor.

The penalized friction traction Ttp is defined on the destination boundary as:

T t, crit
T tp = min  -------------------, 1 T t, trial (3-56)
 T t, trial 

where Tt,trial is defined as

T t, trial = T t – p t s (3-57)

In Equation 3-57, pt is the user-defined friction traction penalty factor. The critical
sliding resistance, Tt,crit, is defined as

T t, crit = min ( μT np + T cohe, T t, max ) (3-58)

In Equation 3-58, μ is the friction coefficient, Tcohe is the user-defined cohesion

sliding resistance, and Tt,max is the user-defined maximum friction traction.

In the following equation δ is the variation (represented by the test() operator in

COMSOL Multiphysics). The contact interaction gives the following contribution to
the virtual work on the destination boundary:

 ( Tnp δg + Ttp ⋅ map ( F )δXm ) dA +  ( wcn δTn + wct ⋅ δTt ) dA

dst dst

where wcn and wct are contact help variables defined as:

w cn = T np, i – T n, i + 1

w ct = ( γ fric ( T tp – ( n ⋅ T tp )n ) ) i – T t, i + 1

where i is the augmented solver iteration number and γfric is a Boolean variable stating
if the parts are in contact or not.

PENA LT Y METHOD
The penalty algorithm is essentially based on the insertion of a stiff spring between the
contacting boundaries. The penalty factor can be interpreted as the stiffness of that

LOADS AND BOUNDARY CONDITIONS | 395

 spring. A high value of the stiffness fulfills the contact conditions more accurately, but
if it is too high, it makes the problem ill-conditioned and unstable. The contact
pressure in the normal direction is computed as


 – p n d g + p 0 if d g < p 0 ⁄ p n
Tn = 
 0 if d g ≥ p 0 ⁄ p n


where dg is the effective gap distance, pn is the penalty factor, and p0 is the pressure at
zero gap. The latter parameter can be used to reduce the overclosure between the
contacting boundaries if an estimate to the contact pressure is known. In the default
case, when p0 = 0, there must always be some overclosure (a negative gap) in order for
a contact pressure to develop.

In case of friction, the tangential force Tt is computed as

T t, crit
T t = min  -------------------, 1 T t, trial
T t, trial

where

T t, trial = – p t s

This can be viewed as a tangential spring giving a force proportional to the sliding
distance. The penalty factor pt can be identified as the spring constant. The sticking
condition is thus replaced by a stiff spring, so that there is a small relative movement
even if the force required for sliding is not exceeded.

The definition of Tt,crit depends on the algorithm used for computing the normal
contact.


 min ( μT np + T cohe, T t, max ) augmented Lagrangian method
T t, crit = 
 min ( μT n + T cohe, T t, max ) penalty method


The same notation as for the description of the augmented Lagrangian method above
is used.

The springs between the contacting boundaries are added using weak expressions on
the destination boundary.

396 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

  ( Tnp δg + Ttp ⋅ map ( F )δXm ) dA
dst

FRICTION MODELS
The friction model is either no friction or Coulomb friction.

The frictional coefficient μ is defined as


 μ + ( μ s – μ d ) exp ( – α dcf v s ) if dynamic friction
μ =  d
 μs otherwise


where μs is the static frictional coefficient, μd is the dynamic frictional coefficient, vs is

the slip velocity, and αdcf is a decay coefficient.

DIRECTIONS OF THE CONTACT FORCES

Since the contact model always makes use of the geometric nonlinearity assumption,
the contact pressure and the friction force represent, respectively, the normal and
tangential components of the nominal traction. This is the force density with
components in the current configuration (that is, on the spatial frame) but related to
the undeformed area of the corresponding surface element.

ADHESION
You can model a situation where two boundaries stick together once the get into
contact by adding an Adhesion subnode to Contact. Adhesion can be modeled when
the penalty contact method is used. The adhesion formulation can be viewed as if a
thin elastic layer is placed between the source and destination boundaries when
adhesion is activated.

The adhesion starts acting when the adhesion criterion is met in the previous time or
parameter step. An auxiliary degree of freedom located at Gauss points is used as an
indicator to whether the adhesion criterion has been met or not.

When adhesion is active, the tensile stress σn in the normal direction is computed from
the effective gap dg as

σn = –Kp dg

The default value for the stiffness in the normal direction Kp is the penalty stiffness of
the contact condition, but you can also enter it explicitly. In compression, the penalty

LOADS AND BOUNDARY CONDITIONS | 397

 stiffness is always used. This leads to a bilinear stiffness in the normal direction when
Normal stiffness is User Defined.

The shearing deformation is resisted using a shear stiffness K τ , related to the normal
stiffness by

Kτ = nτ Kp

where n τ is a coefficient with the default value 0.17. This coefficient can either be
input explicitly, or be computed from a Poisson’s ratio. A plane strain assumption is
used for this conversion, giving

1 – 2ν
n τ = ---------------------
2(1 – ν)

The shear traction is proportional to the relative deformation between the source and
destination boundaries measured from the position when the adhesion criterion was
first fulfilled, so that

t = – K τ sF t

where s is the slip vector.

DECOHESION
When adhesion is active, it is possible to break the bond between the source and
destination boundaries by specifying a Traction separation law. Three different such
decohesion laws are available.

The decohesion behavior in both tension or shear is based on two quantities, the
maximum stress and the energy release rate. Tearing in the normal direction is called
mode I, and shear is called mode II. In the theory below a subscript i can take the
values I and II.

Up to the maximum stress, the adhesive layer is linearly elastic. At higher deformation,
the stress decreases, and a damage is assumed to have occurred. The damage function
is a scalar function of the maximum relative displacement between the two boundaries.
The stiffness in the damaged state is permanently reduced, even on unloading.

In general, the stress state is a combination of tension and shear, so the data for the
two modes must be combined into a multiaxial decohesion rule. For two of the
decohesion laws you can specify details of the multiaxial behavior.

398 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

As a measure of the displacement in the adhesive layer, the mixed mode relative
displacement um, is defined as weighted combination of the normal direction
displacement uI and the tangential displacement uII.

2 2
um =  u I + u II

In this expression, the notation


 x =  x, if x ≥ 0
 0, if x < 0

has been used. This is the displacement which is compared with the various mixed
mode failure displacements described below.

Linear separation law

In the linear separation law, the decreasing part of the stress vs. displacement curve is
linear. The graph in Figure 3-18 is generic, and valid both for mode I and mode II.

σi
σic

ui0 uif

Figure 3-18: Stress vs. displacement curve for the linear separation law

The failure initiation displacement ui0 for each mode is determined from the
maximum stress and the elastic stiffness, as

σ Ic
u I0 = --------
Kp
σ IIc
u II0 = ----------
Kτ

LOADS AND BOUNDARY CONDITIONS | 399

 The decohesion displacement uif for each mode is implicitly determined from the
given energy release rate (which is the area under the stress vs. displacement curve):

2G ic
u if = ------------
σ ic

The mixed mode failure initiation displacement is then defined as

2
um
u 0 = u I0 u II0 ------------------------------------------------
-
2 2 2 2
 u I u II0 + u II u I0

The implication is that compressive displacements are ignored, and only tensile and
tangential displacements contribute.

The final mixed mode failure displacement depends on the selected Failure criterion.
The Power Law failure criterion is defined as

G I  η  G II  η
 --------
- -----------
 G Ic +  G IIc = 1

whereas the Benzeggagh-Kenane failure criterion is defined as

G II η
G Ic + ( G IIc – G Ic )  ---------------------- = G I + G II
G I + G II

The exponent η is called the Mode mixity exponent.

From these relations, the total decohesion failure displacement in the mixed mode can
be computed. For a power law criterion, the expression is

 –1
 2  K η  β2 K   η
η ------
 2 ( 1 + β )-  --------
  - +  ------------- 
p τ
 u f = ----------------------- uI > 0
 u 0  G Ic  G IIc  

 u f = u IIf uI ≤ 0

 β=u II
-------
-
 uI


For the Benzeggagh-Kenane criterion, the corresponding expression is

400 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

  η
 2u m
2   n u2  
τ
 u f = ----------------------------------------------  G Ic + ( G IIc – G Ic )  --------------------------- 
II
uI > 0
 2 2   2 2 
u 0 K p ( u I + n τ u II )   u I + n τ u II 


 u f = u IIf uI ≤ 0

The damage function can be described as

uf u – u0
D ( u ) = min  1, max  0, -----  -----------------  
u uf – u0

The secant stiffness of the damaged adhesive layer (indicated by the red line in
Figure 3-18) is

K I = ( 1 – D )K p
K II = ( 1 – D )K τ

Polynomial separation law

The polynomial separation law differs from the linear separation law in that the
decreasing part of the stress curve is represented by a cubic polynomial. This
polynomial is chosen so that it has zero slope in both ends, which means that the secant
stiffness does not have any discontinuities. The area under the cubic function is the
same as that of the linear function when the failure displacement uif is the same. That
is, for a given energy release rate Gic, the same uif is obtained.

σi
σic

ui0 uif

Figure 3-19: Stress vs. displacement curve for the polynomial separation law

LOADS AND BOUNDARY CONDITIONS | 401

 All expressions presented under Linear separation law are still valid except the damage
function, which now is

u0 u – u0 2 u – u 0
D ( u ) = min  1, max  0, ------  1 +  -----------------  2 ----------------- – 3   
u uf – u0 uf – u0

Multilinear separation law

The multilinear separation law introduces a region of constant stress, similar to a
perfectly plastic material model, as shown in Figure 3-20.

σi
σic

ui0 u uip uif

Figure 3-20: Stress vs. displacement curve for the multilinear separation law

This law requires one more material parameter λ, describing the width of the “plastic”
region. The Shape factor λ is the ratio between the plastic (constant stress) part of Gic
and the total “inelastic” part of Gic:

σ ic ( u ip – u i0 ) σ ic ( u ip – u i0 )
λ = --------------------------------------------------------------------------------- = -----------------------------------
1 1
σ ic ( u ip – u i0 ) + --- σ ic ( u if – u ip ) G ic – --- σ ic u i0
2 2

Note that the shape factor is assumed to be the same for both tension and shear. The
value λ=0 corresponds to the linear separation law.

The plastic displacement uip can thus be expressed as

G ic λ
u ip = λ -------- + u i0  1 – ---
σ ic  2

The decohesion displacement is

402 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

 G ic λu i0
u if = ( 2 – λ ) -------- + ------------
σ ic 2

In this model, a linear mixed mode failure law is assumed:

G I   G II 
 --------
- + ----------- = 1
 G Ic  G IIc

The mixed mode plastic failure displacement up is defined as

2
um
u p = u Ip u IIp ------------------------------------------------
-
2 2 2 2
 u I u IIp + u II u Ip

The mixed mode failure displacement is

2
1+β
- – ( up – u0 )
u f = 2G Ic G IIc -----------------------------------------------------------
2
u 0 ( K p G IIc + β K τ G Ic )

The damage function can be written as

   
   1–u -----0- u < u p  
   u
D ( u ) = min  1, max  0,  
   u0  uf – u  
   1 – ------  ------------------ u ≥ up  
   u uf – up 

LOADS AND BOUNDARY CONDITIONS | 403

 Stress Linearization
Stress linearization is a procedure in which the stress distribution along a line through
the thickness in a solid is approximated with an equivalent linear stress distribution,
similar to what would be the result of an analysis using shell theory. The line is
commonly referred to as a stress classification line, SCL.

First, the computed stresses are transformed into a local orthonormal coordinate
system x1-x2-x3 where the x1 direction is oriented along the SCL. In 3D, you must
specify the x2 direction, and thus implicitly the x3 direction. In 2D and 2D axial
symmetry, the x3 direction is in the out-of-plane direction, that is the Z and azimuthal
directions respectively.

The length of the SCL, which is assumed to be straight, start on one boundary, and
end on the opposite boundary, is denoted L.

The membrane stress tensor is the average of each local stress component along the
SCL:

L
1
σ m, ij 
= ---- σ ij dx 1
L
0

Each component of the bending stress tensor is assumed to have a linear variation along
the SCL, with the value being zero at the midpoint.

2x 1
σ b, ij =  1 – --------- σ b ( max ), ij
 L 

The maximum bending stress is defined so that the linear stress distribution has the
same moment as the true stress distribution.

σ b ( max ), ij = ------2 σ ij  ---- – x 1 dx 1

6 L
L
 2
0

The linearized stress distribution is the sum of the membrane and bending stresses,

σ mb, ij ( x 1 ) = σ m, ij + σ b, ij ( x 1 )

404 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

Finally the peak stress tensor is defined at the two end points of the SCL. It is the
difference between the actual stress tensor, and the linearized stress representation.

σ p ( start ), ij = [ σ ij – σ mb, ij ] x =0
= [ σ ij ] x =0
– σ m, ij + σ b ( max ), ij
1 1

σ p ( end ), ij = [ σ ij – σ mb, ij ] x =L
= [ σ ij ] x =L
– σ m, ij – σ b ( max ), ij
1 1

The stress intensity, also known as the Tresca effective stress, is often the ultimate goal
of a stress linearization. It is computed as a worst case of the effective stress at the two
ends of the SCL. The stress intensity is computed from the principal stresses as

σ int = σ p1 – σ p3

When computing the principal stresses at the end points of the SCL, it is customary to
ignore the bending part of the through-thickness oriented stresses. The principal
stresses and the stress intensity variables are the end points are thus computed using
the following stress tensor:

σ 11 σ m, 11
σ 22 σ mb, 22
σ 33 σ mb, 33
=
σ 12 σ m, 12
σ 13 σ m, 13
σ 23 σ mb, 23

Section forces, similar to what would be computed in a shell or plate analysis, is another
type of result quantities available for each SCL. The in-plane forces are computed from
the membrane stresses as

N 22 σ m, 22
N 33 = L σ m, 33
N 23 σ m, 23

The bending moments are computed from the bending stresses as

M 22 2
σ b ( max ), 22
L
M 33 = ------ σ b ( max ), 33
6
M 23 σ b ( max ), 23

STRESS LINEARIZATION | 405

 The out-of-plane shear forces are computed from the membrane stresses as

Q2 σ m, 12
= L
Q3 σ m, 13

406 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

Energy Quantities
There are several energy quantities computed in the Structural Mechanics interfaces.
This section is a summary of these quantities and how to compute in different
situations.

Elastic Energy

LINEAR ELASTIC MATERIALS

The elastic energy is defined as the recoverable energy stored in an elastic material or
spring. The elastic strain energy density in an elastic material is defined as

ε el

Ws =  σ: dεel (3-59)
0

If the linear elasticity is assumed, then

σ = σ i + C:ε el

where σi is the initial stress. The integration can then be carried out analytically and
the result is

1 1
W s = σ i :ε el + --- ( C:ε el ):ε el = --- ( σ i + σ ):ε el
2 2

This expression is used for the Linear Elastic Material model in a Stationary or Time
Dependent analysis. An implication is that if you modify the linear elastic model in a
way that violates the assumption about stress-strain linearity above, then the computed
strain energy density may be wrong, for example, using a strain dependent Young’s
modulus or a non-constant initial stress.

In the case of Frequency Domain analysis, only the harmonic part is considered. That
is, a constant pre-stress does not contribute to the strain energy density. To emphasize
this, the concept of stored energy is used. The stored energy is the cycle average of the
elastic energy; that is,

ENERGY QUANTITIES | 407

 T
1 1
Wh
T 2 
= ---- --- σ ( t ):ε ( t ) dt
0

The harmonic stress and strain components are generally not in phase with each other,
so the cycle average of the stored energy is computed as

1
W h = --- real ( σ:ε )
4

where the stress and strain are considered as complex quantities, and the overbar
denotes a complex conjugate.

HYPERELASTIC MATERIALS
For a Hyperelastic material, the strain energy density function is the fundamental
quantity from which stresses are derived. The form of the strain energy density
function is determined by the hyperelastic model used.

Nonlinear Elastic Materials

For a Nonlinear Elastic Material, the strain energy density is computed in different
ways depending on the material model selected. If the integration in Equation 3-59
can be performed analytically, then a closed form expression is used, similar to what is
done in the linear elastic material. If not, then the integral is actually computed using
the integrate() operator.

Structural Elements
For structural elements, the strain energy density is split into membrane, bending and
shear parts, which are then summed into a total strain energy density.

The strain energy density for all elastic domains are integrated to give a total elastic
strain energy, which contains all elastic energy stored in a certain physics interface.

Elastic boundary conditions, such as Spring Foundation, Thin Elastic Layer, and
Springs in joints in the Multibody Dynamics interface, also contribute to the total
elastic strain energy variable. In these cases linearity is assumed, so if you enter
nonlinear data, you will probably need to adjust the strain energy expressions.

The Multibody Dynamics interface is available with the addition of the

Multibody Dynamics Module.

408 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

Kinetic Energy
For all materials, a kinetic energy density is computed. For a Time Dependent study,
it is defined as

1 2
W k = --- ρv
2

where ρ is the mass density, and v is the velocity.

In Frequency Domain, the kinetic energy is defined as

1 2 2
W k = --- ρω v
4

which represents the cycle average.

Dissipated Energy
There are many possible mechanisms for energy dissipation in a structure:

• Material damping (Loss factor, Rayleigh damping, or Viscous damping)

• Viscoelasticity
• Dissipation by plasticity or creep
• Viscous damping in boundary conditions, springs and joints
• Friction in mechanical contact

The general form of dissipation loss is

·
Q h = σ:ε

The treatment of dissipated energy is fundamentally different depending on whether

the analysis is in frequency domain or not. For stationary or time dependent cases, the
dissipated energy must be accumulated over parameter ranges or time, which means
extra degrees of freedom must be added. For this reason, you must explicitly select to
compute the mechanical dissipation in these cases. This is done in the Energy
Dissipation section of the material model.

In frequency domain, the dissipated energy per cycle is computed using a closed form
expression. It is always available as postprocessing variables. The expression used is

1
Q h = --- real ( σ: ( iωε ) )
2

ENERGY QUANTITIES | 409

 In the case of Rayleigh damping, the stresses are not directly affected by the damping,
since it is not part of the stress-strain relation. In a problem which is linear elastic
except for the Rayleigh damping, the stresses and strains are in phase with each other,
and does thus not give any damping contribution. This is handled by explicitly adding
an extra contribution to dissipation.

Mechanical Energy Flux

The mechanical energy flux is a vector formed by the multiplication of the stress tensor
and the velocity vector

I = –σ ⋅ v

The reason for the minus sign in the definition is that if you put a pressure on an
external boundary, and it moves in the direction of the load, then a positive power
input in the direction of the load is obtained.

In the time domain, the expression above is used. In frequency domain, two versions
are supplied. The complex mechanical energy flux is the complex vector formed by
multiplying the stress tensor by the complex conjugate of the velocity:

I = – σ ⋅ v˜

The mechanical energy flux is in the frequency domain defined as a real quantity, the
cycle average of the complex mechanical energy flux.

1
I = --- real ( – σ ⋅ v˜ )
2

Energy Variables
The energy variables used in the Structural Mechanics interfaces are summarized in
Table 3-6.
TABLE 3-6: ENERGY VARIABLES USED IN STRUCTURAL MECHANICS

VARIABLE DESCRIPTION SI UNIT PHYSICS COMMENT

INTERFACE
1

phys.Ws Elastic strain J/m3 solid, mbd, shell,

energy density plate, mbrn,
beam, truss
phys.WsM Membrane strain J/m2 shell, plate Contributes to
energy density phys.Ws

410 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

TABLE 3-6: ENERGY VARIABLES USED IN STRUCTURAL MECHANICS

VARIABLE DESCRIPTION SI UNIT PHYSICS COMMENT

1
INTERFACE

phys.WsB Bending strain J/m2 shell, plate Contributes to

energy phys.Ws
phys.WsS Shear strain J/m2 shell, plate Contributes to
energy phys.Ws
phys.WsM Membrane strain J/m beam Contributes to
energy density beam.WsL
phys.WsB Bending strain J/m beam Contributes to
energy beam.WsL
phys.WsS Shear strain J/m beam Contributes to
energy beam.WsL
beam.WsT Torsional strain J/m beam Contributes to
energy beam.WsL
phys.WsL Strain energy J/m beam, truss phys.Ws =
density per unit phys.WsL/area
length
phys.Ws_tot Total elastic strain J solid, mbd, shell, Global variable
energy plate, mbrn, containing
beam, truss integration and
summation of all
phys.Ws
contributions in a
physics interface.
phys.Wk Kinetic energy J/m3 solid, mbd, shell,
density plate, mbrn,
beam, truss
phys.Wk_tot Total kinetic J solid, mbd, shell, Global variable
energy plate, mbrn, containing
beam, truss integration and
summation of all
phys.Wk
contributions in a
physics interface.
phys.Wp Plastic dissipation J/m3 solid Energy density
density dissipated by
plasticity

ENERGY QUANTITIES | 411

 TABLE 3-6: ENERGY VARIABLES USED IN STRUCTURAL MECHANICS

VARIABLE DESCRIPTION SI UNIT PHYSICS COMMENT

1
INTERFACE

phys.Wp_tot Total plastic J solid Global variable

dissipation containing
integration and
summation of all
phys.Wp
contributions in a
physics interface.
phys.Wc Creep dissipation J/m3 solid Energy density
density dissipated by creep.
phys.Wc_tot Total creep J solid Global variable
dissipation containing
integration and
summation of all
phys.Wp
contributions in a
physics interface.
phys.Wv Viscoelastic J/m3 solid Energy density
dissipation density dissipated by
viscosity.
phys.Wv_tot Total viscoelastic J solid Global variable
dissipation containing
integration and
summation of all
phys.Wv
contributions in a
physics interface.
phys.Wed Dielectric J/m3 solid Piezoelectric
dissipation density electrical damping
phys.Wed_tot Total dielectric J solid Global variable
dissipation containing
integration and
summation of all
phys.Wed
contributions in a
physics interface.
phys.Wmd Damping J/m3 solid Piezoelectric
dissipation density mechanical damping

412 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

TABLE 3-6: ENERGY VARIABLES USED IN STRUCTURAL MECHANICS

VARIABLE DESCRIPTION SI UNIT PHYSICS COMMENT

1
INTERFACE

phys.Wmd_tot Total damping J solid Global variable

dissipation containing
integration and
summation of all
phys.Wmd
contributions in a
physics interface.
phys.I* Mechanical W/m2 solid, mbd
energy flux, *
coordinate
phys.Icomplex* Complex W/m2 solid, mbd Frequency domain
mechanical only
energy flux,
*coordinate
1
The availability of physics interfaces is based on the license for a module.

ENERGY QUANTITIES | 413

 References
REFERENCES FOR NONLINEAR STRUCTURAL MODULE MATERIALS
1. G.A. Holzapfel, Nonlinear Solid Mechanics: A Continuum Approach for
Engineering, John Wiley & Sons, 2000.

2. R.S. Rivlin and D.W. Saunders, “Large Elastic Deformations of Isotropic Materials.
VII. Experiments on the Deformation of Rubber,” Phil. Trans. R. Soc. Lond. A, vol.
243, no. 865, pp. 251–288, 1951.

3. E.M. Arruda and M.C. Boyce, “A Three-Dimensional Constitutive Model for the
Large Stretch Behavior of Rubber Elastic Materials,” J. Mech. Phys. Solids, vol. 41, pp.
389–412, 1993.

4. J.C. Simo and T.J.R. Hughes, Computational Inelasticity, Springer, 1998.

5. J. Lubliner, Plasticity Theory, Dover, 2008.

6. K. Grote and E. Antonsson, Springer Handbook of Mechanical Engineering,

Springer, 2009.

7. R. Hertzberg, Deformation and Fracture Mechanics of Engineering Materials,

John Wiley & Sons, 1996.

8. F. Garofalo, “An Empirical Relation Defining the Stress Dependence of Minimum

Creep Rate in Metals,” Trans AIME, vol. 227, no. 351, 1963.

9. S. Brown, K. Kim, and L. Anand, “An Internal Variable Constitutive Model for Hot
Working of Metals,” Int. J. Plasticity, vol. 5, pp. 95–130, 1989.

10. M. Attard, “Finite Strain - Isotropic Hyperelasticity,” Int. J. Solids Struct., vol.
40, pp. 4353–4378, 2003.

11. I. Babuska and M. Suri, “On Locking and Robustness in the Finite Element
Method,” SIAM Journal on Numerical Analysis, vol. 29, pp. 1261–1293, 1992.

12. A.F. Bower, Applied Mechanics of Solids, CRC Press, 2009.

13. A. Gent, “A new constitutive relation for rubbers,” Rubber Chem. Technol., vol.
69, pp. 59–61, 1996.

14. C. Horgan and G. Saccomandi, “Constitutive Models for Compressible

Nonlinearly Elastic Materials with Limiting Chain Extensibility,” J. Elasticity, vol. 7,
pp. 123–138, 2004.

414 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

15. B. Storakers, “On Material Representation and Constitutive Branching in Finite
Compressible Elasticity,” J. Mech. Phys. Solids, vol. 34, pp. 125–145, 1986.

16. Y. Gao, “Large Deformation Field Near a Crack Tip in a Rubber-like Material,”
Theor. Appl. Fract. Mec., vol. 26, pp. 155–162, 1997.

REFERENCES FOR THE GEOMECHANICS MODULE MATERIALS

1. W.F. Chen and E. Mizuno, Nonlinear Analysis in Soil Mechanics: Theory and
Implementation (Developments in Geotechnical Engineering), 3rd ed., Elsevier
Science, 1990.

2. B. Bresler and K.S. Pister, “Strength of Concrete Under Combined Stresses,” ACI
Journal, vol. 551, no. 9, pp. 321–345, 1958.

3. H. Matsuoka and T. Nakai, “Stress-deformation and Strength Characteristics of Soil

Under Three Different Principal Stresses,” Proc. JSCE, vol. 232, 1974.

4. H. Matsuoka and T. Nakai, “Relationship Among Tresca, Mises, Mohr-Coulomb,

and Matsuoka-Nakai Failure Criteria,” Soils and Foundations, vol. 25, no. 4, pp.123–
128, 1985.

5. H.S. Yu, Plasticity and Geotechnics, Springer, 2006.

6. V. Marinos, P. Marinos, and E. Hoek, “The Geological Strength Index: Applications

and Limitations,” Bull. Eng. Geol. Environ., vol. 64, pp. 55–65, 2005.

7. J. Jaeger, N. G. Cook, and R. Zimmerman, Fundamentals of Rock Mechanics, 4th

ed., Wiley-Blackwell, 2007.

8. G. C. Nayak and O. C. Zienkiewicz, “Convenient Form of Stress Invariants for

Plasticity,” J. Struct. Div. ASCE, vol. 98, pp. 949–954, 1972.

9. A.J. Abbo and S.W. Sloan, “A Smooth Hyperbolic Approximation to the

Mohr-Coulomb Yield Criterion,” Computers and Structures, vol. 54, no. 3, pp. 427–
441, 1995.

10. K.J. Willam and E.P. Warnke, “Constitutive Model for the Triaxial Behavior of
Concrete,” IABSE Reports of the Working Commissions, Colloquium (Bergamo):
Concrete Structures Subjected to Triaxial Stresses, vol. 19, 1974.

11. B.H.G. Brady and E.T. Brown, Rock Mechanics for Underground Mining, 3rd
ed., Springer, 2004.

12. H.A. Taiebat and J.P. Carter, “Flow Rule Effects in the Tresca Model,” Computer
and Geotechnics, vol. 35, pp. 500–503, 2008.

REFERENCES | 415
 13. A. Stankiewicz and others, “Gradient-enhanced Cam-Clay Model in Simulation of
Strain Localization in Soil,” Foundations of Civil and Environmental Engineering,
no.7, 2006.

14. D.M. Wood, Soil Behaviour and Critical State Soil Mechanics, Cambridge
University Press, 2007.

15. D.M. Potts and L. Zadravkovic, Finite Element Analysis in Geothechnical

Engineering, Thomas Telford, 1999.

16. W. Tiecheng and others, Stress-strain Relation for Concrete Under Triaxial
Loading, 16th ASCE Engineering Mechanics Conference, 2003.

17. W.F. Chen, Plasticity in Reinforced Concrete, McGraw-Hill, 1982.

18. N. Ottosen, “A Failure Criterion for Concrete,” J. Eng. Mech. Division, ASCE,
vol. 103, no. 4, pp. 527–535, 1977.

19. N. Ottosen and M. Ristinmaa, The Mechanics of Constitutive Modelling, Elsevier,

2005.

REFERENCES TO THE NONLINEAR ELASTIC MATERIAL THEORY

1. Ramberg, W. and Osgood, W. R., “Description of stress-strain curves by three
parameters,” NACA Technical Note, no. 902, 1943.

2. Bower, A. F., Applied Mechanics of Solids, CRC Press, 2009.

3. Hardin, B.O., and Drnevich, V.P., “Shear modulus and damping in soils: Design
equations and curves,” J. Soil Mechanics and Foundations Division, ASCE 98, pp.
667–692, 1972.

4. Duncan, J.M. and Chang, C.Y., “Nonlinear analysis of stress and strain in soils,” J.
Soil Mechanics and Foundations Division, vol. 96, no. 5, pp. 1629–1653, 1970.

5. Duncan, J.M., Byrne, P., Wong, K.S., and Mabry, P., “Strength, stress strain and
bulk modulus parameters for finite element analysis of stresses and movements in soil,”
Geotechnical Engineering Report: UCB/GT/80-01, University of California,
Berkeley, 1980.

6. Selig, E.T., Soil Parameters for Design of Buried Pipelines, Pipeline

Infrastructure, B. A. Bennett, ed., ASCE, New York, pp. 99–116, 1988.

416 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

REFERENCES FOR ELASTOPLASTIC MATERIALS
7. P. Armstrong and C. Frederick, “A Mathematical Representation of the Multiaxial
Bauschinger Effect”, Technical Report RD/B/N731 CEGB, 1966. Reprinted in
Mater. High Temp. 24 (1) pp. 1–26, 2007.

8. M. Jirasek and Z. Bazant, Inelastic Analysis of Structures, Ch.20 General Elastoplastic

Constitutive Models. Wiley, 2001.

9. J. Simo and T. Hughes, Computational Inelasticity, Springer, 1998.

10. J. Simo, “Algorithms for Static and Dynamic Multiplicative Plasticity that Preserve
the Classical Return Mapping Schemes of the Infinitesimal Theory,” Computer
Methods in Applied Mechanics and Engineering, vol. 99, pp. 61–112, 1992.

11. J. Lubliner, Plasticity Theory, Dover, 2008.

12. R. Hill, “A Theory of the Yielding and Plastic Flow of Anisotropic Metals,” Proc.
Roy. Soc. London, vol. 193, pp. 281–297, 1948.

13. N. Ottosen and M. Ristinmaa, The Mechanics of Constitutive Modeling, Elsevier

Science, 2005.

14. S. Shima and M. Oyane. “Plasticity theory for porous metals,” International
Journal of Mechanical Sciences, vol. 18, pp. 285–291, 1976.

15. A. Gurson, “Continuum theory of ductile rupture by void nucleation and growth:
Part I – Yield criteria and flow rules for porous ductile media,” Journal of Engineering
Materials and Technology, vol. 99, pp. 2–15, 1977.

16. V. Tvergaard and A. Needleman, “Analysis of the cup-cone fracture in a round

tensile bar,” Acta Metallurgica, vol. 32, pp. 157–169, 1984.

17. A. Needleman and V. Tvergaard. “An analysis of ductile rupture in notched bars,”
Journal of the Mechanics and Physics of Solids, vol. 32, pp. 461– 490, 1984.

18. K. Nahshon and Z, Xue. “A modified Gurson model and its application to
punch-out experiments,” Engineering Fracture Mechanics, vol. 76, pp. 997– 1009,
2009.

19. N. Fleck, L. Kuhn, and R. McMeeking. “Yielding of metal powder bonded by

isolated contacts,” Journal of the Mechanics and Physics of Solids, vol. 40, pp. 1139–
1162, 1992.

20. P. Redanz. “Numerical modelling of the powder compaction of a cup,” European

Journal of Mechanics - A/Solids, vol. 18, pp.399– 413, 1999.

REFERENCES | 417
 21. J. Cedergren, N. Sorensen, and A. Bergmark. “Three-dimensional analysis of
compaction of metal powder,” Mechanics of Materials vol. 34, pp. 43–59, 2004.

22. S. Tsai and E. Wu. “A general theory of strength for anisotropic materials,” Journal
of Composite Materials, vol. 5, pp. 58-80, 1971.

23. P. Hopkins. “Benchmarks for Membrane and Bending Analysis of Laminated Shells.
Part 2: Strength Analysis,” NAFEMS Ltd, 2005.

REFERENCES FOR THE PIEZOELECTRIC DEVICES MULTIPHYSICS

INTERFACE
1. R. Holland and E. P. EerNisse, Design of Resonant Piezoelectric Devices, Research
Monograph No. 56, The M.I.T. Press, 1969.

2. T. Ikeda, Fundamentals of Piezoelectricity, Oxford University Press, 1990.

3. A.V. Mezheritsky, “Elastic, Dielectric, and Piezoelectric Losses in Piezoceramics:

How it Works all Together,” IEEE Transactions on Ultrasonics, Ferroelectrics, and
Frequency Control, vol. 51, no. 6, 2004.

4. K. Uchino and S. Hirose, “Loss Mechanisms in Piezoelectrics: How to Measure

Different Losses Separately,” IEEE Transactions on Ultrasonics, Ferroelectrics, and
Frequency Control, vol. 48, no. 1, pp. 307–321, 2001.

5. P.C.Y. Lee, N.H. Liu, and A. Ballato, “Thickness Vibrations of a Piezoelectric Plate
with Dissipation,” IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency
Control, vol. 51, no. 1, 2004.

6. P.C.Y. Lee and N.H. Liu, “Plane Harmonic Waves in an Infinite Piezoelectric Plate
with Dissipation,” Frequency Control Symposium and PDA Exhibition, IEEE
International, pp. 162–169, 2002.

7. C.A. Balanis, “Electrical Properties of Matter,” Advanced Engineering

Electromagnetics, John Wiley & Sons, 1989.

8. J. Yang, An Introduction to the Theory of Piezoelectricity, Springer Science and

Business Media, N.Y., 2005.

REFERENCES FOR MAGNETOSTRICTION

1. Chikazumi, Physics of Ferromagnetism, Oxford University Press, New York, 1997.

2. H. Kronmüller, “General Micromagnetic Theory”, in Handbook of Magnetism

and Advanced Magnetic Materials, edited by H. Kronmüller and S. Parkin, Vol. 2:
Micromagnetism, John Wiley & Sons, Chichester, 2007.

418 | CHAPTER 3: STRUCTURAL MECHANICS THEORY

3. X.E. Liu and X.J. Zheng, “A Nonlinear Constitutive Model for Magnetostricite
Materials”, Acta Mech Sinica, vol. 21, pp 278-285, 2005.

4. D.C. Jiles, Introduction to Magnetism and Magnetic Materials, 2nd ed.,

Chapman & Hall, London, 1998.

5. C. H. Sherman and J. L. Butler, Appendix A.7, p.555 in Transducers and Arrays

for Underwater Sound, Springer, New York, 2007.

REFERENCES FOR BOUNDARY CONDITIONS

1. M. Cohen and P.C. Jennings, “Silent Boundary Methods for Transient Analysis,”
Computational Methods for Transient Analysis, vol 1 (editors T. Belytschko and
T.J.R. Hughes), Nort-Holland, 1983.

2. B. Lalanne and M. Touratier, “Aeroelastic Vibrations and Stability in Cyclic

Symmetric Domains,” The International Journal of Rotating Machinery, vol. 6,
no. 6, pp 445–452, 2000.

3. Y.C. Fung, Foundations of Solid Mechanics, Prentice-Hall, 1965, page 99 ff

REFERENCES FOR CONTACT ANALYSIS AND DECOHESION

1. De Lorenzis L., Fernando D., Teng J.G. “Coupled mixed-mode cohesive zone
modeling of interfacial debonding in plated beams”, International Journal of Solids
and Structures, vol. 50, pp. 2477-2494. 2013

2. R.D.S.G. Campilho, M.F.S.F. de Moura, J.J.M.S. Domingues, “Using a cohesive

damage model to predict the tensile behaviour of CFRP single-strap repairs”,
International Journal of Solids and Structures, Volume 45, Issue 5, 1 March 2008,
Pages 1497–1512

3. S T Pinho, L Iannucci, P Robinson, “Formulation and implementation of

decohesion elements in an explicit finite code”, Composites Part A: Applied Science
and Manufacturing, Volume 37, Issue 5, May 2006, Pages 778–789.

REFERENCES | 419
420 | CHAPTER 3: STRUCTURAL MECHANICS THEORY
 4

Solid Mechanics

This chapter describes the Solid Mechanics interface, which is found under the
Structural Mechanics branch ( ) when adding a physics interface.

421
 T he S o li d M e c han i c s In t erfac e
The Solid Mechanics (solid) interface ( ), found under the Structural Mechanics
branch ( ) when adding a physics interface, is intended for general structural
analysis of 3D, 2D, or axisymmetric bodies. In 2D, plane stress or plane strain
assumptions can be used. The Solid Mechanics interface is based on solving the
equations of motion together with a constitutive model for a solid material. Results
such as displacements, stresses, and strains are computed.

The functionality provided by the Solid Mechanics interface depends on the products
you are using. The Acoustics Module, MEMS Module, and Structural Mechanics
Module add several features, for example geometric nonlinearity and advanced
boundary conditions such as contact, follower loads, and non-reflecting boundaries.

The default material is a Linear Elastic Material. With either the Nonlinear Structural
Materials Module or the Geomechanics Module, the physics interface is extended with
more materials, for example, material models for plasticity, hyperelasticity, creep, and
concrete. You can also add your own material models using an External Stress-Strain
Relation

For a detailed overview of the functionality available in each product, visit

http://www.comsol.com/products/specifications/

When this physics interface is added, these default nodes are also added to the Model
Builder — Linear Elastic Material, Free (a boundary condition where boundaries are free,
with no loads or constraints), and Initial Values. Then, from the Physics toolbar, you
can add other nodes that implement, for example, solid mechanics material models,
boundary conditions, and loads. You can also right-click Solid Mechanics to select
physics features from the context menu.

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

422 | CHAPTER 4: SOLID MECHANICS

The default Name (for the first physics interface in the model) is solid.

2D APPROXIMATION

From the 2D approximation list select Plane stress or Plane strain (the
default). Plane stress is relevant for structures which are thin in the
out-of-plane direction, such as a thin plate. Plane strain is relevant when
the 2D model can be considered as a cut through an object which is long
in the out-of-plane direction. For more information see the theory
section.

When modeling using plane stress, the Solid Mechanics interface solves
∂w
for the out-of-plane strain displacement derivative, ------- , in addition to the
∂Z
displacement field u.

When combining Solid Mechanics with other types of physics, there is

often an assumption that the out-of-plane extension is infinitely long. This
is the case in for example Acoustic-Structure interaction problems. In
these cases, Plane strain is usually the correct choice.

Select Out-of-plane mode extension (time-harmonic) to prescribe an

out-of-plane wave number to be used in mode analysis, eigenfrequency,
and frequency domain studies. When selected, enter the Out-of-plane wave
number kz. Note that the input value will only be taken into account in
eigenfrequency and frequency domain studies. For mode analysis, the
out-of-plane wave number is computed as an eigenvalue.

For more information, see Out-of-plane and Circumferential Modes in

the Structural Mechanics Theory chapter.

AXIAL SYMMETRY APPROXIMATION

Select Circumferential mode extension (time-harmonic) to prescribe a

circumferential wave number to be used in eigenfrequency or frequency
domain studies. When selected, enter the Circumferential mode number m.

For more information, see Out-of-plane and Circumferential Modes in

the Structural Mechanics Theory chapter.

THE SOLID MECHANICS INTERFACE | 423

 THICKNESS

For 2D components, enter a value or expression for the Thickness d. The

default value of 1 m is suitable for plane strain models, where it represents
a unit-depth slice, for example. For plane stress models, enter the actual
thickness, which should be small compared to the size of the plate for the
plane stress assumption to be valid.

When manually combining Solid Mechanics with other physics interfaces,

you must make sure that the same thickness assumption is used
everywhere. In most cases, the default settings will be correct since
interfaces which do not have an explicit thickness property will implicitly
assume unit thickness.

Use a Change Thickness node to change thickness in parts of the

geometry if necessary.

S T R U C T U R A L TR A N S I E N T B E H AV I O R
From the Structural transient behavior list, select Include inertial terms (the default) or
Quasi-static. Use Quasi-static to treat the dynamic behavior as quasi-static (with no mass
effects; that is, no second-order time derivatives). Selecting this option gives a more
efficient solution for problems where the variation in time is slow when compared to
the natural frequencies of the system. The default solver for the time stepping is
changed from Generalized alpha to BDF when Quasi-static is selected.

For problems with creep, and sometimes viscoelasticity, the problem can be considered
as quasi-static. This is also the case when the time dependence exists only in some other
physics, like a transient heat transfer problem causing thermal strains.

REFERENCE PO INT F OR MO MENT COMPUTATION

Enter the coordinates for the Reference point for moment computation xref (variable
refpnt). The resulting moments (applied or as reactions) are then computed relative
to this reference point. During the results and analysis stage, the coordinates can be
changed in the Parameters section in the result nodes.

TY P I C A L WA V E S P E E D
The typical wave speed cref is a parameter for the perfectly matched layers (PMLs) if
used in a solid wave propagation model. The default value is solid.cp, the
pressure-wave speed. To use another wave speed, enter a value or expression in the
Typical wave speed for perfectly matched layers field.

424 | CHAPTER 4: SOLID MECHANICS

Note: This section is only available with COMSOL products that include PMLs (see
http://www.comsol.com/products/specifications/)

DISCRETIZATION
To display this section, click the Show button ( ) and select Discretization.

In the Solid Mechanics interface you can choose not only the order of the
discretization, but also the type of shape functions: Lagrange or serendipity. For
highly distorted elements, Lagrange shape functions provide better accuracy than
serendipity shape functions of the same order. The serendipity shape functions will
however give significant reductions of the model size for a given mesh containing
hexahedral, prism, or quadrilateral elements.

The default is to use Quadratic serendipity shape functions for the Displacement field.
Using Linear shape functions will give what is sometimes called constant stress
elements. Such a formulation will for many problems make the model overly stiff, and
many elements may be needed for an accurate resolution of the stresses.

DEPENDENT VA RIA BLES

The physics interface uses the global spatial components of the Displacement field u as
dependent variables. The default names for the components are (u, v, w) in 3D. In 2D
the component names are (u, v), and in 2D axisymmetry they are (u, w). You can
however not use the “missing” component name in the 2D cases as a parameter or
variable name, since it is still used internally.

You can change both the field name and the individual component names. If a new
field name coincides with the name of another displacement field, the two fields (and
the interfaces which define them) share degrees of freedom and dependent variable
component names. You can use this behavior to connect a Solid Mechanics interface
to a Shell directly attached to the boundaries of the solid domain, or to another Solid
Mechanics interface sharing a common boundary.

A new field name must not coincide with the name of a field of another type (that is,
it must contain a displacement field), or with a component name belonging to some

THE SOLID MECHANICS INTERFACE | 425

 other field. Component names must be unique within a model except when two
interfaces share a common field name.

In the COMSOL Multiphysics Reference Manual see Table 2-3 for links
to common sections and Table 2-4 to common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

• Domain, Boundary, Edge, Point, and Pair Nodes for Solid Mechanics
• Solid Mechanics Theory
• Selecting Discretization

• Stresses in a Pulley: Application Library path COMSOL_Multiphysics/

Structural_Mechanics/stresses_in_pulley
• Eigenvalue Analysis of a Crankshaft: Application Library path
COMSOL_Multiphysics/Structural_Mechanics/crankshaft

Domain, Boundary, Edge, Point, and Pair Nodes for Solid Mechanics
The Solid Mechanics Interface has these domain, boundary, edge, point, and pair
nodes and subnodes (listed in alphabetical order), which are available from the Physics
ribbon toolbar (Windows users), Physics context menu (Mac or Linux users), or
right-click to access the context menu (all users).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

FEATURES AVAILABLE FROM SUBMENUS

Many features for the Solid Mechanics interface are added from submenus in the
Physics toolbar groups or context menu (when you right-click the node). The
submenu name is the same in both cases.

The submenus at the Domain level are Material Models, Volume Forces, Mass, Spring, and
Damper, and Domain Constraints.

426 | CHAPTER 4: SOLID MECHANICS

The submenus at the Boundary level are Connections, Pairs, Mass, Spring, and Damper,
and More Constraints.

There are also the Edges and Points submenus.

Note: Some submenus are only present with certain COMSOL products.

FORCE LOADS
Note that you can add force loads acting on all levels of the geometry for the physics
interface. Add a:

• Body Load to domains (to model gravity effects, for example).

• Boundary Load to boundaries (a pressure acting on a boundary, for example).
• Edge Load to edges in 3D (a force distributed along an edge, for example).
• Point Load to points (concentrated forces at points).

If there are subsequent constraints specified on the same geometrical

entity, the last one takes precedence.

For 2D axisymmetric components, COMSOL Multiphysics takes the axial

symmetry boundaries (at r = 0) into account and automatically adds an
Axial Symmetry node to the component that is valid on the axial symmetry
boundaries only.

THE SOLID MECHANICS INTERFACE | 427

 • Added Mass • Magnetostrictive Material
• Antisymmetry • Nonlinear Elastic Material
• Attachment • Periodic Condition
• Beam Connection • Piezoelectric Material
• Body Load • Point Load
• Bolt Pre-Tension • Point Load (on Axis)
• Bolt Selection • Prescribed Acceleration
• Boundary Load • Prescribed Displacement
• Cam-Clay Material • Prescribed Displacement/Rotation
• Change Thickness • Prescribed Velocity
• Contact • Rigid Motion Suppression
• Edge Load • Rigid Connector
• External Strain • Rigid Domain
• External Stress • Ring Load
• External Stress-Strain Relation • Roller
• Fixed Constraint • Rotating Frame
• Free • Shell Connection
• Gravity • Spring Foundation
• Hyperelastic Material • Stress Linearization
• Initial Values • Symmetry
• Linear Elastic Material • Thin Elastic Layer
• Low-Reflecting Boundary • Thin-Film Damping

In the COMSOL Multiphysics Reference Manual, see Table 2-3 for links
to common sections and Table 2-4 for common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

428 | CHAPTER 4: SOLID MECHANICS

These subnodes are available from main parent nodes as indicated in the
documentation:

• Adhesion • Mass and Moment of Inertia (Rigid

• Applied Force (Rigid Connector) Domain)

• Applied Force (Rigid Domain) • Mechanical Damping

• Applied Moment (Rigid • Phase

Connector) • Plasticity
• Applied Moment (Rigid Domain) • Porous Plasticity
• Center of Mass Nodes (Boundary, • Prescribed Displacement/Rotation
Edge, Point) • Predeformation
• Center of Rotation Nodes • Rocks
(Boundary, Edge, Point)
• Safety
• Conduction Loss
• Soil Plasticity
(Time-Harmonic)
• Spring Foundation (Rigid
• Concrete
Connector)
• Creep
• Spring Foundation (Rigid Domain)
• Damping
• Thermal Expansion (for constraints)
• Dielectric Loss
• Thermal Expansion (for materials)
• External Stress
• Thermal Expansion (Attachment)
• Fixed Constraint (Rigid Domain)
• Thermal Expansion (Rigid
• Friction Connector)
• Hygroscopic Swelling • Viscoelasticity
• Initial Stress and Strain • Viscoplasticity
• Initial Values (Rigid Domain)
• Location Nodes (Boundary, Edge,
Point)
• Mass and Moment of Inertia (Rigid
Connector)

THE SOLID MECHANICS INTERFACE | 429

 Initial Values
The Initial Values node adds initial values for the displacement field and structural
velocity field that can serve as an initial condition for a transient simulation or as an
initial guess for a nonlinear analysis. In addition to the default Initial Values node always
present in the interface, you can add more Initial Values nodes if needed.

INITIAL VALUES
Enter values or expressions for the initial values of the Displacement field u (the
displacement components u, v, and w in 3D), and the Structural velocity field ∂u/∂t.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Initial Values
Membrane>Initial Values
Truss>Initial Values

Ribbon
Physics tab with Solid Mechanics selected:
Domains>Solid Mechanics>Initial Values

Physics tab with Membrane selected:

Boundaries>Membrane>Initial Values

Physics tab with Truss selected:

Edges>Truss>Initial Values

Change Thickness
Use the Change Thickness node to model domains with a thickness other than the
overall thickness defined in the physics interface’s Thickness section. The Change
Thickness node is available in:

• Domains in 2D for the Solid Mechanics and Multibody Dynamics interfaces.

• Boundaries in 3D for the Membrane and Shell interfaces. (See also Change
Thickness in shell interface documentation).
• Boundaries in 2D axisymmetry for the Membrane interface.

430 | CHAPTER 4: SOLID MECHANICS

CHANGE THICKNESS
Enter a value for the Thickness d. This value replaces the overall thickness for the
selected domains or boundaries.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Change Thickness
Multibody Dynamics>Change Thickness
Membrane>Change Thickness

Ribbon
Physics tab with Solid Mechanics or Multibody Dynamics selected:
Domains>Solid Mechanics>Change Thickness
Domains>Multibody Dynamics>Change Thickness

Physics tab with Membrane selected:

Boundaries>Membrane>Change Thickness

Linear Elastic Material

The Linear Elastic Material node adds the equations for a linear elastic solid and an
interface for defining the elastic material properties.

By adding the following subnodes to the Linear Elastic Material node you can
incorporate many other effects:

• Thermal Expansion (for materials) • Creep

• Hygroscopic Swelling • Viscoplasticity
• Initial Stress and Strain • Porous Plasticity
• External Stress • Soil Plasticity
• External Strain • Concrete
• Damping • Rocks
• Viscoelasticity • Safety
• Plasticity

Note: Some options are only available with certain COMSOL products (see http://
www.comsol.com/products/specifications/)

THE SOLID MECHANICS INTERFACE | 431

 COORDINATE SYSTEM SELECTION
The Global coordinate system is selected by default. The Coordinate system list contains
any additional coordinate systems that the model includes (except boundary
coordinate systems). The coordinate system is used for interpreting directions of
orthotropic and anisotropic material data and when stresses or strains are presented in
a local system. The coordinate system must have orthonormal coordinate axes, and be
defined in the material frame. Many of the possible subnodes inherit the coordinate
system settings.

LINEAR ELASTIC MATERIAL

To use a mixed formulation by adding the pressure as an extra dependent variable to
solve for, select the Nearly incompressible material check box. For a material with a very
low compressibility, using only displacements as degrees of freedom may lead to a
numerically ill-posed problem.

Define the Solid model and the linear elastic material properties.

Solid Model
Select a linear elastic Solid model: Isotropic (the default), Orthotropic, or Anisotropic.
Select:

• Isotropic for a linear elastic material that has the same properties in all directions.
• Orthotropic for a linear elastic material that has different material properties in
orthogonal directions, so that its stiffness depends on the properties Ei, νij, and Gij.
• Anisotropic for a linear elastic material that has different material properties in
different directions, and the stiffness comes from the symmetric elasticity matrix, D.

Note: The Orthotropic and Anisotropic options are only available with certain
COMSOL products (see http://www.comsol.com/products/specifications/)

• Material Models
• Linear Elastic Material
• Orthotropic and Anisotropic Materials

432 | CHAPTER 4: SOLID MECHANICS

Density
The default Density ρ uses values From material. For User defined enter another value or
expression.

The density is needed for dynamic analysis or when the elastic data is given
in terms of wave speed. It is also used when computing mass forces for
gravitational or rotating frame loads, and when computing mass
properties (Computing Mass Properties).

Specification of Elastic Properties for Isotropic Materials

For an Isotropic Solid model, from the Specify list select a pair of elastic properties for
an isotropic material — Young’s modulus and Poisson’s ratio, Young’s modulus and shear
modulus, Bulk modulus and shear modulus, Lamé parameters, or Pressure-wave and
shear-wave speeds. For each pair of properties, select from the applicable list to use the
value From material or enter a User defined value or expression.

Each of these pairs define the elastic properties and it is possible to convert from one
set of properties to another according to Table 4-1.
TABLE 4-1: EXPRESSIONS FOR THE ELASTIC MODULI.

DESCRIPTION VARIABLE D(E,ν) D(E,G) D(K,G) D(λ,μ)

Young’s E= E E 9KG - 3λ + 2μ
------------------ μ --------------------
modulus 3K + G λ+μ
Poisson’s ν= ν E- λ -
---  1 – ------------------
------- –1 1 3G - --------------------
ratio 2G 2 3K + G 2(λ + μ)

Bulk K= E - EG - K 2μ
----------------------- -------------------------- λ + -------
modulus 3 ( 1 – 2ν ) 3 ( 3G – E ) 3
Shear G= E - G G μ
--------------------
modulus 2(1 + ν)
Lamé λ= Eν G ( E – 2G )- 2G λ
--------------------------------------
- --------------------------- K – --------
parameter ( 1 + ν ) ( 1 – 2ν ) 3G – E 3
λ
Lamé μ= E - G G μ
--------------------
parameter 2(1 + ν)
μ

THE SOLID MECHANICS INTERFACE | 433

 TABLE 4-1: EXPRESSIONS FOR THE ELASTIC MODULI.

DESCRIPTION VARIABLE D(E,ν) D(E,G) D(K,G) D(λ,μ)

Pressure- cp = E(1 – ν) - G ( 4G – E )- K + 4G ⁄ 3- λ + 2μ
wave speed ----------------------------------------- --------------------------- ------------------------- ----------------
ρ ( 1 + ν ) ( 1 – 2ν ) ρ ( 3G – E ) ρ ρ
Shear-wave cs = E - G⁄ρ G⁄ρ μ⁄ρ
speed ------------------------
2ρ ( 1 + ν )
The individual property parameters are:

• Young’s modulus (elastic modulus) E.

• Poisson’s ratio ν.
• Shear modulus G.
• Bulk modulus K.
• Lamé parameter λ and Lamé parameter μ.
• Pressure-wave speed (longitudinal wave speed) cp.
• Shear-wave speed (transverse wave speed) cs. This is the wave speed for a solid
continuum. In plane stress, for example, the actual speed with which a longitudinal
wave travels is lower than the value given.

Specification of Elastic Properties for Orthotropic Materials

When Orthotropic is selected from the Solid model list, the material properties vary in
orthogonal directions only. The Material data ordering can be specified in either
Standard or Voigt notation. When User defined is selected in 3D, enter three values in
the fields for Young’s modulus E, Poisson’s ratio ν, and the Shear modulus G. This
defines the relationship between engineering shear strain and shear stress. It is
applicable only to an orthotropic material and follows the equation

τ ij
ε ij = --------
G ij

νij is defined differently depending on the application field. It is easy to

transform among definitions, but check which one the material uses.

Specification of Elastic Properties for Anisotropic Materials

When Anisotropic is selected from the Solid model list, the material properties vary in
all directions, and the stiffness comes from the symmetric Elasticity matrix, D The
Material data ordering can be specified in either Standard or Voigt notation. When User
defined is selected, a 6-by-6 symmetric matrix is displayed.

434 | CHAPTER 4: SOLID MECHANICS

GEOMETRIC NONLINEARITY
The settings in this section affect the behavior of the selected domains in a
geometrically nonlinear analysis.

If a study step is geometrically nonlinear, the default behavior is to use a large strain
formulation in all domains. Select the Force linear strains check box to always use a
small strain formulation, irrespective of the setting in the study step.

When a geometrically nonlinear formulation is used, the elastic deformations used for
computing the stresses can be obtained in two different ways if inelastic deformations
are present: additive decomposition and multiplicative decomposition. The default is
to use multiplicative decomposition. Select Additive strain decomposition to change to
an assumption of additivity.

THE SOLID MECHANICS INTERFACE | 435

 Note: This section is only available with COMSOL products that support
geometrically nonlinear analysis (see http://www.comsol.com/products/
specifications/).

• There are some cases when a small strain formulation could be useful
for a certain domain, even though the study step is geometrically
nonlinear. One such case is in contact analysis, where the study is always
geometrically nonlinear, but it is possible that a geometrically linear
formulation is sufficient in the material.
• When a multiplicative decomposition is used, the order of the
subnodes to Linear Elastic Material matters. The inelastic deformation
are assumed to have occurred in the same order as the subnodes appear
in the model tree.
• In versions prior to 5.3, only the additive strain decomposition method
was available. If you want to revert to the previous behavior, select
Additive strain decomposition. If the results then differ significantly,
probably the assumption of additivity is questionable, however.
• In models created in a version prior to 4.2a, a check box named Include
geometric nonlinearity may be visible in this section. It is displayed only
if geometric nonlinearity was originally used for the selected domains.
Once the check box is cleared in this Settings window, it is permanently
removed and the study step assumes control over the selection of
geometric nonlinearity.
When Include geometric nonlinearity is selected in this section, it
automatically also selects the Include geometric nonlinearity check box
in the study Settings window.

• Modeling Geometric Nonlinearity

• Inelastic Strain Contributions
• Studies and Solvers in the COMSOL Multiphysics Reference Manual

ENERGY DISSIPATION
You can select to compute and store various energy dissipation variables in a time
dependent analysis. Doing so will add extra degrees of freedom to the model.

To display this section, click the Show button ( ) and select Advanced Physics Options.

436 | CHAPTER 4: SOLID MECHANICS

Select the Calculate dissipated energy check box as needed to compute the energy
dissipated by for example creep, plasticity, viscoplasticity, viscoelasticity, or damping.

• Dissipated Energy
• Energy Variables

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Material Models>Linear Elastic Material
Multibody Dynamics>Linear Elastic Material

Ribbon
Physics tab with Solid Mechanics selected:
Domains>Material Models>Linear Elastic Material

Physics tab with Multibody Dynamics selected:

Domains>Multibody Dynamics>Linear Elastic Material

Nonlinear Elastic Material

The Nonlinear Elastic Material feature is used to model stress-strain relationships which
are nonlinear even at infinitesimal strains. It is available in the Solid Mechanics and
Membrane interfaces. This material model requires either the Nonlinear Structural
Materials Module or the Geomechanics Module.

By adding the following subnodes to the Nonlinear Elastic Material node you can
incorporate many other effects:

• Thermal Expansion (for materials) • Creep

• Hygroscopic Swelling • Viscoplasticity
• Initial Stress and Strain • Porous Plasticity
• External Stress • Soil Plasticity
• External Strain • Concrete
• Damping • Rocks
• Viscoelasticity • Safety
• Plasticity

THE SOLID MECHANICS INTERFACE | 437

 Note: Some options are only available with certain COMSOL products (see http://
www.comsol.com/products/specifications/)

COORDINATE SYSTEM SELECTION

The Global coordinate system is selected by default. The Coordinate system list contains
any additional coordinate systems that the model includes (except boundary
coordinate systems). The coordinate system is used when stresses or strains are
presented in a local system. The coordinate system must have orthonormal coordinate
axes, and be defined in the material frame. Many of the possible subnodes inherit the
coordinate system settings.

NONLINEAR ELASTIC MATERIAL

To use a mixed formulation by adding the pressure as an extra dependent variable to
solve for, select the Nearly incompressible material check box. For a material with a very
low compressibility, using only displacements as degrees of freedom may lead to a
numerically ill-posed problem.

The available material models depend on the COMSOL products you are using.

Nonlinear Structural Materials Module: Select a Material model: Ramberg-Osgood,

Power law, Uniaxial data, Bilinear elastic, or User defined.

Geomechanics Module: Select a Material model: Ramberg-Osgood, Hyperbolic law,

Hardin-Drnevich, Duncan-Chang, Duncan-Selig, or User defined.

Density
All nonlinear elastic material models have density as an input. The default Density ρ
uses values From material. For User defined enter another value or expression.

The density is needed for dynamic analysis or when the elastic data is given
in terms of wave speed. It is also used when computing mass forces for
gravitational or rotating frame loads, and when computing mass
properties (Computing Mass Properties).

Ramberg-Osgood, Power law, Hyperbolic law, Hardin-Drnevich, Duncan-Chang, or

Duncan-Selig
Select from the applicable list to use the value From material or enter a User defined
value or expression.

From the Specify list select a pair of elastic properties for an isotropic material —
Young’s modulus and Poisson’s ratio (the default for Ramberg-Osgood, Power law,

438 | CHAPTER 4: SOLID MECHANICS

Duncan-Chang, and Duncan-Selig) or Bulk modulus and shear modulus (the default for
Hyperbolic law and Hardin-Drnevich).

Then depending on the selections, define the applicable parameters:

• Young’s modulus E.
• Poisson’s ratio ν.
• Shear modulus G.
• Bulk modulus K.
• For Ramberg-Osgood:
- Reference stress σref .
- Reference strain εref.
- Stress exponent n.
• For Power law and Hyperbolic law:
- Reference shear strain γref.
- Strain exponent n.
• For Hardin-Drnevich, define the Reference shear strain γref.
• For Duncan-Chang, define the Ultimate deviatoric stress qult.
• For Duncan-Selig:
- Ultimate deviatoric stress qult.
- Ultimate strain εult.

Uniaxial Data
For Uniaxial data enter a value or expression for the Uniaxial stress function σax as a
function of the uniaxial strain. The default expression is the linear function
2
(210[GPa])*solid.eax N/m , which corresponds to a linear elastic material with
Young’s modulus 210 GPa.

From the Specify list select how to specify the second elastic property for the material
— Bulk modulus or Poisson’s ratio. Then depending on the selection, enter a value or
select from the applicable list to use the value From material or enter a User defined
value or expression:

• Bulk modulus K.
• Poisson’s ratio ν.

When you select Bulk modulus, the Young’s modulus is computed from the tensile part
of the Uniaxial stress function σax. When you select Poisson’s ratio, you can either use

THE SOLID MECHANICS INTERFACE | 439

 the tensile part (default), or use the full tensile-compressive function by selecting the
check box Use nonsymmetric stress-strain data.

Bilinear Elastic
For Bilinear elastic enter a value or select from the applicable list to use the value From
material or enter a User defined value or expression.

• Bulk modulus in tension Kt.

• Bulk modulus in compression Kc.
• Shear modulus G.

User Defined
In the User defined material model, you specify the bulk modulus implicitly by entering
the relation between pressure and volumetric elastic strain. Enter a value or select from
the applicable list to use the value From material (the default) or enter a User defined
value or expression.

• Pressure p. The default expression is (-160[GPa])*solid.eelvol N/m2, which

corresponds to a linear elastic material with bulk modulus 160 GPa.
• Shear modulus G.

GEOMETRIC NONLINEARITY
The settings in this section affect the behavior of the selected domains in a
geometrically nonlinear analysis.

If a study step is geometrically nonlinear, the default behavior is to use a large strain
formulation in all domains. Select the Force linear strains check box to always use a
small strain formulation, irrespective of the setting in the study step.

When a geometrically nonlinear formulation is used, the elastic deformations used for
computing the stresses can be obtained in two different ways if inelastic deformations
are present: additive decomposition and multiplicative decomposition. The default is
to use multiplicative decomposition. Select Additive strain decomposition to change to
an assumption of additivity.

440 | CHAPTER 4: SOLID MECHANICS

 • There are some cases when a small strain formulation could be useful
for a certain domain, even though the study step is geometrically
nonlinear. One such case is in contact analysis, where the study is always
geometrically nonlinear, but it is possible that a geometrically linear
formulation is sufficient in the material.
• When a multiplicative decomposition is used, the order of the
subnodes to Nonlinear Elastic Material matters. The inelastic
deformation are assumed to have occurred in the same order as the
subnodes appear in the model tree.
• In versions prior to 5.3, only the additive strain decomposition method
was available. If you want to revert to the previous behavior, select
Additive strain decomposition. If the results then differ significantly,
probably the assumption of additivity is questionable, however.

• Modeling Geometric Nonlinearity

• Inelastic Strain Contributions
• Studies and Solvers in the COMSOL Multiphysics Reference Manual

ENERGY DISSIPATION
To display this section, click the Show button ( ) and select Advanced Physics Options.

Select the Calculate dissipated energy check box as needed to compute the energy
dissipated by Creep, Plasticity, Viscoplasticity, or Viscoelasticity.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Material Models>Nonlinear Elastic Material
Membrane>Material Models>Nonlinear Elastic Material

Ribbon
Physics tab with Solid Mechanics selected:
Domains>Material Models>Nonlinear Elastic Material

Physics tab with Membrane selected:

Boundaries>Material Models>Nonlinear Elastic Material

THE SOLID MECHANICS INTERFACE | 441

 Hyperelastic Material
The Hyperelastic Material subnode adds the equations for hyperelasticity at large
strains. Hyperelastic materials can be suitable for modeling rubber and other polymers,
biological tissue, and also for applications in acoustoelasticity. The Hyperelastic Material
is available in the Solid Mechanics and Membrane interfaces. This material model
requires the Nonlinear Structural Materials Module.

When a hyperelastic material is included in your model, all studies are geometrically
nonlinear. The Include geometric nonlinearity check box in the study settings is selected
and cannot be cleared.

See also Hyperelastic Material Models in the Structural Mechanics

Theory chapter.

HYPERELASTIC MATERIAL
Select a hyperelastic Material model from the list and then go to the applicable section
for more information.

Density
All hyperelastic material models have density as an input. The default Density ρ uses
values From material. For User defined enter another value or expression.

The density is needed for dynamic analysis or when the elastic data is given
in terms of wave speed. It is also used when computing mass forces for
gravitational or rotating frame loads, and when computing mass
properties (Computing Mass Properties).

• Neo-Hookean • Varga
• St Venant-Kirchhoff • Arruda-Boyce
• Mooney-Rivlin, Two Parameters • Gent
• Mooney-Rivlin, Five Parameters • Blatz-Ko
• Mooney-Rivlin, Nine Parameters • Gao
• Yeoh • Murnaghan
• Ogden • User defined
• Storakers

442 | CHAPTER 4: SOLID MECHANICS

Neo-Hookean
For Neo-Hookean you can use a mixed formulation by adding the negative mean
pressure as an extra dependent variable. For this case, select the Nearly incompressible
material check box. Then if the check box is selected, enter a value for the Initial bulk
modulus κ.

The default values for both Lamé parameter λ and Lamé parameter μ use values From
material.

St Venant-Kirchhoff
For St Venant-Kirchhoff you can use a mixed formulation by adding the negative mean
pressure as an extra dependent variable. For this case, select the Nearly incompressible
material check box.

The default values for both Lamé parameter λ and Lamé parameter μ use values From
material.

Mooney-Rivlin, Two Parameters

For Mooney-Rivlin, two-parameters the Model parameters C10 and C01 both use values
From material. Enter the Initial bulk modulus κ.

Mooney-Rivlin, Five Parameters

For Mooney-Rivlin, five-parameters the Model parameters C10, C01, C20, C02, and C11
all use values From material. Enter the Initial bulk modulus κ.

Mooney-Rivlin, Nine Parameters

For Mooney-Rivlin, nine-parameters the Model parameters C10, C01, C20, C02, C11, C30,
C03, C21, and C12 all use values From material. Enter the Initial bulk modulus κ.

Yeoh
For Yeoh the Model parameters c1, c2, and c3 all use values From material. Enter the
Initial bulk modulus κ.

Ogden
For Ogden you can use a mixed formulation by adding the negative mean pressure as
an extra dependent variable. For this case, select the Nearly incompressible material
check box.

In the table for the Ogden parameters, enter values or expressions in each column: p,
Shear modulus (Pa), and Alpha parameter.

If the Nearly incompressible material check box is selected, also enter the Initial bulk
modulus κ.

THE SOLID MECHANICS INTERFACE | 443

 Storakers
For Storakers, in the table for the Storakers parameters, enter values or expressions in
each column: p, Shear modulus (Pa), Alpha parameter, and Beta parameter.

Varga
For Varga the Model parameters c1, c2, and c3 all use values From material. Enter the
Initial bulk modulus κ.

Arruda-Boyce
For Arruda-Boyce you can use a mixed formulation by adding the negative mean
pressure as an extra dependent variable. For this case, select the Nearly incompressible
material check box.

The default values for the Macroscopic shear modulus μ0 and the Number of segments N
use values From material.

If the Nearly incompressible material check box is selected, also enter the Initial bulk
modulus κ.

Gent
For Gent the default values for the Macroscopic shear modulus μ and the model
parameter jm is to use values From material. Enter the Initial bulk modulus κ.

Blatz-Ko
For Blatz-Ko the Shear modulus μ and the Model parameters β and φ all use values From
material.

Gao
For Gao the Model parameters a and n all use values From material.

Murnaghan
For Murnaghan the Murnaghan third-order elastic moduli constants l, m, and n and the
Lamé parameters λ and μ use values From material.

User defined
For User defined you can use a mixed formulation by adding the negative mean pressure
as an extra dependent variable. In this case, select the Nearly incompressible material
check box.

If the Nearly incompressible material check box is selected, enter the Isochoric strain
energy density Wsiso and the Volumetric strain energy density Wsvol.

444 | CHAPTER 4: SOLID MECHANICS

If the Nearly incompressible material check box is not selected, enter an expression for
the Strain energy density Ws.

For examples of:

• Mooney-Rivlin, two-parameters and Ogden, see Inflation of a Spherical

Rubber Balloon. Application Library path:
Nonlinear_Structural_Materials_Module/Hyperelasticity/balloon_inflation.
• Murnaghan, see Elasto-Acoustic Effect in Rail Steel. Application
Library path: Nonlinear_Structural_Materials_Module/Hyperelasticity/
rail_steel.

ENERGY DISSIPATION
To display this section, click the Show button ( ) and select Advanced Physics Options.

Select the Calculate dissipated energy check box to compute the energy dissipated by
Plasticity.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Material Models>Hyperelastic Material
Membrane>Material Models>Hyperelastic Material

Ribbon
Physics tab with Solid Mechanics selected:
Domains>Material Models>Hyperelastic Material

Physics tab with Membrane selected:

Boundaries>Material Models>Hyperelastic Material

Piezoelectric Material
The Piezoelectric Material node defines the piezoelectric material properties either in
stress-charge form using the elasticity matrix and the coupling matrix, or in
strain-charge form using the compliance matrix and the coupling matrix. It is normally
used together with a Piezoelectric Effect multiphysics coupling node and a
corresponding Charge Conservation, Piezoelectric node in the Electrostatics interface.
This node is added by default to the Solid Mechanics interface when adding a
Piezoelectric Devices interface.

THE SOLID MECHANICS INTERFACE | 445

 This material model requires the Structural Mechanics Module, MEMS Module, or
Acoustics Module.

By adding the following subnodes to the Piezoelectric Material node you can
incorporate many other effects:

• Initial Stress and Strain

• Thermal Expansion (for materials)
• Mechanical Damping
• Coupling Loss
• Dielectric Loss
• Conduction Loss (Time-Harmonic)

When the Piezoelectric Material node is added to the structural mechanics

interface in the absence of an active Piezoelectric Effect multiphysics
coupling node the material behaves similarly to a Linear Elastic Material
node, with elastic properties as corresponding to the elasticity or
compliance matrix entered (see below). The piezoelectric effect is then
not included in the corresponding equation system.

See also Piezoelectric Material in the Structural Mechanics Theory

chapter.

PIEZOELECTRIC MATERIAL PROPERTIES

Select a Constitutive relation — Stress-charge form or Strain-charge form. For each of the
following, the default uses values From material. For User defined enter other values in
the matrix or field as needed.

• For Stress-charge form, select an Elasticity matrix (ordering: xx, yy, zz, yz, xz, xy) (cE).
• For a Strain-charge form, select a Compliance matrix (ordering: xx, yy, zz, yz, xz, xy)
(sE).
• Select a Coupling matrix (ordering: xx, yy, zz, yz, xz, xy) (d).
• Select a Relative permittivity (erS or erT).

446 | CHAPTER 4: SOLID MECHANICS

• Enter values for the Remanent electric displacement Dr.
• Select a Density (p).

For entering these matrices, use the following order (Voigt notation),
which is the common convention for piezoelectric materials: xx, yy, zz, yz,
xz, zy.

The density is needed for dynamic analysis or when the elastic data is given
in terms of wave speed. It is also used when computing mass forces for
gravitational or rotating frame loads, and when computing mass
properties (Computing Mass Properties).

GEOMETRIC NONLINEARITY
If a study step is geometrically nonlinear, the default behavior is to use a large strain
formulation in all domains. There are, however, some cases when the use of a small
strain formulation for a certain domain is needed. In those cases, select the Force linear
strains check box. When selected, a small strain formulation is always used,
independently of the setting in the study step. The check box is not selected by default
to conserve the properties of the model.

• Modeling Piezoelectric Problems

• Modeling Geometric Nonlinearity
• The Electromagnetics Interfaces in the COMSOL Multiphysics
Reference Manual

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Material Models>Piezoelectric Material

Ribbon
Physics tab with Solid Mechanics selected:
Domains>Material Models>Piezoelectric Material

Magnetostrictive Material
The Magnetostrictive Material node defines the magnetostrictive material properties.

THE SOLID MECHANICS INTERFACE | 447

 In case of linear magnetostriction model, the material data can be entered either in the
strain-magnetization form using the elasticity matrix and the coupling matrix, or in
stress-magnetization form using the compliance matrix and the coupling matrix. It is
normally used as part of Magnetostriction multiphysics interface together with a
Magnetostriction multiphysics coupling node and Ampère’s Law, Magnetostrictive node
in the corresponding Magnetic Fields interface. Magnetostrictive Material node is added
by default to the Solid Mechanics interface when adding a Magnetostriction
multiphysics interface. The interface requires the AC/DC Module and at least one of
the following modules: Structural Mechanics, MEMS, Acoustics.

When the Magnetostrictive Material node is added to the Structural

Mechanics interface in the absence of an active Magnetostriction
multiphysics coupling node, the material behaves similarly to a Linear
Elastic Material node with some limitations on the format for the elastic
material data input. All the magnetic material data and coupling data will
have no effect. The magnetostrictive effect is then not included in the
corresponding equation system.

By adding the following subnodes to the Magnetostrictive Material node you can
incorporate other effects:

• Initial Stress and Strain

• Thermal Expansion (for materials)
• Mechanical Damping

See also Magnetostrictive Material in the Structural Mechanics Theory

chapter.

LINEAR MAGNETOSTRICTIVE MATERIAL PROPERTIES

Select a Constitutive relation — Strain-magnetization form or Stress-magnetization form.
For each of the following, the default uses values From material. For User defined, enter
other values in the matrix or field as needed.

• For Strain-magnetization form, select a Compliance matrix (ordering: xx, yy, zz, yz, xz,
xy) (sH).
• For a Stress-magnetization form, select an Elasticity matrix (ordering: xx, yy, zz, yz, xz,
xy) (cH).

448 | CHAPTER 4: SOLID MECHANICS

• Select a Piezomagnetic coupling matrix (ordering: xx, yy, zz, yz, xz, xy) (dHT or eHS).
• Select a Relative permeability (μrT or μrS).
• Select a Density (p).

For entering these matrices, use the following order (Voigt notation),
which is the common convention for magnetostrictive materials: xx, yy,
zz, yz, xz, zy.

The density is needed for dynamic analysis. It is also used when computing
mass forces for gravitational or rotating frame loads, and when computing
mass properties (Computing Mass Properties).

NONLINEAR MAGNETOSTRICTIVE MATERIAL PROPER TIES

See the corresponding the theory section Nonlinear Magnetostriction.

GEOMETRIC NONLINEARITY
If a study step is geometrically nonlinear, the default behavior is to use a large strain
formulation in all domains. There are, however, some cases when the use of a small
strain formulation for a certain domain is needed. In those cases, select the Force linear
strains check box. When selected, a small strain formulation is always used,
independently of the setting in the study step. The check box is not selected by default
to conserve the properties of the model.

• Magnetostriction
• Modeling Magnetostrictive Materials
• Modeling Geometric Nonlinearity
• Ampère’s Law, Magnetostrictive
• The Magnetic Fields Interface in the COMSOL Multiphysics
Reference Manual

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Material Models>Magnetostrictive Material

THE SOLID MECHANICS INTERFACE | 449

 Ribbon
Physics tab with Solid Mechanics selected:
Domains>Material Models>Magnetostrictive Material

Cam-Clay Material
The Cam-Clay Material node adds the equations and interface for defining the material
properties for the modified Cam-clay material. This material model requires the
Geomechanics Module.

The Thermal Expansion (for materials) and Initial Stress and Strain subnodes can be
added to the Cam-Clay Material.

See also Cam-Clay Material Model in the Structural Mechanics Theory

chapter.

MODEL INPUTS
The Pore fluid pressure pfluid is user-defined by default. The default value is 1 atm, but
you can change it to another value or expression for the pore fluid pressure. If there
are other physics interfaces (like Darcy’s law) in the model that make a pressure
variable available, such variables will be available in the list.

CAM-CLAY MATERIAL
To use a mixed formulation by adding the negative mean pressure as an extra
dependent variable to solve for, select the Nearly incompressible material check box.

From the Specify list, define the elastic properties either in terms of Poisson’s ratio or
Shear modulus.

The defaults for the Poisson’s ratio ν or the Shear modulus G, Density ρ, Cam-Clay M
parameter M, Swelling index κ, Compression index λ, and Void ratio at reference pressure
N are taken From material. For User defined enter other values or expressions.

For the Cam-Clay M parameter you can alternatively select Match to

Mohr-Coulomb criterion which then matches the slope of the virgin
consolidation line to the Angle of internal friction. Then select the Angle
of internal friction φ as From material (the default) or User defined.

450 | CHAPTER 4: SOLID MECHANICS

 The density is needed for dynamic analysis or when the elastic data is given
in terms of wave speed. It is also used when computing mass forces for
gravitational or rotating frame loads, and when computing mass
properties (Computing Mass Properties).

Enter a value or expression for the Reference pressure for the parameter N prefN , the
Initial void ratio e0,and the Initial consolidation pressure pc0.

GEOMETRIC NONLINEARITY
If a study step is geometrically nonlinear, the default behavior is to use the
Green-Lagrange strain in all domains. There are, however, some cases when the use of
a small strain formulation for a certain domain is needed. In those cases, select the Force
linear strains check box. When selected, a small strain formulation is always used,
independently of the setting in the study step. The check box is not selected by default
to conserve the properties of the model.

Isotropic Compression with Cam-Clay Material Model: Application

Library path Geomechanics_Module/Verification_Examples/
isotropic_compression

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Material Models>Cam-Clay Material

Ribbon
Physics tab with Solid Mechanics selected:
Domains>Material Models>Cam-Clay Material

Viscoelasticity
Use the Viscoelasticity subnode to add viscous stress contributions to an elastic material
model, This material model is available in the Solid Mechanics and Membrane
interfaces, and can be used together with Linear Elastic Material, Nonlinear Elastic
Material, and Hyperelastic Material.

See also Linear Viscoelastic Materials and Large Strain Viscoelasticity in

the Structural Mechanics Theory chapter.

THE SOLID MECHANICS INTERFACE | 451

 THERMAL EFFECTS
Viscoelastic properties have a strong dependence on the temperature. For
thermorheologically simple materials, a change in the temperature can be transformed
directly into a change in the time scale. Thus, the relaxation time is modified to
aT(T)τm, where aT(T) is a shift function.

Select a Shift function — None (the default), Williams-Landel-Ferry, Arrhenius, or User

defined.

• When the default, None, is kept, the shift function aT(T) is set to unity and the
relaxation time is not modified.
• For Williams-Landel-Ferry enter values or expressions for these properties then the
shift function aT(T) is computed from these parameters and the relaxations time is
shifted according to it:
- Reference temperature TWLF The default is 293.15 K.
- WLF constant 1 C1WLF. The default is 17.44.
- WLF constant 2 C2WLF. The default is 51.6 K.

• For Arrhenius enter values or expressions for these properties then the shift function
aT(T) is computed from these parameters and the relaxations time is shifted
according to it:
- Reference temperature T0. The default is 293.15 K.
- Activation energy Q.

• For User defined enter a value or expression for the shift function aT.

VISCOELASTICITY MODEL
Select a Material model — Generalized Maxwell (the default), Standard linear solid, or
Kelvin-Voigt. Then see the settings for each option that follows.

For any material model, you can select the shear modulus to use when solving a
stationary problem. Choose the Static stiffness for the material model — Long-term
(the default) or Instantaneous.

Generalized Maxwell
For Generalized Maxwell in the table enter the values for the parameters that describe
the viscoelastic behavior as a series of spring-dashpot pairs.

For linear viscoelasticity, in each Branch row enter the stiffness of the spring Gm in the
Shear modulus (Pa) column and the relaxation time constant τm in the Relaxation time
(s) column for the spring-dashpot pair in branch m.

452 | CHAPTER 4: SOLID MECHANICS

For large strain viscoelasticity, in each Branch row enter βm (the energy factor of the
branch) in the Energy factor (1) column and the relaxation time constant τm in the
Relaxation time (s) column for the spring-dashpot pair.

• Use the Add button ( ) to add a row to the table and the Delete button ( ) to
delete a row in the table.
• Using the Load from file button ( ) and the Save to file button ( ) load and
store data for the branches in a text file with three space-separated columns (from
left to right): the branch number, the shear modulus or energy factor, and the
relaxation time for that branch.

Standard Linear Solid

For Standard linear solid enter the values for the parameters that describe the
viscoelastic behavior of the single spring-dashpot branch.

For linear viscoelasticity, select an option from the Relaxation data list and edit the
default as needed:

• Relaxation time τv. The default is 3000 s.

• Viscosity ηv of the dash-pot. The default is 6 x 1013 Pa⋅s.

In the Shear modulus field, enter the stiffness of the spring Gv. The default is
2 x 1010 Pa.

For large strain viscoelasticity, enter the Relaxation time τv, which default is 3000 s, and
the Energy factor βv of the dash-pot. The default is 0.2.

Kelvin-Voigt
For Kelvin-Voigt enter the values for the parameter that describes the viscous behavior
of the single dash-pot.

For linear viscoelasticity, select an option from the Relaxation data list and edit the
default as needed:

• Relaxation time τv. The default is 3000 s.

• Viscosity ηv of the dash-pot. The default is 6 x 1013 Pa⋅s.

For large strain viscoelasticity, enter the Relaxation time τv. The default is 3000 s.

DISCRETIZATION
Select a Shape function type — Discontinuous Lagrange (default) or Gauss point data for
the components of the auxiliary viscoelastic tensor. When the discontinuous Lagrange
discretization is used, the shape function order is selected as one order less than what

THE SOLID MECHANICS INTERFACE | 453

 is used for the displacements. This results in that fewer extra degrees of freedom are
added to the model than when using Gauss point data. The accuracy does in general
not differ much. If you want to enforce that the constitutive law is fulfilled at the
integration points, select Gauss point data.

To compute the energy dissipation caused by viscoelasticity, enable the

Calculate dissipated energy check box in the Energy Dissipation section of
the parent material node.

• Viscoelastic Structural Damper: Application Library path

Structural_Mechanics_Module/Dynamics_and_Vibration/
viscoelastic_damper_frequency
• Viscoelastic Structural Damper—Transient Analysis: Application
Library path Structural_Mechanics_Module/Dynamics_and_Vibration/
viscoelastic_damper_transient

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Linear Elastic Material>Viscoelasticity
Solid Mechanics>Nonlinear Elastic Material>Viscoelasticity
Solid Mechanics>Hyperlastic Material>Viscoelasticity
Membrane>Linear Elastic Material>Viscoelasticity
Membrane>Nonlinear Elastic Material>Viscoelasticity
Membrane>Hyperlastic Material>Viscoelasticity

Ribbon
Physics tab with Linear Elastic Material or Nonlinear Elastic Material node selected in the
model tree:
Attributes>Viscoelasticity

Plasticity
Using the Plasticity subnode you define the properties for modeling elastoplastic
materials. This material model is available in the Solid Mechanics, Membrane and
Truss interfaces, and can be used together with Linear Elastic Material, Nonlinear
Elastic Material, and Hyperelastic Material.

454 | CHAPTER 4: SOLID MECHANICS

The Nonlinear Structural Material Module or the Geomechanics Module are required
for this material model, and the available options depend on the products used. For
details, see http://www.comsol.com/products/specifications/.

See also Elastoplastic Material Models in the Structural Mechanics

Theory chapter.

PLASTICITY MODEL
Use this section to define the plastic properties of the material.

Plasticity Model
Select Small plastic strains or Large plastic strains to apply either an additive or
multiplicative decomposition between elastic and plastic strains.

• When using plasticity together with a hyperelastic material, only the option Large
plastic strains is available.
• When using plasticity in the Membrane and Truss interfaces, only the option Small
plastic strains is available.

Yield Function F
The Yield function F defines the limit of the elastic regime F(σ, σys) ≤ 0.

Select a Yield function F criterion — von Mises stress, Tresca stress, Hill orthotropic
plasticity, or User defined.

• The default is von Mises stress with associate plastic potential.

• Select Tresca stress to use a Tresca yield criterion. The plastic potential can be an
Associated or non associated flow rule with the von Mises stress as plastic potential.
• Select Hill orthotropic plasticity to use Hill’s criterion. For Hill orthotropic plasticity
from the Specify list select either the Initial tensile and shear yield stresses σys0ij or
Hill’s coefficients F, G, H, L, M, and N. The default for either selection uses values
From material (if it exists) or User defined. The principal directions of orthotropy are
inherited from the coordinate system selection in the Linear elastic feature.
• For User defined enter a different value or expression. Write any expression in terms
of the stress tensor variables or its invariants in the φ(σ) field.
- For User defined also select the Plastic potential Q related to the flow rule —
Associated (the default), von Mises, or User defined (non associated). Enter a User
defined value in the Q field as needed.

THE SOLID MECHANICS INTERFACE | 455

 Initial Yield Stress
For all yield criteria, the default Initial yield stress σys0 uses values From material and
represents the stress level where plastic deformation starts.

Isotropic Hardening Model

For all yield criteria, select the type of linear or nonlinear isotropic hardening model
from the Isotropic hardening model list.

• Select Perfectly plastic (ideal plasticity) if the material can undergo plastic
deformation without any increase in yield stress.
• For Linear isotropic hardening the default Isotropic tangent modulus ETiso uses values
From material (if it exists) or User defined. The yield level σys is modified as
hardening occurs, and it is related to the effective plastic strain εpe as
1 1 1
σ ys = σ ys0 + E iso ε pe with ----------- = -------------- – ----
E iso E Tiso E

For the linear isotropic hardening model, the yield stress increases proportionally to
the effective plastic strain εpe. The Young’s modulus E is taken from the elastic
material properties.
• Select Ludwik from the list to model nonlinear isotropic hardening. The yield level
σys is modified by the power-law
n
σ ys ( ε pe ) = σ ys0 + kε pe

the Strength coefficient k and the Hardening exponent n use values From material (if
it exists) or User defined.
• For Swift nonlinear isotropic hardening, the Reference strain ε0 and the Hardening
exponent n use values From material (if it exists) or User defined. The yield level σys
is modified by the power-law
ε pe n
σ ys ( ε pe ) = σ ys0  1 + --------
 ε0 

• Select Voce from the list to model nonlinear isotropic hardening. The yield level σys
is modified by the exponential law

– βε pe
σ ys ( ε pe ) = σ ys0 + σ sat  1 – e 
 

the Saturation flow stress σsat and the Saturation exponent β use values From material
(if it exists) or User defined.

456 | CHAPTER 4: SOLID MECHANICS

• For Hockett-Sherby nonlinear isotropic hardening, the Steady-state flow stress σ∝,
the Saturation coefficient m, and the Saturation exponent n use values From material
(if it exists) or User defined. The yield level σys is increased by the exponential law
n
– mε pe
σ ys ( ε pe ) = σ ys0 + ( σ ∞ – σ ys0 )  1 – e 


• For User defined isotropic hardening the Hardening function σh(εpe) uses values From
material. The yield level σys is modified as

σ ys = σ ys0 + σ h ( ε pe )

This definition implies that the hardening function σh(εpe) must be zero at zero
plastic strain. In other words, σys = σys0 when εpe = 0. With this option it is
possible to fit nonlinear isotropic hardening curves. The hardening function can
depend on more variables than the effective plastic strain, for example the
temperature.

Kinematic Hardening Model

For all yield function criteria, select the type of kinematic hardening model (not
available for hyperelastic materials) from the Kinematic hardening model list.

• Select No kinematic hardening (when either ideal plasticity or an isotropic hardening

model is selected as isotropic hardening model) if it is a material that can undergo
plastic deformation without a shift in the yield surface.
• If Linear kinematic hardening is selected as the Kinematic hardening model, the default
Kinematic tangent modulus Ek uses values From material. This parameter is used to
calculate the back stress σb as plasticity occurs:
2 1 1 1
σ b = C k ⋅ --- ε p with ------- = ------- – ----
3 Ck Ek E

This is Prager’s linear kinematic hardening model, so the back stress σb is

collinear to the plastic strain tensor εp.

• If Armstrong-Frederick is selected from the list, the default Kinematic hardening

modulus Ck and Kinematic hardening parameter γk use values From material. These
parameters are used to calculate the back stress σb from the rate equation

· 2 · ·
σ b = --- C k ε p – γ k ε pe σ b
3

This is Armstrong-Frederik nonlinear kinematic hardening model.

THE SOLID MECHANICS INTERFACE | 457

 • When Chaboche is selected from the Kinematic hardening model list, the default
Kinematic hardening modulus C0 uses values From material. Add branches as needed
to solve N rate equations for the back stresses:

· 2 · ·
σ b,i = --- C i ε p – γ i ε pe σ b,i
3

The total back stress σb is then computed from the sum

N
2
σ b = --- C 0 ε p +
3  σb,i
i=1

For each Branch row, enter Ci (the hardening modulus of the branch i) in the
Hardening modulus (Pa) column and γi (the hardening parameter of the branch i) in
the Hardening parameter (1) column.
Use the Add button ( ) and the Delete button ( ) to add or delete a row in the
table. Use the Load from file button ( ) and the Save to file button ( ) to load
and store data for the branches in a text file with three space-separated columns
(from left to right): the branch number, the hardening modulus for that branch, and
the hardening parameter for that branch.

To compute the energy dissipation caused by plasticity, enable the

Calculate dissipated energy check box in the Energy Dissipation section of
the parent material node.

• Sheet Metal Forming: Application Library path

Nonlinear_Structural_Materials_Module/Plasticity/sheet_metal_forming
• For an example of Large plastic strains, see Necking of an Elastoplastic
Metal Bar: Application Library path
Nonlinear_Structural_Materials_Module/Plasticity/bar_necking.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Linear Elastic Material>Plasticity
Solid Mechanics>Nonlinear Elastic Material>Plasticity
Solid Mechanics>Hyperelastic Material>Plasticity
Membrane>Linear Elastic Material>Plasticity

458 | CHAPTER 4: SOLID MECHANICS

Membrane>Nonlinear Elastic Material>Plasticity
Truss>Linear Elastic Material>Plasticity

Ribbon
Physics tab with Linear Elastic Material, Nonlinear Elastic Material, or Hyperlastic Material
node selected in the model tree:
Attributes>Plasticity

Creep
Use the Creep subnode to define the creep properties of the material model. This
material model is available in the Solid Mechanics and Membrane interfaces, and can
be used together with Linear Elastic Material and Nonlinear Elastic Material.

The Nonlinear Structural Material Module or the Geomechanics Module are required
for this material model, and the available options depend on the products used. For
details, see http://www.comsol.com/products/specifications/.

See also Creep and Viscoplasticity in the Structural Mechanics Theory

chapter.

CREEP DATA

Nonlinear Structural Materials Module

Select a Material model — Norton, Norton-Bailey, Garofalo (hyperbolic sine),
Navarro-Herring, Coble, Weertman, Potential, Volumetric, Deviatoric, or User defined.
Then follow the instructions as below.

Geomechanics Module
Select a Material model — Potential, Volumetric, Deviatoric, or User defined. Then follow
the instructions as below.

Norton
For Norton enter the following settings:

• Creep rate coefficient A.

• Reference creep stress σref. The default is 1 MPa.
• Stress exponent n.
• Select the Include temperature dependency check box to add an “Arrhenius-type”
temperature dependence. Then enter a Creep activation energy Q.

THE SOLID MECHANICS INTERFACE | 459

 Norton-Bailey
For Norton-Bailey enter the following settings:

• Creep rate coefficient A.

• Reference creep stress σref. The default is 1 MPa.
• Stress exponent n.
• Select a Hardening model — Time hardening or Strain hardening.
- For Time hardening, enter the hardening exponent m, the Time shift tshift and the
Reference time tref.
- For Strain hardening, enter the hardening exponent m, the Effective creep strain
shift εshift, and the Reference time tref.

• Select the Include temperature dependency check box to add an “Arrhenius-type”

temperature dependence. Then enter a Creep activation energy Q.

Garofalo (hyperbolic sine)

For Garofalo (hyperbolic sine) enter the following settings:

• Creep rate coefficient A.

• Reference creep stress σref. The default is 1 MPa.
• Garofalo n parameter n.
• Select the Include temperature dependency check box as needed. Then enter a Creep
activation energy Q.

Navarro-Herring
For Navarro-Herring enter the following settings:

• Volume diffusivity Dv.

• Burgers vector b.
• Grain diameter d.

Coble
For Coble enter the following settings:

• Ionic diffusivity Dgb.

• Burgers vector b.
• Grain diameter d.

460 | CHAPTER 4: SOLID MECHANICS

Weertman
For Weertman enter the following settings:

• Diffusivity D0.
• Burgers vector b.
• Stress exponent n.
• Reference creep stress σref. The default is 1 MPa.

Potential
For Potential enter the following settings:

• Rate multiplier η.
• Creep potential Qcr.

Deviatoric
For Deviatoric enter the Creep rate Fcr.

Volumetric
For Volumetric enter the Creep rate Fcr.

User defined
For User defined enter each element for the symmetric Creep rate tensor Fcr. The tensor
components are interpreted in the coordinate system of the parent node.

To compute the energy dissipation caused by creep, enable the Calculate

dissipated energy check box in the Energy Dissipation section of the parent
material node (Linear Elastic Material or Nonlinear Elastic Material).

For an example of Norton and Norton-Baily material models, see

Combining Creep Material Models: Application Library path
Nonlinear_Structural_Materials_Module/Creep/combined_creep.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Linear Elastic Material>Creep
Solid Mechanics>Nonlinear Elastic Material>Creep
Membrane>Linear Elastic Material>Creep
Membrane>Nonlinear Elastic Material>Creep

THE SOLID MECHANICS INTERFACE | 461

 Ribbon
Physics tab with Linear Elastic Material or Nonlinear Elastic Material node selected in the
model tree:
Attributes>Creep

Viscoplasticity
Use the Viscoplasticity subnode to define the viscoplastic properties of the material
model. This material model is available in the Solid Mechanics and Membrane
interfaces, and can be used together with Linear Elastic Material and Nonlinear Elastic
Material.

The Nonlinear Structural Material Module is required for this material model.

See also Creep and Viscoplasticity in the Structural Mechanics Theory

chapter.

VISCOPLASTICITY MODEL
Select a Viscoplasticity model — Anand, Chaboche, or Perzyna. Then follow the
instructions as below.

Anand
For Anand enter the following data:

• Creep rate coefficient A.

• Creep activation energy Q.
• Multiplier of stress ξ.
• Stress sensitivity m.
• Deformation resistance saturation coefficient s0.
• Deformation resistance initial value sinit.
• Hardening constant h0.
• Hardening sensitivity a.
• Deformation resistance sensitivity n.

462 | CHAPTER 4: SOLID MECHANICS

Each of the material properties can either be defined obtained From material, or as User
defined. In the latter case, enter a value or an expression.

Viscoplastic Creep in Solder Joints: Application Library path

Nonlinear_Structural_Materials_Module/Viscoplasticity/
viscoplastic_solder_joints

Chaboche
For Chaboche enter the following settings:

• Viscoplastic rate coefficient A.

• Reference creep stress σref. The default is 1 MPa.
• Stress exponent n.

Yield Function F
The Yield function F defines the limit of the elastic regime F(σ, σys) ≤ 0.

Select a Yield function F criterion — von Mises stress, Tresca stress, Hill orthotropic
plasticity, or User defined.

• The default is von Mises stress with associate plastic potential.

• Select Tresca stress to use a Tresca yield criterion. The plastic potential can be an
Associated or non associated flow rule with the von Mises stress as plastic potential.
• Select Hill orthotropic plasticity to use Hill’s criterion. For Hill orthotropic plasticity
from the Specify list select either the Initial tensile and shear yield stresses σys0ij or
Hill’s coefficients F, G, H, L, M, and N. The default for either selection uses values
From material (if it exists) or User defined. The principal directions of orthotropy are
inherited from the coordinate system selection in the parent feature.
• For User defined enter a different value or expression. In the φ(σ) field write any
expression in terms of the stress tensor components or its invariants.
- For User defined also select the Plastic potential Q related to the flow rule —
Associated (the default), von Mises, or User defined (non associated). Enter a User
defined value in the Q field as needed.

Initial Yield Stress

The default Initial yield stress σys0 uses values From material and represents the stress
level where viscoplastic deformation starts.

THE SOLID MECHANICS INTERFACE | 463

 Isotropic Hardening Model
Select the type of linear or nonlinear isotropic hardening model from the Isotropic
hardening model list.

• Select Perfectly plastic (ideal viscoplasticity) if the material can undergo viscoplastic
deformation without any increase in yield stress.
• For Linear isotropic hardening the default Isotropic tangent modulus ETiso uses values
From material (if it exists) or User defined. The yield level σys is modified as
hardening occurs, and it is related to the effective viscoplastic strain εvpe as
1 1 1
σ ys = σ ys0 + E iso ε vpe with ----------- = -------------- – ----
E iso E Tiso E

For the linear isotropic hardening model, the yield stress increases proportionally to
the effective viscoplastic strain εvpe. The Young’s modulus E is taken from the elastic
material properties.
• Select Ludwik from the list to model nonlinear isotropic hardening. The yield level
σys is modified by the power-law
n
σ ys = σ ys0 + kε vpe

the Strength coefficient k and the Hardening exponent n use values From material (if
it exists) or User defined.
• For Swift nonlinear isotropic hardening, the Reference strain ε0 and the Hardening
exponent n use values From material (if it exists) or User defined. The yield level σys
is modified by the power-law
ε vpe n
σ ys = σ ys0  1 + ----------
 ε0 

• Select Voce from the list to model nonlinear isotropic hardening. The yield level σys
is modified by the exponential law

– βε vpe
σ ys = σ ys0 + σ sat  1 – e 
 

the Saturation flow stress σsat and the Saturation exponent β use values From material
(if it exists) or User defined.
• For Hockett-Sherby nonlinear isotropic hardening, the Steady-state flow stress σ∝,
the Saturation coefficient m, and the Saturation exponent n use values From material
(if it exists) or User defined. The yield level σys is increased by the exponential law

464 | CHAPTER 4: SOLID MECHANICS

 n
– mε vpe
σ ys = σ ys0 + ( σ ∞ – σ ys0 )  1 – e 


• For User defined isotropic hardening the Hardening function σh(εvpe) uses values
From material. The yield level σys is modified as

σ ys = σ ys0 + σ h ( ε vpe )

This definition implies that the hardening function σh(εvpe) must be zero at zero
viscoplastic strain. In other words, σys = σys0 when εvpe = 0. With this option it
is possible to fit nonlinear hardening curves. The hardening function can depend
on more variables than the effective viscoplastic strain, for example the
temperature.

Kinematic Hardening Model

Select the type of kinematic hardening model from the Kinematic hardening model list.

• Select No kinematic hardening when the material can undergo viscoplastic

deformation without a shift in the yield surface.
• If Linear kinematic hardening is selected as the Kinematic hardening model, the default
Kinematic tangent modulus Ek uses values From material. This parameter is used to
calculate the back stress σb as:
2 1 1 1
σ b = C k ⋅ --- ε vp with ------- = ------- – ----
3 Ck Ek E

This is Prager’s linear kinematic hardening model, so the back stress σb is

collinear to the viscoplastic strain tensor εvp.

• If Armstrong-Frederick is selected from the list, the default Kinematic hardening

modulus Ck and Kinematic hardening parameter γk use values From material. These
parameters are used to calculate the back stress σb from the rate equation

· 2 · ·
σ b = --- C k ε vp – γ k ε vpe σ b
3

This is Armstrong-Frederik nonlinear kinematic hardening model.

• When Chaboche is selected from the Kinematic hardening model list, the default
Kinematic hardening modulus C0 uses values From material. Add branches as needed
to solve N rate equations for the back stresses:

THE SOLID MECHANICS INTERFACE | 465

 · 2 · ·
σ b,i = --- C i ε vp – γ i ε vpe σ b,i
3

For each Branch row, enter Ci (the hardening modulus of the branch i) in the
Hardening modulus (Pa) column and γi (the hardening parameter of the branch i) in
the Hardening parameter (1) column.
Use the Add button ( ) and the Delete button ( ) to add or delete a row in the
table. Use the Load from file button ( ) and the Save to file button ( ) to load
and store data for the branches in a text file with three space-separated columns
(from left to right): the branch number, the hardening modulus for that branch, and
the hardening parameter for that branch.
The total back stress σb is then computed from the sum
N
2
σ b = --- C 0 ε vp +
3  σb,i
i=1

Perzyna
For Perzyna enter the following settings:

• Viscoplastic rate coefficient A.

• Reference creep stress σref. The default is 1 MPa.

The other settings are the same as for Chaboche.

To compute the energy dissipation caused by viscoplasticity, enable the

Calculate dissipated energy check box in the Energy Dissipation section of
the parent material node (Linear Elastic Material or Nonlinear Elastic
Material).

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Linear Elastic Material>Viscoplasticity
Solid Mechanics>Nonlinear Elastic Material>Viscoplasticity
Membrane>Linear Elastic Material>Viscoplasticity
Membrane>Nonlinear Elastic Material>Viscoplasticity

466 | CHAPTER 4: SOLID MECHANICS

Ribbon
Physics tab with Linear Elastic Material or Nonlinear Elastic Material node selected in the
model tree:
Attributes>Viscoplasticity

Porous Plasticity
Use the Porous Plasticity subnode to define the properties of a plasticity model for a
porous material.

PO RO US PLAST ICITY MODEL

Use this section to define the plastic properties of the porous material.

Yield Function F
The Yield function F defines the limit of the elastic regime F(σ, σys) ≤ 0.

Select a Yield function F for the porous plasticity criterion — Shima-Oyane, Gurson,
Gurson-Tvergaard-Needleman, Fleck-Kuhn-McMeeking, or FKM-GTN.

Shima-Oyane
For Shima-Oyane enter the following data:

• Initial yield stress σys0.

• Shima-Oyane parameter α.
• Shima-Oyane parameter γ.
• Shima-Oyane parameter m.
• Initial void volume fraction f0.

Gurson
For Gurson enter the following data:

• Initial yield stress σys0.

• Initial void volume fraction f0.

Gurson-Tvergaard-Needleman
For Gurson-Tvergaard-Needleman enter the following data:

• Initial yield stress σys0.

• Tvergaard correction coefficient q1.
• Tvergaard correction coefficient q2.
• Initial void volume fraction f0.

THE SOLID MECHANICS INTERFACE | 467

 • Critical void volume fraction fc.
• Failure void volume fraction ff.

Fleck-Kuhn-McMeeking
For Fleck-Kuhn-McMeeking enter the following data:

• Initial yield stress σys0.

• Initial void volume fraction f0.
• Maximum void volume fraction fmax.

FKM-GTN
For FKM-GTN enter the following data:

• Initial yield stress σys0.

• Tvergaard correction coefficient q1.
• Tvergaard correction coefficient q2.
• Initial void volume fraction f0.
• Maximum void volume fraction fmax.
• Void volume fraction cut-off for GTN model fGTN.
• Void volume fraction cut-off for FKM model fFKM.

Each of the material properties can either be defined obtained From material, or as User
defined. In the latter case, enter a value or an expression.

See also Porous Plasticity in the Structural Mechanics Theory chapter.

To compute the energy dissipation caused by porous compaction, enable

the Calculate dissipated energy check box in the Energy Dissipation section
of the parent material node (Linear Elastic Material or Nonlinear Elastic
Material).

468 | CHAPTER 4: SOLID MECHANICS

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Linear Elastic Material>Porous Plasticity
Solid Mechanics>Nonlinear Elastic Material>Porous Plasticity
Membrane>Linear Elastic Material>Porous Plasticity
Membrane>Nonlinear Elastic Material>Porous Plasticity

Ribbon
Physics tab with Linear Elastic Material or Nonlinear Elastic Material node selected in the
model tree:
Attributes>Porous Plasticity

Soil Plasticity
In the Soil Plasticity subnode you define the properties for modeling materials
exhibiting soil plasticity. This material model can be used together with Linear Elastic
Material and Nonlinear Elastic Material. It is available with the Geomechanics Module.

The yield criteria are described in the theory section:

• Drucker-Prager Criterion
• Mohr-Coulomb Criterion
• Matsuoka-Nakai Criterion
• Lade-Duncan Criterion

SOIL PLASTICITY
Select a Yield criterion — Drucker-Prager, Mohr-Coulomb, Matsuoka-Nakai, or
Lade-Duncan. Most values are taken From material. For User defined choices, enter other
values or expressions.

Drucker-Prager
In the standard Drucker-Prager formulation, the material parameters are given in
terms of the α and k coefficients. Often material data is expressed in the parameters c
and φ used in the Mohr-Coulomb model. You can the choose to use there parameters
instead. If so, select the Match to Mohr-Coulomb criterion check box (see
Mohr-Coulomb Criterion). If this check box is selected, the default values for Cohesion
c and the Angle of internal friction φ are taken From material.

If required, select the Use dilatation angle in plastic potential check box. If this check
box is selected, then enter a value or expression for the Dilatation angle ψ. Alternatively,

THE SOLID MECHANICS INTERFACE | 469

 select From material. The dilatation angle replaces the angle of internal friction when
defining the plastic potential.

If the Match to Mohr-Coulomb criterion check box is not selected, then the default
Drucker-Prager alpha coefficient α and Drucker-Prager k coefficient are taken From
material.

If required, select the Include elliptic cap check box. Select from the list the hardening
model. When Perfectly plastic (no hardening) is selected, enter values or expressions to
define the semi-axes of the ellipse under Elliptic cap parameter pa and Elliptic cap
parameter pb. When Isotropic hardening is selected from the list, the default Isotropic
hardening modulus Kiso, the Maximum plastic volumetric strain εpvol,max, and the Ellipse
aspect ratio R are taken From material (see Elliptic Cap With Hardening). Enter a value
or expression to define the initial semi-axis of the ellipse under the Initial location of the
cap pb0.

Mohr-Coulomb
The default Angle of internal friction φ and Cohesion c are taken From material.

If required, select the Use dilatation angle in plastic potential check box. If this check
box is selected, then enter a value or expression for the Dilatation angle ψ. Alternatively,
select From material. The dilatation angle replaces the angle of internal friction when
defining the plastic potential.

Under Plastic potential select either Drucker-Prager matched at compressive meridian,

Drucker-Prager matched at tensile meridian, or Associated.

If required, select the Include elliptic cap check box. Select from the list the hardening
model. When Perfectly plastic (no hardening) is selected, enter values or expressions to
define the semi-axes of the ellipse under Elliptic cap parameter pa and Elliptic cap
parameter pb. When Isotropic hardening is selected from the list, the default Isotropic
hardening modulus Kiso, the Maximum plastic volumetric strain εpvol,max, and the Ellipse
aspect ratio R are taken From material (see Elliptic Cap With Hardening). Enter a value
or expression to define the initial semi-axis of the ellipse under the Initial location of the
cap pb0.

Matsuoka-Nakai
If required, select the Match to Mohr-Coulomb criterion check box. If this check box is
selected, the default Angle of internal friction φ is taken From material.

If the Match to Mohr-Coulomb criterion check box is not selected, then the default
Matsuoka-Nakai mu coefficient μ is taken From material.

470 | CHAPTER 4: SOLID MECHANICS

Lade-Duncan
If required, select the Match to Mohr-Coulomb criterion check box. If this check box is
selected, then enter a value or expression for the Angle of internal friction φ.
Alternatively, select From material.

If the Match to Mohr-Coulomb criterion check box is not selected, then the default
Lade-Duncan k coefficient k is taken From material.

TE N S I O N C U T - O F F
If required, select the Include tension cut-off check box (see Tension Cut-Off). Enter a
value or expression for the Max tensile stress σt. Use this to constrain the soil plasticity
model with an extra yield surface, which limits the maximum principal stress.

• Deep Excavation: Application Library path Geomechanics_Module/Soil/

deep_excavation
• Flexible and Smooth Strip Footing on Stratum of Clay: Application
Library path Geomechanics_Module/Soil/flexible_footing

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Linear Elastic Material>Soil plasticity
Solid Mechanics>Nonlinear Elastic Material>Soil plasticity

Ribbon
Physics tab with Linear Elastic Material or Nonlinear Elastic Material node selected in the
model tree:
Attributes>Soil plasticity

Concrete
In the Concrete subnode you define the properties for modeling materials with failure
criteria representative of concrete. This material model can be used together with
Linear Elastic Material and Nonlinear Elastic Material. It is available with the
Geomechanics Module.

The failure criteria are described in the theory section:

• Bresler-Pister Criterion

THE SOLID MECHANICS INTERFACE | 471

 • Willam-Warnke Criterion
• Ottosen Criterion

CONCRETE MODEL
Select a Concrete criterion — Bresler-Pister, Willam-Warnke, or Ottosen. The defaults are
taken From material. For User defined choices, enter other values or expressions.

Bresler-Pister
The defaults for the Uniaxial tensile strength σt, Uniaxial compressive strength σc, and
Biaxial compressive strength σb are taken From material.

Willam-Warnke
The defaults for the Uniaxial tensile strength σt, Uniaxial compressive strength σc, and
Biaxial compressive strength σb are taken From material.

Ottosen
The defaults for the Uniaxial tensile strength σc, Ottosen’s parameters a and b , Size
factor k1, and Shape factor k2 are taken From material.

TE N S I O N C U T - O F F
If required, select the Include tension cut-off check box (see Tension Cut-Off). Enter a
value or expression for the Max tensile stress σt. Use this to constraint the concrete
model with an extra yield surface, which limits the maximum principal stress.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Linear Elastic Material>Concrete
Solid Mechanics>Nonlinear Elastic Material>Concrete

Ribbon
Physics tab with Linear Elastic Material or Nonlinear Elastic Material node selected in the
model tree:
Attributes>Concrete

Rocks
In the Rocks subnode you define the properties for modeling materials with failure
criteria representative of rocks. This material model can be used together with Linear
Elastic Material and Nonlinear Elastic Material. It is available with the Geomechanics
Module.

472 | CHAPTER 4: SOLID MECHANICS

The failure criterion are described in the theory section:

• Original Hoek-Brown Criterion

• Generalized Hoek-Brown Criterion

ROCK MODEL
Select a Rock criterion — Original Hoek-Brown or Generalized Hoek-Brown. The defaults
are taken From material. For User defined choices, enter other values or expressions.

Original Hoek-Brown
The defaults for the Uniaxial compressive strength σc, Hoek-Brown m parameter m, and
Hoek-Brown s parameter s are taken From material.

Generalized Hoek-Brown
The defaults for the Uniaxial compressive strength σc, Geological strength index GSI,
Disturbance factor D, and Intact rock parameter mi are taken From material.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Linear Elastic Material>Rocks
Solid Mechanics>Nonlinear Elastic Material>Rocks

Ribbon
Physics tab with Linear Elastic Material or Nonlinear Elastic Material node selected in the
model tree:
Attributes>Rocks

Thermal Expansion (for materials)

Use the Thermal Expansion subnode to add an internal thermal strain caused by changes
in temperature.

The Thermal Expansion subnode is only available with some COMSOL products (see
http://www.comsol.com/products/specifications/).

MODEL INPUTS
From the Temperature T list, select an existing temperature variable from a heat transfer
interface (for example, Temperature (ht)), if any temperature variables exist, or select
User defined to enter a value or expression for the temperature.

THE SOLID MECHANICS INTERFACE | 473

 THERMAL EXPANSION PROPERTIES
Select an Input type to select how the thermal strain is specified. The default is Secant
coefficient of thermal expansion, in which case the thermal strain is given by

ε th = α ( T – T ref )

where α is the secant coefficient of thermal expansion. α can be temperature

dependent.

When Input type is Tangent coefficient of thermal expansion, the thermal strain is given
by

 T 
ε th = exp 
 T ref
α t ( τ ) dτ – 1


where αt is the tangential coefficient of thermal expansion.

When Input type is Thermal strain, enter the thermal strain dL as function of
temperature explicitly.

In all three cases, the default is to take values From material. When entering data as
User defined, select Isotropic, Diagonal or Symmetric to enter one or more components
for a general coefficient of the thermal expansion tensor or the thermal strain tensor.
When a non-isotropic input is used, the axis orientations are given by the coordinate
system selection in the parent node.

Enter a value or expression for the Strain reference temperature Tref which is the
reference temperature that defines the change in temperature together with the actual
temperature.

A heat source term will be created by this node. It can be accessed from
a Thermoelastic Damping node in a heat transfer interface in order to
incorporate the reversed effect that heat is produced by changes in stress.
The heat source term is only present when Structural Transient Behavior
is set to Include inertial terms.

• Thermoelastic Damping
• Entropy and Thermoelasticity

474 | CHAPTER 4: SOLID MECHANICS

 Thermal Stresses in a Layered Plate: Application Library path
Structural_Mechanics_Module/Thermal-Structure_Interaction/layered_plate
MEMS_Module/Actuators/layered_plate

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Linear Elastic Material>Thermal Expansion
Solid Mechanics>Nonlinear Elastic Material>Thermal Expansion
Solid Mechanics>Hyperelastic Material>Thermal Expansion
Solid Mechanics>Piezoelectric Material>Thermal Expansion
Solid Mechanics>Cam-Clay Material>Thermal Expansion
Membrane>Linear Elastic Material>Thermal Expansion
Membrane>Nonlinear Elastic Material>Thermal Expansion
Membrane>Hyperelastic Material>Thermal Expansion
Multibody Dynamics>Linear Elastic Material>Thermal Expansion

Ribbon
Physics tab with Linear Elastic Material, Hyperelastic Material, Nonlinear Elastic Material,
Piezoelectric Material, or Cam-Clay Material node selected in the model tree:
Attributes>Thermal Expansion

Hygroscopic Swelling
Hygroscopic swelling is an internal strain caused by changes in moisture content. This
strain can be written as

ε hs = β h ( c mo – c mo,ref )

where βh is the coefficient of hygroscopic swelling, cmo is the moisture concentration,

and cmo,ref is the strain-free reference concentration.

The Hygroscopic Swelling subnode is only available with some COMSOL products (see
http://www.comsol.com/products/specifications/).

MODEL INPUTS
From the Concentration c list, select an existing concentration variable from another
physics interface, if any concentration variables exist. For User defined enter a value or
expression for the concentration. The unit for the input depends on the setting of

THE SOLID MECHANICS INTERFACE | 475

 Concentration type in the Hygroscopic Swelling Properties section. Only concentration
variables having the chosen physical dimension are available in the Concentration list.

HYGROSCOPIC SWELLING PROPERTIES

In the Concentration type list, select Molar concentration or Mass concentration,
depending on the units used for the concentration.

Enter a Strain reference concentration cref . This is the concentration at which there are
no strains due to hygroscopic swelling.

If Molar concentration is selected as the Concentration type, enter also the Molar mass of
the fluid, Mm. The default value is 0.018 kg/mol, which is the molar mass of water.

The default Coefficient of hygroscopic swelling βh uses values From material. For
User defined, select Isotropic, Diagonal, or Symmetric from the list to enter one or more
components for a general coefficient of hygroscopic swelling tensor βh. The default
value for the User defined case is 1.5e-4 m3/kg. When a non-isotropic coefficient of
hygroscopic swelling is used, the axis orientations are given by the coordinate system
selection in the parent node.

The Include moisture as added mass check box is selected by default. When selected,
the mass of the fluid is included in a dynamic analysis, and when using mass
proportional loads. It will also contribute when computing mass properties.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Linear Elastic Material>Hygroscopic Swelling
Solid Mechanics>Nonlinear Elastic Material>Hygroscopic Swelling
Solid Mechanics>Hyperelastic Material>Hygroscopic Swelling
Membrane>Linear Elastic Material>Hygroscopic Swelling
Membrane>Nonlinear Elastic Material>Hygroscopic Swelling
Membrane>Hyperelastic Material>Hygroscopic Swelling

Ribbon
Physics tab with Linear Elastic Material, Nonlinear Elastic Material, or Hyperelastic
Material node selected in the model tree:
Attributes>Hygroscopic Swelling

476 | CHAPTER 4: SOLID MECHANICS

Initial Stress and Strain
You can add the Initial Stress and Strain subnode to several material models, in order
to specify the stress or strain state in the structure before applying any constraint or
load. The values given are not initial values in the mathematical sense, but rather a
contribution to the constitutive relation.

The Initial Stress and Strain subnode is only available with some COMSOL products
(see http://www.comsol.com/products/specifications/).

In many cases Initial Stress and Strain and External Stress are
interchangeable when prescribing stresses, but you can find some more
options in the latter.

COORDINATE SYSTEM SELECTION

The Global coordinate system is selected by default. The Coordinate system list contains
any additional coordinate systems that the model includes. The given initial stresses
and strains are interpreted in this system.

INITIAL STRESS AND STRAIN

Enter values or expressions for the Initial stress S0 and Initial strain ε0. For both, enter
the diagonal and off-diagonal components (based on space dimension):

• For a 3D Initial stress model, diagonal components S0x, S0y, and S0z and
off-diagonal components S0xy, S0yz, and S0xz, for example.
• For a 3D Initial strain model, diagonal components ε0x, ε0y, and ε0z and off-diagonal
components ε0xy, ε0yz, and ε0xz, for example.

In a geometrically nonlinear analysis, the stresses should be interpreted as Second

Piola-Kirchhoff stresses, and the strains should be interpreted as Green-Lagrange
strains.

• For details about initial stresses and strains, see Inelastic Strain
Contributions and Initial Stresses and Strains.
• For details about the different strain measures, see Deformation
Measures.
• For details about the different stress measures, see Defining Stress.

THE SOLID MECHANICS INTERFACE | 477

 Prestressed Micromirror: Application Library path MEMS_Module/
Actuators/micromirror

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Linear Elastic Material>Initial Stress and Strain
Solid Mechanics>Nonlinear Elastic Material>Initial Stress and Strain
Solid Mechanics>Piezoelectric Material>Initial Stress and Strain
Solid Mechanics>Cam-Clay Material>Initial Stress and Strain
Multibody Dynamics>Linear Elastic Material>Initial Stress and Strain

Ribbon
Physics tab with Linear Elastic Material, Nonlinear Elastic Material, Piezoelectric Material,
or Cam-Clay Material node selected in the model tree:
Attributes>Initial Stress and Strain

External Stress
You can add the External Stress subnode to several material models, in order to specify
an additional stress contribution which is not part of the constitutive relation. The
external stress can be added to the total stress tensor, or act only as an extra load
contribution.

The External Stress subnode is only available with some COMSOL products (see
http://www.comsol.com/products/specifications/).

In many cases External Stress and Initial Stress and Strain are
interchangeable when prescribing stresses. In Initial Stress and Strain, the
given stress is however always added to the stress tensor.

EXTERNAL STRESS
Select a Stress input — Stress tensor (Material), Stress tensor (Spatial) or Pore pressure.

• When Stress tensor (Material) is selected, you enter the external stress in the form of
a Second Piola-Kirchhoff stress tensor. The External stress tensor drop-down list will
contain all stress tensors announced by any physics interface, and also the entry User
defined. When User defined is selected, you can enter the data for the External stress
tensor Sext as Isotropic, Diagonal, or Symmetric depending on the properties of the

478 | CHAPTER 4: SOLID MECHANICS

 tensor. The tensor components are interpreted in the selected coordinate system. If
a stress tensor announced by a physics interface is selected, the coordinate system
setting is ignored — the orientation is handled internally. Choose a Contribution type
— Add to stress tensor or Load contribution only — to determine the effect of the
contribution.
• When Stress tensor (Spatial) is selected, you enter the external stress in the form of
Cauchy stress tensor. The components are interpreted in the selected coordinate
system. Depending on the properties of the tensor, you can enter the data for the
External stress tensor σext as Isotropic, Diagonal, or Symmetric. Choose a Contribution
type — Add to stress tensor or Load contribution only — to determine the effect of
the contribution.
• When Pore pressure is selected, the Absolute pressure drop down menu will always
contain the entry User defined, in which case you manually enter a value or an
expression for the absolute pressure pA. If there are other physics interfaces (like
Darcy’s Law) which compute and announce a pressure variable, such variables are
also present in the list. You can also enter a Reference pressure level pref, which is the
pressure level at which the pore pressure does not give any stress contribution. Enter
the Biot-Willis coefficient aB to specify the fraction of the pore pressure to use. As a
default, its value is taken From material. Choose User defined to enter another value
or expression. When using Pore pressure, there is no contribution to the stress
tensor, the only effect of the pressure is as a load.

• Selecting a stress tensor announced by the same physics interface as

where the External Stress node is added, will result in an error
(‘Circular variable dependency detected’). This operation
would imply that the computed stress depends on itself.
• You can use a stress tensor from the same physics interface, but a
previous solution step. Select the User defined input type and enter
expressions where the withsol operator is used to point to the intended
solution.
• Since all stress tensor representations coincide in a geometrically linear
analysis, Stress tensor (Spatial) is needed only in the case of a
geometrically nonlinear analysis. The stress tensor is entered using a
Cauchy stress tensor representation, and is internally transformed to a
Second Piola-Kirchhoff stress tensor.

THE SOLID MECHANICS INTERFACE | 479

 • For theory, see External Stress.
• For details about the different stress measures, see Defining Stress.

Deep Excavation: Application Library path

Geomechanics_Module/Soil/Deep Excavation

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Linear Elastic Material>External Stress
Solid Mechanics>Nonlinear Elastic Material>External Stress
Solid Mechanics>Hyperelastic Material>External Stress
Membrane>Linear Elastic Material>External Stress
Membrane>Nonlinear Elastic Material>External Stress
Membrane>Hyperelastic Material>External Stress

Ribbon
Physics tab with Linear Elastic Material, Nonlinear Elastic Material, or Hyperelastic
Material node selected in the model tree:
Attributes>External Stress

External Strain
The External Strain subnode allows you to provide inelastic strain contributions to the
material models Linear Elastic Material, Nonlinear Elastic Material, and Hyperelastic
Material on a variety of formats, including using external coded functions.

COORDINATE SYSTEM SELECTION

Select a Coordinate system. All inputs that you give below are interpreted in the given
coordinate system. This also implies that external code should assume that strains are
passed with its local orientations. Deformation gradients are rotated by the local
system in both indices.

EXTERNAL STRAIN
Select the type of Strain input — External material, Strain tensor, Deformation gradient,
Deformation gradient, inverse, or Stretches. The option Strain tensor is not available for
the Hyperelastic Material.

480 | CHAPTER 4: SOLID MECHANICS

External material
For External material, the computation of an additional inelastic strain contribution is
delegated to external code which has been compiled into a shared library. External
libraries must first be imported into an External Material node under Global
Definitions>Materials.

Select an External material from the list of compatible external materials added under
Global Definitions>Materials.

For a material to be compatible with this External Strain subnode, its Interface type
must be set to a type whose required input quantities are all defined in this node.
Allowed required inputs include Green-Lagrange strains, the deformation gradient,
second Piola-Kirchoff stress as well as all standard model inputs. Select Inelastic residual
quantity — Strain or Deformation gradient to define the type of quantity by which you
want communicate with the external code.

• If the Interface type in the External material node is Inelastic residual strain, then
select Strain.
• If the Interface type in the External material node is Inelastic residual deformation,
then select Deformation gradient.

Strain tensor
For Strain tensor, enter an inelastic strain contribution εext. From the Strain tensor list,
you can choose User defined, or any strain tensor which is announced by another
physics interface. If you select User defined, enter values or expressions for the upper
diagonal part of the symmetric strain tensor.

Deformation gradient
For Deformation gradient, enter an inelastic deformation gradient contribution Fext.
From the External deformation gradient list, you can choose User defined, or any
deformation gradient tensor which is announced by another physics interface. If you
select User defined, enter values or expressions for the components of the deformation
gradient tensor.

Deformation gradient, inverse

For Deformation gradient, inverse, enter an inelastic inverse deformation gradient
–1
contribution F ext . From the External deformation gradient inverse list, you can choose
User defined, or any inverse deformation gradient tensor which is announced by
another physics interface. If you select User defined, enter values or expressions for the
components of the inverse deformation gradient tensor.

THE SOLID MECHANICS INTERFACE | 481

 Stretches
For Stretches, enter values or expressions for the three principal stretches. Entering
data on this form is convenient for some simple geometries and strain states, but in
general it is difficult to provide suitable a coordinate system for the principal
orientations.

The External Strain node is only available with some COMSOL products (see http://
www.comsol.com/products/specifications/).

• Selecting a tensor announced by the same physics interface as where

the External Strain node is added, may result in an error (‘Circular
variable dependency detected’). This operation is usually
meaningless.
• You can use a tensor from the same physics interface, but from a
previous solution step. Select the User defined input type and enter
expressions where the withsol operator is used to point to the intended
solution.
• If you select Deformation gradient form the Strain input list, and the
Additive strain decomposition check-box is selected in the parent
material node during a geometrically nonlinear study step, the
Green-Lagrange strain tensor is computed from
1 T
ε ext = --- ( F ext F ext – I )
2

• If you select Deformation gradient form the Strain input list when
engineering strains are expected, the engineering strain tensor is
computed from
1 T
ε ext = --- ( F ext + F ext ) – I
2

This is the case ff the study step is geometrically linear, or when the
Force Linear Strains check-box is selected in the parent material node,
• If you select Strain tensor form the Strain input list, and a the parent
material node operates with multiplicative strain decomposition, the
external strain is converted into a deformation gradient using the
infinitesimal strain assumption
F ext = I + ε ext

482 | CHAPTER 4: SOLID MECHANICS

 See also External Material and Working with External Materials in the
COMSOL Multiphysics Reference Manual.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Linear Elastic Material>External Strain
Solid Mechanics>Nonlinear Elastic Material>External Strain
Solid Mechanics>Hyperelastic Material>External Strain

Ribbon
Physics tab with Linear Elastic Material, Nonlinear Elastic Material, or Hyperelastic
Material node selected in the model builder tree:
Attributes>External Strain

Safety
Use the Safety subnode to set up variables which can be used to check the risk of failure
according to various criteria. It can be used in combination with Linear Elastic Material
and Nonlinear Elastic Material. Four different variables describing the failure risk will
be defined, as described in Table 4-2.

You can add any number of Safety nodes to a single material model. The contents of
this feature will not affect the analysis results as such, so you can add Safety nodes after
having performed an analysis and just do an Update Solution in order to access to the
new variables for result evaluation.
TABLE 4-2: VARIABLES FOR SAFETY FACTOR EVALUATION

VARIABLE DESCRIPTION CRITERION CRITERION

FULFILLED VIOLATED

Failure index, FI For a linear criterion, this is the ratio FI<1 FI>1
between the computed value and the
given limit.
Damage index, DI A binary value, indicating whether DI=0 DI=1
failure is predicted or not. DI is based
on the value of FI.
Safety factor, SF For a linear criterion, this is 1/FI. SF>1 SF<1
Margin of safety, MoS SF-1 MoS>0 MoS<0

For orthotropic and anisotropic failure criteria, the directions are given by the
coordinate system selection in the parent node.

THE SOLID MECHANICS INTERFACE | 483

 The Safety node is only available with some COMSOL products (see http://
www.comsol.com/products/specifications/)

FAILURE MODEL
Select a Failure Criterion. The available choices depend on the physics interface, as
indicated in Table 4-3.
TABLE 4-3: AVAILABLE FAILURE CRITERIA BY PHYSICS INTERFACE

CRITERION SOLID SHELL, MEMBRANE BEAM TRUSS

MECHANICS PLATE

von Mises X X X X X
Tresca X X X X X
Rankine X X X X X
St. Venant X X X X X
Mohr-Coulomb X - - - -
Drucker-Prager X - - - -
Jenkins X X X - -
Waddoups X X X - -
Azzi-Tsai-Hill Plane stress X - - -
Norris Plane stress X - - -
Modified Tsai-Hill Plane stress X - - -
Tsai-Hill X X X - -
Hoffman X X - - -
Tsai-Wu X X - - -
Orthotropic
Tsai-Wu X X - - -
Anisotropic
User defined X X X - -

• When Failure Criterion is von Mises Isotropic, enter Tensile strength σts.
• When Failure Criterion is Tresca Isotropic, enter Tensile strength σts.
• When Failure Criterion is Rankine Isotropic, enter Tensile strength σts and Compressive
strength σcs.
• When Failure Criterion is St. Venant Isotropic, enter Ultimate tensile strain εts.and
Ultimate compressive strain εcs.
• When Failure Criterion is Mohr-Coulomb Isotropic, select Material parameters —
Cohesion and angle of friction or Tensile and compressive strengths to determine the

484 | CHAPTER 4: SOLID MECHANICS

 type of input data.
When Cohesion and angle of friction is used, enter Cohesion c and Angle of internal
friction φ.
When Tensile and compressive strengths is used, enter Tensile strength σts and
Compressive strength σcs.
In either case, you can select Include elliptic cap to limit the allowed compressive
stress. When selected, enter the Elliptic cap parameters pa and pb.
• When Failure Criterion is Drucker-Prager Isotropic, select Material parameters —
Drucker-Prager parameters, Tensile and compressive strengths, or Mohr-Coulomb
parameters to determine the type of input data.
When Drucker-Prager parameters is used, enter Drucker-Prager alpha coefficient α and
Drucker-Prager k coefficient k.
When Tensile and compressive strengths is used, enter Tensile strength σts and
Compressive strength σcs.
When Mohr-Coulomb parameters is used, enter Cohesion c and Angle of internal friction
φ.
In either case, you can select Include elliptic cap to limit the allowed compressive
stress. When selected, enter the Elliptic cap parameters pa and pb.
• When Failure Criterion is Jenkins Orthotropic, enter Tensile strengths σts, Compressive
strengths σcs, and Shear strengths σss. All entries have three components, related to
the principal axes of orthotropy.
• When Failure Criterion is Waddoups Orthotropic, enter Ultimate tensile strains εts,
Ultimate compressive strains εcs, and Ultimate shear strains γss. All entries have three
components, related to the principal axes of orthotropy.
• When Failure Criterion is Azzi-Tsai-Hill Orthotropic, enter Tensile strengths σts,
Compressive strengths σcs, and Shear strengths σss. All entries have three
components, related to the principal axes of orthotropy.
• When Failure Criterion is Norris Orthotropic, enter Tensile strengths σts, Compressive
strengths σcs, and Shear strengths σss. All entries have three components, related to
the principal axes of orthotropy.
• When Failure Criterion is Modified Tsai-Hill Orthotropic, enter Tensile strengths σts,
Compressive strengths σcs, and Shear strengths σss. All entries have three
components, related to the principal axes of orthotropy.
• When Failure Criterion is Tsai-Hill Orthotropic, enter Tensile strengths σts, Compressive
strengths σcs, and Shear strengths σss. All entries have three components, related to
the principal axes of orthotropy.

THE SOLID MECHANICS INTERFACE | 485

 • When Failure Criterion is Hoffman Orthotropic, enter Tensile strengths σts, Compressive
strengths σcs, and Shear strengths σss. All entries have three components, related to
the principal axes of orthotropy.
• When Failure Criterion is Tsai-Wu Orthotropic, enter Tensile strengths σts, Compressive
strengths σcs, and Shear strengths σss. All entries have three components, related to
the principal axes of orthotropy.
• When Failure Criterion is Tsai-Wu Anisotropic, enter Second rank tensor, Voigt notation
f, and Fourth rank tensor F. Enter the components of the tensors with respect to the
directions of the coordinate system in the parent node.
• When Failure Criterion is User defined, you enter two expressions describing the
Failure criterion g(S), used in the failure index, and the Safety factor sf(S)
respectively. As an example, if you would like to replicate the von Mises Isotropic
criterion with tensile strength 350 MPa, you could enter g(S) as solid.mises/
350[MPa]-1 and sf(S) as 350[MPa]/(solid.mises+eps).

For all input fields, the default is to take the value From material. Change to User
defined to enter other values or expressions.

For a detailed description of the various criteria, see Safety Factor

Evaluation in the Structural Mechanics Theory chapter.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Linear Elastic Material>Safety
Solid Mechanics>Nonlinear Elastic Material>Safety
Membrane>Linear Elastic Material>Safety
Membrane>Nonlinear Elastic Material>Safety
Shell>Linear Elastic Material>Safety
Plate>Linear Elastic Material>Safety
Beam>Linear Elastic Material>Safety
Truss>Linear Elastic Material>Safety

Ribbon
Physics tab with Linear Elastic Material or Nonlinear Elastic Material node selected in the
model builder tree:
Attributes>External Strain

486 | CHAPTER 4: SOLID MECHANICS

Damping
Using the Damping subnode, you can add several types of damping to the material
model. Damping can be used in Time Dependent, Eigenfrequency, and Frequency
Domain studies; for other study types the settings in the Damping subnode are ignored.

You can add the Damping subnode to the Linear Elastic Material, the Nonlinear Elastic
Material, and the Hyperelastic Material.

The following types of damping are available:

• Rayleigh Damping
• Isotropic Loss Factor Damping
• Anisotropic Loss Factor Damping
• Orthotropic Loss Factor Damping
• Viscous Damping

The available damping models differs between various COMSOL products (see http:/
/www.comsol.com/products/specifications/).

Mechanical Damping and Losses

The applicability of the different types of damping are summarized in Table 4-4.
TABLE 4-4: AVAILABLE DAMPING TYPES

DAMPING TYPE PHYSICS INTERFACES MATERIAL STUDY TYPES

MODELS

Rayleigh damping All All Time domain,

frequency domain
Isotropic loss factor All All Frequency domain
Anisotropic loss factor All, except Beam Linear Elastic Frequency domain
and Truss
Orthotropic loss factor All, except Beam Orthotropic Frequency domain
and Truss Linear Elastic
Viscous damping Solid Mechanics, All Time domain,
Membrane, frequency domain
Multibody Dynamics

THE SOLID MECHANICS INTERFACE | 487

 DAMPING SETTINGS
Select a Damping type, and enter settings depending on the type.

Rayleigh Damping
In this damping model, the damping parameter ξ is expressed in terms of the mass m
and the stiffness k as

ξ = αdM m + β dK k

That is, Rayleigh damping is proportional to a linear combination of the stiffness and
mass; there is no direct physical interpretation of the mass damping parameter αdM
and the stiffness damping parameter βdM.

Select Input parameters — Alpha and beta — to enter the damping parameters explicitly,
or Damping ratios to derive the damping parameters from the relative damping at two
frequencies.

When Alpha and beta is selected, the Mass damping parameter αdM and the Stiffness
damping parameter βdK.

When Damping ratios is selected, enter two pairs of frequencies, f1 and f2, and the
corresponding damping ratios ζ1 and ζ2 at these frequencies. The Rayleigh damping
parameters are computed as

ς1 f2 – ς2 f1
α dM = 4πf 1 f 2 --------------------------
2 2
-
f2 – f1
ς2 f2 – ς1 f1
β dK = --------------------------
2 2
-
π ( f2 – f1 )

Isotropic Loss Factor Damping

The isotropic loss factor damping is described by the single isotropic loss factor ηs,
which acts on all entries in the elastic constitutive matrix. It can be used for isotropic,
orthotropic, and anisotropic materials.

When Isotropic loss factor is selected, use the Isotropic structural loss factor list to select
the way to enter ηs. The default is to take the value From material. For User defined,
enter another value or expression.

Anisotropic Loss Factor Damping

An elastic material is in general described by a symmetric 6-by-6 elasticity matrix D.
The loss can be isotropic or anisotropic, and is described by either the isotropic loss

488 | CHAPTER 4: SOLID MECHANICS

factor ηs or by a symmetric anisotropic 6-by-6 loss factor matrix ηD or ηDVo. The
orientations are the same as in the parent node.

When Anisotropic loss factor is selected, use the Loss factor for elasticity matrix D list to
select the way to enter ηD or ηDVo. The default is to take the values From material. For
User defined enter the components of ηD or ηDVo in the upper-triangular part of a
symmetric 6-by-6 matrix.

The values for the loss factors are ordered in two different ways,
consistent with the selection of either Standard (XX, YY, ZZ, XY, YZ, XZ)
or Voigt (XX, YY, ZZ, YZ, XZ, XY) notation in the corresponding Linear
Elastic Model. If the values are taken from the material, these loss factors
are found in the Anisotropic or Anisotropic, Voigt notation property group
for the material. For an isotropic material, the anisotropic loss factor is
always given as ηD using the standard notation.

Orthotropic Loss Factor Damping

This option is only available when Orthotropic is selected as the Linear Elastic Material
Solid model.

An orthotropic material is described by three Young’s modulus components (Ex, Ey,

and Ez) and three shear modulus components (Gxy, Gyz, and Gxz). For an orthotropic
material, loss factors can be specified in three different ways:

• Isotropic, as described under Isotropic Loss Factor Damping.

• Anisotropic, as described under Anisotropic Loss Factor Damping.
• Orthotropic, described by three plus three orthotropic loss factors corresponding to
the elastic moduli components for the orthotropic material. The orientations are the
same as in the parent Linear Elastic Material node.
When Orthotropic loss factor is selected, use the Loss factor for orthotropic Young’s
modulus list to select the way to enter ηE. The default is to take the values From
material. For User defined enter other values or expressions.

THE SOLID MECHANICS INTERFACE | 489

 Use the Loss factor for orthotropic shear modulus list to select the way to enter ηG or
ηGVo. The default is to take the values From material. For User defined enter other
values or expressions.

The values for the shear modulus loss factors are ordered in two different
ways, consistent with the selection of either Standard (XX, YY, ZZ, XY, YZ,
XZ) or Voigt (XX, YY, ZZ, YZ, XZ, XY) notation in the corresponding
Linear Elastic Model. If the values are taken from the material, these loss
factors are found in the Orthotropic or Orthotropic, Voigt notation property
group for the material.

Viscous Damping
With viscous damping, the material will get additional stresses proportional to the
strain rate. Enter Bulk viscosity ηb and Shear viscosity ηv to model damping.caused by
volume change and deformation respectively.

• The time-stepping algorithms also add numerical damping, which is

independent of any explicit damping added. For the generalized alpha
time-stepping algorithm you can control the amount of numerical
damping.
• In a study using modal superposition, you can add damping to the
individual eigenmodes.

For an example of Damping, see Heat Generation in a Vibrating

Structure: Application Library path Structural_Mechanics_Module/
Thermal-Structure_Interaction/vibrating_beam

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Linear Elastic Material>Damping
Solid Mechanics>Nonlinear Elastic Material>Damping
Solid Mechanics>Hyperelastic Material>Damping
Membrane>Linear Elastic Material>Damping
Membrane>Nonlinear Elastic Material>Damping
Membrane>Hyperelastic Material>Damping

490 | CHAPTER 4: SOLID MECHANICS

Truss>Linear Elastic Material>Damping
Multibody Dynamics>Linear Elastic Material>Damping

Ribbon
Physics tab with Linear Elastic Material, Hyperelastic Material, or Nonlinear Elastic
Material node selected in the model tree:
Attributes>Damping

Mechanical Damping
The Mechanical Damping subnode allows you to model mechanical losses in the
Piezoelectric Material, either via using the loss factor material data for the stiffness, or
in form of the Rayleigh proportional damping.

DAMPING SETTINGS
Select a Damping type — Loss factor for cE (the default), Isotropic loss factor, or Rayleigh
damping.

Loss Factor for cE

By default the Loss factor for elasticity matrix cE ηcE uses values From material. For User
defined enter other values or expressions in the matrix.

Isotropic Loss Factor

By default the Isotropic structural loss factor ηs uses values From material. For User
defined enter another value or expression.

Rayleigh Damping
Enter the Mass damping parameter αdM and the Stiffness damping parameter βdK.

In this damping model, the damping parameter ξ is expressed in terms of the mass m
and the stiffness k as

ξ = αdM m + β dK k

That is, Rayleigh damping is proportional to a linear combination of the stiffness and
mass; there is no direct physical interpretation of the mass damping parameter αdM
and the stiffness damping parameter βdM. Note that the beta-damping is applied only

THE SOLID MECHANICS INTERFACE | 491

 to the mechanical part of the problem. To include the Rayleigh damping effect for the
piezoelectric coupling terms, add a Coupling Loss subnode.

See also Piezoelectric Losses.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Piezoelectric Material>Mechanical Damping

Ribbon
Physics tab with Piezoelectric Material node selected in the model tree:
Attributes>Mechanical Damping

Coupling Loss
The Coupling Loss subnode allows you to model losses in the piezoelectric coupling in
a Piezoelectric Material, either by using the loss factor material data for the coupling
matrix or as Rayleigh proportional damping.

COUPLING LOSS SETTINGS

Select a Coupling loss — Loss factor for e (the default) or Rayleigh damping.

• For Loss factor for e select a Loss factor for coupling matrix e from the list. Select User
defined to enter values or expressions for ηe in a 3-by-6 matrix.
• For Rayleigh damping enter a Stiffness damping parameter βdC.

See also Piezoelectric Losses.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Piezoelectric Material>Coupling Loss

Ribbon
Physics tab with Piezoelectric Material node selected in the model tree:
Attributes>Coupling Loss

492 | CHAPTER 4: SOLID MECHANICS

Dielectric Loss
The Dielectric Loss subnode allows you to model electrical losses in the Piezoelectric
Material, The losses can be prescribed either by using a loss factor for the electrical
permittivity, or in the form of dielectric dispersion.

DIELECTRIC LOSS SETTINGS

From the Dielectric loss list, select Loss factor for εS (the default) or Dispersion.

For Loss factor for εS select a Loss factor for electrical permittivity εS. Select From
material (the default) to use the value from the material or select User defined to enter
values or expressions for the loss factor in the associated fields. Select Symmetric to
enter the components of ηeS in the upper-triangular part of a symmetric 3-by-3 matrix,
select Isotropic to enter a single scalar loss factor, or select Diagonal. The default values
are 0.

For Dispersion enter the Relaxation time τd, and the Relative permittivity increment
Δε rS in the associated fields. For the latter, you can select Isotropic, Diagonal, or
Symmetric matrix input options.

See also Piezoelectric Losses.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Piezoelectric Material>Dielectric Loss

Ribbon
Physics tab with Piezoelectric Material node selected in the model tree:
Attributes>Dielectric Loss

Conduction Loss (Time-Harmonic)

The Conduction Loss (Time-Harmonic) subnode allows you to model possible conductive
losses in a Piezoelectric Material. The effect is only active in a Eigenfrequency or
Frequency response study.

THE SOLID MECHANICS INTERFACE | 493

 CONDUCTION CURRENT
By default, the Electrical conductivity σe for the media is defined From material. You can
also select User defined or Linearized resistivity.

• For User defined select Isotropic, Diagonal, Symmetric, or Anisotropic depending on

the characteristics of the electrical conductivity, and then enter values or expressions
for the Electrical conductivity σe in the field or matrix.
• For Linearized resistivity the default Reference temperature T0, and Resistivity
temperature coefficient αr, and Reference resistivity ρ0 are taken From material, which
means that the values are taken from the domain (or boundary) material. T is the
current temperature, which can be a value that is specified as a model input or the
temperature from a heat transfer interface. The definition of the temperature field
appears in the Model Inputs section.

See also Piezoelectric Losses.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Piezoelectric Material>Conduction Loss

Ribbon
Physics tab with Piezoelectric Material node selected in the model tree:
Attributes>Conduction Loss

External Stress-Strain Relation

The External Stress-Strain Relation is a special type of material model where the
computation of second Piola-Kirchoff stress is delegated to external code which has
been compiled into a shared library. External libraries must first be imported into an
External Material feature under Global Definitions>Materials.

See also External Material and Working with External Materials in the
COMSOL Multiphysics Reference Manual.

The External Stress-Strain Relation node is only available with some COMSOL
products (see http://www.comsol.com/products/specifications/).

494 | CHAPTER 4: SOLID MECHANICS

MATERIAL
Select an External material from the list of compatible External Materials added under
Global Definitions>Materials. For a material to be compatible with this External Material
material model node, its Interface type must be set to a type whose required input
quantities are all defined by this External Material. Allowed required inputs include
Green-Lagrange strains, the deformation gradient, as well as all standard model inputs.

GEOMETRIC NONLINEARITY
The use of an External Stress-Strain Relation will always force the study to be
geometrically nonlinear. The default behavior is then to use a large strain formulation
in all domains. There are, however, some cases when the use of a small strain
formulation for a certain domain is needed. This will for example be the case if the
material model you have implemented is formulated using engineering strains.

In such cases, select the Force linear strains check box. When selected, a small strain
formulation is always used, independently of the setting in the study step.

• Modeling Geometric Nonlinearity

• Studies and Solvers in the COMSOL Multiphysics Reference Manual

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Material Models>External Stress-Strain Relation

Ribbon
Physics tab with Solid Mechanics selected:
Domains>Material Models>External Stress-Strain Relation

Rigid Domain
Add the Rigid Domain node and select one or more geometrical objects to make them
a rigid body. Rigid Domain is a material model, with only one material property: the
mass density. It can be used for

• Domains in the Solid Mechanics interface (2D and 3D).

• Boundaries in the Shell interface.
• Boundaries (2D) and edges (3D) in the Beam interface.

By default, an Initial Values node is added (see Initial Values (Rigid Domain)).

THE SOLID MECHANICS INTERFACE | 495

 You can add functionality to the rigid domain through the following subnodes:

• Fixed Constraint (Rigid Domain) to fully constrain the rigid domain.

• Prescribed Displacement/Rotation to prescribe the displacement of individual
degrees of freedom.
• Applied Force (Rigid Domain) to apply a force in given point.
• Applied Moment (Rigid Domain) to apply a moment.
• Mass and Moment of Inertia (Rigid Domain) to add extra mass and moment of
inertia in a given point.
• Spring Foundation (Rigid Domain) to add a translational or rotational spring or
damper in a given point.

The Rigid Domain node is only available with some COMSOL products (see http://
www.comsol.com/products/specifications/).

DENSITY
The default Density ρ is taken From material. In this case the material assignment for
the domain supplies the mass density. For User defined enter another value or
expression.

The density is needed for dynamic analysis. It is also used when computing
mass forces for gravitational or rotating frame loads, and when computing
mass properties (Computing Mass Properties).

If the density is set to zero, the center of mass of the rigid domain will be
the same as if a uniform non-zero density had been given.

CENTER OF ROTATION
Select a Center of Rotation — Center of mass, Centroid of selected entities, or User
defined. The center of rotation affects how displacements are interpreted, and is also
used as the default in various subnodes.

• For Center of mass, the center of rotation is taken as the center of mass of the rigid
domain.
• For Centroid of selected entities select an Entity level — Boundary, Edge, or Point. The
available choices depend on physics interface and geometrical dimension. The
center of rotation is located at the centroid of the selected entities, which do not

496 | CHAPTER 4: SOLID MECHANICS

 need to be related to rigid domain itself. As a special case, you can select a single
point, and thus use that point as center of rotation.

Once chosen, a default Center of Rotation: Boundary, Center of

Rotation: Edge, or Center of Rotation: Point subnode is automatically
added.

• For User defined, enter the Global coordinates of center of rotation, Xc, in the table.

Select the Offset check box to add an optional offset vector to the definition of the
center of rotation. Enter values for the offset vector Xoffset.

The center of rotation used is the sum of the vector obtained from any of the input
methods and the offset vector.

X P = X P,input + X offset

Rigid Domain Material Model

Modeling Rigid Bodies: Application Library path

Structural_Mechanics_Module/Connectors_and_Mechanisms/rigid_domain

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Material Models>Rigid Domain
Shell>Material Models>Rigid Domain
Beam>Material Models>Rigid Domain

Ribbon
Physics tab with Solid Mechanics selected:
Domains>Material Models>Rigid Domain

Physics tab with Shell selected:

Boundaries>Material Models>Rigid Domain

THE SOLID MECHANICS INTERFACE | 497

 Physics tab with Beam selected:
Edges>Material Models>Rigid Domain (3D)
Boundaries>Material Models>Rigid Domain (2D)

Initial Values (Rigid Domain)

The Initial Values node is automatically added as a default node to a Rigid Domain
node. You can enter initial values for the rigid body displacement, rotation and
velocities, which can serve as an initial condition for a transient simulation or as an
initial guess for a nonlinear analysis. The initial values that you specify are interpreted
in the selected coordinate system.

CENTER OF ROTATION
Select an option from the list: From parent, Centroid of selected entities, or User defined.
The given initial values are interpreted at the center of rotation.

• For From parent, the center of rotation is taken as the one defined in the parent Rigid
Domain node.
• For Centroid of selected entities select an Entity level — Boundary, Edge, or Point. The
available choices depend on physics interface and geometrical dimension. The initial
conditions act at the centroid of the selected entities, which do not need to be
related to rigid domain itself. As a special case, you can select a single point, and thus
prescribe the initial conditions at that point

Once chosen, a default Center of Rotation: Boundary, Center of

Rotation: Edge, or Center of Rotation: Point subnode is automatically
added.

• For User defined, enter the Global coordinates of center of rotation, Xc, in the table.

Select the Offset check box to add an optional offset vector to the definition of the
center of rotation. Enter values for the offset vector Xoffset.

The center of rotation used is the sum of the vector obtained from any of the input
methods and the offset vector.

X P = X P,input + X offset

I N I T I A L VA L U E S : TR A N S L A T I O N A L
• Displacement at center of rotation u.

498 | CHAPTER 4: SOLID MECHANICS

• Velocity at center of rotation ∂u/∂t.

IN IT IA L VA LUES: RO TATION A L
• For 3D components only: Axis of rotation Ω
• Angle of rotation φ.
• Angular velocity ω (3D components) and ∂φ/∂t (2D components)

LOCATION IN USER INTERFACE

This node is automatically added when the Rigid Domain node is created. It cannot
be added or removed manually.

Fixed Constraint (Rigid Domain)

The Fixed Constraint node adds a condition that makes the entire rigid domain fixed
(fully constrained). The displacements and rotations are zero in all directions. There
are no settings for this node. See Rigid Domain.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Rigid Domain>Fixed Constraint
Shell>Rigid Domain>Fixed Constraint
Beam>Rigid Domain>Fixed Constraint
Multibody Dynamics>Rigid Domain>Fixed Constraint

Ribbon
Physics tab with Rigid Domain node selected in the model tree:
Attributes>Fixed Constraint

When this node has been added once, it is no longer available for selection in the
context menu or ribbon.

The Fixed Constraint and Prescribed Displacement/Rotation. subnodes

cannot be combined for a Rigid Domain.

THE SOLID MECHANICS INTERFACE | 499

 Prescribed Displacement/Rotation
The Prescribed Displacement/Rotation subnode can be used to selectively constrain or
prescribe degrees the of freedom of a Rigid Domain:

• One or several displacement components can be constrained or prescribed at an

arbitrary point.
• One or several rotation components can be constrained.
• The rotation around a given axis can be prescribed.

The prescribed values that you specify are interpreted in the selected coordinate
system.

The Fixed Constraint and Prescribed Displacement/Rotation. subnodes

cannot be combined for a Rigid Domain.

CENTER OF ROTATION
The selection of the center of rotation only affects the prescribed displacements, not
the rotations.

Select an option from the list: From parent, Centroid of selected entities, or User defined.

• For From parent, the center of rotation is taken as the one defined in the parent Rigid
Domain node.
• For Centroid of selected entities select an Entity level — Boundary, Edge, or Point. The
available choices depend on physics interface and geometrical dimension.The
constraints act at the centroid of the selected entities, which do not need to be
related to rigid domain itself. As a special case, you can select a single point, and thus
constrain that point

Once chosen, a default Center of Rotation: Boundary, Center of

Rotation: Edge, or Center of Rotation: Point subnode is automatically
added.

• For User defined, enter the Global coordinates of center of rotation, Xc, in the table.

Select the Offset check box to add an optional offset vector to the definition of the
center of rotation. Enter values for the offset vector Xoffset.

500 | CHAPTER 4: SOLID MECHANICS

The center of rotation used is the sum of the vector obtained from any of the input
methods and the offset vector.

X P = X P,input + X offset

PRESCRIBED DISPLACEMENT AT CENTER OF ROTATION

Select one or several of the Prescribed in x direction, Prescribed in y direction, and
Prescribed in z direction (3D components only) check boxes. Then enter a value or
expression for u0x, u0y, and u0z (3D components).

PRESCRIBED ROTATION
Select an option from the By list: Free (the default), Constrained rotation, or Prescribed
rotation.

For 2D components, and if Prescribed rotation is selected, enter a value or

expression for the Angle of rotation φ0.

For 3D components:

• For Constrained rotation select one or several of the Constrain rotation

around x-axis, Constrain rotation around y-axis, and Constrain rotation
around z-axis check boxes.
• For Prescribed rotation enter values or expressions in the table for the
Axis of rotation Ω . Then enter a value or expression for the Angle of
rotation φ0.

• You can add a Harmonic Perturbation subnode for specifying a

harmonic variation of the values of the prescribed displacements and
rotations in a frequency domain analysis of perturbation type.
• You can activate and deactivate this boundary condition by assigning it
to a constraint group. See Load Cases in the Structural Mechanics
Modeling chapter.
• You can assign the value of the prescribed displacement and rotation to
a load group. See Load Cases in the Structural Mechanics Modeling
chapter.

THE SOLID MECHANICS INTERFACE | 501

 REACTION FORCE SETTINGS
Select Evaluate reaction forces to compute the reaction force caused by the prescribed
motion. The default is to not compute the reaction force. When selected, the
prescribed motion is implemented as a weak constraint.

Select Apply reaction only on rigid body variables to use a unidirectional constraint for
enforcing the prescribed motion. The default is that bidirectional constraints are used.
This setting is useful in a situation where a bidirectional constraint would give an
unwanted coupling in the equations. This would happen if the prescribed value of the
motion is a variable solved for in other equations.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Rigid Domain>Prescribed Displacement/Rotation
Shell>Rigid Domain>Prescribed Displacement/Rotation
Beam>Rigid Domain>Prescribed Displacement/Rotation
Multibody Dynamics>Rigid Domain>Prescribed Displacement/Rotation

Ribbon
Physics tab with Rigid Domain node selected in the model tree:
Attributes>Prescribed Displacement/Rotation

When this node has been added once, it is no longer available for selection in the
context menu or ribbon.

Applied Force (Rigid Domain)

The Applied Force subnode is t is used to apply forces on a rigid domain. The force can
act at an arbitrary position in space. A force implicitly contributes also to the moment
if it is not applied at the center of mass of a rigid domain. The force that you specify is
interpreted in the selected coordinate system.

LOCATION
Select an option from the list: Center of rotation, Centroid of selected entities, or User
defined. This is the location where the force is applied.

• For Center of rotation, the location of the load is taken as the center of rotation as
defined in the parent Rigid Domain node.
• For Centroid of selected entities select an Entity level — Boundary, Edge, or Point. The
available choices depend on physics interface and geometrical dimension. The force

502 | CHAPTER 4: SOLID MECHANICS

 acts at the centroid of the selected entities, which do not need to be related to rigid
domain itself. As a special case, you can select a single point, and thus place the force
at that point

Once chosen, a default Location: Boundary, Location: Edge, or

Location: Point subnode is automatically added.

• For User defined, enter the Location (global coordinates), Xp, in the table.

Select the Offset check box to add an optional offset vector to the definition of the
location. Enter values for the offset vector Xoffset.

The location used is the sum of the vector obtained from any of the input methods and
the offset vector.

X P = X P,input + X offset

APPLIED FORCE
Enter values or expressions for the components of the Applied force F.

• You can add the Phase subnode to specify the phase of this load in a
frequency domain analysis.
• You can specify this load to be a Harmonic Perturbation in a frequency
domain analysis.
• You can assign this load to a load group. See Load Cases in the
Structural Mechanics Modeling chapter.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Rigid Domain>Applied Force
Shell>Rigid Domain>Applied Force
Beam>Rigid Domain>Applied Force
Multibody Dynamics>Rigid Domain>Applied Force

Ribbon
Physics tab with Rigid Domain node selected in the model tree:
Attributes>Applied Force

THE SOLID MECHANICS INTERFACE | 503

 Location Nodes

LOCATION: BOUNDARY
The Location: Boundary subnode is used to select a set of boundaries whose centroid
represents the point of application of a force on a rigid domain.

LOCATION: EDGE
The Location: Edge subnode is used to select a set of edges whose centroid represents
the point of application of a force on a rigid domain.

LOCATION : PO IN T
The Location: Point subnode is used to select a set of points whose centroid represents
the point of application of a force on a rigid domain.

LOCATION IN USER INTERFACE

These nodes are automatically added when Centroid of selected entities is selected in
the parent Applied Force node. They cannot be added or removed manually.

Applied Moment (Rigid Domain)

Use the Applied Moment subnode to apply moments on a rigid domain. The moment
that you specify is interpreted in the selected coordinate system.

APPLIED MOMENT
Enter values or expressions for the components of the Applied moment.

• For 3D components, this is the x, y, and z components for M.

• For 2D components, this is for the applied moment around the z axis Mz.

• You can add the Phase subnode to specify the phase of this load in a
frequency domain analysis.
• You can specify this load to be a Harmonic Perturbation in a frequency
domain analysis.
• You can assign this load to a load group. See Load Cases in the
Structural Mechanics Modeling chapter.

504 | CHAPTER 4: SOLID MECHANICS

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Rigid Domain>Applied Moment
Shell>Rigid Domain>Applied Moment
Beam>Rigid Domain>Applied Moment
Multibody Dynamics>Rigid Domain>Applied Moment

Ribbon
Physics tab with Rigid Domain node selected in the model tree:
Attributes>Applied Moment

Mass and Moment of Inertia (Rigid Domain)

Use the Mass and Moment of Inertia subnode to add an effect of associated abstract rigid
domain, which is physically not modeled and whose inertial properties are known.
These inertial properties can be specified at an arbitrary point in space which is
assumed to be the center of gravity of this extra mass. The moment of inertia tensor
that you specify is interpreted in the selected coordinate system.

CENTER OF MASS
Here you specify the location of the center of mass for the contribution given in this
node. Select an option from the list: Center of rotation, Centroid of selected entities, or
User defined.

• For Center of rotation, the location of the load is taken as the center of rotation as
defined in the parent Rigid Domain node.
• For Centroid of selected entities select an Entity level — Boundary, Edge, or Point. The
available choices depend on physics interface and geometrical dimension.The center
of mass is located at the centroid of the selected entities, which do not need to be
related to rigid domain itself. As a special case, you can select a single point, and thus
position the mass at that point

Once chosen, a default Center of Mass: Boundary, Center of Mass: Edge,

or Center of Mass: Point subnode is automatically added.

• For User defined, enter the Global coordinates of center of mass, Xm, in the table.

Select the Offset check box to add an optional offset vector to the definition of the
location. Enter values for the offset vector Xoffset.

THE SOLID MECHANICS INTERFACE | 505

 The center of mass used is the sum of the vector obtained from any of the input
methods and the offset vector.

X m = X m,input + X offset

MASS AND MOMENT OF INERTIA

Enter values or expressions for the Mass m. Then for the Moment of inertia the axis
directions of the moment of inertia tensor are given by the selected coordinate system.

• For 3D components, select Isotropic (the default), Diagonal, or Symmetric and enter
one or more components for the tensor I.
• For 2D components, enter a value or expression for Iz.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Rigid Domain>Mass and Moment of Inertia
Shell>Rigid Domain>Mass and Moment of Inertia
Beam>Rigid Domain>Mass and Moment of Inertia
Multibody Dynamics>Rigid Domain>Mass and Moment of Inertia

Ribbon
Physics tab with Rigid Domain node selected in the model tree:
Attributes>Mass and Moment of Inertia

Center of Mass Nodes

CENTER OF MASS: BOUNDARY

Use the Center of Mass: Boundary subnode to select a set of boundaries whose centroid
represents the center of mass of a Mass and Moment of Inertia (Rigid Domain).

CENTER OF MASS: EDGE

The Center of Mass: Edge subnode to select a set of edges whose centroid represents the
center of mass of a Mass and Moment of Inertia (Rigid Domain).

CENTER OF MASS: POINT

The Center of Mass: Point subnode to select a set of points whose centroid represents
the center of mass of a Mass and Moment of Inertia (Rigid Domain).

506 | CHAPTER 4: SOLID MECHANICS

LOCATION IN USER INTERFACE
These nodes are automatically added when Centroid of selected entities is selected in
the parent Mass and Moment of Inertia (Rigid Domain) node. They cannot be added
or removed manually.

Spring Foundation (Rigid Domain)

The Spring Foundation subnode is used to specify a spring or damper connecting the
rigid domain to a fixed ground. The spring can act at an arbitrary position in space. A
translational spring implicitly contributes also to the moment if it is not applied at the
center of mass of a rigid domain. The data that you specify is interpreted in the selected
coordinate system.

LOCATION
Select an option from the list: Center of rotation, Centroid of selected entities, or User
defined. This is the location where the spring will be attached.

• For Center of rotation, the location of the spring is taken as the center of rotation as
defined in the parent Rigid Domain node.
• For Centroid of selected entities select an Entity level — Boundary, Edge, or Point. The
available choices depend on physics interface and geometrical dimension. The
spring is attached at the centroid of the selected entities, which do not need to be
related to rigid domain itself. As a special case, you can select a single point, and thus
attach the spring at that point.

Once chosen, a default Location: Boundary, Location: Edge, or

Location: Point subnode is automatically added.

• For User defined, enter the Global coordinates of center of rotation, Xc, in the table.

Select the Offset check box to add an optional offset vector to the definition of the
location. Enter values for the offset vector Xoffset.

The location used is the sum of the vector obtained from any of the input methods and
the offset vector.

X P = X P,input + X offset

SPRING
Select a Spring type — Spring constant or Force as function of extension.

THE SOLID MECHANICS INTERFACE | 507

 When Spring constant is selected, the translational spring matrix can be entered as
Isotropic, Diagonal, Symmetric, or Anisotropic. For Isotropic the same spring constant is
used in all the diagonal elements of the spring matrix.

When Force as function of extension is selected, enter the force vector Fs. It must be a
a function of the built-in variables describing the spring extension. The default value
indicates the correct variable name, for example solid.rd1.uspring1_spf1.

ROTATIONAL SPRING
Select a Spring type — Spring constant or Moment as function of rotation.

When Spring constant is selected, the rotational spring matrix can be entered as
Isotropic, Diagonal, Symmetric, or Anisotropic. For Isotropic the same spring constant is
used in all the diagonal elements of the spring matrix.

When Moment as function of rotation is selected, enter the moment vector Ms. It must
be a a function of the built-in variables describing the spring extension. The default
value indicates the correct variable name, for example solid.rd1.thspring1_spf1.

In 2D the Spring constant is only a single scalar, representing the stiffness

for rotation around the out-of-plane direction. Similarly, the Moment as
function of rotation is a single scalar.

LOSS FACTOR DAMPING

From the Loss factor type list, select Scalar (Same for all components) or Individual
components.

• For Scalar (Same for all components) enter a single Loss factor for spring ηu,s, which
is used to multiply all values of the spring matrix or spring force vector.
• For Individual components select Isotropic, Diagonal, Symmetric, or Anisotropic, then
enter values or expressions in the table for the Loss factor for spring ηu,k or ηf based
on space dimension. The loss factors act on the corresponding components of the
spring matrix or spring force vector. If you select Isotropic, the effect is the same as
when you selecting Diagonal and enter the same value for all diagonal elements.

ROTATIONAL LOSS FACTOR DAMPING

All settings in the Rotational Loss Factor Damping section are analogous to the
corresponding settings in the Loss Factor section. In 2D, only one scalar loss factor,
corresponding to rotation around the out-of-plane axis, is given.

508 | CHAPTER 4: SOLID MECHANICS

VISCOUS DAMPING
Select Isotropic, Diagonal, Symmetric, or Anisotropic, then enter values or expressions for
the damping constants du in the table. If you select Isotropic, the effect is the same as
when you selecting Diagonal and enter the same value for all diagonal elements.

VISCOUS ROTATIONAL DAMPING

All settings in the Viscous Rotational Damping section are analogous to the
corresponding settings in the Viscous Damping section. In 2D, only one scalar viscous
damping, corresponding to rotation around the out-of-plane axis, is given.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Rigid Domain>Spring Foundation
Multibody Dynamics>Rigid Domain>Spring Foundation
Shell>Rigid Domain>Spring Foundation
Beam>Rigid Domain>Spring Foundation

Ribbon
Physics tab with Rigid Domain node selected in the model tree:
Attributes>Spring Foundation

Free
The Free node is the default boundary condition. It means that there are no constraints
and no loads acting on the boundary. When the physics interfaces is added, a default
Free node is added. If you look at the selections for this node, it will show all
boundaries which do not have any boundary conditions applied.

In each physics interface, the Free node applies to a geometric entity which is one level
below the one on which the physics is active:

• For the Solid Mechanics and Multibody Dynamics interfaces, the Free node applies
to boundaries.
• For the Shell and Membrane interfaces, the Free node applies to edges.
• For the Beam and Truss interfaces, the Free node applies to points.

You can manually add Free nodes to override other boundary conditions. This is
however seldom needed.

THE SOLID MECHANICS INTERFACE | 509

 LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Free
Shell>Free
Membrane>Free
Beam>Free
Truss>Free
Multibody Dynamics>Free

Ribbon
Physics tab with a physics interface selected:
Boundaries>Solid Mechanics>Free
Boundaries>Multibody Dynamics>Free
Edges>Shell>Free
Edges>Membrane>Free
Points>Beam>Free
Points>Truss>Free

Prescribed Displacement
The Prescribed Displacement node adds a condition where the displacements are
prescribed in one or more directions to the geometric entity (domain, boundary, edge,
or point).

If a displacement is prescribed in one direction, this leaves the solid free to deform in
the other directions.

You can also define more general displacements as a linear combination of the
displacements in each direction.

If a zero displacement is applied in all directions, this is the same as a Fixed

Constraint.

For details, see Prescribed Displacements, Velocities, and Accelerations.

510 | CHAPTER 4: SOLID MECHANICS

PRESCRIBED DISPLACEMENT
Define the prescribed displacements using a Standard notation (the default) or a General
notation.

Standard Notation
To define the displacements individually, click the Standard notation button.

Select one or all of the Prescribed in x direction, Prescribed in y direction, and for 3D
components, Prescribed in z direction check boxes. Then enter a value or expression for
u0, v0, and for 3D components, w0. For 2D axisymmetric components, select one or
both of the Prescribed in r direction and Prescribed in z direction check boxes. Then
enter a value or expression for u0 and w0.

General Notation
Click the General notation to specify the displacements using a general notation that
includes any linear combination of displacement components. For example, for 2D
components, use the relationship

H u = R
v

For H matrix H select Isotropic, Diagonal, Symmetric, or Anisotropic and then enter
values as needed in the field or matrix. Enter values or expressions for the R vector R.

For example, to achieve the condition u = v, use the settings

H = 1 –1 , R = 0
0 0 0

which force the domain to move only diagonally in the xy-plane.

CONSTRAINT SETTINGS

To display this section, click the Show button ( ) and select Advanced Physics Options.

In the COMSOL Multiphysics Reference Manual:

• Constraint Reaction Terms

• Weak Constraints
• Constraint Settings

THE SOLID MECHANICS INTERFACE | 511

 When Individual dependent variables is selected in the Apply reaction terms
on drop-down list, the constraint forces are applied directly on the degrees
of freedom, which are the displacements along the global coordinate axes.
If you use this setting together with a local coordinate system, the results
will be inconsistent since the constraint forces will not match the
constraint orientation.

• You can add a Harmonic Perturbation subnode for specifying a

harmonic variation of the values of the prescribed displacements in a
frequency domain analysis of perturbation type.
• You can activate and deactivate this boundary condition by assigning it
to a constraint group. See Load Cases in the Structural Mechanics
Modeling chapter.
• You can assign the value of the prescribed displacement to a load
group. See Load Cases in the Structural Mechanics Modeling chapter.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Domain Constraints>Prescribed Displacement
Solid Mechanics>Prescribed Displacement (Boundary)
Solid Mechanics>Edges>Prescribed Displacement
Solid Mechanics>Points>Prescribed Displacement
Membrane>Face Constraints>Prescribed Displacement
Membrane>Prescribed Displacement (Edge)
Membrane>Points>Prescribed Displacement
Truss>Line Constraints>Prescribed Displacement
Truss>Prescribed Displacement (Point)
Multibody Dynamics>Domain Constraints>Prescribed Displacement
Multibody Dynamics>Prescribed Displacement (Boundary)
Multibody Dynamics>Edges>Prescribed Displacement
Multibody Dynamics>Points>Prescribed Displacement

Ribbon
Physics tab with Solid Mechanics selected:
Domains>Domain Constraints>Prescribed Displacement
Boundaries>Solid Mechanics>Prescribed Displacement

512 | CHAPTER 4: SOLID MECHANICS

Edges>Solid Mechanics>Prescribed Displacement
Points>Solid Mechanics>Prescribed Displacement

Physics tab with Membrane selected:

Boundaries>Face Constraints>Prescribed Displacement
Edges>Membrane>Prescribed Displacement
Points>Membrane>Prescribed Displacement

Physics tab with Truss selected:

Edges>Line Constraints>Prescribed Displacement
Points>Truss>Prescribed Displacement

Physics tab with Multibody Dynamics selected:

Domains>Domain Constraints>Prescribed Displacement
Boundaries>Multibody Dynamics>Prescribed Displacement
Edges>Multibody Dynamics>Prescribed Displacement
Points>Multibody Dynamics>Prescribed Displacement

Prescribed Velocity
The Prescribed Velocity node adds a boundary or domain condition where the velocity
is prescribed in one or more directions. The prescribed velocity condition is applicable
for Time Dependent and Frequency Domain studies. It is possible to prescribe a
velocity in one direction, leaving the solid free in the other directions. The Prescribed
Velocity node is a constraint and overrides any other constraint on the same selection.

The Prescribed Velocity node is only available with some COMSOL products (see
http://www.comsol.com/products/specifications/).

For details about prescribed velocities and accelerations, see Prescribed

Displacements, Velocities, and Accelerations.

COORDINATE SYSTEM SELECTION

Select a coordinate system in which the given velocity components will be interpreted.

Coordinate systems with directions that change with time should not be
used for a prescribed velocity.

THE SOLID MECHANICS INTERFACE | 513

 P R E S C R I B E D VE L O C I T Y
Select one or all of the Prescribed in x direction, Prescribed in y direction, and for 3D
components, Prescribed in z direction check boxes. Then enter a value or expression for
vx, vy, and for 3D components, vz. For 2D axisymmetric components, select one or
both of the Prescribed in r direction and Prescribed in z direction check boxes. Then
enter a value or expression for vr and vz.

In a stationary study, this node can either be ignored or treated as a constraint (similar
to a Prescribed Displacement node with zero displacement). To control this, select an
option from the Displacement in stationary study list — Free or Constrained.

• You can add a Harmonic Perturbation subnode for specifying a

harmonic variation of the values of the prescribed displacements in a
frequency domain analysis of perturbation type.
• You can activate and deactivate this boundary condition by assigning it
to a constraint group. See Load Cases in the Structural Mechanics
Modeling chapter.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Domain Constraints>Prescribed Velocity
Solid Mechanics>More Constraints>Prescribed Velocity (Boundary)
Membrane>Face Constraints>Prescribed Velocity
Membrane>More Constraints>Prescribed Velocity (Edge)
Truss>Line Constraints>Prescribed Velocity
Truss>More Constraints>Prescribed Velocity (Point)

Ribbon
Physics tab with Solid Mechanics selected:
Domains>Domain Constraints>Prescribed Velocity
Boundaries>More Constraints>Prescribed Velocity

Physics tab with Membrane selected:

Boundaries>Face Constraints>Prescribed Velocity
Edges>More Constraints>Prescribed Velocity

Physics tab with Truss selected:

Edges>Line Constraints>Prescribed Velocity
Points>More Constraints>Prescribed Velocity

514 | CHAPTER 4: SOLID MECHANICS

Prescribed Acceleration
The Prescribed Acceleration node adds a boundary or domain condition, where the
acceleration is prescribed in one or more directions. The prescribed acceleration
condition is applicable for Time Dependent and Frequency Domain studies. It is
possible to prescribe a acceleration in one direction, leaving the solid free in the other
directions. The Prescribed Acceleration node is a constraint and overrides any other
constraint on the same selection.

The Prescribed Acceleration node is only available with some COMSOL products (see
http://www.comsol.com/products/specifications/).

For details about prescribed velocities and accelerations, see Prescribed

Displacements, Velocities, and Accelerations.

COORDINATE SYSTEM SELECTION

Select a coordinate system in which the given acceleration components will be
interpreted.

Coordinate systems with directions that change with time should not be
used for a prescribed acceleration.

PRESCRIBED ACCELERATION
Select one or all of the Prescribed in x direction, Prescribed in y direction, and for 3D
components, Prescribed in z direction check boxes. Then enter a value or expression for
ax, ay, and for 3D components, az. For 2D axisymmetric components, select one or
both of the Prescribed in r direction and Prescribed in z direction check boxes. Then
enter a value or expression for ar and az.

THE SOLID MECHANICS INTERFACE | 515

 In a stationary study, this node can either be ignored or treated as a constraint (similar
to a Prescribed Displacement node with zero displacement). To control this, select an
option from the Displacement in stationary study list — Free or Constrained.

• You can add a Harmonic Perturbation subnode for specifying a

harmonic variation of the values of the prescribed displacements in a
frequency domain analysis of perturbation type.
• You can activate and deactivate this boundary condition by assigning it
to a constraint group. See Load Cases in the Structural Mechanics
Modeling chapter.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Domain Constraints>Prescribed Acceleration
Solid Mechanics>More Constraints>Prescribed Acceleration (Boundary)
Membrane>Face Constraints>Prescribed Acceleration
Membrane>More Constraints>Prescribed Acceleration (Edge)
Truss>Line Constraints>Prescribed Acceleration
Truss>More Constraints>Prescribed Acceleration (Point)

Ribbon
Physics tab with Solid Mechanics selected:
Domains>Domain Constraints>Prescribed Acceleration
Boundaries>More Constraints>Prescribed Acceleration

Physics tab with Membrane selected:

Boundaries>Face Constraints>Prescribed Acceleration
Edges>More Constraints>Prescribed Acceleration

Physics tab with Truss selected:

Edges>Line Constraints>Prescribed Acceleration
Points>More Constraints>Prescribed Velocity

Fixed Constraint
The Fixed Constraint node adds a condition that makes the geometric entity fixed (fully
constrained); that is, the displacements are zero in all directions. If there are rotational
degrees of freedom, they will also be zero.

516 | CHAPTER 4: SOLID MECHANICS

CONSTRAINT SETTINGS
To display this section, click the Show button ( ) and select Advanced Physics Options.

In the COMSOL Multiphysics Reference Manual:

• Constraint Reaction Terms

• Weak Constraints
• Constraint Settings

You can activate and deactivate this boundary condition by assigning it to

a constraint group. See Load Cases in the Structural Mechanics Modeling
chapter..

Bracket—Frequency-Response Analysis: Application Library path

Structural_Mechanics_Module/Tutorials/bracket_frequency

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Domain Constraints>Fixed Constraint
Solid Mechanics>Fixed Constraint (Boundary)
Solid Mechanics>Edges>Fixed Constraint
Solid Mechanics>Points>Fixed Constraint
Shell>Face Constraints>Fixed Constraint
Shell>Fixed Constraint (Edge)
Shell>Points>Fixed Constraint
Plate>Face Constraints>Fixed Constraint
Plate>Fixed Constraint (Boundary)
Plate>Points>Fixed Constraint
Membrane>Face Constraints>Fixed Constraint
Membrane>Fixed Constraint (Edge)
Membrane>Points>Fixed Constraint
Beam>Line Constraints>Fixed Constraint
Beam>Fixed Constraint (Point)
Multibody Dynamics>Domain Constraints>Fixed Constraint
Multibody Dynamics>Fixed Constraint (Boundary)

THE SOLID MECHANICS INTERFACE | 517

 Multibody Dynamics>Edges>Fixed Constraint
Multibody Dynamics>Points>Fixed Constraint

Ribbon
Physics tab with Solid Mechanics selected:
Domains>Domain Constraints>Fixed Constraint
Boundaries>Solid Mechanics>Fixed Constraint
Edges>Solid Mechanics>Fixed Constraint
Points>Solid Mechanics>Fixed Constraint

Physics tab with Shell selected:

Boundaries>Face Constraints>Fixed Constraint
Edges>Shell>Fixed Constraint
Points>Shell>Fixed Constraint

Physics tab with Plate selected:

Domains>Face Constraints>Fixed Constraint
Boundaries>Plate>Fixed Constraint
Points>Plate>Fixed Constraint

Physics tab with Membrane selected:

Boundaries>Face Constraints>Fixed Constraint
Edges>Membrane>Fixed Constraint
Points>Membrane>Fixed Constraint

Physics tab with Beam selected:

Edges>Line Constraints>Fixed Constraint
Points>Truss>Fixed Constraint

Physics tab with Multibody Dynamics selected:

Domains>Domain Constraints>Fixed Constraint
Boundaries>Multibody Dynamics>Fixed Constraint
Edges>Multibody Dynamics>Fixed Constraint
Points>Multibody Dynamics>Fixed Constraint

Thermal Expansion (for constraints)

Add the Thermal Expansion subnode to a constraint (Fixed Constraint or Prescribed
Displacement) to prescribe a deformation of the constrained boundary caused by
changes in temperature of the surroundings. This makes it possible to reduce the
stresses caused by such boundary conditions.

518 | CHAPTER 4: SOLID MECHANICS

The thermal strain depends on the coefficient of thermal expansion α, the temperature
T, and the strain-free reference temperature Tref as

ε th = α ( T – T ref )

The Thermal Expansion subnode is only available with some COMSOL products (see
http://www.comsol.com/products/specifications/).

THERMAL EXPANSION PROPERTIES

Specify the thermal properties that define the thermal strain. This is a description of
the thermal expansion of surroundings idealized by the constraints.

Select Inherit from domain to take the thermal expansion data from the domain being
constrained. This should only be used when:

• The temperature and the thermal expansion coefficient do not have a spatial
variation.
• The virtual surrounding material has the same thermal expansion as the domain
itself.

When Inherit from domain is not selected, enter:

• A value or expression for Temperature T, specifying the temperature distribution of

the surrounding material. Any temperature variation must be an explicit function of
the material frame coordinates. It is not possible to use a computed temperature
distribution.
• The Coefficient of thermal expansion α. As a default, values From material are used.
This requires that a material has been assigned to the boundaries, edges, or points
where the constraint is active.
For User defined select Isotropic, Diagonal or Symmetric to enter one or more
components for a general coefficient of thermal expansion tensor α. When a
non-isotropic coefficient of thermal expansion is used, the axis orientations are given
by the coordinate system selected in the Coordinate System Selection section.
• A value or expression for the Strain reference temperature Tref that is the
temperature at which there are no thermal displacements at the constraints.

Enter the coordinates of the Reference point, the point where the displacement is zero.
The choice of reference point only affects the rigid body motion. If there are several

THE SOLID MECHANICS INTERFACE | 519

 different constraints with a Thermal Expansion subnode, the same reference point
should usually be selected in all of them.

• Constraints and Thermal Expansion in the Structural Mechanics

Modeling chapter.
• Thermal Expansion of Constraints in the Structural Mechanics Theory
chapter.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Fixed Constraint>Thermal Expansion
Solid Mechanics>Prescribed Displacement>Thermal Expansion
Membrane>Fixed Constraint>Thermal Expansion
Membrane>Prescribed Displacement>Thermal Expansion
Multibody Dynamics>Fixed Constraint>Thermal Expansion
Multibody Dynamics>Prescribed Displacement>Thermal Expansion

Ribbon
Physics tab with Fixed Constraint or Prescribed Displacement node selected in the model
tree:
Attributes>Thermal Expansion

Roller
The Roller node adds a roller constraint as the boundary condition; that is, the
displacement is zero in the direction perpendicular (normal) to the boundary, but the
boundary is free to move in the tangential direction. A Roller condition is equivalent
to a Symmetry condition.

CONSTRAINT SETTINGS
To display this section, click the Show button ( ) and select Advanced Physics Options.

In the COMSOL Multiphysics Reference Manual:

• Constraint Reaction Terms

• Weak Constraints
• Constraint Settings

520 | CHAPTER 4: SOLID MECHANICS

 You can activate and deactivate this boundary condition by assigning it to
a constraint group. See Load Cases in the Structural Mechanics Modeling
chapter.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Roller
Multibody Dynamics>Roller

Ribbon
Physics tab with Solid Mechanics selected:
Boundaries>Solid Mechanics>Roller

Physics tab with Multibody Dynamics selected:

Boundaries>Multibody Dynamics>Roller

Symmetry
The Symmetry node adds a boundary condition that represents symmetry in the
geometry and in the loads. A symmetry condition is free in the plane and fixed in the
out-of-plane direction.

When applied to an edge (in the Membrane interface) the symmetry plane is formed
by the normal to the boundary and the edge tangent.

The Symmetry node is only available with some COMSOL products (see http://
www.comsol.com/products/specifications/).

CONSTRAINT SETTINGS
To display this section, click the Show button ( ) and select Advanced Physics Options.

In the COMSOL Multiphysics Reference Manual:

• Constraint Reaction Terms

• Weak Constraints
• Constraint Settings

THE SOLID MECHANICS INTERFACE | 521

 Symmetry Constraints

You can activate and deactivate this boundary condition by assigning it to

a constraint group. See Load Cases in the Structural Mechanics Modeling
chapter.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>More Constraints>Symmetry
Membrane>More Constraints>Symmetry

Ribbon
Physics tab with Solid Mechanics selected:
Boundaries>More Constraints>Symmetry

Physics tab with Membrane selected:

Edges>More Constraints>Symmetry

Antisymmetry
The Antisymmetry node adds a boundary condition for an antisymmetry boundary. An
antisymmetry condition is fixed in the plane and free in the out-of-plane direction.

When applied to an edge (in the Membrane interface) the antisymmetry plane is
formed by the normal to the boundary and the edge tangent.

The Antisymmetry node is only available with some COMSOL products (see http://
www.comsol.com/products/specifications/).

In a geometrically nonlinear analysis, large rotations must not occur at the

antisymmetry plane because this causes artificial straining.

522 | CHAPTER 4: SOLID MECHANICS

CONSTRAINT SETTINGS
To display this section, click the Show button ( ) and select Advanced Physics Options.

In the COMSOL Multiphysics Reference Manual:

• Constraint Reaction Terms

• Weak Constraints
• Constraint Settings

Symmetry Constraints

You can activate and deactivate this boundary condition by assigning it to

a constraint group. See Load Cases in the Structural Mechanics Modeling
chapter.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>More Constraints>Antisymmetry
Membrane>More Constraints>Antisymmetry

Ribbon
Physics tab with Solid Mechanics selected:
Boundaries>More Constraints>Antisymmetry

Physics tab with Membrane selected:

Edges>More Constraints>Antisymmetry

Rigid Motion Suppression

The Rigid Motion Suppression node adds a minimum number of constraints required to
suppress any rigid body modes. The constraints are selected so that there will be no
reaction forces if the external loads are self-equilibrating.

The constraint will, depending on physics interface and geometrical dimension, be

applied to one, two, or three points as needed. These points are automatically picked
from the selected geometrical objects.

THE SOLID MECHANICS INTERFACE | 523

 If you need to constrain several objects which are not physically connected, you must
add as many Rigid Motion Suppression nodes as there are disconnected objects.

In the COMSOL Multiphysics Reference Manual:

• Constraint Reaction Terms

• Weak Constraints
• Constraint Settings

• Rigid body motion in the Structural Mechanics Modeling chapter.

• Rigid Motion Suppression Boundary Condition in the Structural
Mechanics Theory chapter.

You can activate and deactivate this boundary condition by assigning it to

a constraint group. See Load Cases in the Structural Mechanics Modeling
chapter.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Domain Constraints>Rigid Body Suppression
Shell>Face Constraints>Rigid Body Suppression
Plate>Face Constraints>Rigid Body Suppression
Membrane>Face Constraints>Rigid Body Suppression
Beam>Line Constraints>Rigid Body Suppression
Multibody Dynamics>Domain Constraints>Rigid Body Suppression

Ribbon
Physics tab with Solid Mechanics or Multibody Dynamics selected:
Domains>Domain Constraints>Rigid Body Suppression

Physics tab with Shell or Membrane selected:

Boundaries>Face Constraints>Rigid Body Suppression

Physics tab with Plate selected:

Domains>Face Constraints>Rigid Body Suppression

Physics tab with Beam or Truss selected:

Edges>Line Constraints>Fixed Constraint

524 | CHAPTER 4: SOLID MECHANICS

Body Load
Add a Body Load to domains for modeling volumetric loads.

FORCE
Select a Load type — Force per unit volume (the default), Total force, or for 2D
components, Force per unit area.

Then enter values or expressions for the components in the matrix based on the
selection and the space dimension.

• After selecting a Load type, the Load list normally only contains User
defined. When combining with another physics interface that can
provide this type of load, it is also possible to choose a predefined load
from this list.
• For Total force COMSOL Multiphysics divides the total force by the
volume of the domains where the load is active. For 2D components,
and if Force per unit area is selected, the body load as force per unit
volume is then the value of F divided by the thickness.

LOAD TYPE VARIABLE SI UNIT GEOMETRIC SPACE DIMENSION

ENTITY (COMPONENTS)
LEVEL

Force per unit volume FV N/m3 domains 3D (x, y, z)

2D (x, y)
2D axisymmetric (r, z)
Force per unit area FA N/m2 domains 2D (x, y)
Total force Ftot N domains 3D (x, y, z)
2D (x, y)
2D axisymmetric (r, z)

• You can add the Phase subnode to specify the phase of this load in a
frequency domain analysis.
• You can specify this load to be a Harmonic Perturbation in a frequency
domain analysis.
• You can assign this load to a load group. See Load Cases in the
Structural Mechanics Modeling chapter.

THE SOLID MECHANICS INTERFACE | 525

 LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Volume Forces>Body Load
Membrane>Face and Volume Loads>Body Load
Multibody Dynamics>Volume Forces>Body Load

Ribbon
Physics tab with Solid Mechanics or Multibody Dynamics selected:
Domains>Volume Forces>Body Load

Physics tab with Membrane selected:

Boundaries>Face and Volume Loads>Body Load

Gravity
When you add a Gravity node, gravity forces are applied to all selected features in the
physics interface with a density, mass, or mass distribution. You select objects having
the highest geometrical dimension of the interface, and all objects with a lower
dimensionality which belong to the selection are automatically included.

The gravity acts in a fixed spatial direction. The load intensity is g = ρgeg where g as a
default is the acceleration of gravity (a predefined physical constant). The action of
gravity can also be seen as a linearly accelerated frame of reference.

The Gravity node is only available with some COMSOL products (see http://
www.comsol.com/products/specifications/).

COORDINATE SYSTEM SELECTION

The Global coordinate system is selected by default. The Coordinate system list contains
any additional coordinate systems that the model includes. It can be used when
prescribing the direction of the gravitational forces.

GRAVITY
Enter the components of the Gravity g. The default value is g_const which is the
physical constant having the value 9.8066 m/s2.

526 | CHAPTER 4: SOLID MECHANICS

For 2D components, the default is that the gravity acts in the negative y direction. For
3D and 2D axisymmetric components, the default is that the gravity acts in the
negative z direction.

Rotating Frame and Gravity

Only features which have a geometrical selection contribute to the mass

forces. The Mass and Moment of Inertia nodes are global features and will
not get any contribution from Gravity.

You can assign this load to a load group. See Load Cases in the Structural
Mechanics Modeling chapter.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Volume Forces>Gravity
Shell>Face and Volume Loads>Gravity
Plate>Face and Volume Loads>Gravity
Membrane>Face and Volume Loads>Gravity
Beam>Line and Volume Loads>Gravity
Truss>Line and Volume Loads>Gravity
Multibody Dynamics>Volume Forces>Gravity

Ribbon
Physics tab with Solid Mechanics or Multibody Dynamics selected:
Domains>Volume Forces>Gravity

Physics tab with Shell or Membrane selected:

Boundaries>Face and Volume Loads>Gravity

Physics tab with Plate selected:

Domains>Face and Volume Loads>Gravity

Physics tab with Beam or Truss selected:

Edges>Line and Volume Loads>Gravity

THE SOLID MECHANICS INTERFACE | 527

 Rotating Frame
Centrifugal, Coriolis, and Euler forces are “fictitious” volume forces that need to be
introduced in a rotating frame of reference, since it is not an inertial system. Use a
Rotating Frame node to add the effect of these forces. The forces are generated by all
selected features in the physics interface having a density, mass, or mass distribution.
You select objects having the highest geometrical dimension of the interface, and all
objects with a lower dimensionality which belong to the selection are automatically
included.

The Rotating Frame node is only available with some COMSOL products (see http://
www.comsol.com/products/specifications/).

ROTATING FRAME
Select an Axis of rotation — x-axis, y-axis, z-axis, or User defined. For User defined enter
a Rotation axis base point rbp and Rotation axis direction eax.

Select a Rotational direction — Counterclockwise or Clockwise. The rotational direction

does not make any difference for a centrifugal force.

Select a Rotational frequency — Angular velocity, RPM, or Revolutions per time. Then
enter a value as needed for Angular velocity magnitude Ω , RPM, or Revolutions per time.

For 3D and 2D components, use the Centrifugal force, Coriolis force, or Euler force
check boxes to determine which effects of a rotating frame that are to be incorporated
in the analysis. Only Centrifugal force is selected by default.

For 2D axisymmetric components, the only effect from a rotating frame is the
centrifugal force, which is then always included.

The Spin softening check box is selected by default. When including spin-softening
effects, an extra contribution to the centrifugal force from deformation is taken into
account. The Spin softening check box is only available if Centrifugal force or Euler force
is selected.

Rotating Frame and Gravity

528 | CHAPTER 4: SOLID MECHANICS

 Only features which have a geometrical selection contribute to the mass
forces. The Mass and Moment of Inertia nodes are global features and will
not get any contribution from Rotating Frame.

You can assign this load to a load group. See Load Cases in the Structural
Mechanics Modeling chapter.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Volume Forces>Rotating Frame
Shell>Face and Volume Loads>Rotating Frame
Plate>Face and Volume Loads>Rotating Frame
Membrane>Face and Volume Loads>Rotating Frame
Beam>Line and Volume Loads>Rotating Frame
Truss>Line and Volume Loads>Rotating Frame
Multibody Dynamics>Volume Forces>Rotating Frame

Ribbon
Physics tab with Solid Mechanics or Multibody Dynamics selected:
Domains>Volume Forces>Rotating Frame

Physics tab with Shell or Membrane selected:

Boundaries>Face and Volume Loads>Rotating Frame

Physics tab with Plate selected:

Domains>Face and Volume Loads>Rotating Frame

Physics tab with Beam or Truss selected:

Edges>Line and Volume Loads>Rotating Frame

Boundary Load
Use a Boundary Load to apply tractions or pressure to boundaries.

THE SOLID MECHANICS INTERFACE | 529

 FORCE
Select a Load type — Force per unit area (the default), Pressure, Total force, or for 2D
components, Force per unit length. Then enter values or expressions for the
components in the matrix based on the selection and the space dimension.

• For Force per unit area, the body load as force per unit volume is then the value of
F divided by the thickness.
• For Total force, COMSOL Multiphysics then divides the total force by the area of
the surfaces where the load is active.
• For Pressure, a scalar input is given, and the orientation of the load is given by the
normal to the boundary. The pressure is positive when directed toward the solid. In
a geometrically nonlinear analysis, the load current surface normal is used.

After selecting a Load type, the Load list normally only contains User
defined. When combining with another physics interface that can provide
this type of load, it is also possible to choose a predefined load from this
list.

LOAD TYPE VARIABLE SI UNIT GEOMETRIC SPACE DIMENSION

ENTITY (COMPONENTS)
LEVEL

Force per unit area FA N/m2 boundaries 3D (x, y, z)

2D (x, y)
2D axisymmetric (r, z)
Force per unit length FL N/m boundaries 2D (x, y)
Total force Ftot N boundaries 3D (x, y, z)
2D (x, y)
2D axisymmetric (r, z)
Pressure p Pa boundaries 3D (x, y, z)
2D (x, y)
2D axisymmetric (r, z)

530 | CHAPTER 4: SOLID MECHANICS

 • You can add the Phase subnode to specify the phase of this load in a
frequency domain analysis.
• You can specify this load to be a Harmonic Perturbation in a frequency
domain analysis.
• You can assign this load to a load group. See Load Cases in the
Structural Mechanics Modeling chapter.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Boundary Load
Multibody Dynamics>Boundary Load

Ribbon
Physics tab with Solid Mechanics or Multibody Dynamics selected:
Boundaries>Solid Mechanics>Body Load
Boundaries>Multibody Dynamics>Body Load

Edge Load
Add an Edge Load to 3D components to apply a force distributed along an edge.

FORCE
Select a Load type — Force per unit length (the default) or Total force. Then enter values
or expressions for the components in the matrix based on the selection:

• The load per unit length FL.

• The total force Ftot. COMSOL Multiphysics then divides the total force by the
volume where the load is active.

After selecting a Load type, the Load list normally only contains User
defined. When combining with another physics interface that can provide
this type of load, it is also possible to choose a predefined load from this
list.

THE SOLID MECHANICS INTERFACE | 531

 • You can add the Phase subnode to specify the phase of this load in a
frequency domain analysis.
• You can specify this load to be a Harmonic Perturbation in a frequency
domain analysis.
• You can assign this load to a load group. See Load Cases in the
Structural Mechanics Modeling chapter.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Edges>Edge Load
Multibody Dynamics>Edges>Edge Load

Ribbon
Physics tab with Solid Mechanics or Multibody Dynamics selected:
Edges>Edge Load

Point Load
Add a Point Load to points for concentrated forces at points in 2D and 3D.

FORCE
Enter values or expressions for the components of the point load Fp.

The Load list normally only contains User defined. When combining with
another physics interface that can provide this type of load, it is also
possible to choose a predefined load from this list.

• You can add the Phase subnode to specify the phase of this load in a
frequency domain analysis.
• You can specify this load to be a Harmonic Perturbation in a frequency
domain analysis.
• You can assign this load to a load group. See Load Cases in the
Structural Mechanics Modeling chapter.

532 | CHAPTER 4: SOLID MECHANICS

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Points>Point Load
Multibody Dynamics>Points>Point Load
Truss>Point Load
Multibody Dynamics>Points>Point Load

Ribbon
Physics tab with Solid Mechanics, Multibody Dynamics, or Multibody Dynamics selected:
Points>Point Load

Physics tab with Truss selected:

Points>Truss>Point Load

Ring Load
Add a Ring Load to points located at R > 0 in axially symmetric models. Select this
feature from the Points submenu.

FORCE
Select the Load type — Force per unit length (the default) or Total force. Enter values
or expressions for FL or Ftot.f

• The Load list normally only contains User defined. When combining
with another physics interface that can provide this type of load, it is
also possible to choose a predefined load from this list.
• You can add the Phase subnode to specify the phase of this load in a
frequency domain analysis.
• You can specify this load to be a Harmonic Perturbation in a frequency
domain analysis.
• You can assign this load to a load group. See Load Cases in the
Structural Mechanics Modeling chapter.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Points>Ring Load
Membrane>Ring Load

THE SOLID MECHANICS INTERFACE | 533

 Ribbon
Physics tab with Solid Mechanics selected:
Points>Ring Load

Physics tab with Membrane selected:

Points>Membrane>Ring Load

Point Load (on Axis)

A Point Load (on Axis) node can be added to points located at R = 0 in axially symmetric
models. This is the only true point load an axisymmetric model, since loads applied at
points having non-zero radial coordinates actually represent a Ring Load.

FORCE
Enter values or expressions for the Force Fz in the axial direction.

• The Load list normally only contains User defined. When combining
with another physics interface that can provide this type of load, it is
also possible to choose a predefined load from this list.
• You can add the Phase subnode to specify the phase of this load in a
frequency domain analysis.
• You can specify this load to be a Harmonic Perturbation in a frequency
domain analysis.
• You can assign this load to a load group. See Load Cases in the
Structural Mechanics Modeling chapter.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Points>Point Load (on Axis)
Membrane>Point Load (on Axis)

Ribbon
Physics tab with Solid Mechanics selected:
Points>Point Load (on Axis)

Physics tab with Membrane selected:

Points>Membrane>Point Load (on Axis)

534 | CHAPTER 4: SOLID MECHANICS

Spring Foundation
Use the Spring Foundation node to apply elastic and damping boundary conditions for
domains, boundaries, edges, and points.

By adding the Predeformation subnode, you can prescribe that the spring force is zero
at a non-zero spring extension.

The Spring Foundation and Thin Elastic Layer nodes are similar, with the difference that
a Spring Foundation connects the structural part on which it is acting to a fixed
“ground”, while a Thin Elastic Layer acts between two parts, either on an interior
boundary or between two boundaries forming a pair.

The Spring Foundation node is only available with some COMSOL products (see http:/
/www.comsol.com/products/specifications/).

COORDINATE SYSTEM SELECTION

The spring and damping constants are given with respect to the selected coordinate
system.

SPRING
Select the Spring type and its associated spring constant or force using Table 4-5 as a
guide. The default option is the spring type for the type of geometric entity and space
dimension, and there are different combinations available based on this.

When the option is of the type ‘force as function of extension’, then the built-in
variables describing the spring extension must be used in the expression as described
in Springs and Dampers. The spring matrix can be entered as Isotropic, Diagonal,
Symmetric, or Anisotropic. For Isotropic the same spring constant is used in all the
diagonal elements of the spring matrix.

When Use material data is selected as Spring type, the spring stiffness values are
computed from the given material data and a layer thickness. From the Specify list,
select a pair of elastic properties — Young’s modulus and Poisson’s ratio, Young’s modulus
and shear modulus, or Bulk modulus and shear modulus. Each of these pairs define the
elastic properties and it is possible to convert from one set of properties to another
according to Table 4-6. For the chosen properties, select from the applicable list to use
the value From material or enter a User defined value or expression. In order to use From
material, you must have assigned a material to the selected boundaries.

THE SOLID MECHANICS INTERFACE | 535

 Enter a Layer thickness, ds, to specify the physical thickness of the elastic layer.
TABLE 4-5: SPRING TYPES FOR THE SPRING FOUNDATION FEATURE

SPRING TYPE VARIABLE SI UNITS GEOMETRIC ENTITY SPACE DIMENSION

LEVEL

Spring constant per kV N/(m⋅m3) domains 3D, 2D, and 2D

unit volume axisymmetric
Total spring constant ktot N/m domains, edges 3D, 2D, and 2D
axisymmetric
Spring constant per kA N/(m⋅m)2 boundaries 3D, 2D
unit area
Spring constant per kL N/(m⋅m) edges, boundaries 3D, 2D
unit length (2D)
Spring constant kP N/m points 3D, 2D, and 2D
axisymmetric
Force per volume as FV N/m3 domains 3D, 2D, and 2D
function of axisymmetric
extension
Total force as Ftot N domains, 3D, 2D, and 2D
function of boundaries, edges axisymmetric
extension
Force per area as FA N/m2 domains, 3D, 2D
function of boundaries
extension
Force per length as FL N/m edges 3D
function of
extension
Force as function of FP N points 3D, 2D, and 2D
extension axisymmetric
Use material data N/A N/A boundaries 3D, 2D, and 2D
axisymmetri
TABLE 4-6: EXPRESSIONS FOR THE ELASTIC MODULI.

DESCRIPTION VARIABLE D(E,ν) D(E,G) D(K,G)

Young’s E= E E 9KG -
------------------
modulus 3K + G
Poisson’s ν= ν E-
---  1 – ------------------
3G -
------- –1 1
ratio 2G 2 3K + G

536 | CHAPTER 4: SOLID MECHANICS

TABLE 4-6: EXPRESSIONS FOR THE ELASTIC MODULI.

DESCRIPTION VARIABLE D(E,ν) D(E,G) D(K,G)

Bulk K= E - EG - K
----------------------- --------------------------
modulus 3 ( 1 – 2ν ) 3 ( 3G – E )
Shear G= E - G G
--------------------
modulus 2(1 + ν)

ROTATIONAL SPRING
This section is available for the Beam interface. All settings in the Rotational Spring
section are analogous to the corresponding settings in the Spring section, but with
forces replaced by moments and displacements replaced by rotations.

LOSS FACTOR DAMPING

From the Loss factor type list, select Scalar (Same for all components) or Individual
components.

• For Scalar (Same for all components) enter a single Loss factor for spring ηs which is
used to multiply all values of the spring matrix or spring force vector.
• For Individual components select Isotropic, Diagonal, Symmetric, or Anisotropic, then
enter values or expressions in the table for the Loss factor for spring ηk or ηf based
on space dimension. The loss factors act on the corresponding components of the
spring matrix or spring force vector. If you select Isotropic, the effect is the same as
when you selecting Diagonal and enter the same value for all diagonal elements.

ROTATIONAL LOSS FACTOR DAMPING

This section is available for the Beam interface. All settings in the Rotational Loss Factor
Damping section are analogous to the corresponding settings in the Loss Factor section.

VISCOUS DAMPING
Select the Damping type using Table 4-7 as a guide. The default option is the default
damping type for the type of geometric entity and space dimension, and there are
different combinations available based on this. The damping matrix can be entered as

THE SOLID MECHANICS INTERFACE | 537

 Isotropic, Diagonal, Symmetric, or Anisotropic. For Isotropic the same viscous constant is
used in all the diagonal elements of the damping matrix
TABLE 4-7: DAMPING TYPES FOR THE SPRING FOUNDATION FEATURE

DAMPING TYPE VARIABLE SI UNITS GEOMETRIC ENTITY SPACE DIMENSION

LEVEL

Damping constant dV N⋅s/(m⋅m3) domains, 3D, 2D

per unit volume boundaries (2D)
Damping constant dA N⋅s/(m⋅m2) domains, 3D, 2D, and 2D
per unit area boundaries axisymmetric
Total damping dtot N⋅s/m domains, 3D, 2D, and 2D
constant boundaries, axisymmetric
edges, points
Damping constant dL N⋅s/(m⋅m) edges 3D
per unit length

ROTATIONAL VISCOUS DAMPING

This section is available for the Beam interface. All settings in the Rotational Viscous
Damping section are analogous to the corresponding settings in the Viscous Damping
section, but with forces replaced by moments and velocities replaced by angular
velocities.

You can activate and deactivate this boundary condition by assigning it to

a constraint group. See Load Cases in the Structural Mechanics Modeling
chapter.

• Springs and Dampers in the Structural Mechanics Modeling chapter.

• Spring Foundation and Thin Elastic Layer in the Structural Mechanics
Theory chapter.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Mass, Spring, and Damper>Spring Foundation (Domain, Boundary)
Solid Mechanics>Edges>Spring Foundation
Solid Mechanics>Points>Spring Foundation
Shell>Mass, Spring, and Damper>Spring Foundation (Boundary, Edge)
Shell>Points>Spring Foundation

538 | CHAPTER 4: SOLID MECHANICS

Plate>Mass, Spring, and Damper>Spring Foundation (Domain, Boundary)
Plate>Points>Spring Foundation
Membrane>Mass, Spring, and Damper>Spring Foundation (Boundary, Edge)
Membrane>Points>Spring Foundation
Beam>Mass, Spring, and Damper>Spring Foundation
Truss>Mass, Spring, and Damper>Spring Foundation
Multibody Dynamics>Mass, Spring, and Damper>Spring Foundation (Domain)
Multibody Dynamics>Spring Foundation (Boundary)
Multibody Dynamics>Edges>Spring Foundation
Multibody Dynamics>Points>Spring Foundation

Ribbon
Physics tab with Solid Mechanics selected:
Domains>Mass, Spring, and Damper>Spring Foundation
Boundaries>Mass, Spring, and Damper>Spring Foundation
Edges>Solid Mechanics>Spring Foundation
Points>Solid Mechanics>Spring Foundation

Physics tab with Shell or Membrane selected:

Boundaries>Mass, Spring, and Damper>Spring Foundation
Edges>Mass, Spring, and Damper>Spring Foundation
Points>Shell>Spring Foundation
Points>Membrane>Spring Foundation

Physics tab with Plate selected:

Domains>Mass, Spring, and Damper>Spring Foundation
Boundaries>Mass, Spring, and Damper>Spring Foundation
Points>Plate>Spring Foundation

Physics tab with Beam or Truss selected:

Edges>Mass, Spring, and Damper>Spring Foundation
Points>Mass, Spring, and Damper>Spring Foundation

Physics tab with Multibody Dynamics selected:

Domains>Mass, Spring, and Damper>Spring Foundation
Boundaries>Multibody Dynamics>Spring Foundation
Edges>Multibody Dynamics>Spring Foundation
Points>Multibody Dynamics>Spring Foundation

THE SOLID MECHANICS INTERFACE | 539

 Thin Elastic Layer
Use the Thin Elastic Layer node to apply elastic and damping conditions between two
parts, either on an interior boundary or on a boundary pair.

By adding the Predeformation subnode, you can prescribe that the spring force is zero
at a non-zero spring extension.

The Thin Elastic Layer and Spring Foundation nodes are similar, with the difference that
a Spring Foundation connects the structural part on which it is acting to a fixed
“ground”.

The Thin Elastic Layer node is only available with some COMSOL products (see http:/
/www.comsol.com/products/specifications/).

PAIR SELECTION
If this node is selected from the Pairs menu, choose the pair on which to apply this
condition. An identity pair has to be created first. Ctrl-click to deselect.

COORDINATE SYSTEM SELECTION

The spring and damping constants are given with respect to the selected coordinate
system, with the exception that when the stiffness is specified by Use material data, the
coordinate system selection is not used.

SPRING
Select the Spring type and its associated spring constant or force using Table 4-8 as a
guide. The default option is the spring type for the type of geometric entity and space
dimension, and there are different combinations available based on this.

When the option is of the type ‘force as function of extension’, then the built-in
variables describing the spring extension must be used in the expression as described
in Springs and Dampers. The spring matrix can be entered as Isotropic, Diagonal,
Symmetric, or Anisotropic. For Isotropic the same spring constant is used in all the
diagonal elements of the spring matrix.

When Use material data is selected as Spring type, the spring stiffness values are
computed from material data and layer thickness. From the Specify list, select a pair of
elastic properties — Young’s modulus and Poisson’s ratio, Young’s modulus and shear
modulus, or Bulk modulus and shear modulus. Each of these pairs define the elastic
properties and it is possible to convert from one set of properties to another according
to Table 4-6. For the chosen properties, select from the applicable list to use the value

540 | CHAPTER 4: SOLID MECHANICS

From material or enter a User defined value or expression. In order to use From material,
you must have assigned a material to the selected boundaries.

Enter a Layer thickness, ds, to specify the physical thickness of the elastic layer.
TABLE 4-8: SPRING TYPES FOR THE THIN ELASTIC LAYER FEATURE

SPRING TYPE VARIABLE SI UNITS SPACE DIMENSION

Total spring constant ktot N/m 3D, 2D, and 2D

axisymmetric
Spring constant per unit area kA N/(m⋅m)2 3D, 2D, and 2D
axisymmetric
Spring constant per unit length kL N/(m⋅m) 2D
Total force as function of extension Ftot N 3D, 2D, and 2D
axisymmetric
Force per area as function of FA N/m2 3D, 2D, and 2D
extension axisymmetric
Force per length as function of FL N/m 2D
extension
Use material data N/A N/A 3D, 2D, and 2D
axisymmetric

LOSS FACTOR DAMPING

From the Loss factor type list, select Scalar (Same for all components) or Individual
components.

• For Scalar (Same for all components) enter a single Loss factor for spring ηs which is
used to multiply all values of the spring matrix or spring force vector.
• For Individual components select Isotropic, Diagonal, Symmetric, or Anisotropic and
enter values or expressions in the table for the Loss factor for spring ηk or ηf based
on space dimension. The loss factors act on the corresponding components of the
spring matrix or spring force vector. If you select Isotropic, the effect is the same as
when you selecting Diagonal and enter the same value for all diagonal elements.

VISCOUS DAMPING
Select the Damping type using Table 4-9 as a guide. The default option is the default
damping type for the space dimension. The damping matrix can be entered as Isotropic,

THE SOLID MECHANICS INTERFACE | 541

 Diagonal, Symmetric, or Anisotropic. For Isotropic the same viscous constant is used in
all the diagonal elements of the damping matrix
TABLE 4-9: DAMPING TYPES FOR THE THIN ELASTIC LAYER FEATURE

DAMPING TYPE VARIABLE SI UNITS SPACE DIMENSION

2
Damping constant per unit area dA N⋅s/(m⋅m ) 3D, 2D, and 2D
axisymmetric
Total damping constant dtot N⋅s/m 3D, 2D, and 2D
axisymmetric
Damping constant per unit dL N⋅s/(m⋅m) 2D
length

You can activate and deactivate this boundary condition by assigning it to

a constraint group. See Load Cases in the Structural Mechanics Modeling
chapter.

• Springs and Dampers in the Structural Mechanics Modeling chapter.

• Spring Foundation and Thin Elastic Layer in the Structural Mechanics
Theory chapter.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Mass, Spring, and Damper>Thin Elastic Layer
Solid Mechanics>Pairs>Thin Elastic Layer

Ribbon
Physics tab with Solid Mechanics selected:
Boundaries>Mass, Spring, and Damper>Thin Elastic Layer
Pairs>Solid Mechanics>Thin Elastic Layer

Predeformation
Use the Predeformation subnode to specify that the elastic forces in Spring Foundation
or Thin Elastic Layer are non-zero at zero displacement. Thus you can model cases
where the unstressed state of the spring is in another configuration than the one
described by the geometry.

542 | CHAPTER 4: SOLID MECHANICS

The value of the predeformation can vary during the simulation, so it should not be
interpreted as an initial value.

SPRING PREDEFORMATION
Based on space dimension, enter the values for the Spring Predeformation u0.

You can assign the load caused by the predeformation to a load group.
See Load Cases in the Structural Mechanics Modeling chapter.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Spring Foundation>Predeformation
Solid Mechanics>Thin Elastic Layer>Predeformation
Shell>Spring Foundation>Predeformation
Plate>Spring Foundation>Predeformation
Membrane>Spring Foundation>Predeformation
Beam>Spring Foundation>Predeformation
Truss>Spring Foundation>Predeformation
Multibody Dynamics>Spring Foundation>Predeformation

Ribbon
Physics tab with Spring Foundation or Thin Elastic Layer node selected in the model tree:
Attributes>Predeformation

Added Mass
The Added Mass node is available on domains, boundaries, and edges and can be used
to supply inertia that is not part of the material itself. Such inertia does not need to be
isotropic, in the sense that the inertial effects are not the same in all directions.

The Added Mass node is only available with some COMSOL products (see http://
www.comsol.com/products/specifications/).

COORDINATE SYSTEM SELECTION

The added mass values are given with respect to the selected coordinate directions.

THE SOLID MECHANICS INTERFACE | 543

 ADDED MASS
Select a Mass type using Table 4-10 as a guide. The default option is the type for the
geometric entity. The added mass matrix can be entered as Isotropic, Diagonal, or
Symmetric. For Isotropic the same mass is used in all the diagonal elements of the mass
matrix.
TABLE 4-10: AVAILABLE MASS TYPES BASED ON GEOMETRIC ENTITY

MASS TYPE VARIABLE SI UNITS GEOMETRIC ENTITY LEVEL

Mass per unit volume pV kg/m3 domains

Mass per unit area pA kg/m2 domains, boundaries
Mass per unit length pL kg/m edges
Total mass m kg domains, boundaries, edges

FRAME ACCELERATION FORCES

Select the Exclude contribution check box to switch off the loads that can be caused by
the added mass when the frame is accelerated when using a Gravity or Rotating Frame
feature. The setting will also determine whether the node will contribute when
Computing Mass Properties.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Mass, Spring, and Damper>Added Mass (Domain, Boundary)
Solid Mechanics>Edges>Added Mass
Shell>Mass, Spring, and Damper>Added Mass (Boundary, Edge)
Plate>Mass, Spring, and Damper>Added Mass (Domain, Boundary)
Membrane>Mass, Spring, and Damper>Added Mass (Boundary, Edge)
Beam>Mass, Spring, and Damper>Added Mass (Edge)
Truss>Mass, Spring, and Damper>Added Mass (Edge)
Multibody Dynamics>Mass, Spring, and Damper>Added Mass (Domain)
Multibody Dynamics>Added Mass (Boundary)
Multibody Dynamics>Edges>Added Mass

Ribbon
Physics tab with Solid Mechanics selected:
Domains>Mass, Spring, and Damper>Added Mass
Boundaries>Mass, Spring, and Damper>Added Mass
Edges>Solid Mechanics>Added Mass

544 | CHAPTER 4: SOLID MECHANICS

Physics tab with Shell or Membrane selected:
Boundaries>Mass, Spring, and Damper>Added Mass
Edges>Mass, Spring, and Damper>Added Mass

Physics tab with Plate selected:

Domains>Mass, Spring, and Damper>Added Mass
Boundaries>Mass, Spring, and Damper>Added Mass

Physics tab with Beam or Truss selected:

Edges>Mass, Spring, and Damper>Added Mass

Physics tab with Multibody Dynamics selected:

Domains>Mass, Spring, and Damper>Added Mass
Boundaries>Multibody Dynamics>Added Mass
Edges>Multibody Dynamics>Added Mass

Periodic Condition
Use a Periodic Condition to prescribe that the displacements on two different sets of
boundaries with the same geometrical shape are related, as in a periodic structure.

Several different types of periodicity properties of the solution can be prescribed using
this boundary condition.

• The Continuity, Antiperiodicity, and User defined periodic conditions directly

prescribe relations between displacements and can be used for any type of study.
• The Floquet periodicity can be used for frequency domain problems with a spatial
periodicity of the geometry and solution. The modeled structure is typically a unit
cell of a repetitive structure.
• The Cyclic symmetry is a special case of a Floquet condition, intended for structures
which consist of a number of sectors which are identical when rotated around a
common axis, like in a fan.

The Floquet periodicity and Cyclic symmetry options are available only with some
COMSOL products (see http://www.comsol.com/products/specifications/).

The two sets of boundaries between which there is a periodicity condition are called
the source and destination respectively. It is not required to have the same mesh on
the source and destination, but the local accuracy of the solution at the boundaries will
be better if you use the same mesh.

THE SOLID MECHANICS INTERFACE | 545

 BOUNDARY SELECTION
Select both the source and destination boundaries.

The software automatically identifies the boundaries as either source boundaries or

destination boundaries. This works fine for cases like opposing parallel boundaries. In
more general cases, use the Destination Selection subnode to specify the boundaries
which constitute the destination. By default this node contains the selection that
COMSOL Multiphysics has identified.

In cases where the periodic boundary is split into several boundaries within the
geometry, it might be necessary to apply separate periodic conditions to each pair of
geometry boundaries for the matching to work properly.

PERIODICITY SETTINGS
With Type of periodicity you select the form of periodicity that your solution should
have.

• For Continuity the displacements on the destination are set equal to the
displacements on the source; u ( x d ) = u ( x s ) . If the source and destination
boundaries are rotated with respect to each other, a transformation is automatically
performed, so that corresponding displacement components are connected.
• For Antiperiodicity the displacements on the destination are set equal to the
displacements on the source with the sign reversed; u ( x d ) = – u ( x s ) . If the source
and destination boundaries are rotated with respect to each other, a transformation
is automatically performed, so that corresponding displacement components are
connected.
• For Floquet periodicity enter a k-vector for Floquet periodicity kF. This is the wave
number vector for the excitation.
• For Cyclic symmetry chose how to define the sector angle that the geometry
represents using Sector angle. If Automatic is selected, the program attempts to find
out how many full repetitions of the geometry there will be on a full revolution. If
User defined is selected, enter a value for the sector angle θS. In both cases, also enter
an Azimuthal mode number for the mode to be studied. It can vary from 0 to N/2,
where N is the total number of sectors on a full revolution.
• For User defined select the check box for any of the displacement components as
needed. Then for each selection, choose the Type of periodicity — Continuity or
Antiperiodicity. Each selected displacement component will be connected by
u i ( x d ) = u i ( x s ) or u i ( x d ) = – u i ( x s ) respectively. If the source and destination

546 | CHAPTER 4: SOLID MECHANICS

 boundaries are rotated with respect to each other, a transformation is automatically
performed, so that corresponding displacement components are connected.

CONSTRAINT SETTINGS
To display this section, click the Show button ( ) and select Advanced Physics Options.

In the COMSOL Multiphysics Reference Manual:

• Constraint Reaction Terms

• Weak Constraints
• Constraint Settings

ORIENTATION OF SOURCE
For information about the Orientation of Source section, see Orientation of Source and
Destination in the COMSOL Multiphysics Reference Manual.

Cyclic Symmetry and Floquet Periodic Conditions

Vibrations of an Impeller: Application Library path

Structural_Mechanics_Module/Dynamics_and_Vibration/impeller

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Connections>Periodic Condition
Solid Mechanics>Connections>Periodic Condition>Destination Selection

Ribbon
Physics tab with Solid Mechanics selected:
Boundaries >Connections>Periodic Condition

Physics tab with Periodic Condition node selected in the model tree:
Attributes>Destination Selection

THE SOLID MECHANICS INTERFACE | 547

 Low-Reflecting Boundary
Use the Low-Reflecting Boundary node to let waves pass out from the model without
reflection in time-dependent or frequency domain analysis. As a default, it takes
material data from the domain in an attempt to create a perfect impedance match for
both pressure waves and shear waves. It can be sensitive to the direction of the
incoming wave.

The Low-Reflecting Boundary node is only available with some COMSOL products (see
http://www.comsol.com/products/specifications/).

DAMPING
Select a Damping type — P and S waves (the default) or User defined. For User defined
enter values or expressions for the Mechanical impedance di. The defaults for all values
are 0.5*solid.rho*(solid.cp+solid.cs).

Low-Reflecting Boundary Condition

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Mass, Spring, and Damper>Low-Reflecting Boundary

Ribbon
Physics tab with Solid Mechanics selected:
Boundaries >Mass, Spring, and Damper>Low-Reflecting Boundary

Thin-Film Damping

The MEMS Module also includes the Thin-Film Flow interface, which,
when appropriately coupled with the Solid Mechanics interface, is
equivalent to using the Thin-Film Damping boundary condition.

Use the Thin-Film Damping node to apply boundary loads to the surface of a resonator
that result from squeeze-film or slide-film damping. Squeeze-film damping occurs
when a thin film of gas is “squeezed” between two parallel plates, one of which is in
motion normal to its surface. Slide-film damping occurs in similar situations except

548 | CHAPTER 4: SOLID MECHANICS

that the direction of motion of the moving plate is tangential to its surface. Examples
of common situations in which these types of damping are appropriate are parallel plate
capacitive and comb drive actuators for squeeze-film and slide-film damping,
respectively. The thin-film damping boundary condition can accommodate general
displacements of the moving plates specified by means of the Fluid-Film Properties
node.

The Thin-Film Damping node is only available with some COMSOL products (see
http://www.comsol.com/products/specifications/).

The Fluid-Film Properties node and the Border node are added by default. These
additional subnodes are available: Fluid-Film Properties, Border, Inlet, Outlet, Wall, and
Symmetry.

The Fluid-Film Properties, Inlet, Outlet, Border, Wall, and Symmetry

nodes are all described for The Thin-Film Flow Interfaces.

Also see Theory for the Thin-Film Flow Interfaces as the Thin-Film
Damping boundary condition is equivalent to an appropriately coupled
Thin-Film Flow interface — so its theory is described in full for that
physics interface.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Mass, Spring, and Damper>Thin-Film Damping

Ribbon
Physics tab with Solid Mechanics selected:
Boundaries>Mass, Spring, and Damper>Thin-Film Damping

Rigid Connector
The Rigid Connector is a boundary condition for modeling rigid regions and kinematic
constraints such as prescribed rigid rotations. The selected boundaries will move as a
single rigid object, irrespective of whether they are geometrically adjacent or not.

If the study step is geometrically nonlinear, the rigid connector takes finite rotations
into account. The feature is similar to the rigid connectors in the Beam and Shell
interfaces. Rigid connectors from Beam, Shell, and Solid Mechanics interfaces can be
attached to each other.

THE SOLID MECHANICS INTERFACE | 549

 You can add functionality to the rigid connector through the following subnodes:

• Applied Force (Rigid Connector) to apply a force in given point.

• Applied Moment (Rigid Connector) to apply a moment.
• Mass and Moment of Inertia (Rigid Connector) to add extra mass and moment of
inertia in a given point.
• Spring Foundation (Rigid Connector) to add a translational or rotational spring or
damper in a given point.

The Rigid Connector node is only available with some COMSOL products (see http:/
/www.comsol.com/products/specifications/).

In versions prior to 4.4, it was also possible to add a Rigid Domain subnode
to the Rigid Connector. This node still appears and functions when you
open an old model where it was used, but it is not possible add a new one.

The functionality has been superseded by the more powerful concept of a

Rigid Domain as a material model.

For information about the old functionality, see the documentation for
version 4.3b or earlier.

COORDINATE SYSTEM SELECTION

The Global coordinate system is selected by default. The Coordinate system list contains
any additional coordinate systems that the model includes. Prescribed displacements
and rotations are specified along the axes of this coordinate system. It is also used for
defining the axis directions of the moment of inertia tensor of the Mass and Moment of
Inertia subnode.

PAIR SELECTION
If this node is selected from the Pairs menu, choose the pair to use. An identity pair has
to be created first. The rigid connector applies to the common part of the boundaries,
and makes the parts behave as if there were an infinitely stiff layer between them.

CENTER OF ROTATION
The center of rotation serves two purposes.

• If you prescribe the displacement of the rigid connector, this is the place where it is
fixed.
• Results are interpreted with respect to the center of rotation.

550 | CHAPTER 4: SOLID MECHANICS

Select a Center of rotation—Automatic, Centroid of selected entities, or User defined.

• For Automatic the center of rotation is at the geometrical center of the selected
geometrical objects.
• For Centroid of selected entities select an Entity level — Boundary, Edge, or Point. The
available choices depend on physics interface and geometrical dimension. The
center of rotation is located at the centroid of the selected entities, which do not
need to be related to rigid connector itself. As a special case, you can select a single
point, and thus use that point as center of rotation.

Once chosen, a default Center of Rotation: Boundary, Center of

Rotation: Edge, or Center of Rotation: Point subnode is automatically
added.

• For User defined, in the Global coordinates of center of rotation XC table enter
coordinates based on space dimension.

Select the Offset check box to add an optional offset vector to the definition of the
center of rotation. Enter values for the offset vector Xoffset.

The center of rotation used is the sum of the vector obtained from any of the input
methods and the offset vector.

X C = X C,input + X offset

PRESCRIBED DISPLACEMENT AT CENTER OF ROTATION

To define a prescribed displacement at the center of rotation for each space direction,
select one or several of the available check boxes then enter values or expressions for
the prescribed displacements. The direction coordinate names can vary depending on
the selected coordinate system.

• Prescribed in x direction u0x

• Prescribed in y direction u0y
• For 3D components: Prescribed in z direction u0z

THE SOLID MECHANICS INTERFACE | 551

 PRESCRIBED ROTATION AT CENTER OF ROTATION
Specify the rotation at the center of rotation. Select from the By list: Free (the default),
Constrained rotation, or Prescribed rotation at center of rotation.

For 2D components, the Constrained rotation and Prescribed rotation at

center of rotation is always about the z-axis, so no component selection is
necessary.

Constrained Rotation (3D Components)

For Constrained rotation select one or more of the available check boxes to enforce zero
rotation about the corresponding axis in the selected coordinate system:

• Constrain rotation about x-axis

• Constrain rotation about y-axis
• Constrain rotation about z-axis

Prescribed Rotation at Center of Rotation

For Prescribed rotation at center of rotation enter an Angle of rotation φ0. For 3D
components also enter an Axis of rotation Ω for the x, y, and z coordinates.

• You can add a Harmonic Perturbation subnode for specifying a

harmonic variation of the values of the prescribed displacements and
rotations in a frequency domain analysis of perturbation type.
• You can activate and deactivate the rigid connector by assigning it to a
constraint group. See Load Cases in the Structural Mechanics
Modeling chapter.
• You can assign the value of the prescribed displacement and rotation to
a load group. See Load Cases in the Structural Mechanics Modeling
chapter.

• Rigid Connector Theory

• Harmonic Perturbation
• Load Cases

552 | CHAPTER 4: SOLID MECHANICS

 Assembly with a Hinge: Application Library path
Structural_Mechanics_Module/Connectors_and_Mechanisms/hinge_assembly

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Connections>Rigid Connector
Multibody Dynamics>Rigid Connectior

Ribbon
Physics tab with Solid Mechanics selected:
Boundaries>Connections>Rigid Connector

Physics tab with Multibody Dynamics selected:

Boundaries>Multibody Dynamics>Rigid Connector

Center of Rotation Nodes

For each of the following physics features, a subnode is automatically added as
indicated.

• Rigid Connector
• Initial Values (Rigid Domain)
• Prescribed Displacement/Rotation

CENTER OF ROTATION: BOUNDARY

Use the Center of Rotation: Boundary subnode to select a set of boundaries whose
centroid represents the center of rotation.

CENTER OF ROTATION: EDGE

Use the Center of Rotation: Edge subnode to select a set of edges whose centroid
represents the center of rotation.

CENTER OF ROTA TION : PO INT

Use the Center of Rotation: Point subnode to select a set of points whose centroid
represents the center of rotation.

THE SOLID MECHANICS INTERFACE | 553

 LOCATION IN USER INTERFACE
These nodes are automatically added when Centroid of selected entities is selected in
the parent node. They cannot be added or removed manually.

Thermal Expansion (Rigid Connector)

Add the Thermal Expansion subnode to prescribe a deformation of the rigid connector
caused by changes in temperature. This makes it possible to reduce stresses caused by
the rigid connector being rigid, while there are thermal deformations in the flexible
body to which it is attached.

The thermal strain depends on the coefficient of thermal expansion α, the temperature
T, and the strain-free reference temperature Tref as

ε th = α ( T – T ref )

THERMAL EXPANSION PROPERTIES

Specify the thermal properties that define the thermal strain in the rigid connector.

Select Inherit from domain to take the thermal expansion data from the domain to
which it is attached. This should only be used when:

• The temperature and the thermal expansion coefficient do not have a spatial
variation.
• The virtual material in the rigid connector has the same thermal expansion as the
domain itself.

When Inherit from domain is not selected, enter:

• A value or expression for Temperature T, specifying the temperature distribution of

the rigid connector. Any spatial variation must be an explicit function of the material
frame coordinates. It is not possible to use a computed temperature distribution.
• The Coefficient of thermal expansion α. As a default, values From material are used.
This requires that a material has been assigned to the boundaries of the rigid
connector selection.
For User defined select Isotropic, Diagonal or Symmetric to enter one or more
components for a general coefficient of thermal expansion tensor α. When a

554 | CHAPTER 4: SOLID MECHANICS

 non-isotropic coefficient of thermal expansion is used, the axis orientations are given
by the coordinate system selected in the Coordinate System Selection section.
• A value or expression for the Strain reference temperature Tref which is the
temperature at which there are no thermal displacements in the rigid connector..

• Constraints and Thermal Expansion in the Structural Mechanics

Modeling chapter.
• Thermal Expansion of Constraints in the Structural Mechanics Theory
chapter.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Rigid connector>Thermal Expansion
Multibody Dynamics>Rigid connector>Thermal Expansion

Ribbon
Physics tab with Rigid connector node selected in the model tree:
Attributes>Thermal Expansion

Applied Force (Rigid Connector)

Use the Applied Force to apply a force to the rigid connector. The force can act at an
arbitrary position in space, and if it is not located at the center of rotation, there it is
also a moment contribution. The directions of the force is fixed in space and does not
follow the rotation of the rigid connector.

COORDINATE SYSTEM SELECTION

Select a Coordinate system for specifying the directions of the force.

LOCATION
Select an option from the list: Center of rotation or User defined. This is the location
where the force is applied.

• For Center of rotation, the location of the load is taken as the center of rotation as
defined in the parent Rigid Connector node.
• For User defined, enter the Location (global coordinates), Xp, in the table.

THE SOLID MECHANICS INTERFACE | 555

 Select the Offset check box to add an optional offset vector to the definition of the
location. Enter values for the offset vector Xoffset.

The location used is the sum of the vector obtained from any of the input methods and
the offset vector.

X P = X P,input + X offset

APPLIED FORCE
Enter values or expressions for the components of the Applied force F. The direction
coordinate names can vary depending on the selected coordinate system.

• You can add the Phase subnode to specify the phase of this load in a
frequency domain analysis.
• You can specify this load to be a Harmonic Perturbation in a frequency
domain analysis.
• You can assign this load to a load group. See Load Cases in the
Structural Mechanics Modeling chapter.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Rigid Connector>Applied Force
Shell>Rigid Connector>Applied Force
Beam>Rigid Connector>Applied Force
Multibody Dynamics>Rigid Connector>Applied Force

Ribbon
Physics tab with Rigid Connector node selected in the model tree:
Attributes>Applied Force

Applied Moment (Rigid Connector)

Use the Applied Moment subnode to apply a moment at the center of rotation. The
direction of the moment is fixed in space and does not follow the rotation of the rigid
connector.

556 | CHAPTER 4: SOLID MECHANICS

COORDINATE SYSTEM SELECTION
This section is only available for 3D components. The Global coordinate system is
selected by default. The Coordinate system list contains any additional coordinate
systems that the model includes. Select a Coordinate system for specifying the directions
of the moment.

APPLIED MOMENT
Enter values or expressions for the components of the Applied moment M

• For 3D components, this is the x, y, and z components for M. The direction

coordinate names can vary depending on the selected coordinate system.
• For 2D components, this is for the applied moment in the z direction Mz.

• You can add the Phase subnode to specify the phase of this load in a
frequency domain analysis.
• You can specify this load to be a Harmonic Perturbation in a frequency
domain analysis.
• You can assign this load to a load group. See Load Cases in the
Structural Mechanics Modeling chapter.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Rigid Connector>Applied Moment
Shell>Rigid Connector>Applied Moment
Beam>Rigid Connector>Applied Moment
Multibody Dynamics>Rigid Connector>Applied Moment

Ribbon
Physics tab with Rigid Connector node selected in the model tree:
Attributes>Applied Moment

Mass and Moment of Inertia (Rigid Connector)

Use the Mass and Moment of Inertia subnode to add inertia properties to the rigid
connector for dynamic analysis.

THE SOLID MECHANICS INTERFACE | 557

 CENTER OF MASS
Here you specify the location of the center of mass for the contribution given in this
node. Select an option from the list: Center of rotation or User defined.

• For Center of rotation, the location of the load is taken as the center of rotation as
defined in the parent Rigid Connector node.
• For User defined, enter the Global coordinates of center of mass, Xm, in the table.

Select the Offset check box to add an optional offset vector to the definition of the
location. Enter values for the offset vector Xoffset.

The center of mass used is the sum of the vector obtained from any of the input
methods and the offset vector.

X m = X m,input + X offset

MASS AND MOMENT OF INERTIA

Enter values or expressions for the Mass m. Also enter values or expressions for the
Moment of inertia I. The axis directions of the moment of inertia tensor are given by
the coordinate system selection in the parent Rigid Connector node.

• For 3D components, select Isotropic (the default), Diagonal, or Symmetric and enter
one or more components for the tensor I.
• For 2D components, enter a value or expression for Iz.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Rigid Connector>Mass and Moment of Inertia
Shell>Rigid Connector>Mass and Moment of Inertia
Multibody Dynamics>Rigid Connector>Mass and Moment of Inertia

Ribbon
Physics tab with Rigid Connector node selected in the model tree:
Attributes>Mass and Moment of Inertia

Spring Foundation (Rigid Connector)

Add a Spring Foundation subnode to specify a spring or damper connecting the rigid
connector to a fixed ground. The spring can act at an arbitrary position in space. A
translational spring implicitly contributes also to the moment if it is not applied at the

558 | CHAPTER 4: SOLID MECHANICS

center of rotation of the rigid connector. The data that you specify is interpreted in the
selected coordinate system.

LOCATION
Select an option from the list: Center of rotation or User defined. This is the location
where the spring is attached.

• For Center of rotation, the spring will be attached to the center of rotation as defined
in the parent Rigid Connector node.
• For User defined, enter the Location (global coordinates), Xp, in the table.

Select the Offset check box to add an optional offset vector to the definition of the
location. Enter values for the offset vector Xoffset.

The location used is the sum of the vector obtained from any of the input methods and
the offset vector.

X P = X P,input + X offset

SPRING
Select a Spring type — Spring constant or Force as function of extension.

When Spring constant is selected, the translational spring matrix can be entered as
Isotropic, Diagonal, Symmetric, or Anisotropic. For Isotropic the same spring constant is
used in all the diagonal elements of the spring matrix.

When Force as function of extension is selected, enter the force vector Fs. It must be a
a function of the built-in variables describing the spring extension. The default value
indicates the correct variable name, for example solid.rig1.uspring1_spf1.

ROTATIONAL SPRING
Select a Spring type — Spring constant or Moment as function of rotation.

When Spring constant is selected, the rotational spring matrix can be entered as
Isotropic, Diagonal, Symmetric, or Anisotropic. For Isotropic the same spring constant is
used in all the diagonal elements of the spring matrix.

THE SOLID MECHANICS INTERFACE | 559

 When Moment as function of rotation is selected, enter the moment vector Ms. It must
be a a function of the built-in variables describing the spring extension. The default
value indicates the correct variable name, for example solid.rig1.thspring1_spf1.

In 2D the Spring constant is only a single scalar, representing the stiffness

for rotation around the out-of-plane direction. Similarly, the Moment as
function of rotation is a single scalar.

LOSS FACTOR DAMPING

From the Loss factor type list, select Scalar (Same for all components) or Individual
components.

• For Scalar (Same for all components) enter a single Loss factor for spring ηu,s, which
is used to multiply all values of the spring matrix or spring force vector.
• For Individual components select Isotropic, Diagonal, Symmetric, or Anisotropic, then
enter values or expressions in the table for the Loss factor for spring ηu,k or ηf based
on space dimension. The loss factors act on the corresponding components of the
spring matrix or spring force vector. If you select Isotropic, the effect is the same as
when you selecting Diagonal and enter the same value for all diagonal elements.

ROTATIONAL LOSS FACTOR DAMPING

All settings in the Rotational Loss Factor Damping section are analogous to the
corresponding settings in the Loss Factor section. In 2D, only one scalar loss factor,
corresponding to rotation around the out-of-plane axis, is given.

VISCOUS DAMPING
Select Isotropic, Diagonal, Symmetric, or Anisotropic, then enter values or expressions for
the damping constants du in the table. If you select Isotropic, the effect is the same as
when you selecting Diagonal and enter the same value for all diagonal elements.

VISCOUS ROTATIONAL DAMPING

All settings in the Viscous Rotational Damping section are analogous to the
corresponding settings in the Viscous Damping section. In 2D, only one scalar viscous
damping, corresponding to rotation around the out-of-plane axis, is given.

560 | CHAPTER 4: SOLID MECHANICS

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Rigid Connector>Spring Foundation
Multibody Dynamics>Rigid Connector>Spring Foundation
Shell>Rigid Connector>Spring Foundation
Beam>Rigid Connector>Spring Foundation

Ribbon
Physics tab with Rigid Connector node selected in the model tree:
Attributes>Spring Foundation

Attachment
The Attachment node is used to define a set of boundaries on a flexible domain which
can be used to connect it with other components through a joint in the Multibody
Dynamics interface. All the selected boundaries behave as if they were connected by a
common rigid body.

Attachments can be added to boundaries in a Multibody Dynamics or a Solid

Mechanics interface, to edges in a Shell interface, or to points in a Beam interface. This
makes it possible to use a joint in the Multibody Dynamics interface for connecting
parts modeled in different physics interfaces.

The Attachment node is available with the addition of the Multibody

Dynamics Module. The Attachments is included in the Multibody
Dynamics Module User’s Guide.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Connections>Attachment

Ribbon
Physics tab with Solid Mechanics selected:
Boundaries>Connections>Attachment

THE SOLID MECHANICS INTERFACE | 561

 Thermal Expansion (Attachment)
Add the Thermal Expansion subnode to prescribe a deformation of the attachment
caused by changes in temperature. This makes it possible to reduce stresses caused by
the attachment being rigid, while there are thermal deformations in the flexible body
to which it is attached.

The thermal strain depends on the coefficient of thermal expansion α, the temperature
T, and the strain-free reference temperature Tref as

ε th = α ( T – T ref )

THERMAL EXPANSION PROPERTIES

Specify the thermal properties that define the thermal strain in the attachment.

Select Inherit from domain to take the thermal expansion data from the domain to
which it is attached. This should only be used when:

• The temperature and the thermal expansion coefficient do not have a spatial
variation.
• The virtual material in the attachment has the same thermal expansion as the
domain itself.

When Inherit from domain is not selected, enter:

• A value or expression for Temperature T, specifying the temperature distribution of

the attachment. Any spatial variation must be an explicit function of the material
frame coordinates. It is not possible to use a computed temperature distribution.
• The Coefficient of thermal expansion α. As a default, values From material are used.
This requires that a material has been assigned to the boundaries of the attachment
selection.
For User defined select Isotropic, Diagonal or Symmetric to enter one or more
components for a general coefficient of thermal expansion tensor α. When a

562 | CHAPTER 4: SOLID MECHANICS

 non-isotropic coefficient of thermal expansion is used, the axis orientations are given
by the coordinate system selected in the Coordinate System Selection section.
• A value or expression for the Strain reference temperature Tref which is the
temperature at which there are no thermal displacements in the attachment.

• Constraints and Thermal Expansion in the Structural Mechanics

Modeling chapter.
• Thermal Expansion of Constraints in the Structural Mechanics Theory
chapter.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Attachment>Thermal Expansion
Multibody Dynamics>Attachment>Thermal Expansion

Ribbon
Physics tab with Attachment node selected in the model tree:
Attributes>Thermal Expansion

Beam Connection

The Beam Connection node is obsolete, and has been superseded by the
Solid-Beam Connection multiphysics coupling. It cannot be added in
version 5.3 and later, but may be present in models created by earlier
versions of the software.

You are advised to update your model to use the new multiphysics
coupling instead, since the Beam Connection node will be removed in
future versions.

In 2D, a solid can be connected to a beam by adding a Beam Connection node in the
Solid Mechanics interface and a Solid Connection node in the Beam interface. The
connection can either be from a point on the beam to a boundary on the solid, or
between two boundaries. The first case is intended for modeling a transition from a
beam to a solid where beam theory assumptions are valid on both sides of the
connection. The second case is for adding a beam on top of a solid as stiffener or
cladding.

THE SOLID MECHANICS INTERFACE | 563

 The Beam Connection node is only available with some COMSOL products (see http:/
/www.comsol.com/products/specifications/).

• For more information about coupling different element types, see

Coupling Techniques.
• For details about the formulation of this coupling, see Connection
Between Shells and Solids

BEAM CONNECTION
Select a Connected entity — Beam point or Beam edge.

When Beam point is selected, select the name of the corresponding Solid Connection
node defined at the point level in the Beam interface to specify the connected parts.

Select an option from the Connected area defined by list: Section height (the default),
Selected boundaries, or Distance from beam axis. This parameter determines how much
of the selected solid boundaries are actually connected to the beam. The default is that
a distance from the beam point having the size of the half the beam section height in
both directions is connected. Using Selected boundaries connects the entire selected
boundaries to the beam. If you select Distance from beam axis, enter a Distance d. This
is used instead of the section height for defining the connection distance.

For Beam edge select the name of the corresponding Solid Connection node defined at
the boundary level in the Beam interface to specify the connected parts.

Select a Distance evaluation: Section height (the default), Geometrical distance, or User
defined. This parameter determines how the coupling treats the determination of the
connection distance. The default is that a distance equal to half the section height is
used. If you select Geometrical distance, the connection distance is computed from the
geometrical distance between the solid boundary and the beam boundary. For User
defined enter a Distance d. This defines the connection distance.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Connections>Beam Connection

Ribbon
Physics tab with Solid Mechanics selected:
Boundaries>Connections>Beam Connection

564 | CHAPTER 4: SOLID MECHANICS

Shell Connection

The Shell Connection node is obsolete, and has been superseded by the
Solid-Shell Connection multiphysics coupling. It cannot be added in
version 5.3 and later, but may be present in models created by earlier
versions of the software.

You are advised to update your model to use the new multiphysics
coupling instead, since the Shell Connection node will be removed in future
versions.

In 3D, a solid can be connected to a shell by adding a Shell Connection node in the Solid
Mechanics interface and a Solid Connection node in the Shell interface. The
connection can either be from an edge of the shell to a boundary on the solid, or
between two boundaries. The first case is intended for modeling a transition from a
shell to a solid where shell assumptions are valid on both sides of the connection. The
second case is for adding a shell on top of a solid.

The Shell Connection node is only available with some COMSOL products (see http:/
/www.comsol.com/products/specifications/).

• For more information about coupling different element types, see

Coupling Techniques.
• For details about the formulation of this coupling, see Connection
Between Shells and Solids

SHELL CONNECTION
Select a Connected entity — Shell edge or Shell boundary.

When Shell edge is selected, select the name of the corresponding Solid Connection node
defined at the edge level in the Shell interface to specify the connected parts.

Select an option from the Connected area defined by list: Shell thickness (the default),
Selected boundaries, or Distance from shell midsurface. This parameter determines how
much of the selected solid boundaries are connected to the shell. The default is that a
distance from the shell edge (having the size of the half the shell thickness in both
directions) is connected. Using Selected boundaries connects the entire selected
boundaries to the shell. If you select Distance from shell midsurface, enter a Distance d.
This is used instead of the shell thickness for defining the connection distance.

THE SOLID MECHANICS INTERFACE | 565

 For Shell boundary select the name of the corresponding Solid Connection node defined
at the boundary level in the Shell interface to specify the connected parts.

Select a Boundary type — Shared (the default) or Parallel. For Parallel, select a Distance
evaluation — Shell properties (the default), Geometrical distance, or User defined. This
parameter determines how the coupling treats the determination of the connection
distance. The default is that a distance equal to half the shell thickness is used. If you
select Geometrical distance, the connection distance is computed from the geometrical
distance between the solid boundary and the shell boundary. For User defined enter a
Distance d. This defines the connection distance.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Connections>Shell Connection

Ribbon
Physics tab with Solid Mechanics selected:
Boundaries>Connections>Shell Connection

Contact
In the Contact node, you define the mechanical and numerical properties for a set of
contact pairs in a contact analysis. Use it for modeling structural contact and
multiphysics contact. In the latter case, you will also need to add corresponding pair
conditions in the other participating physics interfaces.

If friction is to be included in the modeling, add a Friction subnode.

Do not select the same contact pair in more than one Contact node. Doing so may
produce error messages or unpredictable results.

566 | CHAPTER 4: SOLID MECHANICS

The Contact node is only available with some COMSOL products (see http://
www.comsol.com/products/specifications/).

In order to specify contact conditions, one or more Contact Pair nodes

must be available in the Definitions branch.

If you have several interfaces with displacement degrees of freedom in

your model, only the last physics interface in the model tree can contain
contact physics features since that is the physics interface in which the
displacements controls the spatial frame.

When a contact feature is present in your model, all studies are

geometrically nonlinear. The Include geometric nonlinearity check box on
the study step Settings window is selected and cannot be cleared.

A Contact node will override all previous nodes in the Model Tree sharing
the same boundary selections. If you want to add a load (such as the
pressure of a surrounding fluid), the best way of doing that is to select a
Boundary Load from the Fallback Features of the Contact node. Such a load
will then act only on the portions of the boundaries not being in contact.

If an ordinary Boundary Load node is added after the Contact node in the
Model Tree, it will contribute to the load on the boundaries, but without
taking the contact state into account.

• Contact Modeling
• Contact Analysis Theory

CONTACT PRESSURE METHOD

Select the algorithm used for computing the contact in the normal direction,
Augmented Lagrangian (default) or Penalty.

Enter a value for Characteristic stiffness Echar. The default is solid.Eequ, the
equivalent Young’s modulus as defined by most materials. The characteristic stiffness
(stored in a variable named solid.<contact_tag>.E_char) is used in expressions for
the default penalty factors for both the Augmented Lagrangian and Penalty methods. It
should be representative for the stiffness of the destination domain material in a

THE SOLID MECHANICS INTERFACE | 567

 direction normal to the boundary. You may for example need to adjust the default
value in for the following cases:

• The material is strongly anisotropic. The default value is based on an average in all
directions.
• The material is highly nonlinear. The default value is based on the stiffness at zero
strain.
• The variable solid.Eequ is not defined by the material. This is the case for some
user defined materials.

PENALTY FACTOR
If Contact Pressure Method is Augmented Lagrangian, select the type of Penalty factor
control — Preset (default), Manual tuning, or User defined. The settings give access to
an increasing level of detailed control of the penalty factor.

In the augmented Lagrangian method, the penalty factor controls how

“hard” the interface surface is during the iterations, but it does not affect
the converged result. You can consider the penalty factor as a spring giving
a resisting force if the boundaries in the contact pair have an interference.
A larger value gives faster, but less stable, convergence to the state where
there is no interference. A too large value can create convergence
difficulties, particularly if the interference is large

In the penalty method, the penalty factor is the actual stiffness of a spring
inserted between the boundaries in the contact pair.

If you select Preset, you have the choices to select Tuned for as Stability (default) and
Speed. If the contact boundaries move towards each other, so that large interferences
can be expected in the initial iterations, then is Stability is the better choice. In many
models, where the contact state does not change much, using Speed gives significant
performance improvements.

The Penalty factor control selection Manual tuning gives you access to a number of
detailed settings for the penalty factor.

Enter a Penalty factor multiplier. The default value is 1. Increasing this factor gives a
higher penalty factor. From Use relaxation, select Always (default), Never, or Conditional.
When using relaxation, the penalty factor is decreased during the first iterations in each
parameter or time step.

568 | CHAPTER 4: SOLID MECHANICS

If relaxation is used, enter the Initial Relaxation Factor. The default is 0.005. This factor
multiplies the penalty factor in the first iteration. Enter Number of iterations with
relaxation. The default value is 4. The penalty factor is gradually increased up to its full
value, which is used in the iteration after the one where the specified number of
iterations with relaxation have been reached.

If Use relaxation is set to Conditional, enter a Suppression criterion. The default is 0,

which means that the relaxation is used for all parameter or time steps. This is a
Boolean expression which, when fulfilled, suppresses the relaxation. If you, for
example, know that your problem needs relaxation only during the initial phase of the
solution, you could enter en expression like load_parameter>0.1. It is also possible
to use expressions based on the solution, for example
abs(solid.cnt1.gap)<0.05*h. This expression is true when the gap or interference
is small compared to the mesh size. It could be taken as an indication that the contact
problem is almost converged, and thus not in need of any relaxation of the penalty
factor.

Using the Penalty factor control selection User defined gives you the possibility to enter
an explicit expression for penalty factor. Edit or use the default Contact pressure penalty
factor pn.The default value is
min(1e-3*5^niterCMP,1)*solid.<contact_tag>.E_char/solid.hmin_dst.
The default value causes the penalty factor to be increased during the iterations and
takes material stiffness and element size at the contact surface into account. The
variable solid.<contact_tag>.E_char is the characteristic stiffness as defined in the
Contact Pressure Method section, and hmin_dst is the minimum element size on the
destination.

When Contact Pressure Method is Penalty, enter the Contact pressure penalty factor pn.
The default value is solid.<contact_tag>.E_char/solid.hmin_dst. Click to
select Offset penalty function if you want the contact pressure to be nonzero when the
gap is zero. In that case, also enter a value for the Contact pressure at zero gap T0.

TR I G G E R C U T B A C K
This section is only available when Contact Pressure Method is Augmented Lagrangian.
Select the Trigger cutback check box to enable additional control over solver cutbacks
in Time Dependent study or Stationary study with a Parametric continuation solver.
Enter a logical expression for Cutback criterion. When this expression evaluates to a
nonzero value, the iterations are immediately terminated, and the solver tries to use a
smaller value of the time or parameter. You can use this setting to avoid that the solver
spends many iterations trying to recover from an unphysical state. As an example, if

THE SOLID MECHANICS INTERFACE | 569

 you know that the destination boundary is confined so that it does not move more
than 1 mm anywhere, an entry like solid.disp > 5[mm] can be useful, since it is
unlikely that a displacement of this size should appear during a successful iteration
history.

In the COMSOL Multiphysics Reference Manual:

• Time Dependent and Stationary

• About the Time-Dependent Solver and About the Parametric Solver

OFFSET AND ADJUSTMENT

Enter a value or expression for Contact surface offset from geometric destination surface
doffset,d. The offset is subtracted from the gap in the normal direction of the
destination surface.

Enter a value or expression for Contact surface offset from geometric source surface
doffset,s. The offset is subtracted from the gap in the normal direction of the source
surface.

Select Force zero initial gap to compensate for any difference caused by irregularities in
geometry or mesh when the two contacting boundaries should exactly touch each
other in the initial state. The gap is adjusted to zero before any offset is added.

Use the offset properties to adjust initial clearances (negative values) or

interference fits (positive values) without having to change the geometry.
These properties are also useful for studying the effects of geometrical
tolerance when the structure is still modeled using its nominal size.

When combined with Force zero initial gap, the offset will be exact in the
sense that it is not affected by mesh irregularities.

The adjustment made by Force zero initial gap does not move any nodes
in the mesh. Effectively it adds an extra hidden offset, which compensates
for any initial distance between source and destination boundaries.

Offset and Adjustment

570 | CHAPTER 4: SOLID MECHANICS

IN IT IA L VA LUES
This section is only available when Contact Pressure Method is Augmented Lagrangian.
Enter an initial value for the Contact pressure Tn.

If not all the contacting parts are fully constrained, it is important to

supply a value here in order to avoid singular problems.

Speed up the convergence by supplying a guess of the correct order of

magnitude of the contact pressure.

DISCRETIZATION
This section is only available when Contact Pressure Method is Augmented Lagrangian.
To display this section, click the Show button ( ) and select Discretization. Select a
shape-function order for Contact pressure — Linear (the default), Quadratic, Cubic,
Quartic, or (in 2D) Quintic. This setting should usually not be changed. Selecting
anything else than Linear requires that the solver sequence is modified manually
because the lumped solver is then no longer optimal for the contact pressure update.

ADVANCED
To display this section, click the Show button ( ) and select Advanced Physics Options.

When analyzing a multiphysics contact problem, the state of being in contact or not
will be passed from the contact analysis in the structural mechanics interface to other
participating physics interfaces. For numerical reasons, the gap value will however not
be exactly zero even when the boundaries are in contact. A certain small positive value
of the gap will thus be considered as being in contact. The default Multiphysics contact
tolerance is Automatic. If you want to explicitly specify the limit of the gap considered
as being in contact, select Manual, and enter the Contact tolerance Δcontact.

Select Add contact status to solver log to get printout about the changes in the solver
log window. Doing this will add extra dependent variables used for tracking the contact
state on the destination boundaries.

If you have more than one contact pair selected in the Contact node, and the
augmented Lagrangian method is used, the check box Group contact variables in solver
per pair will be shown. When selected, a new Lumped Step will be generated in the
segregated solver for each contact pair. This will not affect the solution, but will give
more granular output in the convergence plots.

THE SOLID MECHANICS INTERFACE | 571

 Select Store energy variables to get access to

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Pairs>Contact
Multibody Dynamics>Pairs>Contact

Ribbon
Physics tab with Solid Mechanics or Multibody Dynamics selected:
Pairs>Contact

Friction
Using the Friction subnode, you can add friction to a Contact node.

The selection of the Friction node is the same as that of its parent Contact node. If you
add more than one Friction node under the same parent, the last one will override all
the preceding nodes. If Friction and Adhesion are present under the same Contact node,
the friction settings will be ignored once the adhesion criterion is fulfilled.

FRICTION
Select a Friction model — Static Coulomb friction or Exponential dynamic Coulomb
friction. The latter can only be used in a time dependent analysis, since the value of the
frictional coefficient depends on the slip velocity. Enter the following data:

• Static frictional coefficient μstat to give the static coefficient of friction.

• Cohesion sliding resistance Tcohe. Supply a traction which must be overcome before
sliding can occur.
• Maximum tangential traction Tt,max. When the tangential traction exceeds this value,
slip will occur, independent of the normal pressure. The default expression is Inf,
indicating that no limit on the tangential traction is active.
• When Exponential dynamic Coulomb friction is selected, also enter the Dynamic
frictional coefficient μdyn and the Friction decay coefficient αdcf.

TAN G EN TI AL FO RC E M E TH OD
Select the algorithm used for computing the tangential forces and sliding criteria,
Augmented Lagrangian or Penalty. The method can be selected independently of the
corresponding Contact Pressure Method setting in the parent Contact node.

572 | CHAPTER 4: SOLID MECHANICS

PENA LT Y FACT OR
If Tangential Force Method is Augmented Lagrangian, select the type of Penalty factor
control — From parent (default), Preset, Manual tuning, or User defined. The settings
give access to an increasing level of detailed control of the penalty factor for the sliding.

In the default case the settings are taken from the Contact node to which the current
Friction node is a child. For a description of the settings, see the documentation of
Penalty Factor under Contact. The only difference is that the default value of the User
defined penalty factor is
min(1e-3*5^niterCMP,1)*solid.<contact_tag>.E_char/3/solid.hmin_dst.

If Tangential Force Method is Penalty, select the type of Penalty factor control — From
parent or User defined. The choice From parent is only available if Contact Pressure
Method is set to Penalty in the parent Contact node. In that case, the settings are taken
from the Contact node to which the current Friction node is a child. When User defined
is selected, enter Tangential pressure force factor pt. The default value is
solid.<contact_tag>.E_char/3/solid.hmin_dst.

IN IT IA L VA LUES
If Tangential Force Method is Augmented Lagrangian, enter values or expressions for the
components of the initial force acting on the destination surface as Friction force Tt.

To determine whether friction effects are active when starting the solution or not,
select the Previous contact state — Not in contact (the default) or In contact.

For In contact enter values or expressions for the Previous mapped source coordinates
xm, old. These serve as initial values to compute the tangential slip. The default value
is (X, Y, Z) and indicates that the contacting boundaries are perfectly coincident in the
initial state. The mapped source coordinates are defined as the location on the source
boundary where it is hit by a certain point on the destination boundary.

The entries in this section should be given as components in the material

frame.

DISCRETIZATION
This section can only be displayed if Tangential Force Method is Augmented Lagrangian.
To display it, click the Show button ( ) and select Discretization. Select a
shape-function order for Friction force — Linear (the default), Quadratic, Cubic, Quartic,
or (in 2D) Quintic. This setting should usually not be changed. Selecting anything else

THE SOLID MECHANICS INTERFACE | 573

 than Linear requires that the solver sequence is modified manually because the lumped
solver is then no longer optimal for the update of the tangential forces.

ADVANCED
To display this section, click the Show button ( ) and select Advanced Physics Options.

For numerical reasons, the gap value will not be exactly zero even when the boundaries
are in contact. A certain small positive value of the gap must thus be used to determine
that the boundaries touch each other. so that friction forces can be introduced. Select
a Friction detection — Automatic or Manual. For Manual enter an absolute value for the
Friction detection tolerance Δcontact, which is the gap when the friction becomes active.

You can also request that the total accumulated slip distance it stored by selecting the
Store accumulated slip check-box.

The contact tolerance is the gap distance between the surfaces at which
friction starts to act. The default value is 10−6 times smaller than the
overall size of the model.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Contact>Friction
Multibody Dynamics>Contact>Friction

Ribbon
Physics tab with Contact selected in the Model Builder tree:
Atributes>Friction

Adhesion
Using the Adhesion subnode, you can add adhesion and decohesion properties to a
contact pair. This functionality requires that the contact is modeled using the penalty
method, and that no offset is used in the penalty stiffness function.

The selection of the Adhesion node is the same as that of its parent Contact node. If you
add more than one Adhesion node under the same parent, the last one will override all

574 | CHAPTER 4: SOLID MECHANICS

the preceding nodes. If Adhesion and Friction are present under the same Contact node,
the friction settings will be ignored once the adhesion criterion is fulfilled.

Adhesion is only available in the context menu and ribbon if

• Contact Pressure Method is set to Penalty in the Contact settings.

• Offset penalty function is not selected in the Contact settings.

An existing Adhesion node will be disabled if any of these settings are

changed later.

ADHESIVE ACTIVATION
Select an Activation criterion, to describe the onset of the adhesion between the source
and destination boundaries.

When the criterion is Pressure, enter the minimum contact pressure pn0, at which
adhesion is initiated.

When the criterion is Gap, enter the distance between the source and destination
boundaries δ0, at which adhesion is initiated. A negative value indicates that there must
be a certain overlap before adhesion is activated.

When the criterion is User Defined, enter a boolean expression such that when it
evaluates to true, adhesion is initiated.

When the criterion is Always active, the source and destination boundaries are in always
in adhesive contact, unless broken by decohesion.

ADHESIVE STIFFNESS
In this section, you specify the stiffness of the adhesive layer in the normal and
tangential directions.

Select how to specify the Normal stiffness of the adhesive layer. The default is to use
From contact penalty factor, in which case the stiffness is the same as the one given as
Contact pressure penalty factor in the settings of the parent Contact node.

To use another stiffness, select User defined, and enter the stiffness in the normal
direction Kp explicitly. This value is used in tension only; in compression the penalty
factor is always used.

Select Shear stiffness defined using to be either Normal to shear ratio or Adhesive
Poisson’s ratio.

THE SOLID MECHANICS INTERFACE | 575

 For Normal to shear ratio, enter the ratio between the shear stiffness and the normal
stiffness n τ .

For Adhesive Poisson’s ratio, enter Poisson’s ratio for the adhesive layer, v, explicitly.

DECOHESION
Select a Traction separation law. The default is No separation, in which case no
decohesion will occur.

For Linear separation, Polynomial separation, or Multilinear separation enter:

• Tensile strength σIc. This is the peak stress in pure tension.

• Shear strength τIIc. This is the peak stress in pure shear.
• Tensile energy release rate GIc. This is the energy released during the whole
decohesion process, in pure tension.
• Shear energy release rate GIIc. This is the energy released during the whole
decohesion process, in pure shear.

For Linear separation or Polynomial separation, select the Failure criterion to be either
Power law or Benzeggagh-Kenane. In either case, enter the Mode mixity exponent η.

For Multilinear separation enter the Shape factor λ.

• Adhesion and Decohesion, and Adhesion in the Structural Mechanics

Modeling chapter.
• Adhesion and Decohesion in the Structural Mechanics Theory chapter.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Contact>Adhesion
Multibody Dynamics>Contact>Adhesion

Ribbon
Physics tab with Contact selected in the Model Builder tree:
Atributes>Adhesion

Bolt Pre-Tension
Use the Bolt Pre-Tension node to define the pre-stress force in pre-tensioned bolts. It
is available for 3D models only.

576 | CHAPTER 4: SOLID MECHANICS

A default Bolt Selection subnode is automatically added. Use one such node to select
each individual bolt. Additional Bolt Selection subnodes are available from the context
menu (right-click the parent node) or from the Physics toolbar, Attributes menu.

The Bolt Pre-Tension node is only available with some COMSOL products (see http:/
/www.comsol.com/products/specifications/).

B O L T P RE - TE N S I O N
Select a Pre-tension type — Pre-tension force (the default) or Pre-tension stress.

• For Pre-tension force enter a value or expression for Fp, the pre-tension force in the
bolt.
• For Pre-tension stress enter a value or expression for σp, the pre-tension stress in the
bolt. The pre-tension force is computed through multiplication by the actual area
of each selected bolt.

• Pre-tensioned Bolts in the Structural Mechanics Modeling chapter for

an complete description of how to model pre-tensioned bolts.
• Bolt Pre-Tension Study for information about the Bolt Pre-Tension
study type.

THE SOLID MECHANICS INTERFACE | 577

 SOLVER SUGGESTION

This section is present only in models created in versions prior to 5.3, in

which the new Bolt Pre-Tension study type was introduced. The purpose is
to maintain backwards compatibility for older models.

Select Solve in bolt pre-tension study only check box to make the bolt
pre-tension degrees of freedom behave as in a version 5.3 model, that is
being solved for only in a study step of the Bolt Pre-Tension type. By
default, the check box is cleared, and then the bolt pre-tension degrees of
freedom are solved for in any type of study step, unless you explicitly
suppress that.

The settings of the Solve in bolt pre-tension study only check box only
affects new study sequences being generated. Existing study sequences
will keep the current state for the bolt pre-tension degrees of freedom.

Once you select the Solve in bolt pre-tension study only check box, the
Solver Suggestion section will be hidden. Thus, it is not possible to clear
the check box again. The degrees of freedom created under this Bolt
Pre-Tension node from now on assumes the version 5.3 behavior.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Bolt Pre-Tension

Ribbon
Physics tab with Solid Mechanics selected in the Model Builder tree:
Global>Bolt Pre-Tension

Bolt Selection
The Bolt Selection subnode is automatically added as a default node to the Bolt
Pre-Tension node. It is used for selecting the bolts. One Bolt Selection node is required
for each bolt.

BOUNDARY SELECTION
From the Selection list, choose the boundaries to define a cross section of a single bolt.
This cross section must be an internal boundary. It is the section where the stress in
the bolt is measured.

578 | CHAPTER 4: SOLID MECHANICS

BOLT SELECTION
Enter a Bolt label. The default is Bolt_1. The label is used for identification during
postprocessing.

SYMMETRY DETECTION
If a bolt is cut by a symmetry plane, the force in the modeled part is only half of the
force in the real bolt. When Automatic symmetry detection is selected, this is
compensated for. The program will then automatically detect when a bolt is located in
a symmetry plane, and apply corrections.

• If the input of the pre-tension load in the parent node is by Pre-tension force, the
given force is interpreted as force for the whole bolt.
• All results are given for the whole bolt.

When Automatic symmetry detection is cleared, the inputs and outputs are based on the
modeled cross section area.

As default, Automatic symmetry detection is selected. When opening

models created in versions prior to 5.3, the check box is cleared in order
to be compatible with the previous behavior.

Pre-tensioned Bolts in the Structural Mechanics Modeling chapter.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Bolt Pre-Tension>Bolt Selection

Ribbon
Physics tab with Bolt Pre-Tension selected in the Model Builder tree:
Attributes>Bolt Selection

Phase
You can add a Phase subnode to nodes which define a load in order to prescribe the
phase angle in a frequency domain analysis.

THE SOLID MECHANICS INTERFACE | 579

 For modeling the frequency response the physics interface splits the harmonic load
into two parameters:

• The amplitude, F, which is specified in the node for the load.

• The phase ( φ ), which is specified in the Phase subnode.

Together these define a harmonic load, for which the amplitude and phase shift can
vary with the excitation frequency, f:

F freq = F ( f ) ⋅ cos ( 2πft + φ )

PHASE
Enter the components of Load phase φ in radians (for a pressure the load phase φ is a
scalar value). Add [deg] to a phase value to specify it using degrees.

Typically the load magnitude is a real scalar value. If the load specified in
the parent feature contains a phase (using a complex-valued expression),
the software adds the phase from the Phase node to the phase already
included in the load.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Body Load>Phase
Solid Mechanics>Boundary Load>Phase
Solid Mechanics>Edge Load>Phase
Solid Mechanics>Point Load>Phase
Solid Mechanics>Rigid Connector>Applied Force>Phase
Solid Mechanics>Rigid Connector>Applied Moment>Phase
Solid Mechanics>Rigid Domain>Applied Force>Phase
Solid Mechanics>Rigid Domain>Applied Moment>Phase
Membrane>Body Load>Phase
Membrane>Face Load>Phase
Membrane>Edge Load>Phase
Membrane>Point Load>Phase
Multibody Dynamics>Body Load>Phase
Multibody Dynamics>Boundary Load>Phase
Multibody Dynamics>Edge Load>Phase
Multibody Dynamics>Point Load>Phase

580 | CHAPTER 4: SOLID MECHANICS

Multibody Dynamics>Rigid Connector>Applied Force>Phase
Multibody Dynamics>Rigid Connector>Applied Moment>Phase
Multibody Dynamics>Rigid Domain>Applied Force>Phase
Multibody Dynamics>Rigid Domain>Applied Moment>Phase

Ribbon
Physics tab with Body Load, Boundary Load, Face Load, Edge Load, Point Load, Applied
Force, or Applied Moment selected:
Attributes>Phase

Harmonic Perturbation
Use the Harmonic Perturbation subnode to specify the harmonic part of non-zero
prescribed displacements, rotations, velocities, or accelerations. This node is used if the
study step contains frequency response of a perturbation type.

The settings are the same as in the parent Prescribed Displacement, Prescribed
Velocity, Prescribed Acceleration, Prescribed Displacement/Rotation or Rigid
Connector node. Only degrees of freedom selected as prescribed in the parent node
can be assigned a value.

See Harmonic Perturbation in the Structural Mechanics Modeling

chapter.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Prescribed Displacement>Harmonic Perturbation
Solid Mechanics>Prescribed Velocity>Harmonic Perturbation
Solid Mechanics>Prescribed Acceleration>Harmonic Perturbation
Solid Mechanics>Rigid Domain>Prescribed Displacement/Rotation>Harmonic Perturbation
Solid Mechanics>Rigid Connector>Harmonic Perturbation
Membrane>Prescribed Displacement>Harmonic Perturbation
Membrane>Prescribed Velocity>Harmonic Perturbation
Membrane>Prescribed Acceleration>Harmonic Perturbation
Truss>Prescribed Displacement>Harmonic Perturbation
Truss>Prescribed Velocity>Harmonic Perturbation
Truss>Prescribed Acceleration>Harmonic Perturbation
Multibody Dynamics>Prescribed Displacement>Harmonic Perturbation

THE SOLID MECHANICS INTERFACE | 581

 Multibody Dynamics>Rigid Domain>Prescribed Displacement/Rotation>
Harmonic Perturbation
Multibody Dynamics>Rigid Connector>Harmonic Perturbation

Ribbon
Physics tab with Prescribed Displacement, Prescribed Velocity, Prescribed Acceleration,
Prescribed Displacement/Rotation, or Rigid Connector node selected in the model tree:
Attributes>Harmonic Perturbation

Stress Linearization
Use the Stress Linearization node to define a line (stress classification line; SCL) along
which a linearization of the stress state will be made. A number of result variables are
created for each SCL. Add one Stress Linearization node for each SCL along which you
want to compute linearized stresses.

When a Stress Linearization node has been added, you do not need to compute a new
solution. It is sufficient to perform an Update Solution to make the variables for the new
SCL available.

The line selection can consist of several edges, but they are assumed to form a straight
line from one free boundary to another. For 2D and 2D Axisymmetric geometries, the
only input is the line selection. For 3D, you also need to specify the orientation of the
local coordinate system in which the linearized stresses are represented.

The Stress Linearization node is only available with some COMSOL products (see
http://www.comsol.com/products/specifications/).

SECOND AXIS ORIENTATION

This section is present only in 3D. Here you specify the orientation of the local
coordinate system in which the components of the linearized stress tensor are
represented. The first direction is always along the SCL, and you indicate the second
direction. The third direction is automatically formed by being orthogonal to the line
and the second direction.

Select Reference point or Orientation vector. When a reference point is used, the second
local direction will be in the plane formed by the SCL and the selected point. When an
orientation vector is used, the second local direction will be directed in the plane

582 | CHAPTER 4: SOLID MECHANICS

formed by the SCL and the selected orientation. The orientation is always adjusted so
that it is orthogonal to the SCL.

• For Reference point, select Defined by — Point or Coordinates. When Point is used,
select the reference point in the Second Axis Orientation Reference Point section.
When Coordinates is used, enter the coordinates for the reference point in the
Reference point defining local 2 direction table.
• For Orientation vector select Defined by — Edge or Coordinates. When Edge is used,
select an edge in the geometry as orientation vector in the Second Axis Orientation
Reference Vector section. When Coordinates is used, enter the orientation vector
manually in the Orientation vector defining local 2 direction table.

• Stress Linearization in the Structural Mechanics Modeling Chapter.

• Stress Linearization in the Structural Mechanics Theory Chapter.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Variables>Stress Linearization

Ribbon
Physics tab with Solid Mechanics selected in the Model Builder tree:
Global>Variables>Stress Linearization

Important Variables in the Solid Mechanics Interface

In each node in the physics interface, a number of variables are defined. In Table 4-11,
you can find a list of variables that you as a user may want to use when, for example,
creating your own variables and expressions. This list is far from exhaustive but
contains some of the most commonly used variables. To see all variables defined by a
certain node in the Model Builder tree, enable the Equation View, and examine the
contents in the Equation View node under each physics node in the Model Builder tree.

The scopes of the variables in the table serve as examples only and assume the default
first instance of a certain feature in the Model Builder tree. The actual scope will
depend on the tag of a certain node in the model tree.

In the case of vectors and tensors, only a single component is shown. The actual indices
can depend on the dimensionality of the problem and names of coordinate system axis
directions.

THE SOLID MECHANICS INTERFACE | 583

 Variables created in contact analysis are summarized in Dependent Variables in Contact
Analysis and Important Contact Variables

In the COMSOL Multiphysics Reference Manual:

• For more information about the Equation View node, see Equation
View
• For more information about variable scopes, see Variable Naming
Convention and Namespace.
• For more information about tags in the model tree, see Displaying
Node Names, Tags, and Types in the Model Builder.

TABLE 4-11: IMPORTANT VARIABLES IN THE SOLID MECHANICS INTERFACE

VARIABLE DESCRIPTION DEFINED IN COMMENTS

solid.disp Total displacement All materials1

solid.vel Velocity magnitude All materials1
solid.u_tX Velocity, X All materials1
component
solid.acc Acceleration All materials1
magnitude
solid.u_ttX Acceleration, X All materials1
component
solid.eXX Strain tensor, XX All materials1 Total
component (Green-Lagrange)
strain in global
directions. See
also Deformation
Measures.
solid.el11 Strain tensor, local All materials1 Total strain in the
coordinate system, 11 local directions.
component
solid.eel11 Elastic strain tensor, All materials1
local coordinate
system, 11
component

584 | CHAPTER 4: SOLID MECHANICS

TABLE 4-11: IMPORTANT VARIABLES IN THE SOLID MECHANICS INTERFACE

VARIABLE DESCRIPTION DEFINED IN COMMENTS

solid.eeldev11 Deviatoric elastic All materials1

strain tensor, local
coordinate system, 11
component
solid.eiel11 Inelastic strain tensor, All materials1 See also Inelastic
local coordinate Strain
system, 11 Contributions.
component
solid.FdxX Deformation gradient, All materials1 See also
xX component Deformation
Measures.
solid.Fdlx1 Deformation gradient, All materials1
local, x1 component
solid.J Volume ratio All materials1 Total relative
volume change.
solid.Cel11 Elastic Cauchy-Green All materials1
tensor, local
coordinate system, 11
component
solid.Jel Elastic volume ratio All materials1 Elastic relative
volume change.
solid.Ldxy Rate of strain tensor, All materials1 See also Strain
xy component Rate and Spin.
solid.Lwxy Spin tensor, xy All materials1 See also Strain
component Rate and Spin.
solid.SXX Second All materials1 See also Defining
Piola-Kirchhoff stress, Stress.
XX component
solid.Sl11 Second All materials1
Piola-Kirchhoff stress,
local coordinate
system, 11
component
solid.SdevX Deviatoric second All material1 See also
Piola-Kirchhoff stress, Invariants of the
X component Stress Tensor.
solid.PxX First Piola-Kirchhoff All materials1 See also Defining
stress, xX component Stress.

THE SOLID MECHANICS INTERFACE | 585

 TABLE 4-11: IMPORTANT VARIABLES IN THE SOLID MECHANICS INTERFACE

VARIABLE DESCRIPTION DEFINED IN COMMENTS

solid.sxy Stress tensor, xy All materials1 Cauchy stress,

component See also Defining
Stress.
solid.sdevx Deviatoric stress All materials1 Deviatoric
tensor, x component Cauchy stress,
See also
Invariants of the
Stress Tensor.
solid.mises von Mises stress All materials1 See also
Invariants of the
Stress Tensor.
solid.tresca Tresca stress All materials1 See also
Invariants of the
Stress Tensor.
solid.I1s First principal All materials1 See also
invariant of stress Invariants of the
Stress Tensor.
solid.I2s Second principal All materials1 See also
invariant of stress Invariants of the
Stress Tensor.
solid.I3s Third principal All materials1 See also
invariant of stress Invariants of the
Stress Tensor.
solid.II2s Second invariant of All materials1 See also
stress deviator Invariants of the
Stress Tensor.
solid.II3s Third invariant of All materials1 See also
stress deviator Invariants of the
Stress Tensor.
solid.RFx Reaction force, x All materials1
component
solid.RMx Reaction moment, y All materials1
component
solid.Tax Traction (force/area), All materials1
x component
1 - All materials: Linear Elastic Material, Nonlinear Elastic Material, Hyperelastic Material,
Piezoelectric Material, Magnetostrictive Material, Cam-Clay Material.

586 | CHAPTER 4: SOLID MECHANICS

 5

Shell and Plate

This chapter describes the Shell and Plate interfaces, which are found under the
Structural Mechanics branch ( ) when adding a physics interface.

In this chapter:

• Theory for Shell and Plate Interfaces

• The Shell and Plate Interfaces

587
 Theory for Shell and Plate Interfaces
The theory for the Shell and Plate interfaces is discussed in this section:

• About Shells and Plates

• Theory Background for the Shell and Plate Interfaces
• Reference for the Shell Interface

About Shells and Plates

A shell is a thin-walled structure in 3D where a simple form is assumed for the variation
of the displacement through the thickness. Using this approximation, it is possible to
develop a model for the deformation that is more similar to a 2D plane stress condition
than to a full 3D state.

For a shell to give accurate results it is important that the structure can
really be described as being thin-walled. When modeling using shells you
should in general model the faces at the midplane of the real geometry.
You can use the offset setting if the midsurface of the shell does not
coincide with the boundary of your geometry.

Plates are similar to shells but act in a single plane and usually with only out-of-plane
loads. The plate and shell elements in COMSOL Multiphysics are based on the same
formulation. The Plate interface for 2D models is a specialization of the Shell interface.
In the following, the text fully describes the Shell interface, and the Plate interface is
mentioned only where there are nontrivial differences.

A Shell interface can be active either on free surfaces embedded in 3D or on the

boundary of a solid 3D object. In the latter case, it can be used to model a
reinforcement on the surface of a 3D solid. A Plate interface can only be active on
domains in 2D.

To describe a shell, you provide its thickness, a possible offset, and the elastic material
properties.

The element used for the shell interface is of Mindlin-Reissner type, which means that
transverse shear deformation is accounted for. It can thus also be used for rather thick
shells. It has an MITC formulation where MITC means mixed interpolation of

588 | CHAPTER 5: SHELL AND PLATE

tensorial components. A general description of this element family is in Ref. 1.

The dependent variables are the displacements u, v, and w in the global x, y, and z
directions, and the displacements of the shell normals ax, ay, and az in the global x, y,
and z directions.

Figure 5-1: The degrees of freedom in the shell interface. N is the normal vector in the
original configuration and n is the normal in the deformed state.

The degrees of freedom represent the displacements on the reference surface. The
reference surface is the boundary where the shell element mesh is created. If an offset
property is used, the reference surface differs from the physical shell midsurface. The
displacement vector on the midsurface, u, can be expressed as

u = uR + ζ0 a

where uR is the displacement on the reference surface (the displacement degrees of

freedom) and ζ0 is the offset. The rotational displacement a is the same on both
midsurface and reference surface.

THEORY FOR SHELL AND PLATE INTERFACES | 589

 For input and output, the Shell interface to a large extent replaces the displacements
of the shell normals by the more customary rotations θx, θy, and θz about the global
axes. For a geometrically linear analysis, the relation between normal displacement and
rotation vector is simple: a = θ × n where n is the unit normal of the shell.

For a standard plate analysis only three degrees of freedom are needed: the
out-of-plane displacement w and the displacements of the shell normals ax and ay. It
is also possible to activate all six degrees of freedom, so that any type of analysis of a
shell initially positioned in the xy-plane can be performed using the Plate interface.
Using six degrees of freedom is the default, but three degrees of freedom can be
selected instead for efficiency.

Also for plates, the rotations θx, θy (and possibly θz) are used to a large extent.

When six degrees of freedoms are used in the Plate interface, there must
be enough constraints to suppress any in-plane rigid body motions.

In the Shell interface, the coordinates are usually denoted with lower case
letters (x, y, z). If a Solid Mechanics or Membrane interface is present in
the same model, then it becomes necessary to make a difference between
the material frame and the spatial frame (Material and Spatial
Coordinates). In this case the coordinates in the Shell interface changes
to (X, Y, Z).

Theory Background for the Shell and Plate Interfaces

Several topics are discussed in this section:

• Geometry and Deformation • External Loads

• Strains • Stress and Strain Calculations
• Offset • Local Coordinate Systems
• Rotation Representation • Connection Between Shells and
• The MITC Shell Formulation Solids

• Initial Values and Prescribed Values • Connection Between Shells and

Beams
• Symmetry and Antisymmetry
Boundary Conditions

590 | CHAPTER 5: SHELL AND PLATE

GEOMETRY AND DEFORMATION
Let r be the undeformed shell midsurface position, ξi be element local (possibly
nonorthogonal) coordinates with origin in the shell midsurface, and n be the normal
to the undeformed midsurface. The thickness of the shell is d, which can vary over the
element. The local coordinates ξ1 and ξ2 follow the midsurface, and ξ3 is the
coordinate in the normal direction. The normal coordinate has a value of −d/2 on the
bottom side of the element, and +d/2 on the top side.

The position of the deformed midsurface is r + u, and the normal after deformation is
n + a. To keep the normal a unit vector requires that

n+a = 1 (5-1)

In a geometrically linear analysis Equation 5-1 is replaced by the simpler linearized

form

n⋅a = 0

since the formulation in that case assumes that

a «1

The vectors r, u, n, and a are interpolated by the nth-order Lagrange basis functions.
The basic assumption is that the position of a point within the shell after deformation
has a linear dependence of the thickness coordinate, and thus is

1 2 3 1 2 1 2 3 1 2 1 2
x ( ξ , ξ , ξ ) = r (ξ ,ξ ) + u (ξ ,ξ ) + ξ ( n (ξ ,ξ ) + a (ξ ,ξ ) )

The superscripts indicate contravariant indices, while subscripts indicate covariant

indices.

STRAINS
The in-plane Green-Lagrange strain in the local covariant components can then be
written as

ε αβ = --- ∂ ( r + u + ξ ( n + a ) ) ⋅ ∂ ( r + u + ξ ( n + a ) ) –
1 3 3
2 ∂ ξα ∂ξ
β

∂ r ξ3 ∂ r ξ3 3 3 2
( + n) ⋅ ( + n ) = γ αβ + ξ χ αβ + ( ξ ) κ αβ
α β
∂ξ ∂ξ

The indices α and β range from 1 to 2. The transverse shear strains in local covariant
components are

THEORY FOR SHELL AND PLATE INTERFACES | 591

 ε 3α = ε α3 =

--- ∂ ( r + u + ξ ( n + a ) ) ⋅ ( n + a ) – ∂ ( r + ξ n ) ⋅ n = ζ α + ξ ω α
1 3 3 3
2 ∂ ξα ∂ξ
α

The constitutive relation for the shell elements is a plane stress assumption, as is
customary in shell theory. The strain component in the normal direction ε33 is thus
irrelevant. The different parts of the strain tensors above can be written out as

1 ∂u ∂r ∂r ∂u ∂u ∂u
γ αβ = --- --------α- -------β- + --------α- -------β- + --------α- -------β-
2 ∂ξ ∂ξ ∂ξ ∂ξ ∂ξ ∂ξ

1 ∂r ∂a ∂a ∂r ∂u ∂n ∂n ∂u ∂u ∂a ∂a ∂u
χ αβ = --- --------α- -------β- + --------α- -------β- + --------α- -------β- + --------α- -------β- + --------α- -------β- + --------α- -------β-
2 ∂ξ ∂ξ ∂ξ ∂ξ ∂ξ ∂ξ ∂ξ ∂ξ ∂ξ ∂ξ ∂ξ ∂ξ

1 ∂a ∂n ∂n ∂a ∂a ∂a
κ αβ = --- --------α- -------β- + --------α- -------β- + --------α- -------β-
2 ∂ξ ∂ξ ∂ξ ∂ξ ∂ξ ∂ξ

1 ∂r ∂u ∂u
ζ α = --- --------α- ⋅ a + --------α- ⋅ n + --------α- ⋅ a
2 ∂ξ ∂ξ ∂ξ

1 ∂n ∂a ∂a
ω α = --- --------α- ⋅ a + --------α- ⋅ n + --------α- ⋅ a
2 ∂ξ ∂ξ ∂ξ

In a geometrically linear analysis, the nonlinear terms (products between u, a, and

their derivatives) disappear. In all study types, the contributions from the parts καβ and
ωα are ignored. They are small unless the element has an extremely high ratio between
thickness and radius of curvature, in which case the errors from using shell theory are
large anyway.

Note that ζ α here is a strain component, not to be confused with the

local coordinate in the normal direction or offset.

OFFSET
It is possible to model a shell with a midsurface that is not located at the meshed
surface but at a certain offset from it. The offset is assumed to occur along the normal
of the shell surface. In this case,

1 2 1 2 1 2
r (ξ ,ξ ) = r R (ξ ,ξ ) + ζ o n (ξ ,ξ )

592 | CHAPTER 5: SHELL AND PLATE

where rR is the position of the meshed reference surface and ζo is the offset distance.

Since all geometric derivatives are computed at the mesh on the reference surface, the
following type of expressions are used when evaluating the strains:

∂r- ∂r R ∂n
--------
α
= --------α- + ζ o --------α-
∂ξ ∂ξ ∂ξ

The degrees of freedom are located on the reference surface.

All loads and boundary conditions are assumed to be applied at the midsurface, so a
force acting in the plane of the shell does not cause any bending action if there is an
offset.

The numerical integration of the element is performed over the reference surface. If
the shell is curved, the area of the actual midsurface and the reference surface differ.
This is compensated for by multiplying the weak expressions with an area scale factor,
defined as

∂r ∂r
1
× 2
∂ξ ∂ξ
ASF = ----------------------------
∂r R ∂r R
1
× 2
∂ξ ∂ξ

Any expressions depending on the coordinates are evaluated on the shell reference
surface.

ROTATION REPRESENTATION
In a geometrically linear analysis, a rotation vector is defined as

θ = n×a

In a geometrically nonlinear analysis, the rotation vector axis is defined by

n×a
e θ = -----------------
n×a

while the amplitude of the rotation vector is computed as

θ = acos ( 1 + n ⋅ a )

THEORY FOR SHELL AND PLATE INTERFACES | 593

 This representation is unique only for rotations up to 180 degrees, but since the
rotation vector representation is only an output convenience, it has no impact on the
analysis.

THE MITC SHELL FORMULATION

The MITC formulation (Ref. 1) does not take the strain components directly from the
shape functions of the element. Instead, meticulously selected interpolation functions
are selected for the individual strain components. The values of the interpolated strains
are then at selected points in the element tied to the value that would be computed
from the shape functions. The interpolation functions and tying points are specific to
each element shape and order.

Each contribution to the virtual work of the element is numerically integrated over the
reference surface while the integration in the thickness direction is performed
analytically. The computation of the strain energy from transverse shear deformations
uses a correction factor of 5/6 to compensate for the difference between the assumed
constant average shear strain and the true parabolic distribution.

In regions where the discretized surface is smooth (which is always the case for plates),
the normal of the shell surface is uniquely defined. When two or more shell elements
meet at an angle, each element must however keep its own normal direction. It is thus
not possible to have the same set of degrees of freedom for the displacement of the
normal in such a point. This is automatically handled by the program. The automatic
search for these fold lines compares the normals of all boundaries sharing an edge. If
the angle between the normals is larger than a certain angle (with a default of
3 degrees) it is considered as a fold line. For a fold line, the assumption is that the angle
between the shell normals remains constant. This gives

nj ⋅ nk = ( nj + aj ) ⋅ ( nk + ak )

or

nj ⋅ ak + aj ⋅ nk + aj ⋅ ak = 0 (5-2)

where the values or j and k range over the number of shells elements with different
normals. The third term in Equation 5-2 is relevant only in a large deformation analysis
because it is nonlinear. A special case occurs when two adjacent boundaries are parallel
but their normal vectors have opposite directions. In this case the special constraint

ak = –aj

594 | CHAPTER 5: SHELL AND PLATE

is applied along their common edge.

INTEGRATION
All volume integrals over a shell element are split into a surface integral, which is
performed numerically, and a thickness direction integral which is performed
analytically. It is thus not possible to enter data which depend on the thickness
direction. All material properties are evaluated at the reference surface. Formally this
can be written as

d⁄2

 f(ξ , ξ , ξ    f1 ( ξ , ξ
1 2 3 1 2 3 1 2
) dV = f 1 ( ξ , ξ )f 2 ( ξ ) dς dA = )F 2 ( d ) dA
V A –d ⁄ 2 A

All functions of ζ are assumed to be of the form ζn. Odd powers will integrate to zero,
so typically the cross thickness integration will give factors d (for the case n=0) and
d3/12 (for the case n=2). The thickness d can be a function of the position.

IN IT IA L VA LUES AN D PRES CRIBED VAL UES

Because the normal vector displacements are quantities which are less intuitive than the
more customary nodal rotations, it is possible to specify the prescribed values in terms
of nodal rotations as well as in terms of the normal vector displacement. The
representation by normal vector direction is insensitive to whether the analysis is
geometrically nonlinear or not. The direction of the shell normal is prescribed in the
deformed state, n0. The prescribed values for the actual degrees of freedom, a0, are
internally computed as

n0
a 0 = --------- – n
n0

If the rotation vector input is used, and the analysis is geometrically linear, then

a 0 = Ω0 × n

where Ω0 is the vector of prescribed nodal rotations. This relation is fully defined only
when all three components of Ω0 are given. It is also possible to prescribe only one or
two of the components of Ω0, while leaving the remaining components free. Because
it has no relevance to prescribe the rotation about the normal direction of the shell, it
is only possible prescribe individual rotations in a shell local system. In this case, the
result becomes one or two constraint relations between the components of a0. The
directions are the edge local coordinate system where t1 is the tangent to the edge and

THEORY FOR SHELL AND PLATE INTERFACES | 595

 t2 is perpendicular and inward from the edge, in the plane of the shell. These
constraints are formulated as

t 2 ⋅ a 0 = – Ω 01

t 1 ⋅ a 0 = Ω 02

Here Ω0i is the prescribed rotation around the axis ti.

In a geometrically nonlinear analysis, it is not possible to prescribe individual elements

of the rotation vector. If only one or two components have been specified, the
remaining components are set to zero. The actual degrees of freedom are then
computed as

a 0 = ( R ( Ω 0 ) – I )n

where R is a standard rotation matrix, representing the finite rotation about the given
rotation vector.

Initial velocities are always given using an angular velocity vector ω as

·
a = ω × ( n + a0 )

SYMMETRY AND ANTISYMMETRY BOUNDARY CONDITIONS

It is possible to prescribe symmetry and antisymmetry boundary conditions. As a
default, they are expressed in a shell local coordinate system. If applied to a boundary,
the normal to the shell is assumed to be the normal to the symmetry or antisymmetry
plane. The conditions are

u⋅n = 0

a ⋅ t1 = 0

a ⋅ t2 = 0

for the symmetry case and

u ⋅ t1 = 0

u ⋅ t2 = 0

for the antisymmetry case. Here t1 and t2 are two perpendicular directions in the plane
of the shell.

596 | CHAPTER 5: SHELL AND PLATE

When applied to an edge, there is a local coordinate system where t1 is the tangent to
the edge, and t2 is perpendicular and in the plane of the shell. The assumption is then
that t2 is the normal to the symmetry or antisymmetry plane. The constraints are

u ⋅ t2 = 0

a ⋅ t2 = 0

for the symmetry case and

u⋅n = 0

u ⋅ t1 = 0

a ⋅ t1 = 0

for the antisymmetry case.

When symmetry or antisymmetry conditions are specified in a general coordinate

system with axis directions ei, i = 1, 2, 3 with e1 as the normal to the
symmetry/antisymmetry plane the constraints are

u ⋅ e1 = 0

a ⋅ ( e2 × n ) = 0

a ⋅ ( e3 × n ) = 0

or the symmetry case and

u ⋅ e2 = 0

u ⋅ e3 = 0

a ⋅ ( e1 × n ) = 0

for the antisymmetry case. Using a general coordinate system sometimes leads to
higher accuracy, since there is no element interpolation of the constraint directions
involved.

EXTERNAL LOADS
Contributions to the virtual work from the external load is of the form

THEORY FOR SHELL AND PLATE INTERFACES | 597

 u test ⋅ F + a test ⋅ [ M × ( n + a ) ]

where the forces (F) and moments (M) can be distributed over a boundary or an edge
or concentrated in a point. The contribution from the normal vector displacement a
is only included in a geometrically nonlinear analysis. Loads are always referred to the
midsurface of the element. In the special case of a follower load, defined by its pressure
p, the force intensity is

F = –p ( n + a )

For a follower load, the change in midsurface area is not taken into account, in order
to be consistent with the assumption that thickness changes are ignored.

STRESS AND STRAIN CALCULATIONS

The strains calculated in the element are, as described above, the covariant tensor
components. They have little significance for the user, and are internally transformed
to a Cartesian coordinate system. This system can be global or element local. The
stresses are computed by applying the constitutive law to the thus computed strain
tensor.

Each part of the covariant strain (γαβ, χαβ, ζα) is transformed separately. They
correspond to membrane, bending, and shear action, respectively, and it is thus
possible to separate the stresses from each of these actions. The membrane stress is
defined as

Ni
σ m = D [ γ – γi – α ( T m – T ref ) – β h ( c m – c mo,ref ) ] + ------
d

where D is the plane stress constitutive matrix, Ni are the initial membrane forces, and
γi the initial membrane strains. The influence of thermal strains is included through the
midsurface temperature Tm, and the hygroscopic swelling through the midsurface
moisture concentration, cm.The membrane stress can be considered as the stress at the
mid-surface, or as the average through the thickness.

The bending stress is defined as

Dd ΔT Δc mo 6M i
σ b = -------- χ – χ i – α -------- – β h ------------- + ----------
-
2 d d d
2

where χi is the initial value of the bending part of the strain tensor (actually: the
curvature), and Mi are the initial bending and twisting moments. ΔT is the
temperature difference between the top and bottom surface of the shell, and Δcmo is

598 | CHAPTER 5: SHELL AND PLATE

the difference in moisture concentration between the top and bottom. The bending
stress is the stress at the top surface of the shell if no membrane stress is present.

The average transverse shear stress is defined as

5 Qi
σ s = --- ⋅ 2G ( ζ – ζ i ) + ------
6 d

where G represent the transverse shear moduli, ζi is the initial average shear strain, and
Qi are the initial transverse shear forces. The correction factor 5/6 ensures that the
stresses are averaged so that they correspond to the ratio between shear force and
thickness. The corresponding strains ζ and ζi are averaged in an energy sense.

The actual in-plane stress at a certain level in the element is then

σ = σ m + zσ b

where z is a parameter ranging from −1 (bottom surface) to +1 (top surface). The

computation of the shear stress at a certain level in the element uses the following
analytical parabolic stress distribution:

2
3σ s ( 1 – z )
-----------------------------
2

The shell section forces (membrane forces, bending moments, and shear forces) are
computed from the stresses as

N = dσ m

2
d
M = ------ σ
6 b

Q = dσ s

LOCAL COORDINATE SYSTEMS

Boundaries
Many quantities for a shell can best be interpreted in a local coordinate system aligned
to the shell surface. Material data and initial stresses and stress results are always
represented in this local coordinate system. You specify the orientation of the local
directions in a Shell Local System node under the Linear Elastic Material.

THEORY FOR SHELL AND PLATE INTERFACES | 599

 The local system for stress output coincides with the orientations defined for the
material input. Stresses are also available transformed to the global coordinate system.

The definition of the local shell surface coordinate system is as follows:

1. The local z direction ezl is the positive normal of the shell surface.

2. The local x direction exl is the projection of the first direction in the material
coordinate system (ex1) on the shell surface

e x1 – ( e x1 ⋅ e zl )e zl
e xl = ----------------------------------------------------
e x1 – ( e x1 ⋅ e zl )e zl

This projection cannot be performed if ex1 is normal to the shell. In that case the
second axis ex2 of the material system instead defines exl using the same procedure.
Thus, if

e x1 ⋅ e zl > 0.99

then

e x2 – ( e x2 ⋅ e zl )e zl
e xl = ----------------------------------------------------
e x2 – ( e x2 ⋅ e zl )e zl

3. At last, the second in-plane direction is generated as

e yl = e zl × e xl

This procedure is followed irrespective of whether a global or a local coordinate system

defines the directions.

Note the following:

• When using an isotropic material, the only effect of selecting a local coordinate
system is that the definition directions of local stresses change.
• When defining orthotropic and anisotropic materials, local coordinate systems do
not need to be created so that they exactly follow the shell surface. It is sufficient
that the local system when projected as described above gives the intended in-plane
directions.
• For shells in the X-Y plane, and for plates, the global and local directions coincide
by default.

600 | CHAPTER 5: SHELL AND PLATE

• On curved shells, local stress components can become discontinuous if there is a
location where ex1 becomes perpendicular to the shell surface.
• In this section, every reference to stresses is equally valid for strains.

Local edge system

Many features, such as an edge load, allow input in an edge local coordinate system.
The orthogonal local edge coordinate system directions xl, yl, and zl are defined so
that:

• The first direction (xl) is along the edge. This direction can be visualized by
selecting the Show edge directions arrows check box in the View settings.
• The third direction (zl) is the same as the shell normal direction. The shell normal
direction can be visualized in the default plot named Undeformed geometry, or by
adding a Coordinate System Surface plot and selecting the Shell, Local System.
• The second direction (yl) is in the plane of the shell and orthogonal to the edge. It
is formed by the cross product of zl and xl; y l = z l × x l .

When an edge is shared between two or more boundaries, the directions may not
always be unique. It is then possible to use the control Face Defining the Local Edge
System to select from which boundary the normal direction zl should be picked. The
default is Use face with lowest number. If the geometry selection contains several edges,
the only available option is Use face with lowest number, since the list of adjacent
boundaries would then be different for each edge.

RESULTS EVALUATION
For visualization and results evaluation, predefined variables include all nonzero stress
and strain tensor components, principal stresses and principal strains, in-plane and
out-of-plane forces, moments, and von Mises and Tresca effective stresses. It is
possible to evaluate the stress and strain tensor components and effective stresses at an
arbitrary distance from the midsurface. The parameter zshell (variable name shell.z)
is found in, for example, the Parameters table of the Settings window of for example a
surface plot. Its can be set to a value from −1 (downside) to +1 (upside). A value of 0
means the midsurface of the shell. The default value is given in the Default
through-thickness result location section of the Shell interface.

Stresses and strains are available both in the global coordinate system and in the shell
local system as described in Local Coordinate Systems.

THEORY FOR SHELL AND PLATE INTERFACES | 601

 RIGID DOMAIN FOR SHELLS
The inertial properties mass (m) and moment of inertia tensor (I) of a rigid shell takes
the finite thickness into account. They are computed as:


m = ρ dV = ρd dA 
V A
1 1 1
  
X M = ----- ρX dV = ----- ρ ( r + ςn ) dV = ----- ρdr dA
m m m
V V A


T T
I = ( ( ( X – X M ) ⋅ ( X – X M ) )E 3 – ( X – X M ) ⋅ ( X – X M ) )ρ dV =
V

 ( ( ( r – XM )
T T
⋅ ( r – X M ) )E 3 – ( r – X M ) ⋅ ( r – X M ) )ρd dA +
A
3
T d
 ( E 3 – nn )ρ ------ dA
12
A

where E 3 and X M are the identity matrix and the center of mass of the rigid domain,
respectively. The last term in I is accounting for the finite thickness, and there the fact
T
that n ⋅ n = 1 has been taken into account.

CONNECTION BETWEEN SHELLS AND SOLIDS

This section describes the theory and assumptions behind the multiphysics couplings
Solid-Shell Connection and Solid-Beam Connection. Only the shell version of the
connection is described in detail, since the beam version is a direct specialization to 2D.

Solver Settings

There are three types of connections between a shell and a solid of interest:

• Type 1 connection: The shell connects to the solid in a thin region (having the same
thickness as the shell), so that shell theory is valid on both sides. This connection is
the most important from the application point of view and the most difficult to
create manually.

602 | CHAPTER 5: SHELL AND PLATE

• Type 2 connection: The tangent plane to the shell is perpendicular to the face of a
‘thick’ solid, in which case the physics of the connection can, at best, be
approximative.
• Type 3 connection: The shell acts as cladding on a solid.

The first two cases have similar physics and can be treated more or less as one case.
Usually, the shell should not connect to parts on the solid boundary further away than
what is represented by the shell thickness (or some other user-defined distance).

Shell Perpendicular to Solid

Figure 5-2 illustrates the first two cases. The shell edge can be part of the definition of
the solid but there is no assumption about that.

Figure 5-2: An example of a shell extending perpendicular to the solid boundary.

The connection of a solid to a shell is based on that shell theory is valid on both sides
of the connection. This can be divided into these assumptions:

• The dimension of the solid in the shell thickness direction is the same as the
thickness variable in the shell, shell.d.
• The midsurface of the shell to be connected to the solid is positioned at the
midsurface of the solid. The reference surface of the shell can be placed anywhere.
• The cut in the solid is orthogonal to the tangent of the shell (that is, the shell normal
vector is in the plane of the cut).
• Basic shell theory assumptions are valid, for example, the direct stress in the
transverse direction is negligible.

One basic shell theory assumption is actually not valid in practice: plane sections do
not always remain plane under deformation. A detailed analysis shows that if there is a
transverse shear force in the section, there must be a deviation from planarity to get the
correct shear strain distribution. This is more important as the shell grows thicker, but

THEORY FOR SHELL AND PLATE INTERFACES | 603

 without it, it is not possible to get a perfect connection. In Mindlin plate theory, shear
is related to the difference between rotation and the derivative of displacement, so that
plane sections remain plane, but no longer perpendicular to the midsurface. This gives
an average shear strain, while it is known from analytical solutions that the shear strain
has a parabolic distribution through the thickness.

Consider a cut where the local coordinates are defined as follows:

• x′ is the outward normal to the solid boundary

• z′ is along the shell normal
• y′ is the tangent to the shell edge, directed so that x′-y′-z′ form a right-hand system

Using the plane section assumption in the shell gives the displacements in the solid as:

u' sld (y',z') = u' sh (y',0) + z'a' x (y',0)

v' sld (y',z') = v' sh (y',0) + z'a' y (y',0)

where the subscripts sld and sh represent ‘solid’ and ‘shell’, respectively, and a′ is the
displacement of the shell normal, represented in the local directions.

The values on the solid boundaries should be interpreted as mapped using an extrusion
operator from the shell edge.

A simple connection for the transverse direction can be generated by

w' sld (y',z') = w' sh (y',0)

This connection, however, enforces a ‘plane strain condition’ in the solid, which is not
consistent with shell theory and which causes local unphysical stresses if Poisson’s ratio
is nonzero. This effect disappears within a few elements from the connection, and the
approximation can, in many situations, be acceptable. This constraint is enforced if
Method is set to Rigid in the multiphysics coupling.

A more accurate connection is derived in the following. The first approximation of the
stress state in a moderately curved shell is

604 | CHAPTER 5: SHELL AND PLATE

 
 σ x' = A 1 + B 1 z'
 σ y' = A 2 + B 2 z'

 σ z' = 0

 τ x'y' = B 3 z'

 2
 τ x'z' = C 1 ( 1 – z )

 τ y'z' = C 2 ( 1 – z )
2

2z'
where the relative thickness coordinate z = ------- has been introduced.
d
Assuming that Hooke’s law and plane stress conditions are valid, the transverse direct
strain can be computed as

υ
ε z' = – ---- ( σ x' + σ y' ) = K 1 (x',y') + 2K 2 (x',y')z'
E

The transverse displacement is then, after integration with respect to z′:

2
w' = K 0 (x',y') + K 1 (x',y')z' + K 2 (x',y')z'

Thus, in order to obtain a stress-free transverse displacement, this deformation must

be allowed. Note that the last term makes the solid thinner on one side of the
midsurface and thicker on the other. The constant term—that is, the displacement at
the midsurface of the solid—is known from the shell midsurface displacement. The
other two terms depend on the stress state and thus on derivatives of the displacement,
which are not readily available. The K1 term is caused by the membrane action and the
K2 term by the bending action.

For the transverse shear stress, Hooke’s law gives

∂u' ∂w' τ x'z' C1 2

γ x'z' = ------- + -------- = --------- = ------- ( 1 – z )
∂z' ∂x' G G

which can be reformulated as

∂------
u'- C1 2 ∂w' C1 2 ∂K 0 ∂K 1 ∂K 2 2
= ------- ( 1 – z ) – -------- = ------- ( 1 – z ) – ---------- – ---------- z' – ---------- z'
∂z' G ∂x' G ∂x' ∂x' ∂x'

Note that as the K1 term is related to membrane action, it cannot contribute to the
transverse shear stress. Since the derivative in the x′ direction cannot be controlled by

THEORY FOR SHELL AND PLATE INTERFACES | 605

 a condition on the boundary, it is necessary to make an assumption about u′(z′). An
integration with respect to z′ gives

3
u' = M 0 (x',y') + M 1 (x',y')z' + M 3 (x',y')z'

This shows that a third power of z′ is required in order to be able to represent the
correct shear strain contribution.

It is, however, not possible to directly determine the coefficients in front of the
additional terms, since they depend on the actual stress state. The idea is here to
introduce them as dependent variables in the problem. These variables are defined by
extra shape functions on the shell edge.

The constraints applied on the solid can then be written as

3
u' sld (y',z') = u' sh (y',0) + z'a' x (y',0) + q u (y',0) ( z – k u z )d
v' sld (y',z') = v' sh (y',0) + z'a' y (y',0)
2
w' sld (y',z') = w' sh (y',0) + q w1 (y',0)zd + q w2 (y',0) ( z – k w2 )d

Here qu, qw1, and qw2 are the new dependent variables defined on the shell edge. They
are dimensionless, due to the multiplication with the shell thickness, d. The constants
k u = 3 ⁄ 5 and k w2 = 1 ⁄ 5 are explained below. The variable qw1 is proportional to
the membrane axial strain, the variable qw2 is proportional to the bending strain, and
the variable qu is proportional to the transverse shear strain. Since these variable are
directly related to strains, the shape function order used is one order lower than for the
displacements.

If no extra equations defining qu, qw1, and qw2 are introduced, these variables try to
adapt to proper values through the reaction forces on the solid. The reaction force for
u′ is the traction σx′ and the reaction force for w′ is the traction τx′z′. When taking the
variation of the new dependent variables, these enforce the following constraints:

606 | CHAPTER 5: SHELL AND PLATE

 d
----
2

 (z
3
test ( q u ) – k u z )σ x' dz' = 0
–d
-------
2
d
----
2

test ( q w1 )  zτx'z' dz' = 0

–d
-------
2
d
----
2

 (z
2
test ( q w2 ) – k w2 )τ x'z' dz' = 0
–d
-------
2

The equation with qw1 is trivially fulfilled because the shear stress is an even function
of z′. Inserting the known stress distributions gives equations that can be solved for ku
and kw2.

The constraint expressions must now be formulated in global directions. As a start, the
constraints are written on vector form in local directions as

u' sld = u' sh + z'a' + q (5-3)

where

3
q u (y',0) ( z – k u z )
q = 0
2
q w1 (y',0)z + q w2 (y',0) ( z – k w2 )

The fact that a z' = 0 has been used when formulating Equation 5-3.

All coordinate directions are retrieved from the shell, because the normal to the solid
boundary is not necessarily constant.

The only coordinate value needed is actually z′. For the other two coordinates only the
direction is important. The coordinate in the normal direction can be computed as

z' = ( X sld – X sh ) ⋅ N sh

THEORY FOR SHELL AND PLATE INTERFACES | 607

 This definition of z′ assumes that the thickness of the solid does not change
significantly. Under geometric nonlinearity, the computation should be based on the
current geometry.

z' = ( x sld – x sh ) ⋅ n sh

The latter expression introduces additional nonlinearities in the model because it

depends on deformed position and deformed normal. Also, the position of the shell
midsurface with respect to the solid is actually part of the solution.

Let Φ be the matrix that transforms displacements from the global system to the local
system:

Φ = e x' e y' e z'

The expression for the constraints in global directions then becomes

T T T
u sld = Φ u' sld = Φ ( u' sh + z'a' + q ) = u sh + z'a + Φ q

The only transformation actually needed is thus the projection of the q vector. For a
linear case, the transformation can be written as

Φ = st l × N sh st l N sh

where N is the undeformed shell normal (shell.an) and tl is the shell edge tangent
(shell.tle). The coefficient s is either 1 or −1, and is selected so that the x′ direction
coincides with the outward normal of the solid.

For a geometrically nonlinear case, the corresponding deformed directions are used.

When an offset is used for the shell, it is assumed that the center of the connection is
at the actual shell midsurface.

Beam Perpendicular to Solid

This is the analogous case in 2D. Beam theory assumes that the stress in the
out-of-plane direction is zero. It is thus only physically sound to connect to a model
where the plane stress assumption is used. The derivation above still remains valid with
the following exceptions: the displacements in the local y’ direction are zero and the
tangent direction tl is replaced by the out-of-plane direction.

The shell thickness is replaced by twice the effective radius of the beam in the equations
defining the displacements.

608 | CHAPTER 5: SHELL AND PLATE

Shell Parallel to Solid
The case where the shell is parallel to the boundary of the solid exists in two versions
- Shared and Parallel.

In the shared case, the shell is modeled on a boundary which is a face of the solid. In
this case, it assumed that the names of the displacement degrees of freedom in the solid
and shell interfaces are not the same. If the same names are used, there is no need to
use a connection feature, since the coupling is then automatic. A shell offset can be
used to model an actual distance between the boundaries. For a layer ‘glued’ on the
solid, the offset would equal half the shell thickness.

In the parallel case, a separate boundary is used for modeling the shell. The distance
between the shell and the face of the solid is taken into account when setting up the
constraints, so that

u sld = u shl – ζa

where ζ is a distance from the solid to the shell. The right hand side is mapped from
the shell to the solid using an extrusion operator. The default is that ζ is half the shell
thickness, but you can also use the geometrical distance between the boundaries, or a
user defined distance.

CONNECTION BETWEEN SHELLS AND BEAMS

When connecting elements from the Shell interface with elements from the Beam
interface, the following must be noted:

• The rotational degrees of freedom have different definitions in the two interfaces.
The Beam interface uses a rotation vector representation, whereas the Shell interface
uses the normal vector displacements.
• In general, a shell element locally has three translational, but only two rotational,
degrees of freedom. The rotation around the normal is not part of the shell theory
and is constrained to zero. The corresponding rotational degree of freedom in a
beam should not thus not be connected to the shell.

You can create the appropriate couplings by adding a Solid-Shell Connection

multiphysics coupling. The theory of this connection is outlined below.

THEORY FOR SHELL AND PLATE INTERFACES | 609

 Beam Edge to Shell Edge
This coupling is intended for the common situation where beams are attached along a
plate to act as stiffeners. There are two variants of the coupling:

• The beam is modeled at an edge which is also an edge in the beam interface. This
case is called Shell and beam shared edges in the Shell-Beam Connection node. In
practice, the beam is usually placed on one side of the shell, and this offset plays an
important role in the stiffness of the combined section. The offset, d0, can be given
as a user input.
• The beam is modeled at a separate edge, representing the actual center line. This
case is called hell and beam parallel edges in the Shell-Beam Connection node, and the
closest geometrical distance between the edges directly gives d0. You do not need
to use the same mesh on both lines. Since the constraints are formed for the shell
edge, some parts of the beams could however become unconnected if the beam
elements are very short when compared to the shell element size.

The displacement at the centerline of the beam can then be written in therms of the
degrees of freedom in the shell as

u b = u s + ( d 0 ⋅ n )a

where n is the normal to the shell. The rotation vector in the beam can be expressed
in the shell degrees of freedom as

θb = n × a

For a geometrically linear case, the constraint

n⋅a = 0

is enforced by the shell interface for each shell boundary. This is why there are only two
active rotational degrees of freedom. To avoid propagating this constraint to the beam,
only the those components of the beam that act in the plane of the shell should be
constrained. This can be expressed as

( θb – n × a ) ⋅ ti i = 1, 2

which can be simplified to

θb ⋅ t1 + a ⋅ t2 = 0
θb ⋅ t2 – a ⋅ t1 = 0

610 | CHAPTER 5: SHELL AND PLATE

The constraints are actually formed on the shell edge, and the degrees of freedom are
taken from the beam using a General Extrusion operator which maps values from the
closest point on the beam to the shell.

The definitions of n, t1, and t2 may however be discontinuous over a shell edge. For
this reason, the constraint is formed using values from one boundary only if several
boundaries share the edge. Another complications arises when the edge is a fold line,
that is when the boundaries that meet do not have a common normal direction. On a
fold line all three rotational degrees of freedom do exist in the shell and should then
be connected to the corresponding degrees of freedom in the shell. In this case, also a
third rotational constraint is formed.

Beam Point to Shell Boundary

This coupling is intended for the case when the beam is not in the same plane as the
boundary modeled by shell theory. This case is called Shell boundaries to beam points
in the Shell-Beam Connection node. Physically, this can be seen as a beam with one end
welded to a plate. In order to get a correct stiffness representation of such a
connection, it is necessary that the beam is connected to an area of the shell which is
similar to the actual physical width of the beam. The connected area on the shell does
not have to fit a boundary in the geometry. It is however necessary that the mesh size
is such that there are at least three nodes within the connected area.

You can then select the region to connect using three different criteria.

The connected region is treated as rigid. The displacement of the shell is controlled by
the displacement and rotation of the beam endpoint through

us = ub + θb × ( Xs – Xb )

The coordinates of the shell are evaluated at the reference surface.

As it is only possible to constrain the in-plane rotations of the shell, the continuity in
rotation is projected onto the shell, giving

θb ⋅ t1 + a ⋅ t2 = 0
θb ⋅ t2 – a ⋅ t1 = 0

The rotation of the beam around the normal of the shell, which does not participate
in the rotation constraints, is indirectly connected through the displacement equation,
so it implicitly receives an appropriate stiffness.

THEORY FOR SHELL AND PLATE INTERFACES | 611

 Beam Point to Shell Edge
This case is identical to the previous case, with the only exception that the selection in
the Shell interface is an edge. The beam can have any orientation relative to the shell
edge. This case is called Shell edges to beam points in the Shell-Beam Connection node.

Reference for the Shell Interface

1. D. Chapelle and K.J. Bathe, The Finite Element Analysis of Shells—
Fundamentals, Springer-Verlag, Berlin Heidelberg, 2003.

612 | CHAPTER 5: SHELL AND PLATE

The Shell and Plate Interfaces
The Shell interface is for 3D models.

The Plate interface is for 2D models—domains are selected instead of

boundaries, and boundaries instead of edges. Otherwise the Settings
windows are similar to those for the Shell interface.

The Shell ( ) interface, found under the Structural Mechanics branch ( ) when
adding a physics interface, is used to model structural shells on 3D boundaries. Shells
are thin flat or curved structures, having significant bending stiffness. The interface
uses shell elements of the MITC type, which can be used for analyzing both thin
(Kirchhoff theory) and thick (Mindlin theory) shells. Geometric nonlinearity can be
taken into account.

The Plate ( ) interface, found under the Structural Mechanics branch ( ) when
adding a physics interface, provides the ability to model structural plates in 2D. Plates
are thin flat structures with significant bending stiffness, being loaded in a direction
out of the plane.

The Linear Elastic Material is the only available material model. It adds a linear elastic
equation for the displacements and has a Settings window to define the elastic material
properties.

When this interface is added, these default nodes are also added to the Model Builder—
Linear Elastic Material, Free (a boundary condition where edges are free, with no loads
or constraints), and Initial Values. Then, from the Physics toolbar, add other nodes that
implement, for example, boundary conditions. You can also right-click Shell or Plate to
select physics features from the context menu.

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is shell or plate.

THE SHELL AND PLATE INTERFACES | 613

 THICKNESS
Define the Thickness d by entering a value or expression in the field. The default is
0.01 m. Use the Change Thickness node to define a different thickness in parts of the
shell or plate. The thickness can be variable if an expression is used.

Offset Definition

This section is available for the Shell interface only. For theory, see Offset.

If the actual shell midsurface is not the boundary on which the mesh exists, it is
possible to prescribe an offset in the direction of the surface normal by using an offset
definition. The offset is defined as positive if the shell midsurface is displaced from the
meshed boundary in the direction of the positive shell normal.

Select an option from the Offset definition list—No offset (the default), Physical offset,
or Relative offset. The default No offset means that the modeled boundary coincides
with the shell midsurface.

• For Physical offset enter a value or expression in the zoffset field for the actual
distance from the meshed boundary to the shell midsurface.

Figure 5-3: Meshed boundary indicated in red. The vector n indicates the positive
orientation of the shell normal.

• For Relative offset enter a value or expression in the zrel_offset field for the offset as
the ratio between the offset distance and half the shell thickness. A value of +1 means
that the actual shell bottom surface is located on the meshed boundary, and a value
of −1 means that the shell top surface is located on the meshed boundary.

614 | CHAPTER 5: SHELL AND PLATE

Figure 5-4: Meshed boundary indicated in red.The vector n indicates the positive
orientation of the shell normal.

Use the Change Thickness node to define a different offset in parts of the
shell.

S T R U C T U R A L TR A N S I E N T B E H AV I O R
From the Structural transient behavior list, select Include inertial terms (the default) or
Quasi-static. Use Quasi-static to treat the dynamic behavior as quasi-static (with no mass
effects; that is, no second-order time derivatives). Selecting this option gives a more
efficient solution for problems where the variation in time is slow when compared to
the natural frequencies of the system. The default solver for the time stepping is
changed from Generalized alpha to BDF when Quasi-static is selected.

This is often the case when the time dependence exists only in some other physics, like
a transient heat transfer problem causing thermal strains.

REFERENCE POINT FOR MOMENT COMPUTATION

Enter the default coordinates for the Reference point for moment computation xref. The
resulting moments (applied or as reactions) are then computed relative to this
reference point. During the results and analysis stage, the coordinates can be changed
in the Parameters section in the result nodes.

FOLD-LINE LIMIT ANGLE

This section is available for the Shell interface only. Also see The MITC
Shell Formulation.

The fold-line limit angle α is the smallest angle between the normals of two boundaries
that makes their intersection to be treated as a fold line. The normal to the shell is
discontinuous along a fold-line. Enter a value or expression in the α field. The default

THE SHELL AND PLATE INTERFACES | 615

 value is 0.05 radians (approximately 3 degrees). The value must be larger than 0, and
less than π/2, but angles larger than a few degrees are not usually meaningful.

DEFAULT THROUGH-THICKNESS RESULT LOCATION

Enter a number between -1 and 1 for the Local z-coordinate [-1,1] for
thickness-dependent results Z. The value can be changed from −1 (downside) to +1
(upside). The default is +1. A value of 0 means the midsurface of the shell. This is the
default position for stress and strain evaluation during the results analysis. During the
results and analysis stage, the coordinates can be changed in the Parameters section in
the result features.

ADVANCED SETTINGS
To display this section, click the Show button ( ) and select Advanced Physics Options.
Normally these settings do not need to be changed.

The Use MITC interpolation check box is selected by default, and this interpolation,
which is part of the MITC shell formulation, should normally always be active.

For the Plate interface, the Use 3D formulation check box is used to select whether six
or three variables are used in the formulation. For geometrically nonlinear analyses, or
when in-plane (membrane) forces are active, six variables must be used. This check box
is selected by default.

If the Solve for out-of-plane strain components check box is selected, extra degrees of
freedom will be added for computing the out-of-plane strain components. This
formulation is similar to what is used for plane stress in the Solid Mechanics and
Membrane interfaces and it is computationally somewhat more expensive than the
standard formulation. In the default formulation, the out-of-plane strain in the shell is
explicitly computed from the stress. This may cause circular references of variables if
you for example want the constitutive law to be strain dependent. If you encounter
such problems, use the alternative formulation.

616 | CHAPTER 5: SHELL AND PLATE

DISCRETIZATION
To display this section, click the Show button ( ) and select Discretization.

Select the order of the Displacement field — Linear or Quadratic. The degrees of
freedom for the displacement of the shell normals will always have the same shape
functions as the displacements.

DEPENDENT VA RIA BLES

Both interfaces define two dependent variables (fields)—the displacement field u and
the field of normal displacements ar. The names can be changed but the names of
fields and dependent variables must in general be unique within a model. If you
intentionally use the same name for fields from different physics interfaces, these
degrees of freedom are treated as being the same. This can be used when mixing
different type of structural mechanics interfaces, where you often want the
displacements to be the equal.

• Domain, Boundary, Edge, Point, and Pair Nodes for the Shell and
Plate Interfaces
• Theory for Shell and Plate Interfaces

• Vibrations of a Disk Backed by an Air-Filled Cylinder: Application

Library path
Structural_Mechanics_Module/Acoustic-Structure_Interaction/coupled_vibr
ations_manual
• Pinched Hemispherical Shell: Application Library path
Structural_Mechanics_Module/Verification_Models/pinched_hemispherical_s
hell

THE SHELL AND PLATE INTERFACES | 617

 Domain, Boundary, Edge, Point, and Pair Nodes for the Shell and
Plate Interfaces
The Shell and Plate Interfaces have the following domain, boundary, edge, point, and
pair nodes available from the Physics ribbon toolbar (Windows users), Physics context
menu (Mac or Linux users), or right-click to access the context menu (all users).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

FEATURES AVAILABLE FROM SUBMENUS

Many features for the Shell or Plate interface are added from submenus in the Physics
toolbar groups or context menu (when you right-click the node). The submenu name
is the same in both cases.

The submenus at the Boundary (Shell interface) or Domain (Plate interface) level are

• Material Models
• Face and Volume Loads,
• Mass, Spring, and Damper
• Connections
• Face Constraints

The submenus at the Edge (Shell interface) or Boundary (Plate interface) level are

• Mass, Spring, and Damper

• Connections (Shell interface only)
• More Constraints
• Pairs

There is also the Points submenu.

618 | CHAPTER 5: SHELL AND PLATE

LINKS TO FEATURE NODE INFORMATION
These nodes are described in this section (listed in alphabetical order):

• Antisymmetry • Point Load

• Attachment • Point Mass
• Beam Connection • Point Mass Damping
• Body Load • Prescribed Acceleration
• Change Thickness • Prescribed Displacement/Rotation
• Edge Load • Prescribed Velocity
• External Stress • Rigid Connector
• Face Load • Shell Local System
• Hygroscopic Swelling • Simply Supported
• Initial Values • Solid Connection
• Initial Stress and Strain • Symmetry
• Linear Elastic Material • Thermal Expansion (for constraints)
• No Rotation • Thermal Expansion (for materials)
• Periodic Condition • Thermal Expansion (Attachment)
• Phase • Thermal Expansion (Rigid
• Pinned Connector)

If there are subsequent constraints specified on the same geometrical

entity, the last one takes precedence. The exception is that “Pinned” and
“No Rotation” boundary conditions do not override each other.

THE SHELL AND PLATE INTERFACES | 619

 These nodes are described in the documentation for the Solid Mechanics interface:

• Added Mass • Fixed Constraint2

• Applied Force (Rigid Connector) • Gravity1
• Applied Moment (Rigid Connector) • Predeformation
• Damping • Rigid Domain
• Free • Rotating Frame1
• Mass and Moment of Inertia (Rigid • Spring Foundation
Connector)
1
This is selected from the Face and Volume Loads submenu for this interface.
2
This is selected from the Face Constraints submenu for this interface.

Harmonic Perturbation, Prestressed Analysis, and Small-Signal Analysis

in the COMSOL Multiphysics Reference Manual

In the COMSOL Multiphysics Reference Manual see Table 2-3 for links
to common sections and Table 2-4 to common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

Initial Values
The Initial Values node adds initial values for the translational displacement and velocity
field as well as the for the normal displacement and velocity field. It can serve as an
initial condition for a transient simulation or as an initial guess for a nonlinear analysis.
In addition to the default Initial Values node always present in the interface, you can
add more Initial Values nodes if needed.

INITIAL VALUES
Enter values or expressions for the initial values:

• Displacement field u (the displacement components u, v, and w)

∂u
• Velocity field ------
∂t

620 | CHAPTER 5: SHELL AND PLATE

• Displacement of shell normals ar
∂ar
• Displacement of shell normals, first time derivative ----------
∂t
Context Menus
Shell>Initial Values
Shell>Boundaries>Initial Values
Plate>Initial Values
Plate>Domains>Initial Values

Ribbon
Physics tab with Shell selected:
Boundaries>Shell>Initial Values

Physics tab with Plate selected:

Domains>Plate>Initial Values

Change Thickness
Use the Change Thickness node to model regions with a thickness or offset other than
the overall values defined in the interface Thickness section.

CHANGE THICKNESS
Enter a value for the Thickness d. This value replaces the overall thickness for the
boundaries selected above. The thickness can be variable if an expression is used.

Offset Definition
If the actual shell midsurface is not the boundary on which the mesh exists, it is
possible to prescribe an offset in the direction of the surface normal by using an offset
definition. The offset is defined as positive if the shell midsurface is displaced from the
meshed boundary in the direction of the positive shell normal.

Select an option from the Offset definition list—From parent (the default), No offset,
Physical offset, or Relative offset. The default From parent means that the offset is as
defined in the Thickness section of the shell interface.

• For No offset it means that the modeled boundary coincides with the shell
midsurface.

THE SHELL AND PLATE INTERFACES | 621

 • For Physical offset enter a value or expression in the zoffset field for the actual
distance from the meshed boundary to the shell midsurface.

Figure 5-5: Meshed boundary indicated in red. The vector n indicates the positive
orientation of the shell normal.

• For Relative offset enter a value or expression in the zrel_offset field for the offset as
the ratio between the offset distance and half the shell thickness. A value of +1 means
that the actual shell bottom surface is located on the meshed boundary, and a value
of -1 means that the shell top surface is located on the meshed boundary.]

Figure 5-6: Meshed boundary indicated in red. The vector n indicates the positive
orientation of the shell normal.

For theory, see Offset.

LOCATION IN USER INTERFACE

Context Menus
Shell>Change Thickness
Plate>Change Thickness

Ribbon
Physics tab with Shell selected:
Boundaries>Shell>Change Thickness

Physics tab with Plate selected:

Domains>Plate>Change Thickness

622 | CHAPTER 5: SHELL AND PLATE

Linear Elastic Material
The Linear Elastic Material node adds the equations for a linear elastic shell and an
interface for defining the elastic material properties.

By adding the following subnodes to the Linear Elastic Material node you can
incorporate many other effects:

• Thermal Expansion (for materials)

• Hygroscopic Swelling
• Initial Stress and Strain
• External Stress
• Damping

A Shell Local System subnode is always added. In this node you specify the coordinate
system in which material orientations and results are interpreted. You can add several
Shell Local System nodes in order to control the local directions on different
boundaries.

LINEAR ELASTIC MATERIAL

Select a linear elastic Solid model—Isotropic, Orthotropic, or Anisotropic and enter the
settings as described for the Linear Elastic Material for the Solid Mechanics interface.
Note that:

• For Orthotropic no values for Ez, νyz, or νxz need to be entered due to the shell
assumptions.
• For User defined Anisotropic a 6-by-6 symmetric matrix is displayed. Due to the shell
assumptions, you only need to enter values for D11, D12, D22, D14, D24, D55, D66,
and D56.
• The material orientation is always interpreted in a local coordinate system aligned
with the shell boundary as described in Local Coordinate Systems.

GEOMETRIC NONLINEARITY
In this section there is always one check box. Either Force linear strains or Include
geometric nonlinearity is shown.

If a study step is geometrically nonlinear, the default behavior is to use a large strain
formulation in all domains. There are however some cases when you would still want
to use a small strain formulation for a certain domain. In those cases, select the Force

THE SHELL AND PLATE INTERFACES | 623

 linear strains check box. When selected, a small strain formulation is always used,
independently of the setting in the study step.

The default value is that the check box is cleared, except when opening a model created
in a version prior to 4.3. In this case the state is chosen so that the properties of the
model are conserved.

The Include geometric nonlinearity check box is displayed only if the model
was created in a version prior to 4.3, and geometric nonlinearity was
originally used for the selected domains. It is then selected and forces the
Include geometric nonlinearity check box in the study step to be selected.
If the check box is cleared, it is permanently removed and the study step
assumes control over the selection of geometric nonlinearity.

LOCATION IN USER INTERFACE

Context Menus
Shell>Material Models>Linear Elastic Material
Plate>Material Models>Linear Elastic Material

Ribbon
Physics tab with Shell selected:
Boundaries>Material Models>Linear Elastic Material

Physics tab with Plate selected:

Domains>Material Models>Linear Elastic Material

Shell Local System

The Shell Local System subnode is used for interpreting directions of orthotropic and
anisotropic material data and when stresses or strains are presented in a local system.

624 | CHAPTER 5: SHELL AND PLATE

COORDINATE SYSTEM SELECTION
The Global coordinate system is selected by default. The Coordinate system list contains
any additional coordinate system that the model includes. The coordinate system given
is projected onto the shell surface as described in Local Coordinate Systems.

When a Shell or Plate interface is added, there is also a Shell Local System
node added under Definitions. This coordinate can be used to reference
the local directions selected in a Shell interface. It is used internally in the
Shell interface, and can also be accessed from other physics interfaces. The
coordinate system can also be used in a Coordinate System Surface plot to
visualize the local directions.

Do not edit the Shell Local System node added under Definitions.

LOCATION IN USER INTERFACE

Context Menus
Shell>Linear Elastic Material>Shell Local System
Plate>Linear Elastic Material>Shell Local System

Ribbon
Physics tab with Linear Elastic Material node selected in the model tree:
Attributes>Shell Local System

Thermal Expansion (for materials)

Use the Thermal Expansion subnode to add an internal thermal strain caused by changes
in temperature. The thermal strain depends on the coefficient of thermal expansion
(CTE) α, the temperature T, and the strain-free reference temperature Tref as

ε th = α ( T – T ref )

It is possible to model bending due to a temperature gradient in the thickness direction

of the shell. The temperature is then assumed to vary linearly through the thickness.

MODEL INPUTS
From the Temperature T list, select an existing temperature variable from a heat transfer
interface (for example, Temperature (htsh/sol1)), if any temperature variables exist. For
User defined enter a value or expression for the temperature. This is the mid-surface
temperature of the shell, controlling the membrane part of the thermal expansion. Iff

THE SHELL AND PLATE INTERFACES | 625

 needed, you can add a through-thickness temperature gradient in the Thermal Bending
section.

THERMAL EXPANSION PROPERTIES

Specify the thermal properties that define the thermal strain.

From the Coefficient of thermal expansion α list, select From material to use the
coefficient of thermal expansion from the material, or User defined to enter a value or
expression for α. Select Isotropic, Diagonal or Symmetric to enter one or more
components for a general coefficient of thermal expansion tensor α. When a
non-isotropic coefficient of thermal expansion is used, the axis orientations are given
by the coordinate system selection in the parent node.

Enter a value or expression of the Strain reference temperature Tref, which is the
reference temperature where the thermal strain is zero.

THERMAL BENDING
Enter the Temperature difference in thickness direction ΔTz, which affects the thermal
bending. This is the difference between the temperatures at the top and bottom
surfaces.

LOCATION IN USER INTERFACE

Context Menus
Shell>Linear Elastic Material>Thermal Expansion
Plate>Linear Elastic Material>Thermal Expansion

Ribbon
Physics tab with Linear Elastic Material node selected in the model tree:
Attributes>Thermal Expansion

Hygroscopic Swelling
Hygroscopic swelling is an internal strain caused by changes in moisture content. This
strain can be written as

ε hs = β h ( c mo – c mo,ref )

where βh is the coefficient of hygroscopic swelling, cmo is the moisture concentration,

and cmo,ref is the strain-free reference concentration. It is possible to model bending

626 | CHAPTER 5: SHELL AND PLATE

due to a concentration gradient in the thickness direction of the shell. The
concentration is then assumed to vary linearly through the thickness.

MODEL INPUTS
From the Concentration c list, select an existing concentration variable from another
physics interface, if any concentration variables exist. For User defined enter a value or
expression for the concentration. This is the mid-surface concentration in the shell,
controlling the membrane part of the hygroscopic swelling.

The unit for the input depends on the setting of Concentration type in the Hygroscopic
Swelling Properties section. Only concentration variables having the chosen physical
dimension are available in the Concentration list.

HYGROSCOPIC SWELLING PROPERTIES

In the Concentration type list, select Molar concentration (the default) or Mass
concentration, depending on the units used for the concentration.

Enter a Strain reference concentration cref. This is the concentration at which there are
no strains due to hygroscopic swelling.

If Molar concentration is selected as Concentration type, also enter the Molar mass of the
fluid, Mm. The default value is 0.018 kg/mol, which is the molar mass of water.

The default Coefficient of hygroscopic swelling βh uses values From material. For
User defined select Isotropic (the default), Diagonal, or Symmetric to enter one or more
components for a general coefficient of hygroscopic swelling tensor βh.The default
value for the User defined case is 1.5e-4 m3/kg. When a non-isotropic coefficient of
hygroscopic swelling is used, the axis orientations are given by the coordinate system
selection in the parent node.

The Include moisture as added mass check box is selected by default. When selected,
the mass of the fluid is included in a dynamic analysis, and when using mass
proportional loads. It will also contribute when computing mass properties.

HYGROSCOPIC BENDING
Enter the Concentration difference in thickness direction Δcgy, which affects the
hygroscopic bending. This is the difference between the concentrations at the top and
bottoms surfaces.

THE SHELL AND PLATE INTERFACES | 627

 LOCATION IN USER INTERFACE

Context Menus
Shell>Linear Elastic Material>Hygroscopic Swelling
Plate>Linear Elastic Material>Hygroscopic Swelling

Ribbon
Physics tab with Linear Elastic Material node selected in the model tree:
Attributes>Hygroscopic Swelling

Initial Stress and Strain

You can add the Initial Stress and Strain subnode to the Linear Elastic Material, in order
to specify the stress or strain state in the structure before applying any constraint or
load. The values given are not initial values in the mathematical sense, but rather a
contribution to the constitutive relation

In many cases Initial Stress and Strain and External Stress are
interchangeable when prescribing stresses, but you can find some more
options in the latter.

COORDINATE SYSTEM SELECTION

The given initial stresses and strains are interpreted in this system.

The Shell local system is selected by default. This means that the orientations are the
same as used for the material description.

If you select Projected from coordinate system, you can select any coordinate system
present in the model. The Coordinate system list will then contain any coordinate
system present in the model. The coordinate system given is projected onto the shell
surface as described in Local Coordinate Systems.

INITIAL STRESS
Specify the initial stress as the Initial in-plane force, the Initial moment, and the Initial
out-of-plane shear force. Enter values or expressions in the applicable fields for the:

• Initial in-plane force Ni

• Initial moment Mi
• Initial out-of-plane shear force Qi

628 | CHAPTER 5: SHELL AND PLATE

INITIAL STRAIN
Specify the initial strain as the Initial membrane strain, the Initial bending strain, and the
Initial transverse shear strain. Enter values or expressions in the applicable fields for the:

• Initial membrane strain γi

• Initial bending strain χi
• Initial transverse shear strain ζi

• For definitions of the generalized strains, see Theory for Shell and Plate
Interfaces.
• For details about initial stresses and strains, see Inelastic Strain
Contributions and Initial Stresses and Strains.

LOCATION IN USER INTERFACE

Context Menus
Shell>Linear Elastic Material>Initial Stress and Strain
Plate>Linear Elastic Material>Initial Stress and Strain

Ribbon
Physics tab with Linear Elastic Material node selected in the model tree:
Attributes>Initial Stress and Strain

External Stress
You can add the External Stress subnode to several material models, in order to specify
an additional stress contribution which is not part of the constitutive relation. The
external stress can be added to the total stress tensor, or act only as an extra load
contribution.

In many cases External Stress and Initial Stress and Strain are
interchangeable when prescribing stresses. In Initial Stress and Strain, the
given stress is however always added to the stress tensor.

COORDINATE SYSTEM SELECTION

The Global coordinate system is selected by default. The Coordinate system list contains
any additional coordinate system that the model includes. The user defined external
stresses and strains are interpreted in this system. The coordinate system given is

THE SHELL AND PLATE INTERFACES | 629

 projected onto the shell surface as described in Local Coordinate Systems.

EXTERNAL STRESS
Select a Stress input—Stress tensor or In-plane force.

• When Stress tensor is selected, you enter the external stress in the form of Second
Piola-Kirchhoff stress tensors. The Membrane part of external stress, Bending part of
external stress, and Shear part of external stress drop-down lists all behave the same
way: Depending on the type, they will contain all membrane (bending, shear) stress
tensors announced by any physics interface, and also the entry User defined. When
User defined is selected, you can enter the data for the membrane (bending, shear)
part of the External stress tensor σext,m (σext,b, σext,s) as Isotropic, Diagonal, or
Symmetric depending on the properties of the tensor. The tensor components are
interpreted in the selected coordinate system. If a stress tensor announced by a
physics interface is selected, the coordinate system setting is ignored — the
orientation is handled internally.

Choose a Contribution type—Add to stress tensor or Load contribution only to

determine the effect of the contribution.
• When In-plane force is selected, you enter the external stress in the form of section
forces. Specify the external stress as values or expressions for the In-plane force Next,
the Moment Mext, and the Out-of-plane shear force Qext. Choose a Contribution
type—Add to stress tensor or Load contribution only to determine the effect of the
contribution.

Selecting a stress tensor announced by the same physics interface as where

the External Stress node is added, will result in an error (‘Circular
variable dependency detected’). This operation would imply that
the computed stress depends on itself.

For theory, see External Stress.

LOCATION IN USER INTERFACE

Context Menus
Shell>Linear Elastic Material>External Stress
Plate>Linear Elastic Material>External Stress

630 | CHAPTER 5: SHELL AND PLATE

Ribbon
Physics tab with Linear Elastic Material node selected in the model tree:
Attributes>External Stress

Prescribed Displacement/Rotation
The Prescribed Displacement/Rotation node adds an edge, boundary, domain, or point
condition to a model where the displacements and rotations are prescribed in one or
more directions.

With this condition it is possible to prescribe a displacement in one direction or one of

the rotations, leaving the shell free to deform or rotate in the other directions.

• If zero displacements and rotations are prescribed in all directions, this is the same
as Fixed Constraint.
• If zero displacements are prescribed in all directions, this is the same as Pinned.
• If zero rotations are prescribed in all directions, this is the same as No Rotation.

FACE DEFINING THE LOCAL EDGE SYSTEM

This setting is used in conjunction with a Local edge system. If displacement or
rotations is prescribed for an edge which is shared between boundaries, the edge
system can be ambiguous. Select the boundary which should define the edge system.
The default is Use face with lowest number.

This section is available only for edges in the Shell interface.

COORDINATE SYSTEM SELECTION

Specify the coordinate system to use for specifying the prescribed
displacement/rotation. The coordinate system selection is based on the geometric
entity level.

Boundaries (Plate interface: Domains)

From the Coordinate system list select from:

• Global coordinate system (the default)

• Boundary System (a predefined normal-tangential coordinate system)
• Any additional user-defined coordinate system

THE SHELL AND PLATE INTERFACES | 631

 Edges (Plate interface: Boundaries)
From the Coordinate system list select from:

• Local edge system (the default).

For details about the definition of local edge systems, see Local edge
system.

• Global coordinate system (the standard global coordinate system).

• Any additional user-defined coordinate system.

Points
From the Coordinate system list select from:

• Global coordinate system (the default)

• Any additional user-defined coordinate system

Depending on the selected coordinate system, the displacement and rotation

components change accordingly.

PRESCRIBED DISPLACEMENT
To define a prescribed displacement for each spatial direction (x, y, and z), select one
or all of the Prescribed in x direction, Prescribed in y direction, and Prescribed in
z direction check boxes. Then enter a value or expression for the prescribed
displacements u0, v0, or w0.

PRESCRIBED ROTATIONS
Select a prescribed rotation from the By list—Free, Rotation, or Normal vector. Select:

• Free (the default) to leave the rotations unconstrained.

• Rotation to activate a prescribed rotation in a direction. Enter a value or expression
for the prescribed rotation θ in each row for the local tangential directions t1 and
t2. Under For small strains, select one or both of the Free rotation around t1 direction
and Free rotation around t2 direction check boxes to remove the constraint for the
corresponding rotation component. If unchecked, the rotations are constrained to
either the input value or to the default zero rotation. The status of the check boxes
has no effect when the geometric nonlinearity is activated under the study settings.

632 | CHAPTER 5: SHELL AND PLATE

 This is because the constraints put on different rotation components are not
independent of each other in case of finite rotations.
• Normal vector to describe the rotational degrees of freedom as a prescribed normal
vector. Enter the components of the Prescribed normal vector N0.

CONSTRAINT SETTINGS

To display this section, click the Show button ( ) and select Advanced Physics Options.

In the COMSOL Multiphysics Reference Manual:

• Constraint Reaction Terms

• Weak Constraints
• Constraint Settings

When Individual dependent variables is selected in the Apply reaction terms

on drop-down list, the constraint forces are applied directly on the degrees
of freedom, which are the displacements along the global coordinate axes.
If you use this setting together with a local coordinate system, the results
will be inconsistent since the constraint forces will not match the
constraint orientation.

• You can add a Harmonic Perturbation subnode for specifying a

harmonic variation of the values of the prescribed displacements in a
frequency domain analysis of perturbation type.
• You can activate and deactivate this boundary condition by assigning it
to a constraint group. See Load Cases in the Structural Mechanics
Modeling chapter.
• You can assign the value of the prescribed displacement to a load
group. See Load Cases in the Structural Mechanics Modeling chapter.

LOCATION IN USER INTERFACE

Context Menus
Shell>Face Constraints>Prescribed Displacement/Rotation
Shell>Prescribed Displacement/Rotation (Edge)

THE SHELL AND PLATE INTERFACES | 633

 Shell>Points>Prescribed Displacement/Rotation
Plate>Face Constraints>Prescribed Displacement/Rotation
Plate>Prescribed Displacement/Rotation (Boundary)
Plate>Points>Prescribed Displacement/Rotation

Ribbon
Physics tab with Shell selected:
Boundaries>Face Constraints>Prescribed Displacement/Rotation
Edges>Shell>Prescribed Displacement/Rotation
Points>Shell>Prescribed Displacement/Rotation

Physics tab with Plate selected:

Domains>Face Constraints>Prescribed Displacement/Rotation
Boundaries>Plate>Prescribed Displacement
Points>Plate>Prescribed Displacement

Thermal Expansion (for constraints)

Add the Thermal Expansion subnode to Fixed Constraint to prescribe a deformation of
the constrained edge caused by changes in temperature of the surroundings. This
makes it possible to reduce the stresses caused by such boundary conditions.

The thermal strain depends on the coefficient of thermal expansion α, the temperature
T, and the strain-free reference temperature Tref as

ε th = α ( T – T ref )

THERMAL EXPANSION PROPERTIES

Specify the thermal properties that define the thermal strain. This is a description of
the thermal expansion of the urroundings idealized by the constraints.

Select Inherit from boundary to take the thermal expansion data from the domain being
constrained. This should only be used when:

• The temperature and the thermal expansion coefficient do not have a spatial
variation.
• The virtual surrounding material has the same thermal expansion as the domain
itself.

634 | CHAPTER 5: SHELL AND PLATE

When Inherit from boundary is not selected, enter:

• A value or expression for Temperature T, specifying the temperature distribution of

the surrounding material. Any temperature variation must be an explicit function of
the material frame coordinates. It is not possible to use a computed temperature
distribution.
• The Coefficient of thermal expansion α. As a default, values From material are used.
This requires that a material has been assigned to the boundaries, edges, or points
where the constraint is active.
For User defined select Isotropic, Diagonal or Symmetric to enter one or more
components for a general coefficient of thermal expansion tensor α. When a
non-isotropic coefficient of thermal expansion is used, the axis orientations are given
by the coordinate system selected in the Coordinate System Selection section.
• A value or expression for the Strain reference temperature Tref that is the
temperature at which there are no thermal displacements at the constraints.

Enter the coordinates of the Reference point, the point where the displacement is zero.
The choice of reference point only affects the rigid body motion. If there are several
different constraints with a Thermal Expansion subnode, the same reference point
should usually be selected in all of them.

• Constraints and Thermal Expansion in the Structural Mechanics

Modeling chapter.
• Thermal Expansion of Constraints in the Structural Mechanics Theory
chapter.

LOCATION IN USER INTERFACE

Context Menus
Shell>Fixed Constraint>Thermal Expansion

Ribbon
Physics tab with Fixed Constraint or Prescribed Displacement node selected in the model
tree:
Attributes>Thermal Expansion

THE SHELL AND PLATE INTERFACES | 635

 Prescribed Velocity
The Prescribed Velocity node adds an edge, boundary, domain, or point condition
where the translational or rotational velocity is prescribed in one or more directions.
The prescribed velocity condition is applicable for Time Dependent and Frequency
Domain studies. With this condition it is possible to prescribe a velocity in one
direction, leaving the shell free in the other directions.

The Prescribed Velocity node is a constraint, and overrides any other constraint on the
same selection.

For details about prescribed velocities and accelerations, see Prescribed

Displacements, Velocities, and Accelerations.

COORDINATE SYSTEM SELECTION

Specify the coordinate system to use for specifying the prescribed velocity. The
coordinate system selection is based on the geometric entity level.

Coordinate systems with directions that change with time should not be
used. If you choose another, local coordinate system, the acceleration
components change accordingly.

Boundaries
From the Coordinate system list select from:

• Global coordinate system (the default)

• Boundary System (a predefined normal-tangential coordinate system)
• Any additional user-defined coordinate system

Edges
From the Coordinate system list select from:

• Local edge system (the default)

For details about the definition of local edge systems, see Local edge
system.

636 | CHAPTER 5: SHELL AND PLATE

• Global coordinate system (the standard global coordinate system).
• Any additional user-defined coordinate system.

Points
From the Coordinate system list select from:

• Global coordinate system (the default)

• Any additional user-defined coordinate system

Depending on the selected coordinate system, the velocity components change

accordingly.

FACE DEFINING THE LOCAL EDGE SYSTEM

This setting is used in conjunction with a Local edge system. If the velocity is prescribed
for an edge which is shared between boundaries, the edge system can be ambiguous.
Select the boundary which should define the edge system. The default is Use face with
lowest number.

This section is available only for edges in the Shell interface and is only
visible if the selected coordinate system is Local edge system.

P RE S C R I B E D VE L O C I T Y
To define a prescribed velocity for each spatial direction (x, y, and z), select one or
more of the Prescribed in x direction, Prescribed in y direction, and Prescribed in
z direction check boxes. Then enter a value or expression for the prescribed velocity
components vx, vy, and vz.

P RE S C R I B E D A N G U L A R VE L O C I T Y
To define a prescribed angular velocity for each spatial direction (x, y, and z), select one
or all of the Prescribed around x direction, Prescribed around y direction, and Prescribed
around z direction check boxes and enter a value or expression for in each ∂ θ x ⁄ ∂ t ,
∂ θ y ⁄ ∂ t , or ∂ θ z ⁄ ∂ t field.

You can add a Harmonic Perturbation subnode for specifying a harmonic

variation of the values of the prescribed velocity in a frequency domain
analysis of perturbation type.

THE SHELL AND PLATE INTERFACES | 637

 LOCATION IN USER INTERFACE

Context Menus
Shell>Face Constraints>Prescribed Velocity
Shell>More Constraints>Prescribed Velocity (Edge)
Plate>Face Constraints>Prescribed Velocity
Plate>More Constraints>Prescribed Velocity (Boundary)

Ribbon
Physics tab with Shell selected:
Boundaries>Face Constraints>Prescribed Velocity
Edges>>More Constraints>Prescribed Velocity

Physics tab with Plate selected:

Domains>Face Constraints>Prescribed Velocity
Boundaries>>More Constraints>Prescribed Velocity

Prescribed Acceleration
The Prescribed Acceleration node adds an edge, boundary, domain, or point condition
where the translational or rotational acceleration is prescribed in one or more
directions. The prescribed acceleration condition is applicable for Time Dependent
and Frequency Domain studies. With this condition it is possible to prescribe an
acceleration in one direction, leaving the shell free in the other directions.

The Prescribed Acceleration node is a constraint, and overrides any other constraint on
the same selection.

For details about prescribed velocities and accelerations, see Prescribed

Displacements, Velocities, and Accelerations.

COORDINATE SYSTEM SELECTION

Specify the coordinate system to use for specifying the prescribed acceleration. The
coordinate system selection is based on the geometric entity level.

Coordinate systems with directions that change with time should not be
used. If you choose another, local coordinate system, the acceleration
components change accordingly.

638 | CHAPTER 5: SHELL AND PLATE

Boundaries
From the Coordinate system list select from:

• Global coordinate system (the default)

• Boundary System (a predefined normal-tangential coordinate system)
• Any additional user-defined coordinate system

Edges
From the Coordinate system list select from:

• Local edge system (the default).

For details about the definition of local edge systems, see Local edge
system.

• Global coordinate system (the standard global coordinate system).

• Any additional user-defined coordinate system.

Points
From the Coordinate system list select from:

• Global coordinate system (the default)

• Any additional user-defined coordinate system

Depending on the selected coordinate system, the acceleration components change

accordingly.

FACE DEFINING THE LOCAL EDGE SYSTEM

This setting is used in conjunction with a Local edge system. If the acceleration is
prescribed for an edge which is shared between boundaries, the edge system can be
ambiguous. Select the boundary which should define the edge system. The default is
Use face with lowest number.

This section is available only for edges in the Shell interface and is only
visible if the selected coordinate system is Local edge system.

PRESCRIBED ACCELERATION
To define a prescribed acceleration for each spatial direction (x, y, and z), select one or
more of the Prescribed in x direction, Prescribed in y direction, and Prescribed in

THE SHELL AND PLATE INTERFACES | 639

 z direction check boxes. Then enter a value or expression for the prescribed
acceleration components ax, ay, and az.

PRESCRIBED ANGULAR ACCELERATION

To define a prescribed angular acceleration for each spatial direction (x, y, and z), select
one or all of the Prescribed around x, y, and z direction check boxes and enter a value
2 2 2 2 2 2
or expression for in each ∂ θ x ⁄ ∂ t , ∂ θ y ⁄ ∂ t , or ∂ θ z ⁄ ∂ t field.

You can add a Harmonic Perturbation subnode for specifying a harmonic

variation of the values of the prescribed acceleration in a frequency
domain analysis of perturbation type.

LOCATION IN USER INTERFACE

Context Menus
Shell>Face Constraints>Prescribed Acceleration
Shell>More Constraints>Prescribed Acceleration (Edge)
Plate>Face Constraints>Prescribed Acceleration
Plate>More Constraints>Prescribed Acceleration (Boundary)

Ribbon
Physics tab with Shell selected:
Boundaries>Face Constraints>Prescribed Acceleration
Edges>>More Constraints>Prescribed Acceleration

Physics tab with Plate selected:

Domains>Face Constraints>Prescribed Acceleration
Boundaries>>More Constraints>Prescribed Acceleration

Pinned
The Pinned node adds an edge, boundary, domain, or point condition that fixes the
translations in all directions, that is, all displacements are zero. The rotations are not
constrained.

640 | CHAPTER 5: SHELL AND PLATE

CONSTRAINT SETTINGS
To display this section, click the Show button ( ) and select Advanced Physics Options.

In the COMSOL Multiphysics Reference Manual:

• Constraint Reaction Terms

• Weak Constraints
• Constraint Settings

You can activate and deactivate this boundary condition by assigning it to

a constraint group. See Load Cases in the Structural Mechanics Modeling
chapter.

LOCATION IN USER INTERFACE

Context Menus
Shell>Face Constraints>Pinned
Shell>More Constraints>Pinned (Edge)
Shell>Points>Pinned
Plate>Face Constraints>Pinned
Plate>More Constraints>Pinned (Boundary)
Plate>Points>Pinned

Ribbon
Physics tab with Shell selected:
Boundaries>Face Constraints>Pinned
Edges>More Constraints>Pinned
Points>Shell>Pinned

Physics tab with Plate selected:

Boundaries>Face Constraints>Pinned
Edges>More Constraints>Pinned
Points>Plate>Pinned

No Rotation
The No Rotation node adds an edge, boundary, domain, or point condition that fixes
the rotations around all axes. The translations are not constrained.

THE SHELL AND PLATE INTERFACES | 641

 CONSTRAINT SETTINGS
To display this section, click the Show button ( ) and select Advanced Physics Options.

In the COMSOL Multiphysics Reference Manual:

• Constraint Reaction Terms

• Weak Constraints
• Constraint Settings

You can activate and deactivate this boundary condition by assigning it to

a constraint group. See Load Cases in the Structural Mechanics Modeling
chapter.

LOCATION IN USER INTERFACE

Context Menus
Shell>Face Constraints>No Rotation
Shell>More Constraints>No Rotation (Edge)
Shell>Points>No Rotation
Plate>Face Constraints>No Rotation
Plate>More Constraints>No Rotation (Boundary)
Plate>Points>No Rotation

Ribbon
Physics tab with Shell selected:
Boundaries>Face Constraints>No Rotation
Edges>More Constraints>No Rotation
Points>Shell>No Rotation

Physics tab with Plate selected:

Boundaries>Face Constraints>No Rotation
Edges>More Constraints>No Rotation
Points>Plate>No Rotation

Simply Supported
The Simply Supported node adds an edge condition that constrains the displacement in
the direction perpendicular to the shell. The in-plane rotation perpendicular to the
edge is also constrained. Optionally you can constrain the in-plane displacements.

642 | CHAPTER 5: SHELL AND PLATE

IN-PLANE DISPLACEMENT CONSTRAINTS
Select Along edge to constrain the in-plane translation along the edge.

Select Perpendicular to edge to constrain the in-plane displacement perpendicular to the

edge.

CONSTRAINT SETTINGS
To display this section, click the Show button ( ) and select Advanced Physics Options.

In the COMSOL Multiphysics Reference Manual:

• Constraint Reaction Terms

• Weak Constraints
• Constraint Settings

You can activate and deactivate this boundary condition by assigning it to

a constraint group. See Load Cases in the Structural Mechanics Modeling
chapter.

LOCATION IN USER INTERFACE

Context Menus
Shell>More Constraints>Simply Supported
Plate>More Constraints>Simply Supported

Ribbon
Physics tab with Shell selected:
Edges>More Constraints>Simply Supported

Physics tab with Plate selected:

Boundaries>More Constraints>Simply Supported

Symmetry
The Symmetry node adds an edge or boundary condition that defines a symmetry edge
or boundary.

THE SHELL AND PLATE INTERFACES | 643

 FACE DEFINING THE LOCAL EDGE SYSTEM (EDGES ONLY)
This setting is used in conjunction with a Local edge system. If symmetry is prescribed
for an edge which is shared between boundaries, the edge system can be ambiguous.
Select the boundary which should define the edge system. The default is Use face with
lowest number.

This section is available only for edges in the Shell interface.

COORDINATE SYSTEM SELECTION (EDGES ONLY)

Specify the coordinate system to use for specifying a symmetry edge. From the
Coordinate system list select from:

• Local edge system (the default).

For details about the definition of local edge systems, see Local edge
system.

• Global coordinate system (the standard global coordinate system).

• Any additional user-defined coordinate system.

SYMMETRY (EDGES ONLY)

If another coordinate system than the Local edge system is used, select an Axis to use as
symmetry plane normal. Select 1, 2, or 3 for the first, second, or third axis, respectively.

CONSTRAINT SETTINGS
To display this section, click the Show button ( ) and select Advanced Physics Options.

In the COMSOL Multiphysics Reference Manual:

• Constraint Reaction Terms

• Weak Constraints
• Constraint Settings

Symmetry and Antisymmetry Boundary Conditions

644 | CHAPTER 5: SHELL AND PLATE

 You can activate and deactivate this boundary condition by assigning it to
a constraint group. See Load Cases in the Structural Mechanics Modeling
chapter.

LOCATION IN USER INTERFACE

Context Menus
Shell>Face Constraints>Symmetry
Shell>More Constraints>Symmetry (Edge)
Plate>More Constraints>Symmetry (Boundary)

Ribbon
Physics tab with Shell selected:
Boundaries>Face Constraints>Symmetry
Edges>More Constraints>Symmetry

Physics tab with Plate selected:

Boundaries>More Constraints>Symmetry

Antisymmetry
The Antisymmetry node adds an edge or boundary condition that defines an
antisymmetry edge or boundary.

FACE DEFINING THE NORMAL DIRECTION (EDGES ONLY)

This setting is used in conjunction with a Local edge system. If antisymmetry is
prescribed for an edge which is shared between boundaries, the edge system can be
ambiguous. Select the boundary which should define the edge system. The default is
Use face with lowest number.

This section is available only for edges in the Shell interface.

THE SHELL AND PLATE INTERFACES | 645

 COORDINATE SYSTEM SELECTION (EDGES ONLY)
Specify the coordinate system to use for specifying a symmetry edge. From the
Coordinate system list select from:

• Local edge system (the default).

For details about the definition of local edge systems, see Local edge
system.

• Global coordinate system (the standard global coordinate system).

• Any additional user-defined coordinate system.

ANTISYMMETRY (EDGES ONLY)

If another coordinate system than the Local edge system is used, select an Axis to use as
symmetry plane normal. Select 1, 2, or 3 for the first, second, or third axis, respectively.

CONSTRAINT SETTINGS
To display this section, click the Show button ( ) and select Advanced Physics Options.

In the COMSOL Multiphysics Reference Manual:

• Constraint Reaction Terms

• Weak Constraints
• Constraint Settings

Symmetry and Antisymmetry Boundary Conditions

You can activate and deactivate this boundary condition by assigning it to

a constraint group. See Load Cases in the Structural Mechanics Modeling
chapter.

646 | CHAPTER 5: SHELL AND PLATE

LOCATION IN USER INTERFACE

Context Menus
Shell>Face Constraints>Antisymmetry
Shell>More Constraints>Antisymmetry (Edge)
Plate>More Constraints>Antisymmetry (Boundary)

Ribbon
Physics tab with Shell selected:
Boundaries>Face Constraints>Antisymmetry
Edges>More Constraints>Antisymmetry

Physics tab with Plate selected:

Boundaries>More Constraints>Antisymmetry

Body Load
Add a Body Load to boundaries (for the Plate interface add it to domains). The loads
are defined in the given coordinate system.

FORCE
Enter values or expressions for the components (x. y, z) of the body load FV.

MOMENT
Enter values or expressions for the components (x. y, z) of the moment body load ML.

After selecting a Load type, the Load list normally only contains User
defined. When combining the Shell interface with another physics
interface, it is also possible to select a predefined load from this list.

• You can add the Phase subnode to specify the phase of this load in a
frequency domain analysis.
• You can specify this load to be a Harmonic Perturbation in a frequency
domain analysis.
• You can assign this load to a load group. See Load Cases in the
Structural Mechanics Modeling chapter.

THE SHELL AND PLATE INTERFACES | 647

 LOCATION IN USER INTERFACE

Context Menus
Shell>Face and Volume Loads>Body Load
Plate>Face and Volume Loads>Body Load

Ribbon
Physics tab with Shell selected:
Boundaries>Face and Volume Loads>Body Load

Physics tab with Plate selected:

Domains>Face and Volume Loads>Body Load

Face Load
Add a Face Load to boundaries (for the Plate interface add it to domains), to use it as
a pressure or tangential force acting on a surface. The loads are defined in the given
coordinate system.

FORCE
Select a Load type—Force per unit area, Total force, or Pressure.

LOAD TYPE VARIABLE SI UNIT GEOMETRIC SPACE DIMENSION

ENTITY LEVEL (COMPONENTS)

Force per unit area FA N/m2 boundaries 3D (x, y, z)

domains 2D (x, y, z)
Total force Ftot N boundaries 3D(x, y, z)
domains 2D(x, y, z)
Pressure p Pa boundaries 3D
domains 2D

• A positive pressure is directed in the negative element normal

direction.
• The pressure load is a ‘follower load’. The direction changes with
deformation in a geometrically nonlinear analysis.
• After selecting a Load type, the Load list normally only contains User
defined. When combining with another physics interface, it is also
possible to choose a predefined load from this list.

648 | CHAPTER 5: SHELL AND PLATE

MOMENT
Enter values or expressions for the components of the moment face load MA.

• You can add the Phase subnode to specify the phase of this load in a
frequency domain analysis.
• You can specify this load to be a Harmonic Perturbation in a frequency
domain analysis.
• You can assign this load to a load group. See Load Cases in the
Structural Mechanics Modeling chapter.

LOCATION IN USER INTERFACE

Context Menus
Shell>Face and Volume Loads>Face Load
Plate>Face and Volume Loads>Face Load
Membrane>Face and Volume Loads>Face Load

Ribbon
Physics tab with Shell or Membrane selected:
Boundaries>Face and Volume Loads>Face Load

Physics tab with Plate selected:

Domains>Face and Volume Loads>Face Load

Edge Load
Add an Edge Load as a force or moment distributed along an edge (for the Plate
interface add it to boundaries). The load is defined in the given local coordinate
system.

COORDINATE SYSTEM SELECTION

Specify the coordinate system to use for specifying the load. From the Coordinate
system list select from:

• Global coordinate system (the standard global coordinate system).

THE SHELL AND PLATE INTERFACES | 649

 • Local edge system (Shell only)

For details about the definition of local edge systems, see Local edge
system.

• Any additional user-defined coordinate system.

FACE DEFINING THE LOCAL EDGE SYSTEM

This setting is used in conjunction with a Local edge system. When the load is applied
to an edge which is shared between boundaries, the edge system can be ambiguous.
Select the boundary which should define the edge system. The default is Use face with
lowest number.

This section is available only for edges in the Shell interface and is only
visible if the selected coordinate system is Local edge system.

FORCE
Select a Load type—Force per unit length (the default), Force per unit area, or Total
Force. Enter values or expressions for the components (x, y, z).

LOAD TYPE VARIABLE SI UNIT GEOMETRIC SPACE DIMENSION

ENTITY LEVEL

Force per unit length FL N/m edges 3D

boundaries 2D
2
Force per unit area FA N/m edges 3D
boundaries 2D
Total force Ftot N edges 3D
boundaries 2D

After selecting a Load type, the Load list normally only contains User
defined. When combining the Shell interface with another physics
interface, it is also possible to choose a predefined load from this list.

MOMENT
Select a Load type to define the moment load—Moment per unit length (the default) or
Moment per unit area. Enter values or expressions for the components (x, y, z).

650 | CHAPTER 5: SHELL AND PLATE

This section is available only in the Shell and Plate interfaces.

LOAD TYPE VARIABLE SI UNIT GEOMETRIC SPACE DIMENSION

ENTITY LEVEL

Moment per unit length ML N edges 3D

boundaries 2D
2
Moment per unit area MA N⋅m/m edges 3D
boundaries 2D
Total moment Mtot N⋅m edges 3D
boundaries 2D

• You can add the Phase subnode to specify the phase of this load in a
frequency domain analysis.
• You can specify this load to be a Harmonic Perturbation in a frequency
domain analysis.
• You can assign this load to a load group. See Load Cases in the
Structural Mechanics Modeling chapter.

LOCATION IN USER INTERFACE

Context Menus
Shell>Edge Load
Plate>Edge Load
Membrane>Edge Load

Ribbon
Physics tab with Shell or Membrane selected:
Boundaries>Shell>Edge Load
Boundaries>Membrane>Edge Load

Physics tab with Plate selected:

Domains>Plate>Edge Load

Point Load
Add a Point Load to points for concentrated forces or moments at points. The loads are
defined in the given coordinate system.

THE SHELL AND PLATE INTERFACES | 651

 FORCE
Enter values or expressions for the components (x. y, z) of the point load FP.

The Load list normally only contains User defined. When combining the
Shell interface with another physics interface, it is also possible to choose
a predefined load from this list.

MOMENT
Enter values or expressions for the components (x. y, z) of the point moment MP.

• You can add the Phase subnode to specify the phase of this load in a
frequency domain analysis.
• You can specify this load to be a Harmonic Perturbation in a frequency
domain analysis.
• You can assign this load to a load group. See Load Cases in the
Structural Mechanics Modeling chapter.

LOCATION IN USER INTERFACE

Context Menus
Shell>Points>Point Load
Plate>Points>Point Load

Ribbon
Physics tab with Shell or Plate selected:
Points>Point Load

Point Mass
Use the Point Mass node to model a discrete mass or mass moment of inertia that is
concentrated at a point.

The Point Mass Damping subnode can be added to specify a mass-proportional

damping.

COORDINATE SYSTEM SELECTION

With the Coordinate system list, control the coordinate system around the axis of which
the principal mass moment of inertias are defined.

652 | CHAPTER 5: SHELL AND PLATE

PO IN T MA SS
Enter a Point mass m.

Enter a single value for an isotropic Mass moment of inertia J, or select Diagonal or
Symmetric to enter a full moment of inertia tensor.

FRAME ACCELERATION FORCES

Click to select the Exclude contribution check box to switch off the loads that can be
caused by the point mass when the frame is accelerated when using a Gravity or
Rotating Frame feature. The setting will also determine whether the node will
contribute when Computing Mass Properties.

LOCATION IN USER INTERFACE

Context Menus
Shell>Points>Point Mass
Plate>Points>Point Mass

Ribbon
Physics tab with Shell or Plate selected:
Points>Shell>Point Mass
Points>Plate>Point Mass

Point Mass Damping

Use the Point Mass Damping subnode to add damping to a Point Mass parent node.

PO IN T MASS DAMPIN G

Enter a Mass damping parameter αdM. This is the mass proportional term of a Rayleigh
damping.L O C A T I O N I N U S E R I N T E R F A C E

Context Menus
Shell>Point Mass>Point Mass Damping
Plate>Point Mass>Point Mass Damping

Ribbon
Physics tab with Point Mass node selected in the model tree:
Attributes>Point Mass Damping

THE SHELL AND PLATE INTERFACES | 653

 Periodic Condition
Use a Periodic Condition to prescribe that the displacements and rotations on two
different sets of edges with the same geometrical shape are related, as in a periodic
structure. In the Plate interface the connection is between boundaries rather than
edges as is the case in the Shell interface.

Several different types of periodicity properties of the solution can be prescribed using
this boundary condition.

• The Continuity, Antiperiodicity, and User defined periodic conditions directly

prescribe relations both between displacements and between rotations. They can be
used for any type of study.
• The Floquet periodicity can be used for frequency domain problems with a spatial
periodicity of the geometry and solution. The modeled structure is typically a unit
cell of a repetitive structure.
• The Cyclic symmetry is a special case of a Floquet condition, intended for structures
which consist of a number of sectors which are identical when rotated around a
common axis, like in a fan.

The two sets of edges between which there is a periodicity condition are called the
source and destination respectively. It is not required to have the same mesh on the

654 | CHAPTER 5: SHELL AND PLATE

source and destination, but the local accuracy of the solution near the edges will be
better if you use the same mesh.

For periodic conditions on shells the periodicity condition acts on edges,

as opposed to solids and plates where it acts on boundaries. This means
that the orientation cannot be determined automatically. You must
provide coordinate systems using the Orientation of Source and Orientation
of Destination sections respectively. The default coordinate system is the
Global coordinate system, which works well if the edges are parallel. In
other cases, you need to add a Destination Selection subnode, in order to
supply the coordinate system for the destination.

In cases of rotational symmetry, you can assign the same cylindrical

coordinate system to both source and destination, as long as the edges do
not intersect at the longitudinal axis.

When there is a common point on the axis of revolution, you should

explicitly constrain it to remain on the axis and to have no rotations using
Prescribed Displacement/Rotation.

EDGE SELECTION
Select both the source and destination edges.

The software automatically identifies the edges as either source edges or destination
edges. This works fine for cases like opposing parallel edges. In more general cases, use
the Destination Selection subnode to specify the edges which constitute the destination.
By default this node contains the selection that COMSOL Multiphysics has identified.

In cases where the periodic edge is split into several edges within the geometry, it
might be necessary to apply separate periodic conditions to each pair of geometry
edges for the matching to work properly.

In the Plate interface, Edge Selection is replaced by Boundary Selection.

THE SHELL AND PLATE INTERFACES | 655

 PERIODICITY SETTINGS
With Type of periodicity you select the form of periodicity that your solution should
have.

• For Continuity the displacements and rotations on the destination are set equal to
their counterparts on the source; u ( x d ) = u ( x s ) and a ( x d ) = a ( x s ) . If the
source and destination objects are rotated with respect to each other, a
transformation is performed using the selected coordinate systems, so that
corresponding components of the degrees of freedom are connected.
• For Antiperiodicity the displacements and rotations on the destination are set equal
to their counterparts on the source but with the sign reversed; u ( x d ) = – u ( x s )
and a ( x d ) = – a ( x s ) . If the source and destination objects are rotated with respect
to each other, a transformation is performed using the selected coordinate systems,
so that corresponding components of the degrees of freedom are connected.
• For Floquet periodicity enter a k-vector for Floquet periodicity kF. This is the wave
number vector for the excitation.
• For Cyclic symmetry the settings differ slightly between the Plate and Shell interfaces.

In the Plate interface, chose how to define the sector angle that the
geometry represents using Sector angle. If Automatic is selected, the
program attempts to find out how many full repetitions of the geometry
there will be on a full revolution. If User defined is selected, enter a value
for the sector angle θS.

If any point on the edges having the periodic condition is located on the
axis of cyclic symmetry, enter the Axis direction vector, tc. This orientation
of the axis of cyclic symmetry is then needed for eliminating conflicting
constraints.

In the Shell interface, you always must enter a value for the sector angle
θS.

In either case, also enter an Azimuthal mode number for the mode to be studied. It
can vary from 0 to N/2, where N is the total number of sectors on a full revolution.
• For User defined select the check box for any of the displacement or rotation
components as needed. Then for each selection, choose the Type of periodicity—
Continuity or Antiperiodicity. Each selected displacement component will be

656 | CHAPTER 5: SHELL AND PLATE

 connected by u i ( x d ) = u i ( x s ) or u i ( x d ) = – u i ( x s ) respectively. Each selected
rotation component will be connected by a i ( x d ) = a i ( x s ) or a i ( x d ) = – a i ( x s )
If the source and destination objects are rotated with respect to each other, a
transformation is performed using the selected coordinate systems so that
corresponding components of the degrees of freedom are connected.

CONSTRAINT SETTINGS
To display this section, click the Show button ( ) and select Advanced Physics Options.

In the COMSOL Multiphysics Reference Manual:

• Constraint Reaction Terms

• Weak Constraints
• Constraint Settings

ORIENTATION OF SOURCE
In Transform to intermediate map, select a coordinate system representing the
orientation of the degrees of freedom on the source selection. The corresponding
setting for the destination is given in the Destination Selection subnode.

• Cyclic Symmetry and Floquet Periodic Conditions in the Structural

Mechanics Theory chapter.
• Orientation of Source and Destination in the COMSOL Multiphysics
Reference Manual.

Vibrations of an Impeller: Application Library path

Structural_Mechanics_Module/Dynamics_and_Vibration/impeller

LOCATION IN USER INTERFACE

Context Menus
Shell>Connections>Periodic Condition
Plate>Periodic Condition

Shell>Connections>Periodic Condition>Destination Selection

Plate>Periodic Condition>Destination Selection

THE SHELL AND PLATE INTERFACES | 657

 Ribbon
Physics tab with Shell selected:
Edges>Connections>Periodic Condition

Physics tab with Plate selected:

Boundaries>Periodic Condition

Physics tab with Periodic Condition node selected in the model tree:
Attributes>Destination Selection

Solid Connection

The Solid Connection node is obsolete, and has been superseded by the
Solid-Shell Connection multiphysics coupling. It cannot be added in
version 5.3 and later, but may be present in models created by earlier
versions of the software.

You are advised to update your model to use the new multiphysics
coupling instead, since the Solid Connection node will be removed in
future versions.

In 3D, a shell can be coupled to a solid by adding a Solid Connection node in the Shell
interface and a Shell Connection node in the Solid Mechanics interface. The
connection can either be from an edge of the shell to a boundary on the solid, or
between two boundaries. The first case is intended for modeling a transition from a
shell to a solid where shell assumptions are valid on both sides of the connection. The
second case is for adding a shell on top of a solid. Select this feature from the
Connections submenu.

SOLID CONNECTION (EDGES ONLY)

Select a Connection type—Softened or Simplified. When using Softened (the default),
three extra degrees of freedom are added to each selected point. This allows for a more
accurate description of the transition, but the model can in some cases become
underconstrained if the mesh on the solid is very coarse. The Simplified version of the
connection just adds constraints to the boundary of the solid, and in general causes
local disturbances of the stress field.

658 | CHAPTER 5: SHELL AND PLATE

When using the Solid Connection node for a boundary, the connection is intended for
placing a shell as a cladding on the boundary of a solid. No other settings except the
boundary selection are required.

• For more information about coupling different element types, see

Coupling Techniques.
• For details about the formulation of this coupling, see Connection
Between Shells and Solids

LOCATION IN USER INTERFACE

Context Menus
Shell>Connections>Solid Connection

Ribbon
Physics tab with Shell selected:
Boundaries>Connections>Solid Connection
Edges>Connections>Solid Connection

Beam Connection

The Beam Connection node is obsolete, and has been superseded by the
Shell-Beam Connection multiphysics coupling. It cannot be added in
version 5.3 and later, but may be present in models created by earlier
versions of the software.

You are advised to update your model to use the new multiphysics
coupling instead, since the Beam Connection node will be removed in
future versions.

A shell can be connected to a beam by adding a Beam Connection node in the Shell
interface and a Shell Connection node in the Beam interface.

The connection can be between:

• Two edges, one in each interface

THE SHELL AND PLATE INTERFACES | 659

 • A point in the Beam interface and a boundary in the Shell interface
• A point in the Beam interface and an edge in the Shell interface

• For more information about coupling different element types, see

Coupling Techniques.
• For details about the formulation of this coupling, see Connection
Between Shells and Beams

BEAM CONNECTION (EDGES)

Select a Connected entity—Beam point or Beam edge.

When Beam point is selected, select the name of the corresponding Shell Connection
node defined at the point level in the Beam interface to specify the connected parts.

For Beam point select an option from the Connected region list—Selected edges (the
default), Distance (automatic), Distance (manual), or Connection criterion.

• Using Selected edges makes all selected edges rigidly connected to the point on the
beam.
• For Distance (automatic), all parts on the shell edge, which are within the default
distance from the beam point, are connected. This distance is determined by the
cross section properties of the beam. It is contained in the variable beam.re in The
Beam Interface.
• For Distance (Manual) enter a Connection radius rc. All parts on the shell edge, which
are within the given distance from the beam point, are connected.
• For Connection criterion enter a Boolean expression in the text field. The beam is
connected to the selected shell edge wherever the expression has a nonzero value.
The default value is 1, which is equivalent to using the Selected edges option.

For Beam edge select the name of the corresponding Shell Connection node defined at
the edge level in the Beam interface to specify the connected parts. Select an Edge
type—Shared (the default) or Parallel.

• For Shared select an Offset definition—Along shell normal (the default) or Offset
vector. For Along shell normal enter an Offset ζ. For Offset vector enter values for d0
in the table. The offset is the vector from the reference surface of the shell to the

660 | CHAPTER 5: SHELL AND PLATE

 actual position of the beam. The Offset vector is interpreted in the coordinate system
selected in the Coordinate System Selection section.
• For Parallel enter a value for the Parallelism tolerance Δθ. The default is 1 degree.
The the two edges are connected only where they are parallel within the specified
tolerance.

BEAM CONNECTION (BOUNDARIES)

Select the name of the corresponding Shell Connection node defined at the point level
in the Beam interface to specify the connected parts.

Select an option from the Connected region list—Selected boundaries (the default),
Distance (automatic), Distance (manual), or Connection criterion.

• Using Selected boundaries makes all selected boundaries rigidly connected to the
point on the beam.
• For Distance (automatic), all parts on the shell boundary, which are within the a
default distance from the beam point, are connected. This distance is determined by
the cross section properties of the beam. It is contained in the variable beam.re in
the Beam interface.
• For Distance (manual) enter a Connection radius rc. All parts on the shell edge, which
are within the given distance from the beam point, are connected.
• For Connection criterion enter a Boolean expression in the text field. The beam is
connected to the selected shell boundary wherever the expression has a nonzero
values. The default value is 1, which is equivalent to using the Selected edges option.

LOCATION IN USER INTERFACE

Context Menus
Shell>Connections>Beam Connection

Ribbon
Physics tab with Shell selected:
Boundaries>Connections>Beam Connection
Edges>Connections>Beam Connection

Rigid Connector
The Rigid Connector is a boundary condition for modeling rigid regions and kinematic
constraints such as prescribed rigid rotations.The selected shell edges will move as a
single rigid object, irrespective of whether they are geometrically adjacent or not.

THE SHELL AND PLATE INTERFACES | 661

 If the study step is geometrically nonlinear, the rigid connector takes finite rotations
into account. The feature is similar to the rigid connectors in the Solid Mechanics and
Beam interfaces. Rigid connectors from the Shell, Beam and Solid Mechanics
interfaces can be attached to each other.

You can add functionality to the rigid connector through the following subnodes:

• Applied Force (Rigid Connector) to apply a force in given point.

• Applied Moment (Rigid Connector) to apply a moment.
• Mass and Moment of Inertia (Rigid Connector) to add extra mass and moment of
inertia in a given point.
• Spring Foundation (Rigid Connector) to add a translational or rotational spring or
damper in a given point.

COORDINATE SYSTEM SELECTION

The Global coordinate system is selected by default. The Coordinate system list contains
any additional coordinate systems that the model includes. Prescribed displacements
or rotations are specified along the axes of this coordinate system. It is also used for
defining the axis directions of the moment of inertia tensor of the Mass and Moment of
Inertia subnode.

CENTER OF ROTATION
The center of rotation serves two purposes.

• If you prescribe the displacement of the rigid connector, this is the place where it is
fixed.
• Results are interpreted with respect to the center of rotation.

Select a Center of rotation—Automatic, Centroid of selected entities, or User defined.

• For Automatic the center of rotation is at the geometrical center of the selected
edges. The constraints are applied at the center of rotation.
• For Centroid of selected entities select Entity level—Edge or Point. A subnode for
selection of the entities is added to the Model Builder. The center of rotation is
located at the centroid of the selected entities, which do not need to be related to

662 | CHAPTER 5: SHELL AND PLATE

 the edges to which the rigid connector is attached. As a special case, you can select
a single point, and thus locate the center of rotation at a certain point.
• For User defined, in the Global coordinates of center of rotation XC table enter
coordinates based on space dimension.

Once Centroid of selected entities is chosen, a default Center of Rotation:

Edge or Center of Rotation: Point subnode is added, depending on the
setting of Entity level.

Select the Offset check box to add an optional offset vector to the definition of the
center of rotation. Enter values for the offset vector Xoffset.

The center of rotation used is the sum of the vector obtained from any of the input
methods and the offset vector.

X C = X C,input + X offset

PRESCRIBED DISPLACEMENT AT CENTER OF ROTATION

To define a prescribed displacement for each spatial direction x, y, and z select one or
all of the Prescribed in X, Prescribed in Y, and Prescribed in Z direction check boxes. Then
enter a value or expression for the prescribed displacements u0, v0, or w0.

PRESCRIBED ROTATION AT CENTER OF ROTATION

Select an option from the By list—Free (the default), Constrained rotation, or Prescribed
rotation at center of rotation.

• For Constrained rotation select one or more of the Constrain rotation about X,
Constrain rotation about Y, and Constrain rotation about Z axis check boxes in order

THE SHELL AND PLATE INTERFACES | 663

 to enforce zero rotation about the corresponding axis in the selected coordinate
system.
• For Prescribed rotation at center of rotation enter an Axis of rotation Ω and an Angle
of rotation φ . The axis of rotation is given in the selected coordinate system.

• You can add a Harmonic Perturbation subnode for specifying a

harmonic variation of the values of the prescribed displacements and
rotations in a frequency domain analysis of perturbation type.
• You can activate and deactivate the rigid connector by assigning it to a
constraint group. See Load Cases in the Structural Mechanics
Modeling chapter.
• You can assign the value of the prescribed displacement and rotation to
a load group. See Load Cases in the Structural Mechanics Modeling
chapter.

• Rigid Connector Theory

• Harmonic Perturbation
• Load Cases

LOCATION IN USER INTERFACE

Context Menus
Shell>Connections>Rigid Connector

Ribbon
Physics tab with Shell selected:
Edges>Connections>Rigid Connector

Thermal Expansion (Rigid Connector)

Add the Thermal Expansion subnode to prescribe a deformation of the rigid connector
caused by changes in temperature. This makes it possible to reduce stresses caused by
the rigid connector being rigid, while there are thermal deformations in the flexible
shell to which it is attached.

The thermal strain depends on the coefficient of thermal expansion α, the temperature
T, and the strain-free reference temperature Tref as

664 | CHAPTER 5: SHELL AND PLATE

 ε th = α ( T – T ref )

THERMAL EXPANSION PROPERTIES

Specify the thermal properties that define the thermal strain in the rigid connector.

Select Inherit from boundary to take the thermal expansion data from the boundary to
which it is attached. This should only be used when:

• The temperature and the thermal expansion coefficient do not have a spatial
variation.
• The virtual material in the rigid connector has the same thermal expansion as the
domain itself.

When Inherit from boundary is not selected, enter:

• A value or expression for Temperature T, specifying the temperature distribution of

the rigid connector. Any spatial variation must be an explicit function of the material
frame coordinates. It is not possible to use a computed temperature distribution.
• The Coefficient of thermal expansion α. As a default, values From material are used.
This requires that a material has been assigned to the boundaries of the rigid
connector selection.
For User defined select Isotropic, Diagonal or Symmetric to enter one or more
components for a general coefficient of thermal expansion tensor α. When a
non-isotropic coefficient of thermal expansion is used, the axis orientations are given
by the coordinate system selected in the Coordinate System Selection section.
• A value or expression for the Strain reference temperature Tref which is the
temperature at which there are no thermal displacements in the rigid connector..

• Constraints and Thermal Expansion in the Structural Mechanics

Modeling chapter.
• Thermal Expansion of Constraints in the Structural Mechanics Theory
chapter.

LOCATION IN USER INTERFACE

Context Menus
Shell>Rigid connector>Thermal Expansion

THE SHELL AND PLATE INTERFACES | 665

 Ribbon
Physics tab with Rigid connector node selected in the model tree:
Attributes>Thermal Expansion

Attachment
The Attachment node is used to define a set of edges on a flexible shell which can be
used to connect it with other components through a joint in the Multibody Dynamics
interface. All the selected edges behave as if they were connected by a common rigid
body.

Attachments can be added to boundaries in a Multibody Dynamics or a Solid

Mechanics interface, to edges in a Shell interface, or to points in a Beam interface. This
makes it possible to use a joint in the Multibody Dynamics interface for connecting
parts modeled in different physics interfaces.

The Attachment node is available with the addition of the Multibody

Dynamics Module. The Attachments is included in the Multibody
Dynamics Module User’s Guide.

LOCATION IN USER INTERFACE

Context Menus
Shell>Connections>Attachment

Ribbon
Physics tab with Shell selected:
Edges>Connections>Attachment

Thermal Expansion (Attachment)

Add the Thermal Expansion subnode to prescribe a deformation of the attachment
caused by changes in temperature. This makes it possible to reduce stresses caused by
the attachment being rigid, while there are thermal deformations in the flexible shell
to which it is attached.

The thermal strain depends on the coefficient of thermal expansion α, the temperature
T, and the strain-free reference temperature Tref as

ε th = α ( T – T ref )

666 | CHAPTER 5: SHELL AND PLATE

THERMAL EXPANSION PROPERTIES
Specify the thermal properties that define the thermal strain in the attachment.

Select Inherit from boundary to take the thermal expansion data from the boundary to
which it is attached. This should only be used when:

• The temperature and the thermal expansion coefficient do not have a spatial
variation.
• The virtual material in the attachment has the same thermal expansion as the
boundary itself.

When Inherit from boundary is not selected, enter:

• A value or expression for Temperature T, specifying the temperature distribution of

the attachment. Any spatial variation must be an explicit function of the material
frame coordinates. It is not possible to use a computed temperature distribution.
• The Coefficient of thermal expansion α. As a default, values From material are used.
This requires that a material has been assigned to the boundaries of the attachment
selection.
For User defined select Isotropic, Diagonal or Symmetric to enter one or more
components for a general coefficient of thermal expansion tensor α. When a
non-isotropic coefficient of thermal expansion is used, the axis orientations are given
by the coordinate system selected in the Coordinate System Selection section.
• A value or expression for the Strain reference temperature Tref which is the
temperature at which there are no thermal displacements in the attachment.

• Constraints and Thermal Expansion in the Structural Mechanics

Modeling chapter.
• Thermal Expansion of Constraints in the Structural Mechanics Theory
chapter.

LOCATION IN USER INTERFACE

Context Menus
Solid Mechanics>Attachment>Thermal Expansion
Multibody Dynamics>Attachment>Thermal Expansion

Ribbon
Physics tab with Attachment node selected in the model tree:
Attributes>Thermal Expansion

THE SHELL AND PLATE INTERFACES | 667

 Phase
You can add a Phase subnode to nodes which define a load in order to prescribe the
phase angle in a frequency domain analysis.

For modeling the frequency response the physics interface splits the harmonic load
into two parameters:

• The amplitude, F, which is specified in the node for the load.

• The phase ( φ ), which is specified in the Phase subnode.

Together these define a harmonic load, for which the amplitude and phase shift can
vary with the excitation frequency, f:

F freq = F ( f ) ⋅ cos ( 2πft + φ )

PHASE
Add the phase angle Fph for harmonic loads. Enter the phase for each component of
the load in the corresponding fields.

MOMENT LOAD PHASE

Add the phase for the moment load Mph for harmonic loads. Enter the phase for each
component of the moment load in the corresponding fields.

LOCATION IN USER INTERFACE

Context Menus
Shell>Body Load>Phase
Shell>Face Load>Phase
Shell>Edge Load>Phase
Shell>Point Load>Phase
Shell>Rigid Connector>Applied Force>Phase
Shell>Rigid Connector>Applied Moment>Phase

Ribbon
Physics tab with Body Load, Face Load, Edge Load, Point Load, Applied Force, or Applied
Moment selected:
Attributes>Phase

668 | CHAPTER 5: SHELL AND PLATE

Harmonic Perturbation
Use the Harmonic Perturbation subnode to specify the harmonic part of non-zero
prescribed displacements. This node is used if the study step is frequency response of
a perturbation type.

The settings are the same as in the parent Prescribed Displacement/Rotation,

Prescribed Velocity, Prescribed Acceleration, or Rigid Connector node. Only degrees
of freedom selected as prescribed in the parent node can be assigned a value.

See Harmonic Perturbation in the Structural Mechanics Modeling

chapter.

LOCATION IN USER INTERFACE

Context Menus
Shell>Prescribed Displacement/Rotation>Harmonic Perturbation
Shell>Prescribed Velocity>Harmonic Perturbation
Shell>Prescribed Acceleration>Harmonic Perturbation
Shell>Rigid Connector>Harmonic Perturbation
Plate>Prescribed Displacement/Rotation>Harmonic Perturbation
Plate>Prescribed Velocity>Harmonic Perturbation
Plate>Prescribed Acceleration>Harmonic Perturbation

Ribbon
Physics tab with Prescribed Displacement/Rotation, Prescribed Velocity, Prescribed
Acceleration, or Rigid Connector node selected in the model tree:
Attributes>Harmonic Perturbation

THE SHELL AND PLATE INTERFACES | 669

670 | CHAPTER 5: SHELL AND PLATE
 6

Membrane

This chapter describes the Membrane (mbrn) interface, which is found under the
Structural Mechanics branch ( ) when adding a physics interface.

In this chapter:

• Theory for the Membrane Interface

• The Membrane Interface

671
 Theory for the Membrane Interface
The theory for the Membrane interface is described in this section:

• About Membranes
• Theory Background for the Membrane Interface

About Membranes
Membranes can be considered as plane stress elements in 3D with a possibility to
deform both in the in-plane and out-of-plane directions. The difference between a
shell and a membrane is that the membrane does not have any bending stiffness. If the
ratio between the thickness and the dimensions in the other directions becomes very
small, a membrane formulation is numerically better posed than a shell formulation.

Theory for the Membrane Interface supports the same study types as the Solid
Mechanics interface except it does not include the Linear Buckling study type.

To describe a membrane, provide its thickness and the material properties. All
properties can be variable over the element. All elemental quantities are integrated only
at the midsurface. This is a good approximation since by definition a membrane is thin.

The physics interface is intended to model either prestressed membranes or a thin

cladding on top of a solid.

STIFFNESS IN THE NORMAL DIRECTION

When membrane elements are used separately, not supported by other structural
elements, a prestress is necessary in order to avoid a singularity. The unstressed
membrane has no stiffness in the normal direction. It is the geometrically nonlinear
effects (stress stiffening) which supply the out-of-plane stiffness. A prestress can be
given either through initial stress and strain or through a tensile boundary load.
Prestress is not necessary in cases where inertia effects are included in a dynamic
analysis. A small prestress can, however, still be useful to stabilize the analysis in the
initial state. In order to obtain the prestress effect, you must select Include geometric
nonlinearity in the settings for the study step.

MEMBRANES FOR 3D MODELS

The Membrane interface in 3D can be active on internal and external boundaries of a
domain, as well as on boundaries not adjacent to any domain.

672 | CHAPTER 6: MEMBRANE

The dependent variables are the displacements u, v, and w in the global x, y, and z
directions, and the displacement derivative unn in the direction normal to the
membrane.

MEMBRANES FOR 2D AXISYMMETRIC MODELS

The Membrane interface for 2D axisymmetric models can be active on internal and
external boundaries of a solid, as well as on edges that not adjacent to a solid.

The dependent variables are the displacements u and w in the global r and z directions,
and the displacement derivative unn in the direction normal to the membrane in r-z
plane.

Theory Background for the Membrane Interface

A 3D membrane is similar to a shell but it has only translational degrees of freedom
and the results are constant in the thickness direction.

The thickness of the membrane is d, which can vary over the element. The
displacements are interpolated by Lagrange shape functions.

A 2D axisymmetric membrane is similar to the 3D membrane and it has a nonzero

circumferential strain in the out-of-plane direction.

LOCAL COORDINATE SYSTEMS

Many quantities for a membrane can best be interpreted in a local coordinate system
aligned to the membrane surface. Material data, initial stresses-strains, and constitutive
laws are always represented in the local coordinate system.

This local membrane surface coordinate system is defined by the boundary coordinate
system (t1, t2, n).

The quantities like stresses and strains are also available as results in the global
coordinate system after a transformation from a local (boundary) system.

STRAIN-DISPLACEMENT RELATION
The kinematic relations of the membrane element are first expressed along the global
coordinate axes. The strains are then transformed to the element local direction. Since
the membrane is defined only on a boundary, derivatives in all spatial directions are not
directly available. This makes the derivation of the strain tensor somewhat different
from what is used in solid mechanics.

THEORY FOR THE MEMBRANE INTERFACE | 673

 The deformation gradient F is in general defined as the gradient of the current
coordinates with respect to the original coordinates:

∂x ∂u
F = ------- = I + -------
∂X ∂X

In the Membrane interface, a tangential deformation gradient is computed as

t
F T = I – N ⋅ N + ∇T u

Here ∇Tu is a displacement gradient computed using the tangential derivative

operator, and N is the normal vector to the undeformed membrane. FT now contains
information about the stretching in the plane of the membrane.

The Right Cauchy-Green tensor C is generally defined as

T
C = F F

Since the tangential deformation gradient does not contain any information about the
transversal stretch λn, it must be augmented when creating C.

t 2 t t 2 t
C = F T F T + λ n N ⋅ N = F T F T + ( 1 + u nn ) N ⋅ N

Similar augmentation is also needed in the full deformation gradient tensor F.

t
F = I + ∇ T u + u nn N ⋅ N

From C, the Green-Lagrange strains are computed using the standard expression

1
E = --- ( C – I )
2

The local tangential strains in the membrane are calculated by transformation of this
strain tensor into the local tangent plane coordinate system.

The Jacobian J is the ratio between the current volume and the initial volume. In full
3D it is defined as

J = det ( F )

In the membrane, only the C tensor is available, so instead the following expression is
used:

J = det ( C )

674 | CHAPTER 6: MEMBRANE

The area scale factor is also computed as

J
J A = -------------------
1 + u nn

In the case of geometrically linear analysis, a linearized version of the strain tensor is
used.

CONSTITUTIVE RELATION AND WEAK EXPRESSIONS

The constitutive relations for the membrane on the reference surface are similar to
those used in the Solid Mechanics interface.

The thermal strains and initial stresses-strains (only for the in-plane directions of the
membrane) are added in the constitutive relation in a similar manner as it is done in
Solid Mechanics.

The weak expressions in the Membrane interface are similar to that of linear elastic
continuum mechanics.

See also Analysis of Deformation in the documentation of the Solid

Mechanics interface.

EXTERNAL LOADS
Contributions to the virtual work from the external load are of the form

u test ⋅ F

where the forces (F) can be distributed over a boundary or an edge or be concentrated
in a point. In the special case of a follower load, defined by its pressure p, the force
intensity is F = – pn where n is the normal in the deformed configuration.

For a follower load, the change in midsurface area is taken into account, and
integration of the load is done in the spatial frame.

STRESS CALCULATIONS
The stresses are computed by applying the constitutive law to the computed strains.

The membrane does not support transverse and bending forces so the only section
forces it support is the membrane force defined as:

N = d⋅s

THEORY FOR THE MEMBRANE INTERFACE | 675

 where s is the local stress tensor and contains only in-plane stress components.

676 | CHAPTER 6: MEMBRANE

The Membrane Interface
The Membrane (mbrn) interface ( ), found under the Structural Mechanics
branch ( ) when adding a physics interface, is mainly used to model prestressed
membranes, but can also be used to model a thin cladding on a solid. Membranes can
be considered as plane stress elements on boundaries in 3D with a possibility to deform
both in the in-plane and out-of-plane directions. There is also a version of the
membrane interface for 2D axisymmetric problems. The membrane interface is then
applicable to lines since that is what represents boundaries.

The difference between a shell and a membrane is that the membrane does not have
any bending stiffness. In most applications, the membrane is used by itself and not as
a cladding. A tensile prestress is then necessary in order to avoid singularity because a
membrane with no stress or compressive stress has no transverse stiffness. To include
the prestress effect, you must enable geometric nonlinearity for the study step.

The Linear Elastic Material is the default material, which adds a linear elastic equation
for the displacements and has a Settings window to define the elastic material
properties. This material model can also be combined with viscoelasticity.

With the Nonlinear Structural Materials Module, you can also model Nonlinear Elastic
and Hyperelastic materials, and add options such as Plasticity, Creep, and
Viscoplasticity.

For a detailed overview of the functionality available in each product, visit

http://www.comsol.com/products/specifications/

When this physics interface is added, these default nodes are also added to the Model
Builder: Linear Elastic Material, Free (a condition where edges are free, with no loads or
constraints), and Initial Values. In the case if axial symmetry, an Axial Symmetry node is
also added. From the Physics toolbar, you can then add other nodes that implement,
for example, loads and constraints. You can also right-click Membrane to select physics
features from the context menu.

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern

THE MEMBRANE INTERFACE | 677

 <name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is mbrn.

THICKNESS
Define the Thickness d by entering a value or expression in the field. The default is
0.0001 m. Use the Change Thickness node to define a different thickness in parts of
the membrane. The thickness can be variable if an expression is used.

S T R U C T U R A L TR A N S I E N T B E H AV I O R
From the Structural transient behavior list, select Include inertial terms (the default) or
Quasi-static. Use Quasi-static to treat the elastic behavior as quasi-static (with no mass
effects; that is, no second-order time derivatives). Selecting this option gives a more
efficient solution for problems where the variation in time is slow when compared to
the natural frequencies of the system.

REFERENCE PO INT F OR MO MENT COMPUTATION

Enter the default coordinates for the Reference point for moment computation xref. The
resulting moments (applied or as reactions) are then computed relative to this
reference point. During the results and analysis stage, the coordinates can be changed
in the Parameters section in the result nodes.

DEPENDENT VARIABLES
The dependent variable (field variable) is for the Displacement field u which has three
components (u, v, and w). The name can be changed but the names of fields and
dependent variables must be unique within a model.

DISCRETIZATION
To display this section, click the Show button ( ) and select Discretization.

In the Membrane interface you can choose not only the order of the discretization, but
also the type of shape functions: Lagrange or serendipity. For highly distorted
elements, Lagrange shape functions provide better accuracy than serendipity shape
functions of the same order. The serendipity shape functions will however give
significant reductions of the model size for a given mesh containing quadrilateral
elements.

678 | CHAPTER 6: MEMBRANE

The default is to use Quadratic serendipity shape functions for the Displacement field.

• Boundary, Edge, Point, and Pair Nodes for the Membrane Interface
• Theory for the Membrane Interface
• Selecting Discretization

Vibrating Membrane: Application Library path

Structural_Mechanics_Module/Verification_Examples/vibrating_membrane

Boundary, Edge, Point, and Pair Nodes for the Membrane Interface
The Membrane Interface has these boundary, edge, point, and pair nodes available
from the Physics ribbon toolbar (Windows users), Physics context menu (Mac or Linux
users), or right-click to access the context menu (all users)..

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

FEATURES AVAILABLE FROM SUBMENUS

Many features for the Membrane interface are added from submenus in the Physics
toolbar groups or context menu (when you right-click the node). The submenu name
is the same in both cases.

• The submenus at the Boundary level are Material Models, Face and Volume Loads, Mass,
Spring, and Damper, and Face Constraints.
• The submenus at the Edge (3D) or Points (2D axisymmetric) level are Mass, Spring,
and Damper, More Constraints, and Pairs.
• For 3D components, there is also a Points submenu.

LINKS TO FEATURE NODE INFORMATION

These nodes (and subnodes) are described in this section (listed in alphabetical order):

• Initial Stress and Strain • Linear Elastic Material

• External Stress

THE MEMBRANE INTERFACE | 679

 These nodes are described for the Solid Mechanics interface:

• Added Mass • Plasticity

• Antisymmetry • Point Load
• Body Load1 • Point Load (on Axis)
• Creep • Predeformation
• Damping • Prescribed Acceleration
• Edge Load3 • Prescribed Displacement2
• External Stress • Prescribed Velocity
3
• Face Load • Ring Load
• Fixed Constraint • Rotating Frame1
• Free • Spring Foundation
1
• Gravity • Symmetry
• Hygroscopic Swelling • Thermal Expansion (for constraints)
• Hyperelastic Material • Thermal Expansion (for materials)
• Initial Values • Viscoelasticity
• Nonlinear Elastic Material • Viscoplasticity
• Phase
1
This is selected from the Face and Volume Loads submenu for this interface.
2
At the boundary level, this is selected from the Face Constraints submenu for this
interface.
3
Described for the Shell interface.

Harmonic Perturbation, Prestressed Analysis, and Small-Signal Analysis

in the COMSOL Multiphysics Reference Manual

In the COMSOL Multiphysics Reference Manual see Table 2-3 for links
to common sections and Table 2-4 to common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

680 | CHAPTER 6: MEMBRANE

Linear Elastic Material
The Linear Elastic Material node adds the equations for a linear elastic membrane and
an interface for defining the elastic material properties.

By adding the following subnodes to the Linear Elastic Material node you can
incorporate many other effects:

• Thermal Expansion (for materials)

• Hygroscopic Swelling
• Initial Stress and Strain
• External Stress
• Damping
• Viscoelasticity
• Plasticity
• Creep
• Viscoplasticity

Note: Some options are only available with certain COMSOL products (see http://
www.comsol.com/products/specifications/)

COORDINATE SYSTEM SELECTION

The built in Boundary System 1 is selected by default. The Coordinate system list
contains any additional boundary coordinate systems that the model includes. The
coordinate system is used for interpreting directions of orthotropic and anisotropic
material data and when stresses or strains are presented in a local system. Many of the
possible subnodes inherit the coordinate system settings.

LINEAR ELASTIC MATERIAL

To use a mixed formulation by adding the pressure as an extra dependent variable to
solve for, select the Nearly incompressible material check box. For a material with a very
low compressibility, using only displacements as degrees of freedom may lead to a
numerically ill-posed problem.

Define the Solid model and the linear elastic material properties.

Solid Model
To use a mixed formulation by adding the pressure as an extra dependent variable to
solve for, select the Nearly incompressible material check box.

THE MEMBRANE INTERFACE | 681

 Select a linear elastic Solid model—Isotropic (the default), Orthotropic, or Anisotropic.
Select:

• Isotropic for a linear elastic material that has the same properties in all directions.
• Orthotropic for a linear elastic material that has different material properties in
orthogonal directions, so that its stiffness depends on the properties Ei, νij, and Gij.
• Anisotropic for a linear elastic material that has different material properties in
different directions, and the stiffness comes from the symmetric elasticity matrix, D.

• Material Models
• Linear Elastic Material
• Orthotropic and Anisotropic Materials

Density
The default Density ρ uses values From material. For User defined enter another value or
expression.

The density is needed for dynamic analysis or when the elastic data is given
in terms of wave speed. It is also used when computing mass forces for
gravitational or rotating frame loads, and when computing mass
properties (Computing Mass Properties).

Specification of Elastic Properties for Isotropic Materials

For an Isotropic Solid model, from the Specify list select a pair of elastic properties for
an isotropic material—Young’s modulus and Poisson’s ratio, Young’s modulus and shear
modulus, Bulk modulus and shear modulus, Lamé parameters, or Pressure-wave and
shear-wave speeds. For each pair of properties, select from the applicable list to use the
value From material or enter a User defined value or expression.

Each of these pairs define the elastic properties and it is possible to convert
from one set of properties to another (see Table 3-1 in the theory
chapter).

The individual property parameters are:

• Young’s modulus (elastic modulus) E.

• Poisson’s ratio ν.

682 | CHAPTER 6: MEMBRANE

• Shear modulus G.
• Bulk modulus K.
• Lamé parameter λ and Lamé parameter μ.
• Pressure-wave speed (longitudinal wave speed) cp.
• Shear-wave speed (transverse wave speed) cs. This is the wave speed for a solid
continuum. In plane stress, for example, the actual speed with which a longitudinal
wave travels is lower than the value given.

Specification of Elastic Properties for Orthotropic Materials

When Orthotropic is selected from the Solid model list, the material properties vary in
orthogonal directions only. The Material data ordering can be specified in either
Standard or Voigt notation. When User defined is selected in 3D, enter three values in
the fields for Young’s modulus E, Poisson’s ratio ν, and the Shear modulus G. This
defines the relationship between engineering shear strain and shear stress. It is
applicable only to an orthotropic material and follows the equation

τ ij
ε ij = --------
G ij

νij is defined differently depending on the application field. It is easy to

transform among definitions, but check which one the material uses.

Specification of Elastic Properties for Anisotropic Materials

When Anisotropic is selected from the Solid model list, the material properties vary in
all directions, and the stiffness comes from the symmetric Elasticity matrix, D The
Material data ordering can be specified in either Standard or Voigt notation. When User
defined is selected, a 6-by-6 symmetric matrix is displayed.

GEOMETRIC NONLINEARITY
If a study step is geometrically nonlinear, the default behavior is to use a large strain
formulation in all boundaries. There are, however, some rare cases when the use of a
small strain formulation for a certain boundary is needed.

THE MEMBRANE INTERFACE | 683

 In such cases, select the Force linear strains check box. When selected, a small strain
formulation is always used, independently of the setting in the study step.

• Modeling Geometric Nonlinearity

• Studies and Solvers in the COMSOL Multiphysics Reference Manual

ENERGY DISSIPATION
The section is available when you also have the Nonlinear Structural Materials Module.
Then, to display this section, click the Show button ( ) and select Advanced Physics
Options.

Select the Calculate dissipated energy check box as needed to compute the energy
dissipated by Creep, Plasticity, Viscoplasticity or Viscoelasticity.

LOCATION IN USER INTERFACE

Context Menus
Membrane>Material Models>Linear Elastic Material

Ribbon
Physics tab with Membrane selected:
Boundaries>Material Models>Linear Elastic Material

Initial Stress and Strain

You can add the Initial Stress and Strain subnode to the Linear Elastic Material, in order
to specify the stress or strain state in the structure before applying any constraint or
load. The values given are not initial values in the mathematical sense, but rather a
contribution to the constitutive relation..

In many cases Initial Stress and Strain and External Stress are
interchangeable when prescribing stresses, but you can find some more
options in the latter.

For details about initial stresses and strains, see Initial Stresses and Strains.

684 | CHAPTER 6: MEMBRANE

COORDINATE SYSTEM SELECTION
The given initial stresses and strains are interpreted in this system.

For details about how local coordinate systems are used in the Membrane
interface, see Local Coordinate Systems.

INITIAL STRESS AND STRAIN

Specify the initial stress as the Initial local in-plane force N0 and the initial strain as the
Initial local in-plane strain ε0 (dimensionless). If you know the stress, rather than the
force per unit length, type in the stress multiplied by the membrane thickness
mbrn.d.In a geometrically nonlinear analysis, the stresses should be interpreted as
Second Piola-Kirchhoff stresses, and the strains should be interpreted as
Green-Lagrange strains.

• For details about initial stresses and strains, see Inelastic Strain
Contributions and Initial Stresses and Strains.
• For details about the different strain measures, see Deformation
Measures.
• For details about the different stress measures, see Defining Stress.

Prestressed Micromirror: Application Library path MEMS_Module/

Actuators/micromirror

LOCATION IN USER INTERFACE

Context Menus
Membrane>Linear Elastic Material>Initial Stress and Strain
Membrane>Nonlinear Elastic Material>Initial Stress and Strain

Ribbon
Physics tab with Linear Elastic Material or Nonlinear Elastic Material node selected in the
model tree:
Attributes>Initial Stress and Strain

THE MEMBRANE INTERFACE | 685

 External Stress
You can add the External Stress subnode to several material models, in order to specify
an additional stress contribution which is not part of the constitutive relation. The
external stress can be added to the total stress tensor, or act only as an extra load
contribution.

In many cases External Stress and Initial Stress and Strain are
interchangeable when prescribing stresses. In Initial Stress and Strain, the
given stress is however always added to the stress tensor.

EXTERNAL STRESS
Select a Stress input—Stress tensor (Material), Stress tensor (Spatial) or In-plane force.

• When Stress tensor (Material) is selected, you enter the external stress in the form of
a Second Piola-Kirchhoff stress tensor. The External stress tensor drop-down list will
contain all stress tensors announced by any physics interface, and also the entry User
defined. When User defined is selected, you can enter the data for the External stress
tensor Sext as Isotropic, Diagonal, or Symmetric depending on the properties of the
tensor. The tensor components are interpreted in the selected coordinate system. If
a stress tensor announced by a physics interface is selected, the coordinate system
setting is ignored — the orientation is handled internally. Choose a Contribution
type—Add to stress tensor or Load contribution only to determine the effect of the
contribution. Stress tensor components which are not in the plane of the membrane
will be ignored.
• When Stress tensor (Spatial) is selected, you enter the external stress in the form of
Cauchy stress tensor. The components are interpreted in the selected coordinate
system. Depending on the properties of the tensor, you can enter the data for the
External stress tensor σext as Isotropic, Diagonal, or Symmetric. Choose a Contribution
type—Add to stress tensor or Load contribution only to determine the effect of the
contribution. Stress tensor components which are not in the plane of the membrane
will be ignored.
• When In-plane force is selected, you enter the external stress in the form of section
forces. Specify the initial stress as values or expressions for the In-plane force Next,
which has components in the plane of the membrane, along the directions given by

686 | CHAPTER 6: MEMBRANE

 the selected local coordinate system. Choose a Contribution type—Add to stress
tensor or Load contribution only to determine the effect of the contribution.

Since all stress tensor representations coincide in a geometrically linear

analysis, Stress tensor (Spatial) is needed only in the case of a geometrically
nonlinear analysis. The stress tensor is entered using a Cauchy stress
tensor representation, and is internally transformed to a Second-Piola
stress tensor.

• For theory, see External Stress.

• For details about the different stress measures, see Defining Stress.

LOCATION IN USER INTERFACE

Context Menus
Membrane>Linear Elastic Material>External Stress
Membrane>Nonlinear Elastic Material>External Stress
Membrane>Hyperelastic Material>External Stress

Ribbon
Physics tab with Linear Elastic Material, Nonlinear Elastic Material, or Hyperelastic
Material node selected in the model tree:
Attributes>External Stress

THE MEMBRANE INTERFACE | 687

688 | CHAPTER 6: MEMBRANE
 7

Beam

This chapter describes the Beam interface, which you find under the Structural
Mechanics branch ( ) when adding a physics interface.

In this chapter:

• Theory for the Beam Interface

• The Beam Interface

689
 Theory for the Beam Interface
The Beam Interface theory is described in this section:

• About Beams • Stress-Strain Relation

• In-Plane Beams • Thermal Strain
• 3D Beam • Hygroscopic Swelling
• Geometric Variables • Initial Load and Strain
• Shape Functions • Implementation
• Geometric Nonlinearity • Stress Evaluation
• Strain-Displacement/Rotation • Common Cross Sections
Relation

Beam Cross Section

About Beams
A beam is a slender structure that can be fully described by its cross section properties
such as area, moments of inertia, and torsional constant. Beams are the choice for
modeling reinforcements in 3D solids and shell structures, as well as in 2D solids under
the plane stress assumption. Naturally, they can also model lattice works, both planar
and three-dimensional.

Beams can sustain forces and moments in any direction, both distributed and on
individual nodes. The beam’s ends and interconnections can be free, simply supported,
or clamped. In fact, the simplified boundary conditions are usually responsible for
most of the difference that can be found between a beam solution and a full 3D solid
simulation of the same structure. Point constraints on beams are well-behaved, in
contrast to the solid case, and it is possible to use discrete point masses and mass
moments of inertia.

The Beam interface is based on the principle of virtual work. The resulting equation
can equivalently be viewed as a weak formulation of an underlying PDE. The Beam

690 | CHAPTER 7: BEAM

interface uses special shape function classes to define stresses and strains in the beams
using either Euler-Bernoulli or Timoshenko theory.

In-Plane Beams
Use the Beam interface in 2D to analyze planar lattice works of uniaxial beams.

In-plane beams are defined on edges in 2D. They can be used separately or as stiffeners
to 2D solid elements.

VA R I A BL E S A N D S P A C E D I M E N S I O N S
The degrees of freedom (dependent variables) are the global displacements u and v in
the global x and y directions and the rotation θ about the global z-axis.

THEORY FOR THE BEAM INTERFACE | 691

 3D Beam
Use the Beam interface in 3D to model three-dimensional frameworks of uniaxial
beams.

3D beams are defined on edges in 3D. They can be used separately or as stiffeners to
2D solid or shell elements.

VAR IA BL ES AN D SPA C E DIM E NS IO N S

The degrees of freedom (dependent variables) are the global displacements u, v, w in
the global x, y, z directions and the global rotations θx, θy, and θz about the global x-,
y-, and z-axes.

692 | CHAPTER 7: BEAM

Figure 7-1: The degrees of freedom in the Beam interface,

Geometric Variables
The beam formulations are based on that the cross section data are known. You can
either enter them explicitly, or, for a number of standard cross sections, get them
computed internally. In Table 7-1 the important geometric variables used in the Beam
interface are summarized.
TABLE 7-1: GEOMETRIC VARIABLE IN BEAM INTERFACE

PROPERTY VARIABLE NAME DESCRIPTION

A beam.area Cross section area

Izz beam.Izz Moment of inertia around local z-axis
ez beam.ez Distance to shear center in local z-direction
μy beam.muy Max shear stress factor in local y-direction
κy beam.kappay Shear correction factor along local y-axis
rgy beam.rgy Radius of gyration, local y-direction

THEORY FOR THE BEAM INTERFACE | 693

 TABLE 7-1: GEOMETRIC VARIABLE IN BEAM INTERFACE

PROPERTY VARIABLE NAME DESCRIPTION

Iyy beam.Iyy Moment of inertia around local y-axis

ey beam.ey Distance to shear center in local y-direction
μz beam.muz Max shear stress factor in local z-direction
κz beam.kappaz Shear correction factor along local z-axis
rgz beam.rgz Radius of gyration, local z-direction
J beam.J_beam Torsional constant
Wt beam.Wt Torsional section modulus
re beam.re Equivalent radius
beam.yBeam1 Stress evaluation point 1, local coordinates
beam.zBeam1
beam.yBeam2 Stress evaluation point 2, local coordinates
beam.zBeam2
beam.yBeam3 Stress evaluation point 3, local coordinates
beam.zBeam3
beam.yBeam4 Stress evaluation point 4, local coordinates
beam.zBeam4

Shape Functions
The beam element has different shape functions for representing the displacements in
different directions.

• The axial extension is represented by a linear shape function.

• The twist around the beam axis (3D only) is represented by a linear shape function.
• The bending displacement and corresponding rotation is represented by cubic shape
functions, usually called Hermitian shape functions. These will supply exact
solutions to the underlying beam equations as long as distributed loads do not vary
with position.

The shape functions for bending depend on whether Timoshenko theory is employed
or not.

In the beam local system, the displacements, u, and rotations, θ, are interpolated as

694 | CHAPTER 7: BEAM

  
 u1 
 
 u   θ1 
  = [N ] 
 θ   u2 
 
 θ2 
 

where the subscript refers to the two nodes of the element, and N is a matrix of shape
functions.

N1 0 0 0 0 0 N2 0 0 0 0 0
0 N3 0 0 0 N5 0 N4 0 0 0 N6
0 0 N3 0 –N5 0 0 0 N4 0 –N6 0
[N] =
0 0 0 M1 0 0 0 0 0 M2 0 0
0 0 M3 0 M5 0 0 0 M4 0 M6 0
0 –M3 0 0 0 M5 0 –M4 0 0 0 M6

The shape functions for the Euler-Bernoulli case are expressed in the local coordinate
ξ, ranging from 0 to 1, as

N1 = 1 – ξ N2 = ξ
2 3 2 3 2 3 2 3
N 3 = 1 – 3ξ + 2ξ N 4 = 3ξ – 2ξ N 5 = L ( ξ – 2ξ + ξ ) N6 = L ( – ξ + ξ )
M1 = 1 – ξ M2 = ξ
6 2 6 2 2 2
M 3 = – ---- ( ξ – ξ ) M 4 = ---- ( ξ – ξ ) M 5 = 1 – 4ξ + 3ξ M 6 = – 2ξ + 3ξ
L L

where L is the length of the beam element.

For the Timoshenko case, the shape functions are modified, so that they depend on
the degree of shear flexibility. Define

12EI zz 24 ( 1 + ν )I zz
Φ 2 = ----------------------2 = -------------------------------
-
Gκ y AL κ y AL 2
12EI yy 24 ( 1 + ν )I yy
Φ 3 = ---------------------2- = --------------------------------
-
Gκ z AL κ z AL 2

which represent the ratios between bending and shear stiffness in the two principal
directions. The shape functions are then modified so that

THEORY FOR THE BEAM INTERFACE | 695

 ˆ
i ( N3 + Φi N3 ) i ( N 4 + Φ i Nˆ 4 )
N 3 = --------------------------------- N 4 = ---------------------------------
1 + Φi 1 + Φi
( N 5 + Φ i Nˆ 5 ) ˆ
i i ( N6 + Φi N6 )
N 5 = --------------------------------- N 6 = ---------------------------------
1 + Φi 1 + Φi
( M 5 + Φ i Mˆ 5 ) ˆ
i i ( M6 + Φi M6 )
M 5 = ---------------------------------- M 6 = ----------------------------------
1 + Φi 1 + Φi

where

ˆ ˆ
N3 = 1 – ξ N4 = ξ
L L
Nˆ 5 = ---- ( ξ – ξ ) Nˆ 6 = – ---- ( ξ – ξ )
2 2
2 2
Mˆ 5 = 1 – ξ Mˆ 6 = ξ

The superscript i, indicates that the shape functions for bending are no longer the same
in the two principal directions. The shape functions with i = 2 are used for bending in
the local y direction, and the shape functions with i = 3 are used for bending in the
local z direction.

Geometric Nonlinearity
You can use the beam interface for modeling problems with large displacements and
rotations, but small strains. A so called co-rotational formulation is used. The
displacement of each individual beam element is decomposed into a rigid body
translation and rotation, and a local response of the rotated element which is linear.

The assumption that the individual element behaves linearly implies that you must use
a fine mesh if the curvature of the deformed beam is large. The difference in rotation
between the end points of the individual element must not be larger than it would be
possible to analyze it using linear theory.

Different coordinate systems are needed for describing the beam configurations. The
initial configuration of the beam can be described by a triad of orthogonal unit vectors
0
r i . The first vector is parallel to the beam, and the second and third vectors point in
the local y and z directions respectively. The origin of the local system is taken to be
the midpoint of the element. This system translates and rotates with the rigid body
motion of the beam, and the new directions of the axes are called r i .

696 | CHAPTER 7: BEAM

The rotation of the beam is represented by rotation vectors, θ . The rotation at the
midpoint is approximated as the arithmetic mean of the rotations at the nodes,

θ1 + θ2
θ M = ------------------
2

The rigid rotation is then represented by a rotation matrix Rr, corresponding to this
midpoint rotation. It is given by

sin θ M 1 – cos θ M 2
R r = I + -------------------- Θ M + ----------------------------
- ΘM
θM θ
2
M

where Θ M is the skew symmetric representation of the midpoint rotation vector.

0 – θ Mz θ My
ΘM = θ Mz 0 – θ Mx
– θ My θ Mx 0

The axis directions of the co-rotated coordinate system can now be computed as

r1 r2 r3 = Rr r0 r0 r0
1 2 3

The position of a point on the rigidly rotated axis of the element can be obtained as

x R = x M + R r ( X 2 – X 1 )  ξ – ---
1
 2

where the local coordinate ξ ranges from 0 to 1, and Xi denotes original node
coordinates. xM is the midpoint position, computed as the average of the two nodes,

1
x M = --- ( X 1 + u 1 + X 2 + u 2 )
2

In addition to the rigid body motion described so far, there are the local deformational
displacements with respect to the local rotated beam axes. The deformational
displacement can be computed as the difference between the current position and the
rigid body position

u = X + u – xR = u – uR

THEORY FOR THE BEAM INTERFACE | 697

 Here, and in the following, an over-bar denotes a deformational quantity. The
deformational rotation at the nodes is approximated by

θ2 – θ1
θ 1 = – -----------------
2
θ2 – θ1
θ 2 = -----------------
2

These local deformations are interpolated by the same shape functions as described in
the previous section:

 
 u1 
   
 u   θ1 
  = [N] 
 θ   u2 
   
 θ2 
 

The total displacement and rotation vectors can be expressed in term of the rigid
motion of the local axes, followed by the deformational motion relative to these axes.

u = uR + u

A deformational rotation matrix R can be defined as

R = I+Θ

where Θ is the skew-symmetric representation of the deformational rotation vector.

The total rotation vector is computed from a total rotation matrix, R. The total
rotation matrix is first composed from the rigid body rotation and the incremental
rotation.

R = RR r

The total rotation vector can now be extracted from the total rotation matrix. The
magnitude of the rotation vector is first computed as

trace ( R ) – 1
θ = acos  ---------------------------------
2

The full rotation vector is then computed as

698 | CHAPTER 7: BEAM

 R 32 – R 23
θ
θ = ---------------- R 13 – R 31
2 sin θ
R 21 – R 12

To avoid singularity problems when the angle is close to zero, the gamma function is
actually used in the expressions, since

θ - θ θ
----------- = γ  1 + --- γ  1 – ---
sin θ  π  π

Strain-Displacement/Rotation Relation
The axial strain depends on the rotation derivative (curvature) and axial displacement
derivative defined by the shape function and the transversal coordinate in the beam.
For the 3D case it becomes

∂θ ly ∂θ lz ∂u axi
ε = z l ---------- – y l ---------- + -------------
∂s ∂s ∂s

The coordinates from the beam center line in the local transversal directions are
denoted zl and yl respectively. In the 2D case, the first term is omitted, and the local z
direction is always directed out of the plane.

The total strain ε consists of thermal (εth), hygroscopic (εhs), initial (εi), and elastic
strains(εel)

ε = ε el + ε th + ε hs + ε i

Stress-Strain Relation
The stress-strain relation for the axial deformation in the beam is described by

σ = Eε el + σ i

where E is Young’s modulus, and σi is the initial stress.

The stress strain relation for the torsional and shear deformation is

τ = Gγ

THEORY FOR THE BEAM INTERFACE | 699

 where τ is the shear stress, γ is the shear strain and G is the shear modulus. Often the
material data is given in terms of Young’s modulus and Poisson’s ratio ν, in which case
G is defined as

E
G = ---------------------
2(1 + ν)

Thermal Strain
The temperature is assumed to vary linearly across the beam’s cross section. For the
3D beam it becomes

T = T m + T gz z l + T gy y l

Tm is the temperature at the beam center line while Tgz and Tgy are the temperature
gradients in the two local transversal directions. The thermal strain is thus

ε th = α ( T m + T gz z l + T gy y l – T ref )

For the 2D beam, the term depending on zl disappears.

Hygroscopic Swelling
The moisture concentration is assumed to vary linearly across the beam’s cross section.
For the 3D beam it becomes

c mo = c m + c gz z l + c gy y l

cm is the concentration at the beam center line while cgz and cgy are the concentration
gradients in the two local transversal directions. The strain from hygroscopic swelling
is then

ε th = β h ( c m + c gz z l + c gy y l – c mo,ref )

where βh is the coefficient of hygroscopic swelling, and cmo,ref is the strain-free

reference moisture concentration.

For the 2D beam, the term depending on zl disappears.

700 | CHAPTER 7: BEAM

Initial Load and Strain
The initial stress means the stress before any loads, displacements, and initial strains
have been applied.

The initial stress distribution is given as initial forces and moments in the local section
directions.


M yi = σ i z l dA
A
M zi = – σ i y l dA
A


N i = σ i dA
A


M xi = ( τ ixz y l – τ iyz z l ) dA
A

In 2D the x and y components of moments disappear.

In a Timoshenko beam, the initial shear forces appear as independent quantities, and
can also be specified.


T yi = τ ixy dA
A


T zi = τ ixz dA
A

The initial strain is the strain before any loads, displacements, and initial stresses have
been applied. The initial axial strain distribution is given as initial curvature and initial
axial strain

∂θ ly ∂θ lz ∂u axi
ε i = z l  ---------- – y l  ---------- +  -------------
∂s i ∂s i ∂s i

In 2D the zl dependent term disappears. As initial strain for the torsional degree of
freedom, the derivative of the twist angle,

∂θ lx
 ----------
 ∂s  i

is used.

THEORY FOR THE BEAM INTERFACE | 701

 In a Timoshenko beam it is also possible to set the initial values for the shear strains.

Implementation
The implementation is based on the principle of virtual work, which states that the sum
of virtual work from internal strains and external loads equals zero:


T
δW = δ ( – ε el σ + u F dV ) = 0
V

The beam elements are formulated in terms of the stress resultants (normal force,
bending moments and twisting moment).

The normal force is defined as

 
N = σ dA = ( Eε el + σ i ) dA =
A A

 z ∂θ ∂θ lz ∂u axi ∂θ ly ∂θ lz ∂u axi
  E – y l ---------- + ------------- –  z l  ---------- – y l  ---------- +  -------------  –
ly
----------
 l ∂s ∂s ∂s   ∂s i ∂s i ∂s i
A

α ( T m + T gz z l + T gy y l – T ref ) – β h ( c m + c gz z l + c gy y l – c mo,ref ) + σ i  dA =


 ∂u axi  ∂u axi 
  E -------------
 ∂s
– ------------- – α ( T m – T ref ) – β h ( c m – c mo,ref )  dA + σ i dA =
 ∂s  i  
A A
∂u axi ∂u axi
EA  ------------- –  -------------  – α ( T m – T ref ) – β h ( c m – c mo,ref ) + N i
 ∂s ∂s i

Because the local coordinates are defined with their origin at the centroid of the cross
section, any surface integral of an odd power of a local coordinate evaluates to zero.

The beam bending moments are defined as

702 | CHAPTER 7: BEAM

 M ly =  σzl dA =  zl ( Eεel + σi ) dA =
A A
∂θ ly ∂θ lz ∂u axi ∂θ ly ∂θ lz ∂u axi
 z l  E  z l ---------- – y l ---------- + ------------- –  z l  ---------- – y l  ---------- +  -------------  –
∂s ∂s ∂s ∂s i ∂s i ∂s i
A

α ( T m + T gz z l + T gy y l – T ref ) – β h ( c m + c gz z l + c gy y l – c mo,ref ) + σ i dA =

2 ∂θ ly ∂θ ly
 zl E ---------- –  ---------- – αT gz – β h c gz  dA + σ i z l dA = 
 ∂s  ∂s  i 
A A
∂θ ly ∂θ ly
EI yy ---------- –  ---------- – αT gz – β h c gz + M iy
∂s ∂s i

 
M lz = – σy l dA = – y l ( Eε el + σ i ) dA =
A A
∂θ ly ∂θ lz ∂u axi ∂θ ly ∂θ lz ∂u axi
– y l  E  z l ---------- – y l ---------- + ------------- –  z l  ---------- – y l  ---------- +  ------------- 
   ∂s ∂s ∂s    ∂s  i  ∂s  i  ∂s  i
A

– α ( T m + T gz z l + T gy y l – T ref ) – β h ( c m + c gz z l + c gy y l – c mo,ref ) + σ i dA =

∂θ lz ∂θ lz
– y l  E – ---------- +  ---------- – αT gy – β h c gy  dA – σ i y l dA =
 
2
∂s ∂s i
A A
∂θ lz ∂θ lz
EI zz ---------- –  ---------- + αT gy + β h c gy + M iz
∂s  ∂s  i

Mly is present only in 3D, and so is the torsional moment Mlx described below. The
torsional stiffness of the beam is defined using the torsional constant J given by

M
J = -------- l
Gθ

In a similar way as for the bending part a torsional moment is then defined as

∂θ lx ∂θ lx
M lx = GJ  ---------- –  ----------  + M ix
 ∂s  ∂s  i

Using the beam moment and normal force the expression for the virtual work becomes
very compact:

∂θ ly ∂θ lz ∂u axi ∂θ lx
δW = –  M ly δ ---------- + M lz δ ---------- + N δ ------------- + M lx δ ----------  dx

  ∂s   ∂s   ∂s   ∂s  
L

THEORY FOR THE BEAM INTERFACE | 703

 For 2D, the first and fourth terms are omitted. For the case of Timoshenko beam,
there is also a shear stress contribution added,


δW = – ( T ly δγ xy + T lz δγ xz ) dx
L

were the second term is present only in 3D.

A special feature of some unsymmetrical cross sections is that they twist under a
transversal load that is applied to beam centerline. As an example, this would be the
case for a U-profile under self-weight, loaded in the stiff direction. It is only a load
applied at the shear center which causes a pure deflection without twist. This effect can
be incorporated by supplying the coordinates of the shear center in the local coordinate
system (ey, ez). A given transversal load (flx, fly, flz), which is defined as acting along
the centerline, is then augmented by a twisting moment given by

m lx = f ly e z – f lz e y

Stress Evaluation
Since the basic result quantities for beams are the integrated stresses in terms of section
forces and moment, special considerations are needed for the evaluation of actual
stresses.

The normal stress from axial force is constant over the section, and computed as

N
σ n = ----
A

The normal stress from bending is computed in four user-selected points (ylk, zlk) in
the cross section as

M ly z lk M lz y lk
σ bk = ----------------- – -----------------
I yy I zz

In 2D, only two points, specified by their local y-coordinates are used.

The total normal stress in these points is then

σ k = σ bk + σ n

The peak normal stress in the section is defined as

704 | CHAPTER 7: BEAM

 σ max = max ( σ k )

When using the built in common cross sections, a special method is used for the
Circular and Pipe sections. Since there are no extreme positions around a circle, a
maximum bending stress is computed as

2 2
d o M ly + M lz
σ b, max = ------------------------------------
2I zz

where do is the outer diameter. This stress then replaces the stress from the stress
evaluation points in maximum stress expressions. This ensures that the correct peak
stress is evaluated irrespective of where it appears along the circumference.

The shear stress from twist in general has a complex distribution over the cross section.
The maximum shear stress due to torsion is defined as

M lx
τ t, max = -------------
Wt

where Wt is the torsional section modulus. This result is available only in 3D.

The section shear forces are computed in two different ways depending on the beam
formulation. For Euler-Bernoulli theory, the section forces proportional to the third
derivative of displacement, or equivalently, the second derivative of the rotation.

2
∂ θ ly
T lz = EI yy ------------
2
-
∂s
2
∂ θ lz
T ly = – E I zz ------------
2
-
∂s

where Tlz is available only in 3D. In the case of Timoshenko theory shear force is
computed directly from the shear strain.

The average shear stresses are computed from the shear forces as

T lz
τ sz, ave = --------
A
(7-1)
T ly
τ sy, ave = --------
A

THEORY FOR THE BEAM INTERFACE | 705

 Since the shear stresses are not constant over the cross section, the maximum shear
stresses are also available, using section dependent correction factors:

τ sz, max = μ z τ sz, ave

(7-2)
τ sy, max = μ y τ sy, ave

As the directions and positions of maximum shear stresses from shear and twist are not
known in a general case, upper bounds to the shear stress components are defined as

τ xz, max = τ sz, max + τ t, max

τ xy, max = τ sy, max + τ t, max

The maximum von Mises effective stress for the cross section is the defined as

2 2 2
σ mises = σ max + 3τ xy, max + 3τ xz, max

Since the maximum values for the different stress components in general occur at
different positions over the cross section, the effective stress thus computed is a
conservative approximation.

Common Cross Sections

The cross section data for the common cross sections can be computed internally in
COMSOL Multiphysics. In this section, the expressions used are summarized.

Beam Cross Section

706 | CHAPTER 7: BEAM

RECTANGULAR SECTION

Figure 7-2: Geometry of a rectangular cross section. The diagram also displays in
COMSOL Multiphysics when this option is selected.

TABLE 7-2: RECTANGULAR SECTION CONSTANTS

PROPERTY EXPRESSION REMARKS

A hy hz

Izz 3
hy hz
------------
12
ez 0
μy 1.5
κy 5/6
Iyy 3
hz hy
------------
-
12
ey 0
μz 1.5
κz 5/6
J 2 q=
h y h z q  
2 2
( 2n – 1 )π- 
------------------------------ tanh  ------------------------
192q
----------------- 1 –
3   5
π ( 2n – 1 )
5  2q  
min(hy/hz,hz/hy)
 n=1 

THEORY FOR THE BEAM INTERFACE | 707

 TABLE 7-2: RECTANGULAR SECTION CONSTANTS

PROPERTY EXPRESSION REMARKS

Wt J q=
---------------------------------------------------------------------------------------------------------------------------
 2  min(hy/hz,hz/hy)
 8 
hy hz q  1 –
  ---------------------------------------------------------------------------
2 2  ( 2n – 1 )π 
 n = 1 π ( 2n – 1 ) cosh  ------------------------- 
2q

p1 – h y – h z
 --------
-, ---------
 2 2 
p2
h –h
-----y- --------z-
2, 2 

p3
h h
-----y- -----z-
 2, 2

p4 – h y h z
 --------
-, ------
 2 2
re hy + hz
------------------
4

BOX SECTION

Figure 7-3: Geometry of a box shaped cross section. The diagram also displays in COMSOL
Multiphysics when this option is selected.

708 | CHAPTER 7: BEAM

TABLE 7-3: BOX SECTION CONSTANTS

PROPERTY EXPRESSION REMARKS

A 2 ( hy tz + hz ty ) –4 ty tz

Izz 3 3 2
t z h y + t y ( h z – 2t z ) t y ( h z – 2t z ) ( h y – t y )
------------------------------------------------ + -------------------------------------------------------
6 2
ez 0
μy 2 2
( h y h z – ( h y – 2t y ) ( h z – 2t z ) )A
-------------------------------------------------------------------------------
-
16t z I zz

κy 2h y t z
---------------
A
Iyy 3 3 2
t y h z + t z ( h y – 2t y ) t z ( h y – 2t y ) ( h z – t z )
------------------------------------------------- + -------------------------------------------------------
6 2
ey 0
μz 2 2
( h z h y – ( h z – 2t z ) ( h y – 2t y ) )A
-------------------------------------------------------------------------------
-
16t y I yy

κz 2h z t y
--------------
A
J 2 2 Thin-walled
2 ( hy – ty ) ( hz – tz )
----------------------------------------------------- approximation
hy – ty hz – tz
----------------- + ----------------
tz ty

Wt 2 ( h y – t y ) ( h z – t z ) ⋅ min ( t y, t z ) Thin-walled
approximation
p1 – h y – h z
 --------
 2 , 2 
- ---------

p2
h –h
-----y- --------z-
2, 2 

p3
h h
-----y-, -----z-
 2 2

THEORY FOR THE BEAM INTERFACE | 709

 TABLE 7-3: BOX SECTION CONSTANTS

PROPERTY EXPRESSION REMARKS

p4 – h y h z
 --------
 2 , 2
- ------

re hy + hz
------------------
4

CIRCULAR SECTION

Figure 7-4: Geometry of a circular cross section. The diagram also displays in COMSOL
Multiphysics when this option is selected.

TABLE 7-4: CIRCULAR SECTION CONSTANTS

PROPERTY EXPRESSION REMARKS

A 2
πd o
----------
4
Izz 4
πd
---------o-
64
ez 0
μy 4
---
3
κy 0.9
Iyy I zz

ey 0

710 | CHAPTER 7: BEAM

TABLE 7-4: CIRCULAR SECTION CONSTANTS

PROPERTY EXPRESSION REMARKS

μz μy
κz κy
J 4
πd
---------o-
32
Wt 3
πd
---------o-
16
p1 do 
 –--------
 2 , 0
-

p2 d o
 0, –--------
-
 2 
p3
d
-----o-, 0
2 
p4
 0, d
-----o-
 2

re do
------
2

PIPE SECTION

Figure 7-5: Geometry of a pipe cross section. The diagram also displays in COMSOL
Multiphysics when this option is selected.

THEORY FOR THE BEAM INTERFACE | 711

 TABLE 7-5: PIPE SECTION CONSTANTS

PROPERTY FORMULA REMARKS

A 2 2
π ( do – di )
---------------------------
4
Izz 4 4
π ( do – di )
---------------------------
64
ez 0
μy 3 3
( d o – d i )A
------------------------------------
-
12 ( d o – d i )I zz

κy di
1 + ------
do

Iyy I zz

ey 0
μz μy
κz κy
J
4 4
π ( do – di )
---------------------------
32
Wt 4 4
π ( do – di )
---------------------------
16d o

p1 –do 
 --------
-, 0
 2 
p2 – d o
 0, --------
-
 2 
p3
d
-----o- 
 2 , 0

712 | CHAPTER 7: BEAM

TABLE 7-5: PIPE SECTION CONSTANTS

PROPERTY FORMULA REMARKS

p4
 0, d
-----o-
 2

re do
------
2

H-PROFILE SECTION

Figure 7-6: Geometry of a an H-profile cross section. The diagram also displays in
COMSOL Multiphysics when this option is selected.

TABLE 7-6: H-PROFILE SECTION CONSTANTS

PROPERTY EXPRESSION REMARKS

A 2h z t y + t z ( h y – 2t y )

Izz 3 3 2
2h z t y + t z ( h y – 2t y ) ty hz ( hy – ty )
------------------------------------------------------
- + -----------------------------------
-
12 2
ez 0
μy 2
[ 4h z t y ( h y – t y ) + t z ( h y – 2t y ) ]A
--------------------------------------------------------------------------------------
8t z I zz

κy hy tz
-----------
A

THEORY FOR THE BEAM INTERFACE | 713

 TABLE 7-6: H-PROFILE SECTION CONSTANTS

PROPERTY EXPRESSION REMARKS

Iyy 3 3
2t y h z + t z ( h y – 2t y )
----------------------------------------------------
12
ey 0
μz 2 2
( h z – t z )A
--------------------------
8I yy

κz 8h z t y
-------------
-
3A
J 3 3 Thin-walled
2t y h z + t z ( h y – 2t y )
---------------------------------------------------- approximation
3
Wt J Thin-walled
-----------------------------
-
max ( t y, t z ) approximation

p1 – h y – h z
 --------
-, ---------
 2 2 
p2
h –h
-----y-, --------z-
2 2 
p3
h h
-----y- -----z-
 2, 2

p4 – h y h z
 --------
 2 , 2
- ------

re hy + hz
------------------
4

714 | CHAPTER 7: BEAM

U-PROFILE SECTION

Figure 7-7: Geometry of a U-profile cross section. Also displays in COMSOL Multiphysics
when this option is selected.

TABLE 7-7: U-PROFILE SECTION CONSTANTS

PROPERTY EXPRESSION REMARKS

A h y t z + 2 ( h z – t z )t y

Izz 3 3 2
t z h y + 2t y ( h z – t z ) t y ( h z – t z ) ( h y – t y )
- + ----------------------------------------------------
-----------------------------------------------
12 2
zCG 2
h y t z ( 2h z – t z ) + 2t y ( h z – t z )
----------------------------------------------------------------------------
2A
ez 2 2
t z t y ( 2h z – t z ) ( h y – t y )
h z – ---- + ---------------------------------------------------------- – z CG
2 16I zz

μy 2 2
( h y h z – ( h y – 2t y ) ( h z – t z ) )A
---------------------------------------------------------------------------
-
8t z I zz

κy hy tz
-----------
A
Iyy 3 3 2
8t y h z + t z ( h y – 2t y ) + 3t z ( h y – 2t y ) ( 2h z – t z )
-------------------------------------------------------------------------------------------------------------------------
-
12
2
– z CG A

ey 0

THEORY FOR THE BEAM INTERFACE | 715

 TABLE 7-7: U-PROFILE SECTION CONSTANTS

PROPERTY EXPRESSION REMARKS

μz 2
z CG A
-----------------
2I yy

κz 8h z t y
-------------
-
3A
J 3 3 Thin-walled
2t y h z + t z ( h y – 2t y )
---------------------------------------------------- approximation
3
Wt J Thin-walled
-----------------------------
-
max ( t y, t z ) approximation

p1 –hy
 --------
-, – z CG
 2 
p2
h
-----y- z 
 2 , – CG

p3
h
-----y- h 
 2 , z – z CG

p4 –hy
 --------
-, h z – z CG
 2 
re hy + hz
------------------
4

716 | CHAPTER 7: BEAM

T- P R O F I L E S E C T I O N

Figure 7-8: Geometry of a T-profile cross section. Also displays in COMSOL Multiphysics
when this option is selected.

TABLE 7-8: T-PROFILE SECTION CONSTANTS

PROPERTY EXPRESSION REMARKS

A h z t y + ( h y – t y )t z

yCG 2
t z ( h y – t y ) + t y h z ( 2h y – t y )
-------------------------------------------------------------------------
2A
Izz 3 3 2
4t z ( h y – t y ) + h z t y + 3t y h z ( 2h y – t y ) 2
----------------------------------------------------------------------------------------------------- – y CG A
12
ez 0
μy 2
y CG A
------------------
2I zz

κy hy tz
-----------
A
Iyy 3 3
ty hz + tz ( hy – ty )
--------------------------------------------
-
12
ey ty
h y – ---- – y CG
2

THEORY FOR THE BEAM INTERFACE | 717

 TABLE 7-8: T-PROFILE SECTION CONSTANTS

PROPERTY EXPRESSION REMARKS

μz 2 2
( h z – t z )A
--------------------------
8I yy

κz 4h z t y
-------------
-
3A
J 3 3 Thin-walled
ty hz + tz ( hy – ty )
--------------------------------------------- approximation
3
Wt J Thin-walled
---------------------------
-
max ( t y, t z ) approximation

p1 t z
 – y , –-------
 CG 2 
p2 h z
 h – y , –--------
-
 y CG 2 

p3
 h –y , h
-----z-
 y CG 2 

p3
 – y , t---z-
 CG 2 
re hy + hz
------------------
4

718 | CHAPTER 7: BEAM

The Beam Interface
The Beam interface ( ), found under the Structural Mechanics branch ( ) when
adding a physics interface, is used for modeling slender structural elements, having a
significant bending stiffness. The formulation allows geometric nonlinearity, with large
rotations and small strains, and beams can be modeled on 2D boundaries and 3D
edges.

Two-noded straight elements with an Hermitian formulation are used. Two different
assumptions about the physics can be used:

• Euler (or Euler-Bernoulli) theory. This formulation is intended for slender beams,
and do not take shear deformations into account.
• Timoshenko theory. In this formulation that extends the beam theory to ‘thick’
beams, shear deformations are taken into account. In a dynamic analysis, inertial
effects from rotation are also included.

Among the computed results are displacements, rotations, stresses, strains, and section
forces. In addition to giving the beam properties explicitly in terms of area, moment
of inertia, and so on, several predefined common cross-section types are available.
Cross section data to be used in Cross Section Data settings can be computed using
Theory for the Beam Cross Section Interface.

The Linear Elastic Material node is the only available material model.

When this physics interface is added, these default nodes are also added to the Model
Builder: Linear Elastic Material, Cross Section Data, Free (a condition where points are
free, with no loads or constraints), and Initial Values. Then, from the Physics toolbar,
add other nodes that implement, for example, loads and constraints. You can also
right-click Beam to select physics features from the context menu.

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is beam.

THE BEAM INTERFACE | 719

 BEAM FORMULATION
Select Euler-Bernoulli or Timoshenko to use the appropriate beam theory.

For COMSOL Multiphysics models created in versions 4.3b and earlier,

the only option available is Euler-Bernoulli. See the Backward
Compatibility section to make changes to such a model.

BACKWARD COMPATIBILITY

The formulation of the beam element was significantly changed as of

COMSOL Multiphysics version 4.4. For this reason, by default, models
created in previous versions default to using the Euler-Bernoulli theory.

This section only displays if a model was created in an earlier version. You
also need to click the Show button ( ) and select Advanced Physics
Options to edit the section and change the Beam Formulation to
Timoshenko (if required).

Click to clear the Use pre 4.4 formulation check box to convert a model into the new
formulation. For most models, the only difference is that it will then be possible to
select Timoshenko theory in the Beam Formulation section.

There are, however, some situations where you need to make additional changes to
your model when migrating to the new formulation:

• If you have made manual adjustments to the solver sequence, such as scaling the
dependent variables or adjusting groups in a segregated solver. These settings might
have to be redone.
• If you are using multiple physics interfaces and use the same degree of freedom
names for the displacements in the Beam interface and other structural mechanics
interfaces. In this case, you need to rename the degrees of freedom in the Beam
interface and create a different type of connection between the beams and the other
physics interface, for example using a Solid Connection or Shell Connection.

REFERENCE PO INT F OR MO MENT COMPUTATION

Enter the default coordinates for the Reference point for moment computation xref. The
resulting moments (applied or as reactions) are then computed relative to this
reference point. During the results and analysis stage, the coordinates can be changed
in the Parameters section in the result nodes.

720 | CHAPTER 7: BEAM

DEPENDENT VA RIA BLES
The Beam interface has these dependent variables (fields):

• The displacement field u, which has two components (u, v) in 2D and three
components (u, v, and w) in 3D.
• The rotation angle θ, which has one component in 2D (th) and three components
in 3D (thx, thy, and thz).

The names can be changed but the names of fields and dependent variables must be
unique within a model.

The dependent variable names remain same in both a geometrically linear

and a geometrically nonlinear analysis. Under geometric nonlinearity, the
dependent variables are however not defined though shape functions. The
equivalent shape function variables are (beam.uLinx, beam.uLiny,
beam.uLinz) and (beam.thLinx, beam.thLiny, beam.thLinz). In this
case, you will see the latter names under Dependent variables in the Solver
tree.

If needed, these shape variables can be used to write any extra

contributions in the Beam interface.

If a physics interface that separates the material and spatial frame (Solid
Mechanics is one such example) is added to the model, the coordinate
indices change from (x, y, z) to (X, Y, Z) in the name of these variables.

DISCRETIZATION
The discretization cannot be changed. The element has different shape functions for
the axial and transversal degrees of freedom. The axial displacement and twist are
represented by linear shape functions, while the bending is represented by a cubic
shape function (“Hermitian element”).

• Boundary, Edge, Point, and Pair Nodes for the Beam Interface
• Theory for the Beam Interface
• Theory for the Beam Cross Section Interface

THE BEAM INTERFACE | 721

 • Channel Beam: Application Library path
Structural_Mechanics_Module/Verification_Examples/channel_beam
• Instability of a Space Arc Frame: Application Library path
Structural_Mechanics_Module/Verification_Examples/
space_frame_instability

Boundary, Edge, Point, and Pair Nodes for the Beam Interface
The Beam Interface has these boundary, edge, point, and pair nodes available from the
Physics ribbon toolbar (Windows users), Physics context menu (Mac or Linux users),
or right-click to access the context menu (all users).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

FEATURES AVAILABLE FROM SUBMENUS

Many features for the Beam interface are added from submenus in the Physics toolbar
groups or context menu (when you right-click the node). The submenu name is the
same in both cases.

The submenus at the Edge level (3D) or Boundary level (2D) are

• Material Models
• Line and Volume Loads,
• Mass, Spring, and Damper
• Connections
• Line Constraints.

The submenus at the Point level are

• Mass, Spring, and Damper

• Connections
• More Constraints
• Pairs.

722 | CHAPTER 7: BEAM

LINKS TO FEATURE NODE INFORMATION
These nodes (and subnodes) are described in this section (listed in alphabetical order):

• Antisymmetry • Point Mass

• Attachment • Point Mass Damping
• Cross Section Data • Prescribed Acceleration
• Edge Load • Prescribed Displacement/Rotation
• External Stress • Prescribed Velocity
• Hygroscopic Swelling • Rigid Connector
• Initial Stress and Strain • Section Orientation
• Initial Values • Shell Connection
• Linear Elastic Material • Solid Connection
• No Rotation • Symmetry
• Pinned • Thermal Expansion (for materials)
• Point Load

These nodes are described for the Solid Mechanics interface:

• Added Mass • Gravity1

• Damping • Predeformation
• Fixed Constraint • Rigid Domain
• Free • Rotating Frame1
• Spring Foundation
1
This is selected from the Line and Volume Loads submenu for this interface.

If there are subsequent constraints specified on the same geometrical

entity, the last one takes precedence. The exception is that the “Pinned”
and “No Rotation” boundary conditions do not override each other since
the degrees of freedom that they constrain are mutually exclusive.

THE BEAM INTERFACE | 723

 In the COMSOL Multiphysics Reference Manual see Table 2-3 for links
to common sections and Table 2-4 to common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

Initial Values
The Initial Values node adds an initial values for the displacement field, the velocity
field, the rotations, and the angular velocity. It serves as initial conditions for a transient
simulation or as an initial guess for a nonlinear analysis. In addition to the default Initial
Values node always present in the interface, you can add more Initial Values nodes if
needed.

INITIAL VALUES
Enter values or expressions for the following based on space dimension:

• Displacement field u
du
• Velocity field
dt
• Rotation field θ
dθ
• Angular velocity
dt
Context Menus
Beam>Initial Values
Beam>Edges>Initial Values

Ribbon
Physics tab with Solid Mechanics selected:
Edges>Beam>Initial Values

Cross Section Data

In the Cross Section Data node you specify the geometric properties of the beam’s cross
section. In addition, some stress evaluation properties can be defined.

For 3D models, a default Section Orientation subnode is added, in which you specify
the orientation of the principal axes of the section. You can add any number of Section

724 | CHAPTER 7: BEAM

Orientation subnodes if the same section appears with different spatial orientations in
the structure.

This is required input data.

• Common Cross Sections

• Beam Cross Section

CROSS SECTION DEFINITION

The default is User defined. Select Common sections to choose from predefined sections.

For User defined go to Basic Section Properties and Stress Evaluation Properties to
continue defining the cross section.

For Common sections select a Section type—Rectangle, Box, Circular, Pipe, H-profile,
U-profile, or T-profile. Then go to the relevant section below to continue defining the
section. Each Section type also has a figure showing the section and its defining
dimensions.

For equations and a figure see:

• Rectangular Section
• Box Section
• Circular Section
• Pipe Section
• H-Profile section
• U-Profile section
• T-Profile section

Examples of how to work with cross sections: Application Library path

Structural_Mechanics_Module/Verification_Examples/channel_beam and
Structural_Mechanics_Module/Civil_Engineering/pratt_truss_bridge

THE BEAM INTERFACE | 725

 Rectangle
Enter values or expressions for the following.

• Width in local y-direction hy

• Width in local z-direction hz

Box
Enter values or expressions for the following.

• Width in local y-direction hy

• Width in local z-direction hz
• Wall thickness in local y-direction ty
• Wall thickness in local z-direction tz

Circular
Enter a value or expression for the Diameter do.

Pipe
Enter values or expressions for the following.

• Outer diameter do
• Inner diameter di

H-profile, U-profile, or T-profile

Enter values or expressions for the following.

• Section height hy
• Flange width hz
• Flange thickness ty
• Web thickness tz

BASIC SECTION PROPERTIES

This section is only available if User defined is selected as the Cross Section
Definition.

726 | CHAPTER 7: BEAM

The following table lists the basic section properties (some apply in 3D only). Enter
values for these properties in the associated fields. The default values correspond to a
circular cross section with a diameter of 0.1 m:

COMMENT DESCRIPTION PARAMETER SI UNIT

2D and 3D Area of cross section A m2

2D and 3D Moment of inertia about local z-axis Izz m4
2D and 3D, Shear correction factor along local y-axis κy 1
Timoshenko
beam
3D only Distance to shear center in local z-direction ez m
3D only Moment of inertia about local y-axis Iyy m4
3D only Distance to shear center in local y-direction ey m
3D only Torsional constant J m4
3D only, Shear correction factor along local z-axis κz 1
Timoshenko
beam

For 3D models, the orientation of the cross section is given in Section

Orientation. If the beam’s cross section is a square or circle (solid or tube),
the area moments of inertia are the same independent of direction, so the
beam is totally symmetric and the orientation of the principal axes of the
cross section is not a problem unless you are interested in looking at
results defined using the local coordinate system. Such results are bending
moments, shear forces, local displacements and rotations.

STRESS EVALUATION PROPERTIES

This section is only available if User defined is selected as the Cross Section
Definition.

Select the Bending stress evaluation points—From section heights (the default or From
specified points.

Stress evaluation using only section heights is meaningful only when the cross section
is symmetric.

THE BEAM INTERFACE | 727

 The max shear stress factor determines the ratio between the peak and the average
shear stress over the cross section as described by Equation 7-1 and Equation 7-2.

From Section Heights

For From section heights enter values in each field for the following parameters as
needed for the space dimension:

COMMENT DESCRIPTION PARAMETER SI UNIT

2D and 3D Section height in local y direction hy m

2D and 3D Max shear stress factor in local y direction μy 1
3D only Section height in local z direction hz m
3D only Torsional section modulus Wt m3
3D only Max shear stress factor in local z direction μz 1

From Specified Points

For From specified points enter values in the Evaluation points in local system table as
needed for the space dimension. Then enter the following parameters in the applicable
fields.

COMMENT DESCRIPTION PARAMETER SI UNIT

2D and 3D Max shear stress factor in local y direction μy 1

3D only Torsional section modulus Wt m3
3D only Max shear stress factor in local z direction μz 1

LOCATION IN USER INTERFACE

Context Menus
Beam>Cross Section Data

Ribbon
Physics tab with Beam selected:
Edges>Beam>Cross Section Data

Section Orientation
Use the Section Orientation subnode to define the orientation of a beam cross section
using a reference point or an orientation vector. There is always one Section Orientation
subnode for each cross section, and as many Section Orientation subnodes as needed

728 | CHAPTER 7: BEAM

can be added if the same section appears with different spatial orientations in the
structure.

This node is available for 3D components.

ORIENTATION METHOD
Select the Reference point (the default) or Orientation vector. For Reference point For
enter a Reference point defining local y direction P.

The coordinate system is defined as follows:

The local x direction is in the edge direction. The positive edge direction can be
checked by vector plotting the local edge tangent direction. The coordinates of the
reference point define the local xy-plane together with the beam axis. The local
coordinate system (exl, eyl, ezl) is formed using the following algorithm:

v zl = e xl × ( p – m )

v zl
e zl = ----------
v zl

e yl = e zl × e xl

Here, p is the reference point, and m is the midpoint of the beam element. The
definition of the local coordinate system is illustrated in Figure 7-9.

Figure 7-9: Local beam coordinate system defined by a reference point

THE BEAM INTERFACE | 729

 For the creation of a local coordinate system to be possible, the point cannot coincide
with the edge or the edge extension. If this is attempted, an error message is generated.

The settings for the global coordinates of the point are

[1000,1000,1000]. This is useful only for symmetric cross sections.

Often a number of edges in a plane have the same orientation. It is then easy to select
all edges and specify a point anywhere in the same plane, not coinciding with an edge
or an edge extension.

For Orientation vector enter a Orientation vector defining local y direction V and
optionally the Rotation of vector around beam axis φ. The beam orientation is defined
similarly to what is described above, with the difference that in this case the direction
vector is explicitly defined whereas when an orientation point is used, the direction
vector is obtained as the vector from the beam axis to the specified point. The local
coordinate system (exl, eyl, ezl) is formed using the following algorithm:

v zl = e xl × V

v zl
e zl = ----------
v zl

e yl = e zl × e xl

The Rotation of vector around beam axis has the effect of rotating the given vector
around the beam axis (using the right-hand rule) before it is used to define the local
xy-plane. This simplifies the input for some cross sections, such as L-shaped profiles,
where the principal axes have a direction which is skewed relative to a more natural
modeling position. This can be written as

e yl = e' yl cos φ – e' zl sin φ

e zl = e' zl cos φ + e' yl sin φ

Here the directions denoted with a prime are unrotated beam axis orientations
obtained by the procedure described above.

LOCATION IN USER INTERFACE

Context Menus
Beam>Cross Section Data>Section Orientation

730 | CHAPTER 7: BEAM

Ribbon
Physics tab with Cross Section Data node selected in the model tree:
Attributes>Section Orientation

Linear Elastic Material

The Linear Elastic Material node adds the equations for a linear elastic beam and an
interface for defining the elastic material properties.

By adding the following subnodes to the Linear Elastic Material node you can
incorporate many other effects:

• Thermal Expansion (for materials)

• Hygroscopic Swelling
• Initial Stress and Strain
• External Stress
• Damping

LINEAR ELASTIC MATERIAL

Define the linear elastic material properties.

Specification of Elastic Properties for Isotropic Materials

From the Specify list, select a pair of elastic properties for an isotropic material. Select:

• Young’s modulus and Poisson’s ratio to specify Young’s modulus (elastic modulus) E
and Poisson’s ratio ν. Poisson’s ratio is used for computing the torsional stiffness,
and is thus important only for 3D beams.
• Young’s modulus and shear modulus to specify Young’s modulus (elastic modulus) E
and the shear modulus G. The shear modulus is used for computing the torsional
stiffness, and is thus important only for 3D beams.
• Bulk modulus and shear modulus to specify the bulk modulus K and the shear
modulus G.

THE BEAM INTERFACE | 731

 • Lamé parameters to specify the Lamé parameters λ and μ.
• Pressure-wave and shear-wave speeds to specify the pressure-wave speed
(longitudinal wave speed) cp and the shear-wave speed (transverse wave speed) cs.

This is the wave speed for a solid continuum. In a truss or beam element,
the actual speed with which a longitudinal wave travels is lower than the
value given. When using this type of input the density must also be given.

For each pair of properties, select from the applicable list to use the value From material
or enter a User defined value or expression.

Each of these pairs define the elastic properties, and it is possible to convert from one
set of properties to another.

Density
Define the Density ρ of the material. Select From material to take the value from the
material or User defined to enter a value for the density.

The density is needed for dynamic analysis or when the elastic data is given
in terms of wave speed. It is also used when computing mass forces for
gravitational or rotating frame loads, and when computing mass
properties (Computing Mass Properties).

Thermally Loaded Beam: Application Library path

Structural_Mechanics_Module/Verification_Examples/thermally_loaded_beam

LOCATION IN USER INTERFACE

Context Menus
Beam>Material Models>Linear Elastic Material

Ribbon
Physics tab with Beam selected:
Edges>Material Models>Linear Elastic Material

732 | CHAPTER 7: BEAM

Thermal Expansion (for materials)
Use the Thermal Expansion subnode to add an internal thermal strain caused by changes
in temperature. The thermal strain depends on the coefficient of thermal expansion
(CTE) α, the temperature T, and the strain-free reference temperature Tref as

ε th = α ( T – T ref )

It is possible to model bending due to a temperature gradient in the transverse

directions of the beam. The temperature is then assumed to vary linearly through the
thickness.

MODEL INPUTS
From the Temperature T list, select an existing temperature variable from a heat transfer
interface. For User defined enter a value or expression for the temperature (the default
is 293.15 K). This is the centerline temperature of the beam, controlling the axial part
of the thermal expansion.

THERMAL EXPANSION PROPERTIES

Specify the thermal properties that define the thermal strain.

From the Coefficient of thermal expansion α list, select From material to use the
coefficient of thermal expansion from the material, or User defined to enter a value or
expression for α.

Enter a value or expression of the Strain reference temperature Tref, which is the
reference temperature where the thermal strain is zero.

THERMAL BENDING
Enter the Temperature gradient in local y-direction Tgy (in 2D and 3D) and in the
Temperature gradient in local z-direction Tgz (in 3D), which affects the thermal
bending. If beam cross section dimensions have been defined at Bending stress
evaluation points—From section heights, these could be used in an expression
containing the temperature difference.

LOCATION IN USER INTERFACE

Context Menus
Beam>Linear Elastic Material>Thermal Expansion

THE BEAM INTERFACE | 733

 Ribbon
Physics tab with Linear Elastic Material node selected in the model tree:
Attributes>Thermal Expansion

Hygroscopic Swelling
Hygroscopic swelling is an internal strain caused by changes in moisture content. This
strain can be written as

ε hs = β h ( c mo – c mo,ref )

where βh is the coefficient of hygroscopic swelling, cmo is the moisture concentration,

and cmo,ref is the strain-free reference concentration. It is possible to model bending
due to a concentration gradient in the transverse directions of the beam. The
concentration is then assumed to vary linearly through the thickness.

MODEL INPUTS
From the Concentration c list, select an existing concentration variable from another
physics interface, if any concentration variables exist. For User defined enter a value or
expression for the concentration. This is the centerline concentration of the beam,
controlling the axial part of the hygroscopic swelling.

The unit for the input depends on the setting of Concentration type in the Hygroscopic
Swelling Properties section. Only concentration variables having the chosen physical
dimension are available in the Concentration list.

HYGROSCOPIC SWELLING PROPERTIES

In the Concentration type list, select Molar concentration (the default) or Mass
concentration, depending on the units used for the concentration.

Enter a Strain reference concentration cref. This is the concentration at which there are
no strains due to hygroscopic swelling.

If Molar concentration is selected as Concentration type, also enter the Molar mass of the
fluid, Mm. The default value is 0.018 kg/mol, which is the molar mass of water.

The default Coefficient of hygroscopic swelling βh uses values From material. For
User defined select Isotropic (the default), Diagonal, or Symmetric to enter one or more
components for a general coefficient of hygroscopic swelling tensor βh.The default
value for the User defined case is 1.5e-4 m3/kg.

734 | CHAPTER 7: BEAM

HYGROSCOPIC BENDING
Enter the Concentration gradient in local y-direction cgy (in 2D and 3D) and in the
Concentration gradient in local z-direction cgz (in 3D), which affects the hygroscopic
bending. If beam cross section dimensions have been defined at Bending stress
evaluation points—From section heights, these could be used in an expression
containing the concentration difference.

LOCATION IN USER INTERFACE

Context Menus
Beam>Linear Elastic Material>Hygroscopic Swelling

Ribbon
Physics tab with Linear Elastic Material node selected in the model tree:
Attributes>Hygroscopic Swelling

Initial Stress and Strain

You can add the Initial Stress and Strain to the Linear Elastic Material node, in order
to specify the stress or strain state in the structure before applying any constraint or
load. The values given are not initial values in the mathematical sense, but rather a
contribution to the constitutive relation.

In many cases Initial Stress and Strain and External Stress are
interchangeable when prescribing stresses, but you can find some more
options in the latter.

INITIAL STRESS
Specify the initial stress as the:

• Initial axial force Ni.

• Initial bending moment Miz and for 3D models, Miy.
• For 3D models: Initial torsional moment Mix.
• For Timoshenko beam: Initial shear force Tiy and for 3D models, Tiz.

INITIAL STRAIN
Specify the initial strain as the:

• Initial axial strain eni.

THE BEAM INTERFACE | 735

 • Initial curvature θsiz and for 3D models, and θsiy.
• For 3D models: Initial twist θsix.
• For Timoshenko beam: Initial shear strain γsiy and for 3D models, γsiz.

For details about initial stresses and strains, see Inelastic Strain
Contributions and Initial Stresses and Strains.

Context Menus
Beam>Linear Elastic Material>Initial Stress and Strain

Ribbon
Physics tab with Linear Elastic Material node selected in the model tree:
Attributes>Initial Stress and Strain

External Stress
You can add the External Stress subnode to several material models, in order to specify
an additional stress contribution which is not part of the constitutive relation. The
external stress can be added to the total stress tensor, or act only as an extra load
contribution.

In many cases External Stress and Initial Stress and Strain are
interchangeable when prescribing stresses. In Initial Stress and Strain, the
given stress is however always added to the stress tensor.

EXTERNAL STRESS
Specify the external stress as section forces:

• External axial force Next.

• External bending moment Mz,ext and for 3D models, My,ext.
• For 3D models: External torsional moment Mx,ext.
• For Timoshenko beam: External shear force Ty,ext and for 3D models, Tz,ext.

736 | CHAPTER 7: BEAM

For each quantity, you can select User defined, in which case you enter values or
expressions explicitly. The drop-down lists will also contain any section forces
announced from a physics interface.

Selecting section forces announced by the same physics interface as where

the External Stress node is added, will result in an error (‘Circular
variable dependency detected’). This operation would imply that
the computed force depends on itself.

Context Menus
Beam>Linear Elastic Material>External Stress

Ribbon
Physics tab with Linear Elastic Material node selected in the model tree:
Attributes>External Stress

Prescribed Displacement/Rotation
The Prescribed Displacement/Rotation node adds an edge (3D), boundary (2D), or
point (2D and 3D) condition where the displacements and rotations are prescribed in
one or more directions.

• If a prescribed displacement or rotation is not activated in any direction, this is the

same as a Free constraint.
• If zero displacements and rotations are prescribed, this is the same as a Fixed
Constraint.
• If only zero displacements are prescribed, this is the same as a Pinned constraint.
• If only zero rotations are prescribed, this is the same as a No rotation constraint.

COORDINATE SYSTEM SELECTION

Specify the coordinate system to use for specifying the prescribed
displacement/rotation. See the section Coordinate System Selection for Edge Load.

PRESCRIBED DISPLACEMENT
Select one or all of the Prescribed in x direction, Prescribed in y direction, and for 3D
models, Prescribed in z direction check boxes. Then enter a value or expression for u0,
v0, and for 3D models, w0.

THE BEAM INTERFACE | 737

 PRESCRIBED ROTATION

For 2D models, to define a prescribed rotation select the Prescribed in out

of plane direction check box and enter a value or expression for θ0z.

For 3D models, to define a prescribed rotation for each space direction (x,
y, and z), select one or all of the Prescribed around x direction, Prescribed
around y direction, and Prescribed around z direction check boxes and enter
a value or expression for in each θ0x, θ0y, or θ0z field.

In a geometrically nonlinear analysis in 3D, you should prescribe all three

components of the rotation vector. Prescribing only one or two
components may not give unique results, since finite rotations are not
commutative.

CONSTRAINT SETTINGS

To display this section, click the Show button ( ) and select Advanced Physics Options.

In the COMSOL Multiphysics Reference Manual:

• Constraint Reaction Terms

• Weak Constraints
• Constraint Settings

• You can add a Harmonic Perturbation subnode for specifying a

harmonic variation of the values of the prescribed displacements in a
frequency domain analysis of perturbation type.
• You can activate and deactivate this boundary condition by assigning it
to a constraint group. See Load Cases in the Structural Mechanics
Modeling chapter.
• You can assign the value of the prescribed displacement to a load
group. See Load Cases in the Structural Mechanics Modeling chapter.

738 | CHAPTER 7: BEAM

LOCATION IN USER INTERFACE

Context Menus
Beam>Line Constraints>Prescribed Displacement
Beam>Prescribed Displacement (Point)

Ribbon
Physics tab with Beam selected:
Edges>Line Constraints>Prescribed Displacement
Points>Beam>Prescribed Displacement

Prescribed Velocity
The Prescribed Velocity node adds an edge (3D), boundary (2D), or point (2D and
3D) that prescribes the translational or rotational velocity in one or more directions.
The prescribed velocity condition is applicable for Time Dependent and Frequency
Domain studies. With this condition it is possible to prescribe a velocity in one
direction, leaving the beam free in the other directions.

The Prescribed Velocity node is a constraint, and overrides any other constraint on the
same selection.

For details about prescribed velocities and accelerations, see Prescribed

Displacements, Velocities, and Accelerations.

COORDINATE SYSTEM SELECTION

Coordinate systems with directions that change with time should not be
used.

Specify the coordinate system to use for specifying the prescribed

translational/rotational velocity. See the section Coordinate System Selection for Edge
Load. Depending on the selected coordinate system, the velocity components change
accordingly.

P RE S C R I B E D VE L O C I T Y
Select one or all of the Prescribed in x direction, Prescribed in y direction, and for 3D
models, Prescribed in z direction check boxes. Then enter a value or expression for vx,
vy, and for 3D models, vz.

THE BEAM INTERFACE | 739

 P R E S C R I B E D A N G U L A R VE L O C I T Y

For 2D models, to define a prescribed angular velocity select the

Prescribed in out of plane direction check box and enter a value or
expression for ∂θ0z/∂t.

For 3D models, to define a prescribed angular velocity for each space

direction (x, y, and z), select one or all of the Prescribed around x direction,
Prescribed around y direction, and Prescribed around z direction check
boxes and enter a value or expression for in each ∂θ0x/∂t, ∂θ0y/∂t, or
∂θ0z/∂t field.

You can add a Harmonic Perturbation subnode for specifying a harmonic

variation of the values of the prescribed velocity in a frequency domain
analysis of perturbation type.

LOCATION IN USER INTERFACE

Context Menus
Beam>Line Constraints>Prescribed Velocity
Beam>More Constraints>Prescribed Velocity (Point)

Ribbon
Physics tab with Beam selected:
Edges>Line Constraints>Prescribed Velocity
Points>More Constraints>Prescribed Velocity

Prescribed Acceleration
The Prescribed Acceleration node adds a boundary or domain condition where the
acceleration is prescribed in one or more directions. The prescribed acceleration
condition is applicable for Time Dependent and Frequency Domain studies. With this
boundary condition it is possible to prescribe a acceleration in one direction, leaving
the beam free in the other directions.

740 | CHAPTER 7: BEAM

The Prescribed Acceleration node is a constraint, and overrides any other constraint on
the same selection.

For details about prescribed velocities and accelerations, see Prescribed

Displacements, Velocities, and Accelerations.

COORDINATE SYSTEM SELECTION

Coordinate systems with directions that change with time should not be
used.

Specify the coordinate system to use for specifying the prescribed

translational/rotational acceleration. See the section Coordinate System Selection for
Edge Load. Depending on the selected coordinate system, the acceleration
components change accordingly.

PRESCRIBED ACCELERATION
Select one or all of the Prescribed in x direction, Prescribed in y direction, and for 3D
models, Prescribed in z direction check boxes. Then enter a value or expression for ax,
ay, and for 3D models, az.

PRESCRIBED ANGULAR ACCELERATION

For 2D models, to define a prescribed angular acceleration select the

Prescribed in out of plane direction check box and enter a value or
expression for ∂2θ0z/∂t2.

For 3D models, to define a prescribed angular acceleration for each space

direction (x, y, and z), select one or all of the Prescribed around x direction,
Prescribed around y direction, and Prescribed around z direction check
boxes and enter a value or expression for in each 2θ0x/∂t2, 2θ0y/∂t2, or
2
θ0z/∂t2 field.

THE BEAM INTERFACE | 741

 You can add a Harmonic Perturbation subnode for specifying a harmonic
variation of the values of the prescribed acceleration in a frequency
domain analysis of perturbation type.

LOCATION IN USER INTERFACE

Context Menus
Beam>Line Constraints>Prescribed Acceleration
Beam>More Constraints>Prescribed Acceleration (Point)

Ribbon
Physics tab with Beam selected:
Edges>Line Constraints>Prescribed Acceleration
Points>More Constraints>Prescribed Acceleration

Pinned
The Pinned node adds an edge (3D), boundary (2D), or point (2D and 3D) condition
that makes all nodes on the selected objects to have zero displacements; that is, all
translations are fixed while still allowing rotations.

CONSTRAINT SETTINGS
To display this section, click the Show button ( ) and select Advanced Physics Options.

In the COMSOL Multiphysics Reference Manual:

• Constraint Reaction Terms

• Weak Constraints
• Constraint Settings

You can activate and deactivate this boundary condition by assigning it to

a constraint group. See Load Cases in the Structural Mechanics Modeling
chapter.

742 | CHAPTER 7: BEAM

LOCATION IN USER INTERFACE

Context Menus
Beam>Line Constraints>Pinned(Line)
Beam>Pinned (Point)

Ribbon
Physics tab with Beam selected:
Edges>Line Constraints>Pinned
Points>Beam>Pinned

Thermal Expansion (for constraints)

Add the Thermal Expansion subnode to a constraint (Fixed or Prescribed Displacement)
to prescribe a deformation of the constraint caused by changes in temperature of the
surroundings. This makes it possible to reduce stresses caused by the boundary
conditions.

The thermal strain depends on the coefficient of thermal expansion α, the temperature
T, and the strain-free reference temperature Tref as

ε th = α ( T – T ref )

THERMAL EXPANSION PROPERTIES

Specify the thermal properties that define the thermal strain. This is a description of
the thermal expansion of surroundings idealized by the constraints.

Select Inherit from edge to take the thermal expansion data from the edge being
constrained. This should only be used when:

• The temperature and the thermal expansion coefficient do not have a spatial
variation.
• The virtual surrounding material has the same thermal expansion as the edge itself.

When Inherit from edge is not selected, enter:

• A value or expression for Temperature T, specifying the temperature distribution of

the surrounding material. Any spatial variation must be an explicit function of the
material frame coordinates. It is not possible to use a computed temperature
distribution.
• The Coefficient of thermal expansion α. As a default, values From material are used.
This requires that a material has been assigned to the points where the constraint is

THE BEAM INTERFACE | 743

 active.
For User defined enter the coefficient of thermal expansion α.
• A value or expression for the Strain reference temperature Tref which is the
temperature at which there are no thermal displacements at the constraints.

Enter the coordinates of the Reference point, the point where the displacement and
rotation is zero. The choice of reference point only affects the rigid body motion. If
there are several different constraints with a Thermal Expansion subnode, the same
reference point should usually be selected in all of them.

• Constraints and Thermal Expansion in the Structural Mechanics

Modeling chapter.
• Thermal Expansion of Constraints in the Structural Mechanics Theory
chapter.

LOCATION IN USER INTERFACE

Context Menus
Beam>Fixed>Thermal Expansion
Beam>Prescribed Displacement>Thermal Expansion

Ribbon
Physics tab with Fixed or Prescribed Displacement node selected in the model tree:
Attributes>Thermal Expansion

No Rotation
The No Rotation node adds an edge (3D), boundary (2D), or point (2D and 3D)
condition that prevents all rotation at the selected objects while still allowing
translational motion.

744 | CHAPTER 7: BEAM

CONSTRAINT SETTINGS
To display this section, click the Show button ( ) and select Advanced Physics Options.

In the COMSOL Multiphysics Reference Manual:

• Constraint Reaction Terms

• Weak Constraints
• Constraint Settings

You can activate and deactivate this boundary condition by assigning it to

a constraint group. See Load Cases in the Structural Mechanics Modeling
chapter.

LOCATION IN USER INTERFACE

Context Menus
Beam>Line Constraints>No Rotation (Line)
Beam>More Constraints>No Rotation (Point)

Ribbon
Physics tab with Beam selected:
Edges>Line Constraints>No Rotation
Points>More Constraints>No Rotation

Symmetry
The Symmetry node adds an edge (3D), boundary (2D), or point (2D and 3D)
condition that defines asymmetry edge, boundary, or point.

COORDINATE SYSTEM SELECTION

Specify the coordinate system to use for specifying the orientation of the symmetry
plane. See the section Coordinate System Selection for Edge Load.

SYMMETRY
Select an Axis to use as symmetry plane normal to specify the direction of the normal to
the symmetry plane in the selected coordinate system.

• For 2D models, select 1 or 2 for the first or second axis, respectively.

• For 3D models, select 1, 2, or 3 for the first, second, or third axis, respectively.

THE BEAM INTERFACE | 745

 CONSTRAINT SETTINGS
To display this section, click the Show button ( ) and select Advanced Physics Options.

In the COMSOL Multiphysics Reference Manual:

• Constraint Reaction Terms

• Weak Constraints
• Constraint Settings

Symmetry Constraints

You can activate and deactivate this boundary condition by assigning it to

a constraint group. See Load Cases in the Structural Mechanics Modeling
chapter.

LOCATION IN USER INTERFACE

Context Menus
Beam>Line Constraints>Symmetry
Beam>More Constraints>Symmetry (Point)

Ribbon
Physics tab with Beam selected:
Edges>Line Constraints>Symmetry
Points>More Constraints>Symmetry

Antisymmetry
The Antisymmetry node adds an edge (3D), boundary (2D), or point (2D and 3D)
condition that defines an antisymmetry edge, boundary, or point.

COORDINATE SYSTEM SELECTION

Specify the coordinate system to use for specifying the antisymmetry. See the section
Coordinate System Selection for Edge Load.

746 | CHAPTER 7: BEAM

ANTISYMMETRY
Select an Axis to use as anti-symmetry plane normal to specify the direction of the
normal to the symmetry plane in the selected coordinate system.

• For 2D models, select 1 or 2 for the first or second axis, respectively.

• For 3D models, select 1, 2, or 3 for the first, second, or third axis, respectively.

CONSTRAINT SETTINGS
To display this section, click the Show button ( ) and select Advanced Physics Options.

In the COMSOL Multiphysics Reference Manual:

• Constraint Reaction Terms

• Weak Constraints
• Constraint Settings

Symmetry Constraints

You can activate and deactivate this boundary condition by assigning it to

a constraint group. See Load Cases in the Structural Mechanics Modeling
chapter.

LOCATION IN USER INTERFACE

Context Menus
Beam>Line Constraints>Antisymmetry (Line)
Beam>More Constraints>Antisymmetry (Point)

Ribbon
Physics tab with Beam selected:
Edges>Line Constraints>Antisymmetry
Points>More Constraints>Antisymmetry

Edge Load
Add an Edge Load as a force or moment distributed along an edge.

THE BEAM INTERFACE | 747

 COORDINATE SYSTEM SELECTION
Specify the coordinate system to use for specifying the load. From the Coordinate
system list select from the following based on space dimension:

For 2D models (boundaries):

• Global coordinate system (the default)

• Boundary System (a predefined normal-tangential coordinate system)
• Any additional user-defined coordinate system

For 3D models (edges):

• Global coordinate system (the default; the standard global coordinate

system).
• Local edge system. This is the coordinate system defined by the beam
cross section orientation as defined in the Section Orientation subnode
under Cross Section Data. The local edge coordinate systems directions
are available as variables for plotting using an Arrow Line plot, for
example: Under Beam>Beam Local System in the plot settings lists of
predefined expressions, select Base vector (beamsys) x,
Base vector (beamsys) y, or Base vector (beamsys) z.
• Any additional user-defined coordinate system.

748 | CHAPTER 7: BEAM

FORCE
Select a Load type—Frce per unit length (the default), Force per unit volume, or Total
force. Enter values or expressions for the components.

LOAD TYPE VARIABLE SI UNIT GEOMETRIC SPACE DIMENSION

ENTITY LEVEL (COMPONENTS)

Force per unit length FL N/m edges 3D (x, y, z)

boundaries 2D (x, y)
Force per unit volume FV N/m3 edges 3D (x, y, z)
boundaries 2D (x, y)
Total force Ftot N edges 3D (x, y, z)
boundaries 2D (x, y)

• When Force per unit volume is selected, the given load is multiplied by
the cross section area. This option is useful for modeling body loads
like gravity or centrifugal loads.
• After selecting a Load type, the Load list normally only contains User
defined. When combining the Beam interface with another physics
interface, it is also possible to choose a predefined load from this list.

MOMENT
Enter values or expressions for the components of the moment edge load ML (3D) or
Mlz (2D).

• You can add the Phase subnode to specify the phase of this load in a
frequency domain analysis.
• You can specify this load to be a Harmonic Perturbation in a frequency
domain analysis.
• You can assign this load to a load group. See Load Cases in the
Structural Mechanics Modeling chapter.

LOCATION IN USER INTERFACE

Context Menus
Beam>Line and Volume Loads>Edge Load
Beam Rotor>Edge Load

THE BEAM INTERFACE | 749

 Ribbon
Physics tab with Beam selected:
Edges>Line and Volume Loads>Edge Load

Physics tab with Beam Rotor selected:

Edges>Edge Load

Point Load
Add a Point Load to points for concentrated forces or moments at points. The loads are
defined in the given coordinate system.

POINT LOAD
Enter values or expressions for the components (x, y, z for 3D models and x, y for 2D
models) of the point load FP.

POINT MOMENT
Enter values or expressions for the components (x, y, z for 3D models and z for 2D
models) of the point moment MP (3D) or Mlz (2D).

• The FP list normally only contains User defined. When combining the
Beam interface with another physics interface, it is also possible to
choose a predefined load from this list.
• You can add the Phase subnode to specify the phase of this load in a
frequency domain analysis.
• You can specify this load to be a Harmonic Perturbation in a frequency
domain analysis.
• You can assign this load to a load group. See Load Cases in the
Structural Mechanics Modeling chapter.

LOCATION IN USER INTERFACE

Context Menus
Beam>Point Load
Beam Rotor> Point Load

Ribbon
Physics tab with Beam or Beam Rotor selected:
Points>Beam>Point Load

750 | CHAPTER 7: BEAM

Point Mass
Use the Point Mass node to model a discrete mass or mass moment of inertia which is
concentrated at a point.

The Point Mass Damping subnode can be added to specify a mass-proportional

damping.

COORDINATE SYSTEM SELECTION

With the Coordinate system list, select the coordinate system in which the mass
moment of inertia tensor is defined.

PO IN T MA SS
Enter a Point mass m.

Enter a value or values for the Mass moment of inertia. For 2D models, enter one value
for Jz. For 3D models, enter a single value for an isotropic moment of inertia tensor,
or select Diagonal or Symmetric to enter a full moment of inertia tensor.

FRAME ACCELERATION FORCES

Click to select the Exclude contribution check box to switch off the loads that can be
caused by the point mass when the frame is accelerated when using a Gravity or
Rotating Frame feature. The setting will also determine whether the node will
contribute when Computing Mass Properties.

LOCATION IN USER INTERFACE

Context Menus
Beam>Mass, Spring, and Damper>Point Mass

Ribbon
Physics tab with Beam selected:
Points>Mass, Spring, and Damper>Point Mass

Point Mass Damping

Use the Point Mass Damping subnode to add damping to a Point Mass parent node.

PO IN T MASS DAMPIN G
Enter a Mass damping parameter αdM. This is the mass proportional term of a Rayleigh
damping.

THE BEAM INTERFACE | 751

 LOCATION IN USER INTERFACE

Context Menus
Beam>Point Mass>Point Mass Damping

Ribbon
Physics tab with Point Mass node selected in the model tree:
Attributes>Point Mass Damping

Solid Connection

The Solid Connection node is obsolete, and has been superseded by the
Solid-Beam Connection multiphysics coupling. It cannot be added in
version 5.3 and later, but may be present in models created by earlier
versions of the software.

You are advised to update your model to use the new multiphysics
coupling instead, since the Solid Connection node will be removed in
future versions.

Use a Solid Connection node for connecting beams to solids in 2D models. To complete
the connection, a corresponding Beam Connection node must also be added in the Solid
Mechanics interface. The Solid Connection node can be added to either points or
boundaries.

When using it for a point, the connection is intended for a transition from a beam to
a solid, so that beam theory is assumed to be valid on both sides of the connection.

When using it for a boundary, the beam acts as a cladding or stiffener on the solid.

SOLID CONNECTION (POINTS ONLY)

Select a Connection type—Softened or Simplified. When using Softened (the default),
three extra degrees of freedom are added to each selected point. This allows for a more
accurate description of the transition, but the model can in some cases become
underconstrained if the mesh on the solid is very coarse. The Simplified version of the
connection just adds constraints to the boundary of the solid, and will in general cause
local disturbances of the stress field.

752 | CHAPTER 7: BEAM

When using the Solid Connection node for a boundary, the connection is intended for
placing a beam as a stiffener or cladding along the boundary of a solid. No other
settings than the boundary selection are required.

• For more information about coupling different element types, see

Coupling Techniques.
• For details about the formulation of this coupling, see Connection
Between Shells and Solids

LOCATION IN USER INTERFACE

Context Menus
Beam>Connections>Solid Connection

Ribbon
Physics tab with Shell selected:
Boundaries>Connections>Solid Connection
Points>Connections>Solid Connection

Shell Connection

The Shell Connection node is obsolete, and has been superseded by the
Shell-Beam Connection multiphysics coupling. It cannot be added in
version 5.3 and later, but may be present in models created by earlier
versions of the software.

You are advised to update your model to use the new multiphysics
coupling instead, since the Shell Connection node will be removed in future
versions.

A beam can be connected to a shell by adding a Shell Connection node in the Beam
interface and a Beam Connection node in the Shell interface. The connection can be
between:

• two edges, one in each physics interface,

• a point in the Beam interface and a boundary in the Shell interface, or
• a point in the Beam interface and an edge in the Shell interface.

THE BEAM INTERFACE | 753

 There are no settings for the Shell Connection node except the selections of points or
edges.

• For more information about coupling different element types, see

Coupling Techniques.
• For details about the formulation of this coupling, see Connection
Between Shells and Beams

LOCATION IN USER INTERFACE

Context Menus
Beam>Connections>Shell Connection

Ribbon
Physics tab with Shell selected:
Edges>Connections>Shell Connection
Points>Connections>Shell Connection

Rigid Connector
The Rigid Connector is a boundary condition for modeling rigid regions and kinematic
constraints such as prescribed rigid rotations. The selected points will move as a single
rigid object.

If the study step is geometrically nonlinear, the rigid connector takes finite rotations
into account. The feature is similar to the rigid connectors in the Solid Mechanics and
Shell interfaces. Rigid connectors from Beam, Shell, and Solid Mechanics interfaces
can be attached to each other.

You can add functionality to the rigid connector through the following subnodes:

• Applied Force (Rigid Connector) to apply a force in given point.

• Applied Moment (Rigid Connector) to apply a moment.
• Mass and Moment of Inertia (Rigid Connector) to add extra mass and moment of
inertia in a given point.
• Spring Foundation (Rigid Connector) to add a translational or rotational spring or
damper in a given point.

754 | CHAPTER 7: BEAM

COORDINATE SYSTEM SELECTION
The Global coordinate system is selected by default. The Coordinate system list contains
any additional coordinate systems that the model includes. Prescribed displacements
or rotations are specified along the axes of this coordinate system. It is also used for
defining the axis directions of the moment of inertia tensor of the Mass and Moment of
Inertia subnode.

CENTER OF ROTATION
The center of rotation serves two purposes.

• If you prescribe the displacement of the rigid connector, this is the place where it is
fixed.
• Results are interpreted with respect to the center of rotation.

Select a Center of rotation — Automatic, Centroid of selected entities, or User defined.

• For Automatic the center of rotation is at the geometrical center of the selected
points. The constraints are applied at the center of rotation.
• For Centroid of selected entities a subnode for selection of the points is added to the
Model Builder. The center of rotation is located at the centroid of the selected
points, which do not need to be related to the points to which the rigid connector
is attached. As a special case, you can select a single point, and thus locate the center
of rotation at a certain point.
• For User defined, in the Global coordinates of center of rotation XC table enter
coordinates based on space dimension.

Once Centroid of selected entities is chosen, a default Center of Rotation:

Point subnode is added.

Select the Offset check box to add an optional offset vector to the definition of the
center of rotation. Enter values for the offset vector Xoffset.

The center of rotation used is the sum of the vector obtained from any of the input
methods and the offset vector.

X C = X C,input + X offset

PRESCRIBED DISPLACEMENT AT CENTER OF ROTATION

To define a prescribed displacement at the center of rotation for each space direction,
select one or several of the available check boxes then enter values or expressions for

THE BEAM INTERFACE | 755

 the prescribed displacements. The direction coordinate names can vary depending on
the selected coordinate system.

• Prescribed in x direction u0x

• Prescribed in y direction u0y
• For 3D components: Prescribed in z direction u0z

PRESCRIBED ROTATION AT CENTER OF ROTATION

Specify the rotation at the center of rotation. Select from the By list: Free (the default),
Constrained rotation, or Prescribed rotation at center of rotation.

For 2D components, the Constrained rotation and Prescribed rotation at

center of rotation is always about the z-axis, so no component selection is
necessary.

Constrained Rotation (3D Components)

For Constrained rotation select one or more of the available check boxes to enforce zero
rotation about the corresponding axis in the selected coordinate system:

• Constrain rotation about x-axis

• Constrain rotation about y-axis
• Constrain rotation about z-axis

Prescribed Rotation at Center of Rotation

• For Prescribed rotation at center of rotation enter an Angle of rotation φ0. For 3D
components also enter an Axis of rotation Ω for the x, y, and z coordinates.

• You can add a Harmonic Perturbation subnode for specifying a

harmonic variation of the values of the prescribed displacements and
rotations in a frequency domain analysis of perturbation type.
• You can activate and deactivate the rigid connector by assigning it to a
constraint group. See Load Cases in the Structural Mechanics
Modeling chapter.
• You can assign the value of the prescribed displacement and rotation to
a load group. See Load Cases in the Structural Mechanics Modeling
chapter.

756 | CHAPTER 7: BEAM

 • Rigid Connector Theory
• Harmonic Perturbation
• Load Cases

LOCATION IN USER INTERFACE

Context Menus
Beam>Connections>Rigid Connector

Ribbon
Physics tab with Beam selected:
Points>Connections>Rigid Connector

Attachment
The Attachment node is used to define a point or a set of points on a flexible beam
which can be used to connect it with other components through a joint in the
Multibody Dynamics interface. All the selected points behave as if they were connected
by a common rigid body.

Attachments can be added to boundaries in a Multibody Dynamics or a Solid

Mechanics interface, to edges in a Shell interface, or to points in a Beam interface. This
makes it possible to use a joint in the Multibody Dynamics interface for connecting
parts modeled in different physics interfaces.

The Attachment node is available with the addition of the Multibody

Dynamics Module. The Attachments is included in the Multibody
Dynamics Module User’s Guide.

LOCATION IN USER INTERFACE

Context Menus
Beam>Connections>Attachment

Ribbon
Physics tab with Beam selected:
Points>Connections>Attachment

THE BEAM INTERFACE | 757

 Phase
You can add a Phase subnode to nodes which define a load in order to prescribe the
phase angle in a frequency domain analysis.

For modeling the frequency response the physics interface splits the harmonic load
into two parameters:

• The amplitude, F, which is specified in the node for the load.

• The phase ( φ ), which is specified in the Phase subnode.

Together these define a harmonic load, for which the amplitude and phase shift can
vary with the excitation frequency, f:

F freq = F ( f ) ⋅ cos ( 2πft + φ )

PHASE
Add the phase load Fph for harmonic loads. Enter the phase for each component of
the load in the corresponding fields.

MOMENT LOAD PHASE

Add the phase for the moment load Mph for harmonic loads. Enter the phase for each
component of the moment load in the corresponding fields.

LOCATION IN USER INTERFACE

Context Menus
Beam>Edge Load>Phase
Beam>Point Load>Phase

Ribbon
Physics tab with Edge Load or Point Load selected:
Attributes>Phase

Harmonic Perturbation
Use the Harmonic Perturbation subnode to specify the harmonic part of non-zero
prescribed displacements or rotations. This node is used if the study step is frequency
response of a perturbation type.

758 | CHAPTER 7: BEAM

The settings are the same as in the parent Prescribed Displacement/Rotation,
Prescribed Velocity, or Prescribed Displacement/Rotation node. Only degrees of
freedom selected as prescribed in the parent node can be assigned a value.

See Harmonic Perturbation in the Structural Mechanics Modeling

chapter.

LOCATION IN USER INTERFACE

Context Menus
Beam>Prescribed Displacement/Rotation>Harmonic Perturbation
Beam>Prescribed Velocity>Harmonic Perturbation
Beam>Prescribed Acceleration>Harmonic Perturbation

Ribbon
Physics tab with Prescribed Displacement/Rotation, Prescribed Velocity, or Prescribed
Acceleration node selected in the model tree:
Attributes>Harmonic Perturbation

THE BEAM INTERFACE | 759

760 | CHAPTER 7: BEAM
 8

Beam Cross Section

This chapter describes the Beam Cross Sections interface, which you find under
the Structural Mechanics branch ( ) when adding a physics interface.

In this chapter:

• Using the Beam Cross Section Interface

• Theory for the Beam Cross Section Interface
• The Beam Cross Section Interface

761
 Using the Beam Cross Section
Interface
In this section:

• About Beams and Cross Section Data

• Using the Beam Cross Section Interface

About Beams and Cross Section Data

A beam is a slender structure that can be fully described by cross section properties like
area, moments of inertia. In COMSOL Multiphysics there are two physics interfaces
for analyzing beam structures, one in 2D and one in 3D.

APPROACHES FOR OBTAINING THE CROSS SECTION DATA

The values for the cross section properties can be found in different ways. By:

• looking up tabulated values,

• inserting dimensions into handbook formulas,
• using the built-in common sections in The Beam Interface, and
• using Theory for the Beam Cross Section Interface.

For the common built-in cross sections available with the Beam interface
(Rectangular, Box, Circular, Pipe, H-section, U-section, and T-section),
using the Common sections option in the Cross Section Data settings
provides you with the most efficient input alternative. See Common
Cross Sections for details.

For non-standard cross sections with non-trivial shapes, using the Beam Cross Section
interface is the only realistic alternative. This physics interface also provides you with
more accurate data than the other methods, since it does not rely on common
engineering approximations.

The cross section data obtained from the different approaches does, in general, differ
somewhat. Two common sources for this difference is that the influence of fillets in the

762 | CHAPTER 8: BEAM CROSS SECTION

geometry can be excluded, and that approximations such as assuming thin walled
sections have been used.

STRESS COMPUTATION
Since the Beam interface is only aware of the cross section properties and not of the
true shape of the cross section, the stresses computed there can only be approximate.
The approximation used by COMSOL is conservative in the sense that all stresses are
assumed to interact in the worst possible way.

In the Beam Cross Section interface the true stress distribution can be computed given
a set of section forces. This can be important in cases where the stresses obtained in the
Beam interface exceed allowed values.

Using the Beam Cross Section Interface

The Beam Cross Section interface can either be used separately or as a separate model
node and geometry in the same model file where the actual beam problem is solved.

In most cases, it is easiest to compute the beam cross section properties in

a separate model.

There are some cases when using one or more Beam Cross Section interfaces and a
Beam interface together in the same model file can be advantageous. For example,
when changes in the cross sections can be anticipated. There are however a number of
things to pay attention to in this case:

• Usually the determination of the cross section data is more computationally

expensive than the actual analyses of the beam structure. Then it is best to use either
separate studies or study steps for the two tasks.
• When two separate studies or study steps are used, then the Values of variables not
solved for must be set in the second study step where only the beam problem is
solved. Under Dependent Variables you can also click to clear the Store in output
check box for the beam cross section degrees of freedom in order to save space.
• When referencing the beam cross section properties in the input fields of the Beam
interface, you must use a fully qualified variable name, like mod2.bcs3.A for the
area.
• If a Beam interface is added after a Beam Cross Section interface, the only study type
shown when adding a physics interface is Stationary. In this case, under Custom

USING THE BEAM CROSS SECTION INTERFACE | 763

 Studies, select Preset Studies for Some Physics Interfaces to find the other study types
available for the beam analysis.

Values of Dependent Variables and Physics and Variables Selection in the

COMSOL Multiphysics Reference Manual

Since the Beam Cross Section interface is active in 2D, the cross sections are analyzed
in the xy-plane. However, the Beam interface uses a notation where the local x-axis is
along the beam, and the cross section is described in a local yz-plane.

In order to avoid confusion, the cross section properties are described in local x1-x2
coordinates (see Figure 8-1). When data is transferred to a Beam interface, you must
keep track of the coordinates that correspond to the local y and z directions,
respectively.

Channel Beam: Application Library path Structural_Mechanics_Module/

Verification_Models/channel_beam

COMPUTING THE CROSS SECTION DATA

In a 2D model, the geometry of the cross section is drawn. If the section is simply
connected (that is, has no internal holes), then usually nothing else needs to be done
before running the analysis.

The default mesh density is tuned for thin-walled sections. For solid sections an
unnecessarily large model is obtained when using the default mesh.

If the section is not simply connected, add one Hole node for each internal hole. In
that node, select all boundaries around the hole.

The computed cross section data is stored in the variables listed in Table 8-1:
TABLE 8-1: VARIABLES CONTAINING CROSS SECTION DATA

VARIABLE DESCRIPTION INPUT TO BEAM INTERFACE LINK TO THEORY

SECTION

bcs.A Area Area (A) Area

bcs.CGx Center of gravity, x Implicit, since it affects the
coordinate positioning of the beam
Center of
centerline.
Gravity
bcs.CGx Center of gravity, y As above
coordinate

764 | CHAPTER 8: BEAM CROSS SECTION

TABLE 8-1: VARIABLES CONTAINING CROSS SECTION DATA

VARIABLE DESCRIPTION INPUT TO BEAM INTERFACE LINK TO THEORY

SECTION

bcs.x1 First coordinate in Implicit, since it can be

principal axes system used for determining stress
Local
evaluation locations
Coordinates
bcs.x2 Second coordinate in As above
principal axes system
bcs.I1 Largest principal moment Moment of inertia around
of inertia local y/z-axis (Izz/Iyy)
bcs.I2 Smallest principal As above
moment of inertia
bcs.Ixx Moment of inertia around -
Moments of
x-axis
Inertia
bcs.Iyy Moment of inertia around -
y-axis
bcs.Ixy Deviatoric moment of -
inertia in xy system
bcs.rg Radius of gyration -
bcs.alpha Angle from x-axis to first Rotation of vector around Directions of
principal axis beam axis (φ) Principal Axes
bcs.mu1 Max shear stress factor, Max shear stress factor in
x1 direction local y/z direction (μy/μz)
bcs.mu2 Max shear stress factor, As above
x2 direction
bcs.kappa1 Shear correction factor, -
x1 direction Bending Shear
bcs.kappa2 Shear correction factor, - Stresses
x2 direction
bcs.ei1 Shear center location, Distance to shear center in
first local coordinate local y/z direction.
bcs.ei2 Shear center location, As above
second local coordinate
bcs.J Torsional constant Torsional constant (J)
bcs.Wt Torsional section Torsional section modulus Torsion
modulus (Wt)

USING THE BEAM CROSS SECTION INTERFACE | 765

 TABLE 8-1: VARIABLES CONTAINING CROSS SECTION DATA

VARIABLE DESCRIPTION INPUT TO BEAM INTERFACE LINK TO THEORY

SECTION

bcs.Cw Warping constant -

bcs.Kw Warping section modulus -
Warping
bcs.sw Non-uniform torsion -
parameter

COMPUTING DETAILED STRESSES

If you have set of section forces (axial force, shear forces, bending moments, and
twisting moments), it is possible to display the stresses it causes. To do this, enter the
values in the Section Forces section. You can also add your own acceptance criteria by
adding one or more Safety nodes.

The stresses are available in the variables listed in Table 8-2.

After changing data in this section, you do not need to compute the study
for this physics interface again once is has been solved. It is sufficient to
do an Update Solution to get the stress plots updated.

Studies and Solvers in the COMSOL Multiphysics Reference Manual

TABLE 8-2: VARIABLES CONTAINING STRESS DISTRIBUTIONS

VARIABLE DESCRIPTION AND LINK TO THEORY

SECTION

bcs.sN Stress from axial force Axial Stress

bcs.sM1 Bending stress from moment around x1 axis
Bending Axial Stresses
bcs.sM2 Bending stress from moment around x2 axis

766 | CHAPTER 8: BEAM CROSS SECTION

TABLE 8-2: VARIABLES CONTAINING STRESS DISTRIBUTIONS

VARIABLE DESCRIPTION AND LINK TO THEORY

SECTION

bcs.tT1x Shear stress from force in x1 direction, x

component
bcs.tT1y Shear stress from force in x1 direction, y
component
bcs.resT1 Shear stress from force in x1 direction,
resultant Bending Shear
bcs.tT2x Shear stress from force in x2 direction, x Stresses
component
bcs.tT2y Shear stress from force in x2 direction, y
component
bcs.resT2 Shear stress from force in x2 direction,
resultant
bcs.tMtx Shear stress from twisting moment, x
component
bcs.tMty Shear stress from twisting moment, y Torsional Shear
component Stresses
bcs.resMt Shear stress from twisting moment,
resultant
bcs.mises von Mises effective stress Effective Stress

USING THE BEAM CROSS SECTION INTERFACE | 767

 Theory for the Beam Cross Section
Interface
In this section the theory for the Beam Cross Section interface is described:

• Cross Section Properties

• Computation of Stresses

Cross Section Properties

The following cross section properties computed by the Beam Cross Section interface
are described in this section:

• Area • Local Coordinates

• Center of Gravity • Bending Shear Stresses
• Moments of Inertia • Torsion
• Directions of Principal Axes • Warping

AREA
The area is computed as:

A =  dA
A

CENTER OF GRAVITY
The center of gravity is computed as:

1

x CG = ---- x dA
A
A

1

y CG = ---- y dA
A
A

MOMENTS OF INERTIA
The moments of inertia in the XY coordinate system are:

768 | CHAPTER 8: BEAM CROSS SECTION

  ( y – yCG )
2
I xx = dA
A

 ( x – xCG )
2
I yy = dA
A

I xy =  ( x – xCG ) ( y – yCG ) dA
A

Since the input data required by the Beam interface is the principal moments of inertia,
these must also be computed. Using the radius of the Mohr’s circle:

I xx – I yy 2
 --------------------- 2
R = - + I xy
 R 

the principal moments of inertia can be expressed as:

I xx + I yy
I 1 = ---------------------- + R
2

I xx + I yy
I 2 = ---------------------- – R
2

As an auxiliary variable, the radius of gyration is computed, using the expression

I1 + I2
rg = ----------------
A

DIRECTIONS OF PRINCIPAL AXES

The angle needed to rotate the x-axis to the axis of the largest principal moment of
inertia (x1) is denoted α. From the definition of Mohr’s circle, the angle is:

– 2I xy
α = --- atan  ----------------------
1
2 I xx – I yy

When implemented using the atan2 function, the angle can be correctly evaluated for
all rotations, and returns in the interval -π < α < π.

THEORY FOR THE BEAM CROSS SECTION INTERFACE | 769

 Figure 8-1: Local coordinate system and rotation angle.

LOCAL COORDINATES
The local coordinate system, having its origin in the center of gravity, and orientation
given by the principal moments of inertia is given by:

x 1 = ( x – x CG ) cos α + ( y – y CG ) sin α
x 2 = ( y – y CG ) cos α – ( x – x CG ) sin α

BENDING SHEAR STRESSES

The shear stresses caused by bending cannot be given a simple closed form solution,
but must be solved using two independent partial differential equations, one for the
force in each direction. The complete derivation is given at the end of this section. First
some quantities computed from the shear stresses are defined.
(1)
The following notation is used: τ 2z . This is a shear stress in the x2 direction (acting
on the plane with z as normal) caused by a unit shear force acting in the x1 direction.

770 | CHAPTER 8: BEAM CROSS SECTION

Max Shear Stress Factor
The max shear stress factor is the ratio between the maximum shear stress in the cross
section and the average shear stress. For a shear force in the x1 direction, the definition
is:

(1)
max ( τ ) (1)
μ 1 = --------------------------- = max ( τ ) ⋅ A
1
----
A

where

(1) (1) 2 (1) 2

τ = ( τ 1z ) + ( τ 2z )

is the resulting shear stress from a unit load in the x1 direction. Similarly:

(2)
μ 2 = max ( τ )⋅A

Shear Correction Factor

The shear correction factor is also computed. The shear correction factor is a
multiplier which makes the strain energy from the average shear stress and shear strain
in the cross section equal to the true shear energy in the cross section. The shear
correction factor can be introduced through the concept shear area. The shear area is
the reduced area which should replace the true area when computing the shear
deformation of a beam. In terms of the shear correction factor it can be written as:

(1)
As = κ1 A

where κ 1 is the shear correction factor for a shear force in the x1 direction. Thus, for
a shear flexible beam, the constitutive relation for the average shear is

τ T T
γ = -------- = ------------ = -----------
κG κGA GA s

To compute the shear correction factor, the true strain energy based on the actual
stress and strain distribution is set equal to the strain energy from the average shear
stress, when acting over the shear area. The full energy expression is

1 (1) (1) (1) (1)


--- ( τ 1z γ 1z + τ 2z γ 2z ) dA
2
A

The strain energy based on the averaged shear stress and shear strain is

THEORY FOR THE BEAM CROSS SECTION INTERFACE | 771

 1 T (1) T 2
--- τ 1z γ 1z A = --- ------1 γ 1z A = --------------  ------1 A
1 (1) (1) 1
2 2A 2κ 1 G A  

giving

2
1- (1) 2 (1) 2 T1

-------
2G
( ( τ 1z ) + ( τ 2z ) ) dA = -------------------
2Gκ 1 A
A

Since T1 is a unit shear force, the shear correction factor can be computed as

 (1) 2  –1
 
κ 1 =  A ( τ ) dA

A

Similarly:

 (2) 2  –1
 
κ 2 =  A ( τ ) dA

A

Shear Center Distance

The shear center (or, equivalently, the center of rotation) is the point around which
the shear stresses from bending has no torque. In COMSOL it is represented as the
distance from the center of gravity of the cross section in the principal axes coordinate
system. The torque can be computed as:


0 = ( – τ 1z ( e 2 – x 2 ) + τ 2z ( e 1 – x 1 ) ) dA =
A

 ( τ1z x2 + –τ2z x1 ) dA + e1  τ2z dA – e2  τ1z dA

A A A

Since there are two separate solutions for the shear stresses, it is possible to split the
determination of the two shear center coordinates into:

(2) (2)

e 1 = – ( τ 1z x 2 + – τ 2z x 1 ) dA
A
(1) (1)

e 2 = ( τ 1z x 2 + – τ 2z x 1 ) dA
A

Here the fact that the shear force resultant has a unit value has been used.

772 | CHAPTER 8: BEAM CROSS SECTION

Derivation of the Equations for Computing the Shear Stresses
Basic beam theory assumptions gives the following stress components:

σ1 = 0
σ2 = 0
M1 x2 M2 x1 (8-1)
σ z = -------------- – --------------
I1 I2
τ 12 = 0

The shear forces are related to the bending moments through

∂M 2
T 1 = – -----------
∂z
∂M 1
T 2 = -----------
∂z

The static equilibrium equations are:

σ 1, 1 + τ 12, 2 + τ 1z,z = 0
τ 12, 1 + σ 2, 2 + τ 2z,z = 0
τ 1z,1 + τ 2z,2 + σ z,z = 0

Insertion of the known stresses into the equilibrium equations gives:

τ 1z,z = 0
τ 2z,z = 0
T2 x2 T1 x1
τ 1z,1 + τ 2z,2 + ------------- + ------------- = 0
I1 I2

The two first equations simply state that the shear stresses are independent of z,
whereas the third equation is the one on which to focus the interest. Assume that the
shear stresses can be derived from a scalar stress potential ψ , through:

2
νT 1 x 2
τ 1z = ψ ,1 + ---------------------------
2 ( 1 + ν )I 2
2
νT 2 x 1
τ 2z = ψ ,2 + ---------------------------
2 ( 1 + ν )I 1

THEORY FOR THE BEAM CROSS SECTION INTERFACE | 773

 Insertion of this assumption into the third equilibrium equation gives the Poisson type
equation:

T2 x2 T1 x1
Δψ = –  ------------- + ------------- (8-2)
 I1 I2 

In addition to equilibrium, also compatibility must be fulfilled. The Beltrami-Michell

form of compatibility equations includes the assumption of an isotropic linear elastic
material, and then states that:

1 ∂( trace ( σ ) )
Δσ ij + ------------ -------------------------------- = 0
1 + ν ∂x i ∂x j

Given the stress state from Equation 8-1, the only two non-trivial equations are:

1 T1
Δτ 1z + ------------ ------ = 0
1 + ν I2
1 T2
Δτ 2z + ------------ ------ = 0
1 + ν I1

Inserting the assumed stress components gives:

νT 1 1 T1 T1
Δψ ,1 + ----------------------- + ------------ ------ = Δψ ,1 + ------ = 0
( 1 + ν )I 2 1 + ν I 2 I2
νT 2 1 T2 T2
Δψ ,2 + ----------------------- + ------------ ------ = Δψ ,2 + ------ = 0
( 1 + ν )I 1 1 + ν I 1 I1

Integration of the first equation with respect to x1 and the second equation with
respect to x2 gives:

T1 x1
Δψ + ------------- + f 2 ( x 2 ) = 0
I2
T2 x2
Δψ + ------------- + f 1 ( x 1 ) = 0
I1

Combining these two equations results in:

T2 x2 T1 x1
Δψ = –  ------------- + -------------
 I1 I2 

774 | CHAPTER 8: BEAM CROSS SECTION

This is the same equation as Equation 8-2. It is thus possible to fulfill equilibrium,
compatibility, and the constitutive relation with a single equation of Poisson type.

On all free boundaries, the stress must be zero, giving the condition:

τ 1z n 1 + τ 2z n 2 = 0

Using the assumed shear stresses, this can be converted into the Neumann condition:

2 2
 νT 1 x 2 νT 2 x 1 
∇ψ ⋅ n = –  --------------------------- n 1 + --------------------------- n 2
 2 ( 1 + ν )I 2 2 ( 1 + ν )I 1 

It must also be determined that the resultant of the shear stresses actually match the
applied shear forces, that is:

 τ1z dA = T1
A

 τ2z dA = T2
A

The proofs for the two components are analogous, so it is shown only for the x1
direction:

2
 νT 1 x 2  νT 1 I 1
 τ1z dA =   ψ ,1 + --------------------------- dA =
 2 ( 1 + ν )I 2  ψ,1 dA + --------------------------
2 ( 1 + ν )I 2
-
A A A

To compute the integral of the x1 derivative of ψ , a term containing the differential

equation itself is added. This is a zero contribution, but it makes further simplifications
possible.

THEORY FOR THE BEAM CROSS SECTION INTERFACE | 775

 T2 x2 T1 x1
 ψ,1 dA =   ψ,1 + x1  Δψ + ------------
I1
- + -------------  dA =
I2  
A A
∂ ∂
 ( ψ,1 + x1 Δψ ) dA + T1 =   ∂ x1 ( x1 ψ,1 ) + ∂ x2 ( x1 ψ,2 ) dA + T1 =
A A

°Γ ( ( x1 ψ,1 )n1 + ( x1 ψ,2 )n2 ) ds + T1 = °Γ x1 ( ∇ψ ⋅ n )ds + T1 =

2 2
  νT 1 x 2 νT 2 x 1 
°Γ x 1  –  --------------------------- n 1 + --------------------------- n 2  ds + T 1 =
  2 ( 1 + ν )I 2 2 ( 1 + ν )I 1  

2
 νT 1 x 2  –ν T1 I1
  – --------------------------- dA + T 1 = --------------------------- + T 1
 2 ( 1 + ν )I 2 2 ( 1 + ν )I 2
A

In the transformations above these facts are used:

• The area integral of x1, x1 or x1x2 are zero since the coordinate system is positioned
at the center of gravity and is oriented along the principal axes.
2
• The area integral of x 2 is I1.
• The divergence theorem is used for transformation between area and surface
integrals.
• The Neumann condition on the boundary is inserted into the integral along the
boundary.

This proves that the assumed stress field also produces the correct resultants.

When internal holes are present, it is necessary ensure compatibility in the sense that
the displacement is single valued when going around the hole:

°Γ du1 = °Γ du2 = °Γ duz = 0

The displacements can be derived from the strains, which are given by the stress state:

776 | CHAPTER 8: BEAM CROSS SECTION

 ∂u 1
ε 1 = --------- = – νε z
∂x 1
∂u 2
ε 2 = --------- = – νε z
∂x 2
∂u z M 1 x 2 M 2 x 1
ε z = --------- = -------------- – --------------
∂z EI 1 EI 2
∂u 1 ∂u 2
γ 12 = --------- + --------- = 0
∂x 2 ∂x 1
2
∂u 1 ∂u z 1  νT 1 x 2 
γ 1z = --------- + --------- = ----  ψ ,1 + ---------------------------
∂z ∂x 1 G  2 ( 1 + ν )I 2
2
∂u 2 ∂u z 1  νT 2 x 1 
γ 2z = --------- + --------- = ----  ψ ,2 + ---------------------------
∂z ∂x 2 G  2 ( 1 + ν )I 1

Integration of the direct strains gives:

2
νM 1 x 1 x 2 νM 2 x 1
u 1 = – ------------------------ + ------------------ + g 1 ( x 2, z )
EI 1 2EI 2
2
νM 1 x 2 νM 2 x 1 x 2
u 2 = – ------------------ + ------------------------ + g 2 ( x 1, z )
2EI 1 EI 2

Since the only part of the displacement that is relevant for the bending shear stresses is
independent of the z coordinate, the functions g1 and g2 can be considered as
independent of z.

∂u 1
 - dx + --------- dx  = ∂u 1
°Γ du1 = °Γ  --------
∂x 1 1 ∂x 2 2

  νM 1 x 2 νM 2 x 1 νM 1 x 1 ∂g 1
- – ------------------  dx 1 +  – ------------------ + --------- dx 2 =
°Γ   – -----------------
EI 1 EI 2    EI 1 ∂x 2 

 νM 1 x 1 x 2  νM 1 x 1 x 2
 
 ------------------------ dA –  ------------------------ dA = 0
 EI 1   EI 1 
A A

In the last transformation Green’s theorem is used. The uniqueness of the u2

displacement can be shown in the same way.

THEORY FOR THE BEAM CROSS SECTION INTERFACE | 777

 The uniqueness of the out-of-plane component of the displacement is shown in
Equation 8-3:

∂u z ∂u z ∂u 1
 - dx + --------- dx  =   γ – --------- dx +  γ – --------- dx  = ∂u 2
°Γ duz = °Γ  --------
∂x 1 1 ∂x 2 2 °    1z ∂z  1  2z ∂z  2
Γ
2 2
1 νT 1 x 2   νT 1 x 1 x 2 νT 2 x 1 
°Γ  ----  ψ ,1 + --------------------------- –  ---------------------- + ----------------  dx 1 +
G
  2 ( 1 + ν )I 2  EI 1 2EI 2  

2 2
1 νT 2 x 1   νT 1 x 2 νT 2 x 1 x 2  (8-3)
°Γ  ----  ψ ,2 + --------------------------- –  ---------------- + ----------------------  dx 2 =
G 2 ( 1 + ν )I 1  2EI 1 EI 2  

νT 1 x 2 νT 1 x 1
–  ----  ψ ,12 + ----------------------- – ---------------- dA +
1
G  ( 1 + ν )I 2  EI 1 
A
νT 2 x 1 νT 2 x 2
  ---G-  ψ,21 + ----------------------
- – ---------------- dA = 0
1
( 1 + ν )I 1 EI 2 
A

In the last step of Equation 8-3 all integrals are zero since the coordinate system is
located at the center of gravity of the section. This proves that all displacement
components are unique.

When solving the problem, the shear stresses caused by a unit force in each of the two
principal directions must be separated, so two separate problems are solved. For the
force in the x1 direction it is formulated as:

(1) x1
Δψ = – -----
I2

with

2
(1) –ν x2
∇ψ ⋅ n = --------------------------- n 1
2 ( 1 + ν )I 2

on all boundaries. The stresses are computed as:

2
(1) (1) νx 2
τ 1z = ψ ,1 + ---------------------------
2 ( 1 + ν )I 2
(1) (1)
τ 2z = ψ ,2

778 | CHAPTER 8: BEAM CROSS SECTION

The corresponding problem for the x2 direction is:

(2) x2
Δψ = – -----
I1

2
(2) –ν x1
∇ψ ⋅ n = --------------------------- n 1
2 ( 1 + ν )I 1

(2) (2)
τ 1z = ψ ,1
2
(2) (2) νx 1
τ 2z = ψ ,2 + ---------------------------
2 ( 1 + ν )I 1

TO R S I O N
The torsional properties cannot in general be computed using a closed form
expression. Determining the torsional rigidity requires the solution of a PDE over the
cross section. There are two ways to do this: Using a warping function or using the
Prandtl stress function. The Prandtl stress function approach is used in COMSOL
since it gives easier boundary conditions.

The general torsion theory includes the shear modulus and angle of twist, but these
properties are not needed to determine the torsional rigidity, so both parameters are
treated as having the value 1. In that case the equation to be solved can be simplified
to:

Δφ = – 2

where φ is the stress function. For a singly connected region the boundary condition
is φ = 0 along the whole boundary. Having solved this problem the torsional rigidity
can be computed as:

J = 2 φ dA 
A

The shear stresses are defined as:

∂φ
τ xz = ------
∂y
∂φ
τ yz = – ------
∂x

THEORY FOR THE BEAM CROSS SECTION INTERFACE | 779

 The torsional modulus can be determined as:

J
W t = ----------------------------
max ( ∇φ )

In the case that there are internal holes in the section the situation is slightly more
complex. The φ = 0 condition is now applicable only to the external boundary,
whereas each boundary of an internal hole i needs a Dirichlet boundary condition:

φ = Hi

where H i is a constant to be determined. The constant value of the stress function

fulfills the stress free boundary conditions. There is also a compatible condition that
must be fulfilled: the displacements must be single valued when going around each
hole along its boundary Γ i . This is trivially fulfilled for the in-plane displacements, but
the out-of-plane displacement, w, generates the necessary equations to determine H i :

∂------ ∂w  γ –  ∂-----
u-   ∂v
dx +  γ yz –  ------   dy
w-
0=
°Γ dw = Γ° ∂x
dx + ------- dy =
∂y °
Γ
 xz  ∂z     ∂z  
i i i

 ∂-----φ- + y dx +  – ∂-----
φ-   ∂-----
φ- ∂φ
+ y t x –  ------ + x t y ds
=
°
Γ
 ∂y   ∂x 
– x dy =
°
Γ
 ∂y   ∂x 
i i

Here it has been used so that the strains are equal to the stresses since the shear
modulus is set to 1. The kinematic assumption that the in-plane displacements can be
written as:

u = – yz
v = xz

is employed. This assumption implies that the origin of the coordinate system is at the
center of rotation. This is true only for doubly symmetric sections, but adding a
constant offset to the x and y coordinates does not contribute to the integral.

Since the gradient of φ depends linearly on the yet unknown variables H i , the values
of which can be solved by adding one equation:

 ∂-----φ- + y t –  ∂-----
φ-  ds = 0
°
Γ
 ∂y  x  ∂x + x t y
i

for each hole.

780 | CHAPTER 8: BEAM CROSS SECTION

The expression for the torsional rigidity must in this case be augmented to:


J = 2 φ dA + 2  Hi Ai
A i

where A i is the area of hole i.

WARP I N G
The warping properties of the cross section are not used by the Beam interface in
COMSOL Multiphysics, since an assumption of pure St Venant torsion is used. The
data can still be useful to do manual estimates.

The warping function ω (x,y) describes the out-of-plane deformation related to

torsion. It fulfills the Laplace equation Δω = 0 .

The boundary conditions giving stress-free boundaries:

∇ω ⋅ n = ( x – e x )t x + ( y – e y )t y

The offset by the shear center coordinates (ex, ey) is introduced since the torsion
theory assumes that the coordinate system has it origin in the center of rotation (which
is the same as the shear center).

The level of the warping function must also be fixed by adding a Dirichlet condition
in a point. The actual value is however difficult to set. Instead it is easier to solve for a
shifted warping function:

ω s (x,y) = ω (x,y) + C

The shifted warping function can be set to zero in an arbitrary point. The true warping
function is then computed as:

1

ω (x,y) = ω s (x,y) – ---- ω s (x,y) dA
A
A

This criterion expresses that the average of the warping function must be zero since the
axial stresses induced by torsion should not have a resultant.

The warping constant, which is used in analysis of non-uniform torsion, is defined as:

 ω (x,y)
2
Cw = dA
A

THEORY FOR THE BEAM CROSS SECTION INTERFACE | 781

 The axial stress caused by non-uniform torsion is:

B ω (x,y)
σ w (x,y) = ---------------------
Cw

where B is the bimoment. The maximum axial stress is:

B max ( ω (x,y) ) B
σ w, max = -------------------------------------------- = --------
Cw Kw

The warping modulus is then defined as:

max ( ω (x,y) )
K w = ------------------------------------
Cw

Given the warping constant, it is possible to compute a non-dimensional number that

can be used to characterize the influence of non-uniform torsion in a beam with a
certain length L. This number is:

2 2
π EC w 2 ( 1 + ν )π C w
------------------- = -----------------------------------
-
2 2
GJL JL

Since the length is independent of the cross section, the sensitivity number is defined
as:

2
2 ( 1 + ν )π C w
s w = ------------------------------------
J

It has the physical dimension length squared.

Computation of Stresses
The stresses are computed using the following expressions:

• Axial Stress • Torsional Shear Stresses

• Bending Axial Stresses • Effective Stress
• Bending Shear Stresses

AXIAL STRESS
The axial stress is computed as:

782 | CHAPTER 8: BEAM CROSS SECTION

 N
σ N = ----
A

where N is the axial force.

BENDING AXIAL STRESSES

The bending stresses are computed as:

M1 x2
σ M1 = --------------
I1
M2 x1
σ M2 = – --------------
I2

Where M1 is the moment around the first principal axis and M2 is the moment around
the second principal axis.

BENDING SHEAR STRESSES

The components of the shear stresses caused by a shear force T1 along the x1 axis are:

(1) (1)
τ T1x = ( τ 1z cos α – τ 2z sin α ) ⋅ T 1
(1) (1)
τ T1y = ( τ 2z cos α + τ 1z sin α ) ⋅ T 1

The components of the shear stresses caused by a shear force T2 along the x2 axis are:

(2) (2)
τ T2x = ( τ 1z cos α – τ 2z sin α ) ⋅ T 2
(2) (2)
τ T2y = ( τ 2z cos α + τ 1z sin α ) ⋅ T 2

In both cases, the resultants are also computed as:

2 2 2 2
τ T1 = τ T1x + τ T1y and τ T2 = τ T2x + τ T2y

TO R S I O N A L S H E A R S T R E S S E S
The components of the shear stress caused by Saint-Venant torsion are:

THEORY FOR THE BEAM CROSS SECTION INTERFACE | 783

 M t φ ,y
τ Mtx = ---------------
J
M t φ ,x
τ Mty = – --------------
J

where Mt is the twisting moment. The resultant is computed as:

2 2
τ Mt = τ Mtx + τ Mty

EFFECTIVE STRESS
The von Mises effective stress is computed from the stress components defined above
using the expression:

2 2 2
σ vM = ( σ N + σ M1 + σ M2 ) + 3 ( τ T1x + τ T2x + τ Mtx ) + 3 ( τ T1y + τ T2y + τ Mty )

784 | CHAPTER 8: BEAM CROSS SECTION

The Beam Cross Section Interface
The Beam Cross Section interface ( ), found under the Structural Mechanics
branch ( ) when adding a physics interface, is used for computing cross-section
properties for beams and also for a detailed evaluation of stresses in a beam when the
section forces to which it is subjected are known. It is mainly used in conjunction with
The Beam Interface.

As input you give a 2D sketch of the cross section, and computed results are area,
moments of inertia, center of gravity, principal axis directions, torsional rigidity, shear
center location, shear correction factors, warping constant, and stress distributions for
different load types.

• Using the Beam Cross Section Interface

• Theory for the Beam Cross Section Interface
• Studies and Solvers in the COMSOL Multiphysics Reference Manual

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is bcs.

DOMAIN SELECTION
The default setting is to include All domains in the model to define the cross section.
To select specific domains, select Manual from the Selection list.

MATERIAL PROPERTIES
Enter a value for Poisson’s ratio v (dimensionless). The default is 0.3. This value only
influences the detailed distribution of shear stresses caused by a transversal load.

THE BEAM CROSS SECTION INTERFACE | 785

 SECTION FORCES
If you want to perform a detailed computation of the stress state in a beam cross
section, enter the section forces.

After changing data in this section, you do not need to compute the study
for this physics interface again once is has been solved. It is sufficient to
do an Update Solution to get the stress plots updated.

Enter values or expressions for the following:

• Axial force N (SI unit: N).

• Bending moment around 1-axis M1.
• Shear force along 2-axis T2.
• Bending moment around 2-axis M2.
• Shear force along 1-axis T1.
• Twisting moment Mt.

DISCRETIZATION
To display this section, select click the Show button ( ) and select Discretization.
Select an Element order—Linear, Quadratic (the default), Cubic, Quartic, or Quintic.

Hole
Use the Hole feature to define internal holes for the beam cross section. One Hole
feature must be added for each internal hole.

BOUNDARY SELECTION
You should select all boundaries around a certain hole from the Selection list (Manual).
The All boundaries option is not relevant.

• Using the Beam Cross Section Interface

• Theory for the Beam Cross Section Interface

LOCATION IN USER INTERFACE

Context Menus
Beam Cross Section>Hole

786 | CHAPTER 8: BEAM CROSS SECTION

Ribbon
Physics tab with Beam Cross Section interface selected:
Boundaries>Beam Cross Section>Hole

Safety
Use the Safety node to set up variables which can be used to check the risk of failure
according to various criteria. Four different variables describing the failure risk will be
defined, as described in Table 8-3. You can add any number of Safety nodes.
TABLE 8-3: VARIABLES FOR SAFETY FACTOR EVALUATION

VARIABLE DESCRIPTION CRITERION CRITERION

FULFILLED VIOLATED

Failure index, FI For a linear criterion, this is the FI<1 FI>1

ratio between the computed value
and the given limit.
Damage index, DI A binary value, indicating whether DI=0 DI=1
failure is predicted or not. DI is
based on the value of FI.
Safety factor, SF For a linear criterion, this is 1/FI. SF>1 SF<1
Margin of safety, MoS SF-1 MoS>0 MoS<0

FAILURE MODEL
Select a Failure Criterion — von Mises Isotropic, Tresca Isotropic, or Rankine Isotropic.

• When Failure Criterion is von Mises Isotropic, enter Tensile strength σts.
• When Failure Criterion is Tresca Isotropic, enter Tensile strength σts.
• When Failure Criterion is Rankine Isotropic, enter Tensile strength σts and Compressive
strength σcs.
• When Failure Criterion is User defined, you enter two expressions describing the
Failure criterion g(S), used in the failure index, and the Safety factor sf(S)
respectively. As an example, if you would like to replicate the von Mises Isotropic
criterion with tensile strength 350 MPa, you could enter g(S) as solid.mises/
350[MPa]-1 and sf(S) as 350[MPa]/(solid.mises+eps).

THE BEAM CROSS SECTION INTERFACE | 787

 For all input fields, the default is to take the value From material. Change to User
defined to enter other values or expressions.

For a detailed description of the various criteria, see Safety Factor

Evaluation in the Structural Mechanics Theory chapter.

LOCATION IN USER INTERFACE

Context Menus
Beam Cross Section>Safety

Ribbon
Physics tab with Beam Cross Section selected:
Domains>Safety

788 | CHAPTER 8: BEAM CROSS SECTION

 9

Truss

This chapter describes the Truss interface, which is found under the Structural
Mechanics branch ( ) when adding a physics interface.

In this chapter:

• Modeling with Truss Elements

• Theory for the Truss Interface
• The Truss Interface

789
 Modeling with Truss Elements
Truss elements are elements that can only sustain axial forces. They have displacements
as degrees of freedom. Truss elements are sometimes referred to as bars or spars. They
live on boundaries in 2D and edges in 3D. Typical uses of truss elements are:

• Trusses
• Cables and wires
• Reinforcement bars
• Two-point springs and dashpots. A special material model called Spring-Damper is
available for this purpose.
• Thin ‘strain gauges’ attached to for example solid elements

The Truss interface supports the same study types as the Solid Mechanics interface.

Dependent Variables
The degrees of freedom (dependent variables) are the global displacements u, v, and
w (3D only) in the global x, y, and z directions, respectively.

MODELING WIRES AND CABLES

You can use the Truss interface for modeling wires and cables, possibly sagging under
gravity or other external loads. Below are some suggestions for how to model such
structures efficiently:

• It is generally recommended that you use first order shape functions when modeling
wire-like structures. This is the default in the Truss interface.
• If the cable is to be allowed to sag, you must not use the Straight Edge Constraint
for those edges. When using first order shape functions, this constraint is disabled
as default, so this is an issue only if you use higher order shape functions.
• Most cable problems are geometrically nonlinear. A wire which is not in tension is
not numerically stable. Physically, it wrinkles in an unpredictable manner. In order
to start the analysis, you either have to add an initial stress or some weak springs.
• If there are no line or volume loads, the wire is straight, as long as it is in tension.
In this case, only one element is needed for the whole wire, since you can handle the
zero-stiffness in compression in the material data. Enter the modulus of elasticity so

790 | C H A P T E R 9 : TR U S S
 that it depends on the axial stress, and has very small value in compression. Such an
expression could for example be E0*(1-0.9999*(truss.en<0)).
• In some problems, there are large deformations, but low tensile stresses. This would
for example be the case if you model a wire hanging free under self weight (‘the
catenary problem’). Such problems are numerically ill-conditioned, but can still be
solved as long as you use linear shape functions for the truss elements, and use tight
tolerances for the nonlinear solver.

M O D E L I N G W I T H TR U S S E L E M E N T S | 791
 Theory for the Truss Interface
Theory Background for the Truss Interface
Trusses are modeled using Lagrange shape function. The Lagrange shape function
makes it possible to specify both normal strains and Green-Lagrange strains to handle
small strains as well as large deformations.

STRAIN-DISPLACEMENT RELATION
The axial strain εn is calculated by expressing the global strains in tangential derivatives
and projecting the global strains on the edge.

t
ε n = t ε gT t

where t is the edge tangent vector and εgT is defined as

ε xT ε xyT ε xzT
ε gT = ε xyT ε yT ε yzT
ε xzT ε yzT ε zT

The strains can be expressed as either engineering strains for small displacements or
Green-Lagrange strains for large displacements. The Green-Lagrange strain tensor
used for large displacements is defined as

1  ∂u i ∂u j ∂u k ∂u k 
ε ijT = ---  + + ⋅ 
2  ∂ xj ∂ xi ∂ xi ∂ xj 
T T T T

The engineering strain tensor used for small displacements is defined as

1  ∂u i ∂u j 
ε ijT = ---  +  (9-1)
2  ∂ xj ∂ xi 
T T

The axial strain written out becomes

792 | C H A P T E R 9 : TR U S S
 ε n = t x ( ε xT t x + ε xyT t y + ε xzT t z ) +
t y ( ε xyT t x + ε yT t y + ε yzT t z ) +
t z ( ε xzT t x + ε yzT t y + ε zT t z )

In the Truss interface, the coordinates are usually denoted with lower case
letters (x, y, z). If a Solid Mechanics or Membrane interface is present in
the same model, then it becomes necessary to make a difference between
the material frame and the spatial frame (Material and Spatial
Coordinates). In this case the coordinates in the Truss interface change
to (X, Y, Z).

STRESS-STRAIN RELATION
The constitutive relation for a truss is uniaxial. The axial stress, σn, is computed as

σ n = σ ex + Eε n, el = σ ex + E ( ε n – ε inel )

ε inel = ε 0 + ε th + ε hs + ε pl

σ ex = σ 0 + σ ext

where

• E is the modulus of elasticity

• εn is the total axial strain
• εn,el is the elastic axial strain
• εinel is the sum of all inelastic strain contributions:
- Initial strain, ε0
- Thermal strain, εth
- Hygroscopic strain, εhs
- Plastic strain, εpl
• σex is the sum of all extra strain contributions:
- Initial stress σ0
- External stress σext

In a geometrically nonlinear analysis, this equation should be interpreted as a relation

between Second Piola-Kirchhoff stresses and Green-Lagrange strains.

T H E O R Y F O R T H E TR U S S I N T E R F A C E | 793
 For output, the First Piola-Kirchhoff stress Pn is computed from the Second
Piola-Kirchhoff stress using

P n = S n ⋅ s'

where s’ is the ratio between current and initial length. The axial force in the element
is then computed as

N = Pn ⋅ A0

where A0 is the undeformed cross-section area. The engineering (Cauchy) stress is

defined by

A0
σ n = P n -------
A

where A is the deformed area of the element. For a geometrically linear analysis, the
change in area is ignored, so that A = A0.

For a geometrically nonlinear analysis, the area change is computed based on an

assumption about a linear elastic material with Poisson’s ratio ν. The area change is

A0
------- = ( 1 – 2νε n ) –1
A

This is the only occasion where the Truss interface makes use of the value of ν.

In a geometrically linear analysis all the stress representations have the same value.

IMPLEMENTATION
Using the principle of virtual work results in the following weak formulation

  u FPi
t t
δW = d ( – ε n σ n + u F V ) dV +
V i

where the summation stands for summation over all points in the geometry. Replacing
the integration over the cross section with the cross-sectional area (A) and the volume
forces with line forces, the equation becomes

 ( – εntest σn A + utest FL ) dL +  utest FPi

t t
δW =
L i

794 | C H A P T E R 9 : TR U S S
In the case of geometric nonlinearity, the stress and strain should be interpreted as
Second Piola-Kirchhoff stress and Green-Lagrange strains

THEORY FOR STRAIGHT EDGE CONSTRAINT

The optional constraint to enforce the nodes to lie on the straight line between the end
points of the edge are formulated as follows:

Starting with the large displacement case, let xd1 and xd2 be the deformed position of
the two end points of the edge

x di = u i + x i (9-2)

where ui is the displacement, and xi is the coordinate (undeformed position) at end

point i. The equation for the straight line through the end points is

x + u = x d1 + ta (9-3)

where t is a parameter along the line, and a is the direction vector for the line. a is
calculated from the deformed position of the end points as

a = x d2 – x d1

The constraints for the edge is derived by substituting the parameter t from one of the
scalar equations in Equation 9-3 into the remaining ones. In 2D the constraint
equations become

( x + u – x d1 )a y – ( y + v – y d1 )a x = 0

In 3D the two constraints equations become

( x + u – x d1 )a z – ( z + w – z d1 )a x = 0
( y + v – y d1 )a z – ( z + w – z d1 )a y = 0

To avoid problems when the edge is directed in one of the coordinate axes directions,
a third constraint is added. This constraint is a linear combination of the two earlier
constraints:

( y + v – y d1 )a x – ( x + u – x d1 )a y = 0

This constraint is nonlinear, since a depends on the displacement.

A linear constraint is needed in the case of a geometrically linear problem to become

independent of the solver. The linear relation for the displacement is

T H E O R Y F O R T H E TR U S S I N T E R F A C E | 795
 u 1 ( x n2 – x n ) + u 2 ( x n – x n1 )
u = ------------------------------------------------------------------------ + u ax ( x 2 – x 1 ) (9-4)
( x n2 – x n1 )

where uax is the axial displacement along the edge, and xn is a linear parameter along
the edge

x ( x2 – x1 ) + y ( y2 – y1 ) + z ( z2 – z1 )
x n = --------------------------------------------------------------------------------------------
2 2 2
( x2 – x1 ) + ( y2 – y1 ) + ( z2 – z1 )

Eliminating uax from Equation 9-4 results in the following linear constraint in 2D

u 1 ( x n2 – x n ) + u 2 ( x n – x n1 )
------------------------------------------------------------------------ – u ( y 2 – y 1 ) –
( x n2 – x n1 )
v 1 ( x n2 – x n ) + v 2 ( x n – x n1 )
---------------------------------------------------------------------- – v ( x 2 – x 1 ) = 0
( x n2 – x n1 )

and the following three linear constraints in 3D:

u 1 ( x n2 – x n ) + u 2 ( x n – x n1 )
------------------------------------------------------------------------ – u ( z 2 – z 1 ) –
( x n2 – x n1 )
w 1 ( x n2 – x n ) + w 2 ( p – x n1 )
----------------------------------------------------------------------- – w ( x 2 – x 1 ) = 0
( x n2 – x n1 )
v 1 ( x n2 – x n ) + v 2 ( x n – x n1 )
---------------------------------------------------------------------- – v ( z 2 – z 1 ) –
( x n2 – x n1 )
(9-5)
w 1 ( x n2 – x n ) + w 2 ( x n – x n1 )
-------------------------------------------------------------------------- – w ( y 2 – y 1 ) = 0
( x n2 – x n1 )
v 1 ( x n2 – x n ) + v 2 ( x n – x n1 )
---------------------------------------------------------------------- – v ( x 2 – x 1 ) –
( x n2 – x n1 )
u 1 ( x n2 – x n ) + u 2 ( x n – x n1 )
- – u ( y2 – y1 ) = 0
-----------------------------------------------------------------------
( x n2 – x n1 )

THEORY FOR SPRING-DAMPER

Introduction
You can use a Spring-Damper to connect two points by an elastic spring, a viscous
damper, or both. Such springs can be used in any structural mechanics physics
interface, by adding a Truss interface. You can then set the degree of freedom names
in the two interface to the same name, in order to share the same displacement fields.

796 | C H A P T E R 9 : TR U S S
The two points can move relative to each other in an arbitrary way as long as they do
not coincide. The spring and damping forces act along the line between the current
positions of the two connection points.

Figure 9-1: Conceptual sketch of a Spring-Damper.

Connection Points and Spring Length

The current position of the two end points, x1 and x2 can be written as

x1 = X1 + u 1
x2 = X2 + u 2

where X1 and X1 are the original positions of the two points, and u1 and u2 are their
respective displacements. The initial spring length, l0, is

l0 = X2 – X1 = ( X2 – X1 ) ⋅ ( X2 – X1 )

The current spring length, l, is

l = x2 – x1

In the case of a geometrically linear analysis, the current spring length is linearized to

( x2 – x1 ) ⋅ ( X2 – X1 )
l = -----------------------------------------------------
l0

In addition to the initial geometrical distance between the two points you can specify
an initial spring extension Δl0, so that the free length of the spring is

l f = l 0 – Δl 0

You can also specify the free length of the spring explicitly.

T H E O R Y F O R T H E TR U S S I N T E R F A C E | 797
 The spring extension Δl is computed as the difference between the current spring
length and the free length,

Δl = l – l f

Deactivation
You can specify that the Spring-Damper should be deactivated under certain
conditions. It can for example be active only in tension, or break at a certain
elongation. In terms of implementation, this means that many expressions are
multiplied by an activation indicator, iac. The activation indicator has the value 1
when the component is active, and 0 when deactivated.

Spring and Damping Forces

The spring force is proportional to the spring constant k:

F s = kΔl

If k depends on the extension, so that the spring is nonlinear, it should be interpreted

as a secant stiffness, that is

F s = k ( Δl ) ⋅ Δl

You can also specify the spring force as function of extension explicitly, as

F s = F s ( Δl )

To create the expression for the function, use the built-in variable for the spring
extension. It has the form <physicsTag>.<SpringNodeTag>.dl, for example
truss.spd1.dl.

In a dynamic analysis, the viscous damping force is computed as

F d = c d ( Δl )
dt

where c is the viscous damping coefficient. In frequency domain, it is also possible to

specify a loss factor η, and the total damping force will then be

F d = iωcΔl + iηF s

The magnitude of the total force is

F = Fs + Fd

798 | C H A P T E R 9 : TR U S S
The total forces in the global coordinate system, acting on the connection points are

( x2 – x1 )
F 2 = – F 1 = – F -----------------------
l

In a geometrically linear case, the orientation of the force is kept fixed, so that

( X2 – X1 )
F 2 = – F 1 = – F -------------------------
l

The contribution to the virtual work is

F 2 ⋅ ( δu 2 – δu 1 )

Spring and Damping Energies

In stationary and time dependent analysis, the elastic energy in the spring is computed
as

Δl


W s = i ac F s d( Δl )
0

In a time dependent analysis, the energy dissipated in the damper, Wd, is computed
using an extra degree of freedom. The following equation is added:

dW ( Δl )-
-----------d- = i ac F d
-------------
dt d dt

In a frequency domain analysis, the elastic energy in the spring and the energy
dissipated in the damper are computed as

1
W s = --- i ac Re ( lindev ( F s ) ⋅ conj ( lindev ( Δl ) ) )
4
1
W d = --- i ac Re ( lindev ( F d ) ⋅ conj ( lindev ( Δl ) ) )
4

These energy quantities represent the cycle average, and only the perturbation terms
are included.

T H E O R Y F O R T H E TR U S S I N T E R F A C E | 799
 The Truss Interface
The Truss interface ( ), found under the Structural Mechanics branch ( ) when
adding a physics interface, is used for modeling slender elements that can only sustain
axial forces. It can be used for analyzing truss works where the edges are straight, or
to model sagging cables like the deformation of a wire exposed to gravity. It is available
in 3D and 2D. Geometric nonlinearity can be taken into account.

The Truss interface can also be used for modeling springs and dashpots.

The default material model is Linear Elastic Material. With the Nonlinear Structural
Materials Module, you can also model Plasticity.

For a detailed overview of the functionality available in each product, visit

http://www.comsol.com/products/specifications/

When this physics interface is added, these default nodes are also added to the Model
Builder: Linear Elastic Material, Cross Section Data, Free (a condition where points are
free, with no loads or constraints), Straight Edge Constraint (to ensure that the points
lie on a straight line between the end points of the edge or boundary), and Initial
Values. Then, from the Physics toolbar, you can add other nodes that implement, for
example, loads and constraints. You can also right-click Truss to select physics features
from the context menu.

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is truss.

REFERENCE PO INT F OR MO MENT COMPUTATION

Enter the default coordinates for the Reference point for moment computation xref. The
resulting moments (applied or as reactions) are then computed relative to this

800 | C H A P T E R 9 : TR U S S
reference point. During the results and analysis stage, the coordinates can be changed
in the Parameters section in the result nodes.

DEPENDENT VA RIA BLES

The dependent variable (field variable) is for the Displacement field u which has two
components (u, v) in 2D and three components (u, v, and w) in 3D. The name can be
changed but the names of fields and dependent variables must be unique within a
model.

DISCRETIZATION
To display this section, click the Show button ( ) and select Discretization. The
default is to use linear shape functions. If the truss elements share an edge with another
structural mechanics interface, you may want to use quadratic shape functions instead,
in order to get displacement compatibility.

• Boundary, Edge, Point, and Pair Nodes for the Truss Interface
• Edge Load
• Theory for the Truss Interface

• Vibrating String: Application Library path

Structural_Mechanics_Module/Verification_Examples/vibrating_string
• In-Plane and Space Truss: Application Library path
Structural_Mechanics_Module/Verification_Examples/
inplane_and_space_truss

Boundary, Edge, Point, and Pair Nodes for the Truss Interface
The Truss Interface has these boundary, edge, point, and pair nodes are available from
the Physics ribbon toolbar (Windows users), Physics context menu (Mac or Linux
users), or right-click to access the context menu (all users).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

T H E TR U S S I N T E R F A C E | 801
 FEATURES AVAILABLE FROM SUBMENUS
Many features for the Truss interface are added from submenus in the Physics toolbar
groups or context menu (when you right-click the node). The submenu name is the
same in both cases.The submenus at the Edge level (3D) or Boundary level (2D) are

• Material Models
• Line and Volume Loads
• Mass, Spring, and Damper
• Line Constraints

The submenus at the Point level are

• Mass, Spring, and Damper

• More Constraints
• Pairs

LINKS TO FEATURE NODE INFORMATION

These nodes (and subnodes) are described in this section (listed in alphabetical order):

• Antisymmetry • Point Mass

• Cross Section Data • Point Mass Damping
• Edge Load • Spring-Damper
• External Stress • Straight Edge Constraint
• Hygroscopic Swelling • Symmetry
• Initial Stress and Strain • Thermal Expansion (for constraints)
• Linear Elastic Material • Thermal Expansion (for materials)
• Phase
• Pinned

802 | C H A P T E R 9 : TR U S S
These nodes are described for the Solid Mechanics or Beam interface:

• Added Mass • Predeformation

• Damping • Prescribed Acceleration
• Free • Prescribed Displacement
• Gravity1 • Prescribed Velocity
• Initial Values • Rotating Frame1
• Plasticity • Spring Foundation
• Point Load
1
This is selected from the Line and Volume Loads submenu for this interface.

If there are subsequent constraints on the same geometrical entity, the last
one takes precedence.

In the COMSOL Multiphysics Reference Manual see Table 2-3 for links
to common sections and Table 2-4 to common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

Cross Section Data

Use the Cross Section Data node to enter the cross section area for the truss elements
on the selected edges.

This is required input data.

BASIC SECTION PROPERTIES

Enter an Area A.

LOCATION IN USER INTERFACE

Context Menus
Truss>Cross Section Data

T H E TR U S S I N T E R F A C E | 803
 Ribbon
Physics tab with Truss selected:
Edges>Truss>Cross Section Data

Straight Edge Constraint

The Straight Edge Constraint controls the addition of an additional constraint, forcing
the edge to be straight. The default is to add this constraint to all edges unless linear
shape functions have been chosen in the Discretization section of the interface. Using
this additional constraint removes the need to use a mesh with only one element per
edge. Internal nodes will make the model singular because the truss element only has
stiffness in the axial direction in a geometrically linear problem. The same problem is
present also when using higher-order shape functions, even if there is only one element
along an edge since there are internal nodes.

In the case of geometric nonlinearity, there is a stiffness in the transverse directions as

long as the axial force is tensile.

The additional constraints increase the solution time, especially for large 3D and
transient problems. The default mesh, when using Physics-controlled mesh in the Mesh
node, is to use one element per edge only, so that the extra constraints are not needed
unless the shape functions are of higher order.

ACTIVATION CONDITION
Select the Disable for linear discretization order check box to suppress the addition of
the straight edge constraints when linear shape functions are used. This is the default
state.

See also Theory for Straight Edge Constraint and Modeling Wires and
Cables.

LOCATION IN USER INTERFACE

Context Menus
Truss>Line Constraints>Straight Edge Constraint

Ribbon
Physics tab with Truss selected:
Edges>Line Constraints>Straight Edge Constraint

804 | C H A P T E R 9 : TR U S S
Linear Elastic Material
The Linear Elastic Material node adds the equations for a linear elastic truss element,
and an interface for defining the elastic material properties.

By adding the following subnodes to the Linear Elastic Material node you can
incorporate many other effects:

• Thermal Expansion (for materials)

• Hygroscopic Swelling
• Initial Stress and Strain
• External Stress
• Damping
• Plasticity

LINEAR ELASTIC MATERIAL

Define the linear elastic material properties. These settings are the same as described
under Linear Elastic Material for the beam interface.

GEOMETRIC NONLINEARITY
In this section there is always one check box. Either Force linear strains or Include
geometric nonlinearity is shown.

If a study step is geometrically nonlinear, the default behavior is to use a large strain
formulation for all boundaries or edges. There are however some cases when you
would still want to use a small strain formulation in a part of the structure. In those
cases, select the Force linear strains check box. When selected, a small strain
formulation is always used, independently of the setting in the study step. The default
value is that the check box is cleared, except when opening a model created in a version
prior to 4.3. In this case the state is chosen so that the properties of the model are
conserved.

The Include geometric nonlinearity check box is displayed only if the model was created
in a version prior to 4.3, and geometric nonlinearity was originally used for the selected
boundaries or edges. It is then selected and forces the Include geometric nonlinearity
check box in the study step to be selected. If the check box is cleared, the check box is
permanently removed and the study step assumes control over the selection of
geometric nonlinearity.

T H E TR U S S I N T E R F A C E | 805
 ENERGY DISSIPATION
To display this section, click the Show button ( ) and select Advanced Physics Options.

Select the Calculate dissipated energy check box as needed to compute the energy
dissipated by Plasticity.

LOCATION IN USER INTERFACE

Context Menus
Truss>Material Models>Linear Elastic Material

Ribbon
Physics tab with Truss selected:
Edges>Material Models>Linear Elastic Material

Thermal Expansion (for materials)

Use the Thermal Expansion subnode to add an internal thermal strain caused by changes
in temperature.

MODEL INPUTS
From the Temperature T list, select an existing temperature variable from another
physics interface, if any such temperature variables exist, or select User defined to enter
a value or expression for the temperature.

THERMAL EXPANSION PROPERTIES

Select an Input type to select how the thermal strain is specified. The default is Secant
coefficient of thermal expansion, in which case the thermal strain is given by

ε th = α ( T – T ref )

where α is the secant coefficient of thermal expansion. α can be temperature

dependent.

When Input type is Tangent coefficient of thermal expansion, the thermal strain is given
by

 T 
ε th = exp 
 T ref
α t ( τ ) dτ – 1


where αt is the tangential coefficient of thermal expansion.

806 | C H A P T E R 9 : TR U S S
When Input type is Thermal strain, enter the thermal strain dL as function of
temperature explicitly.

In all three cases, the default is to take values From material. Use User defined to enter
an expression for the coefficient of thermal expansion or thermal strain.

Enter a value or expression for the Strain reference temperature Tref which is the
reference temperature that defines the change in temperature together with the actual
temperature.

LOCATION IN USER INTERFACE

Context Menus
Truss>Linear Elastic Material>Thermal Expansion

Ribbon
Physics tab with Linear Elastic Material node selected in the model tree:
Attributes>Thermal Expansion

Hygroscopic Swelling
Hygroscopic swelling is an internal strain caused by changes in moisture content. This
strain can be written as

ε hs = β h ( c mo – c mo,ref )

where βh is the coefficient of hygroscopic swelling, cmo is the moisture concentration,

and cmo,ref is the strain-free reference concentration. The temperature is assumed to
be constant over the cross section of the truss element.

The settings for the Truss interface are the same as described for the Beam
interface (excluding the hygroscopic bending options). See Hygroscopic
Swelling in the documentation for the Beam interface.

LOCATION IN USER INTERFACE

Context Menus
Truss>Linear Elastic Material>Hygroscopic Swelling

T H E TR U S S I N T E R F A C E | 807
 Ribbon
Physics tab with Linear Elastic Material node selected in the model tree:
Attributes>Hygroscopic Swelling

Initial Stress and Strain

You can add the Initial Stress and Strain subnode to the Linear Elastic Material, in order
to specify the stress or strain state in the structure before applying any constraint or
load. The values given are not initial values in the mathematical sense, but rather a
contribution to the constitutive relation

In many cases Initial Stress and Strain and External Stress are
interchangeable when prescribing stresses, but you can find some more
options in the latter.

INITIAL STRESS AND STRAIN

Enter an Initial axial strain eni and Initial axial stress σni.

For details about initial stresses and strains, see Inelastic Strain
Contributions and Initial Stresses and Strains.

LOCATION IN USER INTERFACE

Context Menus
Truss>Linear Elastic Material>Initial Stress and Strain

Ribbon
Physics tab with Linear Elastic Material node selected in the model tree:
Attributes>Initial Stress and Strain

External Stress
You can add the External Stress subnode to the Linear Elastic Material, in order to
specify an additional stress contribution which is not part of the constitutive relation.

808 | C H A P T E R 9 : TR U S S
The external stress can be added to the total stress tensor, or act only as an extra load
contribution.

In many cases External Stress and Initial Stress and Strain are
interchangeable when prescribing stresses. In Initial Stress and Strain, the
given stress is however always added to the stress tensor.

EXTERNAL STRESS
Select a Stress input—Stress tensor or Axial stress.

• When Stress tensor is selected, you enter the external stress in the form of Second
Piola-Kirchhoff stress tensors. In the External stress tensor drop-down list, stress
tensors announced by any physics interface will be shown, and also the entry User
defined. When User defined is selected, you can enter the data for the External stress
tensor Sext as Isotropic, Diagonal, or Symmetric depending on the properties of the
tensor. The tensor components are interpreted in the global coordinate system, and
are projected onto the tangential direction of the truss element. If a stress tensor
announced by a physics interface is selected, the coordinate system setting is ignored
— the orientation is handled internally. Choose a Contribution type—Add to stress
tensor or Load contribution only to determine the effect of the contribution.
• When Axial stress is selected, you enter a value or an expression for the Axial stress
Sn,ext. Choose a Contribution type—Add to stress tensor or Load contribution only to
determine the effect of the contribution.

Selecting a stress tensor announced by the same physics interface as where

the External Stress node is added, will result in an error (‘Circular
variable dependency detected’). This operation would imply that
the computed stress depends on itself.

For theory, see External Stress.

LOCATION IN USER INTERFACE

Context Menus
Truss>Linear Elastic Material>External Stress

T H E TR U S S I N T E R F A C E | 809
 Ribbon
Physics tab with Linear Elastic Material node selected in the model tree:
Attributes>External Stress

Spring-Damper
Use a Spring-Damper to model — between two points—an elastic spring, a viscous
damper, or both. By adding a Truss interface with this material model, you can add
springs and dashpots to any other structural mechanics physics interface.

SPRING-DAMPER
Select a Spring type—Spring constant or Force as function of extension.

• For Spring constant enter a value for the spring constant k.

• For Force as function of extension enter an expression for the spring force Fs. The
expression must be a function of the extension of the spring. The built-in variable
for the spring extension has the form <physicsName>.<SpringNodeTag>.dl, for
example truss.spd1.dl. The default expression is (1[N/m])*truss.spd1.dl,
which corresponds to a linear spring with the stiffness 1 N/m.

To add viscous damping in a dynamic analysis, enter a value or expression for the
Damping coefficient c.

To add loss factor damping, enter a value for the Loss factor damping η.

FREE LENGTH
The free length is the distance between the connection points when there is no force
in the spring. Select an option from the list—Specify initial extension or Specify free
length.

• For Specify initial extension enter a value for Δl 0 . The free length is computed as
l f = l 0 – Δl 0 , where l0 is the initial distance between the connection points.
• For Specify free length enter a value for lf.

ACTIVATION CONDITION
Select a Spring action—Bidirectional, Tension only, or Compression only.

Select the Deactivation condition check box to enter a Deactivation indicator expression
idac. The expression is treated as a boolean expression, so that when it is evaluated to
a nonzero value, the spring or damper is deactivated.

810 | C H A P T E R 9 : TR U S S
Select the Permanently deactivate check box if the spring is supposed to be removed
permanently from the simulation once the deactivation condition is fulfilled for the
first time.

As an example, if the spring should break at a certain extension, you can write an
expression like truss.spd1.dl>0.12[m], and select the Permanently deactivate check
box. If the check box is not selected, the spring would become active again when its
connection points come close enough to each other.

LOCATION IN USER INTERFACE

Context Menus
Truss>Material Models>Spring-Damper

Ribbon
Physics tab with Truss selected:
Edges>Material Models>Spring-Damper

Pinned
The Pinned node adds an edge (3D), boundary (2D), or point (2D and 3D) condition
that makes the edge, boundary, or point fixed; that is, the displacements are zero in all
directions.

CONSTRAINT SETTINGS
To display this section, click the Show button ( ) and select Advanced Physics Options.

In the COMSOL Multiphysics Reference Manual:

• Constraint Reaction Terms

• Weak Constraints
• Constraint Settings

You can activate and deactivate this boundary condition by assigning it to

a constraint group. See Load Cases in the Structural Mechanics Modeling
chapter.

T H E TR U S S I N T E R F A C E | 811
 LOCATION IN USER INTERFACE

Context Menus
Truss>Line Constraints>Pinned (Line)
Truss>Pinned (Point)

Ribbon
Physics tab with Truss selected:
Edges>Line Constraints>Pinned
Points>Truss>Pinned

Thermal Expansion (for constraints)

Add the Thermal Expansion subnode to a constraint (Pinned or Prescribed Displacement)
to prescribe a deformation of the constraint caused by changes in temperature of the
surroundings. This makes it possible to reduce stresses caused by the boundary
conditions.

The thermal strain depends on the coefficient of thermal expansion α, the temperature
T, and the strain-free reference temperature Tref as

ε th = α ( T – T ref )

THERMAL EXPANSION PROPERTIES

Specify the thermal properties that define the thermal strain. This is a description of
the thermal expansion of surroundings idealized by the constraints.

Select Inherit from edge to take the thermal expansion data from the domain being
constrained. This should only be used when:

• The temperature and the thermal expansion coefficient do not have a spatial
variation.
• The virtual surrounding material has the same thermal expansion as the edge itself.

When Inherit from domain is not selected, enter:

• A value or expression for Temperature T, specifying the temperature distribution of

the surrounding material. Any spatial variation must be an explicit function of the
material frame coordinates. It is not possible to use a computed temperature
distribution.
• The Coefficient of thermal expansion α. As a default, values From material are used.
This requires that a material has been assigned to the points where the constraint is

812 | C H A P T E R 9 : TR U S S
 active.
For User defined enter the coefficient of thermal expansion α.
• A value or expression for the Strain reference temperature Tref which is the
temperature at which there are no thermal displacements at the constraints.

Enter the coordinates of the Reference point, the point where the displacement is zero.
The choice of reference point only affects the rigid body motion. If there are several
different constraints with a Thermal Expansion subnode, the same reference point
should usually be selected in all of them.

• Constraints and Thermal Expansion in the Structural Mechanics

Modeling chapter.
• Thermal Expansion of Constraints in the Structural Mechanics Theory
chapter.

LOCATION IN USER INTERFACE

Context Menus
Truss>Pinned>Thermal Expansion
TrussPrescribed Displacement>Thermal Expansion

Ribbon
Physics tab with Pinned or Prescribed Displacement node selected in the model tree:
Attributes>Thermal Expansion

Symmetry
The Symmetry node adds an edge (3D), boundary (2D), or point (2D and 3D)
condition that defines a symmetry edge, boundary, or point.

COORDINATE SYSTEM SELECTION

The Global coordinate system is selected by default. The Coordinate system list contains
any additional coordinate systems that the model includes. The coordinate system is
used in conjunction with the Axis to use as normal direction setting.

SYMMETRY
Select an Axis to use as normal direction. This specifies the direction of the normal to
the symmetry plane. Select 1, 2, or 3 for the first, second, or third axis in the selected
coordinate system.

T H E TR U S S I N T E R F A C E | 813
 CONSTRAINT SETTINGS
To display this section, click the Show button ( ) and select Advanced Physics Options.

In the COMSOL Multiphysics Reference Manual:

• Constraint Reaction Terms

• Weak Constraints
• Constraint Settings

Symmetry Constraints

You can activate and deactivate this boundary condition by assigning it to

a constraint group. See Load Cases in the Structural Mechanics Modeling
chapter.

LOCATION IN USER INTERFACE

Context Menus
Truss>Line Constraints>Symmetry (Line)
Truss>More Constraints>Symmetry (Point)

Ribbon
Physics tab with Truss selected:
Edges>Line Constraints>Symmetry
Points>More Constraints>Symmetry

Antisymmetry
The Antisymmetry node adds an edge (3D), boundary (2D), or point (2D and 3D)
condition that defines an antisymmetry edge, boundary, or point.

COORDINATE SYSTEM SELECTION

The Global coordinate system is selected by default. The Coordinate system list contains
any additional coordinate systems that the model includes. The coordinate system is
used in conjunction with the Axis to use as normal direction setting.

814 | C H A P T E R 9 : TR U S S
ANTISYMMETRY
Select an Axis to use as normal direction. This specifies the direction of the normal to
the antisymmetry plane.

• For 2D models, select 1 or 2 for the first or second axis, respectively.

• For 3D models, select 1, 2, or 3 for the first, second, or third axis, respectively.

CONSTRAINT SETTINGS
To display this section, click the Show button ( ) and select Advanced Physics Options.

In the COMSOL Multiphysics Reference Manual:

• Constraint Reaction Terms

• Weak Constraints
• Constraint Settings

Symmetry Constraints

You can activate and deactivate this boundary condition by assigning it to

a constraint group. See Load Cases in the Structural Mechanics Modeling
chapter.

LOCATION IN USER INTERFACE

Context Menus
Truss>Line Constraints>Antisymmetry (Line)
Truss>More Constraints>Antisymmetry (Point)

Ribbon
Physics tab with Truss selected:
Edges>Line Constraints>Antisymmetry
Points>More Constraints>Antisymmetry

Edge Load
Add an Edge Load as a force distributed along an edge (3D models) or boundary (2D
models).

T H E TR U S S I N T E R F A C E | 815
 FORCE
Select a Load type—Force per unit length, Force per unit volume, or Total force. Enter
values or expressions for the components.

LOAD TYPE VARIABLE SI UNIT GEOMETRIC SPACE DIMENSION

ENTITY LEVEL (COMPONENTS)

Force per unit length FL N/m edges 3D (x, y, z)

boundaries 2D (x, y)
Force per unit volume FV N/m3 edges 3D (x, y, z)
boundaries 2D (x, y)
Total force Ftot N edges 3D (x, y, z)
boundaries 2D (x, y)

• When Force per unit volume is selected, the given load is multiplied by
the cross section area. This option is useful for modeling body loads
like gravity or centrifugal loads.
• After selecting a Load type, the Load list normally only contains User
defined. When combining the Truss interface with another physics
interface, it is also possible to choose a predefined load from this list.
• You can add the Phase subnode to specify the phase of this load in a
frequency domain analysis.
• You can specify this load to be a Harmonic Perturbation in a frequency
domain analysis.
• You can assign this load to a load group. See Load Cases in the
Structural Mechanics Modeling chapter.

LOCATION IN USER INTERFACE

Context Menus
Truss>Line and Volume Loads>Edge Load

Ribbon
Physics tab with Truss selected:
Edges>Line and Volume Loads>Edge Load

Point Mass
Use the Point Mass node to add a discrete mass that is concentrated at a point.

816 | C H A P T E R 9 : TR U S S
The Point Mass Damping subnode can be added to specify a mass-proportional
damping.

PO IN T MA SS
Enter a Point mass m.

LOCATION IN USER INTERFACE

Context Menus
Truss>Mass, Spring, and Damper>Point Mass

Ribbon
Physics tab with Truss selected:
Edges>Mass, Spring, and Damper>Point Mass

Point Mass Damping

Use the Point Mass Damping subnode to add damping to a Point Mass parent node.

PO IN T MASS DAMPIN G
Enter a Mass damping parameter αdM. This is the mass proportional term of a Rayleigh
damping.

LOCATION IN USER INTERFACE

Context Menus
Truss>Point Mass>Point Mass Damping

Ribbon
Physics tab with Point Mass node selected in the model tree:
Point Mass Damping

Phase
You can add a Phase subnode to nodes which define a load in order to prescribe the
phase angle in a frequency domain analysis.

For modeling the frequency response the physics interface splits the harmonic load
into two parameters:

• The amplitude, F, which is specified in the node for the load.

• The phase ( φ ), which is specified in the Phase subnode.

T H E TR U S S I N T E R F A C E | 817
 Together these define a harmonic load, for which the amplitude and phase shift can
vary with the excitation frequency, f:

F freq = F ( f ) ⋅ cos ( 2πft + φ )

PHASE
Add the phase load Fph for harmonic loads. Enter the phase for each component of
the load in the corresponding fields.

LOCATION IN USER INTERFACE

Context Menus
Truss>Edge Load>Phase
Truss>Point Load>Phase

Ribbon
Physics tab with Edge Load or Point Load node selected in the model tree:
Attributes>Phase

818 | C H A P T E R 9 : TR U S S
 10

Multiphysics Interfaces and Couplings

The Structural Mechanics Module contains predefined multiphysics interfaces to

facilitate easy set up of models with the most commonly occurring couplings. Four
of the physics interfaces are found under the Structural Mechanics branch ( )
when adding a physics interface — Thermal Stress, Joule Heating and Thermal
Expansion, Piezoelectric Devices, and Magnetostriction. The Fluid-Structure
Interaction interfaces are found under the Fluid Flow branch ( ).

• The Thermal Stress Interface

• The Fluid-Structure Interaction Interface
• The Fluid-Structure Interaction, Fixed Geometry Interface
• Theory for the Fluid-Structure Interaction Interface
• The Joule Heating and Thermal Expansion Interface
• The Piezoelectric Devices Interface
• The Magnetostriction Interface
• Hygroscopic Swelling Coupling
• Couplings Between Structural Mechanics Interfaces

819
 T he T he r m a l St ress In t erface
The Thermal Stress ( ) interface combines a Solid Mechanics interface with a Heat
Transfer in Solids interface. The coupling occurs on the domain level, where the
temperature from the Heat Transfer interface acts as a thermal load for the Solid
Mechanics interface, causing thermal expansion.

When a predefined Thermal Stress interface is added from the Structural Mechanics
branch ( ) of the Model Wizard or Add Physics windows, Solid Mechanics and Heat
Transfer in Solids interfaces are added to the Model Builder.

In addition, the Multiphysics node is added, which automatically includes the

multiphysics coupling features Thermal Expansion and Temperature Coupling.

On the Constituent Physics Interfaces

The Solid Mechanics interface is intended for general structural analysis of 3D, 2D, or
axisymmetric bodies. In 2D, plane stress or plane strain assumptions can be used. The
Solid Mechanics interface is based on solving Navier’s equations, and results such as
displacements, stresses, and strains are computed.

The Heat Transfer in Solids interface provides features for modeling heat transfer by
conduction, convection, and radiation. A Heat Transfer in Solids model is active by
default on all domains. All functionality for including other domain types, such as a
fluid domain, is also available. The temperature equation defined in solid domains
corresponds to the differential form of the Fourier’s law that may contain additional
contributions like heat sources.

In previous versions of COMSOL Multiphysics, a specific physics

interface called Thermal Stress was added to the Model Builder. Now, a
predefined multiphysics coupling approach is used, improving the
flexibility and design options for your modeling. For specific details, see
The Multiphysics Node and Multiphysics Modeling Approaches in the
COMSOL Multiphysics Reference Manual.

For information about the constitutive equations including thermal

expansion in the section dealing with the theory background, see
Structural Mechanics Theory.

820 | CHAPTER 10: MULTIPHYSICS INTERFACES AND COUPLINGS

SETTINGS FOR PHYSICS INTERFACES AND COUPLING FEATURES
When physics interfaces are added using the predefined couplings, for example Thermal
Stress, specific settings are included with the physics interfaces and the coupling
features.

However, if physics interfaces are added one at a time, followed by the coupling
features, these modified settings are not automatically included.

For example, if single Solid Mechanics and Heat Transfer in Solids interfaces are added,
an empty Multiphysics node appears in the model tree. You can choose from the
available coupling features but the settings in the constituent interfaces are nor
modified.

Coupling features are available from the context menu (right-click the
Multiphysics node) or from the Physics toolbar, Multiphysics menu.

TABLE 10-1: MODIFIED SETTINGS FOR A THERMAL STRESS INTERFACE

PHYSICS INTERFACE OR MODIFIED SETTINGS

COUPLING FEATURE

Solid Mechanics For the Solid Mechanics interface, under Structural

Transient Behavior the Structural transient behavior is
set to Quasi-static.
Heat Transfer in Solids For the Heat Transfer in Solids interface, under
Discretization, the shape function order for the
temperature is set to Linear.
Thermal Expansion The Domain Selection is the same as that of the
participating physics interfaces.
The corresponding Solid Mechanics and Heat Transfer in
Solids interfaces are preselected in the Thermal
Expansion section.
Temperature Coupling The corresponding Solid Mechanics and Heat Transfer in
Solids interfaces are preselected in the Temperature
Coupling section (described in the COMSOL
Multiphysics Reference Manual).

PHYSICS INTERFACES AND COUPLING FEATURES

Use the online help in COMSOL Multiphysics to locate and search all the
documentation. All these links also work directly in COMSOL
Multiphysics when using the Help system.

THE THERMAL STRESS INTERFACE | 821

 Coupling Features
• The Thermal Expansion (Multiphysics Coupling) coupling feature node is
described in this section.
• The Temperature Coupling feature node is described for The Joule Heating
Interface in the COMSOL Multiphysics Reference Manual.

Fuel Cell Bipolar Plate: Application Library path

Structural_Mechanics_Module/Thermal-Structure_Interaction/bipolar_plate

Thermal Expansion in a MEMS Device: Application Library path

MEMS_Module/Sensors/thermal_expansion

Thermal Expansion (Multiphysics Coupling)

The Thermal Expansion coupling is similar to the Thermal Expansion (for materials)
node ( ) that can be added under, for example, the Linear Elastic Material or
Hyperelastic Material for the Solid Mechanics interface. The purpose is the same, and if
both nodes are used for the same selection, the settings in the coupling node takes
precedence.

SETTINGS
The Label is the default multiphysics coupling feature name.

The Name is used primarily as a scope prefix for variables defined by the coupling node.
Refer to such variables in expressions using the pattern <name>.<variable_name>. In
order to distinguish between variables belonging to different coupling nodes or physics
interfaces, the name string must be unique. Only letters, numbers, and underscores (_)
are permitted in the Name field. The first character must be a letter.

The default Name (for the first multiphysics coupling feature in the model) is te1.

DOMAIN SELECTION
When nodes are added from the context menu, you can select Manual (the default)
from the Selection list to choose specific domains to define the coefficient of thermal
expansion and the different temperatures that cause thermal stress, or select All domains
as needed.

822 | CHAPTER 10: MULTIPHYSICS INTERFACES AND COUPLINGS

When Thermal Expansion is added as an effect of adding a Thermal Stress interface, the
selection is the same as for the participating physics interfaces.

Only domains that are active in the physics interfaces selected in the Coupled
Interfaces section can be selected.

THERMAL EXPANSION PROPERTIES

Select an Input type to select how the thermal strain is specified. The default is Secant
coefficient of thermal expansion, in which case the thermal strain is given by

ε th = α ( T – T ref )

where α is the secant coefficient of thermal expansion. α can be temperature

dependent.

When Input type is Tangent coefficient of thermal expansion, the thermal strain is given
by

 T 
ε th = exp 
 T ref
α t ( τ ) dτ – 1


where αt is the tangential coefficient of thermal expansion.

When Input type is Thermal strain, enter the thermal strain dL as function of
temperature explicitly.

In all three cases, the default is to take values From material. When entering data as
User defined, select Isotropic, Diagonal or Symmetric to enter one or more components
for a general coefficient of the thermal expansion tensor or the thermal strain tensor.
When a non-isotropic input is used, the axis orientations are given by the coordinate
system selection in the parent node.

Enter a value or expression for the Strain reference temperature Tref which is the
reference temperature that defines the change in temperature together with the actual
temperature.

Select Thermoelastic damping to include the reverse coupling where the changes in
stress act as a heat source in the heat transfer analysis. Thermoelastic damping is only
active when Structural Transient Behavior is set to Include inertial terms.

COUPLED INTERFACES
This section defines the physics involved in the multiphysics coupling. The Heat
transfer and Solid mechanics lists include all applicable physics interfaces.

THE THERMAL STRESS INTERFACE | 823

 The default values depend on how the coupling node is created.

• If it is added from the Physics ribbon (Windows users), Physics contextual toolbar
(Mac and Linux users), or context menu (all users), then the first physics interface
of each type in the component is selected as the default.
• If it is added automatically when a multiphysics interface is selected in the Model
Wizard or Add Physics window, then the two participating physics interfaces are
selected.

You can also select None from either list to uncouple the Thermal Expansion node from
a physics interface. If the physics interface is removed from the Model Builder, for
example Heat Transfer in Solids is deleted, then the Heat transfer list defaults to None as
there is nothing to couple to.

If a physics interface is deleted and then added to the model again, then
in order to re-establish the coupling, you need to choose the physics
interface again from the Heat transfer or Solid mechanics lists. This is
applicable to all multiphysics coupling nodes that would normally default
to the once present interface. See Multiphysics Modeling Approaches in
the COMSOL Multiphysics Reference Manual.

• Thermoelastic Damping
• Entropy and Thermoelasticity

824 | CHAPTER 10: MULTIPHYSICS INTERFACES AND COUPLINGS

T he Flui d- S tr uc t u re In t erac t i on
Interface
Using the Fluid-Structure Interaction (fsi) interface ( ), found under the Fluid Flow
branch ( ) when adding a physics interface, you can model phenomena where a fluid
and a deformable solid affect each other. The physics interface models both the fluid
domain and the solid domain (structure) and includes a predefined condition for the
interaction at the fluid-solid boundaries. An ALE formulation is used for incorporating
the geometrical changes of the fluid domain.

The fluid can be either compressible or incompressible. The flow regime can be
laminar or turbulent (if you have a license for the CFD Module). The solid domain has
the same options as in a Solid Mechanics interface, including contact conditions and
also nonlinear materials if the Nonlinear Structural Materials Module or Geomechanics
Module is available.

Using a stationary or a time-dependent study, the Fluid-Structure Interaction interface

models two-way coupling between solids and fluids. There are, however, also special
study steps available in order to model one-way coupled fluid-structure interaction.

The Fluid-Structure Interaction interface is available for 3D, planar 2D, and 2D
axisymmetric geometries.

In planar 2D, the physics interface uses the assumption that the structures deform in
the plane strain regime. This means that the interpretation of the results are values “per
meter thickness,” and there is no specific thickness to specify.

When the Fluid-Structure Interaction interface is added, the following default nodes
are added to the Model Builder—Fluid Properties, Linear Elastic Material, and Free
Deformation (for the mesh movement and default boundary conditions) in the
domains; Wall (for the fluid), Prescribed Mesh Displacement (for the mesh movement),
and Free (for the solid mechanics, which initially is not applicable to any boundary
because the default settings assume a fluid domain) as default boundary conditions;
and Initial Values.

In addition, for the fluid-solid boundary, a Fluid-Solid Interface Boundary node adds
the fluid-structure interaction. This node is only applicable to interior fluid-solid
boundaries.

THE FLUID-STRUCTURE INTERACTION INTERFACE | 825

 From the Physics toolbar, add other nodes that implement, for example, loads,
constraints, and other nonlinear materials for the solid domain. You can also right-click
Fluid-Structure Interaction to select physics features from the context menu.

The Settings window contains the following sections plus additional sections that are
similar to those for physics interface nodes Settings windows for fluid flow, solid
mechanics, and moving mesh interfaces.

The Fluid-Structure Interaction interface default is to treat all domains as fluid. The
Linear Elastic Material node, which is the default node for the solid domain, initially
has an empty selection. When a solid mechanics material is added to the solid domains,
the physics interface automatically identifies the fluid-solid interaction boundaries and
assigns the Fluid-Solid Interface Boundary condition to those boundaries. Two materials
are typically defined in an FSI model: one for the fluid and one for the solid.

For an overview of available variables for monitoring and plotting the

moving mesh, see Predefined Variables in Deformed Geometry and
Moving Mesh in the COMSOL Multiphysics Reference Manual. Note,
however, that the variables in the Fluid-Structure Interaction interface use
the Name fsi instead of ale.

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is fsi.

FREE DEFORMATION SETTINGS

Select a Mesh smoothing type — Winslow (the default), Hyperelastic, Laplace, or Yeoh.
For the Yeoh mesh smoothing type, also specify a Stiffening factor (default: 100). See
Smoothing Methods in the COMSOL Multiphysics Reference Manual for more
information.

826 | CHAPTER 10: MULTIPHYSICS INTERFACES AND COUPLINGS

PHYSICAL MODEL
By default the physics interface uses the Compressible flow (Ma<0.3) formulation of the
Navier-Stokes equations. Select Incompressible flow to use the incompressible (constant
density) formulation.

Enter a Reference pressure level pref (SI unit: Pa). The default value is 1[atm].

REFERENCE POINT FOR MOMENT COMPUTATION

Enter the coordinates for the Reference point for moment computation, xref
(SI unit: m). All moments are then computed relative to this reference point.

S T R U C T U R A L TR A N S I E N T B E H AV I O R
Select a Structural transient behavior—Include inertial terms (the default) or
Quasi-static.

DEPENDENT VA RIA BLES

The dependent variable (field variables) include the following. The turbulence
variables are only active if the fluid flow part uses a turbulence model. The name can
be changed but the names of fields and dependent variables must be unique within a
model.

• Pressure p (SI unit: Pa)

• Turbulent dissipation rate ep (SI unit: m2/s3)
• Turbulent kinetic energy k (SI unit: m2/s3)
• Reciprocal wall distance G (SI unit: 1/m)
• Displacement field usolid (SI unit: m)
• Velocity field ufluid (SI unit: m/s)

ADVANCED SETTINGS
To display this section, click the Show button ( ) and select Advanced Physics Options.
Normally these settings do not need to be changed.

The Use pseudo time stepping for stationary equation form check box is active per
default. It adds pseudo time derivatives to the momentum equations when the
Stationary equation form is used. When selected, also choose a CFL number expression—
Automatic (the default) or Manual. Automatic sets the local CFL number (from the
Courant–Friedrichs–Lewy condition) to the built-in variable CFLCMP which in turns

THE FLUID-STRUCTURE INTERACTION INTERFACE | 827

 trigger a PID regulator for the CFL number. For Manual enter a Local CFL number
CFLloc.

• Pseudo Time Stepping for Laminar Flow Models and Pseudo Time
Stepping in the COMSOL Multiphysics Reference Manual
• Domain, Boundary, Edge, Point, and Pair Nodes for the
Fluid-Structure Interaction Interface
• Theory for the Fluid-Structure Interaction Interface
• Basic Modeling Steps for Fluid-Structure Interaction

Peristaltic Pump: Application Library path Structural_Mechanics_Module/

Fluid-Structure_Interaction/peristaltic_pump

Domain, Boundary, Edge, Point, and Pair Nodes for the

Fluid-Structure Interaction Interface
The Fluid-Structure Interaction Interface has these domain, boundary, edge, point,
and pair nodes, listed in alphabetical order, available from the Physics ribbon toolbar
(Windows users), Physics context menu (Mac or Linux users), or right-click to access
the context menu (all users).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

These nodes are unique to the physics interface and described in this section:

• Initial Values
• Wall
• Fluid-Solid Interface Boundary

• If you also have a CFD Module or Heat Transfer Module, the Interior Wall
boundary condition is also available and is documented in the CFD Module User’s
Guide or Heat Transfer Module User’s Guide, respectively. This boundary

828 | CHAPTER 10: MULTIPHYSICS INTERFACES AND COUPLINGS

 condition is useful for avoiding meshing of thin wall structures by using no-slip
conditions on interior curves and surfaces.
• With the addition of a module that enhances various fluid flow interfaces, the Line
Mass Source and Point Mass Source features are also available and described in the
COMSOL Multiphysics Reference Manual.

THE FLUID-STRUCTURE INTERACTION INTERFACE | 829

 These nodes are described for the Solid Mechanics interface:These nodes are described

• Added Mass • Point Load

• Antisymmetry • Predeformation
• Body Load • Prescribed Acceleration
• Boundary Load • Prescribed Displacement
• Contact • Prescribed Velocity
• Edge Load • Rigid Connector
• Fixed Constraint • Roller
• Free • Spring Foundation
• Linear Elastic Material • Symmetry
• Viscoelasticity • Thin Elastic Layer
• Periodic Condition

for the Laminar Flow or Moving Mesh interfaces in the COMSOL Multiphysics
Reference Manual (listed in alphabetical order):

• No Viscous Stress • Periodic Flow Condition

• Fixed Mesh • Prescribed Deformation
• Flow Continuity • Prescribed Mesh Displacement
• Fluid Properties • Pressure Point Constraint
• Free Deformation • Symmetry
• Inlet • Volume Force
• Open Boundary
• Outlet

In the COMSOL Multiphysics Reference Manual see Table 2-3 for links
to common sections and Table 2-4 to common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

830 | CHAPTER 10: MULTIPHYSICS INTERFACES AND COUPLINGS

Initial Values
The Initial Values node adds initial values for pressure, turbulent dissipation rate,
turbulent kinetic energy, reciprocal wall distance, displacement field, and velocity field.

These variables can serve as an initial condition for a transient simulation or as an initial
guess for a nonlinear analysis. If more than one set of initial values is required, add
additional Initial Values nodes from the Physics toolbar.

IN IT IA L VA LUES
Enter the initial values as values or expressions. The variables for turbulence are only
valid for fluid flow using a turbulence model.

• Pressure p (SI unit: Pa)

• Turbulent dissipation rate εp (ep) (SI unit: m2/s3)
• Turbulent kinetic energy k (SI unit: m2/s3)
• Reciprocal wall distance G (SI unit: 1/m)
• Displacement field usolid (SI unit: m)
• Velocity field ufluid (SI unit: m/s)

Wall
The Wall node includes a set of boundary conditions describing fluid-flow conditions
at stationary, moving, and leaking walls. For turbulent flow, the description may
involve wall functions and/or asymptotic expressions for certain turbulence variables.

BOUNDARY CONDITION
Select a Boundary condition for the wall.

• No Slip1 • Moving Wall

• Slip • Leaking Wall
• Sliding Wall
• Moving Wall

If you also have a CFD Module, Heat Transfer Module or Microfluidics Module
additional options become available as documented in the CFD Module User’s Guide,
Heat Transfer Module User’s Guide or Microfluidics Module User’s Guide
respectively.

THE FLUID-STRUCTURE INTERACTION INTERFACE | 831

 No Slip
No slip is the default boundary condition for a stationary solid wall for laminar flow
(and SST, Low Re k-ε, Algebraic yPlus, L-VEL, and Spalart-Allmaras turbulence
models). The condition prescribes u = 0; that is, the fluid at the wall is not moving.

Slip
The Slip option prescribes a no-penetration condition, u·n=0. It is implicitly assumed
that there are no viscous effects at the slip wall and hence, no boundary layer develops.
From a modeling point of view, this can be a reasonable approximation if the main
effect of the wall is to prevent fluid from leaving the domain.

Sliding Wall
The Sliding wall boundary condition is appropriate if the wall behaves like a conveyor
belt; that is, the surface is sliding in its tangential direction. A velocity is prescribed at
the wall and the boundary itself does not have to actually move relative to the reference
frame.

• For 3D components, values or expressions for the Velocity of sliding wall uw should
be specified. If the velocity vector entered is not in the plane of the wall, COMSOL
Multiphysics projects it onto the tangential direction. Its magnitude is adjusted to
be the same as the magnitude of the vector entered.
• For 2D components, the tangential direction is unambiguously defined by the
direction of the boundary. For this reason, the sliding wall boundary condition has
different definitions in different space dimensions. A single entry for the Velocity of
the tangentially moving wall Uw should be specified in 2D.
• For 2D axisymmetric components when Swirl flow is selected in the physics interface
properties, the Velocity of moving wall, ϕ component vw may also be specified.

Moving Wall
For an arbitrary wall movement, the condition u = uw may be prescribed. In this case,
the components of the Velocity of moving wall uw should be specified.

Specifying this boundary condition does not automatically cause the associated wall to
move. An additional Moving Mesh interface needs to be added to physically track the
wall movement in the spatial reference frame.

Leaking Wall
This boundary condition may be used to simulate a wall where fluid is leaking into or
leaving the domain with the velocity u = ul through a perforated wall. The
components of the Fluid velocity ul on the leaking wall should be specified.Constraint
Settings

832 | CHAPTER 10: MULTIPHYSICS INTERFACES AND COUPLINGS

This section is displayed by clicking the Show button ( ) and selecting Advanced
Physics Options.

• Wall in the COMSOL Multiphysics Reference Manual

• Moving Mesh Interface in the COMSOL Multiphysics Reference
Manual

Fluid-Solid Interface Boundary

The Fluid-Solid Interface Boundary condition defines the fluid load on the structure and
how structural displacements affect the fluid’s velocity.

Basic Modeling Steps for Fluid-Structure Interaction

The following steps describe the basics of how to set up a model of fluid-structure
interaction:

1 In the Model Wizard or Add Physics window, select the Fluid-Structure Interaction
interface. Click Next.
2 Select a study option—Stationary, Stationary, One-Way Coupled, Time Dependent, or
Time Dependent, One-Way Coupled from the Preset Studies branch on the Select Study
Type list; then click Finish.

If you also have the CFD Module, you can add a Stationary, One-Way
Coupled with Initialization or Transient, One-Way Coupled with Initialization
study for turbulence models requiring the wall distance. See Stationary,
One-Way Coupled with Initialization and Transient, One-Way Coupled
with Initialization in the COMSOL Multiphysics Reference Manual for
more information.

3 Create the geometry, which should contain a fluid domain and a solid domain.
4 Add the materials, typically a fluid and a solid. Then assign the material added last
to the domains that represents the solid (or the fluid, if the solid material was added
first).

THE FLUID-STRUCTURE INTERACTION INTERFACE | 833

 5 By default, the Fluid-Structure Interaction interface adds a Fluid Properties node for
the fluid domain as well as a Free Deformation node for the mesh displacements in
the moving mesh to all domains in the geometry.
For the solid domain, the default is the Linear Elastic Material node with the setting
to include geometric nonlinearity and an initially empty selection.
To use another material model for the solid, from the Physics toolbar, Domains>Solid
Mechanics submenu, select a material node for the solid. Add the domains that
represent the solid to its selection. That selection automatically overrides the Fluid
Properties and Free Deformation nodes in the solid.

Typically, fluid-structure interaction means that there are large

deformations. In this case, the Include geometric nonlinearity check box
should be selected in the Study Settings section of the Settings window for
the study.

6 Verify that the default boundary conditions are correctly assigned for the three types
of boundaries in the model: the Wall node for all fluid boundaries (and a Prescribed
Mesh Displacement node for zero mesh displacements on the same boundaries), the
Free node for all solid boundaries, and the Fluid-Solid Interface Boundary node on the
interior boundaries between the fluid and the solid. The Fluid-Solid Interface
Boundary node implements the coupling from the force exerted on the solid
boundary by the fluid as well as the as the structural velocities acting on the fluid as
a moving wall.
7 Add additional boundary conditions as needed. Typically the fluid domain needs an
Inlet node and an Outlet node for the inflow and outflow boundaries, respectively.
To add these nodes from the Physics toolbar, Boundaries>Laminar Flow submenu,
select Inlet and Outlet (if the fluid is laminar). The solid domain needs some
constraint such as a Fixed Constraint at some boundary.
8 Create the mesh and check that it resolves the domains sufficiently. A finer mesh
might be needed other than what the default mesh settings provide.
9 To solve the problem, from the Study toolbar click Compute. The solver settings
might require some adjustments depending on the characteristics of the model.

834 | CHAPTER 10: MULTIPHYSICS INTERFACES AND COUPLINGS

10 Also add additional physics features to the model such as Thermal Expansion or
Conjugate Heat Transfer, if applicable.

Studies and Solvers in the COMSOL Multiphysics Reference Manual

THE FLUID-STRUCTURE INTERACTION INTERFACE | 835

 The Fluid-Structure Interaction,
Fixed Geometry Interface
The Fluid-Structure Interaction, Fixed Geometry ( ) interface can be used to model
phenomena where a fluid and a deformable solid structure affect each other. Both the
fluid loading on the structure and the structural velocity transmission to the fluid can
be taken into account. The interface models situations where the displacements of the
solid are assumed to be small enough for the geometry of the fluid domain to be
considered as fixed during the interaction.

When a predefined Fluid-Structure Interaction, Fixed Geometry interface is added from

the Fluid Flow branch ( ) of the Model Wizard or Add Physics windows, Laminar Flow
and Solid Mechanics interfaces are added to the Model Builder.

In addition, the Multiphysics node is added, which automatically includes the

multiphysics coupling feature Fluid-Structure Interaction, Fixed Geometry.

On the Constituent Physics Interfaces

The Laminar Flow interface solves the Navier-Stokes equations, for conservation of
momentum, and the continuity equation, for conservation of mass, to compute the
fluid velocity and pressure. The interface supports incompressible flows and
compressible flows at low Mach numbers (typically less than 0.3). When the CFD
Module is available, turbulent flow is supported through the inclusion of a number of
turbulence models.

The Solid Mechanics interface is intended for general structural analysis of 3D, 2D, or
axisymmetric bodies. In 2D, plane stress or plane strain assumptions can be used. The
Solid Mechanics interface is based on solving Navier’s equations, and results such as
displacements, stresses, and strains are computed.

SETTINGS FOR PHYSICS INTERFACES AND COUPLING FEATURES

When physics interfaces are added using the predefined couplings, all specific settings
are included with the physics interfaces and the coupling features.

However, if physics interfaces are added one at a time, followed by the coupling
features, these modified settings are not automatically included.

836 | CHAPTER 10: MULTIPHYSICS INTERFACES AND COUPLINGS

For example, if single Laminar Flow and Solid Mechanics interfaces are added, COMSOL
adds an empty Multiphysics node. You can choose from the available coupling features,
but the modified settings are not included.

Coupling features are available from the context menu (right-click the
Multiphysics node) or from the Physics toolbar, Multiphysics menu.

PHYSICS INTERFACES AND COUPLING FEATURES

Use the online help in COMSOL Multiphysics to locate and search all the
documentation. All these links also work directly in COMSOL
Multiphysics when using the Help system.

Coupling Features
The Fluid-Structure Interaction, Fixed Geometry coupling feature node is described
in this section.

Physics Interface Features

Physics nodes are available from the Physics ribbon toolbar (Windows users), Physics
context menu (Mac or Linux users), or right-click to access the context menu (all
users).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

• The available physics features for The Single-Phase Flow, Laminar Flow Interface
are listed in the section Domain, Boundary, Pair, and Point Nodes for Single-Phase
Flow in the COMSOL Multiphysics Reference Manual.
• The available physics features for The Solid Mechanics Interface are listed in the
Domain, Boundary, Edge, Point, and Pair Nodes for Solid Mechanics section.

THE FLUID-STRUCTURE INTERACTION, FIXED GEOMETRY INTERFACE | 837

 Fluid-Structure Interaction, Fixed Geometry
Use Fluid-Structure Interaction, Fixed Geometry ( ) coupling when the deformations
of the structure are so small that they do not significantly affect the fluid domain. Two
effects can be modeled:

• The load from the fluid on the structure

• The structural velocity can act as a boundary condition on the fluid.

SETTINGS
The Label is the default multiphysics coupling feature name.

The Name is used primarily as a scope prefix for variables defined by the coupling node.
Refer to such variables in expressions using the pattern <name>.<variable_name>. In
order to distinguish between variables belonging to different coupling nodes or physics
interfaces, the name string must be unique. Only letters, numbers and underscores (_)
are permitted in the Name field. The first character must be a letter.

The default Name (for the first multiphysics coupling feature in the model) is fsifg1.

C O U P L I N G TY P E
Select Fully coupled (the default), Fluid loading on structure, or Velocity transmission to
fluid.

FLUID-STRUCTURE INTERACTION, FIXED GEOMETRY

This section defines the physics interfaces involved in the Fluid-Structure Interaction,
Fixed Geometry coupling. The Fluid flow and Solid mechanics lists include all applicable
physics interfaces.

The default values depend on how the node is created.

• If it is added from the Physics ribbon (Windows users), Physics contextual toolbar
(Mac and Linux users), or context menu (all users), then the first physics interface
of each type in the component is selected as the default.
• If it is added automatically when the physics interface is chosen in the Model Wizard
or Add Physics window, then the two participating physics interfaces are selected.

You can also select None from either list to uncouple the Fluid-Structure Interaction,
Fixed Geometry node from a physics interface. If the physics interface is removed from

838 | CHAPTER 10: MULTIPHYSICS INTERFACES AND COUPLINGS

the Model Builder, for example Solid Mechanics is deleted, then the Solid mechanics list
defaults to None as there is nothing to couple to.

If a physics interface is deleted and then added to the model again, then
in order to re-establish the coupling, you need to choose the physics
interface again from the lists. This is applicable to all multiphysics
coupling nodes that would normally default to the once present physics
interface. See Multiphysics Modeling Approaches in the COMSOL
Multiphysics Reference Manual.

THE FLUID-STRUCTURE INTERACTION, FIXED GEOMETRY INTERFACE | 839

 Theory for the Fluid-Structure
Interaction Interface
The Fluid-Structure Interaction Interface combines fluid flow with solid mechanics to
capture the interaction between the fluid and the solid structure. A Solid Mechanics
interface and a Single-Phase Flow interface model the solid and the fluid, respectively.
The Fluid-Structure Interaction (FSI) couplings appear on the boundaries between
the fluid and the solid. The physics interface uses an arbitrary Lagrangian-Eulerian
(ALE) method to combine the fluid flow formulated using an Eulerian description and
a spatial frame with solid mechanics formulated using a Lagrangian description and a
material (reference) frame.

The fluid flow is described by the Navier-Stokes equations, which provide a solution
for the velocity field ufluid. The total force exerted on the solid boundary by the fluid
is the negative of the reaction force on the fluid,

 
f = n ⋅  – pI +  μ ( ∇u fluid + ( ∇u fluid ) T ) – --- μ ( ∇ ⋅ u fluid )I 
2
(10-1)
 3 

where p denotes pressure, μ the dynamic viscosity for the fluid, n the outward normal
to the boundary, and I the identity matrix. Because the Navier-Stokes equations are
solved in the spatial (deformed) frame while the Solid Mechanics interface is defined
in the material (undeformed) frame, a transformation of the force is necessary. This is
done according to

dv
F = f ⋅ --------
dV

where dv and dV are the mesh element scale factors for the spatial frame and the
material (reference) frame, respectively.

The coupling in the other direction consists of the structural velocities

∂u solid
∂t

(the rate of change for the displacement of the solid), which act as a moving wall for
the fluid domain. The predefined Fluid-Solid Interface boundary condition includes
these couplings for bidirectionally coupled FSI simulations.

840 | CHAPTER 10: MULTIPHYSICS INTERFACES AND COUPLINGS

The solid mechanics formulation supports geometric nonlinearity (large
deformations). The spatial frame also deforms with a mesh deformation that is equal
to the displacements usolid of the solid within the solid domains. The mesh is free to
move inside the fluid domains, and it adjusts to the motion of the solid walls. This
geometric change of the fluid domain is automatically accounted for in COMSOL
Multiphysics by the ALE method.

ON E- WAY F LU I D-S TR U CT URE IN TE RA CTIO N

For small values of the solid displacement and its rate of change, the Fluid-Structure
Interaction interface includes one-way coupled model formulations. The one-way
coupled models sequentially solve for the fluid flow, compute the load from
Equation 10-1, and then apply it in the solution for the solid displacement. Since these
methods are unidirectional, the mesh deformation is excluded from the solution.
When applicable, the one-way coupled versions offer computationally cheaper
alternatives to a fully coupled counterpart.

Stationary and Time Dependent one-way coupled studies are available for selection
from the Preset Studies branch when adding a study. These studies include a Fluid study
step and a Solid study step. When an additional physics interface is added to the model,
it is by default added to both study steps.

In this case the one-way coupled study steps display under Preset Studies for Some
Physics Interfaces branch since the Fluid study step does not solve for the solid
displacement and vice versa. When using a turbulence model requiring the distance to
the closest wall, the Preset Studies includes a Wall Distance Initialization study step.

When solving a transient one-way coupled FSI model, besides saving the solution from
the Fluid study step with adequate frequency, it is advisable to save the solution from
the Solid study step at the same times as the fluid solution. This way, all the information
from the Fluid study step is used in the Solid study step.

Studies and Solvers in the COMSOL Multiphysics Reference Manual

THEORY FOR THE FLUID-STRUCTURE INTERACTION INTERFACE | 841

 The Joule Heating and Thermal
Expansion Interface
The Joule Heating and Thermal Expansion ( ) interface combines thermal, electric,
and structural multiphysics effects. The predefined interaction adds the
electromagnetic losses from the electric field as a heat source. In addition, the
temperature from the Heat Transfer in Solids interface acts as a thermal load for the
Solid Mechanics interface, causing thermal expansion.

You can use this multiphysics coupling for coupled thermal, electrical, and structural
analysis of, for example, the movement of some actuator, where an electric current
causes a temperature increase, which in turn leads to a displacement through thermal
expansion.

When a predefined Joule Heating and Thermal Expansion interface is added from the
Structural Mechanics branch ( ) of the Model Wizard or Add Physics windows, Solid
Mechanics, Electric Currents, and Heat Transfer in Solids interfaces are added to the
Model Builder.

In addition, the Multiphysics node is added, which automatically includes the

multiphysics coupling features Thermal, Expansion Electromagnetic Heat Source,
Boundary Electromagnetic Heat Source, and two Temperature Coupling nodes.

On the Constituent Physics Interfaces

The Electric Currents interface computes electric field, current and potential
distributions in conducting media under conditions where inductive effects are
negligible; that is, when the skin depth is much larger than the studied device.
Depending on the licensed products, time and frequency domain formulations that
account for capacitive effects are also provided. The Electric Currents interface solves
a current conservation equation based on Ohm's law using the scalar electric potential
as the dependent variable.

The Solid Mechanics interface is intended for general structural analysis of 3D, 2D, or
axisymmetric bodies. In 2D, plane stress or plane strain assumptions can be used. The
Solid Mechanics interface is based on solving Navier's equations, and results such as
displacements, stresses, and strains are computed.

The Heat Transfer in Solids interface provides features for modeling heat transfer by
conduction, convection, and radiation. A Heat Transfer in Solids model is active by

842 | CHAPTER 10: MULTIPHYSICS INTERFACES AND COUPLINGS

default on all domains. All functionality for including other domain types, such as a
fluid domain, is also available. The temperature equation defined in solid domains
corresponds to the differential form of the Fourier's law that may contain additional
contributions like heat sources.

In previous versions of COMSOL Multiphysics, a specific physics

interface called Joule Heating and Thermal Expansion was added to the
Model Builder. Now, a predefined multiphysics coupling approach is
used, improving the flexibility and design options for your modeling. For
specific details, see The Multiphysics Node and Multiphysics Modeling
Approaches in the COMSOL Multiphysics Reference Manual.

SETTINGS FOR PHYSICS INTERFACES AND COUPLING FEATURES

When physics interfaces are added using the predefined couplings, for example Joule
Heating and Thermal Expansion, specific settings are included with the physics interfaces
and the coupling features.

However, if physics interfaces are added one at a time, followed by the coupling
features, these modified settings are not automatically included.

For example, if single Solid Mechanics, Electric Currents, and Heat Transfer in Solids
interfaces are added, COMSOL adds an empty Multiphysics node. You can then choose
from the available coupling features, Thermal Expansion, Electromagnetic Heat Source,
Boundary Electromagnetic Heat Source, and Temperature Coupling, but the modified
settings are not included.

Coupling features are available from the context menu (right-click the
Multiphysics node) or from the Physics toolbar, Multiphysics menu.

TABLE 10-2: MODIFIED SETTINGS FOR A JOULE HEATING AND THERMAL EXPANSION INTERFACE

PHYSICS INTERFACE OR MODIFIED SETTINGS

COUPLING FEATURE

Solid Mechanics For the Solid Mechanics interface, under Structural

Transient Behavior, the Structural transient behavior is set
to Quasi-static.
Electric Currents No changes.
Heat Transfer in Solids For the Heat Transfer in Solids interface, under
Discretization, the shape function order for the
temperature is set to Linear.

THE JOULE HEATING AND THERMAL EXPANSION INTERFACE | 843

 TABLE 10-2: MODIFIED SETTINGS FOR A JOULE HEATING AND THERMAL EXPANSION INTERFACE

PHYSICS INTERFACE OR MODIFIED SETTINGS

COUPLING FEATURE

Thermal Expansion The Domain Selection is the same as that of the

participating physics interfaces.
The corresponding Solid Mechanics and Heat Transfer in
Solids interfaces are preselected in the Thermal Expansion
section.
Electromagnetic Heat The Domain Selection is the same as that of the
Source participating physics interfaces.
The corresponding Electric Currents and Heat Transfer in
Solids interfaces are preselected in the Electromagnetic
Heat Source section.
Boundary Electromagnetic The Boundary Selection contains all the boundaries of the
Heat Source participating physics interfaces.
The corresponding Electric Currents and Heat Transfer in
Solids interfaces are preselected in the Boundary
Electromagnetic Heat Source section.
Temperature Coupling Two Temperature Coupling nodes (described in the
COMSOL Multiphysics Reference Manual) are generated,
one for the transfer of temperatures to the Solid
Mechanics interface and one for the transfer of
temperatures to the Electric Currents interface.

PHYSICS INTERFACES AND COUPLING FEATURES

Use the online help in COMSOL Multiphysics to locate and search all the
documentation. All these links also work directly in COMSOL
Multiphysics when using the Help.

Coupling Features
• The Thermal Expansion (Multiphysics Coupling) coupling feature node is
described for The Thermal Stress Interface.
• The Electromagnetic Heat Source, Boundary Electromagnetic Heat Source, and
Temperature Coupling coupling feature nodes are described for The Joule Heating
Interface in the COMSOL Multiphysics Reference Manual.

844 | CHAPTER 10: MULTIPHYSICS INTERFACES AND COUPLINGS

Physics Interface Features
Physics nodes are available from the Physics ribbon toolbar (Windows users), Physics
context menu (Mac or Linux users), or right-click to access the context menu (all
users).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

• The available physics features for The Solid Mechanics Interfaceare listed in the
section Domain, Boundary, Edge, Point, and Pair Nodes for Solid Mechanics.
• The available physics features for The Heat Transfer in Solids Interface are listed in
the section Feature Nodes for the Heat Transfer in Solids Interface.
• The available physics features for The Electric Currents Interface are listed in the
section Domain, Boundary, Edge, Point, and Pair Nodes for the Electric Currents
Interface in the COMSOL Multiphysics Reference Manual.

If you have an add-on module, such as the Heat Transfer Module or AC/
DC Module, there are additional specialized physics nodes available and
described in the individual module documentation.

THE JOULE HEATING AND THERMAL EXPANSION INTERFACE | 845

 The Piezoelectric Devices Interface
The Piezoelectric Devices ( ) interface combines Solid Mechanics and Electrostatics
together with the constitutive relationships required to model piezoelectrics. Both the
direct and inverse piezoelectric effects can be modeled, and the piezoelectric coupling
can be formulated using either the strain-charge or stress-charge forms.

When a predefined Piezoelectric Devices interface is added from the Structural Mechanics
branch ( ) of the Model Wizard or from Add Physics windows, Solid Mechanics and
Electrostatics interfaces are added to the Model Builder.

In addition, the Multiphysics node is added, which automatically includes the

multiphysics coupling feature Piezoelectric Effect.

The participating Solid Mechanics interface includes the default Piezoelectric Material
feature with its selection set to all domains. The Electrostatics interface has a default
Charge Conservation, Piezoelectric feature with similar settings.

Such features can be also added manually to their corresponding interfaces similar to
any other material model therein.

The multiphysics node Piezoelectric Effect can be active only on the selection, where
both features Piezoelectric Material and Charge Conservation, Piezoelectric are active.

You input both the mechanical and electrical material data under the Piezoelectric
Material node. The data can be presented in either stress-charge or strain-charge form.

When it is used without an active Piezoelectric Effect coupling feature, the

Piezoelectric Material node works similarly to a Linear Elastic Material feature with the
material data input limited to anisotropic form using Voigt notations. All the electric
material data has no effect.

You use the Charge Conservation, Piezoelectric feature under Electrostatics to select
those domains, where the material is supposed to experience piezoelectric coupling.
When used without a counterpart under the Solid Mechanics interface (and/or
without the coupling feature) Charge Conservation, Piezoelectric node acts as an
ordinary Charge Conservation feature with its material data input limited to the
electric permittivity only.

All solid mechanics and electrostatics functionality for modeling is also accessible to
include surrounding elastic solids or air domains. For example, add any solid
mechanics material for other solid domain, a dielectric model for air (via Charge

846 | CHAPTER 10: MULTIPHYSICS INTERFACES AND COUPLINGS

Conservation feature), or a combination. Note that in order to model a non-solid
dielectric domain, you need to remove such domain from the domain selection for the
entire Solid Mechanics interface. This is because all material models under that
interface represent solid materials (with the Linear Elastic Material node being always
present and active in all those domains, where it is not explicitly overridden by any
other material model).

In 2D and 2D axial symmetry, adding a Piezoelectric Devices interface also adds

predefined base-vector coordinate systems for the material’s (in the plane 2D case)
XY-, YZ-, ZX-, YX-, XZ-, and XY-planes. These additional coordinate systems are
useful for simplifying the material orientation for the piezoelectric material.

On the Constituent Physics Interfaces

The Solid Mechanics interface is intended for general structural analysis of 3D, 2D, or
axisymmetric bodies. In 2D, plane stress or plane strain assumptions can be used. The
Solid Mechanics interface is based on solving Navier’s equations, and results such as
displacements, stresses, and strains are computed.

The Electrostatics interface is used to compute the electric field, the electric
displacement field and potential distributions in dielectrics under conditions where the
electric charge distribution is explicitly prescribed. The formulation is stationary but
for use together with other physics interfaces, also eigenfrequency, frequency-domain,
small-signal analysis and time-domain modeling are supported in all space dimensions.

The physics interface solves Gauss’ law for the electric field using the scalar electric
potential as the dependent variable.

In previous versions of COMSOL Multiphysics, a specific physics

interface called Piezoelectric Devices was added to the Model Builder. Now,
a predefined multiphysics coupling approach is used, improving the
flexibility and design options for your modeling. For specific details, see
The Multiphysics Node and Multiphysics Modeling Approaches in the
COMSOL Multiphysics Reference Manual.

SETTINGS FOR PHYSICS INTERFACES AND COUPLING FEATURES

When physics interfaces are added using the predefined couplings, for example
Piezoelectric Devices, specific settings are included with the physics interfaces and the
coupling features.

THE PIEZOELECTRIC DEVICES INTERFACE | 847

 However, if physics interfaces are added one at a time, followed by the coupling
features, these modified settings are not automatically included.

For example, if both Solid Mechanics and Electrostatics interfaces are added, COMSOL
adds an empty Multiphysics node. You can choose the available coupling feature
Piezoelectric Effect but the modified settings are not included.

Coupling features are available from the context menu (right-click the
Multiphysics node) or from the Physics toolbar, Multiphysics menu.

PHYSICS INTERFACES AND COUPLING FEATURES

Use the online help in COMSOL Multiphysics to locate and search all the
documentation. All these links also work directly in COMSOL
Multiphysics when using the Help system.

Coupling Feature
The Piezoelectric Effect coupling feature node is described in this section.

Additional Features
Additional nodes and subnodes available with this multiphysics interface are described
with the interfaces where they are available. Coupling Loss, Dielectric Loss, Mechanical
Damping, and Conduction Loss (Time-Harmonic) subnodes are available for Piezoelectric
Material under the Solid Mechanics interface. The Charge Conservation, Piezoelectric
feature is described for the Electrostatics interface.

• Piezoelectric Material
• Modeling Piezoelectric Problems

Piezoelectric Shear-Actuated Beam: Application Library path

Structural_Mechanics_Module/Piezoelectric_Effects/shear_bender

Surface Acoustic Wave Gas Sensor: Application Library path

MEMS_Module/Sensors/saw_gas_sensor

848 | CHAPTER 10: MULTIPHYSICS INTERFACES AND COUPLINGS

Piezoelectric Effect
The Piezoelectric Effect multiphysics coupling node ( ) passes the appropriate
relative permittivity from the Piezoelectric Material node in the Solid Mechanics interface
(where it is specified together with the other material properties of the piezoelectric)
to the Charge Conservation, Piezoelectric node in the Electrostatics interface. The
Charge Conservation, Piezoelectric node implements the domain level electrostatics
equations, and requires no user settings when it is coupled with the multiphysics node.

SETTINGS
The Label is the default multiphysics coupling feature name.

The Name is used primarily as a scope prefix for variables defined by the coupling node.
Refer to such variables in expressions using the pattern <name>.<variable_name>. In
order to distinguish between variables belonging to different coupling nodes or physics
interfaces, the name string must be unique. Only letters, numbers and underscores (_)
are permitted in the Name field. The first character must be a letter.

The default Name (for the first multiphysics coupling feature in the model) is pze1.

DOMAIN SELECTION
The domain selection is locked so that all applicable domains are selected. Only
domains that have both Charge Conservation, Piezoelectric selected in the Electrostatics
interface and Piezoelectric Material selected in the Solid Mechanics interface are selected.

PIEZOELECTRIC EFFECT
This section defines the physics involved in the Piezoelectric Effect multiphysics
coupling. The Solid mechanics and Electrostatics lists include all applicable physics
interfaces.

The default values depend on how the Piezoelectric Effect node is created.

• If it is added from the Physics ribbon (Windows users), Physics contextual toolbar
(Mac and Linux users), or context menu (all users), then the first physics interface
of each type in the component is selected as the default.
• If it is added automatically when a Piezoelectric Devices interface is selected in the
Model Wizard or Add Physics window, then the participating Solid Mechanics and
Electrostatics interfaces are selected.

You can also select None from either list to uncouple the Piezoelectric Effect node from
a physics interface. If the physics interface is removed from the Model Builder, for

THE PIEZOELECTRIC DEVICES INTERFACE | 849

 example Solid Mechanics is deleted, then the list defaults to None as there is nothing to
couple to.

If a physics interface is deleted and then added to the model again, then
in order to re-establish the coupling, you need to choose the physics
interface again from the lists. This is applicable to all multiphysics
coupling nodes that would normally default to the once present interface.
See Multiphysics Modeling Approaches in the COMSOL Multiphysics
Reference Manual.

Coupling Piezoelectric Devices with Acoustics

Using piezoelectric materials for an acoustic application is common, such as in sonars,
microphones, sensors, and so forth. This is why coupling piezoelectric devices with
acoustic domains is of particular interest for these applications.

The multiphysics couplings between Solid Mechanics and the acoustics

interfaces require the Acoustics Module.

Compared to a single piezoelectric model, you need to add a pressure acoustics physics
interface, for example, Pressure Acoustics, Frequency Domain or Pressure Acoustics,
Transient (depending on which study type you want to use) plus an Acoustic-Structure
Boundary coupling under the Multiphysics node. You can also directly create the nodes
that are needed for coupling by adding an Acoustic-Piezoelectric Interaction interface
from the Model Wizard or Add Physics windows. If solid and acoustic domains are
correctly defined, then the right coupling boundaries are automatically selected. Then
specify domains of application for each physics.

• Select solid domains and Piezoelectric Material domains in Solid Mechanics.

• Select electrostatic domains and Charge Conservation, Piezoelectric domains in
Electrostatics.
• Select acoustic domains in a Pressure Acoustics node.
• Under the Multiphysics branch, confirm that selections for Piezoelectric Effect and
Acoustic-Structure Boundary are the right ones. If several Pressure Acoustics, Solid

850 | CHAPTER 10: MULTIPHYSICS INTERFACES AND COUPLINGS

 Mechanics, or Electrostatics interfaces are present, select the right ones that need to
be coupled in the multiphysics interfaces.
• Continue the modeling process by entering the settings for each physics interface
and feature and define materials.

THE PIEZOELECTRIC DEVICES INTERFACE | 851

 The Magnetostriction Interface
The Magnetostriction ( ) interface combines Solid Mechanics and Magnetic Fields
together with the constitutive relationships required to model magnetostrictive
materials and devices. Both the direct and inverse magnetostriction effects can be
modeled.

When a predefined Magnetostriction interface is added from the Structural Mechanics

branch ( ) of the Model Wizard or from Add Physics windows, Solid Mechanics and
Magnetic Fields interfaces are added to the Model Builder.

In addition, the Multiphysics node is added, which automatically includes the

multiphysics coupling feature Magnetostriction.

The participating Solid Mechanics interface includes the default Magnetostrictive

Material feature with its selection set to all domains. The Magnetic Fields interface gets
a default Ampère’s Law, Magnetostrictive feature with similar settings.

Such features can be also added manually to their corresponding interfaces similar to
any other material models therein.

The multiphysics node Magnetostriction can be active only on the selection, where
both features Magnetostrictive Material and Ampère’s Law, Magnetostrictive are
active.

You input both the mechanical and magnetic material data under the Magnetostrictive
Material, and the electrical properties under Ampère’s Law, Magnetostrictive.

When it is used without an active Magnetostriction coupling feature, the

Magnetostrictive Material node works similarly to a Linear Elastic Material feature with
some limitations on the format for the elastic material data input. All the magnetic
material data and coupling data will have no effect.

You use the Ampere’s Law, Magnetostrictive feature under Magnetic Fields to select
those domains, where the material is supposed to experience magnetostrictive
coupling. When used without a counterpart under the Solid Mechanics interface
(and/or without the coupling feature) Ampère’s Law, Magnetostrictive node acts as
an ordinary Ampère’s Law feature with its material data input limited to the electric
properties only. The magnetic permittivity of free space will be assumed.

All solid mechanics and magnetics functionality for modeling is also accessible to
include surrounding elastic solids or air domains. For example, add any solid

852 | CHAPTER 10: MULTIPHYSICS INTERFACES AND COUPLINGS

mechanics material for other solid domain, a non-solid model for air (via Ampere’s Law
feature), or a combination. Note that in order to model a non-solid magnetic material,
you need to remove such domain from the domain selection for the entire Solid
Mechanics interface. This is because all material models under that interface represent
solid materials (with the Linear Elastic Material node being always present and active
in all those domains, where it is not explicitly overridden by any other material model).

In 2D and 2D axial symmetry, adding a Magnetostriction interface also adds

predefined base-vector coordinate systems for the material’s (in the plane 2D case)
XY-, YZ-, ZX-, YX-, XZ-, and XY-planes. These additional coordinate systems are
useful for simplifying the material orientation for non isotropic magnetostrictive
materials.

On the Constituent Physics Interfaces

The Solid Mechanics interface is intended for general structural analysis of 3D, 2D, or
axisymmetric bodies. The Solid Mechanics interface is based on solving Navier’s
equations, and results such as displacements, stresses, and strains are computed.

The Magnetic Fields interface solves the Ampère’s law for the magnetic flux density
field by using the vector magnetic potential as the dependent variable.

SETTINGS FOR PHYSICS INTERFACES AND COUPLING FEATURES

When physics interfaces are added using the predefined couplings, for example
Magnetostriction, specific settings are included with the physics interfaces and the
coupling features.

However, if physics interfaces are added one at a time, followed by the coupling
features, these modified settings are not automatically included.

For example, if both Solid Mechanics and Magnetic Fields interfaces are added,
COMSOL adds an empty Multiphysics node. You can choose the available coupling
feature Magnetostriction but the modified settings are not included.

Coupling features are available from the context menu (right-click the
Multiphysics node) or from the Physics toolbar, Multiphysics menu.

THE MAGNETOSTRICTION INTERFACE | 853

 PHYSICS INTERFACES AND COUPLING FEATURES

Use the online help in COMSOL Multiphysics to locate and search all the
documentation. All these links also work directly in COMSOL
Multiphysics when using the Help system.

Coupling Feature
The Magnetostriction coupling feature node is described in this section.

Additional Features
Additional nodes and subnodes available with this multiphysics interface are described
with the interfaces where they are available. Mechanical Damping and Thermal Expansion
subnodes are available for Magnetostrictive Material under the Solid Mechanics
interface. The Ampère’s Law, Magnetostrictive feature is described for the Magnetic
Fields interface.

• Magnetostrictive Material
• Ampère’s Law, Magnetostrictive
• Modeling Magnetostrictive Materials

Nonlinear Magnetostrictive Transducer: Application Library path

Structural_Mechanics_Module/Magnetostrictive_Devices/
nonlinear_magnetostriction

Magnetostriction
The Magnetostriction multiphysics coupling node ( ) passes the appropriate
magnetization contribution from the Magnetostrictive Material node in the Solid
Mechanics interface (where it is defined together with the material properties of the
magnetostrictive material) to the Ampère’s Law, Magnetostrictive node in the Magnetic
Fields interface. It also passes the mechanics stress contribution due to the applied
magnetic field back to the Magnetostrictive Material node.

COUPLING TYPE
From the list, choose one of these coupling types:

854 | CHAPTER 10: MULTIPHYSICS INTERFACES AND COUPLINGS

• Joule effect, also known as direct magnetostrictive effect, to include only the
deformation of the material caused by its magnetization in response to the applied
magnetic field.
• Villari effect, also known as inverse magnetostrictive effect, to include only the
change in the material magnetization as a results of applied mechanical stress or
strain.

• Fully coupled (the default) to include both the direct and inverse magnetostrictive
effects.
In 2D and 2D axial symmetry geometries only, select the Lorentz force contribution
check box to include the electric field changes and the body load caused by the material
motion in a magnetic field. This has no effect under stationary studies.

• Velocity (Lorentz Term)

THE MAGNETOSTRICTION INTERFACE | 855

 Hygroscopic Swelling Coupling
Hygroscopic Swelling
The Hygroscopic Swelling multiphysics coupling node ( ) is for moisture
concentration coupling between the Solid Mechanics interface and either the
Transport of Diluted Species or Transport of Diluted Species in Porous Media
interfaces.

Hygroscopic swelling is an internal strain caused by changes in moisture content. This

strain can be written as

ε hs = β h ( c mo – c mo,ref )

where βh is the coefficient of hygroscopic swelling, cmo is the moisture concentration,

and cmo,ref is the strain-free reference concentration.

SETTINGS
The Label is the default multiphysics coupling feature name.

The Name is used primarily as a scope prefix for variables defined by the coupling node.
Refer to such variables in expressions using the pattern <name>.<variable_name>. In
order to distinguish between variables belonging to different coupling nodes or physics
interfaces, the name string must be unique. Only letters, numbers, and underscores (_)
are permitted in the Name field. The first character must be a letter.

The default Name (for the first multiphysics coupling feature in the model) is hs1.

HYGROSCOPIC SWELLING PROPERTIES

Enter a Strain reference concentration cmo,ref . This is the concentration at which there
are no strains due to hygroscopic swelling.

Enter the Molar mass of the fluid, Mm. The default value is 0.018 kg/mol, which is the
molar mass of water.

The default Coefficient of hygroscopic swelling βh uses values From material. For
User defined select Isotropic (the default), Diagonal, or Symmetric from the list to enter
one or more components for a general coefficient of hygroscopic swelling tensor
βh.The default value for the User defined case is 1.5e-4 m3/kg. When a non-isotropic
coefficient of hygroscopic swelling is used, the axis orientations are given by the

856 | CHAPTER 10: MULTIPHYSICS INTERFACES AND COUPLINGS

coordinate system selection in the structural mechanics material node to which it
contributes.

The Include moisture as added mass check box is selected by default. When selected,
the mass of the fluid is included in a dynamic analysis, and when using mass
proportional loads. It will also contribute when computing mass properties.

HYGROSCOPIC SWELLING
This section defines the physics involved in the Hygroscopic Swelling multiphysics
coupling. The Moisture concentration and Structure lists include all applicable physics
interfaces.

You can select None from either list to uncouple the coupling node from a physics
interface. If the physics interface is removed from the Model Builder, for example Solid
Mechanics is deleted, then the list defaults to None as there is nothing to couple to.

If a physics interface is deleted and then added to the model again, then
in order to re-establish the coupling, you need to choose the physics
interface again from the lists. This is applicable to all multiphysics
coupling nodes that would normally default to the once present interface.
See Multiphysics Modeling Approaches in the COMSOL Multiphysics
Reference Manual.

HYGROSCOPIC SWELLING COUPLING | 857

 Couplings Between Structural
Mechanics Interfaces
Some of the structural mechanics interfaces are formulated using different types of
degrees of freedom. This needs to be taken into account in models where such
interfaces interact. You can easily set up common transitions using the predefined
multiphysics couplings. These are:

• Solid-Shell Connection
• Solid-Beam Connection
• Shell-Beam Connection

Solid-Shell Connection
Add the Solid-Shell Connection node to create transitions between domains modeled
using the Solid Mechanics or Multibody Dynamics interfaces and boundaries modeled
using the Shell interface.

The Solid-Shell Connection node is only available with some COMSOL products (see
http://www.comsol.com/products/specifications/)

CONNECTION SETTINGS
Select the Connection type — Solid boundaries to shell edges,
Solid and shell shared boundaries, or Solid and shell parallel boundaries.

For the two first options, there is an automatic search for possible adjacent geometrical
objects. If you want to modify the selections, select the Manual control of selections
check box. If to deselect it, the selections will be replaced by the automatic ones.

Solid boundaries to shell edges

If the automatic selection is not sufficient, select the Manual control of selections check
box. Then, select the connected solid boundaries in the Boundary Selection, Solid
section and the corresponding shell edges in the Edge Selection, Shell section.

Select Connected area defined by — Shell thickness, Selected solid boundaries, or Distance
from shell midsurface. This parameter determines how much of the selected solid
boundaries that are connected to the shell. The default is that a distance from the shell
edge having the size of the half the shell thickness in both perpendicular directions is

858 | CHAPTER 10: MULTIPHYSICS INTERFACES AND COUPLINGS

connected. Using Selected solid boundaries connects the selected boundaries in their
entirety to the shell. If you select Distance from shell midsurface, enter a Distance d. This
value is used instead of the half the shell thickness for defining the connection distance.

Select a Method — Rigid or Flexible. The Rigid version of the coupling only adds
constraints to the boundary of the solid, which in general causes local disturbances of
the stress field, since the thickness cannot change. When using the Flexible coupling,
three extra degrees of freedom are added along the shell edge. This allows for a more
accurate description of the transition, but the model can in some cases become
underconstrained if the mesh on the solid is very coarse.

You can modify the Connection tolerance Δ when either of the options Shell thickness or
Distance from shell midsurface are used. The search distance is slightly extended by the
distance Δ. The default value is 0.5% of the shell thickness, which allows for small
inaccuracies on for example a curved geometry.

Solid and shell shared boundaries

If the automatic selection is not sufficient, select the Manual control of selections check
box. Then, select the connected boundaries in the Boundary Selection section.

Solid and shell parallel boundaries

Select the connected solid boundaries in the Boundary Selection, Solid section and the
corresponding shell edges in the Boundary Selection, Shell section.

Select Distance evaluation — Shell properties, Geometrical distance, or User defined. This
parameter determines how the coupling treats the determination of the connection
distance, used when expressing th coupling between translation and rotation. When
Shell properties is selected, the distance is based on the shell thickness and offset, so
that a distance equal to half the shell thickness is used. If you select Geometrical
distance, the connection distance is computed from the geometrical distance between
the selected boundaries. For User defined, enter a Distance d, which defines the
connection distance.

• For more information about coupling different element types, see

Coupling Techniques.
• For details about the formulation of this coupling, see Connection
Between Shells and Solids

COUPLINGS BETWEEN STRUCTURAL MECHANICS INTERFACES | 859

 An example of couplings between shells and solids is shown in
Connecting Shells and Solids: Application Library path
Structural_Mechanics_Module/Tutorials/shell_solid_connection

Solid-Beam Connection

CONNECTION SETTINGS
Select the Connection type — Solid boundaries to beam points, or Solid and beam parallel
boundaries.

Solid boundaries to beam points

In this case, there is an automatic search for possible adjacent geometrical objects. If
you want to modify the selections, select the Manual control of selections check box. If
to deselect it, the selections will be replaced by the automatic ones.

For a manual selection, select the connected solid boundaries in the Boundary Selection,
Solid section and the corresponding beam points in the Point Selection, Beam section.

Select Connected area defined by — Section height, Selected boundaries, or Distance from
beam axis. This parameter determines how much of the selected solid boundaries that
are actually connected to the beam. The default is that a distance from the beam point
having the size of the half the beam section height in each direction is connected.
Using Selected boundaries connects the entire selected boundaries to the beam. If you
select Distance from beam axis, enter a Distance d. This is used instead of the section
height for defining the connection distance.

Select a Method — Rigid or Flexible. The Rigid version of the coupling only adds
constraints to the boundary of the solid, which in general causes local disturbances of
the stress field, since the thickness cannot change. When using the Flexible coupling,
three extra degrees of freedom are added to each beam point. This allows for a more
accurate description of the transition, but the model can in some cases become
underconstrained if the mesh on the solid is very coarse.

If needed, modify the Connection tolerance Δ.The search distance is slightly extended
by the distance Δ. The default value is 0.5% of the beam height.

Solid and beam parallel boundaries

Select the connected solid boundaries in the Boundary Selection, Solid section and the
corresponding shell edges in the Boundary Selection, Beam section.

860 | CHAPTER 10: MULTIPHYSICS INTERFACES AND COUPLINGS

Select Distance evaluation — Section height, Geometrical distance, or User defined. This
parameter determines how the coupling treats the determination of the connection
distance, used when expressing th coupling between translation and rotation. The
default is that a distance equal to half the section height is used. If you select
Geometrical distance, the connection distance is computed from the geometrical
distance between the solid boundary and the beam boundary. For User defined enter a
Distance d. This defines the connection distance.

• For more information about coupling different element types, see

Coupling Techniques.
• For details about the formulation of this coupling, see Connection
Between Shells and Solids

Shell-Beam Connection
Add the Shell-Beam Connection node to create transitions between boundaries or edges
modeled using the Shell interface, and edges or points modeled using the Beam
interface.

The Shell-Beam Connection node is only available with some COMSOL products (see
http://www.comsol.com/products/specifications/)

CONNECTION SETTINGS
Select the Connection type — Shell edges to beam points, Shell and beam shared edges,
Shell and beam parallel edges, or Shell boundaries to beam points.

For all options except Shell and beam parallel edges, there is an automatic search for
possible adjacent geometrical objects. If you want to modify the selections, select the
Manual control of selections check box. If to deselect it, the selections will be replaced
by the automatic ones.

Shell edges to beam points

If the automatic selection is not sufficient, select the Manual control of selections check
box. Then, select the connected shell edges in the Edge Selection, Shell section and the
corresponding beam points in the Point Selection, Beam section.

COUPLINGS BETWEEN STRUCTURAL MECHANICS INTERFACES | 861

 Select a Connected region — Selected edges, Distance (automatic), Distance (manual), or
Connection criterion.

• Using Selected edges makes all selected edges rigidly connected to the point on the
beam.
• For Distance (automatic), all parts on the shell edge, which are within the default
distance from the beam point, are connected. This distance is determined by the
cross section properties of the beam. It is contained in the variable beam.re as
described in the documentation for The Beam Interface.
• For Distance (Manual) enter a Connection radius rc. All parts on the shell edge, which
are within the given distance from the beam point, are connected.
• For Connection criterion enter a Boolean expression in the text field. The beam is
connected to the selected shell edge wherever the expression has a nonzero value.
The default value is 1, which is equivalent to using the Selected edges option.

Shell and beam shared edges

If the automatic selection is not sufficient, select the Manual control of selections check
box. Then, select the connected edges in the Edge Selection section.

Select an Offset definition — Along shell normal or Offset vector. For Along shell normal
enter an Offset ζ. For Offset vector enter values for d0 in the table. The offset is the
vector from the reference surface of the shell to the actual position of the beam. The
Offset vector is interpreted in the coordinate system selected in the Coordinate System
Selection section.

Shell and beam parallel edges

Enter a value for the Parallelism tolerance Δθ. The default is 1 degree. The two edges
are connected only where they are parallel within the specified tolerance.

Shell boundaries to beam points

If the automatic selection is not sufficient, select the Manual control of selections check
box. Then, select the connected shell boundaries in the Boundary Selection, Shell
section and the corresponding beam points in the Point Selection, Beam section.

Select an option from the Connected region list — Selected boundaries,

Distance (automatic), Distance (manual), or Connection criterion.

• Using Selected boundaries makes all selected boundaries rigidly connected to the
point on the beam.
• For Distance (automatic), all parts on the shell boundary, which are within the a
default distance from the beam point, are connected. This distance is determined by

862 | CHAPTER 10: MULTIPHYSICS INTERFACES AND COUPLINGS

 the cross section properties of the beam. It is contained in the variable beam.re as
described in the documentation for The Beam Interface.
• For Distance (manual) enter a Connection radius rc. All parts on the shell edge, which
are within the given distance from the beam point, are connected.
• For Connection criterion enter a Boolean expression in the text field. The beam is
connected to the selected shell boundary wherever the expression has a nonzero
values. The default value is 1, which is equivalent to using the Selected boundaries
option.

• For more information about coupling different element types, see

Coupling Techniques.
• For details about the formulation of this coupling, see Connection
Between Shells and Beams

Examples of all types of couplings between shells and beams are shown in
Connecting Shells and Beams: Application Library path
Structural_Mechanics_Module/Tutorials/shell_beam_connection

COUPLINGS BETWEEN STRUCTURAL MECHANICS INTERFACES | 863

864 | CHAPTER 10: MULTIPHYSICS INTERFACES AND COUPLINGS
 11

Glossary

This Glossary of Terms contains finite element modeling terms in a structural

mechanics context. For mathematical terms, and geometry and CAD terms specific
to the COMSOL Multiphysics® software and documentation, see the glossary in
the COMSOL Multiphysics Reference Manual. For references to more
information about a term, see the index.

865
 Glossary of Terms
anisotropy Variation of material properties with direction. Both global and local
user-defined coordinate systems can be used to define anisotropic material properties.

arbitrary Lagrangian-Eulerian (ALE) method A technique to formulate equations in a

mixed kinematic description. An ALE referential coordinate system is typically a mix
between the material (Lagrangian) and spatial (Eulerian) coordinate systems.

augmented Lagrangian method A method for solving contact problems.

Augmentation components are introduced for the contact pressure and the
components of the friction traction vector. Additional iteration levels are added where
the displacement, contact pressure and traction variables are solved separately. The
algorithm repeats this procedure until it fulfills a convergence criterion.

axial symmetry Symmetry in both load and geometry, solves for the radial (r) and
axial (z) displacement.

bar A line element that only has translational degrees of freedom, capable of
sustaining axial forces, with no bending moments, torsional moments, or shear forces.
Can be used on lines in 2D and 3D. Also called spar or truss element. In COMSOL
Multiphyiscs the term truss element is used.

beam A line element having both translational and rotational degrees of freedom.
Capable of sustaining axial forces, bending moments, torsional moments, and shear
forces. Can be used on lines in 2D and 3D.

body forces Forces distributed through the volume of a body.

buckling The sudden collapse or reduction in stiffness of a structure under a critical

combination of applied loads.

cable A tension-only truss member used to model large deformation including sag.

Cauchy stress The most fundamental stress measure defined as force/deformed area
in fixed directions not following the body.

compliance matrix The inverse of the elasticity matrix. See elasticity matrix.

866 | CHAPTER 11: GLOSSARY

constitutive equations The equations formulating the stress-strain relationship of a
material.

constraint Constrains the displacement or rotations to zero or a specified value.

contact model The mathematical method to model bodies that come into contact
with each other.

contact pair A pair that consists of some source boundaries and destination
boundaries and is used for contact modeling.

coordinate system Global Cartesian, local geometrical, application specific, and

user-defined coordinate systems. Loads, constraints, material properties, and variables
are defined in a specific coordinate system.

damping Dissipation of energy in a vibrating structure. A common assumption is

viscous damping where the damping is proportional to the velocity. See also Rayleigh
damping.

deformation gradient Tensor containing the complete information about the local
straining and rotation of the material. It is a positive definite second rank tensor.

destination boundary One side of a contact pair; the destination boundary is

prohibited to penetrate the source boundary.

double dogleg solver The default nonlinear solver for mechanical contact. This solver
is also useful for highly nonlinear simulations such as large plastic deformation or
hyperelastic materials.

eigenfrequency study Solving for the damped or undamped natural frequencies and
vibration modes of a structure.

elasticity matrix The matrix D relating strain to stresses:

σ = Dε

equilibrium equation The equation expressing the equilibrium formulated in the stress
components.

Eulerian Model described and solved in a coordinate system that is fixed (spatial
frame). See also Lagrangian and arbitrary Lagrangian-Eulerian method.

G L O S S A R Y O F TE R M S | 867
 first Piola-Kirchhoff stress A stress measure used when geometric nonlinearities arise.
All forces in COMSOL Multiphysics in case of geometric nonlinearity are of this type.

flexibility matrix The inverse of the elasticity matrix. See elasticity matrix.

free vibration The undamped vibration of a structure after it is displaced from the
equilibrium position and released. See also eigenfrequency analysis.

frequency response A harmonic analysis solving for the steady-state response from a
harmonic excitation. Typically a frequency sweep is performed, solving for many
excitation frequencies at one time.

geometric nonlinearity In solid mechanics, the deformation state characterized by

finite (or large displacements) but small to moderate strains. Not all material models
are suitable for large strain analysis, even though the displacement and rotation can be
large.

Green-Lagrange strain Nonlinear strain measure used in large-deformation analysis.

In a small strain, large rotation analysis, the Green-Lagrange strain corresponds to
the engineering strain, with the strain values interpreted in the original directions. The
Green-Lagrange strain is a natural choice when formulating a problem in the
undeformed state. The conjugate stress is the second Piola-Kirchhoff stress.

initial strain The strain in a stress-free structure before it is loaded.

initial stress The stress in a non-deformed structure before it is loaded.

isotropic material A material where the material properties are independent of

direction.

Lagrangian Model described and solved in a coordinate system that moves with the
material. See also Eulerian and arbitrary Lagrangian-Eulerian method.

large deformation The deformations are so large so the nonlinear effect of the change
in geometry or stress stiffening need to be accounted for. See also geometric
nonlinearity.

linear buckling analysis Solves for the linear buckling load using the eigenvalue solver.

linear viscoelasticity See visoelastic material.

868 | CHAPTER 11: GLOSSARY

load multiplier A load used in linearized buckling analysis for estimating the critical
load at which a structure becomes unstable.

mass damping parameter Rayleigh damping parameter, the coefficient in front of the
mass matrix.

mass participation factors A measure of the sensitivity of a certain eigenmode to a

uniform acceleration.

mixed formulation A formulation used for nearly incompressible materials, where the
mean stress have been added as a dependent variable to avoid numerical problems.

nonlinear geometry See large deformations.

orthotropic material An orthotropic material has at least two orthogonal planes of

symmetry, where material properties are independent of direction within each plane.
Such materials require nine elastic constants in the constitutive equations.

parametric study A study that finds the solution dependence due to the variation of a
specific parameter.

pinned A constraint condition where the displacement degrees of freedom are fixed
but the rotational degrees of freedom are free, typically used for frames modeled using
beam and truss elements.

plane strain An assumption on the strain field where all out-of-plane strain
components are assumed to be zero.

plane stress An assumption on the stress field, all out-of-plane stress components are
assumed to be zero.

plate Thin plane structure loaded in the normal direction.

principle of virtual work States that the variation in internal strain energy is equal to
the work done by external forces.

principal stresses/strains Normal stresses/strains with no shear components that act

on the principal planes. The magnitude of the principal stresses/strains are
independent of the coordinate system used.

G L O S S A R Y O F TE R M S | 869
 rate of strain tensor The rate at which the strain tensor changes with respect to time
in time-dependent studies.

quasi-static transient study The loads vary slow enough for the inertia terms to be
negligible. A transient thermal study coupled with a structural analysis can often be
treated as quasi-static.

Rayleigh damping A viscous damping model where the damping is proportional to the
mass and stiffness through the mass and stiffness damping parameters.

rotational degrees of freedom Degrees of freedom associated with a rotation around

an axis. Beams, rigid connectors, rigid domains, and shells have rotational degrees of
freedom.

second Piola-Kirchhoff stress Conjugate stress to Green-Lagrange strain used in

large deformation analysis. The orientations of the stress components follow the
material directions.

shell elements A thin element where both bending and membrane effects are
included.

source boundary One side of a contact pair; the destination boundary is prohibited
to penetrate the source boundary.

spar see bar.

spin tensor The skew-symmetric part of the velocity gradient tensor.

stationary study A study where the loads and constraints are constant in time. Also
called static.

strain Relative change in length, a fundamental concept in structural mechanics.

stress Internal forces in the material, normal stresses are defined as forces/area normal
to a plane, and shear stresses are defined as forces/area in the plane. A fundamental
concept in structural mechanics.

stress stiffening The geometrically nonlinear effect which supplies the out-of-plane
stiffness for membranes, for example.

870 | CHAPTER 11: GLOSSARY

stiffness damping parameter Rayleigh damping parameter, the coefficient in front of
the stiffness matrix.

strain energy The energy stored by a structure as it deforms under load. Also called
elastic energy.

time dependent study A time-dependent or transient study shows how the solution
varies over time, taking into account mass, mass moment of inertia, and damping.

Tresca stress An effective stress measure that is equal to the maximum shear stress.

truss See bar.

viscoelastic material Viscoelastic materials have a time-dependent response, even if

the loading is constant. Many polymers and biological tissues exhibit such a behavior.
Linear viscoelasticity is a commonly used approximation where the stress depends
linearly on the strain and its time derivatives.

G L O S S A R Y O F TE R M S | 871
872 | CHAPTER 11: GLOSSARY
I n d e x
A absolute values 316 geometric nonlinearity 137
absolute-tolerance parameters 193 harmonic perturbation 48
abstract rigid domain 355 initial stress and strain 478, 480, 685
acceleration loads 76 large plastic strains 458
acoustic-structure interaction, frequency linear buckling study 55
domain interface 163 linear elastic material 732
added mass (node) 543 load cases 73–74
added mass, theory 383 membrane interface 679
adding MEMS materials database 95
connections 66 modal mass 38, 152, 173
additive strain decomposition (check model superposition 44, 46
box) 435, 440 Mooney-Rivlin, two parameters 445
adhesion (node) 574 Ogden 445
alpha coefficient 470 piezoelectric devices 848, 854
Anand viscoplasticity 332 piezoelectric materials database 95
angle of internal friction 289, 297, 470 plasticity models 458
angular acceleration and velocity 351 prestressed bolts 55, 183
anisotropic materials rigid connector 553
defining 241 sharing edges 68, 860, 863
elastic properties 434, 683 shells 617
antisymmetry (node) soil plasticity 471
beam interface 746 solid mechanics 426
shell and plate interfaces 645 spring foundation 161
solid mechanics 522 thermal expansion 475
truss interface 814 thermal stress 822
Application Libraries window 28 trusses 801
application library examples viscoelastic material 454
beam cross sections 764 viscoplasticity models 463
beams 722 applied force (node) 502, 555
Cam-Clay material 451 theory 357
creep models 461 applied moment (node) 504, 556
cross section data 725 theory 357
damping 490 applying
eigenfrequency analysis 38 loads 71
fixed constraint 517 moments 75
fluid-structure interaction 828 Arbitrary Lagrangian-Eulerian (ALE)

INDEX| 873
 method 133 stress evaluation 704
area, beam cross sections 768 thermal expansion 733
Arruda-Boyce material 275 thermal strain 700
associated flow rule 308 Beltrami-Michell 774
attachment (node) bending shear stress 783
shells 666, 757 bending shear stresses 770
solid mechanics 561 bending stress 783
theory 356 biaxial compression 313
augmented Lagrangian method 21, 143, biaxial data 315
392 biaxial tension 472
axial stress 782 bimoment 782
axial symmetry Blatz-Ko material 277
constraints and 81 body load (node)
initial stress and strain 228 shell and plate interfaces 647
axisymmetric models solid mechanics 525
solid mechanics 58 bolt pre-tension (node) 576
azimuthal wave number 389 bolt pre-tension, theory 178
bolt selection (node) 578
B back stress 304, 333
bolted joints 178
beam
boundary conditions
coupling to a solid 67
contact pairs 147
beam connection (node) 752
shell and plate interfaces 596
shells 659
boundary load (node) 529
solid mechanics 563
boundary nodes
beam cross section interface 785
beam interface 722
theory 768
fluid-structure interaction 828
beam interface 719
membrane interface 679
theory 690
shell and plate interfaces 618
beams
solid mechanics 426
cross section data 724
truss interface 801
initial loads and strains 701
box sections, beams 725
initial stresses and strains 735
Bresler-Pister criterion 312
linear elastic material 731
buckling 20
loads applied 751
built-in couplings 66
prescribed acceleration 740
bulk modulus
prescribed displacement/rotation 737
elastic moduli 237, 433, 537
prescribed velocity 739
Burgers vector 330
section orientation 728
strain-displacement/rotation 699 C calcite 316

874 | I N D E X
calculating stress and strain 598 cohesionless soils 296
Cam-Clay material (node) 450 cohesive-frictional materials 313
Cam-clay model 318 common sections, beam interface 725
canonical systems 201 common settings 25
carbonate rocks 316 complementarity 308
Cauchy stress 222 complex mechanical energy flux 410
Cauchy stress tensor 130 complex modulus 125
Cayley-Hamilton theorem 211, 225 compressive meridians 226, 290, 313–314
center of gravity, beam cross sections concrete 312, 314
768 concrete (node) 471
center of mass conduction loss (time-harmonic) (node)
boundary (node) 506 493
edge (node) 506 connection
point (node) 506 beam edge to shell edge 610
rigid domains 351 beam perpendicular to solid 608
center of rotation 772 beam point to shell boundary 611
boundary (node) 553 beam point to shell edge 612
edge (node) 553 connections
point (node) 553 beams, shells 609
rigid domains 353 beams, shells, solids 602
centrifugal acceleration loads 76 shell perpendicular to solid 603
ceramics 313 consistency parameter 308
CFD Module 828, 831 constitutive relation, membranes 675
CFL number, pseudo time stepping, and constraints 78
827 contact (node) 566
change thickness (node) contact formulation 152
shell and plate interfaces 621 contact help variables 395
solid mechanics 430 contact modeling, friction 397
charge conservation, piezoelectric converse piezoelectric effect 96
(node) 445 coordinate system, beam cross sections
circle, Mohr 289 770
circular sections, beams 725 coordinate systems
Coble creep 331 constraints and 79
coefficient of thermal expansion loads and 71
beams 733 local edge system 636
shells and plates 626 solid mechanics theory 201
cohesion 289, 470 coordinate systems, membranes 673
cohesion sliding resistance 395 Coulomb friction 572

INDEX| 875
 coupling deformation resistance 332
beam to a shell 67 saturation coefficient 333
beam to a solid 67 sensitivity 333
shell to a solid 67 destinations and sources 144
coupling loss (node) 492 deviatoric creep 327, 330
coupling operator 68 deviatoric stress 284
creep (node) 459 dielectric loss (node) 493
creep dissipation rate density 334 dielectric loss factor 108
creep strain rate 323 diffusional creep 330
critical load factor 54 dilatational contributions 266
cross section (node) 724 direct piezoelectric effect 96, 337
cross section data (node) 803 dislocation creep 331
crystal cleavage 316 dispersion curves 40
crystal cut standards 99 displacement field, defining 128
crystalline solids and creep 324 displacement gradients 204
curvature 701 displacement variables
cyclic symmetry, theory 387 element types and 65
cylindrical coordinate systems 58 dissipated energy 409
dissipated energy density 311, 334
D damped eigenfrequency study 125
dissipated energy density rate 257
damping
dissipation, piezoelectric materials 340
equation of motion, and 120
distortional contributions 266
loss factors 124
disturbance factor 318
losses and 119
documentation 27
point mass 653, 751, 817
dolomite 316
solid mechanics 488
domain nodes
viscoelastic materials, and 261
fluid-structure interaction 828
damping (node)
solid mechanics 426
solid mechanics 487
double dogleg nonlinear solver 154
damping models 124
Drucker-Prager criterion 291
decohesion 398
ductile materials 286
defining
Dulong-Petit law 243
anisotropic materials 241
dynamic cyclic symmetry 388
constraints 78
dynamic frictional coefficients 397
isotropic materials 237
multiphysics models 162 E edge load (node)
orthotropic materials 239 beam interface 747
thermoelastic materials 243 shell and plate interfaces 649
deformation gradient 204 solid mechanics 531

876 | I N D E X
 truss interface 815 explicit damping 125
edge nodes external loads, shell and plate interfaces
beam interface 722 597
fluid-structure interaction 828 external strain (node) 480, 482
membrane interface 679 external stress (node) 478, 629, 686, 736,
solid mechanics 426 808
truss interface 801 external stress-strain relation (node) 494
effective creep strain rate 327
F face load (node) 648
effective plastic strain rate 283
failure surfaces 282
effective stress 784
first Piola-Kirchhoff stress 222
effective stress tensor 317
fixed constraint (node) 499, 516
eigenfrequency study 36
fixed constraint, theory 356
solid mechanics 370
Floquet periodicity, theory 387
eigenvalue solvers 37
flow rule 307
eight-chain model 276
fluid pore pressure 295
elastic deformation tensor 213
fluid-solid interface boundary (node) 833
elastic energy 407
fluid-structure interaction 836, 838
elastic Green-Lagrange strain tensor 214
setting up a model 833
elastic material properties 433–434, 682–
fluid-structure interaction interface 825
683
theory 840
elastic moduli 237
fold lines 594
elastic right Cauchy-Green tensor 214
fold-line limit angle 615
elastic strain energy 280
follower loads 675
elastic volumetric strain variable 321
force linear strains (check box) 435, 440,
elasticity matrix 237
447, 449, 495, 684
elastoplastic materials 282, 313
free (node) 509
elcontact variable 393
free-free modes 38
element types 65
frequency domain study
elliptic cap 293
solid mechanics 369
elplastic 308, 310
frequency response study
emailing COMSOL 29
loss factor damping 125
energy dissipation 409
viscous damping 125
energy function 263
friction (node) 572
energy quantities 407
friction forces 573
equation of motion, damping and 120
friction in contact modeling 397
equivalent viscous damping 125
friction models 572
evanescent modes 39
friction traction penalty factor 395
excitation frequency 125
friction, angle 289, 297, 470

INDEX| 877
 frictionless materials 290, 292 hexagonal prism 285
Hill orthotropic plasticity 298
G Gao material 278
Hill’s effective stress 300
gap distance variable 394
Hoek-Brown criterion 316
Garofalo law 329
hole (node) 786
Gauss points 308, 310
H-profile sections, beams 725
general extrusion operator 68
hydrostatic axis 226
generalized Hoek-Brown criterion 317
hydrostatic pressure 223, 289
generalized Maxwell model 255
hydrostatic stress 296
Gent material 276
hygroscopic swelling 77
geological strength index (GSI) 317
hygroscopic swelling (node) 475
geometric nonlinearity 133
hygroscopic swelling (node), multiphys-
membranes and 672
ics 856
micromechanics, and 127
hyperbolic sinus 329
piezoelectric devices 137
hyperelastic material 408
glass transition temperature 260
hyperelastic material (node) 442
global
hyperelastic materials 269
coordinate systems 201
nearly incompressible 268
GMRES iterative solvers 191
theory 263
gradient displacements 204
hysteretic loss 105
gravity (node) 526
Green’s theorem 777 I I.R.E. standard, for material orientation
Green-Lagrange strain 128 97
Green-Lagrange tensor 207 ideal plasticity 301
IEEE standard, for material orientation
H Haigh–Westergaard coordinates 226
97
hardening constant 332
imperfection sensitivity 54
hardening models
implementation
theory 300
beams 702
hardening sensitivity 332
trusses 794
harmonic loads 369
include geometric nonlinearity (check
harmonic perturbation 46
box) 436
harmonic perturbation (node)
inelastic deformation tensor 214
truss interface 581, 669, 758
inertial effects, contact modeling 152
hear center 772
inertial forces 352
hear correction factor 771
initial bulk modulus 272
heat dissipation 125
initial loads and strains, beams 701
Heat Transfer Module 828, 831
initial stress and strain 341
Hencky plastic strain 310
theory 227
Hermitian matrices 191

878 | I N D E X
 initial stress and strain (node) 477 Kelvin–Voigt viscoelastic model 259
beam interface 735 kinematic constraints 84
membrane interface 684 kinematic hardening 303
shell and plate interfaces 628 kinematics, rigid domain 349
solid mechanics 477 kinetic energy 409
truss interface 808 knowledge base, COMSOL 29
initial values (node) Kuhn-Tucker conditions 308
beam interface 724
L Lade-Duncan criterion 296
fluid-structure interaction 831
Lagrange shape functions, trusses and
rigid domains 498
792
rigid domains, theory 353
Lagrangian formulations 203
shell and plate interfaces 620
Lamé parameters 238, 433
solid mechanics 430
large deformation modeling 203
initial yield stress 301–302
large deformations 20
interior wall (node) 828
piezoelectric materials 137
internet resources 27
large plastic strain 307
invariants 264
large strain plasticity 265
inverse piezoelectric effect 337
leaking wall, wall boundary condition 832
isochoric
left Cauchy-Green tensor 208
contributions 266
Lie derivative 309
process 205
limestone 316
strain energy density 268
limiting chain extensibility 277
isochoric strain energy 280
linear buckling study 20, 137
isochoric-elastic
linear elastic material (node) 431, 624,
deformation 266
681
Green-Lagrange strain tensor 266
beam interface 731
right Cauchy-Green tensor 266
shell and plate interfaces 623
isotropic hardening 301
truss interface 805
isotropic materials
linear elastic materials 92
defining 237
linear viscoelasticity 253
elastic properties 433, 682
linearized buckling analysis 53
isotropic plasticity 282
linper operator 46
isotropic rocks 317
load cases 73
iterative solvers 191
load multiplier 54
J joints, bolted 178 loads
Joule heating and thermal expansion in- acceleration 76
terface 163, 842 applied to beams 751
pressure 76
K k coefficient 471

INDEX| 879
 singular 74 mass matrix scaling 50
total 77 mass moment of inertia 751
local CFL number 827 material coordinates 200
local coordinate systems 202 material frame 128
local edge system 636 material models 90
location materials
boundary (node) 504 nearly incompressible 93
edge (node) 504 piezoelectric 337
point (node) 504 viscoelastic 253
locking 269 materials, hyperelastic 263
Lode angle 225 Matsuoka-Nakai criterion 295
Hoek-Brown criterion 317 max scaling 51
Mohr-Coulomb criterion 290 max shear stress factor 771
Tresca criterion 285 mechanical damping (node) 491
Willam-Warnke criterion 313 mechanical energy flux 410
logarithmic decrement 120 membrane interface 677
logarithmic plastic strain 310 theory 672
long-term shear modulus 255 meridians, tensile and compressive 226,
loss factor damping 290, 313–314
solid mechanics and 124 metal plasticity 283
solid mechanics theory 124 metals 286, 291
springs, and 161 metals and creep 324
loss modulus 125, 261 Microfluidics Module 831
losses and damping 119 Mindlin plate theory 604
low-reflecting boundary (node) 548 Mindlin-Reissner type shell 588
low-reflecting boundary, theory 386 MITC shell formulation 588, 594
mixed formulation 450
M macroscopic shear modulus 275
mixed formulations 94
magnetostriction 341
mobilized planes 295
magnetostriction (node) 854
mode analysis study 39–40
magnetostriction multiphysics interface
modeling fluid-structure interaction 833
852
modeling, large deformations 203
magnetostrictive material 447
modified Cam-clay model 318
magnetostrictive material (node) 447
modified Mohr-Coulomb criterion 298
marble 316
modified tensors 267
mass and moment of inertia (node) 505,
Mohr’s circle 769
557
Mohr-Coulomb criterion 289
theory 355
moment computations 86
mass density 205

880 | I N D E X
 moments 352 Neo-Hookean material 269
beams 727 Neumann boundary conditions
shells and plates 598 applied force 502
solid mechanics and 75 applied moment 504
moments of inertia 690, 762 no rotation (node)
moments of inertia, beam cross sections beam interface 744
768 shell and plate interfaces 641
Mooney-Rivlin material no slip, wall boundary condition 832
five parameters 271 nodes, common settings 25
nine parameters 272 nominal stress 230
two parameters 270 non-associated flow rule 308
moving mesh interface, piezoelectric de- nonlinear elastic material 408
vices and 140 nonlinear elastic material (node) 437
MPH-files 28 Norton equation 327
mu coefficient 470 Norton-Bailey model 328
Mukherjee-Bird-Dorn equation 331
O octahedral normal stress 313
multiaxial stress states 312
octahedral plane 226
multibody dynamics interface 561
Ogden material 273
multiphysics
one-way coupled model formulations
hygroscopic swelling 856
841
magnetostriction 852, 854
orientation, piezoelectric material 97
piezoelectric devices 846
orthotropic materials
piezoelectric effect 849
defining 239
multiphysics coupling
elastic properties 434, 683
fluid-structure interaction 836, 838
loss factor damping, and 489
Joule heating and thermal expansion
Ottosen criterion 314
842
over-consolidation pressure 319
thermal expansion (node) 822
P pair nodes
thermal stress 820
beam interface 722
multiphysics modeling 162
fluid-structure interaction 828
MUMPS direct solvers 191
membrane interface 679
Murnaghan material 278
solid mechanics 426
N Navarro-Herring creep 330
truss interface 801
Navier-Stokes equations 840
parametric analysis 20
nearly incompressible hyperelastic mate-
penalized friction traction 395
rials 268
penalty factor relaxation 148
nearly incompressible material 450
penalty factors
nearly incompressible materials 93, 228
contact node, and 569

INDEX| 881
 contact pairs and 147 plastic multiplier 308
theory 395 plastic potential 308
penalty method 392 plasticity (node) 454
perfectly elastoplastic materials 301 plasticity models 455
perfectly plastic hardening 301 plate interface 613
periodic boundary conditions 545, 654 theory 588
periodic condition (node) plates
shell 654 external loads 597
solid mechanics 545 initial stresses and strains 628
periodic conditions, theory 387 linear elastic material 623
phase (node) 579 MITC shell formulation 594
beam interface 758 prescribed acceleration 638
shell and plate interfaces 668 prescribed displacement/rotation 631
solid mechanics 579 prescribed velocity 636
truss interface 817 stress and strain calculations 598
physics interfaces symmetry and antisymmetry 596
connecting to DOF 65 thermal expansion 625
physics interfaces, common settings 25 point load (node) 532
piezoelectric coupling 846 beam interface 750
piezoelectric crystal cut 97 shell and plate interfaces 651
piezoelectric devices multiphysics inter- point load on axis (node) 534
face 846 point loads example 74
piezoelectric effect (node) 849 point mass (node)
piezoelectric losses 105 beam interface 751
piezoelectric material (node) 445 shell and plate interfaces 652
piezomagnetic coupling 449 truss interface 816
piezomagnetic coupling matrices 342 point mass damping (node)
pinned (node) beam interface 751
beam interface 742 shell interface 653
shell and plate interfaces 640 truss interface 817
truss interface 811 point nodes
pipe sections, beams 725 beam interface 722
plane stress and strain 57, 423 fluid-structure interaction 828
planes, symmetry and constraints 82 membrane interface 679
plastic deformation gradient 215 solid mechanics 426
plastic element 308, 310 truss interface 801
plastic flow rule 310 Poisson type equation 774
plastic Green-Lagrange strain 215 Poisson’s ratio 94, 237, 433, 536

882 | I N D E X
 polynomial hyperelastic material 271 rate independent plasticity 308
power law 327 rate of strain tensor 212, 309
Prager’s hardening rule 304 Rayleigh damping 122
Prandtl stress function 779 rectangle sections, beams 725
predeformation (node) 542 reference coordinates 200
prescribed acceleration (node) 515 reference point for moment computa-
beam interface 740 tion 86
shell and plate interfaces 638 refpnt variable 424
prescribed displacement (node) relaxation, of penalty factor 148
solid mechanics 510 renaming
prescribed displacement, theory 356 displacement DOF 66
prescribed displacement/rotation (node) resonant frequency 122
500 results evaluation, for shells 601
beam interface 737 right Cauchy-Green deformation tensor
shell and plate interfaces 631 207, 263
theory 356 right polar decomposition 206
prescribed velocity (node) 513 right stretch tensor 206
beam interface 739 rigid body 349
shell and plate interfaces 636 rigid body suppression (node) 523
solid mechanics 513 rigid connection type, shells 604
pressure loads 76 rigid connector 391
pressure-wave speeds 238, 434 rigid connector (node) 549, 661, 754
prestressed analysis, eigenfrequency rigid connector theory 356
study 137 rigid connectors
prestressed analysis, frequency domain kinematic constraints and 84
study 137 moments and 75
primary creep 323 rigid domain (node) 495
principal stresses 224, 226 theory 349
principal stretches 264 rigid domains
principle of virtual work 367 angular acceleration 352
Prony series 255 angular velocity 351
propagating modes 39 dynamics 352
pseudo time stepping ring load (node) 533
advanced settings 827 rock mass 318
rock types 316
Q quaternion constraint 350
rocks 316
quaternion representation, of rigid con-
rocks (node) 472
nector 391
roller (node) 520
R radius of gyration 769

INDEX| 883
 rotated coordinate system 100 symmetry and antisymmetry 596
rotating frame (node) 528 thermal expansion 625
rotation matrix 350 shift function 452
rotation, rigid body 349 sign conventions 223
rotational degrees of freedom 691 simply supported (node) 642
rotational joints, beams 85 singular loads 74
size factors 315
S safety (node) 359, 483, 787
skew-symmetric part 212
Saint-Venant torsion 783
sliding wall, wall boundary condition 832
saturated cohesive soils 286
slip, wall boundary condition 832
scaling of eigenvectors 37
slit boundary 178
second Piola-Kirchhoff stress 130, 202,
SLS model 258
222
small plastic strain 307
secondary creep 323
soil deformation theory 320
section orientation (node) 728
soil plasticity (node) 469
selecting
solid
solvers 190
coupling to a beam 67
shape factors 315
coupling to a shell 67
shear area 771
solid connection (node) 658
shear modulus expression 237, 433, 537
solid mechanics
shear stresses 224, 289, 312
damping 488
shear-wave speeds 238, 434
edge loads 531
shell
initial stresses and strains 477
coupling to a beam 67
prescribed acceleration 515
coupling to a solid 67
prescribed velocity 513
shell connection (node)
solid mechanics interface 422
beams 753
theory 198
solid mechanics 565
solid-shell connection (node) 858, 861
shell interface 613
solver methods
theory 588
augmented Lagrangian 392
shells
penalty method 392
external loads 597
solver parameters 190
initial stresses and strains 628
solver settings 190
linear elastic material 623
SOR line solvers 192
MITC shell formulation 594
sources and destinations 144
prescribed acceleration 638
spatial coordinates 200
prescribed displacement/rotation 631
spatial stress tensor 130
prescribed velocity 636
spatially mobilized planes (STP) 295
stress and strain calculations 598

884 | I N D E X
spin tensor 212 beams 699
spring constant 160 trusses 793
spring foundation 507 study steps, geometric nonlinearity and
spring foundation (node) 507, 535, 558 133
spring foundation, solid mechanics 160 study types
spring foundation, theory 379 eigenfrequency 370
spring-damper (node) 810 frequency domain, solid mechanics 369
St Venant torsion 781 parametric 20
St Venant-Kirchhoff material 270 surface traction and reaction forces 88
St. Venant’s principle 74 symmetric matrices 190
standard linear solid model 258 symmetry (node)
standard settings 25 beam interface 745
static frictional coefficients 397 shell and plate interfaces 643
stationary solvers 190 solid mechanics 521
steady-state creep 323 truss interface 813
steady-state stiffness 255 symmetry constraints 81
storage modulus 125, 261
T tangent modulus 302
Storakers material 274
tangential strains 674
stored energy 407
technical support, COMSOL 29
straight edge constraint (node) 804
temperature loads 76
straight edges 795
tensile meridian 226
strain energy density 280, 408
tensile meridians 290, 292, 297, 313–314
strain-displacement, trusses 792
tension cut-off 297
strain-displacement/rotation 699
tertiary creep 323
strains, membranes 673
theory
stress
beam cross sections 768
Cachy 222
beam interface 690
first Piola-Kirchhoff 222
fluid-structure interaction 840
second Piola-Kirchhoff 222
membrane interface 672
stress and strain, piezoelectric devices
shell and plate interfaces 588
118, 337, 342
solid mechanics interface 198
stress components, beam cross sections
thermal expansion
773
loads and 76
stress evaluation, beams 704
thermal expansion (multiphysics cou-
stress linearization (node) 582
pling) 822
stress stiffening 672
thermal expansion (node) 473
stresses, membranes 673
beam interface 733
stress-strain relation
shell and plate interfaces 625

INDEX| 885
 solid mechanics 473, 518, 634, 743, 812 linear elastic material 805
thermal expansion, hyperelastic materi- straight edge 795
als and 264 strain-displacement 792
thermal expansion, Joule heating and 842 two-point tensor 205
thermal strain
U undrained shear strength 285
beams 700
uniaxial compression 315–316
thermal stress interface 820
uniaxial compressive strength 312
modeling 162
uniaxial tension 284, 472
thermal-electric-structural interaction
units, loads and 71
163
u-profile sections, beams 725
thermoelastic materials, defining 243
user defined material 279
thermorheologically simple viscoelastic
using
materials 259
coordinate systems 201
thin elastic layer (node) 540
predefined variables 86
thin elastic layer, solid mechanics 160
spatial and material coordinates 200
thin elastic layer, theory 380
weak constraints 87
thin-film damping (node) 548
uspring variable 160
three-parameter model 258
V Varga material 274
torional shear stress 783
variables
torsion, beam cross sections 779
cross section data, beams 764
torsional constants and moments 703
deformation gradient tensor 206
total Lagrangian formulation 20
density 205
total loads 77
effective creep strain 327
t-profile sections, beams 725
effective plastic strain 283, 304
tractions 130
elastic Green-Lagrange tensor 215
transient creep 323
elastic right Cauchy-Green tensor 215
translation, rigid body 349
elastic volumetric strain 321
Tresca effective stress variable 285
elastic, inelastic, and total volume ratio
Tresca stress 284
215
Tresca yield criterion 284
elcontact 393
triaxial conditions 313
Green-Lagrange tensor 208
triaxial data 315
invariants 279
TRS material 260
isochoric elastic Green-Lagrange
true stress tensor 130
strain tensor 268
truss interface 800
isochoric-elastic Cauchy-Green defor-
trusses
mation tensor 267
harmonic perturbation 581, 669, 758
isochoric-elastic right Cauchy-Green
initial stresses and strains 808
deformation tensor 267

886 | I N D E X
 material and spatial coordinates 201 fluid structure interaction 831
predefined 86 warping function 779
principal elastic stretches 279 warping, beam cross sections 781
refpnt 424 wave speeds 434, 683, 732
right Cauchy-Green deformation ten- waveguide 39
sor 208 weak constraints, using 87
right stretch tensor 206 websites, COMSOL 29
rotation tensor 206 Weertman creep 331
stresses, beams 766 Willam-Warnke criterion 313
thermal stretch and thermal volume WLF shift functions 260, 452
ratio 265
Y yield function 283
Tresca effective stress 285
yield functions 289
volumetric plastic strain 283, 319
yield stress levels 456, 463
vdamper variable 161
yield surface 308
Vicat softening temperature 261
yield surfaces 281, 291
Villari effect 341
Young’s modulus expression 237, 433,
viscoelastic materials
536
definition 93
Z Zener model 258
frequency domain analysis and damp-
ing 261
temperature effects 259
theory 253
viscoelasticity (node) 451
viscoplasticity (node) 462, 467
viscous damping 125, 161
Voigt form 98
Voigt notation 239, 490
volume-preserving contributions 266
volumetric contributions 266
volumetric creep strain rate 326
volumetric plastic strain rate 283
volumetric plastic strain variable 319
volumetric strain energy 280
volumetric strain energy density 268
von Mises criterion 283–284
von Mises effective stress 784
von Mises stress 283

W wall (node)

INDEX| 887
888 | I N D E X