Collapse Advanced

Version 15.1

Trademark Notice
Bentley and the "B" Bentley logo are either registered or unregistered trademarks or service
marks of Bentley Systems, Incorporated. All other marks are the property of their respective
owners.

Copyright Notice
Copyright ©2021, Bentley Systems, Incorporated. All Rights Reserved.
 TABLE OF CONTENTS
1 Introduction .................................................................................................................................... 9
1.1 Overview.................................................................................................................................. 9
1.2 Program Features..................................................................................................................... 9
2 Collapse Advanced Program .......................................................................................................... 11
2.1 Beam Element........................................................................................................................ 11
2.2 Plate Elements ....................................................................................................................... 17
2.3 Tubular Connections .............................................................................................................. 18
2.4 Element Distributed Loading .................................................................................................. 18
2.5 Foundations ........................................................................................................................... 18
2.6 Solution Techniques ............................................................................................................... 19
2.7 Analysis Considerations .......................................................................................................... 23
2.7.1 Progressive Collapse Analysis.......................................................................................... 23
2.7.2 Ship Impact Analysis ....................................................................................................... 24
3 Collapse Modelling and Input ........................................................................................................ 26
3.1 Analysis Options ..................................................................................................................... 27
3.1.1 Modelling Options: CLPOPT ............................................................................................ 27
3.1.1.1 Member Local Buckling ........................................................................................... 27
3.1.1.2 Joint Flexibility ........................................................................................................ 27
3.1.1.3 Tubular Connection Capacity Check ........................................................................ 27
3.1.1.4 Strain Hardening ..................................................................................................... 28
3.1.1.5 All Members or Plates Elastic .................................................................................. 28
3.1.1.6 Considering Skipped Elements Plastically ................................................................ 28
3.1.1.7 Pile Plasticity........................................................................................................... 28
3.1.1.8 Collapse Critical Displacement ................................................................................ 28
3.1.1.9 Creating a SACS Model File at Final Step ................................................................. 28
3.1.2 Solver and Convergence Options: CLPOPT ...................................................................... 28
3.1.2.1 Number of Member Sub-segments (finite elements)............................................... 28
3.1.2.2 Continue if Maximum Number of Iterations Exceeded ............................................ 29
3.1.2.3 Sub-incrementation Solver...................................................................................... 29
3.1.2.4 Arc-length Solver .................................................................................................... 29
3.1.2.5 Global Stiffness Iterations and Convergence ........................................................... 30
3.1.2.6 Member Iterations and Displacement Convergence ................................................ 30
3.1.3 Additional Member and Plate Options: CLPOP2 .............................................................. 31
3.1.3.1 Additional Member Options.................................................................................... 31
3.1.3.2 Additional Plate Options ......................................................................................... 32
3.1.4 Additional Solver and Convergence Options: CLPOP2 ..................................................... 32
3.1.4.1 Convergence Control Parameters............................................................................ 32
3.1.4.2 Relaxed Iteration and Convergence Criteria ............................................................ 32
3.2 Output Reports ...................................................................................................................... 33
3.2.1 Reports Options: CLPRPT ................................................................................................ 33
3.2.1.1 Joint Displacements ................................................................................................ 33
3.2.1.2 Selecting Joints for Displacement Report ................................................................ 33

SACS® Collapse Advanced Release 15.1

 3.2.1.3 Joint Reactions........................................................................................................ 34
3.2.1.4 Pilehead Reactions Report ...................................................................................... 34
3.2.1.5 Elements (Plates and Beam Members) Internal Loads and Stresses ......................... 34
3.2.1.6 Selecting Members for Internal Loads and Stress Report......................................... 34
3.2.1.7 Selecting Plates for Reports .................................................................................... 34
3.2.1.8 Excluding Elastic Members ...................................................................................... 34
3.2.1.9 Print Von Mises Stresses on Integration Points........................................................ 34
3.2.1.10 Collapse Summary Report ....................................................................................... 34
3.2.1.11 Member Summary Report ...................................................................................... 35
3.2.1.12 Element Warning Messages .................................................................................... 35
3.2.1.13 Von Mises Stress Check for Elastic Plates ................................................................ 35
3.2.1.14 Designating Minimum Plasticity .............................................................................. 35
3.2.1.15 Member Strain Report ............................................................................................ 35
3.3 Applying Load ........................................................................................................................ 36
3.3.1 Load Sequence: LDSEQ ................................................................................................... 36
3.3.1.1 Defining a Load Sequence ....................................................................................... 36
3.3.1.2 Load Sequences with More than Three Load Steps ................................................. 36
3.3.2 Removing a Member: MEMREM..................................................................................... 36
3.3.3 Load for Ship Impact, Dropped Object, and Blast Wall .................................................... 36
3.4 Joint Connection Options ....................................................................................................... 37
3.4.1 Tubular Connection Strength and Flexibility Parameters: ................................................ 37
3.4.1.1 Tubular Connection Options: JSOPT ........................................................................ 37
3.4.1.2 Resistance Factor Data: RSFAC/RSFACO .................................................................. 37
3.4.2 MSL Joint Flexibility and Strength Formulation: MSLOPT................................................. 37
3.4.2.1 Joint Flexibility ........................................................................................................ 38
3.4.2.2 Joint Strength ......................................................................................................... 38
3.4.2.3 Fracture Criteria...................................................................................................... 38
3.4.2.4 MSL Assessment Factor of Safety ............................................................................ 38
3.4.2.5 Plasticity in MSL ...................................................................................................... 38
3.4.3 Joint Strength/Flexibility Selection: JSSEL ........................................................................ 38
3.5 Designating Elements as Elastic .............................................................................................. 40
3.5.1 Beam Elements............................................................................................................... 40
3.5.1.1 Elastic Members: MEMELA ..................................................................................... 40
3.5.1.2 Elastic Member Groups: GRPELA............................................................................. 40
3.5.2 Plate Elements................................................................................................................ 40
3.5.2.1 Elastic Plates Elements: PLTELA ............................................................................... 40
3.5.2.2 Elastic Plate Groups: PGRELA .................................................................................. 40
3.6 Nonlinear Springs ................................................................................................................... 40
3.6.1 Nonlinear Spring Supports: NLSPRG ................................................................................ 41
3.6.2 Joint to Joint Nonlinear Springs: NLSPJJ .......................................................................... 41
3.6.3 Corotational Joint to Joint Nonlinear Springs: NLSPST ..................................................... 41
3.7 Material Properties ................................................................................................................ 42
3.7.1 Multilinear Stress-Strain Curve: ...................................................................................... 42
3.7.2 Member Ductility Limits: MEMDUC ................................................................................ 43

SACS® Collapse Advanced Release 15.1

 3.7.3 Yield Stress Overrides ..................................................................................................... 43
4 Troubleshooting ............................................................................................................................ 44
4.1 Converting Current Collapse to Collapse Advanced ................................................................ 45
4.2 Collapse Advanced Messages ................................................................................................. 47
4.2.1 Information Messages in Collapse Advanced .................................................................. 47
4.2.2 Error Messages in Collapse Advanced ............................................................................. 49
4.2.3 Warning Messages in Collapse Advanced........................................................................ 49
4.2.4 Warning Messages in Members ...................................................................................... 50
4.2.5 Warning Message in Piles ............................................................................................... 52
4.2.6 Warning Messages for Non-convergence........................................................................ 52
4.2.7 Numerical Instability Warnings ....................................................................................... 53
4.2.8 Divergence Warning Messages ....................................................................................... 54
4.2.9 Arc-length Warning Message .......................................................................................... 54
4.3 Post-buckling Analysis ............................................................................................................ 56
4.4 Force Tolerance ..................................................................................................................... 58
4.5 Non-convergence in Secondary Elements ............................................................................... 60
5 Collapse Advanced Samples .......................................................................................................... 61
5.1 Beam Element Verification and Benchmarks .......................................................................... 61
5.1.1 Sample 1: Elastic Behavior of a Restrained Simply Supported Beam Subjected to Uniform
Load 61
5.1.2 Sample 2: Elastic Buckling of Toggle Beam ...................................................................... 62
5.1.3 Sample 3: Elastic Cantilever Beam .................................................................................. 63
5.1.3.1 Sample 3.1: Buckling under Axial Force ................................................................... 63
5.1.3.2 Sample 3.2: End Lateral Force ................................................................................. 65
5.1.3.3 Sample 3.3: End Bending Moment .......................................................................... 66
5.1.4 Sample 4: Elastoplastic Cantilever with Tubular Section .................................................. 68
5.1.5 Sample 5: Elastoplastic Buckling of Columns ................................................................... 69
5.1.5.1 Sample 5.1: Elastoplastic of Tubular Simple Column with Euler-Bernoulli Theory .... 69
5.1.5.2 Sample 5.2: Elastoplastic of Tubular Simple Column with Timoshenko Bending
Theory 70
5.1.5.3 Comparison of Elastoplastic Buckling of Columns with API ...................................... 72
5.1.6 Sample 6: Buckling of Wide Flange Sections .................................................................... 73
5.1.6.1 Sample 6.1 and Sample 6.2: Lateral-Torsional Buckling of Wide Flange Columns with
Wagner Nonlinear Strain........................................................................................................... 73
5.1.6.2 Comparison of Elastoplastic Lateral-Torsional Buckling of Wide Flange under Pure
Bending with AISC 2010 ............................................................................................................ 77
5.1.7 Sample 7: Harrison’s Space Beam with Tubular Sections ................................................. 78
5.1.8 Sample 8: Elastoplastic Portal Beams .............................................................................. 79
5.1.9 Sample 9: Space Beam with Wide Flange ........................................................................ 80
5.1.10 Sample 10: Six-story Building with Wide Flange Section .................................................. 81
5.1.11 Sample 11: Buckling of Hinged Right-angle Beam............................................................ 84
5.1.12 Sample 12: Two-story 3D Beam with Rectangular Cross-Sections.................................... 86
5.1.13 Sample 13: Buckling of a Beam Dome with Rectangular Cross-section ............................ 87
5.1.14 Sample 14: Collapse Analysis of 2D Frame with Tubular Braces ....................................... 88

SACS® Collapse Advanced Release 15.1

 5.1.15 Sample 15: Four-leg Jackets ............................................................................................ 89
5.1.16 Sample 16: Tee Section ................................................................................................... 91
5.1.16.1 Sample 16.1: Elastic lateral torsional buckling of simple beam with Tee section ...... 91
5.1.16.2 Sample 16.2: Elastoplastic flexural buckling of simple column with Tee section....... 92
5.1.17 Sample 17: Channel Section............................................................................................ 94
5.1.17.1 Sample 17.1: Elastic lateral torsional buckling of simple beam with Channel section94
5.1.17.2 Sample 17.2: Elastoplastic flexural buckling of simple column with Channel section 96
5.1.18 Sample 18: Angle Section ............................................................................................... 98
5.1.18.1 Sample 18.1: Elastic lateral torsional buckling of simple beam with Angle section ... 98
5.1.18.2 Sample 18.2: Elastoplastic flexural buckling of simple column with Angle section . 100
5.1.19 Sample 19: Box Section................................................................................................. 102
5.1.19.1 Sample 19.1: Elastic lateral torsional buckling of simple beam with Box section .... 102
5.1.19.2 Sample 19.2: Elastoplastic flexural buckling of simple column with Box section .... 104
5.1.20 Sample 20: Conical Section ........................................................................................... 106
5.1.21 Sample 21: Concentric (Double) Tubular Section .......................................................... 108
5.1.22 Sample 22: Launch Runner and Special Launch Runner Section..................................... 110
5.1.23 Sample 23: Rectangular Tube ....................................................................................... 112
5.1.23.1 Sample 23.1: Elastic lateral torsional buckling of simple beam with Rectangular Tube
section 112
5.1.23.2 Sample 23.2: Elastoplastic flexural buckling of simple column with Rectangular Tube
section 114
5.1.24 Sample 24: Double Web Plate Girder Section................................................................ 116
5.1.25 Sample 25: Boxed Plate Girder Section ......................................................................... 118
5.1.26 Sample 26: Unsymmetrical Plate Girder........................................................................ 120
5.1.26.1 Sample 26.1: Elastic lateral torsional buckling of simply supported Unsymmetrical
Plate Girder ............................................................................................................................ 120
5.1.26.2 Sample 26.2: Elastoplastic flexural buckling of simple column with Unsymmetrical
Plate Girder ............................................................................................................................ 121
5.1.27 Sample 27: Double Angle .............................................................................................. 123
5.2 Plate Element Benchmarks and Samples .............................................................................. 126
5.2.1 Sample 1: Elastic Restrained Beam Subjected to Uniform Load ..................................... 126
5.2.2 Sample 2: Elastic Cantilever Beam ................................................................................ 127
5.2.2.1 Sample 2.1: End Axial Force .................................................................................. 127
5.2.2.2 Sample 2.2: End Lateral Force ............................................................................... 128
5.2.2.3 Sample 2.3: End Bending Moment ........................................................................ 129
5.2.3 Sample 3: Elastic Clamped Hinged Deep Arc ................................................................. 131
5.2.4 Sample 4: Elastic Slit Ring Shape Plate with Lifting Line Load ......................................... 133
5.2.5 Sample 5: Elastic Hemispherical Shell ........................................................................... 134
5.2.6 Sample 6: Elastic Hinged Cylindrical Roof ...................................................................... 135
5.2.7 Sample 7: Elastic Semi-Cylindrical Shell Subjected to Vertical Point Load ...................... 137
5.2.8 Sample 8: Elastic Open-ended Cylindrical Shell with Radial Pulling Forces ..................... 138
5.2.9 Sample 9: Elastic Pinched Cylindrical Shell with Rigid End Diaphragms .......................... 140
5.2.10 Sample 10: Elastoplastic Cantilever Beam ..................................................................... 142
5.2.11 Sample 11: Elastoplastic Simply-supported Plate under uniform Pressure .................... 143

SACS® Collapse Advanced Release 15.1

 5.2.12 Sample 12: Elastoplastic Pinched Cylindrical Shell with Rigid End Diaphragms .............. 144
5.2.13 Sample 13: Elastoplastic Buckling of Cylindrical Roof .................................................... 145
5.3 Joints ................................................................................................................................... 147
5.3.1 Sample 1: T and Y Joint Flexibility ................................................................................. 147
5.3.2 Sample 2: Frame with X Joint ........................................................................................ 152
5.3.3 Sample 3: Frame with K joint ........................................................................................ 154
5.4 Local Buckling ...................................................................................................................... 156
5.4.1 Sample 1: Local Buckling in Tubular Members .............................................................. 156
5.4.2 Sample 2: Local Buckling in Conical Members ............................................................... 157
5.5 Multilinear Elastoplastic Materials ....................................................................................... 159
5.5.1 Sample 1: Multilinear Elastoplastic Materials for Members .......................................... 159
5.5.2 Sample 2: Multilinear Elastoplastic Materials for Plates ................................................ 160
5.5.3 Sample 3: Multilinear Elastoplastic Materials for Piles .................................................. 161
5.6 Impact Analysis .................................................................................................................... 162
5.6.1 Sample 1: Dropped Object ............................................................................................ 162
5.6.2 Sample 2: Ship Impact .................................................................................................. 165
5.7 Dynamic Response Samples ................................................................................................. 168
5.7.1 Ship Impact .................................................................................................................. 168
5.7.2 Dropped Object ............................................................................................................ 172
5.7.3 Blast ............................................................................................................................. 176
6 Commentary ............................................................................................................................... 181
6.1 Introduction ......................................................................................................................... 181
6.2 Finite Element Formulation: Basics ....................................................................................... 182
6.3 Beam Element...................................................................................................................... 183
6.3.1 Beam Element - Corotational Approach for Geometric Nonlinearity ............................. 184
6.3.1.1 Basics.................................................................................................................... 184
6.3.1.2 Unit Quaternions and Rotation Matrix .................................................................. 184
6.3.1.3 Local Displacement, Element and Joints Rotation Matrices ................................... 185
6.3.1.4 Transformation Matrix .......................................................................................... 187
6.3.1.5 Geometric Stiffness............................................................................................... 189
6.3.2 Beam Element Local Stiffness Matrix and Load Vector .................................................. 192
6.3.2.1 Euler-Bernoulli Beam Element .............................................................................. 193
6.3.2.2 Timoshenko Beam Element................................................................................... 194
6.3.2.3 Wagner Nonlinear Torsion .................................................................................... 196
6.3.3 Beam Element Numerical Integration ........................................................................... 197
6.4 Plate Element....................................................................................................................... 198
6.4.1 Plate Element Corotational Formulation for Geometric Nonlinearity ............................ 198
6.4.1.1 Local Displacement, Element, and Joints Rotation Matrices .................................. 198
6.4.1.2 Transformation Matrix .......................................................................................... 200
6.4.1.3 Geometric Stiffness............................................................................................... 202
6.4.2 Plate Element Local Stiffness Matrix and Load Vector ................................................... 205
6.4.2.1 Kirchhoff Triangular Plate Element ........................................................................ 208
6.4.2.2 Mindlin Triangular Plate Element .......................................................................... 210
6.4.2.3 Kirchhoff Quadrilateral Plate Element ................................................................... 211

SACS® Collapse Advanced Release 15.1

 6.4.2.4 Mindlin Quadrilateral Plate Element ..................................................................... 213
6.4.3 Plate Internal Forces and Moments .............................................................................. 215
6.4.4 Plate Element Numerical Integration ............................................................................ 215
6.5 Elastoplastic Material ........................................................................................................... 216
6.5.1 Elastoplastic for Beam Element .................................................................................... 217
6.5.2 Elastoplastic for Plate Element ..................................................................................... 219
6.5.3 Numerical Samples for Elastoplastic Calculation ........................................................... 220
6.6 Arc-length Method ............................................................................................................... 223
6.6.1 Introduction ................................................................................................................. 223
6.6.2 The Cylindrical Arc-Length Method ............................................................................... 224
6.6.3 The Spherical Arc-Length Method ................................................................................. 228
6.6.4 Arc-Length Parameter Calculations and User-defined factors ....................................... 229
6.7 Element Offsets.................................................................................................................... 230
6.8 Element End-Releases .......................................................................................................... 232
6.9 Joint Strength ....................................................................................................................... 233
6.9.1 Applicability Ranges...................................................................................................... 234
6.9.2 API RP 2A-LRFD............................................................................................................. 235
6.9.3 ISO 19902 ..................................................................................................................... 237
6.9.4 NORSOK ....................................................................................................................... 240
6.9.4.1 NORSOK Revision 1 ............................................................................................... 240
6.9.4.2 NORSOK Revision 2 ............................................................................................... 242
6.9.4.3 NORSOK Revision 3 ............................................................................................... 243
6.9.5 MSL .............................................................................................................................. 244
6.9.5.1 MSL Fracture ........................................................................................................ 246
6.10 Joint Flexibility ..................................................................................................................... 247
6.10.1 Applicability Ranges...................................................................................................... 248
6.10.2 Joint Flexibility Implementation .................................................................................... 249
6.10.3 Fessler Joint Flexibility .................................................................................................. 252
6.10.4 Buitrago Joint Flexibility................................................................................................ 252
6.11 MSL Joint Flexibility .............................................................................................................. 253
6.11.1 MSL Plastic Interaction Function ................................................................................... 254
6.11.2 MSL Uncoupled Joint Flexibility .................................................................................... 254
6.11.3 MSL Joint Flexibility Coefficients ................................................................................... 255
6.11.4 MSL Coupled Joint Flexibility ........................................................................................ 256
6.11.4.1 Strain Hardening ................................................................................................... 256
6.11.4.2 Consistency Rule ................................................................................................... 257
6.11.4.3 Flow Rule .............................................................................................................. 258
6.11.4.4 Elastoplastic Stiffness Matrix................................................................................. 258
6.11.4.5 Integration............................................................................................................ 259
6.11.4.6 Unloading ............................................................................................................. 260
6.12 Local Buckling ...................................................................................................................... 261
6.12.1 Marshall and Gates ....................................................................................................... 261
6.12.2 API Bulletin 2U ............................................................................................................. 261
6.12.3 API RP 2A-LRFD............................................................................................................. 262

SACS® Collapse Advanced Release 15.1

 6.12.3.1 API RP 2A-LRFD: Conical Members ........................................................................ 262
6.12.4 ISO 19902 ..................................................................................................................... 262
6.12.4.1 ISO 19902: Conical Members ................................................................................ 263
6.12.5 Implementation............................................................................................................ 263
7 REFERENCES................................................................................................................................ 264
8 INPUT LINES ................................................................................................................................ 267

SACS® Collapse Advanced Release 15.1

1 Introduction
1.1 Overview
SACS Collapse Advanced is a large deflection, large displacement/rotation, elastoplastic, nonlinear finite
element solver for structures. Collapse Advanced provides new enhancements and improvement over
previous the Collapse module. The program is fully integrated into the SACS suite of programs and uses
the same input data as that for a standard SACS IV/PSI analysis. No new modeling is required to conduct
a full plastic collapse analysis of a structure.

Users can select Collapse Advanced as their default analysis engine by making the appropriate selection
under Settings/Analysis Settings in the SACS Executive. In addition, a run file generated by a previous
version of SACS should be updated by opening it in the Analysis Generator in the SACS Executive and
then re-saving to apply new changes.

1.2 Program Features

The Collapse program requires no special modeling and only minimal additional input specified in a
Collapse input file. Collapse Advanced currently supports following features:

1. Full Newton-Raphson nonlinear iterations with automatic sub-incrementation

2. Arc-length method for unloading and post-buckling analysis

3. Automatic load sub-incrementation to improve convergence

4. Elastic and Elastoplastic material behavior with strain hardening, bilinear and multilinear stress-
strain curves

5. General multilinear stress-strain curve for elastoplastic calculation with strain hardening or
strain softening for beam members, plates and piles

6. Distributed plasticity to model gradual plastification of beam element (both along the element
and beam cross section) and plates (including thickness)

7. Geometric nonlinearity due to large displacement and rotation for both beam elements and
plates.

8. Euler and Timoshenko bending theory for beam elements

9. Second order strains for beam elements

10. Advanced options to include lateral-torsional buckling modes in wide flanges and other open
thin-walled sections

11. Elastoplastic thin (Kirchhoff) and thick (Mindlin) plate bending theory

12. Inclusion of member and plate offsets and their effects on geometric stiffness in large rotations

13. Accounts for segmented elements automatically

14. Local buckling and ductility check for beam sub-segments

9

SACS® Collapse Advanced Release 15.1

 15. Various joint flexibility methods (both linear and nonlinear) and joint strength failure

16. Nonlinear springs

17. Sequential load stacking capability with user-controlled load incrementation, includes both
loading and unloading capabilities

18. Load cases may contain loading and/or specified displacements

19. The program creates analysis results file that is read by Collapse View program which shows
failure progression and the gradual plastification and collapse mechanism graphically

20. The program supports pile-soil-structure interaction (PSI) using user-defined T-Z, P-Y, and end-
bearing resistance curves, automatically generated resistance curves based on API
recommendation (including CPT method), axial adhesion, torsion adhesion, and soil liquefaction.

Following features are currently included in Collapse Advanced but they are not thoroughly verified:

1. Elastoplastic calculation Dented tubular with grout

2. Thermal loads in beam elements

The following features are currently under development and will be included in future releases:

1. Corotational joint to joint nonlinear spring

2. Thermal and space load for plate elements

3. Thermal degradation of plate elements

10

SACS® Collapse Advanced Release 15.1

2 Collapse Advanced Program
The basic procedure used by the Collapse Advanced program to perform a nonlinear analysis is
discussed below. The details for finite element formulation are discussed in section 6.

2.1 Beam Element

Beam element stiffness is determined using second order strains and large displacement/rotation to
account for geometric nonlinear effects assuming elastoplastic material properties. Each beam is
automatically discretized by using sub-segments along the member length. Additionally, elastoplastic
stress-strain is calculated at discrete points distributed through the beam cross-section. Next, forces and
stiffness are calculated using numerical integration for each given sub-segment.

When sub-segment stiffness and forces are determined, the beam element is treated as a superelement
whose stiffness is defined by the its sub-elements stiffness, joint flexibility, member end offsets, and end
releases (see section 2.6). While the intermediate nodes along a beam element are reduced for the
elemental stiffness matrix and member end forces, all sub-segments represent the deflected shape of
the element. Figure 1 illustrates the beam element model.

By default, non-segmented (elastic or elastoplastic) beam elements are divided into 8 sub-segments
along the length of the element while segmented beam elements are divided into sub-segments
according to changes in the cross-section. Collapse Advanced uses a single sub-segment for linear
members or for members with aspect ratio (height/length) greater than one - i.e. very short members.
The default number of sub-segments can be changed between 1 and 30 on the CLPOPT input line. In
general, a minimum of 4 is recommended.

Figure 1: Collapse Advanced Beam Element

11

SACS® Collapse Advanced Release 15.1

 Table 1: Beam cross-sections and integration points

Section Section partitions and Gauss–Legendre integration points per each part Total (𝑁𝑝 )
Tubular, cone Parts 1st quarter 2nd quarter 3rd quarter 4th quarter 16
Concentric
Tubular (each Points 4 4 4 4 (32)
tubular)
Parts top flange bottom flange Web fillets
Wide Flange 29
Points 5 for each half (2x5) 5 for each half (2x5) 5 1 for each fillet (4x1)
Parts along local y along local z
Rectangular 25
Points 5 5
Parts flange left half flange right half Web
Tee 14
Points 4 4 6
Parts 1st side 2nd side
Angle 12
Points 6 6
Parts top flange bottom flange Web fillets
Channel 18
Points 5 5 6 1 for each fillet (2x1)
Parts 1st side 2nd side 3rd side 4th side
Box 12
Points 3 3 3 3
Parts 1st quarter 2nd quarter 3rd quarter 4th quarter Solid
Jack-up Leg 16 32
Points 4 4 4 4
(4x4)
Launch Runner Parts Tubular quarters Two side plates Bottom plate
and Special 28
Launch Runner Points 4 (each) 4 (each) 4

Rectangular Parts 1st quarter 2nd quarter 3rd quarter 4th quarter
24
Tube Points 2 for arc 4 for flat 2 for arc 4 for flat 2 for arc 4 for flat 2 for arc 4 for flat

Unsymmetrical Parts top flange bottom flange Web

25
Plate Girder Points 5 for each half (2x5) 5 for each half (2x5) 5

Boxed Plate Parts top/bottom plates right/left plates web

30
Girder Points 6 (each) 6 (each) 6

Double Web Parts right side of flanges middle of flanges left side of flanges webs
36
Plate Girder Points 3 (each) 6 (each) 3 (each) 6 (each)
Parts horizontal sides vertical sides
Double Angles 24
Points 6 (each) 6 (each)

Dented Tubular Parts top circular part bottom circular part top dented part bottom dented part grout
27
with grout Points 6 6 3 3 9

12

SACS® Collapse Advanced Release 15.1

Collapse Advanced monitors stress and strain history at multiple points (called integration points) along
the element and within the cross-section. Two sets of integration points are used along each sub-
segment located at Gauss–Legendre quadrature integration points (see section 6.3.3 for more details).
Each set contains various integration points to model stress-strain distributed across the beam cross-
section. The number of integration points through a cross-section of a beam element are dependent on
the cross-section shape of the beam element. For example, 16 points are used for a tubular section (4
per each quarter) and 29 points are used for a wide flange: 10 for each flange, 5 on the web and 1 for
each of the 4 fillets. Therefore, for a given tubular beam element with 8 sub-segments, elastoplastic
calculations are carried out at total number of 256 points (8x2x16). Similarly, for a wide flange beam
element, elastoplastic calculations are carried out at 464 (8x2x29) points. Table 1 shows the cross-
section parts for various cross section types and the corresponding number of integration points in each
part. Figure 2 illustrates location of integration points for various section types.

At a given iteration, each point is checked for plasticity using a von Mises stress surface (i.e. J2
plasticity). The internal member strains are based on second order strain theory and the stress-strain
properties are defined by a bilinear model with user defined strain hardening. Elastoplastic nonlinear
iterative calculations are performed using the Backward Euler method [1]. When the stresses at a point
exceed the material elastic limit, the point is considered to be in a plastic state, thus allowing for gradual
plastification of the beam cross section. For more details on elastoplastic calculations, please refer to
section 6.5.

Local tubular buckling is determined from the total strain in the cross section and is included as a
permanent hinge at the point of occurrence in the model. Fracture (ductility limit) is checked at each
integration points to model a gradual fracture. If the strain exceeds the ductility limit at all integration
points, the beam is assumed to be fractured.

Collapse Advanced supports both Euler-Bernoulli (without shear deformation) and Timoshenko (with
shear deformation) beam bending theories. Elastoplastic calculations based upon Euler-Bernoulli theory
are based upon normal stresses only while Timoshenko formulation utilizes both normal stresses and
shear stresses.
NOTE: Elastoplastic calculations based upon Timoshenko’s approach assume thin-walled beam sections.

The geometrical effects (geometric stiffness and loads) due to the large rotation and displacement are
modeled through a corotational [2] approach. The corotational approach updates the element local
coordinate systems gradually during the nonlinear iterative process and enables beam elements to
undergo very large rotations and displacements. Hence, the beam deflected shape is accurately
calculated at the member ends and along its length at each sub-segment. Member elastic and plastic
buckling is automatically calculated using the beam deflected shape and the plasticity of the member
sub-segments.

13

SACS® Collapse Advanced Release 15.1

The following sections are supported for full elastoplastic plastic behavior including second order strains
and corotational large rotation/displacement:
• Tubular
• Wide Flange
• Prismatic (rectangular) – (currently only Euler-Bernoulli bending theory)
• Tee
• Angle
• Channel
• Box
• Cone
• Jack-up Leg – (currently only Euler-Bernoulli bending theory)
• Concentric tubular
• Launch Runner and Special Launch Runner
• Rectangular Tube
• Unsymmetrical Plate Girder
• Boxed Plate Girder
• Double Web Plate Girder
• Double Angles
• Dented tubular with grout (experimental)
If the section properties (e.g. cross section area, inertia, etc.) for the above section types does not match
the calculated values based upon cross section dimensions, then the member is treated as elastic,
however, second order strains and geometric corotational effects will be included. Other SACS section
types not included in the above list are also treated as elastic but without second order strains or
geometric corotational affects.
Details of beam element formulation can be found in section 6.3.

14

SACS® Collapse Advanced Release 15.1

 Figure 2: Integration points for various beam sections

15

SACS® Collapse Advanced Release 15.1

 Figure 2 : Continue.

16

SACS® Collapse Advanced Release 15.1

2.2 Plate Elements
Collapse Advanced considers isotropic (6 degrees of freedom) plate elements as fully elastoplastic with
large displacement and rotations. The stress-strain history is monitored at discrete integration points
within the plate surface and through the plate thickness. For triangular plate, the elastoplastic
calculations are performed at three integration points across the plate and 6 integration points through
the thickness – a total of 18 integration points. For quadrilateral plates, there are a total of 24
integration points (4 across the plate and 6 through the thickness). This detailed analysis allows Collapse
to accurately model gradual plastification of the plates both across the plate and through its thickness.
Figure 3 illustrates the location of integration points for triangular and rectangular plate elements.

Thickness
Plate

Figure 3: Integration points for plate elements

Elastoplastic calculations for plates are performed using the Backward Euler method with a J2 yield
surface. Collapse Advanced supports both thin (Kirchhoff) and thick (Mindlin) plate theories. For thin
plates two normal stress and the in-plane shear stress components are considered. Thick plate bending
theory accounts for out-of-plane shear stresses in addition to the normal stress and the shear stress
components – a total of five stress components. By default, thin plate theory is assumed. Thick bending
theory can be selected by making the appropriate selection on the CLPOP2 input line.

Large displacement and rotation, geometric nonlinearities, are modeled using the corotational
formulation [1, 2]. The corotational formulation also accounts for the effects of plate offsets on both
geometric and elemental stiffness matrices.

Currently, only isotropic plates can be modelled as fully elastoplastic. Other SACS plate element types
such as membrane, shear only, corrugated, and stiffened plate elements are considered as elastic
however, corotational large displacement/rotation effects are still included for these plate elements.

Details of plate element formulation can be found in section 6.4.

17

SACS® Collapse Advanced Release 15.1

2.3 Tubular Connections
Tubular joint flexibility options include:

1) Fessler's empirical formulas [3, 4]

2) Single brace formulation due to Buitrago [5]

3) MSL formulation [6].

Tubular connection failure criterion can be introduced via:

1) Modified ultimate LRFD strength formulation as per API code of practice [7]

2) Joint strength check as per NORSOK code of practice [8]

3) MSL formulation [6].

The brace stiffness is automatically removed from the analysis at the onset of a connection failure.

2.4 Element Distributed Loading

Applied distributed loads are converted to element nodal loads by integrating over the sub-segments of
a member or the plate element. Changes to the sub-segment and plate positions resulting from the
deformation of the structure are also accounted for, however, during deformation, the direction of all
member loads are kept constant with respect to the global coordinate system, this is also true for
member loads which are defined in the member local coordinate system. This type of load is also
referred to as a conservative load and as such may produce axial load effects during deformation of the
structure. As an example, consider a horizontal member with a distributed load along global Z axis. At
zero load step, the applied load is perpendicular to all sub-segments and it does not produce any
internal axial load in the member. However, while the analysis is progressing, the orientation of sub-
segments will change due to deformation and the applied member load, the direction of which is kept
constant with respect to the global coordinate system, will result in axial load effects in the member.

Similarly, for plate elements, the direction of joint loads and hydrostatic pressure loads is always kept
constant with respect to the global coordinate system. The direction of plate pressure loads (constant or
varying between joints) is always kept constant with respect to the element local coordinates system.

Note: Currently thermal loads are only supported for beam elements.

Note: To reduce computational time, element loads are only updated at beginning of each load
increment – instead of each iteration.

2.5 Foundations
The collapse solution includes the effects of a nonlinear pile/soil foundation. Tubular pile elements are
treated the same as regular tubular members and are segmented along the length and around the
circumference. Soil data is represented with standard T-Z and P-Y data in the PSI format.

18

SACS® Collapse Advanced Release 15.1

2.6 Solution Techniques
The nonlinear solver for Collapse Advanced is based on a two-step iteration scheme. The first step
involves the solution of all beam elements, piles, and plates for a given initial state and known
incremental nodal displacements and the formation of a tangent stiffness matrix. The second step
involves the solution of the entire structure using the current tangent stiffness matrix and an updated
incremental displacement vector.

For plates, the stiffness and forces are determined via a nonlinear elastoplastic finite element approach
with the inclusion of a co-rotational formulation. In comparison, the solution for a beam element is
computationally much more intensive due to the beam element internal sub-segmentation and the
inclusion of member end-releases and joint flexibility (see Figure 1). Collapse Advanced treats each
element as a small-scale nonlinear problem in which the displacements at member ends are known and
the solution is obtained from an iterative process involving the convergence of internal forces. A similar
nonlinear solution procedure is used for the foundation piles but with the following differences: 1) Only
pilehead displacements are given (instead of both member ends), 2) The number of sub-segments for
each pile are based on the PSI input file, 3) Pile sub-segments are embedded into nonlinear elastic
springs associated with soil P-Y and T-Z curves, and 4) Collapse Advanced only solves for the pilehead
stiffness matrix and force (instead of both member ends).

The Collapse Advanced nonlinear solver is based on a full Newton-Raphson iterative approach in which
the load increment size can be automatically reduced to improve the convergence rate via a sub-
incrementation scheme. The sub-incrementation scheme gradually reduces a given load increment by
factor of 2 until convergence is achieved or the incrementation limit is reached. Once convergence is
achieved for a few consecutive load increments using the reduced load increment, the sub
incrementation scheme will then start to gradually increase the load increment. In the case of buckling,
the Arc-length [1, 2] method is used to predict unloading behavior during the post-buckling response.
The workflow is summarized in Figure 4. Details of formulation can be also found in section 6.6.

19

SACS® Collapse Advanced Release 15.1

For a given load increment repeat following steps until convergence:
Step 1: Elements
Plate Elements: For all plates elements
i. Offset (if any): Adjust plate nodal displacements
ii. Given initial and current displacement vector, update plate orientation and corotational
transformation matrix
iii. Calculate local stiffness matrix and force vector through elastoplastic finite element
formulation
iv. Calculate geometric stiffness, global stiffness, and force vector
v. Offset (if any): Adjust global stiffness matrix and force vector
vi. Insert element stiffness matrix into model stiffness and update model residual vector
Beam Elements: For all beam members
Members without end-release or joint flexibility:
i. Offsets (if any): Adjust member-end displacements (see section 6.7)
ii. Solve for member internal nodes for given displacement
Step 1: For all sub-segments (i.e. finite elements)
• Form co-rotational transformation matrix
• Determine local stiffness matrix including elastoplastic effects, fracture
(ductility), and local buckling
• Form global and geometric stiffness matrix
Step 2: Assemble element stiffness matrices and iteratively solve for internal nodal
displacement
Step 3: Check for convergence
• NO: go to Step 1
• YES: Condense out all internal nodes from stiffness matrix and form 12x12
member stiffness matrix and 12x1 load vector
iii. Offset (if any): Adjust 12x12 global stiffness matrix and 12x1 force vector
iv. Insert element stiffness matrix into model stiffness and update model residual vector

Figure 4: Collapse Advanced nonlinear solver work flow

20

SACS® Collapse Advanced Release 15.1

 Member with end-release:
i. Offsets (if any): Adjust member-end displacements (see section 6.7)
ii. Solve for member internal nodes for given displacement considering special sub-segments
for end-releases (see section 6.8)
Step 1: For all sub-segments
• Form co-rotational transformation matrix
• Determine local stiffness matrix including elastoplastic effects, fracture
(ductility), and local buckling
• Form global and geometric stiffness matrix
Step 2: Assemble element stiffness matrices and iteratively solve for internal nodal
displacement
Step 3: Check for convergence
• NO: go to Step 1
• YES: Condense out all internal nodes from stiffness matrix and form member
stiffness matrix and load vector
iii. Offset (if any): Adjust 12x12 global stiffness matrix and 12x1 force vector
iv. Insert element stiffness matrix into model stiffness and update model residual vector
Member with joint flexibility:
i. Offsets (if any): Adjust member-end displacements (see section 6.7)
ii. Repeat following steps until convergence
a. Update joint flexibility stiffness parameter and adjust member-end displacements
b. Solve for member internal nodes for given displacements
Step 1: For all sub-segments
• Form co-rotational transformation matrix
• Determine local stiffness matrix including elastoplastic effects, fracture
(ductility), and local buckling
• Form global and geometric stiffness matrix
Step 2: Assemble element stiffness matrices and iteratively solve for internal nodal
displacement
Step 3: Check for convergence
• NO: go to Step 1
• YES: Condense out all internal nodes from stiffness matrix and form 12x12
member stiffness matrix and 12x1 load vector
c. Check for convergence: NO (go to Step a) YES (go to step iii)
iii. Offset (if any): Adjust 12x12 global stiffness matrix and 12x1 force vector
iv. Insert stiffness matrix into model stiffness and update model residual vector

Figure 4: Continue.

21

SACS® Collapse Advanced Release 15.1

 Foundation: For all piles
i. Solve for pile internal nodes for given pilehead displacement considering nonlinear elastic
stiffness of soil medium (i.e. nonlinear elastic springs based on P-Y and T-Z curves)
Step 1: For all Pile sub-segments (i.e. finite elements)
• Determine local stiffness matrix including elastoplastic effects
• Form global stiffness matrix for pile sub-segment
Step 2: Assemble element stiffness matrices and iteratively solve for internal nodal
displacements
Step 3: Check for convergence
• NO: go to Step 1
• YES: Condense out all internal nodes from stiffness matrix and form 6x6
pilehead stiffness matrix and 6x1 pilehead load vector
ii. Insert pilehead stiffness matrix into model stiffness and update residual vector

Step 2: Entire Model

Step 1 Newton Iteration: Solve a Newton iteration for new displacement increment using
updated tangent stiffness and residual vector determined in Scale 1.
Check for convergence:
a. NO: go to Scale 1 and repeat. If maximum number of iterations have been reached
then go to Step 2 Newton with sub-incrementation.
b. YES: Perform the required checks such as joint strength checks, report the results and
go to next load increment.
Step 2 Newton with sub-incrementation: Solve a Newton iteration with reduced load
increment (sub-incrementation). Check for convergence:
a. NO: go to Scale 1 and repeat with reduced load increment. If maximum number of
iterations and sub-incrementation have been reached then go to Step 3 Arc-length.
b. YES: Perform required checks such as joint strength check, report the results and go to
next the load increment. If the last few reduced load steps have converged using the
reduced load increment then start increasing the load increment.
Step 3 Arc-length: Solve using Arc-length method with given value for arc length. Check for
convergence:
a. NO: go to Scale 1 and repeat with reduced load increment. If maximum number of
iterations reduce arc-length value. If arc-length value is too small, stop analysis.
YES: Perform required checks such as joint strength check, report the results and go to
next load increment. If everything is going well, switch back to Newton iteration.

Figure 4: Continue.

22

SACS® Collapse Advanced Release 15.1

2.7 Analysis Considerations
The Collapse module is capable of handling most structural problems where plasticity may occur
through large deflections. Some obvious applications include progressive analysis (pushover analysis),
ship impact, dropped object, blast analysis, and general safety case studies. Basic considerations in
conducting such analyses are outlined below.

2.7.1 Progressive Collapse Analysis

The 'Plastic Collapse' mode of assessment offers an improved design concept over linear ‘Elastic’ theory
for the analysis/re-analysis of structures. The basic concept of a plastic collapse analysis is as follows:

The load is applied to the structure incrementally. The nodal displacements and element forces are
calculated for each load step and the stiffness matrix is updated. When the stress in a member reaches
the yield stress, plasticity is introduced. The introduction of plasticity reduces the stiffness of the
structure and additional loads due to subsequent load increments will be redistributed to members
adjacent to the those that have gone plastic. This phenomenon (progressive collapse of members) will
continue until the structure is no longer able to sustain any increase in loading.

For large offshore structures, the analysis can be highly CPU intensive since each element is subdivided
into eight sub segments to allow for a development of a plastic hinge anywhere along the member
length, for tubular elements, elastoplastic stresses are calculated at 16 integration points across the
cross section to allow for a gradual development of a plastic hinge throughout the cross section.
Collapse run time can be decreased by modeling parts of the structure which have little or no
contribution to the overall stiffness of the structure (such as boat landings) as dummy structures. All
elements contained in a dummy structure are removed by the Seastate module and the loads on the
dummy structure are transferred to the main structure before the Collapse analysis is initiated.

Elements whose stiffness may be of significance to the overall behavior of the structure, but which are
not structurally important (such as conductors and conductor guides, wishbone elements, topsides
elements ...etc.) should be kept elastic throughout the loading history.

Further reductions in run time can be achieved by pre-combining loads wherever possible to minimize
the number of loads in a load sequence. Also, a structure undergoing a high level of nonlinear behavior
can require an increasing number of iterations for the solution to converge. In such cases it is better to
reduce the step size than to increase the maximum iteration limit. Reducing the step size effectively
linearizes the problem which decreases the required number of iterations and therefore decreases the
runtime. A sub incrementation scheme has been implemented in Collapse to automatically reduce the
step size in areas of high non- linearity where convergence becomes an issue.

23

SACS® Collapse Advanced Release 15.1

2.7.2 Ship Impact Analysis
A ship impact scenario involves transference of a ships kinetic energy into strain energy resulting from:

a. Local deformation of the impacted member due to denting and beam bending.

b. Global deformation of the entire structure.

c. Deformation of the ship structure.

Local deformation of the impacted member due to beam bending and the global deformation of the
structure is readily accounted for by Collapse. To account for localized denting it is recommended that
the impacted member is modeled using isotropic plate elements. The SACS module Precede has the
facility to generate a tubular finite element plate mesh for a given member. Alternatively, the local
denting energy of an impacted member may also be considered in accordance with the Ellinas or Furnes
approaches outlined in the API RP2A-WSD code of practice by selecting the appropriate option on the
IMPACT input line. The maximum dent and maximum energy absorbed by the member can be limited by
options provided on IMPACT input line. Ellinas and Furnes force and energy formula are show in Figure
5.

NOTE the latter approach does not account for any geometric nonlinearities resulting from local
indentations.

A joint force, together with the total kinetic energy or the mass and velocity of the impacting object, can
be used to simulate an impact via the IMPACT input line. Collapse also allows for automatic unloading
for post impact analysis – for more details see section 5.6 for IMPACT samples.

The SHPIND input line may be used to input user defined ship indentation curves. In addition, the
IMPACT input line allows the user to select DNV RP-C204 force displacement curves for a 5000-ton ship
and a 1.5m and 10m diameter infinitely stiff cylindrical column similar to those shown in Figure 6.
Collapse Advanced assumes further energy is not absorbed by the ship once the maximum ship force
has been exceeded.

Collapse Advanced also provides a new feature to assume member and ship absorbed energies as plastic
(unrecoverable) energy. With this new option, the ship and member dent and energies remain
unchanged during unloading – i.e. plastic dent deformation.

Finally, Collapse View can be used to produce reports and plots of the energy absorbed by the structure,
the member local dent, and the ship

24

SACS® Collapse Advanced Release 15.1

 Figure 5: API RP2A-WSD Ellinas and Furnes Member Indentation Curves

Figure 6: DNV Force – Indentation Curves for a 5000-ton Ship

25

SACS® Collapse Advanced Release 15.1

3 Collapse Modelling and Input
The Collapse program requires a SACS model file and a Collapse input file. A standard SACS model may
be used as the model input for the nonlinear analysis without any change. the nonlinear plastic analysis
options are defined in a Collapse input file. Collapse Advanced input are divided into six groups:

1) Analysis Options: Collapse Advanced options are specified in two input lines CLPOPT line (primary)
and CLPOP2 line (secondary and optional). Additional collapse analysis options on CLPOP2 are
typically used for refining the analysis parameters or troubleshooting analyses which doesn’t
converge. There are also three additional input lines associated with different features in Collapse
Advanced including: FRCTOL (force tolerance for convergence), SUBINC (Newton iteration with sub-
incrementation) ARCLEN (arc-length method for post-buckling and unloading iteration). See section
3.1 for details.
2) Report Options: Output report options may be specified on the CLPRPT line. Also, input lines JTSEL,
MEMSEL and PLTSEL can be used to print report for specific joints, beam elements and plates,
respectively. See section 3.2 for details.
3) Loading: load sequence can be defined on LDSEQ input line. See section 3.3 for details.
4) Joint Connection Options: Options for joints (flexibility and strength) can be entered on JSOPT.
5) Elastic Elements: Some elements in offshore structures which may contribute to overall stiffness of
the model but may not be structurally significant (such as conductors and conductor guides, wish
bone elements, etc.). These elements can be kept as elastic to reduce computational time for a
large-scale collapse analysis. Collapse Advanced provides various option to assign specific elements
or group of elements as elastic.
6) Nonlinear Springs: there are two types of nonlinear springs in collapse input: 1) joint to ground
(support) and 2) joint to joint. The springs specifications may be entered on NLSPRG NLSPJJ input
lines.
7) Material Properties: Collapse input has various options to override yield stress for member and
plates. Also, ductility can be overridden for specific members using MEMDUC input line.

Following sections discuss specific inputs.

26

SACS® Collapse Advanced Release 15.1

3.1 Analysis Options
3.1.1 Modelling Options: CLPOPT
3.1.1.1 Member Local Buckling
Local buckling of the member cross section may be considered by specifying analysis option ‘LB’ in one
of the analysis options fields. The criteria used for local buckling is specified on columns 52-53 as ‘MG’
for Marshall & Gates [9] lower limit of critical strain, ‘2U’ for API Bulletin 2U recommendations, ‘LR’ for
API ultimate strength code criteria and ‘IS’ for ISO 19902 Section 12.2.3.3. A moment free hinge is
introduced along the member at the onset of local buckling.

Local buckling can be skipped for given members and groups via the MEMSKP and GRPSKP input lines
respectively.

3.1.1.2 Joint Flexibility

The effects of tubular connection flexibility may be accounted for by specifying the analysis option ‘JF’ in
columns 34-35 to use Fessler's empirical formulas for joint flexibility. Single brace formulation by
Buitrago can be selected for joint flexibility by entering ‘JB’ in columns 34-35.

Formulation for connection flexibility developed by MSL Engineering Limited (UK) for JIP project
‘Assessment Criteria, Reliability and Reserve Strength of Tubular Joints’ can be accessed by entering ‘JF’
on the CLPOPT line and adding MSLOPT input line to the collapse input line. The formulation can be
specified with analysis option ‘MF’ for mean level or ‘CF’ for characteristic level on the input line
MSLOPT in columns 8-9. By default, the program uses elastoplastic (coupled) MSL joint flexibility
formulation. The MSL flexibility can be set to uncoupled formulation by entering ‘NMP’ on column 78-80
of MSLOPT.

3.1.1.3 Tubular Connection Capacity Check

Joint strength check based upon API RP 2A-LRFD recommendations for tubular joints can be
implemented by specifying ‘JS’ in one of the analysis options field between columns 26-41. Entering ‘ND’
implements joint strength check in accordance to NORSOK-revision 2 standard for the design of steel
structures. Similarly, joint strength check in accordance NORSOK-revisions 1 and 3 are available by
entering ‘N1’, and ‘N3’, accordingly.

Joint Capacity check in accordance to ISO 19902 can be implemented by specifying ‘IS’ in columns 26-41.

Formulation for capacity check developed by MSL Engineering Limited (UK) for JIP ‘Assessment Criteria,
Reliability and Reserve Strength of Tubular Joints’ includes mean level and characteristic level options
specified with analysis option ‘MS’ or ‘CS’, respectively, in columns 10-11 on the MSLOPT line.

Once the joint strength check criterion has been exceeded the connection is considered to have failed
and the brace connected sub-segment stiffness is progressively reduced.

Note: Since coupled/elastoplastic MSL Joint flexibly implicitly assumes connection softening, the brace
sub-segment softening is not considered when elastoplastic MSL flexibility is implemented.

27

SACS® Collapse Advanced Release 15.1

3.1.1.4 Strain Hardening
Strain hardening effects can be included at the onset of plasticity by entering a strain hardening ratio in
columns 76-80. The strain hardening ratio is defined as the ratio of the slope of the plastic portion of the
stress-strain curve to the slope of the elastic portion. The default value for strain hardening is 0.002.

3.1.1.5 All Members or Plates Elastic

All members can be treated as elastic by entering ‘ME’ in columns 40-41. Similarly, all plates can be
treated as elastic plates by entering ‘PE’.

3.1.1.6 Considering Skipped Elements Plastically

By default, any element or element group designated in the model file to be skipped for post processing
purposes is considered as an elastic throughout the analysis. Skipped elements may be considered to
have plastic material properties by specifying the analysis option ‘NS’ in columns 26-27.

Note: Skipped beam elements are designated in the model file by ‘SK’ in columns 20-21 on the MEMBER
line defining the member or by specifying member class ‘9’ in column 47 on the GRUP line defining the
group to which it is assigned. Skipped plates are designated by ‘SK’ in columns 31-32 on the PLATE line
defining it.
3.1.1.7 Pile Plasticity

When executing a nonlinear plastic analysis including the pile/soil foundation, the pile elements material
properties may be treated as elastic or plastic. Enter ‘PP’ in columns 36-37 to use plastic material
properties for pile elements.

3.1.1.8 Collapse Critical Displacement

The critical displacement or the maximum deflection allowed before the structure is assumed to have
collapsed or has failed may be specified in columns 71-75.

3.1.1.9 Creating a SACS Model File at Final Step

A SACS model file with joint coordinates that reflect the final displaced position of the joint may be
created by inputting ‘SF’ in columns 38-39.

Note: The user may generate a deformed SACS model for any given load increment in Collapse View.

3.1.2 Solver and Convergence Options: CLPOPT

3.1.2.1 Number of Member Sub-segments (finite elements)
By default, members with plastic material properties are divided into 8 sub-segments along the member
length, linear elements or very short members are modeled using as single element. The number of sub-
segments for members may be specified in columns 14-16. A minimum of 4 sub-segments is
recommended for accurate results.

Additional options for member sub-segmentation are also available on the MEMSEG and GRPSEG input
lines. The MEMSEG input line can be used to override the member sub-segment definition on the
CLPOPT line for a specific member by specifying the number of sub-segments and the start and end

28

SACS® Collapse Advanced Release 15.1

joints of a member. As many MEMSEG lines as required may be specified. Similarly, the GRPSEG input
line can be used to override the sub-segment definition for a group of members by specifying the
number of sub-segments and the group ID. As many GRPSEG lines as required may be specified.

Note: The sub-segment length is determined by dividing the total member length by the maximum
number of sub-segments designated. For segmented members, any sub-segment which has a change in
property is further divided into two constant property sub-segments at the point at which the section
property changes. Therefore, segmented members may have more sub-segments than the maximum
specified.

3.1.2.2 Continue if Maximum Number of Iterations Exceeded

Collapse Advanced terminates when the maximum number of iterations is exceeded. Continue ‘CN’
option is no longer supported in Collapse Advanced. It is recommended to use sub-incrementation (‘SI’
in columns 44-45), Arc-length method (‘AL’ in columns 42-43), relaxed convergence (see CLPOP2). For
more details, see Troubleshooting section.

3.1.2.3 Sub-incrementation Solver

Sub-incrementation may be used to improve convergence rate by specifying ‘SI’ in columns 44-45 for
Newton-Raphson solver. The sub-incrementation solver automatically reduces load increment by a
factor of 2 (up to a specified level) until convergence is achieved. If convergence is achieved in few
consecutive increments, the solver will gradually increase (accelerate) the load increment.

The sub-incrementation options may be entered on the SUBINC line. The default values of 5 and 4 for
the maximum number of sub-incrementation levels and the maximum acceleration of the sub-
incrementation may be overridden in columns 8-9 and columns 11-12, respectively.

Section 4.2.6 discusses the details how to utilize Collapse Advanced sub-incrementation feature in
nonlinear analysis.

3.1.2.4 Arc-length Solver

The arc-length solver may be used for post-buckling analysis by specifying the analysis option ‘AL’ in
columns 42-43.

The arc-length solver options may be entered on the ARCLEN line. Enter the type of arc-length iteration
in columns 8-10. There are two types of arc-length iteration methods available in Collapse Advanced,
cylindrical (CYL) and spherical (SPH). The cylindrical method uses deflection increments to determine the
arc-length parameter while the spherical arc-length method utilizes a factor (columns 31-36) to combine
both force and deflection increment. Automatic arc-length parameter can be scaled by a factor in
columns 24-29 (the default value is 1.0) – i.e. if the arc-length parameter is too large, a value smaller
than 1 can be entered or vice versa.

29

SACS® Collapse Advanced Release 15.1

Arc-length method also utilizes a sub-incrementation scheme if it is required. If this option is left blank
(i.e. default), Collapse Advanced will attempt to solve using the following order:

1. Cylindrical method
2. Cylindrical method with sub-incrementation
3. Spherical method
4. Spherical method with sub-incrementation
5. Cylindrical method with increasing arc-length

By default, the maximum number of arc-length steps is 500 and may be overridden in columns 12-18.
The default maximum number of 10 sub-incrementation level may be overridden in columns 20-22.

Details of Arc-length formulation can be found in section 6.6. Also, section 4.2.9 discusses how to utilize
Collapse Advanced Arc-length method for post-buckling.

3.1.2.5 Global Stiffness Iterations and Convergence

For any load increment, a beam-column solution is performed for each plastic member using the cross
section sub-element details. The global stiffness iteration is then performed which can include the
effects of connection flexibility and nonlinear pile/soil foundation effects. The deflected shape of the
structure is then determined and compared against the displacements of the previous global stiffness
iteration. The stiffness iterations are repeated until the displacements and rotations satisfy the
displacement and rotation convergence tolerances or the maximum number of iterations has been met.

By default, the maximum number of global stiffness iterations per load increment is 20 but may be
overridden in columns 11-13 of the CLPOPT input line. The default displacement and rotation
convergence tolerances are 0.01 inch or 0.01cm and 0.001 radians and may be overridden in columns
56-60 and 61-65 of the CLPOPT input line respectively.

In addition to displacement and rotation tolerances, the force convergence tolerance criteria are also
considered by using the FRCTOL input line. The default force and moment tolerances are both 0.001 and
may be overridden in columns 8-13 and 15-20 of FRCTOL input line respectively. 0.001 tolerance is
assumed to produce the most accurate results. See section 4.4 in Troubleshooting for best practices to
increase the default value.

Note: At a given load increment, if the residual is very small relative to the initial value but the
convergence does not achieve, Collapse Advanced may automatically doubles the maximum number of
iterations to improve the convergence rate.

3.1.2.6 Member Iterations and Displacement Convergence

For any load increment, a beam-column solution is performed for each plastic member using the cross
section sub-element details. Member stiffness iterations continue until the displacements of member
sub-segment joints for two successive iterations meet the member displacement tolerance or until the
maximum number of member iterations has been met. The default number of member iterations is 20
and may be overridden in columns 17-19 of the CLPOPT input line. The default member displacement
tolerance is 0.01 inch or 0.01cm and may be overridden in columns 66-70.

30

SACS® Collapse Advanced Release 15.1

Note: The maximum number of member iterations may be increased when member solution has not
converged.

3.1.3 Additional Member and Plate Options: CLPOP2

3.1.3.1 Additional Member Options

Enter the member eccentricity ratio in columns 8-13. Enter the maximum ductility for any member in
columns 15-20. Members that exceed this limit are assumed to be fractured.

By default, the co-rotational 2nd-order terms are only considered for open thin-walled sections. Enter
‘ITC’ in columns 22-24 to include the co-rotational 2nd-order terms for tubulars and other closed
sections.

Note: In most applications, utilizing the co-rotational 2nd-order terms for tubulars increases
computational time without significant improvement in the results. See Sample 3 for beam elements in
section 5.1.3.3 for application of ‘ITC’ option in collapse analysis

Enter ‘EXC’ option in columns 22-24 to exclude co-rotational 2nd-order terms for all section types. Details
of corotational formulation can be found in section 6.3.1.5. The overall runtime can be reduced by
excluding the 2nd order corotational terms for open sections members which are not subject to large
loads or large deformations (for example, secondary elements which are restrained against buckling).
Sample 10 in section 5.1.10 illustrates the use of the ‘EXC’ option.

There are three options available for second order strains in Collapse Advanced.

1. Axial-Bending 2nd-order Strains: this is the default option.

2. Wagner Nonlinear Torsion: this option is associated with thin-walled open section beams which
are prone to torsional deformation. In most cases, this option does not affect structural
response, but it will increase overall computation time. The Wagner nonlinear torsion can be
added by entering ‘IWT’ in columns 26-28. For further information, see section 6.3.2.3 for
Wagner torsion formulation and sample 6.1 for beams in section 5.1.6.1 for the use of this
option.
3. Excluding 2nd order strains: Enter ‘NLS’ in columns 26-28 to exclude all 2nd-order (large) strains.

Member out-of-straightness in the form of a harmonic function can be included for all members using
the member sub-segments by entering an out-of-straightness ratio in columns 30-35 – the default value
is zero (i.e. no out-of-straightness).

The member end-releases is by default considered for the collapse analysis. However, to improve
convergence rate in special circumstances, the end-releases can be ignored by entering ‘NR’ on columns
55-56. This option has no effect on wishbones and they are always considered for the analysis.

To ensure accuracy of elastoplastic calculation, Collapse Advanced compares the input section
properties of all members with calculated properties. If the input properties of a given member are
differ from calculated properties, by default the program sets the member to elastic and reports a
warning message. By entering ‘C’ on column 54 on CLPOP2 line, the program overrides the input
properties with calculated properties and assumes the member as elastoplastic.

31

SACS® Collapse Advanced Release 15.1

For joint flexibility or joint strength check, Collapse Advanced first computes the properties of brace-
chord connections. If properties of a given brace-chord connection is out of applicability range of
selected methods, Collapse Advanced by default excludes the connection from joint calculation. The
out-of-range connection can be included by entering ‘IJ’, ‘IF’, or ‘IS’ on columns 57-58 to include it for
both flexibility and strength, only for flexibility or only for strength, respectively. For details on
applicability range of all methods, please see sections 6.9 and 6.10 in Commentary.

Note: Including out-of-range connections may lead to non-convergence or my slow down the
convergence significantly.

3.1.3.2 Additional Plate Options

By default, shear deformation effects are not considered for plates – i.e. Plate bending theory is based
on thin (Kirchhoff) theory. Enter ‘MPT’ to include Mindlin thick plate theory deformation effects in
plates. When using Mindlin thick plate theory for triangular plates, shear-locking may be prevented by
applying a correction factor. The triangular plate shear-locking factor is 2.00 by default, this may be
overridden in columns 48-53.

Note: Thick (Mindlin) plate quad element does not need a correction factor for shear-locking. See section
6.4 for details on finite element formulation of Mindlin plate bending theory.

All plate elements include an extra term commonly referred to as “drilling stiffness” to prevent
instability of plates against torsional twist (for details see 6.4.2). The drilling factor is set to 0.05 by
default, this may be overridden in columns 41-46.

Note: Drilling stiffness rarely affects analysis convergence and results, however it becomes effective in
the presence of a large in-plane twist of the plate surface.

3.1.4 Additional Solver and Convergence Options: CLPOP2

3.1.4.1 Convergence Control Parameters
By default, convergence is achieved if both displacement tolerances and force tolerances are satisfied. In
addition to the convergence checks, a force residual is monitored at a given load increment to check for
divergence. Two checks are performed for the force residual at a given iteration:

1) residual ratio to initial residual should be always less than 10𝑒𝑥𝑝1

2) residual value itself should always less than 10𝑒𝑥𝑝2

If any of these tests fail, the iteration is aborted, and load increment is reduced using a sub-
incrementation scheme. The default values of exp1 and exp2 are 4 and 6 respectively and can be revised
in columns 59 and 61, respectively. See section 4.2.8 for troubleshooting.

3.1.4.2 Relaxed Iteration and Convergence Criteria

Collapse Advanced has various features to improve convergence rate and improve accuracy for post-
buckling analysis. The relaxed (weakened) convergence criterion can be implemented by entering ‘CRX’
in columns 70-72. This relaxed criterion is defined as the residual ratio to initial residual less than 10𝑒𝑥𝑝3
and the default value of -5, can be overridden on columns 63-64. A similar relaxed convergence criterion

32

SACS® Collapse Advanced Release 15.1

can be applied for beam element iterations by entering ‘MCR’ in columns 78-80. For further information,
see sections 4.1 and 4.2 for troubleshooting.

Note: Relaxed convergence criteria should be used with caution. The convergence criteria can increase
the chance of convergence for some load increments, however, it may also lead to divergence (or
significant error) for other load increments due to numerical error accumulation during the iterative
process.

A framework for a relaxed Newton iterative process is also available. In this method, the current
displacement increment is scaled by a factor and is then used to perform the next Newton iteration. This
feature may increase the chance of convergence in cases where there is significant deformation in a
single load increment (e.g. just after buckling). The relaxation factor is automatically calculated. Enter
‘IRX’ in columns 66-68 of the CLPOP2 input line to relax the iteration criteria for the entire model.

Many safeguards have been implemented in Collapse Advanced to prevent the analysis from jumping
over an unstable post-buckling path – also known as snap-through response. However, in rare occasions,
some models still display this behavior. The Relaxed Iteration method completely prevents a snap-
through response by controlling progress of the displacement increment in the analysis. For more
details, see section 4.3 in for troubleshooting and sample 2 in section 5.1.2 for the application of the
Relaxed Iteration approach.

Similarly, the relaxed iteration approach can be used for member calculation by entering ‘MIR’ in
columns 74-76 on CLPOP2 line. See 4.2.4 in Troubleshooting for ‘MIR’ usage.

Note: The relaxed iteration approach may increase the chance of convergence, however, it will also
increase number of iterations required for convergence and hence the computation time.

3.2 Output Reports

Output reports including joint deflections, joint reactions, member internal loads and stresses, collapse
summary and member summary reports are available. Report data may be generated based on the final
analysis results or each load increment.

3.2.1 Reports Options: CLPRPT

3.2.1.1 Joint Displacements

Joint displacements may be reported for the structure’s final position or for each load increment by
specifying ‘P0’ or ‘P1’ respectively in columns 8-9 on the CLPRPT input line.

Note: P2 option (deflection report at each iteration) is deprecated and will be overridden as P1.

3.2.1.2 Selecting Joints for Displacement Report

By default, the displacements for each joint in the model is reported in the joint displacement report.
The user may designate the joints to be reported in the joint displacement report on the JTSEL line.
There is no limit to the number of joints that may be designated.

Note: If joints are designated using the JTSEL line, only joints specified are included in the joint
displacement report.

33

SACS® Collapse Advanced Release 15.1

3.2.1.3 Joint Reactions
Joint reactions may be reported for the structure’s final position or for each load increment by
specifying ‘R0’ or ‘R1’ respectively in columns 10-11 on the CLPRPT input line.

Note: R2 option (reaction report at each iteration) is deprecated and will be overridden as R1.

3.2.1.4 Pilehead Reactions Report

The pilehead reactions may be reported for the structure's final position, for each load increment by
specifying 'F0' or 'F1' respectively in columns 26-27 on the CLPRPT input line.

Note: F2 option (pilehead reaction report at each iteration) is deprecated and will be overridden as F1.

3.2.1.5 Elements (Plates and Beam Members) Internal Loads and Stresses
Member internal loads and stresses may be reported for the structure’s final position or for each load
increment by specifying ‘M0’ or ‘M1’ respectively in columns 12-13 on the CLPRPT input line.

Note: M2 option (internal loads and stresses report at each iteration) is deprecated and will be
overridden as M1.
3.2.1.6 Selecting Members for Internal Loads and Stress Report
By default, the internal loads and stresses will be reported for all members in the model. To avoid large
reports the user may select specific members to be reported by using the MEMSEL line. There is no limit
to the number of members that may be designated.

3.2.1.7 Selecting Plates for Reports

By default, reports will be produced for all plates. The user can request reports on specific plates by
using the PLTSEL line. There is no limit to the number of plates that may be selected.

3.2.1.8 Excluding Elastic Members

Members whose properties remain elastic may be excluded from the internal loads and stress reports
by selecting the ‘MP’ option. The report will thus contain internal loads and stresses only for plastic
members.

3.2.1.9 Print Von Mises Stresses on Integration Points

Von Mises stress can be reported at all integration points by entering ‘SP’ in columns 16-17. Number of
integration points depends on beam element section type or plate element type. Figure 2 and Figure 3
show position of integration pints for beam elements and plates, respectively.

3.2.1.10 Collapse Summary Report

The Collapse solution summary report containing the load case, load factor, force summation, and
maximum displacement and rotation for each load increment may be obtained by specifying report
option ‘SM’.

34

SACS® Collapse Advanced Release 15.1

3.2.1.11 Member Summary Report
Select the ‘MS’ option to obtain a plastic member summary report including the plasticity ratio and
member internal loading for each load increment.

3.2.1.12 Element Warning Messages

Select the ‘PW’ option to print all warning message for beam members, plates and pile into the listing
file and collapse log file in addition to collapse troubleshooting file.

3.2.1.13 Von Mises Stress Check for Elastic Plates

Select the ‘VM’ option to print Von Mises check for all plates designated as elastic.

3.2.1.14 Designating Minimum Plasticity

A minimum plasticity ratio for the member stress report may be specified in columns 32-36 on the
CLPRPT line. If a minimum plasticity ratio is specified, only members with sub-elements that have
plasticity ratios greater than the ratio specified are reported. Similar inputs can be entered in columns
38-42 for piles and 44-48 for plates.

Note: Plate strain reports (strain and strain rates) are not supported by Collapse Advanced.
3.2.1.15 Member Strain Report
Following options can be used to generate strain reports for members in Collapse Advanced:

• Enter ‘EN’ on columns 58-59 to report normal strains to the listing file at integration points for
member sub-segments.
• Enter ‘EP’ on columns 62-63 to report plastic strains to the listing file at integration points for
member sub-segments.
• By default, the maximum normal strain will be saved in Collapse Advanced database (clbdb
folder). By entering ‘EP’ on columns 64-65, the maximum plastic strain will be saved instead. The
maximum strain report (normal or plastic) can be generated by selecting Collapse View -> File ->
Generate Report -> Member Strains.

35

SACS® Collapse Advanced Release 15.1

3.3 Applying Load
Unlike standard linear analysis, the Collapse Advanced program analyzes a set of load cases applied step
by step or sequentially rather than simultaneously. The Collapse Advanced program allows for multiple
load sequences to be defined where each load sequence analyzed it treated as an independent
nonlinear analysis. Collapse Advanced also allows the removal of a user defined beam member at a
given load step. This option can be used to model significant damage to the model.

3.3.1 Load Sequence: LDSEQ

3.3.1.1 Defining a Load Sequence
A load sequences defines a set of load steps that will be applied in the sequence or order specified by
the user using LDSEQ lines. Enter the load sequence name in columns 7-10 of the first LDSEQ line
defining the sequence.

Each load sequence may contain from one to fifty SACS IV basic load case definitions (including repeated
load cases) in columns 21-80 on the LDSEQ line. A load step defines the basic load case to be applied,
the number of increments over which to apply the load case, the initial load case factor, and the final
load case factor. For a given load step, the magnitude of each load increment is constant and is
determined by:
(𝐸𝑛𝑑 𝑓𝑎𝑐𝑡𝑜𝑟 − 𝐵𝑒𝑔𝑖𝑛 𝑓𝑎𝑐𝑡𝑜𝑟)
𝐿𝑜𝑎𝑑 𝐼𝑛𝑐𝑟𝑒𝑚𝑒𝑛𝑡 =
𝑁𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑖𝑛𝑐𝑟𝑒𝑚𝑒𝑛𝑡𝑠
Note: The order in which loading is applied in the sequence may have a significant effect on the analysis
results. For example, dead loading or self-weight should be applied before any environmental loading.

3.3.1.2 Load Sequences with More than Three Load Steps

Multiple LDSEQ lines may be used to define load sequences consisting of more than three load steps.
For each subsequent LDSEQ line, leave the load sequence ID in columns 7-10 blank to designate that the
load steps defined are a continuation of the current load sequence. A maximum of seventeen LDSEQ
lines may be used to define a load sequence. Up to fifty load steps may be used for each load sequence.

3.3.2 Removing a Member: MEMREM

Members may be removed from the analysis at a specified load step. Enter the begin and end joints for
each member to be removed and the load increment on the MEMREM line.

3.3.3 Load for Ship Impact, Dropped Object, and Blast Wall

There are two methods available to model impact loads for ship impact, dropped object, and blast wall
analysis:

1. Utilizing IMPACT input line with the user-defined load cases given in SACS model file
2. Automatically generating impact loads with Dynamic Response Time-History Analysis

In the first method, the impact load case is a user-defined load given in SACS model file, the analysis
sequence is defined by LDSEQ input line. The total energy absorbed by the structure, member or the
ship is entered on IMPAC input line or using ENERGY input file. The analysis continues until total energy
36

SACS® Collapse Advanced Release 15.1

is absorbed and then the structure is automatically unloaded. For more details, see impact samples in
section 5.6.

In the second method, the load sequence is defined by two input lines: LDAPL and LDAPC. LDAPL is used
to add static loads (like model weight) to the collapse load sequence while LDAPC is automatically
generated by the dynamic analysis to add dynamic load history for the collapse analysis.

3.4 Joint Connection Options

3.4.1 Tubular Connection Strength and Flexibility Parameters:
3.4.1.1 Tubular Connection Options: JSOPT
Joint strength and flexibility options used for the tubular connections can be implemented using the
JSOPT line. Using JSOPT line following options can be implemented:

• Relief option to check the brace offsets

• Revise the default tolerance to check brace offsets
• Minimum and maximum gap allowed for K connections
• Effective thickness for grouted members (for more details check Section 2.1.3. in Joint Can user
manual)
• ISO 19902 brace utilization option (see section 6.9.3 in Commentary for details)
• Print level for the unity check values
• Tolerance to check offsets of brace-chord connections
• Option to print connection geometric parameters (β, γ, and τ) to the listing file.

This line is optional in any collapse analysis. If this line is omitted, then default options will be used.

3.4.1.2 Resistance Factor Data: RSFAC/RSFACO

RSFAC and RSFACO input lines can be used to modify joint resistance factors for API RP 2A-LRFD (see
section 6.9.2 in Commentary) or the additional user-defined resistance factors for ISO 19902, NORSOL
and MSL (see sections 6.9.3, 6.9.4, and 6.9.5 in Commentary). RSFAC revises the factors for all
connection and braces while RSFACO revises the factors for selected brace.

In addition to resistance factors, the user can revise API Yield Factor, NORSOK Material factor, or ISO
Yield and Extra resistance factors on RSFAC line.

This line is optional in any collapse analysis. If this line is omitted, then default options will be used.

3.4.2 MSL Joint Flexibility and Strength Formulation: MSLOPT

The MSL joint flexibility and strength formulation developed for the JIP ‘Assessment Criteria, Reliability
and Reserve Strength of Tubular Joints’ can be implemented via the MSLOPT line.

Two levels of tubular connection capacity, ‘mean’ level and ‘characteristic’ level, can be defined. The
‘mean’ level corresponds to a 50% probability of survival while the ‘characteristic’ level corresponds to a
95% probability of survival.

37

SACS® Collapse Advanced Release 15.1

3.4.2.1 Joint Flexibility
The effects of tubular connection flexibility may be accounted for by specifying analysis option ‘MF’ or
‘CF’ for mean or characteristic level, respectively, in columns 8-9.

By default, a convergence tolerance of 0.001 is assumed for joint distortion and rotation. The joint
distortion tolerance can be specified in columns 15-19. The joint rotation tolerance can be specified in
columns 20-24.

3.4.2.2 Joint Strength

The tubular connection strength at ‘mean’ level can be implemented by specifying analysis option ‘MS’
in columns 10-11. Alternatively, the connection strength may be implemented at the characteristic level
by specifying ‘CS’ in columns 10-11.

3.4.2.3 Fracture Criteria

The ductility limits for tension loaded joints may be accounted for by specifying analysis option ‘MT’ at
mean level, and ‘CT’ at characteristic level in columns 12-13.

3.4.2.4 MSL Assessment Factor of Safety

The user can enter MSL assessment factor of safety on columns 25-29 of MSLOPT line. For details see
section 6.9.5 in Commentary. The program uses default value of 1.0 if no input is entered.

3.4.2.5 Plasticity in MSL

The Collapse Advanced program, by default, uses MSL coupled elastoplastic joint flexibility method (see
section 6.11 in Commentary for details). The MSL plasticity model can be ignored by entering ‘NMP’ on
columns 78-80 of MSLOPT line.

3.4.3 Joint Strength/Flexibility Selection: JSSEL

Individual joints may be chosen for joint strength or joint flexibility analysis. The option used, either joint
strength ‘JS’ or joint flexibility ‘JF’, must be specified with CLPOPT analysis options. With the ‘JS’ option
specified on the CLPOPT line, a joint or group of joints may be chosen for joint strength analysis with the
JSSEL line. This means that all braces connected to the joints specified will be included or excluded from
the joint strength analysis. The line either includes or excludes the joints specified in columns 9-77 based
on the entry in column 7. Specifying ‘I’ in column 7 will mean that the joints named are included in the
joint strength analysis; specifying ‘X’ in column 7 will mean that all joints except those named are
included in the joint strength analysis.

In the same manner, joints may be chosen for joint flexibility analysis with the JFSEL line. With either
JSSEL or JFSEL, the include or exclude option is mutually exclusive. Therefore, if multiple lines are used
to include or exclude joints, each line must have the same option specified in column 7.

In the following example, joints 101 and 102 are excluded from joint flexibility analysis. All other joints
will be analyzed.

38

SACS® Collapse Advanced Release 15.1

If the choice of a single joint for joint strength or joint flexibility analysis is not sufficiently restrictive, the
BSSEL and BFSEL allow the user to restrict strength or flexibility analysis to individual brace/chord
connections. The option used, either joint strength ‘JS’ or joint flexibility ‘JF’, must be specified with
CLPOPT analysis options. With the ‘JS’ option specified on the CLPOPT line, a brace/chord connection
joint may be chosen for joint strength analysis with the BSSEL line. The first brace member joints are
specified in columns 9-12 (begin joint) and columns 13-16 (end joint). The strength analysis will be
calculated at the brace/chord connection joint, which is either the begin joint or the end joint of the
brace member and is specified in columns 17-20 for the first brace. Up to five braces may be specified
on the BSSEL line. As in the JSSEL line, brace/chord connections may be included or excluded from
strength analysis by specifying ‘I’ or ‘X’ in column 7.

Equivalently, joint flexibility for individual brace/chord connections is specified with the BFSEL line. With
either BSSEL or BFSEL, the include or exclude option is mutually exclusive. Therefore, if multiple lines are
used to include or exclude brace/chord connection joints, each line must have the same option specified
in column 7.

In the following example, brace/chord connection joint 101 of brace member 101-401 is excluded from
brace strength analysis. All other brace/chord connections will be analyzed.

The resistance factor specified for a brace/chord connection may be modified using the RSFACO line.
This line allows the user to override joint resistance factor values specified on RSFAC lines. The line
specifies the brace member in columns 8-11 (begin joint) and columns 12-15 (end joint). The
brace/chord connection joint, which is either the begin joint or the end joint, is specified in columns 16-
19. The resistance factors (axial tension, axial compression, in-plane bending, out-of-plane bending,
yield stress) are specified in columns 21-45. Optionally, the connection type may be specified in column
47, with choices being ‘X’ (X or cross connection), ‘Y’ (T or Y connection), or ‘K’ (K brace connection). Any
of the resistance factors left unspecified or given the value 0.0 will be replaced by values specified for
the connection joint on previous RSFAC lines.

In the following example, brace/chord connection joint 201 of brace member 201-501 will have an in-
plane bending resistance factor of 3.81 and an out-of-plane resistance factor of 3.61. The values for the
axial tension, axial compression and yield stress resistance factors are the values specified earlier on
RSFAC lines for joint 201.

39

SACS® Collapse Advanced Release 15.1

3.5 Designating Elements as Elastic
By default, members and groups designated as skipped for post processing are treated as large
deflection elements with elastic material properties. Additionally, members or member groups may be
designated by the user as elastic elements using the MEMELA and GRPELA input lines, respectively.
Similarly, plate elements and plate groups can be designated as elastic elements using the PLTELA and
PGRELA input lines respectively.

Note: Designating elements to remain elastic can significantly reduce the run time for a collapse analysis.
Also, certain element types including wishbones, non-structural framing, i.e. framing representing risers,
boat landings, anodes, etc. and dummy framing should be treated as elastic elements for the nonlinear
analysis.

Note: Designating elements to remain elastic does NOT imply they do not experience large deformation.
For example, if the member is subjected to a localized point load, it may be elastically buckled during
analysis. To prevent unfavorable buckling for elastic elements, the user may use MEMSEG or GRPSEG to
set a single sub-segment to the elastic member. For more details see section 4.5 in Troubleshooting.

3.5.1 Beam Elements

3.5.1.1 Elastic Members: MEMELA
Specify the start and end joints of any member that is to be considered as a large deflection elastic
element on the MEMELA input lines. As many MEMELA lines as required may be specified.

3.5.1.2 Elastic Member Groups: GRPELA

Specify member groups to which all elements assigned are to be considered as large deflection elastic
elements on the GRPELA input line. As many GRPELA lines as required may be specified.

3.5.2 Plate Elements

3.5.2.1 Elastic Plates Elements: PLTELA
Specify the plate ID’s of plate elements that are to be considered as large deflection elastic elements on
the PLTELA input lines. As many PLTELA lines as required may be specified.

3.5.2.2 Elastic Plate Groups: PGRELA

Specify plate group names that are to be considered as large deflection elastic elements on the PGRELA
input line. As many PGRELA lines as required may be specified.

3.6 Nonlinear Springs

The Collapse program supports nonlinear springs and nonlinear spring supports.

40

SACS® Collapse Advanced Release 15.1

3.6.1 Nonlinear Spring Supports: NLSPRG
A general nonlinear spring to ground element is available in Collapse. The spring elements have six
uncoupled degrees of freedom. The force deflection characteristics of the spring for each degree of
freedom are defined by discrete Force-Displacement points in the input line NLSPRG. Up to four points
may be used to define the spring Force-Displacement characteristics. As many NLSPRG input lines as
required may be specified.

3.6.2 Joint to Joint Nonlinear Springs: NLSPJJ

Nonlinear springs can be assigned between existing joints. The force deflection characteristics of the
spring for each degree of freedom are defined by discrete Force-Displacement points in the input line
NLSPJJ. As many points as required may be used to define the spring Force-Displacement characteristics.
As many NLSPJJ input lines as required may be specified.

3.6.3 Corotational Joint to Joint Nonlinear Springs: NLSPST

This option is not currently supported in Collapse Advanced and will be implemented in a future release.

41

SACS® Collapse Advanced Release 15.1

3.7 Material Properties
3.7.1 Multilinear Stress-Strain Curve:
Multilinear yield stress-strain can be specified using inputs lines MATPRP. This input line can be used to
model complex strain-hardening and/or strain softening after the yield. The post-yield material
behavior is given as function of plastic strain and stress factor. Plastic strain (defined as total strain
minus yield strain) and Stress factor (defined as actual stress divided by the yield stress) values
beginning at the yield point are entered in MATPRP PLAS lines to define the post-yield behavior of the
material.

Input lines MATGRP, MATPGR, and MATPLG can used to assign the multilinear plastic material to
member groups, plate groups and pile groups, respectively. If this option is not selected, post-yield
behavior remains linear and is governed by the strain hardening ratio specified in the CLPOPT line – i.e.
bilinear elastoplastic. Input lines MATGRP, MATPGR, and MATPLG should be entered after MATPRP
input line.

Figure 7 shows a hypothetical material with strain hardening and strain softening. Table 2 shows how to
calculate the Plastic Strain and Stress Factor values from the actual stress-strain curve.

Figure 7: Sample multilinear elastoplastic with strain hardening and softening

𝑘𝑁 𝑘𝑁 𝜎𝑦
Table 2: Multilinear Plastic Material: 𝜎𝑦 = 24.8 𝑐𝑚 2
𝐸 = 20000 𝑐𝑚 2
𝜀𝑦 = 𝐸
= 0.00124

Strain Stress Plastic Strain Stress factor

𝑘𝑁
𝜀 𝜎 𝑐𝑚2 𝜀 − 𝜀𝑦 𝜎⁄𝜎𝑦
0.0 0.0
Yield 0.00124 24.8 0.0 1.00
0.00224 29.26 0.001 1.18
0.00324 29.76 0.002 1.20
0.00624 29.76 0.005 1.20
0.01124 27.28 0.010 1.10
0.05124 22.32 0.050 0.90

42

SACS® Collapse Advanced Release 15.1

 1 2 3 4 5 6 7 8
12345678901234567890123456789012345678901234567890123456789012345678901234567890
* define the multilinear plastic
MATPRP HEAD MAT1
MATPRP PLAS 1.0 0.001 1.18 0.02 1.20 0.005 1.20 0.01 1.10
MATPRP PLAS 0.05 0.90
* assign to member groups
MATGRP MAT1 GR1 GR2 GR3 GR4
* assign to plate groups
MATPGR MAT1 PG1 PG2 PG3 PG4
* assign to pile groups
MATPLG MAT1 PL1 PL2 PL3 PL4

Note: The plastic strain values should be in the monotonically increasing order.

Note: For calculating the stress value corresponding to a plastic strain value greater than the maximum
value specified in the material model (beyond the last data point in the material model), Collapse
Advanced linearly extrapolates the input curve.

Note: it is recommended to start the plastic strain-stress curve from 0.0 plastic strain and 1.0 stress
factor. Otherwise, the program automatically inserts (0.0,1.0) at the beginning of the plastic stress-strain
curve.

Note: Input lines MATGRP, MATPGR, and MATPLG should be entered after MATPRP input line.

3.7.2 Member Ductility Limits: MEMDUC

By default, post yield ductility is not limited and may increase indefinitely. Member ductility limits may
be imposed using the begin and end joints of each member and the ductility limit in percent on the
MEMDUC line.

3.7.3 Yield Stress Overrides

A yield stress factor may be applied which modifies the yield stress entered in the SACS model file using
the YSFACT line. The yield stress for the entire collapse model may be overridden using the YSUOVR line.
Specific yield stress values may be overridden with YSUMOD line. Member and plate groups may be
individual overridden using the YSMGOV and YSPGOV lines, respectively.

The inner tubular of concentric tubular sections may have a different yield stress than the outer tubular.
Specify the start and end joints of any member with an inner tubular yield stress override on the
GRMSEL line.

When multiple yield stress overrides are applied to a beam or plate elements, the following hierarchy
determines the yield stress:
1. YSMGOV (YSPGOV) – Yield Stress Member (Plate) Group Override transcends:
2. YSUMOD – Yield Stress Modification, which transcends:
3. YSUOVR – Yield Stress Universal Override, which transcends:
4. YSFACT – Yield Stress Factor

43

SACS® Collapse Advanced Release 15.1

4 Troubleshooting
Collapse Advanced results are reported in following forms:

1. Collapse listing file (‘clplst’) which contains outputs (printed based on options on CLPRPT
line), error messages and important warning messages.
2. Collapse database folder or ‘clpdb’ which is used to visually review collapse results in
PRECEDE.
3. Collapse analysis log file (‘clplog’) which contains analysis iterations, important event (such
as joint failure, etc.) and important warning messages.
4. Collapse troubleshooting log file (‘clptrbl’) which contains detailed information about
convergence, residual value at all iterations for all load increments, detailed warning messages,
and important events at each load increment.

Collapse troubleshooting log file is the main source needed to identify and resolve collapse analysis
issues in order to improve convergence and accuracy. Figure 8 and Figure 9 illustrate different outputs in
the troubleshooting log file associated with Newton iteration and Arc-length method, respectively.

Figure 8: Newton iteration outputs in the troubleshooting log file

44

SACS® Collapse Advanced Release 15.1

 Figure 9: Arc-length iteration outputs in the troubleshooting log file

4.1 Converting Current Collapse to Collapse Advanced

By default, the Collapse solver is used for any nonlinear SACS analyses. Collapse Advanced may be
enabled in the SACS System Settings:

1. In the SACS Executive, at the top left select Settings

2. Under Analysis Settings, select ’Yes’ for option ‘Use Collapse Advanced’ (see Figure 10). The
SACS environment is now set up to use Collapse Advanced.

Figure 10: SACS System Settings for Collapse Advanced

45

SACS® Collapse Advanced Release 15.1

Collapse Advanced generates new output files which require an updated analysis run file. An existing
analysis run file may be updated via two options:

a) Generate a new analysis run file using the SACS Executive Analysis Generator.
b) If the run file is already generated by an older version of SACS, open the run file in the Analysis
Generator and simply re-save it.

Existing collapse input files must be updated with Collapse Advanced input options to use the new
Collapse Advanced capabilities for nonlinear analysis:

✓ Essential Inputs: the sub-incrementation and the Arc-length method are necessary for the
majority of Collapse Advanced simulations. The sub-incrementation method can be selected by
entering ‘SI’ in columns 44-45 on CLPOPT input line and the Arc-length method is set by entering
‘AL’ in columns 42-42 on CLPOPT input line.

For many cases, including the sub-incrementation and the Arc-length method is adequate to carry out
the analysis using the Collapse Advanced program and no further input is required. However, to improve
the analysis or resolve any issues, the following options can be considered:

• Relaxed Convergence: the continue option (‘CN’ on CLPOPT line) is no longer supported by
Collapse Advanced. Instead, Collapse Advanced supports a relaxed (weakened) convergence
criteria to improve the convergence rate. The relaxed convergence can be implemented on by
entering CRX and MCR on CLPOP2 input line – see section 3.1.4.2 for input line details. Section
4.4 discusses how to choose the relaxed convergence parameter (i.e. the exponent on columns
63-64) based on the residual convergence rate given in the troubleshooting section.
• Relaxed Iteration: As mentioned in section 3.1.4.2 in rare occasions, the relaxed iteration is
required to prevent the snap-through buckling behavior. The relaxed iteration can be
implemented by entering IRX and MIR on CLPOP2 input line. Please see section 4.3 for more
discussion on post-buckling analysis.
• Improve Arc-length: Arc-length default parameters can be modified on the ARCLEN input line.
For discussion on how to modify Arc-length parameters, please see section 4.2.9 and discussion
on post-buckling analysis in section 4.3.
• Improve sub-incrementation: If the sub-incrementation levels must be increased to improve
convergence, the default value can be revised on the SUBINC input line.
• Force Tolerance: If the default force tolerance is not suitable for the analysis, it can be revised
by entering the FRCTOL input line.

To revert to the current Collapse program:

1. On SACS Systems Setting, set User Collapse Advanced to No.

2. Re-generate or re-save the analysis run file in the SACS Analysis Generator.
3. Remove additional input lines associated with Collapse Advanced from collapse input file.

46

SACS® Collapse Advanced Release 15.1

4.2 Collapse Advanced Messages
To troubleshoot a collapse run, three message types are output:

1. Information (*** INFO): these messages report important events during a Collapse run. The
events include: plasticity in elements, joint failure, local buckling, pile failure, energy absorption
for ship impact, or other similar significant events in the collapse analysis. All INFO messages
are written to the output listing file and troubleshooting log file, in addition, important events
(such as pile pull-out or punch-through) are also reported to the Collapse log file.
2. Errors (*** ERROR): Error messages report any problems encountered with a collapse input
file, SACS model file or any other input. All errors must be resolved before collapse analysis may
proceed. Error messages are written to the Collapse output listing file.
3. Warnings (*** WARNING): these messages pertain to existing issues with the input to a
Collapse analysis and issues arising from iterative calculations. A warning message about an
input does not prevent the analysis from running and are reported in the output listing file.
Warning messages on the iterative procedure, such as non-convergence/singularity or other
issues related to Collapse calculations are reported in the troubleshooting log file. If the non-
convergence/singularity issue prevent the collapse analysis from running, then these messages
are reported in both the log file and the output listing file.

The following sections provide more details on Collapse Advanced messages.

4.2.1 Information Messages in Collapse Advanced

Following information messages may be reported to collapse outputs.

Maximum displacement reaches the user-defined maximum values on CLPOPT line:

*** INFO: Maximum deflection for joint ‘joint name’ has been exceeded
for degree-of-freedom ‘degree of the freedom ID’

Nonlinear Solvers Information:

*** INFO: Commencing level ‘sub-level#’ subincrementation for load
condition ‘load condition name’ at load factor ‘load factor’
*** INFO: Arc length analysis has commenced for load condition ‘load
condition name’ at load factor ‘load factor’
*** INFO: Arc length analysis has been terminated successfully.
*** INFO: Commencing cylindrical arc length iterations with minimum
angle criterion (level 'sub-level#') with arc length = 'Arc-length
parameter'
*** INFO: Commencing spherical arc length iterations with minimum
angle criterion (level 'sub-level#') with arc length = 'Arc-length
parameter'

When the maximum joint deflection reaches during Arc-length steps:

*** INFO: Arc length analysis has been terminated successfully.

47

SACS® Collapse Advanced Release 15.1

If the program encounters an incomplete load sequence (or it does not reach to the user-defined
maximum deflection), the following information message is printed to the output listing file and log file:
'*** INFO: Analysis has been terminated for load sequence ‘load
sequence name’. See Collapse Troubleshooting Log file for details.

The above message is reported when the program exhausts all sub-incrementation levels (SUBINC input
line) and Arc-length steps (ARCLEN input line).

The following information message is issued on convergence when using the “Relaxed Convergence”
option on the CLPOP2 input line and if the load increment is converged using relaxed (weakened)
criterion:
*** INFO: Load increment has converged using RELAXED convergence
criterion (Residual < 10^ ‘exponent’ Initial Residual)

The forces balance of member must be checked if numerous load steps have converged when using this
option. The residual values can be reviewed in the troubleshooting log file to adjust the exponent.

Note: For the Arc-length method, Collapse Advanced calculates the relative residual based on the
residual in the second iteration. The first residual is associated with the initial tangent stiffness matrix.

Beam Elements:
*** INFO: The onset of plasticity has occurred on member ‘member ID’
*** INFO: Local buckling has taken place on member ‘member ID’ sub-
segment ‘sub-segment#’ at load step ‘load step’
*** INFO: The onset of ductility has occurred on member ‘member ID’ at
location ‘location in the member’, strain = ‘strain value’
*** INFO: The member ‘member ID’ has fractured at location ‘location
in the member’, strain = ‘strain value’

Plate Elements:
*** INFO: The onset of plasticity has occurred on plate ‘plate ID’
*** INFO: Elastic plate ‘plate ID’ has a Von Mises ratio > 1.0

Joints:
*** INFO: The onset of joint connection plasticity has occurred at
joint ‘joint name’ on brace ‘brace ID’)
*** INFO: Joint connection failure has occurred at joint ‘joint name’
on brace ‘brace ID’ at load increment ‘load step’
*** INFO: Brace joint connection pull-out has occurred at joint ‘joint
name’ on brace ‘brace ID’ at load increment ‘load step’

Piles:
*** INFO: The onset of pile pull-out has occurred on the pile located
at joint ‘pilehead joint’

48

SACS® Collapse Advanced Release 15.1

*** INFO: The onset of pile punch-through has occurred on the pile
located at joint ‘pilehead joint’
*** INFO: The onset of plasticity has occurred on the pile located at
joint ‘pilehead joint’

Inputs:
*** INFO: Joint flexibility selections are working in ''include''
mode.'
*** INFO: Joint flexibility selections are working in ''exclude''
mode.'
*** INFO: Joint strength selections are working in ''include'' mode.
*** INFO: Joint strength selections are working in ''exclude'' mode.
*** INFO: Brace ‘brace ID’ at joint ‘joint name’ has been excluded
from the joint strength check.
*** INFO: Brace ‘brace ID’ at joint ‘joint name’ has been selected for
the joint strength check.
*** INFO: Brace ‘brace ID’ at joint ‘joint name’ has been excluded
from the joint flexibility calculation.
*** INFO: Brace ‘brace ID’ at joint ‘joint name’ has been selected for
the joint flexibility calculation.'

4.2.2 Error Messages in Collapse Advanced

Error messages are self-explanatory and provide detailed information regarding existing issues with the
input files. All error messages must be resolved in order for a Collapse analysis to run.

In addition to the input error messages, there are also two other error messages as follows:

*** ERROR - Access denied to ‘folder or file name’ this error message indicates
that the Collapse program cannot read or write to the specified file or folder – for example the files are
read-only or the user does not have write privileges on the machine.
*** ERROR: MEMORY ALLOCATION ERROR ‘Collapse Advanced program section
label’ this error message indicates that the analysis is running out the memory and only occurs for
very large models, if the machine has limited memory or the machine operating system limits memory
usage.

4.2.3 Warning Messages in Collapse Advanced

There are two groups of warning messages in Collapse Advanced. The first group illustrates possible
issues in the input. These issues are not necessarily problematic for a collapse analysis and do not
prevent it from running. For example, for joint flexibility/strength, we may get the following messages:
*** WARNING: A number of braces have properties exceeded the
applicable range for selected joint flexibility formulation.

49

SACS® Collapse Advanced Release 15.1

*** WARNING: The parameter ‘parameter name’ at joint ‘joint name’ is
exceeded the applicable range of K-connection for selected joint
strength

These messages indicate the joint specifications are out of the applicable range of the selected joint
flexibility method. In this case the analysis still uses the selected method, but the results may not be
accurate. Warning messages corresponding to input issues are only reported to the collapse output
listing file.

The second group of warning messages is related to non-convergence or other computational issues. All
non-convergence messages are reported to the collapse troubleshooting log file. If a non-convergence
issue prevents the analysis from continuing (or has significant implication in the results), a warning
message is also reported to the listing file and log file.

Other important warning messages which may be output to the listing file and the log file include:

• Convergence or other solver issues at the first load increment (initial convergence issue)
Whether the sub-incrementation method is selected or not selected, if the nonlinear solver
encounters an issue at the first load increment, the program outputs a warning message to the
listing file and the log file. The warning message indicates the user-defined load increment size is
too large for the model. However, in most of the cases, if the sub-incrementation method is
selected on CLPOPT line, the program will automatically resolve the convergence issue and no
further action is required. For additional details, please see section 4.2.6 and the plate samples 4
in section 5.2.4.
• Convergence or other solver issues near the load peak
If the sub-incrementation is NOT selected but the Arc-length method is selected, the program
outputs a warning message near the load factor peak (the limit point). The warning message
simply indicates the transition from Newton Solver to Arc-length Solver without sub-
incrementation. If the analysis continues without any issue, no further action is required. For
additional details, please see section 4.2.6 and the plate sample 6 in section 5.2.6.

The following sections discuss various warning messages. The use of different methods and options
available to improve a collapse analysis based on the warning messages are also presented.

4.2.4 Warning Messages in Members

The following warnings may occur for member calculation:

*** WARNING: Convergence has not been achieved for member ‘member
name’
*** WARNING: Decomposition error whilst solving for member ‘member
name’
*** WARNING: Lateral torsional buckling suspected for member ‘member
name’
*** WARNING: Plasticity calculation error whilst solving for member
‘member name’
*** WARNING: Calculation for joint-flexibility brace ‘Brace ID’ has
encountered a problem.

50

SACS® Collapse Advanced Release 15.1

'*** WARNING: Solution for member ‘member name’ failed

The program will try to resolve the above issues by switching to sub-incrementation and Arc-length
methods. However, if these issues occur multiple time, the analysis may slow down or eventually stop
before completing the load sequence. The following remedial actions may be considered:

1) Reduce the convergence tolerance.

2) Increase the maximum number of iterations for members.
Note: If too many members have convergence issues, this option may slow down the analysis.
3) Check to see if it the member is too short. By default, if the member aspect ratio
(height/length) is greater than one, Collapse will only use a single sub-segment. However, if the
member aspect ratio is slightly less than 1, Collapse uses user-defined number of sub-segments
which may lead to non-convergence or calculation failure. In cases such as these the MEMSEG
and GRPSEG input lines can be used to override number of sub-segments for very short
members.
4) Check the yield stress and strain hardening ratio for warning messages associated with
plasticity calculations.
5) If there is a significant deformation (for example resulting from buckling or large member
loads), the number of sub-segments may be increased by entering new values on MEMSEG or
GRPSEG input lines. Also, the problematic member may be directly sub-divided into shorter
members in the SACS model file.
6) Use the Member Iteration Relaxed option by entering ‘MIR’ in columns 74-76 on CLPOP2 line to
gradually apply the member deformation. This method reduces the displacement increment by
an automatic factor to prevent snap-through response of the member during buckling and
ensures the member follows a more stable path. However, this may increase the overall
number of iterations and computational time. As a good practice, when using the ‘MIR’ option
the maximum number of member iterations should also be increased.
7) User Member Convergence Relaxed option by entering ’MCR’ in columns 78-80 of the CLPOP2
input line to use a more relaxed convergence criterion. This option should be used with caution
as it may lead to unbalanced forces in the members.
8) Check the joint flexibility approach being used if the warning message is related to joint
flexibility. Non-convergence issues in member calculations may occur if the joint specification is
out of the applicable range of the selected joint flexibility method. In such cases, JFSEL can be
used to exclude the joint from the joint flexibility calculation.
9) Check member releases and the magnitude of the applied load on the member. For large
deformation analysis, large member loads on members with end releases may lead to non-
convergence.

51

SACS® Collapse Advanced Release 15.1

4.2.5 Warning Message in Piles
The following warning messages may be output for pile and soil analysis:
*** WARNING: Calculation for pile at joint ‘pilehead joint’ has not
converged.

*** WARNING: Convergence has not been achieved for the pile attached
at joint ‘pilehead joint’

These warning messages occur when the procedure used to calculate the stiffness and plasticity of a pile
has failed to converge.

The determination of stiffness and plasticity of a pile with non-linear soils requires an iterative solution.
The iterative solution is dependent on the displacement convergence criteria specified on the PSIOPT
line of the PSI input file.

The number of iterations used to solve each pile is automatically set to 40 or the number of iterations
defined on the CLPOPT line in columns 11-13 – whichever is greater. If convergence is not achieved prior
to the maximum number of iterations, then the above warning messages will be displayed in the
collapse output log file and the program will switch to the sub-incrementation and then Arc-length
methods if these options have been turned on. If the convergence is still not achieved after exhausting
all sub-incrementation levels and Arc-length iterations, then the program will issue a warning message
noting the termination of the load sequence.

For cases where the program cannot automatically overcome convergence issues associated with the
piled foundation then the following checks should be carried out to help improve convergence:

1. Make the displacement and force convergence tolerance levels equal.

2. Check instabilities in the supported structure. The user should conduct the analysis without the
piled foundation to ensure the stability of the structure.

3. Check the pile capacity using PSI or the single PILE program to ensure the foundation is not
unrealistically weak.

4. Check the plasticity ratios in the piles for the formation of a mechanism.

4.2.6 Warning Messages for Non-convergence

The following messages may be reported at the first load increment:
*** WARNING: Structural instability has been detected during the first
load increment. Increase the number of load increments for this load
step.
*** WARNING: Initial convergence problems encountered with member
‘member ID’.
*** WARNING: Member ‘member ID’ has exhibited high plasticity during
the first load increment.

The above messages indicate that the first load increment may be too large and the load should be
gradually applied to improve convergence. To overcome this, the sub-incrementation option on the
52

SACS® Collapse Advanced Release 15.1

CLPOPT line should be used. However, if the solution still fails to converge then the number of load
increments should be increased. See plate sample 4 in section 5.2.4 for the implementation of the sub-
incrementation method where it is used to improve convergence for the initial load increments.

Sub- incrementation option can also be used to help improve convergence in the near vicinity of a limit
load (e.g. the peak load.). Non- convergence in the near vicinity of a limit load will be indicated by the
following warning message:
*** WARNING: Load increment has failed to converge.

For an example, please see the plate sample 6 in section 5.2.6.

The sub-incrementation option will automatically reduce the load factor and the following message will
be output:
*** INFO: Commencing level ‘sub-level#’ subincrementation for load
condition ‘load condition name’ at load factor ‘load factor’.

The analysis may slow down significantly if the sub-incrementation process occurs over many load
increments. In cases such as these, the following remedial actions can be considered:

1. Reduce the user-defined load increment size: the load increment size can be reduced by
increasing number of load steps on LDSEQ input line for static analysis. For dynamic analysis,
reducing time step size (Output Time Interval) has similar effects.
2. Increase the maximum number of iterations: In some instances, the convergence may be
achieved by slightly increasing maximum number of iterations without significantly increasing
the number of sub-incrementations. The maximum number of iterations can be chosen by
reviewing the convergence rate provided in the troubleshooting log file.

Note: At a given load increment, if the residual is very small relative to the initial value but the
convergence is not achieved, Collapse Advanced automatically doubles the maximum number of
iterations to improve the convergence rate.

4.2.7 Numerical Instability Warnings

The following warning messages may be output by the nonlinear solver in reference to the stiffness
matrix inversion instability:
*** WARNING: Load increment has been aborted due to a problem with the
global decomposition procedure.
*** WARNING: Load increment has been aborted due to severe ill-
conditioning.
*** WARNING: Load increment has been aborted due to a decomposition
error.
*** WARNING: Load increment has been aborted due to a solution error.
*** WARNING: It is recommended that the constraints of the model are
checked.

53

SACS® Collapse Advanced Release 15.1

If above warning messages are output during the application of a load case where the structure should
behave in a linear manner, for example, a load case associated with an initial gravity load, then this is
usually an indication that there is an issue with the SACS model itself. However, if the warning messages
are output in the presence of significant plasticity or other possible failure modes despite using sub-
incrementation or the Arc-length approach, then this indicates that an overall failure mechanism has
formed and the structure has collapsed. To improve results under these circumstances, a tighter
convergence criterion with smaller increments should be used.

4.2.8 Divergence Warning Messages

Checks for divergence associated with the Newton and Arc-length methods are performed from the
third iteration to prevent unnecessary computation. The following warning messages are associated
with divergence:
*** WARNING: Load increment has been aborted due to very large
residual norm ‘residual value’ (> 10^ ‘exponent’)
*** WARNING: Load increment has been aborted due to very large
residual norm ratio ‘residual ratio’ (> 10^ ‘exponent’)

where exponents are user-defined values. The exponents can be updated on CLPOP2 input line if the
output log file shows the load increment has been aborted at an early stage – see section 3.1.3 for
details.

If the divergence warning is output over a large number of load increments, the following actions may
be considered:

1. Implement the sub-incrementation method by entering ‘SI’ on CLPOPT line

2. Reduce user-defined load increment size: the load increment size can be reduced by increasing
the number of load steps on LDSEQ input line for static analysis. For dynamic analysis, reducing
time step size (Output Time Interval) has similar effects.
3. Updating the divergence exponent: if the residual values given in the troubleshooting log file
reveals that the program aborts the iterative procedure at an early stage then the divergence
exponent can be increased to allow the program to continue. However, if the program reports
divergence issues too late, then the exponent can be reduced to prevent unnecessary
computation to help reduce the total runtime.

Note: For the Arc-length method, Collapse Advanced calculates the relative residual based on the
residual in the second iteration. The first residual is associated with the initial tangent stiffness matrix.

4.2.9 Arc-length Warning Message

The use of the Arc-length method may result in the following warning messages:
*** WARNING: Arc length iteration is terminated due to error initial
tangent stiffness calculation.
*** WARNING: Arc length iteration is terminated. No solution is
available in arc length quadratic equation at initial load factor.
*** WARNING: Arc length iteration is terminated. No solution is
available in arc length quadratic equation at initial load factor.
54

SACS® Collapse Advanced Release 15.1

*** WARNING: Arc length iteration is terminated. No solution is
available in arc length quadratic equation at current load factor.
*** WARNING: Arc length iteration is terminated because of an error in
stiffness matrix decomposition.
*** WARNING: Arc length iteration is terminated because of an error in
linear system solution.
*** WARNING: Arc length method did not converge within maximum
iterations
*** WARNING: Arc length iteration is terminated because of very small
increment norm ‘norm value’ < 1.0E-9.

The above warning messages indicate that the automatically calculated arc-length parameter may be
too large and Arc-length method has difficulty finding a stable path. The Arc-length method utilizes sub-
incrementation to reduce this parameter – automatic arc-length value and convergence details can be
found in troubleshooting log file. The run time may increase, or the program may stop if Arc-length sub-
incrementation occurs during too many load steps. To speed up the analysis, the user should enter a
scaling factor (less than one) in columns 24-29 on ARCLEN input line to reduce the arc-length parameter.

The following warning message will be reported if maximum number of sub-incrementation levels have
been exceeded:
*** WARNING: Maximum arc length sub incrementation level has been
exceeded.

Under these circumstances, the following steps are recommended to test the results:
1. If there is already significant deformation and plasticity in the model, it is very likely that the
Arc-length method cannot find any stable path for unloading and a mechanism may have
formed in the model. In other words, the structure may have collapsed. To improve the results
further, the number of load increments may be increased, or a tighter convergence tolerance
can be used.
2. If this warning message is reported without any significant plasticity of deformation, then the
likely causes are:
a) Convergence issue in members, piles, or other parts of the analysis. The troubleshooting
log file may provide more information about non-convergence issues.
b) The load increment may be too large (especially if this warning occurs during the initial
load steps).
c) A large deformation increment may have occurred during a single load step. In this case,
the use of the Relaxed Iteration option on the CLPOP2 input line may help to mitigate
this problem.
d) Tighter convergence criteria may be required.

If all arc-length steps are exhausted either before reaching the target load factor as specified in the load
sequence or the maximum displacement specified on the CLPOPT input line, Collapse Advanced will
issue the following warning messages
*** WARNING: Maximum number of arc length steps has been exceeded.
*** WARNING: Arc length analysis has been terminated unsuccessfully.

55

SACS® Collapse Advanced Release 15.1

Three possible scenarios may occur depending on the Arc-length convergence rate reflected in the
troubleshooting log file:

1. The Arc-length approach has difficulty in achieving convergence at majority of the load steps
when using the automatic parameter and therefore resulting in excessive sub-incrementation
process. This indicates the automatic arc-length parameter may be too large and should be
reduced by entering a scaling factor on the ARCLEN input line –see discussion in section 4.3 on
how to choose the correct scaling factor.
2. The Arc-length method converges during few iterations (for example 3 iterations) for most of
the load steps. This indicates that the automatic arc-length parameter is too small and it may be
increased by entering a scaling factor on ARCLEN input line –see discussion in section 4.3 on
how to choose a correct scaling factor. Alternatively, the maximum number of Arc-length steps
may be increased however, this may lead to longer run times.
3. The Arc-length method achieves convergence for many load steps with a reasonable number of
iterations but without sub-incrementation. Under these circumstances, the maximum number
of Arc-length steps on the ARCLEN input line should be increased.

4.3 Post-buckling Analysis

There are two typical behaviors for post-buckling responses in offshore structures:

• Snap-through where the load factor reduces while the displacement increases.
• Snap-back where the load factor and the displacement both decrease.

Figure 11-a shows typical snap-through response where the structure jumps from one stable
configuration to another. The Relaxed Iteration option (‘IRX’ on the CLPOP2 input line) completely
prevents snap-through buckling response and ensures the analysis switches to Arc-length method for
unloading during the post-buckling analysis. This is shown in Figure 11-b (more details, see sample 2 in
section 5.1.2). To have optimal post-buckling response, the arc-length parameter (which is automatically
calculated) can be scaled by a factor on ARCLEN input line. Figure 11-c illustrates the case where the Arc-
length parameter is very small resulting in too many load steps to reach the target load factor and/or
target maximum displacement. Figure 11-d represents a case where arc-length parameter is too large
resulting in an approximate response oscillating around the optimal solution.

Figure 12-a shows typical snap-back post-buckling response where both load and displacement
decrease. A common example for this behavior is elastic buckling of braces in an offshore structure. The
key identifying feature of snap-back buckling is a sharp discontinuity in the load-displacement response
curve.

Due to the complex response of snap-back buckling, determining the optimal arc-length parameter is
extremely critical. If the arc-length parameter is too small, the algorithm cannot distinguish between
two consecutive Arc-length solutions (for more details on Arc-length method see section 6.6). In this
instance, the analysis cannot progress after the buckling load (caught in a local minimum) or it may
unload on the same path of loading.

56

SACS® Collapse Advanced Release 15.1

The following actions are recommended to improve Arc-length iteration for snap-back response:

1. Avoid using a very small load increment size. A larger load increment size will lead to a wider
search radius (i.e. larger arc-length parameter) for the Arc-length method.
2. Scale the arc-length parameter by a factor on ARCLEN input line.
3. In some instances, Relaxed Iteration procedure (IRX on CLPOP2 input line) may prevent the Arc-
length method from distinguishing between two consecutive Arc-length solutions by limiting its
search radius. Therefore, its recommended to avoid using the Relaxed Iteration option.

As per snap-through behavior, if the arc-length parameter is too large then the program will not able to
model snap-back response correctly as shown in Figure 12-b.

(a) (b)

(c) (d)

Figure 11: Different scenarios in snap-though post-buckling a) unstable path b) optimal Arc-length
solution c) very small arc-length parameter, and d) very large arc-length parameters

(a) (b)

Figure 12: a) snap-back post-buckling response with optimal arc-length parameter b) large arc-length
parameter

57

SACS® Collapse Advanced Release 15.1

4.4 Force Tolerance
The default value for the force convergence tolerance is 0.001 (kN, kips, ton depending on the model
units). The default value is determined such that the Collapse program predicts the most accurate
results. However, the default tolerance for some offshore analyses may result in a reduction in
performance through an increase in the number iterations or the number of sub-incrementation steps
without improvement in results for the following reasons:

1. Large initial residual

✓ The initial residual is approximately represented by the incremental load applied to the
model at a given load step. If the initial residual is too large, then the Collapse program
could require many iterations to reach convergence using the default tolerance.
2. Slow convergence rate with relatively small residual
✓ This scenario is characterized by a large residual reduction rate over the first few
iterations followed by a continuously reducing reduction rate later. In this case, the
Collapse program may also need many iterations (or sub-incrementations) to achieve
convergence using the default tolerance value.
3. Local minimum
✓ The occurrence of a local minimum is common in analyses exhibiting a high degree of
nonlinearity, usually where there is significant plasticity, buckling, and joint failure.
✓ In the presence of a local minimum, the residual reduction rate can be high over the
first few iterations after which the solution could start to oscillate as the residual
becomes smaller. Under these circumstances, the Collapse program may require a high
number of iterations (or sub incrementations) using the default tolerance value.

The user can review the residual values printed in the Troubleshooting Log file to see if the analysis
suffers from any one of the above issues. Figure 13 below illustrates different types of convergence
rates that may be encountered.

Figure 14 shows a sample of a convergence rate in the presence of a local minimum. As shown in the
graph, the residual drops very quickly to small value of 0.1326 from large initial value of 56526 within 5
iterations. However, the convergence slows down later and starts to oscillate due to the presence of a
local minimum. For this sample, if the force tolerance is increased to 0.1 using FRCTOL input line, the
program will achieve convergence within 10 iterations without loss of accuracy – i.e. the residual of 0.1
is still relatively very small compared to the initial residual value of 56526.

The relaxed iteration convergence option (‘CRX’ on CLPOP2 line) can also be used to achieve similar
result without overriding the default force tolerance. In this sample, the ratio of the residual at iteration
5 to the initial residual is less than 10-5(10-5 is the default value for ‘CRX’ option). For this case, the use of
the ‘CRX’ option will result in convergence being achieved at the 5th iteration.

58

SACS® Collapse Advanced Release 15.1

 Figure 13: Convergence rate samples

Large Initial residual

Rapid reduction in residual

Slow convergence rate with

local minimum oscillation

Figure 14: Collapse Advanced program residual with slow convergence rate and local minimum

59

SACS® Collapse Advanced Release 15.1

4.5 Non-convergence in Secondary Elements
To improve performance, certain secondary element types whose overall stiffness is not important to
the overall behavior of the structure can be retained as elastic throughout a Collapse analysis. For
example, the following elements may be retained as elastic for an overall pushover analysis:

1. Wishbones
2. Non-structural framing
✓ framing representing risers
✓ boat landings
✓ anodes
3. The secondary elements in the top sides for which the lateral restrains are not modelled

These elements have negligible effects on the overall performance of the structure and retaining them
as elastic will improve the performance without altering the structural response.

To verify if the non-convergence is resulting from issues associated with the secondary elements, the
user can review following in Troubleshooting Log File:

• Check if the non-convergence warning messages are associated with the secondary members
• Check if the program requires too many iterations to reach the convergence especially in early
stages of the analysis
• Check if the program requires too many sub-incrementation (or Arc-length iteration) to reach
the convergence in early stages of the analysis
• Check if analysis stops prematurely before significant deformation, plasticity, or if pile/joint
failure has occurred.

If the non-convergence is due to the secondary elements, the user can utilize MEMELA and GRPELA
input lines to designate those members and member group to remain elastic.

In some cases, assuming the secondary elements to remain elastic may not resolve all non-convergence
issues. Note, designating elements to remain elastic does NOT imply they do not experience large
deformations. For example, if the member is subjected to a localized point load (such as earthquake or
equipment load), it can still buckle elastically during the analysis. To prevent unfavorable buckling of
elastic elements, the user may use MEMSEG or GRPSEG to set a single sub-segment for the elastic
member.

Note: The program currently does not support tension only or compression only elements. These
elements should be modeled as elastic using a single sub-segment.

60

SACS® Collapse Advanced Release 15.1

5 Collapse Advanced Samples
This chapter includes detailed verification and validation benchmark samples for the Collapse Advanced
program in addition to some engineering sample problems. The first two sections provide validation of
results for problems comprising beam and plate element respectively. The results from Collapse
Advanced are compared against existing references for problems depicting different modes of behavior.
The third and fourth sections present verification of tubular local buckling and joint flexibility. The two
final sections provide a few samples associated with different features in Collapse Advanced such as ship
impact, dropped object and blast wall. SACS model file and collapse input are provided in SACS
installation folder under Collapse Advanced samples.

5.1 Beam Element Verification and Benchmarks

5.1.1 Sample 1: Elastic Behavior of a Restrained Simply Supported Beam Subjected to
Uniform Load
This benchmark highlights the ability of the program to predict the effects of membrane action in a
simply-supported beam under a uniformly distributed load, restrained against both axial and torsion
deformation at both ends, as shown in Figure 6. Results are presented for a beam with tubular and wide
flange cross sections. The material properties of the beams are based on SACS default values. Four sub-
segments per member are used and all members are set to remain elastic by entering ‘ME’ CLPOPT line.
As illustrated in Figure 16, Collapse Advanced accurately predicts nonlinear response compared to
analytical solution provided in [10]. No special option, such as sub-incrementation or Arc-length are
used for this sample.

𝐾𝑁
𝑞൬ ൰
𝑚2
𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛 𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛
𝑤(𝑐𝑚 )
𝐿 = 5𝑚 𝐿 = 5𝑚
Tubular: 40 × 2𝑐𝑚 Wide Flange: 𝑊12 × 120
Figure 15: Restrained beam with plates

Figure 16: Restrained beam mid-point vertical displacement

61

SACS® Collapse Advanced Release 15.1

5.1.2 Sample 2: Elastic Buckling of Toggle Beam
This benchmark highlights the ability of the program to predict elastic snap through buckling. Consider a
toggle beam as shown in Figure 17. The beam reaches the limit point when the beam becomes
completely flat. The toggle beam is modeled using beam elements with both tubular and a wide flange
cross sections and the results are compared with analytical solution presented in [11]. To accurately
model buckling and post-buckling behavior, both sub-incrementation and Arc-length method are
included on CLPOPT line. Four sub-segments were used to model each member and the material
properties are based on SACS default values (all members are set to elastic by entering ‘ME’ option on
CLPOPT line).

Figure 9 shows a snap-through response during post-buckling analysis which cannot be predicted using
the standard Collapse Advanced options. As discussed in sections 3.1.4.2 and 4.2.9, the Relaxed Iteration
option on the CLPOP2 input line (IRX on columns 66-68) can be used to predict the unloading behavior
during post buckling. It can be seen the results are in close agreement with the analytical solution [11].

𝑃 (𝐾𝑁)

𝐻 = 70𝑐𝑚
−𝑤(𝑐𝑚 )

𝐿 = 5𝑚 𝐿 = 5𝑚

Tubular: 40 × 2𝑐𝑚 Wide Flange: 𝑊12 × 120

Figure 17: Toggle Beam

Figure 18: Toggle beam buckling response

62

SACS® Collapse Advanced Release 15.1

5.1.3 Sample 3: Elastic Cantilever Beam
This benchmark involves an elastic cantilever beam subjected to three different end loads 1) an axial
load with small lateral load to produce buckling, 2) lateral load and 3) end moment. Details of each case
are presented in following sections.

5.1.3.1 Sample 3.1: Buckling under Axial Force

Consider an elastic cantilever subjected to a horizonal axial load at its tip. To induce elastic buckling we
𝐿
assume the tip is slightly relocated by small perturbation of 1000 in the vertical direction as shown in
Figure 19. This small perturbation may represent the effects of member self-weight/dead load and/or
any out-of-straightness. The applied loads are gradually increased until the cantilever has buckled. Both
sub-incrementation and the Arc-length method options are employed to determine buckling and post-
buckling responses. The cantilever beam is modelled using a single beam element comprised of four
sub-segments. The member is specified as elastic by entering ‘ME’ on CLPOPT line. The cantilever beam
was modeled using both tubular and a wide flange sections and the material properties were set to SACS
default values. Since this beam experiences a very large rotation, the higher order corotational terms
were included for the case with the tubular section by entering ‘ITC’ on CLPOP2 line. It is worth noting
that by default the higher order corotational terms are assumed for wide flanges and there is no need to
enter additional options. The horizontal and vertical displacements of cantilever tip show excellent
correlation with analytical solutions [12] in Figure 20 and Figure 21. Figure 22 shows beam deformation
at last load step.

𝑤𝑡𝑖𝑝 (𝑐𝑚 )
−𝑢𝑡𝑖𝑝 (𝑐𝑚 )
𝑃 (𝐾𝑁)
𝐿
𝐿 = 10𝑚 1000
Tubular: 40 × 2𝑐𝑚 Wide Flange: 𝑊12 × 120

Figure 19: Cantilever beam with axial load

Figure 20: Cantilever tubular beam tip displacement under axial load buckling

63

SACS® Collapse Advanced Release 15.1

 Figure 21: Cantilever wide flange beam tip displacement under axial load buckling

Figure 22: Deformed cantilever beam at last load step

64

SACS® Collapse Advanced Release 15.1

5.1.3.2 Sample 3.2: End Lateral Force
This benchmark involves a cantilever beam as in the previous example but with a vertical end load as
shown in Figure 23. This problem does not include large rotations and therefor higher order corotational
terms were omitted by excluding the ‘ITC’ option on CLPOP2 input line. The Arc-length method was also
not considered in this example. All other inputs are as per previous sample. The beam response shows
very good correlation when compared against existing analytical solution [13] for both tubular and wide
flange cross sections as seen in Figure 24.

−𝑢𝑡𝑖𝑝 (𝑐𝑚) 𝑤𝑡𝑖𝑝 (𝑐𝑚)

𝑃(𝐾𝑁)
𝐿 = 10𝑚

Tubular: 40 × 2𝑐𝑚 Wide Flange: 𝑊12 × 120

Figure 23: Cantilever beam with lateral load

Figure 24: Cantilever beam tip displacement under lateral load

65

SACS® Collapse Advanced Release 15.1

5.1.3.3 Sample 3.3: End Bending Moment
This benchmark involves a cantilever beam with an end bending moment. The bending moment is
gradually increased until the beam tip rotates 720˚ - i.e. beam is folded twice into two full circles. Due to
large rotation involved in this benchmark, the second order corotational terms are included in the
analysis using ‘ITC’ option on the CLPOP2 input line. To prevent lateral buckling, the beam tip is
restrained against any out-of-plane displacement. In addition, 12 sub-segments are used for this sample
because the beam will undergo very large deformation. It can be seen that the beam tip displacements
are in very good agreement with existing analytical solution [14] as seen in Figure 26. Figure 27 shows
the deformed shapes.

𝑤𝑡𝑖𝑝 (𝑐𝑚 )
−𝑢𝑡𝑖𝑝 (𝑐𝑚 )

𝑀(𝐾𝑁. 𝑚)
𝐿 = 10𝑚

Tubular: 40 × 2𝑐𝑚 Wide Flange: 𝑊12 × 120

Figure 25: Cantilever beam with end moment

Figure 26: Cantilever beam tip displacement with end moment

66

SACS® Collapse Advanced Release 15.1

Figure 27: Deformed cantilever beam. top) 360˚rotation – first circle, bottom) 720˚rotation – second
circle.

67

SACS® Collapse Advanced Release 15.1

5.1.4 Sample 4: Elastoplastic Cantilever with Tubular Section
This sample involves an elastoplastic analysis of a cantilever beam with a tubular cross section subject to
a vertical end load as shown in Figure 28. The cantilever was modeled using a single beam element
comprised of 8 sub-segments. The material properties were assumed elastic-perfectly plastic and were
modelled using two different yield stresses as shown in figure 19. In this benchmark, the beam
experiences large rotations and deformations due to the applied end load. Therefore, both sub-
incrementation and arc-length options are included in the analysis by entering ‘SI’ and ‘AL’ on the
CLPOPT input line. Collapse Advanced results are compared with numerical results presented in
references [15], [16] for the vertical displacement at cantilever tip as shown in Figure 29.

𝑃 (𝐾𝑁)

𝑤(𝑐𝑚 )
𝐿 = 4𝑚
𝐾𝑁
𝐸 = 20000 , 𝐷 = 35.546 𝑐𝑚 𝑡𝑖𝑐ℎ𝑘𝑛𝑒𝑠𝑠 = 0.38
𝑐𝑚 2
𝐾𝑁 𝐾𝑁
𝐹𝑦 = 825 2
𝑎𝑛𝑑 1650 𝑒𝑙𝑎𝑠𝑡𝑖𝑐 − 𝑝𝑒𝑟𝑓𝑒𝑐𝑡𝑙𝑦 𝑝𝑙𝑎𝑠𝑡𝑖𝑐
𝑐𝑚 𝑐𝑚 2

Figure 28: Elastoplastic cantilever with tubular section

Figure 29: Vertical displacement of cantilever tip

68

SACS® Collapse Advanced Release 15.1

5.1.5 Sample 5: Elastoplastic Buckling of Columns
The benchmarks included in this section depict typical elastoplastic buckling behavior of columns.
The first set of benchmarks show the ability of the program to predict elastoplastic buckling behavior of
simple columns with a tubular cross section using Euler-Bernoulli bending theory. The second set of
benchmarks assumes Timoshenko bending theory to account for shear deformation. The third set
compares the elastoplastic buckling loads against the API design code for both wide flange and tubular
section columns.

5.1.5.1 Sample 5.1: Elastoplastic of Tubular Simple Column with Euler-Bernoulli Theory
This benchmark involves the elastoplastic buckling of simply-supported tubular column with various
𝐿
slenderness 𝑟 values: 80, 120, 160. The column is torsionally restrained at both ends and has a small
imperfection in the form of mid-point perturbation as shown in Figure 30. The imperfection is defined as
𝛿
𝑒 = 𝐿 with values of 0.0001, 0.001, 0.01, and 0.05. The column was modelled using 2 elements each
comprised of 8 sub-segments. To account for buckling behavior, both sub-incrementation procedure
and the arc-length method were included by selecting ‘SI’ and ‘AL’ options on the CLPOPT input line. To
improve the analysis, the arc-length parameter is increased (columns 24-29 of ARCLEN line) while the
maximum number of arc-length iterations is reduced (columns 12-18 of ARCLEN line).

The axial response (horizontal displacement at the free end) and lateral deformation (vertical
displacement at mid-point) as shown in Figure 30 and are compared with existing finite element solution
[17] as shown in Figure 31 to Figure 33. It is seen that Collapse Advanced accurately predicts the
elastoplastic response for all cases.

Note: Collapse Advanced analysis is carried out for very large deformation. For comparison purposes,
only the initial portion of the response are shown below.

Note: In Figure 31, the horizontal axis is the initial imperfection (i.e. 𝛿) plus the lateral displacement. For
other cases, the horizontal axis is the lateral displacement.

𝑤(𝑖𝑛𝑐ℎ )
−𝑢 (𝑖𝑛𝑐ℎ ) 𝑃(𝑘𝑖𝑝𝑠)

𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛 𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛

𝛿 = 𝑒𝐿

Tubular: 4.5 × 0.09375 𝑖𝑛𝑐ℎ, 𝐸 = 29000, 𝐹𝑦 = 36 𝐾𝑆𝐼

Figure 30: Simple column with imperfection

69

SACS® Collapse Advanced Release 15.1

 𝐿
Figure 31: Euler-Bernoulli column response for 𝑟 = 80.

𝐿
Figure 32: Euler-Bernoulli column response for = 120
𝑟

𝐿
Figure 33: Euler-Bernoulli column response for 𝑟 = 160

5.1.5.2 Sample 5.2: Elastoplastic of Tubular Simple Column with Timoshenko Bending Theory
This benchmark is same as sample 5.1 but with the inclusion of shear deformation effects in accordance
to Timoshenko’s bending theory. The imperfection values assumed were 0.00001, 0.0001, 0.001 and
0.01, the shear module was assumed to be 𝐾𝑆𝐼 𝐺 = 1115 𝐾𝑆𝐼. The results were compared with
existing numerical solution [18] in following figures and are seen to be in very good agreement.

Note: Collapse Advanced analysis is carried out for very large deformation. For comparison purposes,
only the initial portion of the response are shown below.

70

SACS® Collapse Advanced Release 15.1

 𝐿
Figure 34: Timoshenko column response for 𝑟 = 80

𝐿
Figure 35: Timoshenko column response for 𝑟 = 120

𝐿
Figure 36: Timoshenko column response for 𝑟 = 160

71

SACS® Collapse Advanced Release 15.1

5.1.5.3 Comparison of Elastoplastic Buckling of Columns with API
Similar analyses were repeated for various tubular and wide flange sections with elastoplastic material
and with various slenderness values. The stresses at buckling load were compared with critical axial
stress provided in API design code without the safety factor in Figure 37 and Figure 38. It can be seen in
that the program predicts buckling stress very accurately for all cases.

Tubular Simple Column

30
Buckling Stress (KN/cm^2)

20

API (without safety factor)

10
24X0.375 (Collapse Advanced)
36X0.375 (Collapse Advanced)
48X0.375 (Collapse Advanced)
Elastic/Inelastic Limit
0
0 50 100 150 200

KL/r

Figure 37: Collapse Advanced buckling stress versus API for tubular columns

Wide Flange Simple Column

30
Buckling Stress (KN/cm^2)

20

API (without safety factor)

10
W12X120 (Collapse Advanced)
W24X192 (Collapse Advanced)
W36X302 (Collapse Advanced)
Elastic/Inelastic Limit
0
0 50 100 150 200

KL/r

Figure 38: Collapse Advanced buckling stress versus API for wide flange columns

72

SACS® Collapse Advanced Release 15.1

5.1.6 Sample 6: Buckling of Wide Flange Sections
This benchmark involves of buckling analysis of wide flange beam and columns with the inclusion of
Wagner’s nonlinear strains for torsional deformation. The first benchmark looks at elastic torsional
buckling of a wide flange column. The second part compares the elastoplastic lateral-torsional buckling
of a simple beam subjected to pure bending with AISC 2010.

5.1.6.1 Sample 6.1 and Sample 6.2: Lateral-Torsional Buckling of Wide Flange Columns with
Wagner Nonlinear Strain
This benchmark demonstrates application of Wagner nonlinear strains for the prediction elastic lateral-
torsional buckling of beams with wide flange sections. Wagner strains have negligible role in majority of
standard analysis where beams comprised of open thin walled sections are restrained against lateral
buckling. However, these beams may undergo large torsional deformations resulting from accidental
loading resulting from a blast or a dropped object, in cases such as these, the contribution of Wagner’s
second-order strain may be significant.

Figure 30 shows a simply-supported beam subjected to the axial load with eccentricity. The beam is
restrained against torsion at both ends. The following wide flange sections sizes were considered for the
analysis: W12×120, W24×192, and W36×302. The analysis was conducted for three beam lengths 𝐿 =
𝐻 𝐻 𝐻 𝐻
10, 20, 𝑎𝑛𝑑 30, 𝑓𝑡. The axial load eccentricity was defined as 𝑒 = 0, 16 , 8 , 4 , 2 , 𝐻 where 𝐻 is the section
height. The material was retained elastic with SACS default values. The beam was modelled using two
elements each one comprised of 8 sub-segments. To induce lateral-torsional buckling, an imperfection
in the form of small cord angle was applied to members to represent lateral loads. The axial load was
gradually increased until there was significant deformation in lateral direction. A very fine load step size
is used to precisely determine buckling load. Both sub-incrementation and Arc-length methods were
included in the analysis.
Wagner nonlinear torsional strains were included by entering ‘IWT’ on CLPOP2 input line and the results
were compared with analytical solution [19] in Figures 31 to 33. The results agree with existing analytical
solution [19] for all cases. We can also see that the inclusion of the Wagner nonlinear strains is essential
where there is significant torsional rotation (i.e. beam with higher aspect ratio such as W36×302). For
beams which are restrained laterally, the Wagner nonlinear strain has a minimal effect.

𝑃 𝑃
𝑒 𝑒

𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛 𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛

𝑠𝑚𝑎𝑙𝑙 𝑖𝑚𝑝𝑒𝑟𝑓𝑒𝑐𝑡𝑖𝑜𝑛 𝑎𝑙𝑜𝑛𝑔 𝑤𝑖𝑑𝑒 𝑓𝑎𝑙𝑛𝑔𝑒 𝑤𝑒𝑎𝑘 𝑎𝑥𝑖𝑠

𝑒𝑙𝑎𝑠𝑡𝑖𝑐 − 𝑝𝑒𝑟𝑓𝑒𝑐𝑡𝑙𝑦 𝑝𝑙𝑎𝑠𝑡𝑖𝑐 𝑤𝑖𝑡ℎ 𝑆𝐴𝐶𝑆 𝑑𝑒𝑓𝑎𝑢𝑙𝑡 𝑣𝑎𝑙𝑢𝑒𝑠

Figure 39: Simple column under axial load with eccentricity

73

SACS® Collapse Advanced Release 15.1

 Figure 40: Wagner nonlinear torsion effect in W12x120 buckling

74

SACS® Collapse Advanced Release 15.1

 Figure 41: Wagner nonlinear torsion effect in W24x912 buckling

75

SACS® Collapse Advanced Release 15.1

 Figure 42: Wagner nonlinear torsion effect in W36x302 buckling

76

SACS® Collapse Advanced Release 15.1

5.1.6.2 Comparison of Elastoplastic Lateral-Torsional Buckling of Wide Flange under Pure
Bending with AISC 2010
Similar analyses were repeated for a simply supported beam with elastoplastic material subject to the
pure bending (see Figure 43). The beam is restrained against torsion at both ends. Three beam sections
were considered: W12×120, W24×192, and W36×302. Beams with lengths 10ft, 20ft, 30ft, 60ft, and 90ft
were considered. The predicted ultimate bending moments were compared with ultimate bending
moment given by AISC 2010 with safety removed in Figure 44. The results predicted by the program are
in good agreement with AISC 2010 for all cases.

𝑀 𝑀

𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛 𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛

𝑠𝑚𝑎𝑙𝑙 𝑖𝑚𝑝𝑒𝑟𝑓𝑒𝑐𝑡𝑖𝑜𝑛 𝑎𝑙𝑜𝑛𝑔 𝑤𝑖𝑑𝑒 𝑓𝑎𝑙𝑛𝑔𝑒 𝑤𝑒𝑎𝑘 𝑎𝑥𝑖𝑠

𝑒𝑙𝑎𝑠𝑡𝑖𝑐 − 𝑝𝑒𝑟𝑓𝑒𝑐𝑡𝑙𝑦 𝑝𝑙𝑎𝑠𝑡𝑖𝑐 𝑤𝑖𝑡ℎ 𝑆𝐴𝐶𝑆 𝑑𝑒𝑓𝑎𝑢𝑙𝑡 𝑣𝑎𝑙𝑢𝑒𝑠
Figure 43: Simple beam under pure bending

Figure 44: Comparison of Collapse Advanced ultimate moment and AISC 2010

77

SACS® Collapse Advanced Release 15.1

5.1.7 Sample 7: Harrison’s Space Beam with Tubular Sections
Figure 36 shows Harrisons [20] space beam with tubular sections. The material properties are
considered to elastic-perfectly plastic. The beam was subjected to vertical and horizontal loads as shown
in Figure 45. Each element used to model the structure was comprised of 8 sub-segments. Both sub-
incrementation and Arc-length options were used on the CLPOPT input line and the maximum deflection
(collapse deflection) was set to 10 inches based on the reference results. The predicted response was
compared against existing analytical and experimental results [20]–[22] as shown in Figure 46. The
results predicted by the program are in very good agreement for all cases.

2𝑓𝑡

1.3𝑃(𝑘𝑖𝑝𝑠) 1𝑃(𝑘𝑖𝑝𝑠)
𝑢 (𝑖𝑛𝑐ℎ )

4𝑓𝑡 𝐸 = 28800
𝐶𝑜𝑙𝑢𝑚𝑛 𝐹𝑦 = 30.60 𝐾𝑆𝐼

𝐵𝑒𝑎𝑚 𝐹𝑦 = 31.10 𝐾𝑆𝐼

4𝑓𝑡

Figure 45: Harrison’s space beam with tubular section

Figure 46: Left) Harrison’s space beam response. Right) plastic contour at last load step

78

SACS® Collapse Advanced Release 15.1

5.1.8 Sample 8: Elastoplastic Portal Beams
Figure 47 shows a portal beam comprised of wide flange section W8×31. The portal beam was subjected
to vertical loads applied at top of columns in addition to a horizontal load as shown in Figure 47. The
material properties were assumed to be elastic-perfectly plastic. Three levels of vertical load 𝑉 were
considered, 0.2𝑃𝑦 , 0.4𝑃𝑦 , 0.6𝑃𝑦 where 𝑃𝑦 is the column yield load. For the analysis, the vertical load 𝑉
was applied first using a single load step. The horizontal load 𝐻 was then applied gradually until collapse
occurred. The sub-incrementation and Arc-length options were selected on the CLPOPT input line to
predict the ultimate load and the unloading path. It was found that the analysis continued using Arc-
length iterative scheme until the lateral displacement measured at the top of the columns reached
0.04𝐿 = 14.12 𝑐𝑚 for 𝑉 = 0.2𝑃𝑦 and 𝑉 = 0.4𝑃𝑦 , and 0.0175𝐿 = 6.18 𝑐𝑚 for 𝑉 = 0.6𝑃𝑦 . The Arc-length
parameter was subsequently increased by using the ARCLEN input line which resulted 0.04𝐿
displacement using fewer Arc-length steps. The resulting beam horizontal response is in good
agreement with existing studies [21], [23] as shown in Figure 48.
𝑉 𝑉
𝐻 𝑢 (𝑐𝑚 )

𝐿 = 3.53𝑚
𝐾𝑁
𝐸 = 19990
𝑐𝑚 2
𝐿 𝐾𝑁
𝐹𝑦 = 24.82
𝑐𝑚 2
𝑒𝑎𝑙𝑎𝑠𝑡𝑖𝑐 − 𝑝𝑒𝑓𝑒𝑐𝑡𝑙𝑦 𝑝𝑙𝑎𝑠𝑡𝑖𝑐

Figure 47: Portal Beam

Figure 48: Portal beam response in different vertical load levels

79

SACS® Collapse Advanced Release 15.1

5.1.9 Sample 9: Space Beam with Wide Flange
Figure 49 shows a 3D space beam comprised of wide flange members subjected to vertical and lateral
loads. For this analysis the load 𝑃 was gradually increased until the structure attained its ultimate load.
The non-symmetrical lateral loads resulted in a failure mode which include both torsional and lateral
deformation. Each element used to model the structure was comprised of 8 sub-segments and both sub-
incrementation and Arc-length options were included. To compare against reference studies, the
maximum deflection (collapse deflection on CLPOPT) was set to 5 cm. The horizontal response of the
model at top level was compared with various existing numerical solutions [24], [25] and are in good
agreement as shown in Figure 50.

4𝑃
4𝑃 𝐿
𝑃
4𝑃
𝑢 (𝑐𝑚 )
𝐾𝑁
𝐿 𝐸 = 21000
𝑐𝑚 2
2𝑃 𝐾𝑁
𝐺 = 8000
2𝑃 0.5𝑃 𝑐𝑚 2

2𝑃 𝐾𝑁
𝐹𝑦 = 24.83
𝑐𝑚 2

𝐿 𝑒𝑎𝑙𝑎𝑠𝑡𝑖𝑐 − 𝑝𝑒𝑓𝑒𝑐𝑡𝑙𝑦 𝑝𝑙𝑎𝑠𝑡𝑖𝑐

Figure 49: Space beam with wide flange

Figure 50: Left) Space beam horizontal response. Right) plastic contour at last load step

80

SACS® Collapse Advanced Release 15.1

5.1.10 Sample 10: Six-story Building with Wide Flange Section
Figure 51 shows a six-story building constructed using wide flange members subjected to gravity and
𝐾𝑁
lateral loads. The applied gravity load is equivalent to a uniformly distributed load of 9.6 2 and is
𝑚
applied using point loads at the top of all columns. The lateral load is applied as a series of point loads
with amplitude of 53.376𝐾𝑁 on all joints on the front face of the building. The material is assumed to
be elastic – perfectly plastic. For the nonlinear analysis, the magnitude of the gravity and lateral loads
was increased gradually until the model attained its ultimate capacity. The Arc-length method was used
to predict the post-buckling response of the building until the maximum joint displacement reached
2.5𝑚. The analysis was conducted using both Euler-Bernoulli (without shear deformation effect) and
Timoshenko (with shear deformation effects) beam theory. The results were compared with various
references [16, 23] including DNV’s USFOS program as shown in Figure 52. The plastic ratio contours at
last load step is shown in Figure 53. It is worth noting that the reference studies stop the analysis when
the model reaches the ultimate load while results from Collapse Advanced continue into the post-
buckling phase.

𝒗 𝐾𝑁
𝐸 = 20000
𝑐𝑚 2
𝐾𝑁
𝐿 𝐺 = 7930
𝑐𝑚 2
𝐾𝑁
𝐹𝑦 = 25
𝑐𝑚 2
𝑒𝑎𝑙𝑎𝑠𝑡𝑖𝑐 − 𝑝𝑒𝑓𝑒𝑐𝑡𝑙𝑦 𝑝𝑙𝑎𝑠𝑡𝑖𝑐

Figure 51: Six-story building model

81

SACS® Collapse Advanced Release 15.1

 Figure 52: Six-story building lateral response

Figure 53: Plastic ratio contour for Six-story building at last load step
82

SACS® Collapse Advanced Release 15.1

The 2nd-order corotational terms can be ignored to reduce run time (see section 6.3.1.5 for details on
corotational terms). These terms are excluded by entering ‘EXC’ on CLPOPT input line. The results in
Figure 54 show neglecting these terms does not affect the prediction of the ultimate load in this case,
even though the run time is reduced. It is also noted that if the 2nd order terms are neglected. the
program has difficulty converging in the vicinity of the limit point (i.e. finer load and Arc-length steps are
used) resulting from significant deformation of the members. In addition, the Arc-length method also
fails to converge prior to reaching the target displacement of 2.5m.

Figure 54: Comparison of model response with and without 2nd-order corotational terms

For the sample considered, excluding 2nd-order corotational terms significantly reduces run time without
compromising accuracy of the ultimate load capacity. To ensure accuracy, the following criterion should
apply when excluding 2nd order terms for wide flange or other open thin-walled sections:

• Wide flange or other open thin-walled sections do not experience large loads or significant
plasticity; Or
• Wide flange or other open thin-walled sections do not experience large deformation – i.e. they
are restrained against buckling; Or
• Wide flange or other open thin-walled sections do not have major contribution to the model
stiffness – basically they are secondary members.

83

SACS® Collapse Advanced Release 15.1

5.1.11 Sample 11: Buckling of Hinged Right-angle Beam
Figure 55 shows a 2D right-angle beam comprised of members with rectangular cross sections. The
beam is hinged to the ground and is subjected to a vertical point load. The analysis is conducted
assuming both elastic and elastoplastic material properties with strain hardening ratio of 0.1. The sub-
incrementation and Arc-length methods are selected on CLPOPT line to accurately determine limit
points (i.e. buckling load) and the post-buckling response. Each element is comprised of 4 sub-segments
to ensure the same number of finite elements are used as reference studies [16], [26]. The beam overall
response is compared against [16] and [26] for both elastic and elastoplastic material properties as
shown in Figure 56 and Figure 57, respectively. The deformed shapes are also depicted in Figure 56 and
Figure 57. The elastoplastic analysis is continued beyond the limits presented in the reference studies to
demonstrate post-buckling behavior. The results indicate that results from the program are in very good
agreement with the reference studies for both elastic and elastoplastic response.

1 4
𝐿 𝐿
5 5

𝑃 (𝐾𝑁)
ℎ
𝑢(𝑐𝑚 ) 𝑏
−𝑤(𝑐𝑚) 𝐿 = 1.2𝑚, 𝑏 = 3𝑐𝑚, ℎ = 2𝑐𝑚
𝐿
𝐾𝑁
Elastic: 𝐸 = 7200 𝑐𝑚2
𝐾𝑁 𝐾𝑁
Elastoplastic: 𝐸 = 7061 𝑐𝑚2 𝐹𝑦 = 102 𝑐𝑚2

𝑆𝑡𝑟𝑎𝑖𝑛 ℎ𝑎𝑟𝑑𝑒𝑛𝑖𝑛𝑔 𝑟𝑎𝑡𝑖𝑜 = 0.1

Figure 55: 2D hinged right angle beam

84

SACS® Collapse Advanced Release 15.1

 Buckling Load
Collapse Advanced 18.5
Simo and Vu-Quoc (1986) 18.532
Alemdar and White (2005) 18.567

Figure 56: Beam overall response, deformed shape, and buckling load (the load at the first peak – limit
point) for elastic material

Figure 57: Beam overall response and deformed shape for elastoplastic material

85

SACS® Collapse Advanced Release 15.1

5.1.12 Sample 12: Two-story 3D Beam with Rectangular Cross-Sections
Figure 49 shows a two-story 3D beam comprised of members with rectangular cross sections. Each
element was comprised of 8 sub-segments. The beam was subjected to horizontal and vertical loads
which were gradually increases until the maximum displacement reached 1m. The material was
assumed to be elastic-perfectly-plastic. Both sub-incrementation and Arc-length method were selected
on the CLPOPT input line to determine the limit point, and to study the behavior of the structure for
large deformation after the ultimate load has been reached. The overall response of the structure was
compared with existing studies [16] as shown in Figure 59. It can be seen the results from the program
are in very good agreement.

𝑃 ⁄2
3𝑃 𝑢(𝑐𝑚 ) 𝑃 ⁄2
𝑃 ⁄2 𝑃 𝑃
𝑃 ⁄2 𝑃 ⁄2
3𝑃 𝐶𝑜𝑙𝑢𝑚𝑛𝑠 𝐵𝑒𝑎𝑚𝑠
4𝑚 𝑃 ⁄2
2𝑃 20 𝑐𝑚 40 𝑐𝑚
𝑃
2𝑃 40 𝑐𝑚 20 𝑐𝑚
𝑃 2𝑃 𝑃 ⁄4
2𝑃

4𝑚 𝑃 ⁄4 𝐾𝑁 𝐾𝑁
𝐸 = 1961 2
, 𝐺 = 838
𝑐𝑚 𝑐𝑚 2
𝐾𝑁
𝐹𝑦 = 9.8 𝑆𝑡𝑟𝑎𝑖𝑛 ℎ𝑎𝑟𝑑𝑒𝑛𝑖𝑛𝑔 𝑟𝑎𝑡𝑖𝑜 = 0.0
3𝑚 𝑐𝑚 2
4𝑚
Figure 58: 3D two-story beam with rectangular sections hinged right angle beam

Figure 59: 3D two-story beam response and plastic ratio at maximum load factor

86

SACS® Collapse Advanced Release 15.1

5.1.13 Sample 13: Buckling of a Beam Dome with Rectangular Cross-section
Figure 60 shows a 3D space beam with members comprised of rectangular cross sections. For this
benchmark, both elastic and elastoplastic material properties were considered. For the elastoplastic
case, the structure was modeled using 8 sub-segments per element and was subjected to a vertical point
load applied at the apex. The load was gradually increased to study the buckling and post buckling
behavior. Both sub-incrementation and Arc-length method were employed to determine the ultimate
load and post-buckling response for the elastoplastic case. The vertical displacement of the structure at
the apex was compared against existing results [16], [18] as shown in Figure 61 and Figure 62,
respectively.

𝑃 = 124 𝑀𝑁
𝐴𝑙𝑙 𝑚𝑚𝑒𝑏𝑒𝑟

−𝑤(𝑚 )
122 𝑐𝑚
76 𝑐𝑚

𝐾𝑁 𝐾𝑁 𝐾𝑁
𝐸 = 2069 2
, 𝐺 = 883 2
𝐹𝑦 = 8 2 𝑆𝑡𝑟𝑎𝑖𝑛 ℎ𝑎𝑟𝑑𝑒𝑛𝑖𝑛𝑔 𝑟𝑎𝑡𝑖𝑜 = 0.0
𝑐𝑚 𝑐𝑚 𝑐𝑚

Figure 60: 3D beam dome subjected to a vertical load

Figure 61: Elastic response of the beam dome and deformed shape at last load increment

Figure 62: Elastoplastic response of the beam dome and the plastic ratio
87

SACS® Collapse Advanced Release 15.1

5.1.14 Sample 14: Collapse Analysis of 2D Frame with Tubular Braces
Figure 63 shows a 2D braced frame which has been extensively studied both experimentally and
numerically by Billington Osborne-Moss Engineering Ltd (BOMEL) [27]. The member section properties
are shown in Figure 63. The material properties were determined from the data provided in reference
[27]. The frame was analyzed with and without joint flexibility (‘JF’ option on CLPOPT line) and the
response was compared against experimental results [27]. Each beam element was comprised of 8 sub-
segments. Both sub-incrementation and Arc-length methods were selected on the CLPOPT input line to
help predict the buckling and post buckling behavior. The frame legs were restrained against out-of-
plane displacement. In addition, an out of straightness of 1⁄1000 was assumed for all members as per
CLPOP2 input line. Analysis was conducted until the top of the beam achieved a maximum lateral
deflection of 16 cm. A larger value for the arc-length factor on ARCLEN was used to ensure that the
analysis reaches the maximum displacement value using fewer Arc-length steps. Figure 64 illustrates
that the program accurately predicts the ultimate beam capacity and the post-failure response.

𝑢(𝑐𝑚)

Figure 63: 2D Beam subjected to lateral load. Left) the model and member sections, Right) deformed
beam at final load step with plastic ratio

Figure 64: 2D Beam overall response with and without joint flexibility
88

SACS® Collapse Advanced Release 15.1

5.1.15 Sample 15: Four-leg Jackets
Figure 56 shows four-leg jacket model which has been extensively studied in reference [28] and USFOS
Verification Document [29]. The material of the structure was assumed to be elastic-perfect plastic with
𝐾𝑁
yield stress of 33 𝑐𝑚2 and section properties were selected as per Model S1 in reference [28]. Three
applied loads were considered:

1) Four vertical nodal loads applied at top joints to model the structure self-weight,

2) Distributed member load to represent wave loading

3) Four horizontal nodal loads applied to top joints to determine the ultimate lateral loads.

For the load sequence, the vertical and member load were applied first. The horizontal load was then
applied and gradually increased until the jacket reached its ultimate capacity.

Different scenarios were considered to study the effect of damage to the structure as summarized in
Table 3. Both sub-incrementation and Arc-length options on the CLPOPT input line were selected to help
predict the response up to and slightly after the limit point. For comparison purposes, the maximum
deflection is entered on CLPOPT based on the results given in reference studies, although Collapse
Advanced analysis can be carried out for larger deformations. Each beam element was comprised of 8
sub-segments in addition to an out of straightness of 0.0015 as per[28]. The results predicted by the
program are compared against [28] (cases 1 to 7) and USFOS (cases 6 and 7) in Figure 66. USFOS results
are given in USFOS Verification Document. It can be seen Collapse Advanced results are in good
agreement for all cases.

Table 3: Four-leg jacket analysis cases

Case Load Type out-of-straightness Description

1 Nodal No Undamaged Structure
2 Nodal Yes Undamaged Structure
3 Nodal Yes Member EI (compression brace) damaged
4 Nodal Yes Member EI (compression brace) removed
5 Nodal Yes Member IF (tension brace) removed
6 Nodal + Member Yes Undamaged Structure
7 Nodal + Member Yes Member IF (tension brace) removed

89

SACS® Collapse Advanced Release 15.1

 𝑢(𝑐𝑚 )

Figure 65: Four-leg jacket model and loading

Figure 66: Four-leg Jacket response in various cases

Figure 67: Deformed jacket with plastic ratio contours for case 1 (left) and case 7 (right)

90

SACS® Collapse Advanced Release 15.1

5.1.16 Sample 16: Tee Section
Following benchmarks illustrate the ability of the Collapse Advanced to predict the buckling load (both
flexural and torsional buckling) of beams and columns with Tee section. The samples are 1) a simple
elastic beam subjected to the pure bending to model lateral torsional buckling of Tee sections and 2) a
simple column under the axial compression load to model elastoplastic flexural buckling.

5.1.16.1 Sample 16.1: Elastic lateral torsional buckling of simple beam with Tee section
This sample involves elastic lateral-torsional buckling of a simply supported Tee beam subjected to the
pure bending as shown in Figure 68. The beam is restrained against torsion at both ends and the
material properties are set to SACS default values. The model consists of two elements with eight sub-
segments per member and all members are set to remain elastic by entering ‘ME’ CLPOPT line. To
induce lateral-torsional buckling, an imperfection in the form of small lateral dislocation was applied to
mid-point representing lateral loads. The beam experience large lateral deformation during bucking.
Therefore, both sub- incrementation and arc-length options are included in the analysis by entering ‘SI’
and ‘AL’ on the CLPOPT input line.

The simulation is carried out for following section width to height ratios: 0.5, 0.6, 0.7, 0.8, 0.9, and 1.0.
Since the Tee section is monosymmetric, both positive bending moment (i.e. the flange in compression)
and negative bending moment (the flange in tension) have been considered. Also, the analysis has been
carried out with and without Wagner strain (‘IWT’ on CLPOP2 line). The resulting buckling bending
moments are given in where they are in good agreement with analytical solution given in [19].

±𝑀 𝐿 = 5𝑚 ±𝑀

𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛 𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛

𝑠𝑚𝑎𝑙𝑙 𝑖𝑚𝑝𝑒𝑟𝑓𝑒𝑐𝑡𝑖𝑜𝑛 𝑎𝑙𝑜𝑛𝑔 𝑇𝑒𝑒 𝑠𝑒𝑐𝑡𝑖𝑜𝑛 ℎ𝑜𝑟𝑖𝑧𝑜𝑛𝑡𝑎𝑙 𝑎𝑥𝑖𝑠

𝑊 (cm)

t = 1 cm (both flange and web)

𝐻 = 20 𝑐𝑚
𝐻 (cm)
𝑊
= 0.5 0.6 0.7 0.8 0.9 1.0
𝐻

Figure 68: Simple beam subjected to the pure bending

91

SACS® Collapse Advanced Release 15.1

 Figure 69: Collapse Advanced buckling moment for Tee section versus the analytical solution [19].

5.1.16.2 Sample 16.2: Elastoplastic flexural buckling of simple column with Tee section
This benchmark focuses on elastoplastic flexural buckling analysis of members with Tee section under
compression load. Figure 70 illustrates a simply-supported column subjected to axial compression load
at one end. The column is restrained against torsion at both ends and has a small imperfection in the
form of mid-point perturbation as lateral dislocation. The mid-point lateral dislocation is 1/10th of Tee
section radius of gyration about the local z-axis. The model consists of two elements with eight sub-
segments per members and the material properties are set to SACS default values. Both sub-
incrementation option (‘SI’ on CLPOPT) and arc-length method (‘AL’ on CLPOPT) are utilized for this
benchmark.

The analysis is carried out for following section width to height ratios: 0.5, 0.6, 0.7, 0.8, 0.9, and 1.0 and
results are presented in Figure 71 and Figure 72. Figure 71 illustrates the lateral response of the columns
with different Tee section. Figure 72 compares buckling loads – the peak load for mid-point deflection
curves in Figure 71 – with ultimate load provided in Section E of AISC 360-16 Specification. The results
show that Collapse Advanced accurately predicts the elastoplastic response for all cases.

92

SACS® Collapse Advanced Release 15.1

 𝐿 = 2.5𝑚 𝑃(𝑘𝑁)

𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛 𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛

𝑟𝑧
𝑖𝑚𝑝𝑒𝑟𝑓𝑒𝑐𝑡𝑖𝑜𝑛 𝑎𝑙𝑜𝑛𝑔 𝑇𝑒𝑒 𝑠𝑒𝑐𝑡𝑖𝑜𝑛 ℎ𝑜𝑟𝑖𝑧𝑜𝑛𝑡𝑎𝑙 𝑎𝑥𝑖𝑠
10
𝑊 (cm)

t = 1 cm (both flange and web)

𝐻 = 20 𝑐𝑚
𝐻 (cm)
𝑊
= 0.5 0.6 0.7 0.8 0.9 1.0
𝐻

Figure 70: Simple columns with Tee section

Figure 71: Lateral displacement of the mid-point for various width to height ratio

93

SACS® Collapse Advanced Release 15.1

 Figure 72: Comparison of elastoplastic buckling load with AISC 360-16

5.1.17 Sample 17: Channel Section

Following benchmarks illustrate the ability of the Collapse Advanced to predict the buckling load (both
flexural and torsional buckling) of beams and columns with Channel section. The samples are 1) a simple
elastic beam subjected to the pure bending to model lateral torsional buckling of Channel sections and
2) a simple column under the axial compression load to model elastoplastic flexural buckling.

5.1.17.1 Sample 17.1: Elastic lateral torsional buckling of simple beam with Channel section
This sample involves elastic lateral-torsional buckling of a simply supported Channel beam subjected to
the pure bending as shown in Figure 73. The beam is restrained against torsion at both ends and the
material properties are set to SACS default values. The model consists of two elements with eight sub-
segments per member and all members are set to remain elastic by entering ‘ME’ CLPOPT line. To
induce lateral-torsional buckling, an imperfection in the form of small lateral dislocation was applied to
mid-point representing lateral loads. The beam experience large lateral deformation during bucking.
Therefore, both sub- incrementation and arc-length options are included in the analysis by entering ‘SI’
and ‘AL’ on the CLPOPT input line.

The simulation is carried out for following section width to height ratios: 0.5, 0.6, 0.7, 0.8, 0.9, and 1.0.
The mid-point lateral deformation is plotted for all cases in Figure 74. The resulting buckling bending
moments are compared with analytical solution derived from [19] and Figure 75 shows they are in good
agreement with the analytical solution.

94

SACS® Collapse Advanced Release 15.1

 𝑀 𝐿 = 5𝑚 𝑀

𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛 𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛

𝑠𝑚𝑎𝑙𝑙 𝑖𝑚𝑝𝑒𝑟𝑓𝑒𝑐𝑡𝑖𝑜𝑛 𝑎𝑙𝑜𝑛𝑔 𝐶ℎ𝑎𝑛𝑛𝑒𝑙 𝑠𝑒𝑐𝑡𝑖𝑜𝑛 ℎ𝑜𝑟𝑖𝑧𝑜𝑛𝑡𝑎𝑙 𝑎𝑥𝑖𝑠

𝑊 (cm)

t = 1 cm (both flange and web)

𝐻 = 20 𝑐𝑚
𝐻 (cm)
𝑊
= 0.5 0.6 0.7 0.8 0.9 1.0
𝐻

Figure 73: Simple beam subjected to the pure bending

Figure 74: Mid-point lateral deformation for Channel section

95

SACS® Collapse Advanced Release 15.1

 Elastic Lateral-Torsional Buckling of Channel Section Simple Beam
300
Analytical
Collapse Advanced

Buckling Moment (kN.m) 200

100

0
0.4 0.5 0.6 0.7 0.8 0.9 1 1.1

Width to Height Ratio

Figure 75: Collapse Advanced buckling moment for Channel section versus the analytical solution [19].

5.1.17.2 Sample 17.2: Elastoplastic flexural buckling of simple column with Channel section
This benchmark focuses on elastoplastic flexural buckling analysis of members with Channel section
under compression load. Figure 76 illustrates a simply-supported column subjected to axial compression
load at one end. The column is restrained against torsion at both ends and has a small imperfection in
the form of mid-point perturbation as lateral dislocation. The mid-point lateral dislocation is 1/10th of
Channel section radius of gyration about the local z-axis. The model consists of two elements with eight
sub-segments per members and the material properties are set to SACS default values. Both sub-
incrementation option (‘SI’ on CLPOPT) and arc-length method (‘AL’ on CLPOPT) are utilized for this
benchmark.

The analysis is carried out for following section width to height ratios: 0.5, 0.6, 0.7, 0.8, 0.9, and 1.0 and
results are presented in Figure 77 and Figure 78. Figure 77 illustrates the lateral response of the columns
with different Channel section. Figure 78 compares buckling loads – the peak load for mid-point
deflection curves in Figure 77 – with ultimate load provided in Section E of AISC 360-16 Specification.
The results show that Collapse Advanced accurately predicts the elastoplastic response for all cases.

96

SACS® Collapse Advanced Release 15.1

 𝐿 = 2.5𝑚 𝑃(𝑘𝑁)

𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛 𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛

𝑟𝑧
𝑖𝑚𝑝𝑒𝑟𝑓𝑒𝑐𝑡𝑖𝑜𝑛 𝑎𝑙𝑜𝑛𝑔 𝐶ℎ𝑎𝑛𝑛𝑒𝑙 𝑠𝑒𝑐𝑡𝑖𝑜𝑛 ℎ𝑜𝑟𝑖𝑧𝑜𝑛𝑡𝑎𝑙 𝑎𝑥𝑖𝑠
10
𝑊 (cm)

t = 1 cm (both flange and web)

𝐻 = 20 𝑐𝑚
𝐻 (cm)
𝑊
= 0.5 0.6 0.7 0.8 0.9 1.0
𝐻

Figure 76: Simple columns with Channel section

Figure 77: Lateral displacement of the mid-point for various width to height ratio

97

SACS® Collapse Advanced Release 15.1

 Figure 78: Comparison of elastoplastic buckling load with AISC 360-16

5.1.18 Sample 18: Angle Section

Following benchmarks illustrate the ability of the Collapse Advanced to predict the buckling load (both
flexural and torsional buckling) of beams and columns with Angle section. The samples are 1) a simple
elastic beam subjected to the pure bending to model lateral torsional buckling of Angle sections and 2) a
simple column under the axial compression load to model elastoplastic flexural buckling.

5.1.18.1 Sample 18.1: Elastic lateral torsional buckling of simple beam with Angle section
This sample involves elastic lateral-torsional buckling of a simply supported Angle beam subjected to the
pure bending as shown in Figure 80. The beam is restrained against torsion at both ends and the
material properties are set to SACS default values. The model consists of two elements with eight sub-
segments per member and all members are set to remain elastic by entering ‘ME’ CLPOPT line. To
induce lateral-torsional buckling, an imperfection in the form of small lateral dislocation was applied to
mid-point representing lateral loads. The beam experience large lateral deformation during bucking.
Therefore, both sub- incrementation and arc-length options are included in the analysis by entering ‘SI’
and ‘AL’ on the CLPOPT input line.

The simulation is carried out for following section dimensions: L = 10.0, 11.0, 12.0, 13.0, 14.0, and 15.0
cm. The mid-point lateral deformation is plotted for all cases in Figure 80. The resulting buckling bending
moments are compared with analytical solution derived from [19], and Figure 81 shows they are in good
agreement with the analytical solution.

98

SACS® Collapse Advanced Release 15.1

 𝑀 𝐿 = 5𝑚 𝑀

𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛 𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛

𝑠𝑚𝑎𝑙𝑙 𝑖𝑚𝑝𝑒𝑟𝑓𝑒𝑐𝑡𝑖𝑜𝑛 𝑎𝑙𝑜𝑛𝑔 𝐴𝑛𝑔𝑙𝑒 𝑠𝑒𝑐𝑡𝑖𝑜𝑛 ℎ𝑜𝑟𝑖𝑧𝑜𝑛𝑡𝑎𝑙 𝑎𝑥𝑖𝑠

𝐿 (cm)

t = 1 cm
𝐿 = 10.0 11.0 12.0 13.0 14.0 15.0

𝐿 (cm)

Figure 79: Simple beam subjected to the pure bending

Figure 80: Mid-point lateral deformation for Angle section

99

SACS® Collapse Advanced Release 15.1

 Elastic Lateral-Torsional Buckling of Angle Section Simple Beam
60
Analytical
Collapse Advanced

Buckling Moment (kN.m)

40

20

0
9 10 11 12 13 14 15 16

Leg length (cm)

Figure 81: Collapse Advanced buckling moment for Angle section versus the analytical solution [19]

5.1.18.2 Sample 18.2: Elastoplastic flexural buckling of simple column with Angle section
This benchmark focuses on elastoplastic flexural buckling analysis of members with Angle section under
compression load. Figure 82 illustrates a simply-supported column subjected to axial compression load
at one end. The column is restrained against torsion at both ends and has a small imperfection in the
form of mid-point perturbation as lateral dislocation. The mid-point lateral dislocation is 1/10th of angle
section radius of gyration about the local z-axis. The model consists of two elements with eight sub-
segments per members and the material properties are set to SACS default values. Both sub-
incrementation option (‘SI’ on CLPOPT) and arc-length method (‘AL’ on CLPOPT) are utilized for this
benchmark.

The analysis is carried out for following section width to height ratios: dimensions: L = 10.0, 11.0, 12.0,
13.0, 14.0, and 15.0 cm and results are presented in Figure 83 and Figure 84. Figure 83 illustrates the
lateral response of the columns with different Angle section with different leg length. Figure 84
compares buckling loads – the peak load for mid-point deflection curves in Figure 83 – with ultimate
load provided in Section E of AISC 360-16 Specification. The results show that Collapse Advanced
accurately predicts the elastoplastic response for all cases.

100

SACS® Collapse Advanced Release 15.1

 𝐿 = 2.5𝑚 𝑃(𝑘𝑁)

𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛 𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛

𝑟𝑧
𝑖𝑚𝑝𝑒𝑟𝑓𝑒𝑐𝑡𝑖𝑜𝑛 𝑎𝑙𝑜𝑛𝑔 𝐴𝑛𝑔𝑙𝑒 𝑠𝑒𝑐𝑡𝑖𝑜𝑛 ℎ𝑜𝑟𝑖𝑧𝑜𝑛𝑡𝑎𝑙 𝑎𝑥𝑖𝑠
10

𝐿 (cm)

t = 1 cm
𝐿 = 10.0 11.0 12.0 13.0 14.0 15.0

𝐿 (cm)

Figure 82: Simple columns with Angle section

Figure 83: Lateral displacement of the mid-point for various width to height ratio

101

SACS® Collapse Advanced Release 15.1

 Figure 84: Comparison of elastoplastic buckling load with AISC 360-16

5.1.19 Sample 19: Box Section

Following benchmarks illustrate the ability of the Collapse Advanced to predict the buckling load (both
flexural and torsional buckling) of beams and columns with Box section. The samples are 1) a simple
elastic beam subjected to the pure bending to model lateral torsional buckling of Box sections and 2) a
simple column under the axial compression load to model elastoplastic flexural buckling.

5.1.19.1 Sample 19.1: Elastic lateral torsional buckling of simple beam with Box section
This sample involves elastic lateral-torsional buckling of a simply supported Box beam subjected to the
pure bending as shown in Figure 85. The beam is restrained against torsion at both ends and the
material properties are set to SACS default values. The model consists of two elements with eight sub-
segments per member and all members are set to remain elastic by entering ‘ME’ CLPOPT line. To
induce lateral-torsional buckling, an imperfection in the form of small lateral dislocation was applied to
mid-point representing lateral loads. The beam experience large lateral deformation during bucking.
Therefore, both sub- incrementation and arc-length options are included in the analysis by entering ‘SI’
and ‘AL’ on the CLPOPT input line.

The simulation is carried out for following section height to width ratios: 1.5, 1.6, 1.7, 1.8, 1.9, and 2.0.
The mid-point lateral deformation is plotted for all cases in Figure 86. The resulting buckling bending
moments are compared with analytical solution derived from [19] and Figure 87 shows they are in good
agreement with the analytical solution.

102

SACS® Collapse Advanced Release 15.1

 𝑀 𝐿 = 5𝑚 𝑀

𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛 𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛

𝑠𝑚𝑎𝑙𝑙 𝑖𝑚𝑝𝑒𝑟𝑓𝑒𝑐𝑡𝑖𝑜𝑛 𝑎𝑙𝑜𝑛𝑔 𝐵𝑜𝑥 𝑠𝑒𝑐𝑡𝑖𝑜𝑛 ℎ𝑜𝑟𝑖𝑧𝑜𝑛𝑡𝑎𝑙 𝑎𝑥𝑖𝑠

𝑊 (cm)

t = 1 cm (both flange and web)

𝑊 = 10 𝑐𝑚
𝐻 (cm)
𝐻
= 1.5 1.6 1.7 1.8 1.9 2.0
𝑊

Figure 85: Simple beam subjected to the pure bending

Figure 86: Mid-point lateral deformation for Box section

103

SACS® Collapse Advanced Release 15.1

 Elastic Lateral-Torsional Buckling of Box Section Simple Beam
1400
Analytical
1200 Collapse Advanced

Buckling Moment (kN.m)

1000

800

600

400

200

0
1.4 1.5 1.6 1.7 1.8 1.9 2 2.1

Height to Width Ratio

Figure 87: Collapse Advanced buckling moment for Box section versus the analytical solution [19]

5.1.19.2 Sample 19.2: Elastoplastic flexural buckling of simple column with Box section
This benchmark focuses on elastoplastic flexural buckling analysis of members with Box section under
compression load. Figure 88 illustrates a simply-supported column subjected to axial compression load
at one end. The column is restrained against torsion at both ends and has a small imperfection in the
form of mid-point perturbation as lateral dislocation. The mid-point lateral dislocation is 1/10th of Box
section radius of gyration about the local z-axis. The model consists of two elements with eight sub-
segments per members and the material properties are set to SACS default values. Both sub-
incrementation option (‘SI’ on CLPOPT) and arc-length method (‘AL’ on CLPOPT) are utilized for this
benchmark.

The analysis is carried out for following section height to width ratios: 1.5, 1.6, 1.7, 1.8, 1.9, and 2.0 and
results are presented in Figure 89 and Figure 90. Figure 89 illustrates the lateral response of the columns
with different Box section. Figure 90 compares buckling loads – the peak load for mid-point deflection
curves in Figure 89 – with ultimate load provided in Section E of AISC 360-16 Specification. As seen in the
plots, the Collapse Advanced predicts slightly higher ultimate loads for box section. As shown in Figure
89, the Box columns deform about 4 cm at the mid-point before the buckling occurs. This relatively large
deformation and corresponding geometrical stiffness contribute to higher ultimate load. In addition, the
AISC 360-16 Specification reduces the analytical Euler elastic buckling load by a factor of 0.877. If the
reduction factor is removed from AISC elastoplastic critical stress calculation, the Collapse Advanced
ultimate loads are well within AISC limits – as show in Figure 90.

104

SACS® Collapse Advanced Release 15.1

 𝐿=5𝑚 𝑃(𝑘𝑁)

𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛 𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛

𝑟𝑧
𝑖𝑚𝑝𝑒𝑟𝑓𝑒𝑐𝑡𝑖𝑜𝑛 𝑎𝑙𝑜𝑛𝑔 𝐵𝑜𝑥 𝑠𝑒𝑐𝑡𝑖𝑜𝑛 ℎ𝑜𝑟𝑖𝑧𝑜𝑛𝑡𝑎𝑙 𝑎𝑥𝑖𝑠
10
𝑊 (cm)

t = 1 cm (both flange and web)

𝑊 = 10 𝑐𝑚
𝐻 (cm)
𝐻
= 1.5 1.6 1.7 1.8 1.9 2.0
𝑊

Figure 88: Simple columns with Box section

Figure 89: Lateral displacement of the mid-point for various height to width ratio

105

SACS® Collapse Advanced Release 15.1

 Figure 90: Comparison of elastoplastic buckling load with AISC 360-16

5.1.20 Sample 20: Conical Section

The following is a benchmark to illustrate accuracy of Conical sections calculation in Collapse Advanced
program. Figure 91 shows a simply-supported column consistent of two conical members and two
tubular members. The column is subjected to a lateral load 𝑃 at mid-point and compression axial load
𝑃 ⁄10 at its end. The column is restrained against torsion at both ends. The analysis is carried out using
eight sub-segments per members and the material properties are set to SACS default values. Both sub-
incrementation option (‘SI’ on CLPOPT) and arc-length method (‘AL’ on CLPOPT) are utilized for this
benchmark.

The analysis carried out for three different Conical sections as shown in Figure 91. Figure 92 shows the
failure mode and plasticity contours for all cases at final load step. Figure 93 compares the vertical
displacement at the mid-point calculated with Conical section and the equivalent segmented tubular
members. For the reference solution, the Conical members are divided to 8 segments and each segment
is modelled using a tubular section with the equivalent diameter. The diameter of equivalent tubular
segments is determined by linearly interpolating between Conical bigger and smaller diameters. As seen
in the graph, the Collapse Advanced Conical section is in a good agreement with the reference solution.

106

SACS® Collapse Advanced Release 15.1

 𝑃(𝑘𝑁)
𝑃(𝑘𝑁)
10
𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛 𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛
𝐿 = 5.0𝑚

t = 1 cm
𝐷1 = 40, 𝐷2 = 30 𝑐𝑚
𝐷1 (cm) 𝐷2 (cm)
𝐷1 = 60, 𝐷2 = 30 𝑐𝑚
𝐷1 = 80, 𝐷2 = 30 𝑐𝑚

Figure 91: Simple columns with Conical section

Figure 92: Plasticity contours for simple column with conical section

107

SACS® Collapse Advanced Release 15.1

 Figure 93: Comparison of conical section with segmented tubular section

5.1.21 Sample 21: Concentric (Double) Tubular Section

This benchmark focuses on elastoplastic flexural buckling analysis of members with Concentric Tubulars
under compression load. Figure 94 illustrates a simply-supported column subjected to axial compression
load at one end. The column is restrained against torsion at both ends and has a small imperfection in
the form of mid-point perturbation as lateral dislocation. The model consists of two elements with eight
sub-segments per members and the material properties are set to SACS default values. Both sub-
incrementation option (‘SI’ on CLPOPT) and arc-length method (‘AL’ on CLPOPT) are utilized for this
benchmark.

The analysis carried out for three different Concentric Tubulars as shown in Figure 94. Figure 95
compares the vertical displacement at the mid-point calculated with Concentric Tubular and the
equivalent two-tubular members. For the reference solution, each Concentric Tubular is modelled with
two tubular members – one for the inner tubular and one for the outer tubular. As seen in this graph,
the Collapse Advanced Concentric Tubular is in a good agreement with reference solution.

108

SACS® Collapse Advanced Release 15.1

 𝐿 = 10.0𝑚
𝑃(𝑘𝑁)
𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛 𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛

𝑠𝑚𝑎𝑙𝑙 𝑖𝑚𝑝𝑒𝑟𝑓𝑒𝑐𝑡𝑖𝑜𝑛 𝑎𝑙𝑜𝑛𝑔 ℎ𝑜𝑟𝑖𝑧𝑜𝑛𝑡𝑎𝑙 𝑎𝑥𝑖𝑠

𝐷𝑜𝑢𝑡 (cm) t = 1 cm

𝐷𝑖𝑛 (cm) 𝐷𝑜𝑢𝑡 = 40, 𝐷𝑖𝑛 = 20 𝑐𝑚

𝐷𝑜𝑢𝑡 = 45, 𝐷𝑖𝑛 = 30 𝑐𝑚
𝐷𝑜𝑢𝑡 = 60, 𝐷𝑖𝑛 = 50 𝑐𝑚

Figure 94: Simple columns with Concentric (Double) Tubulars

Figure 95: Comparison of Concentric Tubular section with two-tubular members

109

SACS® Collapse Advanced Release 15.1

5.1.22 Sample 22: Launch Runner and Special Launch Runner Section
This sample involves elastoplastic flexural buckling analysis of members with Launch Runner and Special
Launch Runner under compression load. Figure 96 illustrates a simply-supported column subjected to
axial compression load at one end. The column is restrained against torsion at both ends and has a small
imperfection in the form of mid-point perturbation as lateral dislocation. The model consists of two
elements with eight sub-segments per members and the material properties are set to SACS default
values. Both sub-incrementation option (‘SI’ on CLPOPT) and arc-length method (‘AL’ on CLPOPT) are
utilized for this benchmark.

The analysis carried out for a launch runner and a special launch runner section as shown in Figure 96.
Figure 97 compares the vertical displacement at the mid-point with tubular member without runner
plates.

𝐿 = 20.0𝑚
𝑃(𝑘𝑁)
𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛 𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛

Tubular: D𝑜𝑢𝑡 = 120 𝑐𝑚, t = 3 𝑐𝑚

Runner Depth = 110 𝑐𝑚
Runner Width = 100 𝑐𝑚
Side Plate Thickness = 2 𝑐𝑚
Bottom Plate Thickness = 2.5 𝑐𝑚

Tubular: D𝑜𝑢𝑡 = 120 𝑐𝑚, t = 3 𝑐𝑚

Runner Depth = 110 𝑐𝑚
Runner Width = 100 𝑐𝑚
Runner Top Width = 100 𝑐𝑚
Runner Bottom Depth = 80 𝑐𝑚
Side Plate Thickness = 2 𝑐𝑚
Bottom Plate Thickness = 2.5 𝑐𝑚

Figure 96: Simple columns with Launch Runner sections

110

SACS® Collapse Advanced Release 15.1

Figure 97: Top: Plastic contour of Launch Runner section, Bottom: Comparison of Launch Runner section
buckling

111

SACS® Collapse Advanced Release 15.1

5.1.23 Sample 23: Rectangular Tube
Following benchmarks illustrate the ability of the Collapse Advanced to predict the buckling load (both
flexural and torsional buckling) of beams and columns with Rectangular Tube section. The samples are
1) a simple elastic beam subjected to the pure bending to model lateral torsional buckling of Box
sections and 2) a simple column under the axial compression load to model elastoplastic flexural
buckling.

5.1.23.1 Sample 23.1: Elastic lateral torsional buckling of simple beam with Rectangular Tube
section

This sample involves elastic lateral-torsional buckling of a simply supported Rectangular Tube beam
subjected to the pure bending as shown in Figure 98. The beam is restrained against torsion at both
ends and the material properties are set to SACS default values. The model consists of two elements
with eight sub-segments per member and all members are set to remain elastic by entering ‘ME’
CLPOPT line. To induce lateral-torsional buckling, an imperfection in the form of small lateral dislocation
was applied to mid-point representing lateral loads. The beam experience large lateral deformation
during bucking. Therefore, both sub- incrementation and arc-length options are included in the analysis
by entering ‘SI’ and ‘AL’ on the CLPOPT input line.

The simulation is carried out for following section height to width ratios: 1.50, 1.75, and 2.00. The mid-
point lateral deformation is plotted for all cases in Figure 99. The resulting buckling bending moments
are compared with analytical solution derived from [19] and Figure 100 shows they are in good
agreement with the analytical solution.

𝑀 𝐿 = 5𝑚 𝑀

𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛 𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛

𝑠𝑚𝑎𝑙𝑙 𝑖𝑚𝑝𝑒𝑟𝑓𝑒𝑐𝑡𝑖𝑜𝑛 𝑎𝑙𝑜𝑛𝑔 𝑅𝑒𝑐𝑡𝑎𝑛𝑔𝑢𝑙𝑎𝑟 𝑇𝑢𝑏𝑒 ℎ𝑜𝑟𝑖𝑧𝑜𝑛𝑡𝑎𝑙 𝑎𝑥𝑖𝑠

𝑊 (cm)

t = 1 cm (both flange and web)

𝑊 = 10 𝑐𝑚
𝐻 (cm)
𝐻
= 1.50 1.75 2.00
𝑊

Figure 98: Simple beam subjected to the pure bending

112

SACS® Collapse Advanced Release 15.1

 Figure 99: Mid-point lateral deformation for Rectangular Tube section

Figure 100: Collapse Advanced buckling moment for Rectangular Tube section versus the analytical
solution [19]

113

SACS® Collapse Advanced Release 15.1

5.1.23.2 Sample 23.2: Elastoplastic flexural buckling of simple column with Rectangular Tube
section
This benchmark focuses on elastoplastic flexural buckling analysis of members with Rectangular Tube
section under compression load. Figure 101 illustrates a simply supported column subjected to axial
compression load at one end. The column is restrained against torsion at both ends and has a small
imperfection in the form of mid-point perturbation as lateral dislocation. The mid-point lateral
dislocation is 1/10th of Rectangular Tube section radius of gyration about the local z-axis. The model
consists of two elements with eight sub-segments per members and the material properties are set to
SACS default values. Both sub-incrementation option (‘SI’ on CLPOPT) and arc-length method (‘AL’ on
CLPOPT) are utilized for this benchmark.

The analysis is carried out for following section height to width ratios: 1.50, 1.75, and 2.00 and results
are presented in Figure 102 and Figure 103. Figure 102 illustrates the lateral response of the columns
with different Rectangular Tube sections. Figure 103 compares buckling loads – the peak load for mid-
point deflection curves in Figure 102 – with ultimate load provided in Section E of AISC 360-16
Specification. As seen in the plots, the Collapse Advanced predicts slightly higher ultimate loads for
Rectangular Tube section. As shown in Figure 102, the Rectangular Tube columns deform about 4 cm at
the mid-point before the buckling occurs. This relatively large deformation and corresponding
geometrical stiffness contribute to higher ultimate load. In addition, the AISC 360-16 Specification
reduces the analytical Euler elastic buckling load by a factor of 0.877. If the reduction factor is removed
from AISC elastoplastic critical stress calculation, the Collapse Advanced ultimate loads are well within
AISC limits – as show in Figure 103.

𝐿=5𝑚 𝑃(𝑘𝑁)

𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛 𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛

𝑟𝑧
𝑖𝑚𝑝𝑒𝑟𝑓𝑒𝑐𝑡𝑖𝑜𝑛 𝑎𝑙𝑜𝑛𝑔 𝑅𝑒𝑐𝑡𝑛𝑎𝑔𝑢𝑙𝑎𝑟 𝑇𝑢𝑏𝑒 𝑠𝑒𝑐𝑡𝑖𝑜𝑛 ℎ𝑜𝑟𝑖𝑧𝑜𝑛𝑡𝑎𝑙 𝑎𝑥𝑖𝑠
10
𝑊 (cm)

t = 1 cm (both flange and web)

𝑊 = 10 𝑐𝑚
𝐻 (cm)
𝐻
= 1.50 1.75 2.00
𝑊

Figure 101: Simple columns with Rectangular Tube section

114

SACS® Collapse Advanced Release 15.1

 Figure 102: Lateral displacement of the mid-point for various height to width ratio

Figure 103: Comparison of elastoplastic buckling load with AISC 360-16

115

SACS® Collapse Advanced Release 15.1

5.1.24 Sample 24: Double Web Plate Girder Section
This sample involves elastoplastic flexural buckling analysis of members with Double Web Plate Girder
under compression load. Figure 104 illustrates a simply supported column subjected to axial
compression load at one end. The column is restrained against torsion at both ends and has a small
imperfection in the form of mid-point perturbation as lateral dislocation. The model consists of two
elements with eight sub-segments per members and the material properties are set to SACS default
values. Both sub-incrementation option (‘SI’ on CLPOPT) and arc-length method (‘AL’ on CLPOPT) are
utilized for this benchmark.

The analysis carried out for three Double Web Plate Girder sections with height to width ratio of 1.50,
1.75, and 2.00 as shown in Figure 104. Figure 105 compares the lateral displacement at the mid-point of
the column.

𝐿 = 10.0𝑚
𝑃(𝑘𝑁)
𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛 𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛

𝑊 (cm)

Flange Thickness = 3cm

Web Thickness = 1cm
𝐻 (cm) Web Spacing = 10cm
W = 40 𝑐𝑚
𝐻
= 1.50 1.75 2.00
𝑊

Figure 104: Simple columns with Double Web Plate Girder sections

116

SACS® Collapse Advanced Release 15.1

Figure 105: Top: Plastic contour of Double Web Plate Girder section, Bottom: Comparison of Double
Web Plate Girder sections buckling

117

SACS® Collapse Advanced Release 15.1

5.1.25 Sample 25: Boxed Plate Girder Section
This sample involves elastoplastic flexural buckling analysis of members with Boxed Plate Girder under
compression load. Figure 106 illustrates a simply supported column subjected to axial compression load
at one end. The column is restrained against torsion at both ends and has a small imperfection in the
form of mid-point perturbation as lateral dislocation. The model consists of two elements with eight
sub-segments per members and the material properties are set to SACS default values. Both sub-
incrementation option (‘SI’ on CLPOPT) and arc-length method (‘AL’ on CLPOPT) are utilized for this
benchmark.

The analysis carried out for three Boxed Plate Girder sections with height to width ratio of 1.50, 1.75,
and 2.00 as shown in Figure 106. Figure 107 compares the lateral displacement at the mid-point of the
column.

𝐿 = 10.0𝑚
𝑃(𝑘𝑁)
𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛 𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛

𝑊 (cm)
Side Plates Thickness = 2.5cm
Flange Thickness = 3cm
Web Thickness = 1cm
𝐻 (cm)
W = 40 𝑐𝑚
𝐻
= 1.50 1.75 2.00
𝑊

Figure 106: Simple columns with Boxed Plate Girder sections

118

SACS® Collapse Advanced Release 15.1

 Figure 107: Top: Plastic contour of Boxed Plate Girder, Bottom: Comparison of Boxed Plate Girder
sections buckling

119

SACS® Collapse Advanced Release 15.1

5.1.26 Sample 26: Unsymmetrical Plate Girder
Following benchmarks illustrate the ability of the Collapse Advanced to predict the buckling load (both
flexural and torsional buckling) of beams and columns with Unsymmetrical Plate Girder. The samples are
1) a simple elastic beam subjected to the pure bending to model lateral torsional buckling of
Unsymmetrical Plate Girder and 2) a simple column under the axial compression load to model
elastoplastic flexural buckling.

5.1.26.1 Sample 26.1: Elastic lateral torsional buckling of simply supported Unsymmetrical
Plate Girder

This sample involves elastic lateral-torsional buckling of a simply supported Unsymmetrical Plate Girder
subjected to the pure bending as shown in Figure 108. The beam is restrained against torsion at both
ends and the material properties are set to SACS default values. The model consists of two elements
with eight sub-segments per member and all members are set to remain elastic by entering ‘ME’
CLPOPT line. To induce lateral-torsional buckling, an imperfection in the form of small lateral dislocation
was applied to mid-point representing lateral loads. The beam experience large lateral deformation
during bucking. Therefore, both sub- incrementation and arc-length options are included in the analysis
by entering ‘SI’ and ‘AL’ on the CLPOPT input line. Also, the analysis has been carried out with and
without Wagner strain (‘IWT’ on CLPOP2 line).

The simulation is carried out for following section height to top width ratios: 1.50, 1.75, and 2.00. The
resulting buckling bending moments are compared with analytical solution derived from [19] and Figure
109 shows they are in good agreement with the analytical solution.

𝑀 𝐿 = 10𝑚 𝑀

𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛 𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛

𝑠𝑚𝑎𝑙𝑙 𝑖𝑚𝑝𝑒𝑟𝑓𝑒𝑐𝑡𝑖𝑜𝑛 𝑎𝑙𝑜𝑛𝑔 𝑈𝑛𝑠𝑦𝑚𝑚𝑒𝑡𝑟𝑖𝑐𝑎𝑙 𝑃𝑙𝑎𝑡𝑒 𝐺𝑖𝑟𝑑𝑒𝑟 ℎ𝑜𝑟𝑖𝑧𝑜𝑛𝑡𝑎𝑙 𝑎𝑥𝑖𝑠

Top Flange Thickness = 2cm

𝑊𝑡𝑜𝑝 (cm)
Web Thickness = 1cm
Bottom Flange Width = 30cm
Bottm Flange Thickess = 1.5cm
𝐻 (cm)
𝑊𝑡𝑜𝑝 = 40 𝑐𝑚
𝐻
= 1.50 1.75 2.00
𝑊𝑡𝑜𝑝

Figure 108: Simple beam subjected to the pure bending

120

SACS® Collapse Advanced Release 15.1

 Figure 109: Collapse Advanced buckling moment for Unsymmetrical Plate Girder versus the analytical
solution [19]

5.1.26.2 Sample 26.2: Elastoplastic flexural buckling of simple column with Unsymmetrical Plate
Girder
This benchmark focuses on elastoplastic flexural buckling analysis of members with Unsymmetrical Plate
Girder under compression load. Figure 110 illustrates a simply supported column subjected to axial
compression load at one end. The column is restrained against torsion at both ends and has a small
imperfection in the form of mid-point perturbation as lateral dislocation. The mid-point lateral
dislocation is 1/10th of Unsymmetrical Plate Girder radius of gyration about the local z-axis. The model
consists of two elements with eight sub-segments per members and the material properties are set to
SACS default values. Both sub-incrementation option (‘SI’ on CLPOPT) and arc-length method (‘AL’ on
CLPOPT) are utilized for this benchmark.

The analysis is carried out for following section height to top width ratios: 1.50, 1.75, and 2.00 and
results are presented in Figure 111 and Figure 112. Figure 111 illustrates the lateral response of the
columns with different Unsymmetrical Plate Girder. Figure 112 compares buckling loads – the peak load
for mid-point deflection curves in Figure 111 – with ultimate load provided in Section E of AISC 360-16
Specification.

121

SACS® Collapse Advanced Release 15.1

 𝐿 = 10 𝑚 𝑃(𝑘𝑁)

𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛 𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛

𝑟𝑧
𝑖𝑚𝑝𝑒𝑟𝑓𝑒𝑐𝑡𝑖𝑜𝑛 𝑎𝑙𝑜𝑛𝑔 Unsymmetrical Plate Girder ℎ𝑜𝑟𝑖𝑧𝑜𝑛𝑡𝑎𝑙 𝑎𝑥𝑖𝑠
10

Top Flange Thickness = 2cm

𝑊𝑡𝑜𝑝 (cm)
Web Thickness = 1cm
Bottom Flange Width = 30cm
Bottm Flange Thickess = 1.5cm
𝐻 (cm)
𝑊𝑡𝑜𝑝 = 40 𝑐𝑚
𝐻
= 1.50 1.75 2.00
𝑊𝑡𝑜𝑝

Figure 110: Simple columns with Unsymmetrical Plate Girder

Figure 111: Lateral displacement of the mid-point for various height to width ratio

122

SACS® Collapse Advanced Release 15.1

 Elastoplastic Buckling of Unsymmetrical Plate Girder
3000
Axial Buckling Load (kN)

2000

1000

AISC 360-16
Collapse Advanced
0
1.4 1.5 1.6 1.7 1.8 1.9 2 2.1

Height to Top Width Ratio

Figure 112: Comparison of elastoplastic buckling load with AISC 360-16

5.1.27 Sample 27: Double Angle

This benchmark focuses on elastoplastic flexural buckling analysis of members with Double Angle under
compression load. Figure 113 illustrates a simply supported column subjected to axial compression load
at one end. The column is restrained against torsion at both ends and has a small imperfection in the
form of mid-point perturbation as vertical dislocation. The mid-point lateral dislocation is 1/10th of
Double Angle radius of gyration about the local y-axis. The model consists of two elements with eight
sub-segments per members and the material properties are set to SACS default values. Both sub-
incrementation option (‘SI’ on CLPOPT) and arc-length method (‘AL’ on CLPOPT) are utilized for this
benchmark.

The analysis is carried out for following section height to width ratios: 1.0, 1.5, and 2.0 and results are
presented in Figure 114 and Figure 115. Figure 114 illustrates the vertical response of the columns with
different Double Angle sections. Figure 115 compares buckling loads – the peak load for mid-point
deflection curves in Figure 114 – with ultimate load provided in Section E of AISC 360-16 Specification.

123

SACS® Collapse Advanced Release 15.1

 𝐿=5𝑚 𝑃(𝑘𝑁)

𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛 𝑓𝑖𝑥𝑒𝑑 𝑡𝑜𝑟𝑠𝑖𝑜𝑛

𝑟𝑧
𝑖𝑚𝑝𝑒𝑟𝑓𝑒𝑐𝑡𝑖𝑜𝑛 𝑎𝑙𝑜𝑛𝑔 Double Angle 𝑣𝑒𝑟𝑡𝑖𝑐𝑎𝑙 𝑎𝑥𝑖𝑠
10
𝑊 (cm)
t = 1cm
Spacing = 5cm
𝐻 (cm)
𝑊 = 10 𝑐𝑚
𝐻
= 1.0 1.5 2.0
𝑊

Figure 113: Simple columns with Double Angle section

Figure 114: Vertical displacement of the mid-point for various height to width ratio

124

SACS® Collapse Advanced Release 15.1

Figure 115: Top: Plastic contour of Double Angle section, Bottom: Comparison of elastoplastic buckling
load with AISC 360-16

125

SACS® Collapse Advanced Release 15.1

5.2 Plate Element Benchmarks and Samples
5.2.1 Sample 1: Elastic Restrained Beam Subjected to Uniform Load
Figure 116 shows a simply-supported beam, axially restrained at both ends, subjected to a uniformly
distributed load. The beam is modeled using plate elements. All plate elements are assumed to remain
elastic by selecting option ‘PE’ on CLPOPT line. Two models were considered 1) The beam was modeled
using 16 quadrilateral plate elements 2) the beam was modeled using 2x16 triangular plate elements. In
addition, the beam was analyzed using both thin (Kirchhoff) and thick (Mindlin) plate bending theories.
The displacement of the beam mid-point was compared with existing analytical solution [10] and is in
good agreement as seen in Figure 117 and Figure 118. The sub-incrementation option was selected to
reduce total number of load increments.

𝐾𝑁
𝑞൬ ൰
𝑚2

𝑤(𝑐𝑚 )

𝐾𝑁
𝐸 = 20000 , 𝜈 = 0.0 , 𝐿 = 10𝑚, 𝑤𝑖𝑑𝑡ℎ = 1 𝑚, 𝑇ℎ𝑖𝑐𝑘𝑛𝑒𝑠𝑠 = 1 𝑐𝑚
𝑐𝑚 2

Figure 116: Restrained beam with plates

Figure 117: Restrained beam mid-point vertical displacement with quadrilateral elements

Figure 118: Restrained beam mid-point vertical displacement with triangular elements

126

SACS® Collapse Advanced Release 15.1

5.2.2 Sample 2: Elastic Cantilever Beam
This benchmark looks at an elastic cantilever beam subjected to three different end loads: 1) an axial
load with small lateral load to produce buckling, 2) lateral load and 3) bending end moment. Details of
each case are presented in following sections.

5.2.2.1 Sample 2.1: End Axial Force

Figure 119 shows a cantilever beam subject to an axial end load and a small lateral load to induce
buckling. The beam was modeled using plate elements. Two models were considered 1) Using 8
quadrilateral plates elements. 2) Using 2x8 triangular elements. The beam model was analyzed using
both Kirchhoff and Mindlin bending theories. Both Arc-length and sub-incrementation methods were
used to help predict buckling. In addition, small load steps were used to capture the buckling mode
accurately. The results predicted for the beam tip axial and lateral displacements were compared
against existing finite element solution [30] and are seen to be in very good agreement as shown in
Figure 121. The results also show that the program is able to predict very large deformations for load
step 50 and the final load step as shown in Figure 121.

−𝑢𝑡𝑖𝑝 (𝑚 ) 𝑤𝑡𝑖𝑝 (𝑚 )

𝑃 (𝐾𝑁)
𝑃
(𝐾𝑁)
1000
𝐾𝑁
𝐸 = 20000 , 𝜈 = 0.3 , 𝐿 = 0.5𝑚, 𝑤𝑖𝑑𝑡ℎ = 7.5 𝑐𝑚, 𝑇ℎ𝑖𝑐𝑘𝑛𝑒𝑠𝑠 = 0.45 𝑐𝑚
𝑐𝑚 2

Figure 119: Cantilever beam with axial load

Figure 120: Cantilever beam tip displacement under axial load buckling

127

SACS® Collapse Advanced Release 15.1

 Figure 121: Deformed cantilever beam at load step 50 (left) and the final load step (right)

5.2.2.2 Sample 2.2: End Lateral Force

Figure 122 shows a cantilever beam with an applied lateral load. The cantilever beam was modeled
using plate elements. Two models were considered: 1) Using 16 quadrilateral plate elements. 2) Using
2x16 triangular plate elements. The analysis was conducted using both Kirchhoff and Mindlin bending
theories. Both sub-incrementation or Arc-length options were not required in this case since buckling
behavior is not involved and the target load factor can be achieved using standard iterations. The results
predicted for the beam tip lateral and axial displacement are compared against existing analytical
solution [14] and are in good agreement as shown in Figure 123.

𝑤𝑡𝑖𝑝 (𝑚 )
−𝑢𝑡𝑖𝑝 (𝑚 )

𝑃(𝐾𝑁)

𝐾𝑁
𝐸 = 120 , 𝜈 = 0.0 , 𝐿 = 10 𝑚, 𝑤𝑖𝑑𝑡ℎ = 1 𝑚, 𝑇ℎ𝑖𝑐𝑘𝑛𝑒𝑠𝑠 = 10 𝑐𝑚
𝑐𝑚 2

Figure 122: Cantilever beam with axial load

128

SACS® Collapse Advanced Release 15.1

 Figure 123: Cantilever beam tip displacement under lateral load

5.2.2.3 Sample 2.3: End Bending Moment

Figure 124 shows a cantilever beam subject to an applied end moment. The beam is modeled using
plate elements. The beam tip (i.e. location of applied moment) is laterally restrained to prevent lateral
buckling. Two models were considered: 1) Using 16 quadrilateral plate elements. 2) Using 2x16
triangular plate elements. The analysis was conducted using both Kirchhoff and Mindlin bending
theories. To help improve convergence, the sub-incrementation option was used. The results predicted
by the program were compared against existing analytical solution [14] for the beam tip displacement as
shown in Figure 125. Figure 126 shows beam deformation for peak vertical displacement, peak of
horizontal displacement and the final load step.

𝑤𝑡𝑖𝑝 (𝑚 )
−𝑢𝑡𝑖𝑝 (𝑚 )

𝑀(𝐾𝑁. 𝑚)

𝐾𝑁
𝐸 = 120 , 𝜈 = 0.0 , 𝐿 = 12 𝑚, 𝑤𝑖𝑑𝑡ℎ = 1 𝑚, 𝑇ℎ𝑖𝑐𝑘𝑛𝑒𝑠𝑠 = 10 𝑐𝑚
𝑐𝑚 2

Figure 124: Cantilever beam with end moment

129

SACS® Collapse Advanced Release 15.1

 Figure 125: Cantilever beam tip displacement with end moment

Figure 126: Deformed cantilever beam. Left) at the peak vertical displacement, Center) at the peak
horizontal displacement, and Right) at the last step.

130

SACS® Collapse Advanced Release 15.1

5.2.3 Sample 3: Elastic Clamped Hinged Deep Arc
Figure 127 shows a model of an arc with clamped and hinged end conditions. The arc is modeled using
plate elements and is subjected to a point load at the apex. The model is restrained against out-of-plane
movement to prevent lateral buckling. Two models were considered: 1) Using 40 quadrilateral plate
elements. 2) Using 2x40 triangular plate elements. Since this test involves buckling, both sub-
incrementation and Arc-length methods were used by selecting ‘SI’ and ‘AL’ options on CLPOPT input
line. In addition, all plates were assumed to remain elastic by entering option ‘PE’ on the CLPOPT input
line. The arc apex displacement is compared against two existing finite element solutions [31, 32] as
shown in Figure 128. Figure 129 shows the buckled deformation of the model.

𝑃 (𝐾𝑁)
−𝑢𝑎𝑝𝑒𝑥 (𝑐𝑚 )

−𝑤𝑎𝑝𝑒𝑥 (𝑐𝑚 )

𝜃 = 145°

𝑅 = 1𝑚

𝐾𝑁
𝐸 = 500000 , 𝜈 = 0.0 , 𝑤𝑖𝑑𝑡ℎ = 24 𝑐𝑚, 𝑇ℎ𝑖𝑐𝑘𝑛𝑒𝑠𝑠 = 1 𝑐𝑚
𝑐𝑚2

Figure 127: Elastic clamped hinged deep arc

131

SACS® Collapse Advanced Release 15.1

 Figure 128: Arc apex displacement

Figure 129: large deformation of elastic clamped hinged deep arc

132

SACS® Collapse Advanced Release 15.1

5.2.4 Sample 4: Elastic Slit Ring Shape Plate with Lifting Line Load
Figure 130 shows an elastic ring with a slit. The ring was modeled using 180 (6x30) quadrilateral plate
elements where one end is clamped and the other end is being lifted by a line load. The sub-
incrementation method was used to improved convergence rate – especially at the earlier stages where
there is a significant deformation. The analysis was conducted using both Kirchhoff and Mindlin bending
theories. Vertical displacement of plate free end (inner point A and outer point B) are compared with
existing finite element results [14, 32] as sown in Figure 131. The large deformation of the model is
shown in Figure 132.

𝐾𝑁
𝑞൬ ൰
𝑚
𝑅𝑖𝑛 = 6 𝑚 𝐴 𝐵
𝑅𝑜𝑢𝑡 = 10 𝑚

𝐾𝑁
𝐸 = 2100 , 𝜈 = 0.0 , 𝑇ℎ𝑖𝑐𝑘𝑛𝑒𝑠𝑠 = 3 𝑐𝑚
𝑐𝑚 2

Figure 130: Slit ring shape plate with lifting line load

Figure 131: Slit ring shape plate free end displacement.

Figure 132: large deformation of elastic slit ring shape plate

133

SACS® Collapse Advanced Release 15.1

5.2.5 Sample 5: Elastic Hemispherical Shell
Figure 133 shows a hemispherical shell with a circular cut at angle of 18˚. The shell is modeled using
16x16 plate mesh and is subjected to inward and outward radial point loads. Due to symmetry of the
model and the loading, only a quarter of the shell was modeled and analyzed by applying appropriate
boundary conditions at lines of symmetry and applying half of the load. Kirchhoff (thin) bending theory
was used for the analysis. The load was applied over 50 increments and sub-incrementation or Arc-
length methods were included for this benchmark. The radial displacement of the shell at two points A
and B were compared against existing finite element results [14] as shown in Figure 134. It can be seen
there is very good agreement between the two sets of results.

𝑓𝑟𝑒𝑒

𝑃
𝐵 (𝐾𝑁)
2
𝑃
(𝐾𝑁) 𝐴 −𝑢𝐵 (𝑚)
2
−𝑣𝐴 (𝑚) 𝑓𝑟𝑒𝑒

𝐾𝑁
𝐸 = 6825 , 𝜈 = 0.3 𝑅 = 10𝑚, , 𝑇ℎ𝑖𝑐𝑘𝑛𝑒𝑠𝑠 = 4 𝑐𝑚
𝑐𝑚2

Figure 133: Elastic hemispherical shell

Figure 134: Elastic hemispherical shell radial displacement

134

SACS® Collapse Advanced Release 15.1

5.2.6 Sample 6: Elastic Hinged Cylindrical Roof
Figure 135 shows a model of a shallow cylindrical roof with two hinged sides subjected to a point load P
at the center. Due to the symmetry of the structure and boundary conditions, only one quarter of the
𝑃
structure is modeled with an applied load of 4
. Four separate models were analyzed using 16×16 mesh
of quadrilateral elements, 16×16×2 of triangular plate element mesh and two plate thickness values of
12.7mm and 6.35mm.
The analysis was conducted using both Kirchhoff and Mindlin bending theories. The Arc-length method
was selected on the CLPOPT input line to predict the post-buckling response and the sub-incrementation
method was not used for this benchmark to improve convergence. Since the sub-incrementation option
is not selected, a non-convergence warning message will be reported near buckling indicating the
program has switched to the Arc-length solver. To improve post-buckling results, the automatically
calculated arc-length parameter is also reduced by a factor of 0.5 on the ARCLEN input line. The results
from the program for the roof center vertical displacement were compared with existing finite element
analysis [14] as shown in Figure 70 and Figure 80.
Results from the models with plate thickness of 12.7 shows standard elastic snap-through buckling
response where the model starts to gain stiffness with increasing the displacement during post-buckling
as shown in Figure 136.
Results from the models with plate thickness of 6.35 mm shows snap-back buckling with decreasing
displacement for portion of the post buckling response as shown in Figure 137.
It can be seen there is very good agreement with existing analytical results for all cases.

𝑓𝑟𝑒𝑒
𝑃 (𝐾𝑁) 𝑃
(𝐾𝑁)
4

𝐿 (𝑐𝑚) −𝑤 (𝑐𝑚)

𝑓𝑟𝑒𝑒
𝑓𝑟𝑒𝑒

𝜃 = 0.1 𝑟𝑎𝑑

𝐾𝑁
𝐸 = 310.275 , 𝜈 = 0.3, 𝑅 = 2.54𝑚, 𝐿 = 50.8𝑐𝑚
𝑐𝑚2

Figure 135: Elastic Hinged Cylindrical Roof

135

SACS® Collapse Advanced Release 15.1

 Figure 136: Vertical displacement at center of hinged cylindrical roof, plate thickness: 12.7 mm

Figure 137: Vertical displacement at center of hinged cylindrical roof, plate thickness: 6.35 mm

136

SACS® Collapse Advanced Release 15.1

5.2.7 Sample 7: Elastic Semi-Cylindrical Shell Subjected to Vertical Point Load
Figure 138 shows a semi-cylindrical elastic shell where one curved end is clamped and the other end is
free. The longitudinal ends of the shell are simply supported. The shell model subjected to a vertical
point load P at the apex. Due to symmetry, only half of the shell is modeled. Two models of the
cylindrical shell were considered: 1) Using 32×32 quadrilateral element mesh. 2) Using 32×32×2
triangular element mesh. For this benchmark, the Arc-length and sub-incrementation methods were not
used. The load was applied over 40 steps with a maximum number of iterations 20. The analysis was
conducted using both thin plate (Kirchhoff) and thick plate (Mindlin) theories. The results from Collapse
Advanced for the vertical displacement at cylinder apex was compared with existing finite element
results [14] as shown in Figure 140. It can be seen there is very good agreement between the two sets
of results for all cases. Figure 139 shows the deformation of the shell at load steps 10, 20 and 40.

𝑃
𝐾𝑁 (𝐾𝑁)
𝐸 = 2068.5 2 2
𝑐𝑚
𝜈 = 0.3
𝑅 = 1.016𝑚 −𝑤 (𝑚)

𝐿 = 3.048𝑚
𝑡ℎ𝑖𝑐𝑘𝑛𝑒𝑠𝑠 = 3𝑐𝑚
𝑓𝑟𝑒𝑒

Figure 138: Elastic semi-cylindrical shell subjected to end pinching force

Figure 139: Semi-cylindrical deformed model at load step 10 (left), 20 (center), 40 (right)
137

SACS® Collapse Advanced Release 15.1

 Figure 140: Vertical displacement of semi-cylindrical apex

5.2.8 Sample 8: Elastic Open-ended Cylindrical Shell with Radial Pulling Forces
Figure 141 shows an elastic open-ended cylindrical shell subjected to radial pulling forces. The shell is
modeled using a 24×16 quadrilateral element mesh. Due to symmetry, only one 1/8th of the cylinder is
modelled. The analysis was conducted using both thin plate (Kirchhoff) and thick plate (Mindlin)
theories. This sub-incrementation method was used to improve convergence during the rapid change of
shell deformation. The rapid change of deformation can be seen at the first load increment where a
non-convergence warning message is reported and the program commences the sub-incrementation
procedure. The resulting deflections predicted by the program at various locations on the shell are
compared against existing analysis results [14] as shown in Figure 142. It can be seen there is excellent
agreement between the two sets of results for all cases. Figure 143 shows deformation of the shell
structure at various load steps.

138

SACS® Collapse Advanced Release 15.1

 𝑃
𝑃 (𝐾𝑁) (𝐾𝑁)
4
𝑤𝐴 (𝑚)

𝑓𝑟𝑒𝑒 𝑓𝑟𝑒𝑒

−𝑢𝐵 (𝑚)
𝐿 (𝑚)
𝑃 (𝐾𝑁)
−𝑢𝐶 (𝑚)
𝐾𝑁
𝐸 = 1050 , 𝜈 = 0.3125, 𝑅 = 4.953𝑚, 𝐿 = 10.35𝑚, 𝑡ℎ𝑖𝑐𝑘𝑛𝑒𝑠𝑠 = 9.4𝑐𝑚
𝑐𝑚 2

Figure 141: Elastic open-ended cylindrical shell subjected to radial pulling forces

Figure 142: Displacement of cylindrical shell with radial pulling forces

Figure 143: Deformation of open-ended cylindrical shell

139

SACS® Collapse Advanced Release 15.1

5.2.9 Sample 9: Elastic Pinched Cylindrical Shell with Rigid End Diaphragms
Figure 144 shows an elastic cylindrical shell subjected to a pinching load. The cylinder has rigid
diaphragms at both ends which prevent any in plane deformation. Due to symmetry, only 1/8th of the
shell is modelled. Two models were considered: 1) Using 60×20 for quadrilateral element mesh. 2) Using
60×20×4 triangular plate element mesh. The analysis was conducted using both thin plate (Kirchhoff)
and thick plate (Mindlin) theories. The maximum number of iterations was set to 40 and sub-
incrementation approach was used is to improve the convergence rate. The results predicted by the
program for the radial displacements of the shell mid-points were compared with existing analytical
results [14] as shown in Figure 145. It can be seen there is excellent correlation between the two sets of
results.
Figure 146 shows the effect of mesh refinement on geometrical locking (sharp deformed edges) due to
large deformations

𝑃
(𝐾𝑁)
4
𝑃 (𝐾𝑁)

−𝑤𝐴 (𝑐𝑚)

𝑟𝑖𝑔𝑖𝑑 𝑟𝑖𝑔𝑖𝑑

𝑢𝐵 (𝑐𝑚)
𝐿 (𝑚)
𝑃 (𝐾𝑁)

𝐾𝑁
𝐸 = 30000 , 𝜈 = 0.3 𝑅 = 1𝑚, 𝐿 = 2𝑚, 𝑡ℎ𝑖𝑐𝑘𝑛𝑒𝑠𝑠 = 1𝑐𝑚
𝑐𝑚 2

Figure 144: Elastic pinched cylindrical shell with rigid end diaphragms

140

SACS® Collapse Advanced Release 15.1

 Figure 145: Displacement of elastic pinched cylindrical with rigid end diaphragms

Figure 146: Deformed pinched cylindrical with rigid end diaphragms. Geometrical lock is marked with a
red circle.

141

SACS® Collapse Advanced Release 15.1

5.2.10 Sample 10: Elastoplastic Cantilever Beam
Figure 147 shows a cantilever beam subjected to a lateral point load at the free end. The beam is
modeled using plate elements. Two models are considered for this analysis: 1) Using 10 quadrilateral
plate elements. 2) Using 10×2 triangular plate elements. The elastoplastic material properties were
assumed as shown in Figure 147. The analysis was conducted using both thin plate (Kirchhoff) and thick
plate (Mindlin) theories. The applied load was gradually increased until the beam tip vertical
displacement reached 2.5 m. Arc-length method was not used for this benchmark since buckling is not
involved in this case. The predicted vertical displacement from the program was compared against
existing finite element results [33] as shown in Figure 148.

𝑤(𝑚 )

𝑃(𝐾𝑁)

𝐾𝑁 𝐾𝑁
𝐸 = 120 𝑐𝑚2 , 𝐸 = 2.4 𝑐𝑚2 , 𝜈 = 0.3 , 𝑠𝑡𝑟𝑎𝑖𝑛 ℎ𝑎𝑟𝑑𝑒𝑛𝑖𝑛𝑔 𝑟𝑎𝑡𝑖𝑜 = 0.01

𝐿 = 10 𝑚, 𝑤𝑖𝑑𝑡ℎ = 1 𝑚, 𝑇ℎ𝑖𝑐𝑘𝑛𝑒𝑠𝑠 = 10 𝑐𝑚

Figure 147: Elastoplastic cantilever beam with end moment

Figure 148: Elastoplastic cantilever end displacement

142

SACS® Collapse Advanced Release 15.1

5.2.11 Sample 11: Elastoplastic Simply-supported Plate under uniform Pressure
Figure 149 shows a simply supported rectangular plate without restraints for in plane displacement and
rotation, subjected to uniform pressure. Due to symmetry, only a quarter of the plate was analyzed
using 24x24 quadrilateral plate elements. The analysis was conducted using both thin plate (Kirchhoff)
and thick plate (Mindlin) theories. The results predicted by the program were compared against existing
analytical results [34] as shown in Figure 150. Figure 151 also shows the plate deformation and plasticity
ratio at load factors of 30 and 60. Due to the large deformations involved in this benchmark, the load
factor was applied over 60 steps and the sub-incrementation method was used to help improve the
convergence rate.

ℎ𝑖𝑛𝑔𝑒𝑑
𝑤 (𝑐𝑚)
𝐿 (𝑐𝑚) 𝐾𝑁
𝑞0 ൬ ൰
𝑚2

𝐿 (𝑐𝑚)
ℎ𝑖𝑛𝑔𝑒𝑑
𝐾𝑁 𝐾𝑁
𝐸 = 6900 2
, 𝜈 = 0.3 , 𝐹𝑦 = 24.8 , 𝐸𝑙𝑎𝑠𝑡𝑖𝑐 − 𝑃𝑒𝑟𝑓𝑒𝑐𝑡 𝑃𝑙𝑎𝑠𝑡𝑖𝑐
𝑐𝑚 𝑐𝑚 2
𝐾𝑁
𝐿 = 0.508𝑚, 𝑡ℎ𝑖𝑐ℎ𝑛𝑒𝑠𝑠 = 0.254𝑐𝑚 , 𝑞0 = 10 2
𝑚

Figure 149: Elastoplastic simply-supported plate with large transverse shear deformation

Figure 150: Vertical displacement of elastoplastic simply-supported plate center

Figure 151: Deformed shape of elastoplastic simply-supported plate and plastic ratio contours. Left) at
load factors 30, Right) at load factor 60
143

SACS® Collapse Advanced Release 15.1

5.2.12 Sample 12: Elastoplastic Pinched Cylindrical Shell with Rigid End Diaphragms
Figure 152 shows an elastic cylindrical shell with rigid diaphragms at both ends which prevent in plane
deformation subjected to a pinching load. Due to symmetry, only 1/8th of the shell was modelled, and
because of geometric locking (as explained in Sample 9), a fine mesh of 60 quadrilateral elements
around the circumference and 40 elements along the width of the cylinder was used. Because of the
large deformations and elastoplastic behavior involved in this benchmark, the load was applied over 100
steps. To improve convergence and to predict local buckling, both sub-incrementation and Arc-length
methods were selected on the CLPOPT input line. The analysis was conducted using both thin plate
(Kirchhoff) and thick plate (Mindlin) theories. The vertical displacement of the cylinder center was
compared against exiting finite element analysis [35] as shown in Figure 153.
Figure 154 shows the deflected cylinder with plastic ratio contours at load factors of 0.2, 0.5 and 1.0.

𝑃
(𝐾𝑁)
𝑃 (𝐾𝑁) 4
−𝑤𝐴 (𝑚)

𝑟𝑖𝑔𝑖𝑑 𝑟𝑖𝑔𝑖𝑑

𝐿 (𝑚)
𝑃 (𝐾𝑁)

𝐾𝑁 𝐾𝑁
𝐸 = 3000 2
, 𝜈 = 0.3, 𝑠𝑡𝑟𝑎𝑖𝑛 ℎ𝑎𝑟𝑑𝑒𝑛𝑖𝑛𝑔 𝑟𝑎𝑡𝑖𝑜 = 0.0909 𝑜𝑟 𝐻 = 300 2
𝑐𝑚 𝑐𝑚
𝑅 = 3𝑚, 𝐿 = 6𝑚 𝑡ℎ𝑖𝑐𝑘𝑛𝑒𝑠𝑠 = 3𝑐𝑚

Figure 152: Elastic pinched cylindrical shell with rigid end diaphragms

Figure 153: Displacement of elastoplastic pinched cylindrical with rigid end diaphragms

144

SACS® Collapse Advanced Release 15.1

 Figure 154: Deformed elastoplastic pinched cylindrical with rigid end diaphragms. Left) load factor 0.2,
Center) load factor 0.5 and Right) load factor 1.0

5.2.13 Sample 13: Elastoplastic Buckling of Cylindrical Roof

Figure 155 shows a model of a shallow cylindrical roof with two hinged sides subjected a uniformly
𝐾𝑁
distributed load of 4 𝑚2 . Due to the symmetry of the structure and boundary conditions, only one
quarter of the structure is modeled. The curved ends of the roof are supported by rigid diaphragms and
the two longitudinal edges are assumed to be free. Two models were considered: 1) Using 20×20
quadrilateral elements. 2) Using 20×20×2 triangular elements. The applied vertical load was gradually
increased and the Arc-length method was used to predict buckling. The analysis was conducted until
mid-point of the free edge reached a maximum vertical displacement of 200 cm. The results from the
analysis were compared against existing finite element results [36] as shown in Figure 156.
Figure 157 shows the deflected configuration with plastic ratio contours at the peak load factor and the
final load step.
𝑟𝑖𝑔𝑖𝑑

𝐾𝑁
𝐿 (𝑚) 4
𝑚2

−𝑤 (𝑐𝑚)

𝑟𝑖𝑔𝑖𝑑
𝐾𝑁 𝑅
𝐸 = 2100 , 𝜈 = 0.0
𝑐𝑚2
𝑅
𝐾𝑁 𝜃 = 40°
𝐹𝑦 = 0.42 , 𝑒𝑙𝑎𝑠𝑡𝑖𝑐 − 𝑝𝑒𝑟𝑓𝑒𝑐𝑡𝑙𝑦 𝑝𝑙𝑎𝑠𝑡𝑖𝑐
𝑐𝑚 2
𝑅 = 7.6𝑚, 𝐿 = 15.2𝑚, 𝑡ℎ𝑖𝑐𝑘𝑛𝑒𝑠𝑠 = 7.6 𝑐𝑚

Figure 155: Cylindrical Roof supported by rigid diaphragms

145

SACS® Collapse Advanced Release 15.1

 Figure 156: Vertical Displacement at free edge

Figure 157: Buckling of the cylindrical roof. Top-Left) load factor 0.5, Top-right) load factor 1.0, Bottom-
left) maximum load factor, Bottom-right) final load step.

146

SACS® Collapse Advanced Release 15.1

5.3 Joints
This section illustrates benchmarks for various joint calculations. The first section compares joint
flexibility methods (including Fessler, Buitrago and MSL) for T/Y joints. The results are compared with
joints which have been meshed using triangular plate elements. The second and third set of samples
compares the performance of the MSL joint calculations for X and K joints, respectively.

5.3.1 Sample 1: T and Y Joint Flexibility

Figure 158 illustrates three T/Y joints whose connection properties are summarized in Table 4. The joints
are subjected to a compressive and tensile axial load at the brace end. In addition to the axial loads, two
lateral loads are applied at the brace tip to produce in-plane and out-of-plane bending in the joints. The
analyses are conducted using Fessler’s (‘JF’ on CLPOPT) and Buitrago’s (‘BF’ on CLPOPT) joint flexibility
methods. For the MSL approach both mean and characteristic levels (‘MF’ and ‘CF’ on MSLOPT) are
considered. This sample focuses on the flexibility of T/Y joint types and joint strength is not considered.
The material properties are set to the SACS default values and eight sub-segments per member are
considered. Both arc-length (‘AL’ on CLPOPT) and sub-incrementation (‘SI’ CLPOPT) methods are
implemented to improve results for large deformations.

Table 4: Properties of T/Y connections

Connection 𝜃 𝛽 𝛾 𝜏
Case 1 (T Joint) 90° 0.500 15.000 0.250
Case 2 (Y Joint) 55° 0.400 16.667 0.667
Case 3 (Y Joint) 70° 0.750 15.000 0.750

The results from the various joint flexibility approaches are compared against those obtained from joints
which have been meshed using triangular plate elements as shown in Figure 158. It should be noted that
a meshed joint provides an accurate representation of the joint by including the chord flexibility, the
chord plasticity and large local deformations. The brace tip (in-plane and out-of-plane) displacements
are compared with results obtained for meshed joints in Figure 159, Figure 160, and Figure 161. The
results in Figure 159 show that all joint flexibility methods perform well for T joints, in particular for the
case where the brace is in tension. For other cases, MSL joint flexibility gives better results than Fessler
and Buitrago methods.

147

SACS® Collapse Advanced Release 15.1

 Case 1: T Joint

Case 2: 55° Y Joint

Case 3: 70° Y Joint

Figure 158: T and Y Joints and meshed joints

148

SACS® Collapse Advanced Release 15.1

 Figure 159: Comparison of brace tip displacement for Case 1 (T joint)

149

SACS® Collapse Advanced Release 15.1

 Figure 160: Comparison of brace tip displacement for Case 2 (Y joint at angle of 55°)

150

SACS® Collapse Advanced Release 15.1

 Figure 161: Comparison of brace tip displacement for Case 3 (Y joint at angle of 70°)

151

SACS® Collapse Advanced Release 15.1

5.3.2 Sample 2: Frame with X Joint
This sample considers the application of MSL approach for X joints with reference to two-story frame
with X bracing subject to a lateral load as shown in Figure 162. The section dimensions for all members
are also labelled in Figure 162. The X joint chord segments in the frame have the yield stress of 35
kN/cm2, the material properties of all other members are set to the SACS default values. Both arc-length
(‘AL’ on CLPOPT) and sub-incrementation (‘SI’ CLPOPT) methods are implemented to improve results for
large deformations. The analysis is carried out for three cases outlined below:

1. Without any joint flexibility or strength check

2. MSL Mean Level: ‘MF’, flexibility, ‘MS’ strength, and ‘MT’ on MSLOPT line
3. Characteristic Level: ‘CF’, flexibility, ‘CS’ strength, and ‘CT’ on MSLOPT line

This analysis focusses on the performance of the X joints in the frame and therefore, the input lines
JSSEL and JFSEL are used to select joints ‘JNT1’ and ‘JNT2’ only, all other joints are ignored.

The results are compared with those obtained from a structural model in which the X joints have been
meshed using triangular plate elements as shown in Figure 162. The structure with the meshed joints is
analyzed for the following two cases:

1. All plates associated the meshed joints are modelled as elastoplastic to include the chord
flexibility, the chord plasticity, and large local deformations.
2. Plates associated with the connection chord are assumed to remains elastic to include the joint
flexibility but ignoring the large deformations of the chord elastoplastic response.

The structural responses for all cases are plotted in Figure 163. It can be seen, that the results obtained
from MSL approach implemented in the program are in good agreement with those obtained from the
meshed joints. The MSL Mean Level analysis reports joint strength failure at a lateral load value of 10
MN whilst the MSL Characteristic Level Analysis predicts the joint failure at a load value of 9 MN.

152

SACS® Collapse Advanced Release 15.1

 𝑢 (𝑐𝑚)

𝑃 (𝑀𝑁)

Figure 162: Two-story frame with cross connections. Left) member sections, right) Meshed joints

Figure 163: Comparison of two-story frame with X connections

153

SACS® Collapse Advanced Release 15.1

5.3.3 Sample 3: Frame with K joint
This sample considers the application of the MSL approach for K joints with reference to a two-story
frame with K bracing subjected to a lateral load as shown in Figure 164. The section dimensions for all
members are labeled in Figure 164, the material properties of all members are set to the SACS default
values. Both arc-length (‘AL’ on CLPOPT) and sub-incrementation (‘SI’ CLPOPT) methods are
implemented to accurately predict results for large deformations. The analysis is carried out for the
following three cases:

1. Without any joint flexibility or strength checks

2. MSL Mean Level: ‘MF’, flexibility, ‘MS’ strength, and ‘MT’ on MSLOPT line
3. Characteristic Level: ‘CF’, flexibility, ‘CS’ strength, and ‘CT’ on MSLOPT line

This analysis focusses on the performance of the K connections and therefore input lines JSSEL and JFSEL
are used to select joints ‘JNT1’ and ‘JNT2’ only, flexibility and strength checks of all other joints are
excluded.

The results from the above three cases are compared against those obtained from a structural model in
which the joints have been meshed using triangular plates as shown in Figure 164. The structure
containing the meshed joints is analyzed for the following two cases:

1. All plates associated the meshed joints are modelled as elastoplastic to include the chord
flexibility, the chord plasticity, and large local deformation.
2. Plates associated with the connection chord are assumed to remain elastic to include the joint
flexibility but ignore large deformations resulting from the chord elastoplastic response.

The structural responses for all cases are plotted in Figure 165. It can be seen that the results predicted
by the MSL approach are in good agreement with those obtained from the structural model containing
the meshed joints.

154

SACS® Collapse Advanced Release 15.1

 𝑢 (𝑐𝑚)

𝑃 (𝑀𝑁)

Figure 164: Two-story frame with K connections. Left) member sections, right) Meshed joints

Figure 165: Comparison of two-story frame with K connections

155

SACS® Collapse Advanced Release 15.1

5.4 Local Buckling
This section considers the implementation of the local buckling calculations for tubular and conical
sections based on following methods: Marshall and Gates [9], API Bulletin 2U, API RP 2A-LRFD, and ISO
19902. Results from the program are compared with hand-calculation based on the formulation given in
section 6.12 in Commentary.

5.4.1 Sample 1: Local Buckling in Tubular Members

The following example looks at the local buckling calculations for tubular members. Figure 166
illustrates a cantilever beam subjected to an axial compressive and lateral load at the free end of the
beam. The analysis is conducted using eight sub-segments, SACS default values are used for the material
properties. Both sub-incrementation option (‘SI’ on CLPOPT) and arc-length methods (‘AL’ on CLPOPT)
are considered to accurately predict the post-local buckling behavior. Table 5 shows the tubular section
dimensions for various local buckling approaches to ensure that local buckling occurs before yield.

Table 5: Tubular sections for local buckling analysis

Method Diameter (cm) Wall Thickness (cm)

Marshall and Gates 120.0 1.0
API Bulletin 2U 130.0 0.8
API LRFD 120.0 1.0
ISO 19902 120.0 1.0

As discussed in Commentary section 6.12.5, the program checks the local buckling at the mid-point
sections of every sub-segment. For this case, the critical section is the mid-point of the first sub-segment
𝐿
which is located at from the fixed end. From hand calculations, assuming small deformations, the
16
𝐿
𝑃 𝑃(𝐿− )𝑅
16
stress at the critical section is given by 𝐴
+ 𝐼
, where 𝑅 is the radius, 𝐴 is the cross-section area,
and 𝐼 is the second moment of the area. Figure 167 shows local buckling results from the program and
those obtained using hand calculations. It can be seen that results from both approaches are in good
agreement.

𝑐𝑟𝑖𝑡𝑖𝑐𝑎𝑙 𝑠𝑒𝑐𝑡𝑖𝑜𝑛 𝑓𝑜𝑟 𝑙𝑜𝑐𝑎𝑙 𝑏𝑢𝑐𝑘𝑙𝑖𝑛𝑔

𝑃 (𝑘𝑁)
𝐿 𝐿 = 5𝑚 𝑃 (𝑘𝑁)
16

Figure 166: Cantilever beam for local buckling analysis

156

SACS® Collapse Advanced Release 15.1

 Figure 167: Comparison of local buckling load for the tubular members

5.4.2 Sample 2: Local Buckling in Conical Members

This benchmark considers the local buckling approach implemented for conical sections. Figure 168
illustrates a tubular cantilever beam comprised with a conical section at the fixed end. The beam is
subjected to an axial compressive and a lateral load at the free end. The analysis is conducted using
seven sub-segments for the tubular member and one sub-segment for the conical section to simplify
hand-calculations. The material properties are set to SACS default values and both sub-incrementation
option (‘SI’ on CLPOPT) and arc-length method (‘AL’ on CLPOPT) are implemented to accurately predict
the post-local buckling response. Table 6 shows the conical section dimensions for various local buckling
approaches to ensure the local buckling occurs before yielding. The dimensions of the tubular member
are assumed as 50cm diameter and 2cm wall thickness.

Table 6: Conical sections for local buckling analysis

Method Larger Diameter (cm) Smaller Diameter (cm) Wall Thickness (cm)
Marshall and Gates 120.0 50 1.0
API Bulletin 2U 130.0 50 0.8
API LRFD 120.0 50 1.0
ISO 19902 120.0 50 1.0

As discussed in Commentary section 6.12.5, the program checks for local buckling at the mid-point
section of every sub-segments. For this sample, the critical section will be the mid-point section of the
𝐿
conical member located at 16 from the fixed end. From the hand calculations, assuming small
𝐿
𝑃 𝑃(𝐿− )𝑅
16
deformations, the stress at the critical section is given by 𝐴
+ 𝐼
where 𝑅 is conical section radius,

157

SACS® Collapse Advanced Release 15.1

𝐴 is the cross-section area, and 𝐼 is the second moment of the area at the critical section. As discussed in
the Commentary sections 6.12.3 and 6.12.4, based on API RP 2A-LRFD and ISO 19902 recommendations,
𝐷𝑠
the program utilizes cos (𝛼)
(where 𝐷𝑠 is diameter at the critical section and 𝛼 is the half of the conical
apex angle) to calculate the local buckling limit stresses. For the Marshall and Gates or API Bulletin 2U
options, the program uses the largest diameter of the cone to calculate the local buckling limit stress.

Figure 169 summarizes the results from the Collapse program for the loads for which the local buckling
occurs. It can be seen the results are in good agreement with hand-calculations.

𝑐𝑟𝑖𝑡𝑖𝑐𝑎𝑙 𝑠𝑒𝑐𝑡𝑖𝑜𝑛 𝑓𝑜𝑟 𝑙𝑜𝑐𝑎𝑙 𝑏𝑢𝑐𝑘𝑙𝑖𝑛𝑔

𝑃 (𝑘𝑁)
𝐿 𝐿 = 5𝑚 𝑃 (𝑘𝑁)
16

Figure 168: Cantilever beam for local buckling analysis

Figure 169: Comparison of local buckling load for the conical member

158

SACS® Collapse Advanced Release 15.1

5.5 Multilinear Elastoplastic Materials
This section considers the implementation of the multilinear elastoplastic material for beam members,
plates and piles in collapse analysis. The results are compared against standard bilinear elastoplastic
material. Also, the beam members and plates results are verified against ADINA finite element software.

5.5.1 Sample 1: Multilinear Elastoplastic Materials for Members

This benchmark considers the elastoplastic analysis of a tubular beam with different elastoplastic
materials. Figure 170 illustrates a beam clamped at its both ends and subjected to a load at mid-point.
Three different elastoplastic materials are considered for this benchmark: 1) standard bilinear with
0.005 strain hardening ratio (defined on CLPOPT line), 2) multilinear elastoplastic with strain hardening,
and 3) multilinear elastoplastic with strain softening. The elastoplastic stress-strain curves are presented
in Figure 171-left.

The model consistent of two members and the analysis is conducted using eight sub-segments per
member. Both sub-incrementation option (‘SI’ on CLPOPT) and arc-length method (‘AL’ on CLPOPT) are
implemented to accurately predict the post-local buckling response. Figure 171-right compares the
vertical displacement at the mid-point for different material models. The models with strain hardening
and softening material are also analyzed ADINA finite element software. It can be seen that Collapse
Advanced results are in good agreement with ADINA.

𝐿 = 10𝑚 𝑤(𝑐𝑚 ) 𝐿 = 10𝑚

𝑃 (𝑘𝑁)
𝑘𝑁 𝑘𝑁
𝐷 = 80 𝑐𝑚 𝑡 = 1 𝑐𝑚 𝐸 = 20000 2
𝜎𝑦 = 24.8
𝑐𝑚 𝑐𝑚 2

Figure 170: A tubular beam with fixed ends and subjected to a mid-point load

Figure 171: Left: Stress-strain curve for different post-yield behavior, Right: the tubular beam vertical
displacement for different material models.

159

SACS® Collapse Advanced Release 15.1

5.5.2 Sample 2: Multilinear Elastoplastic Materials for Plates
This benchmark considers the elastoplastic analysis of a plate with different elastoplastic materials.
Figure 172 illustrates a plate clamped at its one side and subjected to a load at the corner. Three
different elastoplastic materials are considered for this benchmark: 1) standard bilinear with 0.005
strain hardening ratio (defined on CLPOPT line), 2) multilinear elastoplastic with strain hardening, and 3)
multilinear elastoplastic with strain softening. The elastoplastic stress-strain curves are presented in
Figure 173-left.

The model consistent of 5x5 mesh with total of 25 quadrilateral thin plate element. Both sub-
incrementation option (‘SI’ on CLPOPT) and arc-length method (‘AL’ on CLPOPT) are implemented to
accurately predict the post-local buckling response. Figure 173-right compares the vertical displacement
at the mid-point for different material models. The models with strain hardening and softening material
are also analyzed ADINA finite element software. It can be seen that Collapse Advanced results are in
good agreement with ADINA.

𝐿 = 1𝑚
𝐿 = 1𝑚 𝑤(𝑐𝑚 )

𝑃 (𝑘𝑁)

𝑘𝑁 𝑘𝑁
𝑡 = 1 𝑐𝑚 𝐸 = 20000 2
𝜎𝑦 = 24.8
𝑐𝑚 𝑐𝑚 2

Figure 172: A plate clamped at one side and subjected to a corner load

Figure 173: Left: Stress-strain curve for different post-yield behavior, Right: vertical displacement at
plate corner for different material models.

160

SACS® Collapse Advanced Release 15.1

5.5.3 Sample 3: Multilinear Elastoplastic Materials for Piles
This benchmark considers the elastoplastic analysis of a monopile with different elastoplastic materials.
Figure 174 illustrates a monopile subjected to a vertical load and a lateral. The vertical load is first
applied to the monopile and then the lateral load gradually increases until the maximum deflection
reaches to 20 inches. Three different elastoplastic materials are considered for this benchmark: 1)
standard bilinear with 0.005 strain hardening ratio (defined on CLPOPT line), 2) multilinear elastoplastic
with strain hardening, and 3) multilinear elastoplastic with strain softening. The elastoplastic stress-
strain curves are presented in Figure 175-left.

The model consistent a member and a pile with length of 60 feet. Both sub-incrementation option (‘SI’
on CLPOPT) and arc-length method (‘AL’ on CLPOPT) are implemented to accurately predict the post-
local buckling response. Figure 175-right compares the lateral displacement of the pilehead for different
material models.

𝐻(𝑘𝑖𝑝𝑠)
𝑃 (𝑘𝑖𝑝𝑠)
𝑢 (𝑖𝑛) 𝑝𝑖𝑙𝑒 𝑙𝑒𝑛𝑔𝑡ℎ = 60 𝑓𝑡
𝐷 = 30 𝑖𝑛
𝑡 = 0.5 𝑖𝑛
𝐸 = 29000 𝑘𝑠𝑖
𝜎𝑦 = 36 𝑘𝑠𝑖

Figure 174: A monopile subjected to vertical and horizontal loads

Figure 175: Left: Stress-strain curve for different post-yield behavior, Right: lateral displacement of
pilehead for different material models.

161

SACS® Collapse Advanced Release 15.1

5.6 Impact Analysis
This example problem demonstrates the application of the IMPACT input line for dropped object and
ship impact analysis using user-defined loads in SACS model file.

5.6.1 Sample 1: Dropped Object

Following sample considers the use of the IMPACT line for a dropped object analysis. The impact loads
are defined as 9 vertical point loads as shown in Figure 176. The total energy absorbed by the structure
is set to 100 ft-kips.

Collapse Advanced program input file is as follows:

1 2 3 4 5 6 7 8
12345678901234567890123456789012345678901234567890123456789012345678901234567890
CLPOPT 20 8 20 ALSI 0.0010.0010.0011000.0.002
CLPRPT P1 SM
JTSEL E131
LDSEQ IMP1 SELF 1 1.0IMP1 200 0.0 200.0
IMPACT IMP1 100.00
END

The following is a detailed description of the input file:

A. The CLPOPT line specifies the use of both Arc-length (‘AL’) and Sub-incrementation (‘SI’)
methods to increase performance of the nonlinear analysis.
B. The CLPRPT line is used to generate the collapse summary report (‘SM’) and joints deflection
report (‘P1’) for every load increment. JTSEL input line is used to select joint ‘E131’ for deflection
output (model center point).
C. The LDSEQ line defines the load sequence for impact analysis. The self-weight (‘SELF’) load case
is first applied to the structure and it is followed by impact load case ‘IMP1’.
a. The large load factor of 200 is chosen to ensure the analysis continue until the entire
energy absorbed by the structure.
D. The IMPACT line is used to define the impact load case ‘IMP1’. This load case is used by the
program to calculate the energy absorbed by the structure. The maximum impact energy is set
to 100.0 ft-kips.
a. Other impact options are not required for this dropped object analysis.

The dropped object results are given in Figure 177 and Figure 178. The program will continue the
analysis based on the defined load sequence until the maximum impact energy is absorbed by the
structure. Once the maximum energy is absorbed, the program automatically unloads the structure until
the impact load factors returns to zero.

162

SACS® Collapse Advanced Release 15.1

Figure 176: Dropped object model using IMPACT input line. Top) The user-defined impact load given in
SACS model file, Bottom) finite element mesh.

163

SACS® Collapse Advanced Release 15.1

 Figure 177: Dropped object results. Left) force-deflection for vertical displacement at center-point,
Right) Energy absorbed by the structure versus the load step.

Figure 178: Plasticity contour for dropped object analysis.

164

SACS® Collapse Advanced Release 15.1

5.6.2 Sample 2: Ship Impact
This example problem shows the use of the IMPACT line for a ship impact analysis. For this example, the
mass of the ship is assumed to be 1250 tons, impact velocity of 6 m/sec, angle of impact is 135° with
respect to the positive x-axis. The Collapse Advanced program input file is given below:

1 2 3 4 5 6 7 8
12345678901234567890123456789012345678901234567890123456789012345678901234567890
CLPOPT 20 8 20 ALSI 0.010.001 0.01127.00.002
CLPRPT P1 SM
LDSEQ IMP1 IMPC 25 5.0
IMPACT IMPC 0000 DNV1 E 607 0000 8. 20. PD
ENERGY 1250. 1.0 6.
JTSEL 0000
GRPELA 14A 14B 14C 14D 17A 17B 19A 19B 19C 19D 21A 21B 21D 22A 22C
GRPELA 24A 24B 24C 24D 24E 24F 24G 24H BL1 BL2 BL3
GRPELA CAA CAB CA7 CB7 CF9 P00 P14 P17 P19 P1E P21 P22 P24 P2E
GRPELA PC1 PC2 R16 R18 RS1 RS2
GRPELA SLV T01 T02 T06 T08 T09 T11 T12 T13 T14 T15 T16 T19 T20 T25
GRPELA T3A T6A T7A V1A W51 W61 WBN YPL
PGRELA P01 P02 P03 P04 PLT
END

The following is a detailed description of the input file:

A. The CLPOPT line specifies the use of both Arc-length (‘AL’) and Sub-incrementation (‘SI’)
approaches to increase performance of the nonlinear analysis.
B. The CLPRPT input line is used generate summary report (‘SM’) and joints deflection report (‘P1’)
for every load increment. JTSEL input is used to select the impact joint ‘0000’ for deflection
output.
C. The LDSEQ input line defines the load sequence for impact analysis using the impact load case
‘IMPC’.
a. The load factor of 5.0 is used to ensure the analysis continues until the entire energy is
absorbed by the structure.
D. The ENERGY input line is used to define the impact energy:
a. The ship mass is 1250.0 ton, and the added mass coefficient is set to 1.0.
b. The impact velocity is 6 m/sec.
E. The IMPACT defines following analysis options
a. The impact load case ‘IMPC’ is selected, the maximum impact energy is left blank since
the energy is defined by ENERGY input line.
b. The impact joint ‘0000’ selected, the applied forces on this joint are used for
member/ship dent calculation.
c. The ship indentation force-displacement characteristics are defined by the ‘DNV1’
option for a bow impact in accordance to DNV RP-C204.
d. The input option ‘E’ implements the Ellinas dent calculation formulation based on
properties of member 607-0000.
e. The member dent energy and indentation depth are limited by entering 8 for factor B –
i.e. the depth is limited by 1/8th of the member dimeter.
f. The member dent energy is limited by maximum of 20% of the total impact energy.
g. The option ‘PD’ implements the plastic member dent and ship dent energy calculations.

165

SACS® Collapse Advanced Release 15.1

 F. The GRPELA and PGRELA are used to define non-structural elements as elastic to improve
performance.

The impact load case ‘IMPC’ is modelled as two horizontal loads as shown in Figure 184 to represent
135° impact direction.

Figure 179: Ship impact model using IMPACT input line. Left) SACS jacket model, Right) point loads to
model 135° ship impact.

The program will continue the analysis based on the defined load sequence until the total impact energy
is absorbed by the structure. Once the total energy is absorbed, the program automatically unloads the
model until the impact load factors returns to zero. The ship impact results are given in Figure 180 and
Figure 181. The listing file prints following summary report (using ‘SM’ option on CLPOPT line) for ship
impact analysis:

SACS COLLAPSE IMPACT ENERGY ABSORPTION SUMMARY REPORT

-----------------------------------------------------

Load Sequence IMP1

Load Condition IMPC

Impact Summary at Peak Energy Absorption

----------------------------------------------------

Impact energy = 22.5000 (MJ)

Energy absorbed by structure = 11.3331 (MJ)
Energy absorbed by ship = 10.3451 (MJ)
Energy absorbed by member = 0.8660 (MJ)
Total absorbed impact energy = 22.5441 (MJ)
% of total energy absorbed = 100.1962 (%)
Ship indentation = 1.5089 (m)
Member indentation = 0.0993 (m)

166

SACS® Collapse Advanced Release 15.1

Figure 180: Ship impact results. Left) Energy absorption by structure, ship and member dent, Right) The
ship dent depth and member dent depth versus the load step.

Figure 181: Ship impact results. Left) Plasticity contour for ship impact analysis, Right) The force-
deflection curve for the impact joint

167

SACS® Collapse Advanced Release 15.1

5.7 Dynamic Response Samples
5.7.1 Ship Impact
The following is an example of a dynamic ship impact analysis using Dynamic Response and Collapse
Advanced to calculate the non-linear behavior of the jacket when a vessel strikes the jacket leg.

The structure in Figure 182 is comprised of a four-leg jacket with skirt piles and a topside made up of
two decks with a crane pedestal and a helideck. Other modeled appurtenances include a boat landing,
risers, and conductors. The impact member is meshed using plate elements and the impact joint is 31P7.

Figure 182: Ship Impact Model and meshed impact member

A mode shape extraction is performed to generate the Dynpac Modal Solution File (dynmod) and
Dynpac Mass File (dynmas) using 40 modes. Dynamic Response uses these files to calculate the dynamic
loading on the structure. More information about the mode shape extraction analysis and theory can be
found in the Dynpac manual. The following is the Dynamic Response input file used to define the
dynamic behavior and loading of the structure:

1 2 3 4 5 6 7 8
12345678901234567890123456789012345678901234567890123456789012345678901234567890
DROPT SHIP EC+Z
SDAMP 5.0
LOAD
SHIP 1250. 6.0 180.0 1.0 31P7
THLOAD SHIP SDO CLP
JTNUM 31P7 701
TIME 2.00000 0.0100 1.0E-9 1.0000
END

168

SACS® Collapse Advanced Release 15.1

The following is a detailed description of the Dynamic Response input file:

G. The DROPT line specifies the analysis options.

a. A ship impact analysis is to be performed (‘SHIP’ in columns 7-10).
H. The SDAMP line specifies that the overall structural damping is 5.0 percent.
I. The LOAD header line specifies that the loading data is too follow.
J. The SHIP line defines the ship impact parameters:
a. The ship weight, 1250 ton, is entered in columns 9-18.
b. The initial velocity, 6 m/sec, is entered in columns 21-27.
c. The ship direction, 180 degrees from the global X axis, is entered in columns 28-34.
d. The distance before impact, 1.0 m, is entered in columns 42-48.
e. The impact joint, 31P7, is entered in columns 63-66.
K. The THLOAD line defines the time history input parameters:
a. The time history source, SHIP, is defined in columns 9-12.
b. The damping type, SDO – Structural Damping Only, is defined in columns 18-20.
c. Various reporting options can be entered in columns 33-56. It is important to note that
the ‘CLP’ option indicates that Dynamic Response will create a collapse input file with
the time history loads and the ‘ALL’ option indicates that Dynamic Response will create
loads for all time points.
L. The JTNUM line selects joints 31P7 and 701 for reports specified on the THLOAD line.
M. The TIME line specifies time integration control parameters, starting at 0.0 seconds and ending
at 2.0 seconds with an output time interval of 0.01 seconds.

The ship impact analysis will perform the dynamic response analysis, generate a collapse input file using
the loads generate at each time point, and then run a collapse analysis using the generated collapse
input file. However, collapse options and parameters must first be defined in a partial input file. The
following is the partial collapse input file used to define the Collapse analysis behavior:

1 2 3 4 5 6 7 8
12345678901234567890123456789012345678901234567890123456789012345678901234567890
CLPOPT 20 8 20 ALSI 0.010.001 0.01127.00.002
CLPRPT P1R0M0MP
JTSEL 31P7
GRPELA 14A 14B 14C 14D 17A 17B 19A 19B 19C 19D 21A 21B 21D 22A 22C
GRPELA 24A 24B 24C 24D 24E 24F 24G 24H BL1 BL2 BL3 CA3 CA4 CA5 CA7
GRPELA CAA CAB CB5 CB7 CF9 CND P00 P14 P17 P19 P1E P21 P22 P24 P2E
GRPELA PC1 PC2 R08 R12 R16 R18 RS1 RS2 RS3 SLV ST1 ST3 ST4 ST5 ST7
GRPELA ST9 T01 T02 T06 T08 T09 T11 T12 T13 T14 T15 T16 T19 T20 T25
GRPELA T3A T6A T7A V1A W51 W61 WBN YPL
PGRELA P01 P02 P03 P04 PLT
LDAPL DEAD SELF 1 0.0 1.0

The following is a detailed description of the partial Collapse input file:

A. The CLPOPT line defines the collapse options:

a. The maximum iterations per load increment, 20, is defined in columns 11-13.
b. The number of member segments, 8, is defined in columns 14-16.
c. The maximum number of iterations, 20, is defined in columns 17-19.

169

SACS® Collapse Advanced Release 15.1

 d. Arc-length method for post-buckling and sub-incrementation to improve convergence are
selected in columns 42-43 and 44-45, respectively.
e. The deflection tolerance, 0.01, is defined in columns 56-60. The rotation tolerance, 0.001, is
defined in columns 61-65. The member deflection tolerance, 0.01, is entered in columns 66-
70. These values are used to determine whether a load increment has converged.
f. The maximum deflection (collapse deflection) is assumed 127 cm (50 inch) in columns 71-75.
g. The strain hardening ratio, 0.002, is defined in columns 76-80. This value determines the
plastic behavior of elements.
B. The CLPRPT line specifies the collapse reporting options.
C. The GRPELA lines indicate which members remain elastic during the analysis. These are deck
members, conductors, risers, boat landings, and other non-primary structural appurtenances.
The intention is to solely capture the plastic behavior of the jacket primary structure.
D. The PGRELA lines indicate which plates will remain elastic during the analysis. These are the
deck plate and skirt pile connecting plates.
E. The LDAPL line defines the first user-generated load step:
a. The load sequence ID, DEAD, is entered in columns 7-10.
b. The load case name, SELF, is entered in columns 21-24. This must match a load case defined
in the SACS model file used for the collapse analysis.
c. The number of increments, 1, is entered in columns 25-29.
d. The ending load factor, 1.0, is entered in columns 37-43. A blank entry for the starting load
factor indicates that the load will start at 0.0.
F. The LDAPC input lines associated with time-history steps are automatically generated and added
by Dynamic Response.

The ship impact analysis was allowed to continue until X displacement of impact joint 31P7 exceeds the
maximum deflection of 127 cm (50 inch) entered on CLPOPT line. The response of the impact joint 31P7
and plasticity contours are shown in Figure 183 and Figure 184, respectively.

Figure 183: Base shear versus XY displacement of impact joint 31P7

170

SACS® Collapse Advanced Release 15.1

Figure 184: Plasticity contours for ship impact. Top-Left) load step 100, Top-right) load step 150, Bottom-
left) load step 175, Bottom-right) final load step.

171

SACS® Collapse Advanced Release 15.1

5.7.2 Dropped Object
The following is an example of a dropped object analysis using Dynamic Response and Collapse
Advanced to calculate the non-linear behavior of the deck when an object is dropped from a height
above the deck.

The structure in Figure 185 is a 1 cm deck plate supported by 400mm x 200m I beams with 5mm web
and 10mm flange thicknesses spaced at 1 meter. The beams have been meshed with quadrilateral
isotropic plate elements to model the interaction between the beams and the deck plate more
accurately. A finer triangular plate mesh is modeled near the impact location of the dropped object. A
self-weight load has been applied to the structure. The impact joint is located at joint 712
(4.0m,4.4m,0.0m).

Figure 185: Dropped Object Model

A mode shape extraction is performed to generate the Dynpac Modal Solution File (dynmod) and
Dynpac Mass File (dynmas) using 50 modes. Dynamic Response uses these files to calculate the dynamic
loading on the structure. More information about the mode shape extraction analysis and theory can be
found in the Dynpac manual.

The following is the Dynamic Response input file used to define the dynamic behavior and loading of the
structure:

1 2 3 4 5 6 7 8
12345678901234567890123456789012345678901234567890123456789012345678901234567890
DROPT DROP EC+Z
SDAMP 2.0
LOAD
DROP 0.45SO 1.25 712
THLOAD DROP SDO PLTPLMPLSPRTALLMXSCLPJPD
JTNUM 712
TIME 1.00000 0.0010 1.0E-9 1.0000
END
s
172

SACS® Collapse Advanced Release 15.1

The following is a detailed description of the Dynamic Response input file:

A. The DROPT line specifies the analysis options.

a. A dropped object analysis is to be performed (‘DROP’ in columns 7-10).
B. The SDAMP line specifies that the overall structural damping is 2.0 percent.
C. The LOAD header line specifies that the loading data is too follow.
D. The DROP line defines the dropped object parameters:
a. The object weight, 45 ton, is entered in columns 9-18.
b. The initial velocity is left blank, indicating 0.0 m/sec, in columns 21-27.
c. The distance before impact, 1.25 m, is entered in columns 42-48.
d. The impact joint, 712, is entered in columns 63-66.
e. The adhere option, S, will analyze the dropped object analysis as a single impact in
column 19. This is useful for instances where the structure does not remain elastic.
f. The free fall option, O, will omit the objects free fall from the analysis in column 20.
E. The THLOAD line defines the time history input parameters:
a. The time history source, DROP, is defined in columns 9-12.
b. The damping type, SDO – Structural Damping Only, is defined in columns 18-20.
c. Various reporting options are entered in columns 33-56. It is important to note that the
‘CLP’ option indicates that Dynamic Response will create a collapse input file with the
time history loads and the ‘ALL’ option indicates that Dynamic Response will create
loads for all time points.
F. The JTNUM line selects joint 712 for reports specified on the THLOAD line.
G. The TIME line specifies time integration control parameters, starting at 0.0 seconds and ending
at 1.0 seconds with an output time interval of 0.001 seconds.

The dropped object analysis will perform the dynamic response analysis, generate a collapse input file
using the loads generate at each time point, and then run a collapse analysis using the generated
collapse input file. However, collapse options and parameters must first be defined in a partial input file.
The following is the partial collapse input file used to define the Collapse analysis behavior:

1 2 3 4 5 6 7 8
12345678901234567890123456789012345678901234567890123456789012345678901234567890
CLPOPT 20 ALSI 0.010.001 0.01 0.005
CLPRPT P1R0M0
JTSEL 712
LDAPL XXXX SELF 1 0.0 1.0

The following is a detailed description of the partial Collapse input file:

A. The CLPOPT line defines the collapse options:

a. The maximum iterations per load increment, 20, is defined in columns 11-13.
b. The number of member segments, 8, is defined in columns 14-16.
c. Arc-length method for post-buckling and sub-incrementation to improve convergence
are selected in columns 42-43 and 44-45, respectively.
d. The deflection tolerance, 0.01, is defined in columns 56-60. The rotation tolerance,
0.001, is defined in columns 61-65. The member deflection tolerance, 0.01, is entered in

173

SACS® Collapse Advanced Release 15.1

 columns 66-70. These values are used to determine whether a load increment has
converged.
e. The strain hardening ratio, 0.005, is defined in columns 76-80. This value determines the
plastic behavior of elements.
B. The CLPRPT line specifies the collapse reporting options.
C. The LDAPL line defines the first user-generated load step:
a. The load sequence ID, XXXX, is entered in columns 7-10.
b. The load case name, SELF, is entered in columns 21-24. This must match a load case
defined in the SACS model file used for the collapse analysis.
c. The number of increments, 1, is entered in columns 25-29.
d. The starting load factor, 0.0, is entered in columns 30-36. The ending load factor, 1.0, is
entered in columns 37-43.
D. The LDAPC input lines associated with time-history steps are automatically generated and added
by Dynamic Response.

Figure 186 shows the vertical displacement of the impact joint 712. Figure 187 and Figure 188 illustrate
the plasticity at final load step. The torsional-lateral buckling of the support beam is clearly visible in
Figure 188.

Figure 186: Joint Z – Displacement (712) vs Load Step

174

SACS® Collapse Advanced Release 15.1

 Figure 187: XY top view of the plasticity contour at last load step

Figure 188: XY bottom view of the plasticity contour at last load step.

175

SACS® Collapse Advanced Release 15.1

5.7.3 Blast
The following is an example of a blast analysis using Dynamic Response and Collapse Advanced to
calculate the non-linear behavior of a blast wall when it is subjected to a blast pressure.

The structure in Figure 189 is a 5 mm thick corrugated plate with a corrugation height of 20mm and a
corrugation width of 16mm spaced at 64mm. The blast wall is 7.168m tall by 7.2m wide. The plate has
been meshed with quadrilateral isotropic plate elements to capture the non-linear behavior of the
corrugated plate. A self-weight load and a nominal blast load has been applied to the structure. Both
loads are distributed to every joint in the structure. Each perimeter joint is fully fixed against translation
and rotation.

Figure 189: Blast Model

A mode shape extraction is performed to generate the Dynpac Modal Solution File (dynmod) and
Dynpac Mass File (dynmas) using 50 modes. Dynamic Response uses these files to calculate the dynamic
loading on the structure. More information about the mode shape extraction analysis and theory can be
found in the Dynpac manual.

176

SACS® Collapse Advanced Release 15.1

The following is the Dynamic Response input file used to define the dynamic behavior and loading of the
structure:

1 2 3 4 5 6 7 8
12345678901234567890123456789012345678901234567890123456789012345678901234567890
DROPT TCLP +Z
SDAMP 2.
LOAD
FVIB THIS CARD SDO LN PLSPLM ALL JPD
JTNUM 293 696 707 7191075
TIME 2.0 0.020 0.0005 1.0
THFORCE NON BLST
LOADC BLST 0.00 0.00
LOADC BLST 0.08 1.00
LOADC BLST 0.20 0.00
LOADC BLST 2.00 0.00
END
s

The following is a detailed description of the Dynamic Response input file:

A. The DROPT line specifies the analysis options.

a. A force time history with collapse analysis is to be performed (‘TCLP’ in columns 7-10).
B. The SDAMP line specifies that the overall structural damping is 2.0 percent.
C. The LOAD header line specifies that the loading data is too follow.
D. The FVIB line defines the force vibration analysis parameters:
a. The type of analysis is defined as a time history analysis (‘THIS’ in columns 7-10).
b. The time history input source, CARD, indicates that the time history data will be input on
subsequent lines in the dynamic response input file in columns 12-15.
c. The damping type, SDO – Structural Damping Only, is defined in columns 17-19.
d. The time history input interpolation, LN – Linear, is defined in columns 28-29.
e. Various plotting options are defined in columns 32-55.
E. The JTNUM line selects joints 293, 696, 707, 719 and 1075 for reports specified on the FVIB line.
F. The TIME line specifies time integration control parameters, starting at 0.0 seconds and ending
at 2.0 seconds with an output time interval of 0.02 seconds.
G. The THFORCE line defines the subsequent time history input.
a. The time history input form, NON – non-uniform time increments, is defined in columns
11-13.
b. The name of the time history input is defined as BLST in columns 22-25. Note that this
name is for reporting purposes and does not need to match the load case defined in the
SACS model file.
H. The LOAD lines define the blast profile loading.
a. The load case ID, BLST, is entered in columns 8-11. Note that this load case ID must be in
the SACS model file.
b. The time at which the load occurs is entered in columns 12-16. Loads at intermediate
time steps will be interpolated between defined load cases.
c. The load case factor is entered in columns 17-23. The load case specified will be
factored by this value. For this input, the maximum blast pressure occurs at 0.08
seconds.

177

SACS® Collapse Advanced Release 15.1

The dropped object analysis will perform the dynamic response analysis, generate a collapse input file
using the loads generate at each time point, and then run a collapse analysis using the generated
collapse input file. However, collapse options and parameters must first be defined in a partial input file.
The following is the partial collapse input file used to define the Collapse analysis behavior:

1 2 3 4 5 6 7 8
12345678901234567890123456789012345678901234567890123456789012345678901234567890
CLPOPT 20 8 20 ALSI 0.010.001 0.01 0.005
CLPRPT P1
JTSEL 293 696 707 1075
LDAPL XXXX SELF 1 0.0 1.0

The following is a detailed description of the partial Collapse input file:

A. The CLPOPT line defines the collapse options:

a. The maximum iterations per load increment, 20, is defined in columns 11-13.
b. Arc-length method for post-buckling and sub-incrementation to improve convergence
are selected in columns 42-43 and 44-45, respectively.
c. The deflection tolerance, 0.10, is defined in columns 56-60. The rotation tolerance,
0.001, is defined in columns 61-65. The member deflection tolerance, 0.01, is entered in
columns 66-70. These values are used to determine whether a load increment has
converged.
d. The strain hardening ratio, 0.005, is defined in columns 76-80. This value determines the
plastic behavior of elements.
B. The CLPRPT line specifies the collapse reporting options.
C. The LDAPL line defines the first user-generated load step:
a. The load sequence ID, XXXX, is entered in columns 7-10.
b. The load case name, SELF, is entered in columns 21-24. This must match a load case
defined in the SACS model file used for the collapse analysis.
c. The number of increments, 1, is entered in columns 25-29.
d. The starting load factor, 0.0, is entered in columns 30-36. The ending load factor, 1.0, is
entered in columns 37-43.
D. The LDAPC input lines associated with time-history steps are automatically generated and added
by Dynamic Response.

Figure 190 shows Y-displacement of the joint 707 versus load steps and Figure 191 presents the base
shear as a function of XY displacement of joint 707. Figure 192 illustrates the plasticity at last load step.

178

SACS® Collapse Advanced Release 15.1

 Figure 190: Joint Displacements for Joint 707

Figure 191: Sum of Y Force vs Joint Y Displacement (707)

179

SACS® Collapse Advanced Release 15.1

 Figure 192: Plasticity at last load step

180

SACS® Collapse Advanced Release 15.1

6 Commentary
6.1 Introduction
This chapter explains the theoretical formulation for the Collapse Advanced program. Like other
nonlinear finite element solvers, Collapse Advanced ensures equilibrium between applied loads 𝐟𝑎𝑝𝑝
(weight, wave/wind, ship impact, etc.) and internal forces 𝐟𝑖𝑛𝑡 (axial and shear forces, bending
moments, etc.). The equilibrium equation can be written in following form:

𝐫(𝐝) = 𝐟𝑖𝑛𝑡 (𝐝) − 𝐟𝑎𝑝𝑝 = 𝟎 (1)

in which 𝐝 is displacement/rotation vector, 𝐟𝑎𝑝𝑝 is the applied load vector associated with conservative
loading (i.e. where the orientation of the applied load in the global coordinate system is not a function
of displacement). The internal forces are assumed to be a function of displacement due to: a) material
nonlinearity resulting from large strains and b) geometrical nonlinearity due to large displacements and
rotations.

The nonlinear equation (1) may be solved iteratively using the Newton-Raphson method as:

𝐫(𝐝0 + 𝛿𝐝) ≈ 𝐫(𝐝0 ) + 𝐊|𝐝0 𝛿𝐝 = 𝟎 (2)

where 𝐝0 is initial value for displacement and 𝛿𝐝 is iterative change in displacement, 𝐫(𝐝0 ) and 𝐊|𝐝0
are the residual vector and the tangent stiffness matrix respectively, and can be represented as:

𝐫(𝐝0 ) = 𝐟𝑖𝑛𝑡 (𝐝0 ) − 𝐟𝑎𝑝𝑝 (3)

𝛿𝐫 𝛿𝐟𝑖𝑛𝑡
𝐊|𝐝0 = 𝛿𝐝| = 𝛿𝐝
| (4)
𝐝0 𝐝0

which basically states the tangent stiffness is derivative of internal force vector with respect to the
displacement at the initial value. Solving for 𝛿𝐝, the displacement vector 𝐝 can be expressed as:
−1
𝐝 = 𝐝0 + 𝛿𝐝 = 𝐝0 − (𝐊|𝐝0 ) 𝐫(𝐝0 ) (5)

Equation (2) is repeated until residual vector becomes smaller than the predefined tolerance – in other
words, the convergence is assumed to be achieved. It is worth noting that, since equation (1) is highly
nonlinear, loads may be applied incrementally to improve the rate of convergence.

For the Newton Raphson approach, equation 5 requires the inversion of the tangent stiffness matrix to
determine the displacement increment, however, in the vicinity of a limit point (represented by buckling
of the structure) and any unstable behavior during post-buckling, the tangent stiffness matrix can
become singular. To overcome this, Collapse Advanced uses what is known as the Arc-length method to
perform nonlinear iterations at limit points and during post-buckling – see section 6.6 for details on the
Arc-length method.
For the theoretical approach used Collapse Advanced, Section 6.2 outlines the general beam work for
geometric nonlinear stiffness and local force calculations. Sections 6.3 discusses beam element
formulation and section 6.4 presents details on the various plate elements available in Collapse
Advanced.

181

SACS® Collapse Advanced Release 15.1

The treatment of element offsets and releases in a nonlinear analysis with large displacements and
rotations is discussed in sections 6.7 and 6.8 respectively.

To account for material nonlinearities which contribute to the formation of the elemental tangent
stiffness matrix and the internal load vector, Collapse Advanced utilizes the J2 yield surface and
backward Euler method to determine the plastic response. The details for this are presented in section
6.5.

6.2 Finite Element Formulation: Basics

The analysis approach in Collapse Advanced is based upon the corotational method [1, 2] which allows
the element deformation to be expressed in two parts: The first part considers rigid body motion in
global coordinates system associated with very large deformation. The second part considers relatively
small deformations in element local coordinates system. This approach enables two separate
calculations of the local stiffness matrix and geometric nonlinearities.
The general notation used in the finite element formulation for Collapse Advanced is defined below:

𝑥, 𝑦, 𝑧 coordinates of a given point within element.

𝑢, 𝑣, 𝑤 displacement along local axes at given point (𝑥, 𝑦, 𝑧) within the element
𝜃𝑥 , 𝜃𝑦 , 𝜃𝑧 rotation about local axes at given point (𝑥, 𝑦, 𝑧) within the element
𝑢𝑗 , 𝑣𝑗 , 𝑤𝑗 nodal displacement along local axes at jth joint of the element
𝜃𝑥𝑗 , 𝜃𝑦𝑗 , 𝜃𝑧𝑗 nodal rotation about local axes at jth joint of the element
𝐗 (𝑗) jth joint coordinates vector in the undeformed configuration
𝑋𝑗 , 𝑌𝑗 plate element jth joint coordinates in the undeformed plate local axis system
𝐱 (𝑗) jth joint coordinates vector in the deformed (current/iterative) configuration
𝑥𝑖𝑗 , 𝑦𝑖𝑗 plate relative coordinates 𝑥𝑖𝑗 = 𝑋𝑖 − 𝑋𝑗 and 𝑦𝑖𝑗 = 𝑌𝑖 − 𝑌𝑗
𝜉, 𝜂 isoparametric coordinates
𝑁𝐿 , 𝑁𝑄 linear and quadratic shape functions for plates as function (𝜉, 𝜂)
𝑁𝑗𝑛𝑡 number of a joint in an element
𝑀𝑝 , 𝑁𝑝 number of joint integration points in the element
𝑉, 𝐴, 𝐿, 𝑡 element volume, section area, element length, thickness
𝐸, 𝐺, 𝜈 elastic modulus (Young’s modulus), shear modulus, and Poisson’s ratio
𝛆, 𝛔 strain and stress vectors
𝜀, 𝜀𝑥 , 𝜀𝑦 normal strains
𝛾, 𝛾𝑥𝑦 , 𝛾𝑥𝑧 , 𝛾𝑦𝑧 shear strains
𝜎, 𝜎𝑥 , 𝜎𝑦 normal stresses
𝜏, 𝜏𝑥𝑦 , 𝜏𝑥𝑧 , 𝜏𝑦𝑧 shear stresses
𝛿 variational operator
𝑇 matrix transpose operator
𝐝, 𝐟, 𝐊 displacement, force vector, stiffness matrix
𝐩, 𝛉 translation and oration degrees of freedom (DOF)
𝐓, 𝐑 transformation matrix and rotation matrix
𝐈 identity matrix
𝑙𝑜𝑐, 𝑔𝑙𝑏 short form for local coordinate system and global coordinate system
𝑔𝑒𝑜 short form for geometric
182

SACS® Collapse Advanced Release 15.1

𝑡𝑜𝑟 short form for torsional

The basics of corotational method are as follow. At a given iteration for a sub-segment, the local and
global incremental change in displacement vectors, 𝛿𝐝𝑙𝑜𝑐 and 𝛿𝐝𝑔𝑙𝑏 respectively, have the following
relationship:

𝛿𝐝𝑙𝑜𝑐 = 𝐓(𝐝𝑔𝑙𝑏 )𝛿𝐝𝑔𝑙𝑏 (6)

where 𝐓 is the corotational transformation matrix. Contrary to standard small displacement and
rotation finite element analysis, the transformation matrix 𝐓 is not a constant matrix and it is defined as
a function of element nodal displacement and rotations and it is updated throughout the nonlinear
analysis. Following the virtual work principle and using local and global coordinate systems:

𝛿𝐝𝑇𝑙𝑜𝑐 𝐟𝑙𝑜𝑐 = 𝛿𝐝𝑔𝑙𝑏

𝑇
𝐟𝑔𝑙𝑏 (7)

in which the superscript T denotes the matrix/vector transpose, and 𝐟𝑙𝑜𝑐 and 𝐟𝑔𝑙𝑏 are the elemental
load vectors in local and global systems, respectively. Substituting equation (6) in (7):

𝐟𝑔𝑙𝑏 = 𝐓 𝑇 𝐟𝑙𝑜𝑐 (8)

Iterative solvers require a tangent stiffness which is the derivative of the global force vector 𝐟𝑔𝑙𝑏 with
respect to the displacement vector in the global axis system 𝐝𝑔𝑙𝑏 . Differentiating equation (8) with
respect to 𝐝𝑔𝑙𝑏 and using equations (6) and (8), the elemental tangent stiffness matrix in global
coordinate system is given by:
𝛿𝐟𝑔𝑙𝑏
𝐊𝑔𝑙𝑏 = 𝛿𝐝 = 𝐓 𝑇 𝐊 𝑙𝑜𝑐 𝐓 + 𝐊 𝒈𝒆𝒐 (9)
𝑔𝑙𝑏

in which, 𝐊 𝑙𝑜𝑐 is the element local stiffness matrix which is given by:

𝛿𝐟
𝐊 𝑙𝑜𝑐 = 𝛿𝐝𝑙𝑜𝑐 (10)
𝑙𝑜𝑐

and 𝐊 𝒈𝒆𝒐 is the geometric stiffness defined as:

𝑇
𝛿𝐓
𝐊 𝒈𝒆𝒐 = ൬𝛿𝐝 ൰ 𝐟𝑙𝑜𝑐 (11)
𝑔𝑙𝑜𝑏

Once the relationship between local and global displacement vector has been established, the elemental
stiffness matrix and load vector can be calculated using a finite element shape function, second-order
strains and elastoplastic material properties.

6.3 Beam Element

This section discusses the finite element formulation for members. Each sub-segment of a given beam
members is modelled using a single finite element where geometrical nonlinearities are modeled using
the corotational approach [2], and the elemental stiffness matrix and force vector is calculated using
second order strains, elastoplastic material properties in conjunction with numerical integration.

183

SACS® Collapse Advanced Release 15.1

Section 6.3.1 discusses the formation of the transformation matrix and its derivative with respect to the
global displacement matrix. Section 6.3.2 defines the finite element formulation for the elemental
stiffness matrix and the force vector. In addition, section 6.3.3 illustrates the numerical integration for
beam elements including integration points associated with different beam cross section types.
Elastoplastic calculations for beam elements are presented in section 6.5.1.
6.3.1 Beam Element - Corotational Approach for Geometric Nonlinearity
6.3.1.1 Basics

The geometric nonlinearity in Collapse Advanced is based on the corotational method [2]. This method
requires three local coordinates systems for a given beam element as shown in Figure 193, these being:
rotation matrix 𝐄 defining the overall element position in space and rotation matrices 𝐔(1) and 𝐔(2) for
the element end joints. Matrices 𝐄, 𝐔(1) and 𝐔 (2) define the position of the element and joints with
respect to the global coordinate system – i.e. their columns are direction of local coordinate systems.
Figure 193 illustrates the beam element local systems.

𝑧
𝑦

𝑥
𝑼(2)

𝑬
𝐄 𝐔(1)

𝐔(1) undeformed deformed

Figure 193: Collapse Advanced element local coordinate systems

Once the local coordinates systems have been defined, the transformation matrix 𝐓 and local
displacement vector can be calculated using rotation matrices 𝐄, 𝐔(1) , 𝐔(2) and the global displacement
vector. In addition, the transformation matrix 𝐓 is differentiated to form the geometric stiffness matrix.

6.3.1.2 Unit Quaternions and Rotation Matrix

Unit quaternions and their relationship with the rotation angles and the rotation matrix must be defined
before forumulating the corotational transformation matrix. Unit quaternions provide a convenient and
numerically stable way to perform various matrix operations associated with the corotational approach.

Unit quaternions are defined in the following form:

𝑇
{𝑞0 } = [𝑞0 , 𝑞1 , 𝑞2 , 𝑞3 ] (12)
𝐪

where 𝑞02 + 𝑞12 + 𝑞22 + 𝑞32 = 1. The quaternions product is defined as:

184

SACS® Collapse Advanced Release 15.1

 𝑎0 𝑏0 − 𝐚𝑇 𝐛
{𝑎0 } × {𝑏0 } = { } (13)
𝐚 𝐛 𝑎0 𝐛 + 𝑏0 𝐚 + 𝐚 × 𝐛

Rotation angle vectors 𝛉 can be expressed as unit quaternions as per section 16.9 of reference [2]:
|𝛉| |𝛉| 𝛉
𝑞0 = 𝑐𝑜𝑠 ( 2 ) , 𝐪 = 𝑠𝑖𝑛 ( 2 ) |𝛉| (14)

𝑞
and if 𝛉 = 𝟎, { 0 } = [1,0,0,0]𝑇 . The rotation matrix 𝐑 based on unit quaternions as per section 16.9 of
𝐪
reference [2] can be defined as:

𝐑 = (𝑞02 − 𝐪𝑇 𝐪)𝐈 + 2𝐪𝐪𝑇 + 2𝑞0 𝐒(𝐪) (15)

In which 𝐒 is the spin operator matrix as defined below:

0 −v3 −v2
𝐒(𝐯) = [v3 0 −v1 ] (16)
v2 v1 0
The spin operator matrix 𝐒 has following properties with respect to the vector cross product

𝐚 × 𝐛 = 𝐒(𝐚)𝐛 = −𝐒(𝐛)𝐚 (17)

𝛿(𝐚 × 𝐛) = 𝛿 (𝐒(𝐚)𝐛) = −𝐒(𝐛)𝛅𝐚 + 𝐒(𝐚)𝛅𝐛 (18)

Finally, any rotation matrix can be converted to the corresponding quaternions using the rotation matrix
components 𝑅𝑖𝑗 as per section 16.9 of reference [2].

𝐴𝑠𝑠𝑢𝑚𝑒 𝑎 = max(𝑡𝑟𝑎𝑐𝑒(𝐑), 𝑅11 , 𝑅22 , 𝑅33 )

1 𝑅 −𝑅 𝑅 −𝑅 𝑅 −𝑅
𝑖𝑓 𝑎 = 𝑡𝑟𝑎𝑐𝑒(𝐑) ⟹ 𝑞0 = 2 (1 + 𝑎)2 , 𝑞1 = 324𝑞 23 , 𝑞2 = 134𝑞 31 , 𝑞2 = 214𝑞 12
0 0 0

𝑅32−𝑅23 1 1 𝑅21 +𝑅12 𝑅13 +𝑅31

𝑖𝑓 𝑎 = 𝑅11 ⟹ 𝑞0 = 4𝑞1
, 𝑞1 = √2 𝑎 + 4 (1 − 𝑡𝑟𝑎𝑐𝑒(𝐑)) , 𝑞2 = 4𝑞1
, 𝑞3 = 4𝑞0
(19)
𝑅13−𝑅31 𝑅21 +𝑅12 1 1 𝑅13 +𝑅31
𝑖𝑓 𝑎 = 𝑅22 ⟹ 𝑞0 = , 𝑞1 = , 𝑞2 = √ 𝑎 + (1 − 𝑡𝑟𝑎𝑐𝑒(𝐑)) , 𝑞3 =
4𝑞2 4𝑞2 2 4 4𝑞2

𝑅21 −𝑅12 𝑅13 +𝑅31 𝑅32 +𝑅23 1 1

𝑖𝑓 𝑎 = 𝑅33 ⟹ 𝑞0 = 4𝑞3
, 𝑞1 = 4𝑞3
, 𝑞2 = 4𝑞3
, 𝑞3 = √2 𝑎 + 4 (1 − 𝑡𝑟𝑎𝑐𝑒(𝐑))

6.3.1.3 Local Displacement, Element and Joints Rotation Matrices

First step is to determine the local coordinate systems associated with two element joints (i.e. 𝐔 (1) and
𝐔(2) ) based on previous known (fixed) displacement vector 𝐝𝑔𝑙𝑏 0 and the current (iterative)
displacement vector 𝐝𝑔𝑙𝑏 𝑖 – in which 𝑖 denotes ith iteration. The following form for global displacement
vector is assumed:

185

SACS® Collapse Advanced Release 15.1

 (1)
𝐩𝑔𝑙𝑏
(1)
𝛉𝑔𝑙𝑏
𝐝𝑔𝑙𝑏 = (2)
(20)
𝐩𝑔𝑙𝑏
(2)
{ 𝛉𝑔𝑙𝑏 }
(𝑗) (𝑗)
where 𝐩𝑔𝑙𝑏 and 𝛉𝑔𝑙𝑏 are the translational and rotational degrees of freedom, respectively. The
incremental change in rotation given by:
(𝑗) (𝑗) (𝑗)
∆𝛉𝑔𝑙𝑏 = 𝛉𝑔𝑙𝑏 − 𝛉𝑔𝑙𝑏 (21)
𝑖 0

Using equation (14), unit quaternions increment associated with the above rotation increment is given
by:
(𝑗) (𝑗) (𝑗)
(𝑗) |∆𝛉𝑔𝑙𝑏| |∆𝛉𝑔𝑙𝑏| ∆𝛉𝑔𝑙𝑏
∆𝑞0 = 𝑐𝑜𝑠 ( ) , ∆𝐪(𝑗) = 𝑠𝑖𝑛 ( ) (𝑗) (22)
2 2 |∆𝛉𝑔𝑙𝑏|

Current quaternions for element joints can be defined using quaternions product of the above
increment to their initial values – for more details see section 16.9 in reference [2].
(𝑗) (𝑗) (𝑗)
𝑞0 ∆𝑞 𝑞
{ } = { 0 }×{ 0 } (23)
(𝑗) (𝑗)
𝐪 𝑖 ∆𝐪 𝐪(𝑗) 0

and substituting into equation (15) yields the current joints rotation matrices.

Element local x-axis, 𝐞𝟏 or first column of matrix 𝐄 is a unit vector connecting element joints, 𝐞𝟏 may be
defined using joint coordinates in the undeformed configuration (𝐗 (𝑗)) and the current translation DOFs
(𝑗)
(𝐩𝑔𝑙𝑏 )
𝑖
(2) (1)
𝐗 (2) +𝐩𝑔𝑙𝑏 −(𝐗 (1)+𝐩𝑔𝑙𝑏 )
𝑖 𝑖
𝐞1 = (2) (1)
(24)
|𝐗 (2)+𝐩𝑔𝑙𝑏 −൬𝐗 (1)+𝐩𝑔𝑙𝑏 ൰|
𝑖 𝑖

To define 𝐞2 and 𝐞3 , it is assumed that the element local system 𝐄 is located mid-way between element
joints [2]. The rotation matrix at the element mid-way, 𝐑 𝑚 , can be based on the transition from 𝐔(1) at
1st joint to 𝐔(2) at 2nd joint – see section 17.1.4 in [2] for details. Once the rotation matrix 𝐑 𝑚 is
calculated, 𝐞𝟐 and 𝐞𝟑 may be calculated by rotation 𝐑 𝑚 such that its local axis coincides with 𝐞1 .
Therefore as per reference [2]:

2 1𝐫𝑇𝐞 3 1 𝐫𝑇𝐞
𝐞2 = 𝐫2 − 1+𝐫 𝑇 𝐞 (𝐫1 + 𝐞1 ) , 𝐞3 = 𝐫3 − 1+𝐫 𝑇 𝐞 (𝐫1 + 𝐞1 ) (25)
1 1 1 1

in which 𝐑 𝑚 = [𝐫1 𝐫2 𝐫3 ].

Once the element and joint local system have been determined, the local displacement vector can be
calculated. Following the corotational method, the local displacement vector 𝐝𝑙𝑜𝑐 is given by:

𝐝𝑙𝑜𝑐 = [∆𝑙 𝜃𝑥1 𝜃𝑦1 𝜃𝑧1 𝜃𝑥2 𝜃𝑦2 𝜃𝑧2 ]𝑇 (26)

186

SACS® Collapse Advanced Release 15.1

where ∆𝑙 is the element elongation which can determined as follows:
(2) (1)
∆𝑙 = 𝑙𝑖 − 𝑙0 , 𝑙0 = |𝐗 (2) −𝐗 (1) |, 𝑙𝑖 = |𝐗 (2) + 𝐩𝑔𝑙𝑏 − (𝐗 (1) + 𝐩𝑔𝑙𝑏 )| (27)
𝑖 𝑖

Where the local rotations are given by angle between element coordinate system 𝐄 and joint local
systems 𝐔(j) (see reference [2] section 16.14):
(𝑗) (𝑗)
𝜃𝑥𝑗 𝐞𝑇3 𝐮2 − 𝐞𝑇2 𝐮3
(𝑗) 1
𝑠𝑖𝑛 (𝛉𝑙𝑜𝑐 𝑖 ) = 𝑠𝑖𝑛 ({𝜃𝑦𝑗 } ) = {𝐞1𝑇 𝐮(𝑗) 𝑇 (𝑗) }
3 − 𝐞3 𝐮1
(28)
2
𝜃𝑧𝑗 𝑙𝑜𝑐 (𝑗) (𝑗)
𝑖 𝐞𝑇2 𝐮1 − 𝐞1𝑇 𝐮2

It is worth noting that other local DOFs are zero, since the element deformation (both rigid body motion
and local deformation) is fully described by 𝐝𝑙𝑜𝑐 , 𝐄, 𝐔(1) and 𝐔(2) .

6.3.1.4 Transformation Matrix

Recalling equation (6) for transformation matrix, the corotational transformation matrix 𝐓 can be
calculated by differentiating the local DOFs with respect to the element global displacement vector.
Therefore:

𝐭 𝑇∆𝑙
𝐭 𝑇𝜃𝑥1
𝑙𝑜𝑐

𝐭 𝑇𝜃𝑦1
𝑙𝑜𝑐
𝑇
𝐓 = 𝐭 𝜃𝑧1 𝑙𝑜𝑐 (29)
𝐭 𝑇𝜃𝑥2
𝑙𝑜𝑐

𝐭 𝑇𝜃𝑦2
𝑙𝑜𝑐
𝑇
[𝐭 𝜃𝑧2 𝑙𝑜𝑐 ]

where 𝐭 are 12 × 1 vectors.

The variation of the element elongation with respect to global displacement vector is given by:
𝛿𝑙𝑖
𝛿(∆𝑙 ) = 𝛿 (𝑙𝑖 − 𝑙0 ) = 𝛿𝐝𝑔𝑙𝑏 = 𝐭 𝑇∆𝑙 𝛿𝐝𝑔𝑙𝑏 (30)
𝛿𝐝𝑔𝑙𝑏

in which 𝐭 𝑇∆𝑙 is a row associated with ∆𝑙 in the transformation matrix 𝐓. To calculate 𝐭 𝑇∆𝑙 , differentiate 𝑙𝑖
(𝑗)
in equation (27) with respect to 𝐝𝑔𝑙𝑏 . Recalling 𝑙𝑖 is only a function of translation DOFs 𝐩𝑔𝑙𝑏 [2]:

𝐭 𝑇∆𝑙 = [−𝐞1𝑇 𝟎 𝐞1𝑇 𝟎] (31)

Similarly, rows associated with rotational of DOFs are determined by differentiating equation (28) with
respect to 𝐝𝑔𝑙𝑏 . Therefore:

187

SACS® Collapse Advanced Release 15.1

 (𝑗) (𝑗)
2cos (𝜃𝑥𝑗 )𝛿𝜃𝑥𝑗 = 𝛿 (𝐞𝑇3 𝐮2 − 𝐞𝑇2 𝐮3 )
(𝑗) (𝑗)
2cos (𝜃𝑦𝑗 )𝛿𝜃𝑦𝑗 = 𝛿 (𝐞1𝑇 𝐮3 − 𝐞𝑇3 𝐮1 ) (32)
(𝑗) (𝑗)
2cos (𝜃𝑦𝑗 )𝛿𝜃𝑧𝑗 = 𝛿 (𝐞𝑇2 𝐮1 − 𝐞1𝑇 𝐮2 )

Which can be written in a compact form as:

𝐠𝑇
𝛿𝜃𝑙𝑜𝑐 = 𝐭 𝑇𝜃 𝛿𝐝𝑔𝑙𝑏 = 2cos (𝜃
𝜃
𝛿𝐝𝑔𝑙𝑏 (33)
𝑙𝑜𝑐 )

where 𝐠 𝜃 defines the right-hand side of equation (32).

The derivatives of the local coordinates systems (𝐄, 𝐔(1) , 𝐔(2) ) can be obtained with respect to 𝐝𝑔𝑙𝑏 .
The derivation of 𝐞1 , the first column in 𝐄, which is only a function of translation DOFs (see equation
(31))is expressed as (for details see 17.1.2 in [2]):
𝛿𝐞1 𝛿𝐞1 𝛿𝐞1 𝛿𝐞1
(1) = −𝐀 , (1) =𝟎, (2) =𝐀, (𝟐) =𝟎 (34)
𝛿𝐩𝑔𝑙𝑏 𝛿𝛉𝑔𝑙𝑏 𝛿𝐩𝑔𝑙𝑏 𝛿𝛉𝑔𝑙𝑏

where
1
𝐀 = 𝑙 (𝐈 − 𝐞1 𝐞1𝑇 ) (35)
𝑖

The derivatives of 𝐞2 and 𝐞3 are obtained by differentiating equation (25) with respect to 𝐝𝑔𝑙𝑏 :

𝛿𝐞𝟐 𝛿𝐞𝟐 𝛿𝐞𝟐 𝛿𝐞𝟐

(1) = 𝐋1 (𝐫2 )𝑇 , (1) = 𝐋2 (𝐫2 )𝑇 , (2) = −𝐋1 (𝐫2 )𝑇 , (𝟐) = 𝐋2 (𝐫2 )𝑇
𝛿𝐩𝑔𝑙𝑏 𝛿𝛉𝑔𝑙𝑏 𝛿𝐩𝑔𝑙𝑏 𝛿𝛉𝑔𝑙𝑏
(36)
𝛿𝐞3 𝑇 𝛿𝐞3 𝑇 𝛿𝐞3 𝑇 𝛿𝐞3 𝑇
(1) = 𝐋1 (𝐫3 ) , (1) = 𝐋2 (𝐫3 ) , (2) = −𝐋1 (𝐫3 ) , (𝟐) = 𝐋2 (𝐫2 )
𝛿𝐩𝑔𝑙𝑏 𝛿𝛉𝑔𝑙𝑏 𝛿𝐩𝑔𝑙𝑏 𝛿𝛉𝑔𝑙𝑏

This can be represented in a matrix form as:

𝛿𝐞2
= 𝐋(𝐫2 )𝑇
𝛿𝐝𝑔𝑙𝑏
(37)
𝛿𝐞3
= 𝐋(𝐫3 )𝑇
𝛿𝐝𝑔𝑙𝑏

where

𝐋(𝐫)𝑇 = [𝐋1 (𝐫)𝑇 𝐋2 (𝐫)𝑇 −𝐋1 (𝐫)𝑇 𝐋2 (𝐫)𝑇 ]

𝑇 𝑇
𝐫 𝐞1 1
𝐋𝟏 (𝐫) = 2
𝐀 + 2 𝐀𝐫(𝐞1 + 𝐫1 ) (38)
𝐒(𝐫) 𝐫 𝑇 𝐞1 1 𝑇
𝐋𝟐 (𝐫) = 𝟐
− 4
𝐒(𝐫1 ) − 4 𝐒(𝐫)𝐞1 (𝐞1 + 𝐫1 )

The denominator in equation (25) is approximated as 1 + 𝐫1𝑇 𝐞1 ≈ 2.

Following section 17.1.2 of reference [2] the derivatives of 𝐔(1) , 𝐔(2) with respect to 𝐝𝑔𝑙𝑏 (considering
they are only functions of rotational DOF) are expressed as:

188

SACS® Collapse Advanced Release 15.1

 (1) (2)
𝛿𝐮𝑘 (1) 𝛿𝐮𝑘 (2)
(1) = 𝐒(𝐮𝑘 )𝑇 , (2) = 𝐒(𝐮𝑘 )𝑇 (39)
𝛿𝛉𝑔𝑙𝑏 𝛿𝛉𝑔𝑙𝑏

and other derivatives are zero. Substituting the above derivations into equation (32), vectors 𝐠 𝜃 can be
defined as:
𝟎
(1) (1)
𝐒 (𝐮2 ) 𝐞3 − 𝐒 (𝐮3 ) 𝐞2 (1) (1)
𝐠 𝜃𝑥1 = + 𝐋(𝐫3 )𝐮2 − 𝐋(𝐫2 )𝐮3
𝑙𝑜𝑐
𝟎
[ 𝟎 ]
𝟎 −𝐀
(1) (1)
𝐒 (𝐮3 ) 𝐞1 − 𝐒 (𝐮1 ) 𝐞3 (1) 𝟎 (1)
𝐠 𝜃𝑦1 = − 𝐋(𝐫3 )𝐮1 + [ ] 𝐮3
𝑙𝑜𝑐
𝟎 𝐀
[ 𝟎 ] 𝟎
𝟎 −𝐀
(1) (1)
𝐒 (𝐮1 ) 𝐞2 − 𝐒 (𝐮2 ) 𝐞1 (1) 𝟎 (1)
𝐠 𝜃𝑧1 = + 𝐋(𝐫2 )𝐮1 − [ ] 𝐮2
𝑙𝑜𝑐
𝟎 𝐀
[ 𝟎 ] 𝟎
(40)
𝟎
𝟎 (2) (2)
𝐠 𝜃𝑥2 = 𝟎 + 𝐋(𝐫3 )𝐮2 − 𝐋(𝐫2 )𝐮3
𝑙𝑜𝑐
(2) (2)
[𝐒 (𝐮2 ) 𝐞3 − 𝐒 (𝐮3 ) 𝐞2 ]
𝟎 −𝐀
𝟎 (2) 𝟎 (2)
𝐠 𝜃𝑦2 = 𝟎 − 𝐋(𝐫3 )𝐮1 + [ ] 𝐮3
𝑙𝑜𝑐
(2) (2)
𝐀
[𝐒 (𝐮3 ) 𝐞1 − 𝐒 (𝐮1 ) 𝐞3 ] 𝟎
𝟎 −𝐀
𝟎 (2) 𝟎 (2)
𝐠 𝜃𝑧2 = 𝟎 + 𝐋(𝐫2 )𝐮1 − [ ] 𝐮2
𝑙𝑜𝑐
(2) (2)
𝐀
[𝐒 (𝐮1 ) 𝐞2 − 𝐒 (𝐮2 ) 𝐞1 ] 𝟎

6.3.1.5 Geometric Stiffness

Recalling equation (11) and using the above definition for the transformation matrix, the beam element
geometric stiffness is expressed as:

𝛿𝐭 ∆𝑙
𝐊 𝒈𝒆𝒐 = 𝛿𝐝 𝑓∆𝑙 + ∑𝜃𝑙𝑜𝑐 𝑓𝜃𝑙𝑜𝑐 (41)
𝑔𝑙𝑜𝑏

in which 𝑓 denotes the components of the local force vector. Using equations (28) for rotational DOFs,
the above equation may be rewritten as:

𝛿𝐭 ∆𝑙 𝛿𝐠 𝜃 𝑓𝜃𝑙𝑜𝑐
𝐊 𝒈𝒆𝒐 = 𝑓 + ∑𝜃𝑙𝑜𝑐 𝐭 𝜃𝑙𝑜𝑐 𝑡𝑎𝑛(𝜃𝑙𝑜𝑐 )𝐭 𝑇𝜃𝑙𝑜𝑐 𝑓𝜃𝑙𝑜𝑐 + ∑𝜃𝑙𝑜𝑐 𝑙𝑜𝑐
(42)
𝛿𝐝𝑔𝑙𝑜𝑏 ∆𝑙 𝛿𝐝𝑔𝑙𝑜𝑏 2𝑐𝑜𝑠(𝜃𝑙𝑜𝑐 )

𝛿𝐭 ∆𝑙
The first term in the above equation, 𝛿𝐝 , can be determined using equations (31), (34), and (35) as:
𝑔𝑙𝑜𝑏

189

SACS® Collapse Advanced Release 15.1

 𝐀 𝟎 −𝐀 𝟎
𝛿𝐭 ∆𝑙 𝟎 𝟎 𝟎 𝟎
=[ ] (43)
𝛿𝐝𝑔𝑙𝑜𝑏 −𝐀 𝟎 𝐀 𝟎
𝟎 𝟎 𝟎 𝟎
The determination of the second term in equation (42) is straightforward since rows of transformation
matrix, 𝐭 𝜃𝑙𝑜𝑐 are calculated as per equations (33) to (40).

𝛿𝐠 𝜃
𝑙𝑜𝑐
Calculation of the higher order corotational term requires the determination of 𝛿𝐝𝑔𝑙𝑜𝑏
which are
12 × 12 matrices. The derivatives of the first terms in equation (4) (section 17.1.3 in [2]):
(1) (1)
𝛿(𝐒(𝐮2 )𝐞3 −𝐒(𝐮3 )𝐞2 ) (1) (1)
𝛿𝐝𝑔𝑙𝑜𝑏
= [𝟎 𝐒(𝐞3 )𝐒 (𝐮2 ) − 𝐒(𝐞2 )𝐒 (𝐮3 ) 𝟎 𝟎] ⋯
(44)
(1) (1)
+ 𝐒 (𝐮2 ) 𝐋(𝐫3 )𝑇 − 𝐒 (𝐮3 ) 𝐋(𝐫2 )𝑇
(1) (1)
𝛿(𝐒(𝐮3 )𝐞1 −𝐒(𝐮1 )𝐞3 ) (1) (1)
= [𝟎 𝐒(𝐞1 )𝐒 (𝐮3 ) − 𝐒(𝐞3 )𝐒 (𝐮1 ) 𝟎 𝟎] ⋯
𝛿𝐝𝑔𝑙𝑜𝑏
(45)
(1) (1) (1)
+ [−𝐒 (𝐮3 ) 𝐀 𝟎 𝐒 (𝐮3 ) 𝐀 𝟎] − 𝐒 (𝐮1 ) 𝐋(𝐫3 )𝑇
(1) (1)
𝛿(𝐒(𝐮1 )𝐞2 −𝐒(𝐮2 )𝐞1 ) (1) (1)
𝛿𝐝𝑔𝑙𝑜𝑏
= [𝟎 𝐒(𝐞2 )𝐒 (𝐮1 ) − 𝐒(𝐞1 )𝐒 (𝐮2 ) 𝟎 𝟎] ⋯
(46)
(1) (1) (1)
+ 𝐒 (𝐮1 ) 𝐋(𝐫2 )𝑇 − [−𝐒 (𝐮2 ) 𝐀 𝟎 𝐒 (𝐮2 ) 𝐀 𝟎]
(2) (2)
𝛿(𝐒(𝐮2 )𝐞3 −𝐒(𝐮3 )𝐞2 ) (2) (2)
= [𝟎 𝟎 𝟎 𝐒(𝐞3 )𝐒 (𝐮2 ) − 𝐒(𝐞2 )𝐒 (𝐮3 )] ⋯
𝛿𝐝𝑔𝑙𝑜𝑏
(47)
(2) (2)
+ 𝐒 (𝐮2 ) 𝐋(𝐫3 )𝑇 − 𝐒 (𝐮3 ) 𝐋(𝐫2 )𝑇
(2) (2)
𝛿(𝐒(𝐮3 )𝐞1 −𝐒(𝐮1 )𝐞3 ) (2) (2)
= [𝟎 𝟎 𝟎 𝐒(𝐞1 )𝐒 (𝐮3 ) − 𝐒(𝐞3 )𝐒 (𝐮1 )] ⋯
𝛿𝐝𝑔𝑙𝑜𝑏
(48)
(2) (2) (2)
+ [−𝐒 (𝐮3 ) 𝐀 𝟎 𝐒 (𝐮3 ) 𝐀 𝟎] − 𝐒 (𝐮1 ) 𝐋(𝐫3 )𝑇
(2) (2)
𝛿(𝐒(𝐮1 )𝐞2 −𝐒(𝐮2 )𝐞1 ) (2) (2)
𝛿𝐝𝑔𝑙𝑜𝑏
= [𝟎 𝟎 𝟎 𝐒(𝐞2 )𝐒 (𝐮1 ) − 𝐒(𝐞1 )𝐒 (𝐮2 )] ⋯
(49)
(2) (2) (2)
+ 𝐒 (𝐮1 ) 𝐋(𝐫2 )𝑇 − [−𝐒 (𝐮2 ) 𝐀 𝟎 𝐒 (𝐮2 ) 𝐀 𝟎]
(𝑗)
𝛿(𝐋(𝐫)𝐮 )
Defining 𝛿𝐝𝑔𝑙𝑜𝑏
:

(1)
𝛿𝐋1 (𝐫𝑚 , 𝐮𝑛 )
𝐋1 (𝐫𝑚 )
(1)
𝛿(𝐋(𝐫𝑚 )𝐮𝑛 )
(1) 𝛿𝐋2 (𝐫𝑚 , 𝐮𝑛 ) 𝐋 (𝐫 ) (1)
= − 2 𝑚 [𝟎 𝐒 (𝐮𝑛 ) 𝟎 𝟎] (50)
𝛿𝐝𝑔𝑙𝑜𝑏 (1)
−𝛿𝐋1 (𝐫𝑚 , 𝐮𝑛 ) −𝐋 (
1 𝑚𝐫 )
[ 𝐋 2 (𝐫𝑚 ) ]
(1)
[ 𝛿𝐋 2 (𝐫 ,
𝑚 𝑛𝐮 ) ]

190

SACS® Collapse Advanced Release 15.1

 (2)
𝛿𝐋1 (𝐫𝑚 , 𝐮𝑛 )
𝐋1 (𝐫𝑚 )
(2)
𝛿(𝐋(𝐫𝑚 )𝐮𝑛 )
(2) 𝛿𝐋2 (𝐫𝑚 , 𝐮𝑛 ) 𝐋 (𝐫 ) (2)
= − 2 𝑚 [𝟎 𝟎 𝟎 𝐒 (𝐮𝑛 )] (51)
𝛿𝐝𝑔𝑙𝑜𝑏 (2)
−𝛿𝐋1 (𝐫𝑚 , 𝐮𝑛 ) −𝐋 1 (𝐫𝑚 )
(2)
[ 𝐋2 (𝐫𝑚 ) ]
[ 𝛿𝐋 2 (𝐫 ,
𝑚 𝑛𝐮 ) ]

in which 𝑚 = 1,2 , 𝑛 = 1,2,3 and 𝛿𝐋1 and 𝛿𝐋2 are as follows (section 17.1.3 in reference [2]):
1 1
𝛿𝐋1 (𝐫𝑚 , 𝐮) = − 𝐀(𝐮𝐞1𝑇 )𝐒(𝐫𝑚 )[𝟎 𝐈 𝟎 𝑇)
𝐈] − 2 𝐀(𝐮𝐫𝑚 𝐀[𝐈 𝟎 −𝐈 𝟎] …
4
𝑇𝐞
𝐫𝑚
+ 1
(𝐀(𝐮𝐞1𝑇 ) + (𝐞1 𝐮𝑇 )𝐀 + (𝐞1𝑇 𝐮)𝐀)[𝐈 𝟎 −𝐈 𝟎] …
2𝑙𝑖
(𝐫1𝑇 +𝐞𝑇
1 )𝐮
+ (𝐀(𝐫𝑚 𝐞1𝑇 ) + (𝐞1 𝐫𝑚
𝑇 )𝐀
+ (𝐞1𝑇 𝐫𝑚 )𝐀)[𝐈 𝟎 −𝐈 𝟎] … (52)
2𝑙𝑖
1 1
− 4 𝐀 ((𝐫1𝑇 + 𝐞1𝑇 )𝐮) 𝐒(𝐫𝑚 )[𝐈 𝟎 𝑇 )
𝐈 𝟎] − 2 𝐀(𝐫𝑚 𝐮 𝐀[𝐈 𝟎 −𝐈 𝟎] …
1
𝑇 ) ( )[
− 𝐀(𝐫𝑚 𝐮 𝐒 𝐫1 𝐈 𝟎 𝐈 𝟎]
4

1 1
𝑇 ) ( ) ( )[
𝛿𝐋2 (𝐫𝑚 , 𝐮) = 4 𝐒(𝐮)𝐒(𝐫𝑚 )[𝟎 𝐈 𝟎 𝐈] − 8 (𝐫𝑚 𝐮 𝐒 𝐮 𝐒 𝐫1 𝟎 𝐈 𝟎 𝐈] …
1 1
+ 8 𝐒(𝐫1 )(𝐮𝐞1𝑇 )𝐒(𝐫𝑚 )[𝟎 𝐈 𝑇)
𝟎 𝐈] + 4 𝐒(𝐫1 )(𝐮𝐫𝑚 𝐀[𝐈 𝟎 −𝐈 𝟎] …
1
− ((𝐫1𝑇 + 𝐞1𝑇 )𝐮) 𝐒(𝐞1 )𝐒(𝐫𝒎 )[𝟎 𝐈 𝟎 𝐈] … (53)
8
1
+ 4 𝐒(𝐫𝑚 ) ൬((𝐫1𝑇 + 𝐞1𝑇 )𝐮) 𝐀 + (𝐞1 𝐮𝑇 )𝐀൰ [𝐈 𝟎 −𝐈 𝟎] …
1
+ 8 𝐒(𝐫𝒎 )(𝐞1 𝐮𝑇 )𝐒(𝐫1 )[𝟎 𝐈 𝟎 𝐈]

The derivatives of the last terms in equation (40) can then be written as (see section 17.1.3 in reference
[2]):
−𝐀 (1)
𝛿([
𝟎 (1)
]𝐮 ) −𝛿𝐀 (𝐮𝑛 ) −𝐀
𝐀 𝑛
𝟎 1 𝟎 𝟎 (1)
= −𝑙 − [ ] [𝟎 𝐒 (𝐮𝑛 ) 𝟎 𝟎] (54)
𝛿𝐝𝑔𝑙𝑜𝑏 𝑖 (1)
𝛿𝐀 (𝐮𝑛 ) 𝐀
𝟎
[ 𝟎 ]
−𝐀 (2)
𝟎 (2)
𝛿([ ]𝐮𝑛 ) −𝛿𝐀 (𝐮𝑛 ) −𝐀
𝐀
𝟎 1 𝟎 𝟎 (2)
= −𝑙 − [ ] [𝟎 𝟎 𝟎 𝐒 (𝐮𝑛 )] (55)
𝛿𝐝𝑔𝑙𝑜𝑏 𝑖 (2)
𝛿𝐀 (𝐮𝑛 ) 𝐀
𝟎
[ 𝟎 ]
where

𝛿𝐀(𝐮) = (𝐀(𝐮𝐞1𝑇 ) + (𝐞1𝑇 𝐮)𝐀 + (𝐞1 𝐮𝑇 )𝐀)[−𝐈 𝟎 −𝐀 𝟎] (56)

The following comments should be noted regarding the corotational geometric stiffness matrix 𝐊 𝒈𝒆𝒐 in
equation (42)

✓ The first term is defined by the p-δ effect in the beam elements

191

SACS® Collapse Advanced Release 15.1

 ✓ The second term is equivalent to p-δ for rotations.
✓ The third term provides coupling between bending moment and torsion and it enables Collapse
Advanced to determine the lateral-torsional buckling of thin-walled open sections such as wide
flange sections.
✓ The third term has minimal effects on buckling of closed sections such as tubular sections which
are widely used in the offshore industry. This term is optional term for tubular and other closed
sections and can be included in an analysis by selecting it on the CLPOP2 input line.
✓ The third term is the most computationally expensive step for calculating the geometric stiffness
matrix. In other words, if it is included in the analysis, it significantly slows down collapse
analysis of typical offshore structures.
✓ Finally, the third term results in a non-symmetric matrix which is converted to a symmetric form
by averaging off-diagonal components as recommending by [2].

6.3.2 Beam Element Local Stiffness Matrix and Load Vector

There are two main terms associated with the beam elemental stiffness matrix and the force vector in
Collapse Advanced: 1) Terms associated axial and bending strains and 2) Terms associated with torsional
strain
It is assumed axial-bending stresses follow full elastoplastic response while the torsional twist and stress
are assumed to have elastic response.

Utilizing the variational principle, the virtual work for a beam element is given by:

𝛿𝜋 = ∫𝑉 𝛿𝛆𝑇 𝛔𝑑𝑉 + ∫𝑉 𝛿𝛾𝑡𝑜𝑟 𝐺𝛾𝑡𝑜𝑟 𝑑𝑉 (57)

in which the first term is associated with beam element bending and axial virtual energy and the second
term is associated with torsion.

To determine the force vector in the local coordinate system, the local displacement and force vectors
are partitioned as:

𝐝𝑎−𝑏
𝑙𝑜𝑐
𝑎−𝑏
𝐟𝑙𝑜𝑐
𝐝𝑙𝑜𝑐 = { } and 𝐟𝑙𝑜𝑐 = { } (58)
𝐝𝑡𝑜𝑟
𝑙𝑜𝑐
𝑡𝑜𝑟
𝐟𝑙𝑜𝑐

where superscript 𝑎 − 𝑏 and 𝑡𝑜𝑟 donate axial-bending and torsional, respectively. Also, following the
corotational formulation in section 6.4.1.1, the local DOFs are defined as:

𝐝𝑎−𝑏
𝑙𝑜𝑐 = [∆𝑙 𝜃𝑦1 𝜃𝑦2 𝜃𝑧1 𝜃𝑧2 ]𝑇 , 𝐝𝑡𝑜𝑟
𝑙𝑜𝑐 = [𝜃𝑥1 𝜃𝑥2 ]𝑇 (59)

Hence , the virtual work equation (57) can be expressed as:

𝑇
𝛿𝜋 = 𝛿𝐝𝑇𝑙𝑜𝑐 𝐟𝑙𝑜𝑐 = (𝛿𝐝𝑎−𝑏 𝑎−𝑏 𝑡𝑜𝑟 𝑇 𝑡𝑜𝑟
𝑙𝑜𝑐 ) 𝐟𝑙𝑜𝑐 + (𝛿𝐝𝑙𝑜𝑐 ) 𝐟𝑙𝑜𝑐 (60)

In which
𝑇 𝑇
𝑎−𝑏 𝛿𝛆 𝑡𝑜𝑟 𝛿𝛾
𝐟𝑙𝑜𝑐 = ∫𝑉 ൬ ൰ 𝛔𝑑𝑉 , 𝐟𝑙𝑜𝑐 = ∫𝑉 ൬𝛿𝐝𝑡𝑜𝑟
𝑡𝑜𝑟 ൰ 𝐺𝛾𝑡𝑜𝑟 𝑑𝑉 (61)
𝛿𝐝𝑎−𝑏
𝑙𝑜𝑐 𝑙𝑜𝑐

Utilizing equation (10) for the local stiffness matrix, the cross coupling terms for axial-bending and
torsion are as follows:

192

SACS® Collapse Advanced Release 15.1

 𝑇 𝑇
𝛿2 𝛆 𝛿𝛆 𝛿𝛆
𝐊 𝑎−𝑏
𝑙𝑜𝑐 = ∫𝑉 ( 2 ) 𝛔𝑑𝑉 + ∫𝑉 ൬ 𝑎−𝑏 ൰ 𝐂 ൬ ൰ 𝑑𝑉 (62)
𝛿𝐝𝑎−𝑏 𝛿𝐝𝑙𝑜𝑐 𝛿𝐝𝑎−𝑏
𝑙𝑜𝑐
𝑙𝑜𝑐

𝑇
𝛿𝛾 𝛿𝛾
𝐊 𝑡𝑜𝑟 𝑡𝑜𝑟 𝑡𝑜𝑟
𝑙𝑜𝑐 = ∫𝑉 ൬𝛿𝐝𝑡𝑜𝑟 ൰ 𝐺 ൬𝛿𝐝𝑡𝑜𝑟 ൰ 𝑑𝑉 (63)
𝑙𝑜𝑐 𝑙𝑜𝑐

where 𝐂 is the constitutive matrix determined through elastoplastic calculations as discussed in section
6.5.1. Beam bending theory (i.e. Euler or Timoshenko) is used to define the elemental axial-bending
stiffness and force by integrating over the element length and beam cross-section as discussed in the
following sections:

Assuming a linear shape function for 𝜃𝑥 variation along the element length and using linear shear
theory, 𝛾𝑡𝑜𝑟 can be expressed as:
𝑑𝜃𝑥 1 1 𝜃𝑥1
𝛾𝑡𝑜𝑟 = = [−𝐿 ]{ } (64)
𝑑𝑥 𝐿 𝜃𝑥2
Where 𝐿 is element length. Substituting equation (64) in equations (61) and (63):

𝑡𝑜𝑟 𝐺𝐽 1 −1 𝜃𝑥1 𝐺𝐽 1 −1
𝐟𝑙𝑜𝑐 = [ ] { } , 𝐊 𝑡𝑜𝑟
𝑙𝑜𝑐 = 𝐿 [ ] (65)
𝐿 −1 1 𝜃𝑥1 −1 1
in which 𝐽 is section torsional constant.

6.3.2.1 Euler-Bernoulli Beam Element

The normal strain in the beam element can be expressed as:
∆𝑢 𝑑𝜃𝑦 (𝑥) 𝑑𝜃𝑧 (𝑥) 1 𝑑𝑣(𝑥) 2 1 𝑑𝑤(𝑥) 2
𝜀(𝑥, 𝑦, 𝑧) = +𝑧 −𝑦 + ( ) + ( ) (66)
𝐿 𝑑𝑥 𝑑𝑥 2 𝑑𝑥 2 𝑑𝑥

in which 𝑢 represents axial displacement, 𝑣 and 𝑤 are lateral displacements of the beam centroid along
local y and z axes respectively, and 𝜃𝑦 and 𝜃𝑧 represent the rotation of the beam centroid about the
local y and z axes respectively.

The first term in equation represents axial strain, the second and third terms represent bending strains
about the local y and z axis, and the last two terms are second-order strains associated with the axial
strain resulting from displacements 𝑣 and 𝑤.

Following the Euler-Bernoulli bending theory:

𝑑𝑤(𝑥) 𝑑𝑣(𝑥)
𝜃𝑦 (𝑥) = − 𝑑𝑥
, 𝜃𝑧 (𝑥) = 𝑑𝑥
(67)

Using a standard finite element shape functions, the beam displacement may be expressed as a function
of nodal DOFs in equation (59):
∆𝑙 1 0 0 0 0
{ 𝑣(𝑥) } = [0 0 𝑁𝜃1 (𝑥) 𝑁𝜃2 (𝑥)] 𝐝𝑎−𝑏
0 𝑙𝑜𝑐 (68)
𝑤(𝑥) 0 −𝑁𝜃1 (𝑥) −𝑁𝜃2 (𝑥) 0 0
where shape functions 𝑁𝜃1 and 𝑁𝜃2 are defined as:

193

SACS® Collapse Advanced Release 15.1

 𝑥(𝐿−𝑥)2 𝑥 2(𝐿−𝑥)
𝑁𝜃1 (𝑥) = 𝐿2
, 𝑁𝜃2 (𝑥) = − 𝐿2
(69)

Substituting in strain equation (66):

1 𝑇
𝜀(𝑥, 𝑦, 𝑧) = 𝐁𝐿𝑒𝑢𝑙𝑒𝑟 𝐝𝑎−𝑏 𝑒𝑢𝑙𝑒𝑟 𝑎−𝑏 𝑒𝑢𝑙𝑒𝑟 𝑎−𝑏
𝑙𝑜𝑐 + 2 (𝐁𝑁𝐿 𝐝𝑙𝑜𝑐 ) (𝐁𝑁𝐿 𝐝𝑙𝑜𝑐 ) (70)

in which 𝐁𝐿𝑒𝑢𝑙𝑒𝑟 and 𝐁𝑁𝐿

𝑒𝑢𝑙𝑒𝑟
are the linear and nonlinear Euler-Bernoulli strain operators respectively (i.e.
matrices) and can be represented as:
1 𝑑2𝑁𝜃1 𝑑2 𝑁𝜃1 𝑑2 𝑁𝜃1 𝑑 2 𝑁𝜃2
𝐁𝐿𝑒𝑢𝑙𝑒𝑟 (𝑥, 𝑦, 𝑧) = [ 𝑧 𝑧 −𝑦 −𝑦 ]
𝐿 𝑑𝑥 2 𝑑𝑥 2 𝑑𝑥 2 𝑑𝑥 2
𝑑𝑁𝜃1 𝑑𝑁𝜃2
0 0 0 (71)
𝑒𝑢𝑙𝑒𝑟 ( 𝑑𝑥 𝑑𝑥
𝐁𝑁𝐿 𝑥, 𝑦, 𝑧) = [ 𝑑𝑁𝜃1 𝑑𝑁𝜃2
]
0 0 0
𝑑𝑥 𝑑𝑥

Using the above strain operators, the local force vector and stiffness matrix for Euler-Bernoulli beam
element are defined as:
𝑇 𝑇
𝑎−𝑏
𝐟𝑙𝑜𝑐 = ∫𝑉 (𝐁𝐿𝑒𝑢𝑙𝑒𝑟 + (𝐁𝑁𝐿
𝑒𝑢𝑙𝑒𝑟 𝑎−𝑏 𝑒𝑢𝑙𝑒𝑟
𝐝𝑙𝑜𝑐 ) 𝐁𝑁𝐿 ) σ𝑑𝑉 (72)

𝑇 𝑇
𝐊 𝑎−𝑏 𝑒𝑢𝑙𝑒𝑟 𝑒𝑢𝑙𝑒𝑟
𝑙𝑜𝑐 = ∫𝑉 ((𝐁𝑁𝐿 ) 𝐁𝑁𝐿 ) σ𝑑𝑉 + ⋯
𝑇 (73)
𝑒𝑢𝑙𝑒𝑟 𝑎−𝑏 𝑇 𝑒𝑢𝑙𝑒𝑟 𝑒𝑢𝑙𝑒𝑟 𝑎−𝑏 𝑇 𝑒𝑢𝑙𝑒𝑟
∫𝑉 (𝐁𝐿𝑒𝑢𝑙𝑒𝑟 + (𝐁𝑁𝐿 𝐝𝑙𝑜𝑐 ) 𝐁𝑁𝐿 ) 𝐶 (𝐁𝐿𝑒𝑢𝑙𝑒𝑟 + (𝐁𝑁𝐿 𝐝𝑙𝑜𝑐 ) 𝐁𝑁𝐿 ) 𝑑𝑉

For elastic beam elements, 𝐶 = 𝐸 and 𝜎 = 𝐸𝜖. For elastoplastic elements, 𝐶 and 𝜎 are determined
through elastoplastic material iterations (see section 6.5.1).

6.3.2.2 Timoshenko Beam Element

For Timoshenko beam elements, the shear strain can be defined as:

𝛾 = 𝑔(𝑦, 𝑧)𝛾̅𝑥𝑦 + ℎ(𝑦, 𝑧)𝛾̅𝑥𝑧 (74)

in which 𝑔 and ℎ are the shear strain distribution functions over beam cross-section, 𝛾̅𝑥𝑦 and 𝛾̅𝑥𝑧
represent the shear strain at the element centroid. To prevent shear-locking associated with
Timoshenko beam elements, constant shear strain is assumed at the centroid of a cross section which
can be represented as a function of the elements DOFs as:
𝜃𝑧1 +𝜃𝑧2 𝜃𝑦1 +𝜃𝑦2
𝛾̅𝑥𝑦 = − 2
, 𝛾̅𝑥𝑧 = 2
(75)

The shear strain distribution functions for thin-walled sections are defined as:
𝐴𝑠𝑦 (𝐼𝑦 𝑄𝑧 (𝑦,𝑧)−𝐼𝑦𝑧𝑄𝑦 (𝑦,𝑧))
𝑔(𝑦, 𝑧) = 2 )
𝑡(𝑦,𝑧)(𝐼𝑦 𝐼𝑧 −𝐼𝑦𝑧
𝐴𝑠𝑧 (𝐼𝑧 𝑄𝑦 (𝑦,𝑧)−𝐼𝑦𝑧𝑄𝑧 (𝑦,𝑧))
(76)
ℎ(𝑦, 𝑧) = 2 )
𝑡(𝑦,𝑧)(𝐼𝑦 𝐼𝑧−𝐼𝑦𝑧

In which

𝐼𝑦 , 𝐼𝑧 , 𝐼𝑦𝑧 represent the second moments of area for an element cross-section

194

SACS® Collapse Advanced Release 15.1

𝑄𝑦 , 𝑄𝑧 are the first moments of area calculated at a given point (𝑦, 𝑧) within the element cross-
section
𝑡 is the thickness at a given point (𝑦, 𝑧) within the element cross-section
𝐴𝑠𝑦 , 𝐴𝑠𝑧 are the cross-section shear areas

Where the cross-section shear areas, 𝐴𝑠𝑦 and 𝐴𝑠𝑧 can be represented as:
2 ) 2
(𝐼𝑦 𝐼𝑧−𝐼𝑦𝑧
𝐴𝑠𝑦 = ∫𝐴 𝑔2 𝑑𝐴 = 𝐼𝑦 𝑄𝑧 −𝐼𝑦𝑧 𝑄𝑦 2
∫𝐴 ( 𝑡
) 𝑑𝐴
2 ) 2 (77)
2 (𝐼𝑦 𝐼𝑧−𝐼𝑦𝑧
𝐴𝑠𝑧 = ∫𝐴 ℎ 𝑑𝐴 = 𝐼𝑧 𝑄𝑦 −𝐼𝑦𝑧 𝑄𝑧 2
∫𝐴 ( 𝑡
) 𝑑𝐴

𝑥 𝑥
Using linear shape functions 1 − 𝐿 and 𝐿 , 𝜃𝑦 and 𝜃𝑧 are interpolated along the element and rewritten
using the bending terms in Timoshenko’s beam theory as:
𝑑𝑣(𝑥) 1 𝑥 1 𝑥
𝑑𝑥
= 𝛾̅𝑥𝑦 + 𝜃𝑧 (𝑥) = (2 − 𝐿 ) 𝜃𝑧1 − (2 − 𝐿 ) 𝜃𝑧2
𝑑𝑤(𝑥) 1 𝑥 1 𝑥
(78)
𝑑𝑥
= 𝛾̅𝑥𝑧 − 𝜃𝑦 (𝑥) = − (2 − 𝐿 ) 𝜃𝑦1 + (2 − 𝐿 ) 𝜃𝑦2

The strain vector 𝛆 can be written as:

𝜀 1 𝑡𝑖𝑚𝑜 𝑎−𝑏 𝑇 𝑡𝑖𝑚𝑜 𝑎−𝑏

(𝐁𝑁𝐿 𝐝𝑙𝑜𝑐 ) (𝐁𝑁𝐿 𝐝𝑙𝑜𝑐 )}
𝛆 = {𝛾} = 𝐁𝐿𝑡𝑖𝑚𝑜 𝐝𝑎−𝑏
𝑙𝑜𝑐 + {2 (79)
0
In which:
1 𝑧 𝑧 𝑦 𝑦
𝐿
−𝐿 𝐿 𝐿
−𝐿
𝐁𝐿𝑡𝑖𝑚𝑜 (𝑥, 𝑦, 𝑧) =[ ℎ(𝑦,𝑧) ℎ(𝑦,𝑧) 𝑔(𝑦,𝑧) 𝑔(𝑦,𝑧)
]
0 2 2
− 2
− 2
1 𝑥 𝑥 1
(80)
0 0 0 − −
𝑡𝑖𝑚𝑜 ( 2 𝐿 𝐿 2
𝐁𝑁𝐿 𝑥, 𝑦, 𝑧) = [ 𝑥 1 1 𝑥 ]
0 𝐿
−2 2
−𝐿 0 0

The local force vector and stiffness matrix are represented as:
𝑇
𝑡𝑖𝑚𝑜 𝑎−𝑏 𝑇 𝑡𝑖𝑚𝑜 𝜎
𝑎−𝑏
𝐟𝑙𝑜𝑐 = ∫𝑉 (𝐁𝐿𝑡𝑖𝑚𝑜 + [(𝐁𝑁𝐿 𝐝𝑙𝑜𝑐 ) 𝐁𝑁𝐿 ]) { } 𝑑𝑉 (81)
𝟎 𝜏

𝑇 𝑇
𝐊 𝑎−𝑏 𝑡𝑖𝑚𝑜 𝑡𝑖𝑚𝑜
𝑙𝑜𝑐 = ∫𝑉 ((𝐁𝑁𝐿 ) 𝐁𝑁𝐿 ) σ𝑑𝑉 + ⋯

𝑡𝑖𝑚𝑜 𝑎−𝑏 𝑡𝑖𝑚𝑜 𝑇 𝑇

𝑡𝑖𝑚𝑜 𝑎−𝑏 𝑡𝑖𝑚𝑜 𝑇 (82)
∫𝑉 (𝐁𝐿𝑡𝑖𝑚𝑜 + [(𝐁𝑁𝐿 𝐝𝑙𝑜𝑐 ) 𝐁𝑁𝐿 ]) 𝐂 (𝐁𝐿𝑡𝑖𝑚𝑜 + [(𝐁𝑁𝐿 𝐝𝑙𝑜𝑐 ) 𝐁𝑁𝐿 ]) 𝑑𝑉
𝟎 𝟎
In which 𝜏 is shear stress and for elastic material the stress strain relationship is defined as follows:
𝐸 0 𝜎 𝐸 0 𝜀
𝐂=[ ] ,𝛔 = { } = [ ]{ } (83)
0 𝐺 𝜏 0 𝐺 𝛾

195

SACS® Collapse Advanced Release 15.1

In case of elastoplastic material, the property matrix 𝐂 and stress vector 𝛔 are determined though
elastoplastic iterative procedure as described in section 6.5.1.

6.3.2.3 Wagner Nonlinear Torsion

The second order Wagner strain [19, 37, 38], is used to account for large torsional deformation in thin
walled open sections. The Wagner strain is represented by the following expression:
1 𝑑𝜃𝑥 2
𝜀 𝑊𝑎𝑔𝑛𝑒𝑟 = 2 (𝜌(𝑦, 𝑧) 𝑑𝑥
) (84)

In which 𝜌 is distance from section shear center to a given point (𝑦, 𝑧) in the cross section and 𝜃𝑥 is
torsional twist. Assuming a linear shape function to interpolate 𝜃𝑥 :

1 𝜌 𝜌 𝜃𝑥1 𝑇 𝜌 𝜌 𝜃𝑥1
𝜀 𝑊𝑎𝑔𝑛𝑒𝑟 = 2 ൬[− 𝐿 ] {
𝐿 𝜃𝑥2
}൰ ൬[−𝐿 ]{ }൰ (85)
𝐿 𝜃𝑥2
The second order strain operators for the Wagner strain used to define the local force vectors
corresponding to Euler-Bernoulli and Timoshenko beam theories, after replacing 𝐝𝑎−𝑏
𝑙𝑜𝑐 with 𝐝𝑙𝑜𝑐 can be
written as:

Euler-Bernoulli
𝑊𝑎𝑔𝑛𝑒𝑟
𝐁𝐿 (𝑥, 𝑦, 𝑧) = [𝐁𝐿𝑒𝑢𝑙𝑒𝑟 0 0]
𝑒𝑢𝑙𝑒𝑟
𝑊𝑎𝑔𝑛𝑒𝑟 𝐁𝑁𝐿 𝟎2×1 𝟎2×1 (86)
𝐁𝑁𝐿 (𝑥, 𝑦, 𝑧) = [ 𝜌 𝜌 ]
𝟎1×5 −𝐿 𝐿
Timoshenko
𝑊𝑎𝑔𝑛𝑒𝑟
𝐁𝐿 (𝑥, 𝑦, 𝑧) = [𝐁𝐿𝑡𝑖𝑚𝑜 𝟎2×1 𝟎2×1 ]
𝑡𝑖𝑚𝑜
𝑊𝑎𝑔𝑛𝑒𝑟 𝐁𝑁𝐿 𝟎2×1 𝟎2×1 (87)
𝐁𝑁𝐿 ( )
𝑥, 𝑦, 𝑧 = [ 𝜌 𝜌 ]
𝟎1×5 −𝐿 𝐿

196

SACS® Collapse Advanced Release 15.1

6.3.3 Beam Element Numerical Integration
The calculation of the local force vector and stiffness matrix entails integration over the element
volume. The general form of these integrals can be rewritten as:

I = ∬𝑥,𝐴 𝑓 (𝑥, 𝑦, 𝑧)𝑑𝑥𝑑𝐴 (88)

Using the quadrature rule, the above integral can be numerically calculated using following form:
𝑀 𝑁
𝑝
I = ∑𝑖=1 𝑝
∑𝑗=1 𝑓(𝑥𝑖 , 𝑦𝑗 , 𝑧𝑗 )𝑊𝑖𝑥 𝑊𝑗𝐴 (89)

In which 𝑊𝑖𝑥 and 𝑊𝑗𝐴 j are integration weights associated with 𝑥 and area 𝐴, respectively.

The Gauss–Legendre quadrature integration points are used to calculate equation (89).
Two integration points 𝑥1 and 𝑥2 are used along the x axis such that:
1 𝐿√3 𝐿 1 𝐿√3 𝐿
𝑀𝑝 = 2 , (𝑥1 = − , 𝑊1𝑥 = ) , (𝑥2 = + , 𝑊2𝑥 = ) (90)
2 6 2 2 6 2

in which the standard Gauss–Legendre points and weights are transformed for integration limit (−1,1)
to (0, 𝐿).

A similar procedure can be carried out for beam cross section by partitioning the beam cross section
into smaller parts and defining the integration points within each part. Table 1 summarizes the
partitions and the number of Gauss–Legendre points for various beam cross-sections. The red dots in
Figure 2 illustrates the location of the integration points for various cross section types.

It should be noted that Collapse Advanced considers a change in cross-section dimensions for tapered
members (or cone sections) when determining the coordinates of integration points within the cross-
section.

Note: Integration points are located at the centerline of partitions for thin walled sections.

197

SACS® Collapse Advanced Release 15.1

6.4 Plate Element
The geometric nonlinearity for plate elements is defined using corotational method while the local
terms are determined using a standard finite element formulation assuming elastoplastic materials.
Two types plate elements are supported, a 3-node triangular plate element and a 4-node quadrilateral
plate element.
Both Kirchhoff (thin plate) and Mindlin (thick plate) bending theories are supported.

The Discrete Kirchhoff Theory (DKT) is used to enforce thin plate bending assumptions for thin triangular
(DKT) and thin quadrilateral (DKQ) elements. For the Mindlin triangular elements, a correction factor is
used to eliminate shear-locking commonly associated with thick plate elements. For the Mindlin
quadrilateral element the shear strain is formulated in a such a way as to prevent shear-locking without
the use of a correction factor.

6.4.1 Plate Element Corotational Formulation for Geometric Nonlinearity

6.4.1.1 Local Displacement, Element, and Joints Rotation Matrices

Each plate element is considered to have two sets of local coordinate systems, element local coordinate
system 𝐄, and joint local coordinate systems, 𝐔 (𝑗). The local coordinate system associated with joints
are determined using unit quaternions as discussed in section 6.3.1.3.

The matrix associated with element local coordinate system 𝐄 matrix is determined using the
assumption that the element spin (rigid body rotation) is zero [2]. This condition is satisfied through
following equations

Triangular Plate (𝑁𝑗𝑛𝑡 = 3)

(2) (1)
𝐞𝑇3 (𝐱𝑖 − 𝐱𝑖 ) = 0
(3) (1)
𝐞𝑇3 (𝐱𝑖 − 𝐱𝑖 ) = 0 (91)
𝜕𝑢 𝜕𝑣
Ω𝑚 = ( − )| =0
𝜕𝑦 𝜕𝑥 𝑐𝑒𝑛𝑡𝑒𝑟
Quadrilateral Plate (𝑁𝑗𝑛𝑡 = 4)
(3) (1)
𝐞𝑇3 (𝐱𝑖 − 𝐱𝑖 ) = 0
(4) (𝟐)
𝐞𝑇3 (𝐱𝑖 − 𝐱𝑖 ) = 0 (92)
𝜕𝑢 𝜕𝑣
Ω𝑚 = ( − )| =0
𝜕𝑦 𝜕𝑥 𝑐𝑒𝑛𝑡𝑒𝑟

Following [2], 𝐄 matrix can be written in following form

cos (𝛾) −sin (𝛾) 0
𝐄 = [𝐞1 𝐞2 𝐞3 ] = [𝐞̅1 𝐞̅2 𝐞̅3 ] [ sin (𝛾) cos (𝛾) 0] (93)
0 0 1
where [𝐞̅1 𝐞̅2 𝐞̅3 ] corresponds to the joint coordinate system in the deformed configuration which
conform to the first two conditions in equations (91) and (92). The value for 𝛾 is determined such that
the third condition in equation (91) and (92) is satisfied. Hence [𝐞̅1 𝐞̅2 𝐞̅3 ] can be expressed as
follows [2]

198

SACS® Collapse Advanced Release 15.1

For Triangular Plate Elements (𝑁𝑗𝑛𝑡 = 3)
(2) (1) (2) (1) (3) (1)
𝐱𝑖 −𝐱𝑖 (𝐱𝑖 −𝐱𝑖 )×(𝐱𝑖 −𝐱𝑖 )
𝐞̅1 = (2) (1) , 𝐞̅3 = (2) (1) (3) (1) , 𝐞̅2 = 𝐞̅3 × 𝐞̅1 (94)
|𝐱𝑖 −𝐱𝑖 | |(𝐱𝑖 −𝐱𝑖 )×(𝐱𝑖 −𝐱𝑖 )|

For Quadrilateral Plate Elements (𝑁𝑗𝑛𝑡 = 4) (see section 18.5 in [2])

(3) (1) (4) (2) (2) (1) (2) (1) 𝑇

(𝐱𝑖 −𝐱𝑖 )×(𝐱𝑖 −𝐱𝑖 ) (𝐱𝑖 −𝐱𝑖 )−൬(𝐱𝑖 −𝐱𝑖 ) 𝐞̅3 ൰𝐞̅3
𝐞̅3 = (3) (1) (4) (2) , 𝐞̅1 = (2) (1) (2) (1) 𝑇
, 𝐞̅2 = 𝐞̅3 × 𝐞̅1 (95)
|(𝐱𝑖 −𝐱𝑖 )×(𝐱𝑖 −𝐱𝑖 )| |(𝐱𝑖 −𝐱𝑖 )−൬(𝐱𝑖 −𝐱𝑖 ) 𝐞̅3 ൰𝐞̅3 |

The rotation angle 𝛾 about 𝐞̅3 is obtained assuming zero rotation at the element center [2]
𝜕𝑢 𝜕𝑣
𝛾𝑚 = (𝜕𝑦 − 𝜕𝑥 )| =0 (96)
𝑐𝑒𝑛𝑡𝑒𝑟

Using the element shape functions (see section 6.4.2 for element shape function and [2]), the above
equation can be rewritten as
𝐴𝑐𝑜𝑠(𝛾) + 𝐵𝑠𝑖𝑛(𝛾) = 𝐶
(𝑗) (𝑗)
𝑁𝑗𝑛𝑡 𝑇 𝑢
̅𝑖 𝑁𝑗𝑛𝑡 𝑇 𝑣̅ 𝑁𝑗𝑛𝑡 𝑇 𝑋𝑗 − 𝑋1 (97)
(𝑗) (𝑗) (𝑗)
𝐴= ∑𝑗=2 (𝐚 ) { (𝑗) } , 𝐵 = ∑𝑗=2 (𝐚 ) { 𝑖 (𝑗) } , 𝐶 = ∑𝑗=2 (𝐚 ) { }
𝑣̅𝑖 −𝑢̅𝑖 𝑌𝑗 − 𝑌1

(𝑗)
In which 𝐚 represents the linear shape functions derivatives calculated at the element center
(𝑗)
𝜕𝑁𝐿
(𝑗) 𝜕𝑦
𝐚 ={ (𝑗)
}| (98)
𝜕𝑁𝐿
− 𝜕𝑥 𝑐𝑒𝑛𝑡𝑒𝑟

(𝑗) (𝑗)
where 𝑢̅𝑖 and 𝑣̅𝑖 are the joint displacement in [𝐞̅1 𝐞̅2 𝐞̅3 ] corresponding to the ith iteration
(𝑗)
𝑢̅𝑖
(𝑗)
𝑣̅𝑖 = [𝐞̅1 𝐞̅2 𝐞̅3 ]𝑇 (𝐱𝑖(𝑗) − 𝐱𝑖(1) ) (99)
(𝑗)
{𝑤
̅𝑖 }

Once the element and joints local coordinate systems (𝐄 and 𝐔(𝑗)) are defined, the local displacement
DOFs at ith iteration can be expressed as [2]:
(𝑗) (𝑗) (1)
𝐩𝑙𝑜𝑐 = 𝐄𝑇 (𝐱𝑖 − 𝐱𝑖 ) − (𝐗 (𝑗) − 𝐗 (1) )
(𝑗) (𝑗)
𝐞𝑇3 𝐮2 − 𝐞𝑇2 𝐮3
(𝑗) 1
(100)
𝛉𝑙𝑜𝑐 = {𝐞1𝑇 𝐮(𝑗) 𝑇 (𝑗) }
3 − 𝐞3 𝐮1
2
(𝑗) (𝑗)
𝐞𝑇2 𝐮1 − 𝐞1𝑇 𝐮2

Where the second term in equation (100) representing rotations is based on equation (28) for beam
elements assuming small angles. The translation DOFs are calculated relative to the 1st element joint so
the local translation at this point is zero.

199

SACS® Collapse Advanced Release 15.1

6.4.1.2 Transformation Matrix
To simplify plate corotational transformation, the local displacement vector can be expressed in terms
of the local translation and rotation DOFs:

𝛿𝐩𝑙𝑜𝑐 𝐓𝑝𝑝 𝟎 𝛿𝐩𝑔𝑙𝑏

𝛿𝐝𝑙𝑜𝑐 = { }=[ ]{ } (101)
𝛿𝛉𝑙𝑜𝑐 𝐓𝜃𝑝 𝐓𝜃𝜃 𝛿𝛉
𝑔𝑙𝑏

The above equation assumes that the local translation DOFs 𝐩𝑙𝑜𝑐 are only dependent upon 𝛿𝐩𝑔𝑙𝑏 (see
equation (100)). To define the transformation matrix, the iterative change in the local element
coordinate system 𝛿𝐄 can be written as:

𝛿𝐄 = 𝐒(𝛿𝛃)[𝐞1 𝐞2 𝐞3 ] (102)

in which 𝐒 is the spin operator and 𝛿𝛃 is the iterative change in local system positioning which can be
expressed as:
(1)
𝐩𝑔𝑙𝑏
𝑖
𝛿𝛃 = 𝐕 ⋮ , 𝐕 = 𝚽−1 𝚿 (103)
(𝑁𝑗𝑛𝑡 )
{𝐩𝑔𝑙𝑏 𝑖 }

In which matrices 𝚽 and 𝚿 can be expressed as:

Triangular Plate (𝑁𝑗𝑛𝑡 = 3)

𝑁𝑗𝑛𝑡 𝐚 (𝑗) 𝑇
(𝑗) (1)
∑𝑗=2 { } 𝐄𝑇 𝐒 (𝐱𝑖 − 𝐱𝑖 )
0
𝑇
𝚽= (2) (1) (104)
(𝐱𝑖 − 𝐱𝑖 ) 𝐒(𝐞3 )
𝑇
[ (𝐱𝑖(3) − 𝐱𝑖(1) ) 𝐒(𝐞3 ) ]
(𝑗) 𝑇 (2) 𝑇 (3) 𝑇
𝑗𝑛𝑡 𝐚
− {𝐚 } 𝐄 𝑇 − {𝐚 } 𝐄 𝑇
𝑁
∑𝑗=2 { } 𝐄𝑇
𝚿= 0 0 0 (105)
−𝐞𝑇3 𝐞𝑇3 𝟎
[ −𝐞𝑇3 𝟎 𝐞𝑇3 ]
Quadrilateral Plate (𝑁𝑗𝑛𝑡 = 4)

𝑁𝑗𝑛𝑡 𝐚 (𝑗) 𝑇
(𝑗) (1)
∑𝑗=2 { } 𝐄𝑇 𝐒 (𝐱𝑖 − 𝐱𝑖 )
0
𝑇
𝚽= (3) (1) (106)
(𝐱𝑖 − 𝐱𝑖 ) 𝐒(𝐞3 )
𝑇
[ (𝐱𝑖(4) − 𝐱𝑖(2) ) 𝐒(𝐞3 ) ]

200

SACS® Collapse Advanced Release 15.1

 (𝑗) 𝑇 (2) 𝑇 (3) 𝑇 (4) 𝑇
𝑗𝑛𝑡 𝐚
− {𝐚 } 𝐄 𝑇 {𝐚 } 𝐄 𝑇 {𝐚 } 𝐄 𝑇
𝑁
∑𝑗=2 { } 𝐄𝑇
𝚿= 0 0 0 0 (107)
−𝐞𝑇3 𝟎 𝐞𝑇3 𝟎
[ 𝟎 −𝐞𝑇3 𝟎 𝐞𝑇3 ]
Using the spin operator cross-product definition given by equations (16) to (18)

𝛿𝐞𝑛 = −𝐒(𝐞𝑛 )𝐕𝛿𝐩𝑔𝑙𝑏 (108)

For the joint local coordinate systems 𝐔(𝑗), which are a function of rotation only:
(𝑗) (𝑗) (𝑗) (𝑗) (𝑗) (𝑗) (𝑗)
𝛿𝐮𝑛 = 𝐮𝑛 | (𝑗) (𝑗) − 𝐮𝑛 | (𝑗) = 𝐒(𝛿𝛉𝑔𝑙𝑏 ) 𝐮𝑛 = −𝐒(𝐮𝑛 )𝛿𝛉𝑔𝑙𝑏 (109)
𝛉𝑔𝑙𝑏 +𝛿𝛉𝑔𝑙𝑏 𝛉𝑔𝑙𝑏

To calculate 𝐓𝑝𝑝 , the variation of 𝛿𝐩𝑙𝑜𝑐 in equation (101) is used:

(𝑗) (𝑗) (1) (𝑗) (1)
𝛿𝐩𝑙𝑜𝑐 = −𝐄𝑇 𝐒(𝛿𝛃) (𝐱𝑖 − 𝐱𝑖 ) + 𝐄𝑇 𝛿 (𝐱𝑖 − 𝐱𝑖 ) (110)

Using equation (108) and after few manipulation, yields the following matrices:

Triangular Plate (𝑁𝑗𝑛𝑡 = 3)

𝟎
𝑇 (1) (2) 𝟎 𝟎 𝟎
𝐓𝑝𝑝 = [𝐄 𝐒 (𝐱𝑖
− 𝐱 𝑖 )] 𝐕 + [ 𝑇
−𝐄 𝐄𝑇 𝟎] (111)
(3)
𝐄𝑇 𝐒 (𝐱𝑖 − 𝐱𝑖 )
(1) −𝐄𝑇 𝟎 𝐄𝑇
Quadrilateral Plate (𝑁𝑗𝑛𝑡 = 4)
𝟎
𝑇 (2) 𝟎 (1) 𝟎 𝟎 𝟎
𝐄 𝐒 (𝐱𝑖 − 𝐱𝑖 )
−𝐄𝑇 𝐄𝑇 𝟎 𝟎
𝐓𝑝𝑝 = 𝑇 (𝐱 (3) (1) 𝐕 + [ ] (112)
𝐄 𝐒 𝑖 − 𝐱𝑖 ) −𝐄𝑇 𝟎 𝐄𝑇 𝟎
𝑇 (𝐱 (4) (1) −𝐄𝑇 𝟎 𝟎 𝐄𝑇
[𝐄 𝐒 𝑖 − 𝐱 𝑖 ) ]

To calculate 𝐓𝜃𝑝 and 𝐓𝜃𝜃 , the local rotation in equation (100) may be differentiated with respect to the
global displacement vector:
(𝑗) (𝑗) (𝑗) (𝑗)
𝛿𝐞𝑇3 𝐮2 + 𝐞𝑇3 𝛅𝐮2 − 𝛿𝐞𝑇2 𝐮3 − 𝐞𝑇2 𝛿𝐮3
(𝑗) 1 (𝑗) (𝑗) (𝑗) (𝑗)
𝛿𝛉𝑙𝑜𝑐 =
2
𝛿𝐞1𝑇 𝐮3 + 𝐞1𝑇 𝛿𝐮3 − 𝛿𝐞𝑇3 𝐮1 − 𝐞𝑇3 𝛿𝐮1 (113)
(𝑗) (𝑗) (𝑗) (𝑗)
{ 𝛿𝐞𝑇2 𝐮1 + 𝐞𝑇2 𝛿𝐮1 − 𝛿𝐞1𝑇 𝐮2 − 𝐞1𝑇 𝛿𝐮2 }

(𝑗)
Using equations (108) and (109) for 𝛿𝐞𝑛 and 𝛿𝐮𝑛 :

𝛿𝛉𝑙𝑜𝑐 = 𝐓𝜃𝑝 𝛿𝐩𝑔𝑙𝑏 + 𝐓𝜃𝜃 𝛿𝛉𝑔𝑙𝑏 (114)

where
𝐔̃ (𝑗) 𝐒(𝐞1 )
1
𝐓𝜃𝑝 = 2 [ ⋮ ] [𝐒(𝐞2 )] 𝐕 (115)
̃ (𝑁𝑗𝑛𝑡 ) 𝐒(𝐞 )
𝐔 3

201

SACS® Collapse Advanced Release 15.1

 (1)
𝐒(𝐮1 )
𝐄̃ [𝐒(𝐮(1)
2 )
] ⋯ 𝟎
(1)
𝐒(𝐮3 )
1
𝐓𝜃𝜃 = − 2 ⋮ ⋱ ⋮ (116)
(𝑁𝑗𝑛𝑡 )
𝐒(𝐮1 )
(𝑁 )
𝟎 ⋯ 𝐄̃ 𝐒(𝐮2 𝑗𝑛𝑡
)
(𝑁𝑗𝑛𝑡 )
[ [𝐒(𝐮3 )]]
with
(𝑗)
𝟎 𝐞𝑇3 −𝐞𝑇2 𝟎 𝐮𝑇3 −𝐮𝑇2
̃
𝐄 = [−𝐞𝑇3 𝟎 𝑇 ] ̃ (𝑗)
𝐞1 , 𝐔 = [−𝐮𝑇3 𝟎 𝐮1𝑇 ] (117)
𝐞𝑇2 −𝐞1𝑇 𝟎 𝐮𝑇2 −𝐮1𝑇 𝟎
6.4.1.3 Geometric Stiffness

Substituting the transformation matrix given by equation (101) into equation (8) yields the following
geometric stiffness matrix:

𝐟𝑔𝑙𝑏 𝑇 𝑇
𝐓𝑝𝑝 𝐓𝜃𝑝 𝐟𝑙𝑜𝑐
{ }=[ 𝑇
]{ } (118)
𝐦𝑔𝑙𝑏 𝟎 𝐓𝜃𝜃 𝐦𝑙𝑜𝑐

The correspond geometric stiffness 𝐊𝑔𝑒𝑜 matrix is given by:

𝐊𝑔𝑝𝑝 𝐊𝑔𝜃𝑝
𝐊𝑔𝑒𝑜 = [ 𝑻 ] (119)in which
𝐊𝑔𝜃𝑝 𝐊𝑔𝜃𝜃
𝑇 𝑇
𝛿𝐓𝑝𝑝 𝛿𝐓𝜃𝑝
𝐊𝑔𝑝𝑝 = 𝐟𝑙𝑜𝑐 + 𝐦𝑙𝑜𝑐
𝛿𝐩𝑔𝑙𝑏 𝛿𝐩𝑔𝑙𝑏
𝑇
𝛿𝐓𝜃𝑝
𝐊𝑔𝜃𝑝 = 𝐦𝑙𝑜𝑐 (120)
𝛿𝛉𝑔𝑙𝑏
𝑇
𝛿𝐓𝜃𝜃
𝐊𝑔𝜃𝜃 = 𝐦𝑙𝑜𝑐
𝛿𝛉𝑔𝑙𝑏

Which assumes 𝐓𝑝𝑝 is not a function of the rotation.

𝑇
𝛿𝐓𝑝𝑝
Starting with 𝐟 :
𝛿𝐩𝑔𝑙𝑏 𝑙𝑜𝑐

𝑇
𝛿𝐓𝑝𝑝 𝐟𝑙𝑜𝑐 = 𝐀1 𝐟𝑙𝑜𝑐 − 𝐀2 𝐟𝑙𝑜𝑐 − 𝐀3 𝐟𝑙𝑜𝑐 (121)
with [2]
(𝑗)
𝐒 (𝐄𝐟𝑙𝑜𝑐 )
𝐀1 𝐟𝑙𝑜𝑐 = − ⋮ 𝐕𝛿𝐩𝑔𝑙𝑏 = 𝐁1 𝛿𝐩𝑔𝑙𝑏
(𝑁𝑗𝑛𝑡 )
[𝐒 ൬𝐄𝐟𝑙𝑜𝑐 ൰] (122)
𝑇 𝑁𝑗𝑛𝑡 (𝑗) (1) (𝑗)
𝐀2 𝐟𝑙𝑜𝑐 = −𝐕 (∑𝑗=2 𝐒 (𝐱𝑖 − 𝐱𝑖 ) 𝐒 (𝐄𝐟𝑙𝑜𝑐 )) 𝐕𝛿𝐩𝑔𝑙𝑏 = 𝐁2 𝛿𝐩𝑔𝑙𝑏
(𝑗) (𝑁 )
𝐀3 𝐟𝑙𝑜𝑐 = −𝐕 𝑇 [𝐒 (𝐄𝐟𝑙𝑜𝑐 ) ⋯ 𝐒 ൬𝐄𝐟𝑙𝑜𝑐𝑗𝑛𝑡 ൰] 𝛿𝐩𝑔𝑙𝑏 = −𝐁𝟏𝑇 𝛿𝐩𝑔𝑙𝑏

202

SACS® Collapse Advanced Release 15.1

where 𝛿𝐕 is assumed to be negligible and has been shown not to play a significant role in the tangent
stiffness matrix. Thus, considering matrix 𝐁2 is not necessarily a symmetric matrix:
𝑇
𝛿𝐓𝑝𝑝 𝐁2 +𝐁2𝑇
𝐟𝑙𝑜𝑐 = 𝐁1 + 𝐁1𝑇 − 𝟐
(123)
𝛿𝐩𝑔𝑙𝑏

𝑇 𝑇
𝛿𝐓𝜃𝑝 𝛿𝐓𝜃𝑝
To calculate 𝐦𝑙𝑜𝑐 and 𝐦𝑙𝑜𝑐 the following expressions are used:
𝛿𝐩𝑔𝑙𝑏 𝛿𝛉𝑔𝑙𝑏

𝐑1
1
− 2 𝐕 𝑇 [𝐒(𝛿𝐞1 ) 𝐒(𝛿𝐞2 ) 𝐒(𝛿𝐞3 )] {𝐑 2 } + ⋯
𝐑3
(124)
𝛿𝐑1
1
− 2 𝐕 𝑇 [𝐒(𝛿𝐞1 ) 𝐒(𝛿𝐞2 ) 𝐒(𝛿𝐞3 )] {𝛿𝐑 2 }
𝛿𝐑 3
with
𝐑1
𝑁𝑗𝑛𝑡 (𝑗) (𝑗)
{𝐑 2 } = ∑𝑗=1 ̃ 𝐦𝑙𝑜𝑐
𝐔 (125)
𝐑3
𝑇 𝑇
𝛿𝐓𝜃𝑝 𝛿𝐓𝜃𝑝
where the first term is associated with 𝐦𝑙𝑜𝑐 while the second term leads to 𝐦𝑙𝑜𝑐 . Considering
𝛿𝐩𝑔𝑙𝑏 𝛿𝛉𝑔𝑙𝑏

equation (108) for 𝛿𝐞𝑛 [2]:

𝑇
𝛿𝐓𝜃𝑝
𝐦𝑙𝑜𝑐 = 𝐕 𝑇 [𝐒(𝐑1 )𝐒(𝐞1 ) 𝐒(𝐑 2 )𝐒(𝐞2 ) 𝐒(𝐑 3 )𝐒(𝐞3 )]𝐕 (126)
𝛿𝐩𝑔𝑙𝑏

where the middle term is not necessarily a symmetric matrix so it is symmetrized for the analysis.

Next [2],
(𝑗) (𝑁𝑗𝑛𝑡 )
𝛿𝐑1 𝐖1 ⋯ 𝐖1
𝑁𝑗𝑛𝑡 (𝑗) (𝑗𝑁𝑗𝑛𝑡 )
̃ 𝐦𝑙𝑜𝑐 = 𝐖 (𝑗)
{𝛿𝐑 2 } = ∑𝑗=1 𝛿𝐔 (𝑗)
⋯ 𝐖2 𝛿𝛉𝑔𝑙𝑏 (127)
2
𝛿𝐑 3 (𝑗) (𝑁𝑗𝑛𝑡 )
[𝐖3 ⋯ 𝐖3 ]

with
(𝑗) (𝑗) (𝑗) (𝑗) (𝑗)
𝐖1 = 𝑚2𝑙𝑜𝑐 𝐒 (𝐮3 ) − 𝑚3𝑙𝑜𝑐 𝐒 (𝐮2 )
(𝑗) (𝑗) (𝑗) (𝑗) (𝑗)
𝐖2 = −𝑚1𝑙𝑜𝑐 𝐒 (𝐮3 ) + 𝑚3𝑙𝑜𝑐 𝐒 (𝐮1 ) (128)
(𝑗) (𝑗) (𝑗) (𝑗) (𝑗)
𝐖3 = 𝑚1𝑙𝑜𝑐 𝐒 (𝐮2 ) − 𝑚2𝑙𝑜𝑐 𝐒 (𝐮1 )

Hence,
(𝑗) (𝑁𝑗𝑛𝑡 )
𝐖1 ⋯ 𝐖1
𝑇
𝛿𝐓𝜃𝑝 1 (𝑗𝑁𝑗𝑛𝑡 )
𝐦𝑙𝑜𝑐 = − 2 𝐕 𝑇 [𝐒(𝛿𝐞1 ) 𝐒(𝛿𝐞2 ) 𝐒(𝛿𝐞3 )] 𝐖2(𝑗) ⋯ 𝐖2 (129)
𝛿𝛉𝑔𝑙𝑏
(𝑗) (𝑁 )
[𝐖3 ⋯ 𝐖3 𝑗𝑛𝑡 ]

203

SACS® Collapse Advanced Release 15.1

 𝑇
𝛿𝐓𝜃𝜃
The final matrix block is 𝐦𝑙𝑜𝑐 and like previous terms:
𝛿𝛉𝑔𝑙𝑏

𝑁𝑗𝑛𝑡 1 (𝑗) (𝑗) (𝑗) (𝑗)

𝑇
𝛿𝐓𝜃𝜃 𝐦𝑙𝑜𝑐 = ∑𝑗=1 [𝐒 (𝛿𝐮1 ) 𝐒 (𝛿𝐮2 ) 𝐒 (𝛿𝐮3 )] 𝐄̃𝐦𝑙𝑜𝑐 (130)
2

assuming
(𝑗) (𝑗)
𝐐1 = [𝟎 −𝐞3 𝐞2 ]𝐦𝑙𝑜𝑐
(𝑗) (𝑗)
𝐐2 = [𝐞3 𝟎 −𝐞1 ]𝐦𝑙𝑜𝑐 (131)
(𝑗) (𝑗)
𝐐3 = [−𝐞𝟐 𝐞1 𝟎]𝐦𝑙𝑜𝑐

and using spin operator cross-product definition in equations (16) to (18), the final part (a block diagonal
matrix) is calculated as:

𝑇
𝛿𝐓𝜃𝜃
𝐃(𝑗) ⋯ 𝟎
𝐦𝑙𝑜𝑐 =[ ⋮ ⋱ ⋮ ] (132)
𝛿𝛉𝑔𝑙𝑏
(𝑁𝑗𝑛𝑡 )
𝟎 ⋯ 𝐃
where
1 (𝑗) (𝑗) (𝑗) (𝑗) (𝑗) (𝑗)
𝐃(𝑗) = 2 ൬𝐒 (𝐐1 ) 𝐒 (𝐮1 ) + 𝐒 (𝐐2 ) 𝐒 (𝐮2 ) + 𝐒 (𝐐3 ) 𝐒 (𝐮3 )൰ (133)

204

SACS® Collapse Advanced Release 15.1

6.4.2 Plate Element Local Stiffness Matrix and Load Vector

The variational energy expression for a plate element is as follows: 𝛿𝜋 = ∫𝑉 𝛿𝛆𝑇 𝛔𝑑𝑉 +
∫𝑉 𝛿𝛾𝑑𝑟𝑖𝑙𝑙 𝛼𝐺𝛾𝑑𝑟𝑖𝑙𝑙 𝑑𝑉 (134)

where the first energy term is associated with plate membrane, bending, and out-of-plane shear (for
Mindlin bending theory) strains. The second term is drilling energy to stabilize plate elements against in-
plane twist (i.e. drilling) [39].

Membrane strain is given by:

𝜕
0
𝜀𝑚 𝑥 𝜕𝑥
𝜕 𝑢
𝛆𝑚 (𝑥, 𝑦, 𝑧) = { 𝜀𝑚 𝑦 } = 0 { } (135)
𝛾𝑚 𝑥𝑦
𝜕𝑦 𝑣
𝜕 𝜕
[𝜕𝑦 𝜕𝑥 ]

The bending strains are defined as

𝜕
0 𝑧
𝜀𝑏 𝑥 𝜕𝑦
𝜀 𝜕 𝜃𝑥
𝛆𝑏 (𝑥, 𝑦, 𝑧) = { 𝑏 𝑦 } = −𝑧 0 { } (136)
𝛾𝑏 𝑥𝑦 𝜕𝑥 𝜃𝑦
𝜕 𝜕
[−𝑧 𝜕𝑥 𝑧
𝜕𝑦 ]

The drilling term is defined by the in-plane shear strain given by:
1 𝜕𝑢 𝜕𝑣
𝛾𝑑𝑟𝑖𝑙𝑙 = 𝜃𝑧 + ( − ) (137)
2 𝜕𝑦 𝜕𝑥

The drilling stress can be expressed as:

𝜏𝑑𝑟𝑖𝑙𝑙 = 𝛼𝐺𝛾𝑑𝑟𝑖𝑙𝑙 (138)

where 𝛼 is a user-defined factor. Collapse Advanced uses default value of 0.05 for 𝛼which can be revised
on the CLPOP2 input line. It is worth noting that the drilling term does not contribute to elastoplastic
stress-strain calculations and is assumed to remain elastic.

Note: The drilling shear strain and stress are assumed constant with respect to the plate thickness

Note: It has been seen that for most of the analytical problems, the default value of 0.05 produces
accurate results. Reference [39] shows 𝛼 can be as large as 10 and still does not affect accuracy of
results.

Mindlin plate bending theory includes the effect of the out-of-plane shear strain which is defined as:
𝛾𝑥𝑧 𝛾̅𝑥𝑧
𝜸 = { 𝛾 } = 𝑓 (𝑧) { 𝛾̅ } (139)
𝑦𝑧 𝑦𝑧

in which 𝛾̅𝑥𝑧 and 𝛾̅𝑦𝑧 are shear strains at plate mid-plane and the distribution function 𝑓 (𝑧) is defined as:
5 5𝑧 2
𝑓 (𝑧) = − (140)
4 𝑡2

205

SACS® Collapse Advanced Release 15.1

For all plate elements, in-plane displacement and rotation (𝑢, 𝑣, 𝜃𝑧 ) are interpolated using standard
finite element linear shape functions. For triangular plate, the linear shape functions are defined in
isoparametric system as
(1) (2) (3)
𝑁𝐿 (𝜉, 𝜂) = 1 − 𝜉 − 𝜂 , 𝑁𝐿 (𝜉, 𝜂) = 𝜉 , 𝑁𝐿 (𝜉, 𝜂) = 𝜂 (141)

For a quadrilateral plate element:

(1) 1
𝑁𝐿 (𝜉, 𝜂) = 4 (1 − 𝜉 )(1 − 𝜂)
(2) 1
𝑁𝐿 (𝜉, 𝜂) = (1 + 𝜉 )(1 − 𝜂)
4
(3) 1 (142)
𝑁𝐿 (𝜉, 𝜂) = 4 (1 + 𝜉 )(1 + 𝜂)
(4) 1
𝑁𝐿 (𝜉, 𝜂) = 4 (1 − 𝜉 )(1 + 𝜂)

Based on above shape functions, the mapping from nodal element coordinates to isoparametric system
can be defined as:
𝑁
𝑗𝑛𝑡 (𝑗) 𝑗𝑛𝑡 𝑁 (𝑗)
𝑥 (𝜉, 𝜂) = ∑𝑗=1 𝑁𝐿 𝑋𝑗 , 𝑦(𝜉, 𝜂) = ∑𝑗=1 𝑁𝐿 𝑌𝑗 (143)

The relation between derivatives with respect to x and y is given by:

𝜕 𝜕
𝜕𝑥 𝜕𝜉
{𝜕} = 𝐉(𝜉, 𝜂 )−1 {𝜕} (144)
𝜕𝑦 𝜕𝜂

where 𝐉(𝜉, 𝜂) is Jacobian matrix defined as:

𝜕𝑥 𝜕𝑦
𝜕𝜉 𝜕𝜉
𝐉(𝜉, 𝜂) = [𝜕𝑥 𝜕𝑦
] (145)
𝜕𝜂 𝜕𝜂

Using above equations, the drilling shear strain can be represented as:
𝑗𝑛𝑡 𝑁 (𝑗) (𝑗)
𝛾𝑑𝑟𝑖𝑙𝑙 (𝜉, 𝜂) = ∑𝑗=1 𝐁𝑑𝑟𝑖𝑙𝑙 𝐝𝑙𝑜𝑐
(𝑗) 𝜕𝑁𝐿
(𝑗)
1 𝜕𝑁𝐿
(𝑗) (146)
𝐁𝑑𝑟𝑖𝑙𝑙 (𝜉, 𝜂) = [1 −
(𝑗)
0 0 0 𝑁𝐿 ]
2 𝜕𝑦 2 𝜕𝑥

Where 𝐁𝑑𝑟𝑖𝑙𝑙 (𝜉, 𝜂) can be written as:

(1) (2) (𝑁𝑗𝑛𝑡 )
𝐁𝑑𝑟𝑖𝑙𝑙 (𝜉, 𝜂) = [𝐁𝑑𝑟𝑖𝑙𝑙 𝐁𝑑𝑟𝑖𝑙𝑙 ⋯ 𝐁𝑑𝑟𝑖𝑙𝑙 ] (147)

The drilling contribution to the force vector and stiffness matrix is given by the following equations:

𝑑𝑟𝑖𝑙𝑙 𝑇
𝐟𝑙𝑜𝑐 = 𝛼𝐺 ∭(𝜉,𝜂,𝑡) 𝐁𝑑𝑟𝑖𝑙𝑙 𝛾𝑑𝑟𝑖𝑙𝑙 (𝜉, 𝜂)|𝐉|𝑑𝜉𝑑𝜂𝑑𝑧 (148)

𝑇
𝐊 𝑑𝑟𝑖𝑙𝑙
𝑙𝑜𝑐 = 𝛼𝐺 ∭(𝜉,𝜂,𝑡) 𝐁𝑑𝑟𝑖𝑙𝑙 𝐁𝑑𝑟𝑖𝑙𝑙 |𝐉|𝑑𝜉𝑑𝜂𝑑𝑧 (149)

where |𝐉| is the determinant of Jacobian matrix at (𝜉, 𝜂).

206

SACS® Collapse Advanced Release 15.1

The plate membrane strains can be expressed as:
𝑁𝑗𝑛𝑡 (𝑗) 𝑢𝑗
𝛆𝑚 (𝜉, 𝜂) = ∑𝑗=1 𝐁𝑚 { 𝑣 }
𝑗
(𝑗)
𝜕𝑁𝐿
𝜕𝑥
0
(𝑗) (150)
(𝑗) 𝜕𝑁𝐿
𝐁𝑚 (𝜉, 𝜂) = 0 𝜕𝑦
(𝑗) (𝑗)
𝜕𝑁𝐿 𝜕𝑁𝐿
[ 𝜕𝑦 𝜕𝑥 ]
(𝑗) (𝑗)
Similarly bending strain and shear strain operators 𝐁𝑏 and 𝐁𝑠 can be defined by the following
equations:
𝑤𝑗
𝑁𝑗𝑛𝑡 (𝑗)
𝛆𝑏 (𝜉, 𝜂, 𝑧) = ∑𝑗=1 𝐁𝑏 (𝜉, 𝜂, 𝑧) {𝜃𝑥𝑗 } (151)
𝜃𝑦𝑗
𝑤𝑗
𝑁𝑗𝑛𝑡 (𝑗)
𝜸(𝜉, 𝜂, 𝑧) = 𝜃
∑𝑗=1 𝐁𝑠 (𝜉, 𝜂, 𝑧) { 𝑥𝑗 } (152)
𝜃𝑦𝑗

Combining all strain terms, the general form of strain vector can be written as:
𝑁
𝑗𝑛𝑡 (𝑗) (𝑗)
𝛆(𝜉, 𝜂, 𝑧) = ∑𝑗=1 𝐁 𝐝𝑙𝑜𝑐 (153)

The strain vector for Kirchhoff (thin) plate and Mindlin (thick) plate can be defined as 𝛆𝑡ℎ𝑖𝑛 =
{𝜀𝑥 𝜀𝑦 𝛾𝑥𝑦 }𝑇 and 𝛆𝑡ℎ𝑖𝑐𝑘 = {𝜀𝑥 𝜀𝑦 𝛾𝑥𝑦 𝛾𝑥𝑧 𝛾𝑦𝑧 }𝑇 respectively. The corresponding strain
(𝑗)
operators 𝐁 are defined by the following equations:
(𝑗) (𝑗) (𝑗)
𝐁𝑡ℎ𝑖𝑛 (𝜉, 𝜂, 𝑧) = [𝐁𝑚 (𝜉, 𝜂) 𝐁𝑏 (𝜉, 𝜂, 𝑧) 𝟎3×1 ] (154)
(𝑗) (𝑗)
(𝑗) 𝐁 (𝜉, 𝜂) 𝐁𝑏 (𝜉, 𝜂, 𝑧) 𝟎3×1
𝐁𝑡ℎ𝑖𝑐𝑘 (𝜉, 𝜂, 𝑧) = [ 𝑚 (𝑗)
] (155)
𝟎2×2 𝐁𝑠 (𝜉, 𝜂, 𝑧) 𝟎2×1
(𝑗) (𝑗)
The details for bending and shear strain operator, 𝐁𝑏 and 𝐁𝑠 , are presented in sections 6.4.2.1 to
6.4.2.4 for different element type.

The general form of 𝐁(𝜉, 𝜂, 𝑧) can be written as:

(𝑁𝑗𝑛𝑡 )
𝐁(𝜉, 𝜂, 𝑧) = [𝐁(1) 𝐁
(2)
⋯ 𝐁 ] (156)

Which leads to the following expression for force vector and stiffness:

𝐟𝑙𝑜𝑐 = ∭(𝜉,𝜂,𝑡) 𝐁𝑇 𝛔|𝐉|𝑑𝜉𝑑𝜂𝑑𝑧 (157)

𝐊 𝑙𝑜𝑐 = ∭(𝜉,𝜂,𝑡) 𝐁𝑇 𝐂𝐁|𝐉|𝑑𝜉𝑑𝜂𝑑𝑧 (158)

207

SACS® Collapse Advanced Release 15.1

In which the stress vector 𝛔 and elastic constitutive matrix 𝐂 are given by the following expressions for
Kirchhoff and Mindlin plate theories:

Kirchhoff (thin) plate

𝐸 𝐸𝜈
𝜎𝑥 1−𝜈 2 1−𝜈 2
0
𝜎 𝑒𝑙𝑎𝑠𝑡𝑖𝑐
𝛔𝑡ℎ𝑖𝑛 = { 𝑦 } , 𝐂𝑡ℎ𝑖𝑛 = [ 𝐸𝜈 𝑒𝑙𝑎𝑠𝑡𝑖𝑐 𝑒𝑙𝑎𝑠𝑡𝑖𝑐
0 ] , 𝛔𝑡ℎ𝑖𝑛 = 𝐂𝑡ℎ𝑖𝑛 𝛆𝑡ℎ𝑖𝑛 (159)
𝐸
1
𝜏𝑥𝑦 1−𝜈 2 1−𝜈 2
0 0 𝐺
Mindlin (thick) plate
𝐸 𝐸𝜈
𝜎𝑥 0 0 0
1−𝜈 2 1−𝜈 2
𝜎𝑦 𝐸𝜈 𝐸
0 0 0
𝛔𝑡ℎ𝑖𝑐𝑘 = 𝜏𝑥𝑦 , 𝐂𝑡ℎ𝑖𝑐𝑘
𝑒𝑙𝑎𝑠𝑡𝑖𝑐
= 1−𝜈 2 1−𝜈 2 , 𝛔𝑒𝑙𝑎𝑠𝑡𝑖𝑐 𝑒𝑙𝑎𝑠𝑡𝑖𝑐
𝑡ℎ𝑖𝑐𝑘 = 𝐂𝑡ℎ𝑖𝑐𝑘 𝛆𝑡ℎ𝑖𝑐𝑘 (160)
𝜏𝑥𝑧 0 0 𝐺 0 0
{ 𝜏𝑦𝑧 } 0 0 0 𝐺 0
[ 0 0 0 0 𝐺]
For elastoplastic materials, the constitutive matrix and stress vector are calculated through elastoplastic
iteration as explained in section 6.5.2.

6.4.2.1 Kirchhoff Triangular Plate Element

The Kirchhoff triangular plate element in Collapse Advanced is based on the work presented in reference
[40], in which the plate element uses a modified quadratic shape functions to interpolate rotations
(𝜃𝑥 , 𝜃𝑦 ) and enforces discrete Kirchhoff theory at the element corners and mid-points. The interpolation
is defined as
𝑤𝑗 𝑤𝑗
𝑁𝑗𝑛𝑡 (𝑗) 𝑁𝑗𝑛𝑡 (𝑗)
𝜃𝑥 (𝜉, 𝜂) = ∑𝑗=1 𝐇𝑥 𝜃 𝜃
{ 𝑥𝑗 } , 𝜃𝑦 (𝜉, 𝜂) = ∑𝑗=1 𝐇𝑦 { 𝑥𝑗 } (161)
𝜃𝑦𝑗 𝜃𝑦𝑗

where
(𝑗)
𝐇𝑥 = [𝐻𝑥3(𝑗−1)+1 𝐻𝑥3(𝑗−1)+2 𝐻𝑥3(𝑗−1)+3 ]
(𝑗)
(162)
𝐇𝑦 = [𝐻𝑦3(𝑗−1)+1 𝐻𝑦3(𝑗−1)+2 𝐻𝑦3(𝑗−1)+3 ]
(𝑖) (𝑖)
in which the modified quadratic shape function 𝐻𝑥 and 𝐻𝑦 are represented by equations (163) as
follows:

1st Joint:

208

SACS® Collapse Advanced Release 15.1

 (1)
𝐻𝑥 (𝜉, 𝜂) = −1.5 (𝑑6 𝑁𝑄6 (𝜉, 𝜂) − 𝑑5 𝑁𝑄5 (𝜉, 𝜂))
(2)
𝐻𝑥 (𝜉, 𝜂) = 𝑁𝑄1 (𝜉, 𝜂) − 𝑒5 𝑁𝑄5 (𝜉, 𝜂) − 𝑒6 𝑁𝑄6 (𝜉, 𝜂)
(3)
𝐻𝑥 (𝜉, 𝜂) = 𝑏5 𝑁𝑄5 (𝜉, 𝜂) + 𝑏6 𝑁𝑄6 (𝜉, 𝜂)
(1)
(163)
𝐻𝑦 (𝜉, 𝜂) = 1.5 (𝑎6 𝑁𝑄6 (𝜉, 𝜂) − 𝑎5 𝑁𝑄5 (𝜉, 𝜂))
(2)
𝐻𝑦 (𝜉, 𝜂) = 𝑏5 𝑁𝑄5 (𝜉, 𝜂) + 𝑏6 𝑁𝑄6 (𝜉, 𝜂)
(3)
𝐻𝑦 (𝜉, 𝜂) = 𝑁𝑄1 (𝜉, 𝜂) − 𝑐5 𝑁𝑄5 (𝜉, 𝜂) − 𝑐6 𝑁𝑄6 (𝜉, 𝜂)
2nd Joint:
(4)
𝐻𝑥 (𝜉, 𝜂) = −1.5 (𝑑4 𝑁𝑄4 (𝜉, 𝜂) − 𝑑6 𝑁𝑄6 (𝜉, 𝜂))
(5)
𝐻𝑥 (𝜉, 𝜂) = 𝑁𝑄2 (𝜉, 𝜂) − 𝑒6 𝑁𝑄6 (𝜉, 𝜂) − 𝑒4 𝑁𝑄4 (𝜉, 𝜂)
(6)
𝐻𝑥 (𝜉, 𝜂) = 𝑏6 𝑁𝑄6 (𝜉, 𝜂) + 𝑏4 𝑁𝑄4 (𝜉, 𝜂)
(4)
(164)
𝐻𝑦 (𝜉, 𝜂) = 1.5 (𝑎4 𝑁𝑄4 (𝜉, 𝜂) − 𝑎6 𝑁𝑄6 (𝜉, 𝜂))
(5)
𝐻𝑦 (𝜉, 𝜂) = 𝑏6 𝑁𝑄6 (𝜉, 𝜂) + 𝑏4 𝑁𝑄4 (𝜉, 𝜂)
(6)
𝐻𝑦 (𝜉, 𝜂) = 𝑁𝑄2 (𝜉, 𝜂) − 𝑐6 𝑁𝑄6 (𝜉, 𝜂) − 𝑐4 𝑁𝑄4 (𝜉, 𝜂)

3rd Joint:
(7)
𝐻𝑥 (𝜉, 𝜂) = −1.5 (𝑑5 𝑁𝑄5 (𝜉, 𝜂) − 𝑑4 𝑁𝑄4 (𝜉, 𝜂))
(8)
𝐻𝑥 (𝜉, 𝜂) = 𝑁𝑄3 (𝜉, 𝜂) − 𝑒4 𝑁𝑄4 (𝜉, 𝜂) − 𝑒5 𝑁𝑄5 (𝜉, 𝜂)
(9)
𝐻𝑥 (𝜉, 𝜂) = 𝑏4 𝑁𝑄4 (𝜉, 𝜂) + 𝑏5 𝑁𝑄5 (𝜉, 𝜂)
(7)
(165)
𝐻𝑦 (𝜉, 𝜂) = 1.5 (𝑎5 𝑁𝑄5 (𝜉, 𝜂) − 𝑎4 𝑁𝑄4 (𝜉, 𝜂))
(8)
𝐻𝑦 (𝜉, 𝜂) = 𝑏4 𝑁𝑄4 (𝜉, 𝜂) + 𝑏5 𝑁𝑄5 (𝜉, 𝜂)
(9)
𝐻𝑦 (𝜉, 𝜂) = 𝑁𝑄3 (𝜉, 𝜂) − 𝑐4 𝑁𝑄4 (𝜉, 𝜂) − 𝑐5 𝑁𝑄5 (𝜉, 𝜂)
where 𝑁𝑄 is standard quadratic shape functions
1
𝑁𝑄1 (𝜉, 𝜂) = 2(1 − 𝜉 − 𝜂) (2 − 𝜉 − 𝜂)
𝑁𝑄2 (𝜉, 𝜂) = 𝜉 (2𝜉 − 1)
𝑁𝑄3 (𝜉, 𝜂) = 𝜂(2𝜂 − 1) (166)
𝑁𝑄4 (𝜉, 𝜂) = 4𝜉𝜂
𝑁𝑄5 (𝜉, 𝜂) = 4𝜂(1 − 𝜉 − 𝜂)
𝑁𝑄6 (𝜉, 𝜂) = 4𝜉 (1 − 𝜉 − 𝜂)
and

209

SACS® Collapse Advanced Release 15.1

 𝑥 𝑥 𝑥
𝑎4 = 𝐿223 , 𝑎5 = 𝐿231 , 𝑎4 = 𝐿212
23 31 12
3 𝑥23 𝑦23 3 𝑥31 𝑦31 3 𝑥12 𝑦12
𝑏4 = 4 𝐿223
, 𝑏5 = 4 𝐿231
, 𝑏4 = 4 𝐿212
1 2 1 1 2 1 1 2 1 2
𝑥 − 𝑦2 𝑥 − 𝑦2 𝑥 − 𝑦
4 23 2 23 4 31 2 31 4 12 2 12
𝑐4 = 𝐿223
, 𝑐5 = 𝐿231
, 𝑐4 = 𝐿212
𝑦23 𝑦31 𝑦12
(167)
𝑑 =−
4 𝐿223
, 𝑑5 = − 𝐿231
, 𝑑4 = − 𝐿212
1 2 1 2 1 2 1 2 1 2 1 2
𝑦 − 𝑥 𝑦 − 𝑥 𝑦 − 𝑥
4 23 2 23 4 31 2 31 4 12 2 12
𝑒4 = 𝐿223
, 𝑒5 = 𝐿231
,𝑒 = 𝐿212
𝐿2𝑖𝑗 = 𝑥𝑖𝑗
2 2
+ 𝑦𝑖𝑗

It should be noted that (θx , θy ) are equivalent to (−βy , βx ) in [40].

(𝑗)
Substituting the above expression in equation (136), 𝐁𝑏 (𝜉, 𝜂, 𝑧) may be expressed as:
(𝑗)
𝜕𝐇𝑦
𝑧
𝜕𝑥
(𝑗)
(𝑗) 𝜕𝐇𝑥
𝐁𝑏 (𝜉, 𝜂, 𝑧) = −𝑧 (168)
𝜕𝑦
(𝑗) (𝑗)
𝜕𝐇𝑥 𝜕𝐇𝑦
[−𝑧 𝜕𝑥
+𝑧
𝜕𝑦 ]
(𝑗)
Using above expression for 𝐁𝑏 , the local force vector and stiffness matrix for Kirchhoff triangular plate
element can be obtained from equations (157) and (158).

6.4.2.2 Mindlin Triangular Plate Element

The Collapse Advanced Mindlin Triangular plate element is based on element formulation proposed in
(𝑗)
reference [41] where the bending strain operator 𝐁𝑏 is based on a linear triangular shape functions as
represented by equation (141) which can be expressed as:
(𝑗)
𝜕𝑁𝐿
0 0 𝑧
𝜕𝑥
(𝑗)
(𝑗) 𝜕𝑁𝐿
𝐁𝑏 (𝜉, 𝜂, 𝑧) = 0 −𝑧 0 (169)
𝜕𝑦
(𝑗) (𝑗)
𝜕𝑁𝐿 𝜕𝑁𝐿
[0 −𝑧 𝜕𝑥
𝑧 𝜕𝑦 ]

Note that 𝑤𝑗 does not have any contribution to the plate rotation (𝜃𝑥 , 𝜃𝑦 ). Hence, the first column in
(𝑗)
equation (169) is zero. The shear strain operator 𝐁𝑠 is based on reference [41] and is represented as:
(𝑗) (𝑗) (𝑗)
𝜕𝑁𝐿 𝜕𝐻𝜃𝑥 𝜕𝐻𝜃𝑦 (𝑗)
(𝑗) 𝜕𝑥 𝜕𝑥 𝜕𝑥
+ 𝑁𝐿
𝐁𝑠 (𝜉, 𝜂, 𝑧) = 𝜑𝑓(𝑧) [ (𝑗) (𝑗) (𝑗) ] (170)
𝜕𝑁𝐿 𝜕𝐻𝜃𝑥 (𝑗) 𝜕𝐻𝜃𝑦
𝜕𝑦 𝜕𝑦
−𝑁𝐿 𝜕𝑦

In which 𝜑 is shear-locking correction factor, 𝑓(𝑧) is shear distribution function given in equation (140)
(𝑗) (𝑗)
and 𝐻𝜃𝑥 and 𝐻𝜃𝑦 are modified quadradic shape functions defined below. Note that θx are equivalent to
−θx in [41]. Therefore, re-factoring the shape functions associated with θx yields:

210

SACS® Collapse Advanced Release 15.1

1st Joint:
(1) 1
𝐻𝜃𝑥 (𝜉, 𝜂) = 8 (𝑏2 𝑁𝑄6 (𝜉, 𝜂) − 𝑏3 𝑁𝑄4 (𝜉, 𝜂))
(1) 1
(171)
𝐻𝜃𝑦 (𝜉, 𝜂) = (𝑎2 𝑁𝑄6 (𝜉, 𝜂) − 𝑎3 𝑁𝑄4 (𝜉, 𝜂))
8
2nd Joint:
(2) 1
𝐻𝜃𝑥 (𝜉, 𝜂) = 8 (𝑏3 𝑁𝑄4 (𝜉, 𝜂) − 𝑏1 𝑁𝑄5 (𝜉, 𝜂))
(2) 1
(172)
𝐻𝜃𝑦 (𝜉, 𝜂) = 8 (𝑎3 𝑁𝑄4 (𝜉, 𝜂) − 𝑎1 𝑁𝑄5 (𝜉, 𝜂))
3rd Joint:
(3) 1
𝐻𝜃𝑥 (𝜉, 𝜂) = 8 (𝑏1 𝑁𝑄5 (𝜉, 𝜂) − 𝑏2 𝑁𝑄6 (𝜉, 𝜂))
(3) 1
(173)
𝐻𝜃𝑦 (𝜉, 𝜂) = 8 (𝑎1 𝑁𝑄5 (𝜉, 𝜂) − 𝑎2 𝑁𝑄6 (𝜉, 𝜂))
Where 𝑁𝑄 are triangular standard quadratic shape functions as given in equation (166), and
𝑎1 = 𝑥32 , 𝑎2 = 𝑥13 , 𝑎3 = 𝑥21
𝑏1 = 𝑦23 , 𝑏2 = 𝑦31 , 𝑏3 = 𝑦12 (174)

(𝑗) (𝑗)
Using the bending strain operator 𝐁𝑏 and shear strain operator 𝐁𝑠 , the load vector and stiffness
matrix may be calculated using equations (157) and (158).

The last part of Mindlin triangular plate element is the shear-locking correction. Following reference [41]
the shear-locking correction factor 𝜑 is given by the following equation:
1
𝜑=√ 𝛼 (175)
1+
𝐶

In which 𝛼 is defined as a ratio of the diagonal terms associated with rotational DOFs in shear and
bending when 𝜑 = 1 and is expressed as:
(𝜃𝑥𝑗 ,𝜃𝑦𝑗)
∑𝑑𝑖𝑎𝑔𝑜𝑛𝑎𝑙 𝐊 𝑒𝑙𝑎𝑠𝑡𝑖𝑐
𝑠ℎ𝑒𝑎𝑟 (𝜑=1)
𝛼= (𝜃𝑥𝑗 ,𝜃𝑦𝑗) (176)
∑𝑑𝑖𝑎𝑔𝑜𝑛𝑎𝑙 𝐊 𝑒𝑙𝑎𝑠𝑡𝑖𝑐
𝑏𝑒𝑛𝑑𝑖𝑛𝑔 (𝜑=1)

The parameter 𝐶 is a user-defined parameter which can be defined on the CLPOP2 input line. Reference
[41] recommends an optimal values 2.0 for 𝐶 (which is assumed as the default value in Collapse
Advanced) based on various numerical experiments. The authors have also shown that values of 1.94
and 2.04 can be used for simply supported and clamped plates, respectively.

6.4.2.3 Kirchhoff Quadrilateral Plate Element

The Collapse Advanced Kirchhoff quadrilateral plate element is based on an element proposed by
reference [42] which utilizes modified quadratic shape functions to satisfy Kirchhoff bending theory at
the element nodes and mid-points. The shape functions are given for element joints as follow:

Note: (𝜃𝑥 , 𝜃𝑦 ) are equivalent to (−𝛽𝑦 , 𝛽𝑥 ) in [42] where (𝑤𝑗 , −𝜃𝑦𝑗 , 𝜃𝑥𝑗 ) are used as nodal DOFs.
Therefore, after re-ordering and re-factoring, the shape functions can be expressed at the element joints
as:

1st Joint:

211

SACS® Collapse Advanced Release 15.1

 (1) 𝑡 𝑡4 3𝑥24
𝐻𝑥 (𝜉, 𝜂) = − ൬− 41 𝑁𝑄5 (𝜉, 𝜂) + 4
𝑁𝑄8 (𝜉, 𝜂) + 2∆
𝑁𝑄9 (𝜉, 𝜂)൰
(2) 𝑟1 −2 𝑟4 −2 𝑥24 (𝑦21 +𝑦41 )
𝐻𝑥 (𝜉, 𝜂) = − ൬−𝑁𝑄1 (𝜉, 𝜂) + 4
𝑁𝑄5 (𝜉, 𝜂) + 4
𝑁𝑄8 (𝜉, 𝜂) + 4∆
𝑁𝑄9 (𝜉, 𝜂)൰
(3) 𝑞1 𝑞4 𝑥24 (𝑥21 +𝑥41 )
𝐻𝑥 (𝜉, 𝜂) = − 4
𝑁𝑄5 (𝜉, 𝜂) − 4
𝑁𝑄8 (𝜉, 𝜂) + 4∆
𝑁𝑄9 (𝜉, 𝜂)
(177)
(1) 𝑝1 𝑝4 3𝑦42
𝐻𝑦 (𝜉, 𝜂) = 4
𝑁𝑄5 (𝜉, 𝜂) − 4
𝑁𝑄8 (𝜉, 𝜂) + 2∆
𝑁𝑄9 (𝜉, 𝜂)
(2) 𝑞1 𝑞4 𝑦42 (𝑦21 +𝑦41 )
𝐻𝑦 (𝜉, 𝜂) =− 4
𝑁𝑄5 (𝜉, 𝜂) − 4
𝑁𝑄8 (𝜉, 𝜂) + 4∆
𝑁𝑄9 (𝜉, 𝜂)
(3) 𝑠1−2 𝑠4 −2 𝑦42 (𝑥21 +𝑥41 )
𝐻𝑦 (𝜉, 𝜂) = − ൬−𝑁𝑄1 (𝜉, 𝜂) + 𝑁𝑄5 (𝜉, 𝜂) + 𝑁𝑄8 (𝜉, 𝜂) + 𝑁𝑄9 (𝜉, 𝜂)൰
4 4 4∆

2nd Joint:
(4) 𝑡 𝑡2 3𝑥31
𝐻𝑥 (𝜉, 𝜂) = − ൬− 41 𝑁𝑄5 (𝜉, 𝜂) − 4
𝑁𝑄6 (𝜉, 𝜂) + 2∆
𝑁𝑄9 (𝜉, 𝜂)൰
(5) 𝑟1 −2 𝑟2 −2 𝑥31 (𝑦12 +𝑦32 )
𝐻𝑥 (𝜉, 𝜂) = − ൬−𝑁𝑄2 (𝜉, 𝜂) + 4
𝑁𝑄5 (𝜉, 𝜂) + 4
𝑁𝑄6 (𝜉, 𝜂) + 4∆
𝑁𝑄9 (𝜉, 𝜂)൰
(6) 𝑞1 𝑞2 𝑥31 (𝑥12 +𝑥32 )
𝐻𝑥 (𝜉, 𝜂) = − 4
𝑁𝑄5 (𝜉, 𝜂) − 4
𝑁𝑄6 (𝜉, 𝜂) + 4∆
𝑁𝑄9 (𝜉, 𝜂)
(178)
(4) 𝑝1 𝑝2 3𝑦13
𝐻𝑦 (𝜉, 𝜂) = − 4
𝑁𝑄5 (𝜉, 𝜂) + 4
𝑁𝑄6 (𝜉, 𝜂) + 2∆
𝑁𝑄9 (𝜉, 𝜂)
(5) 𝑞1 𝑞2 𝑦13 (𝑦12 +𝑦32 )
𝐻𝑦 (𝜉, 𝜂) =− 𝑁𝑄5 (𝜉, 𝜂) − 𝑁𝑄6 (𝜉, 𝜂) + 𝑁𝑄9 (𝜉, 𝜂)
4 4 4∆
(6) 𝑠1−2 𝑠2 −2 𝑦13 (𝑥12 +𝑥32 )
𝐻𝑦 (𝜉, 𝜂) = − ൬−𝑁𝑄2 (𝜉, 𝜂) + 𝑁𝑄5 (𝜉, 𝜂) + 𝑁𝑄6 (𝜉, 𝜂) + 𝑁𝑄9 (𝜉, 𝜂)൰
4 4 4∆

3rd Joint:

(7) 𝑡 𝑡3 3𝑥42
𝐻𝑥 (𝜉, 𝜂) = − ൬ 2 𝑁𝑄6 (𝜉, 𝜂) − 𝑁𝑄7 (𝜉, 𝜂) + 𝑁𝑄9 (𝜉, 𝜂)൰
4 4 2∆
(8) 𝑟2 −2 𝑟3 −2 𝑥42 (𝑦23+𝑦43 )
𝐻𝑥 (𝜉, 𝜂) = − ൬−𝑁𝑄3 (𝜉, 𝜂) + 4
𝑁𝑄6 (𝜉, 𝜂) + 4
𝑁𝑄7 (𝜉, 𝜂) + 4∆
𝑁𝑄9 (𝜉, 𝜂)൰
(9) 𝑞2 𝑞3 𝑥42 (𝑥23 +𝑥43 )
𝐻𝑥 (𝜉, 𝜂) = − 4
𝑁𝑄6 (𝜉, 𝜂) − 4
𝑁𝑄7 (𝜉, 𝜂) + 4∆
𝑁𝑄9 (𝜉, 𝜂)
(179)
(7) 𝑝2 𝑝3 3𝑦24
𝐻𝑦 (𝜉, 𝜂) =− 4
𝑁𝑄6 (𝜉, 𝜂) + 4
𝑁𝑄7 (𝜉, 𝜂) + 2∆
𝑁𝑄9 (𝜉, 𝜂)
(8) 𝑞2 𝑞3 𝑦24 (𝑦23 +𝑦43 )
𝐻𝑦 (𝜉, 𝜂) =− 𝑁𝑄6 (𝜉, 𝜂) − 𝑁𝑄7 (𝜉, 𝜂) + 𝑁𝑄9 (𝜉, 𝜂)
4 4 4∆
(9) 𝑠2−2 𝑠3 −2 𝑦24 (𝑥23 +𝑥43 )
𝐻𝑦 (𝜉, 𝜂) = − ൬−𝑁𝑄3 (𝜉, 𝜂) + 4
𝑁𝑄6 (𝜉, 𝜂) + 4
𝑁𝑄7 (𝜉, 𝜂) + 4∆
𝑁𝑄9 (𝜉, 𝜂)൰

4th Joint:

212

SACS® Collapse Advanced Release 15.1

 (10) 𝑡 𝑡4 3𝑥13
𝐻𝑥 (𝜉, 𝜂) = − ൬ 3 𝑁𝑄7 (𝜉, 𝜂) − 𝑁𝑄8 (𝜉, 𝜂) + 𝑁𝑄9 (𝜉, 𝜂)൰
4 4 2∆
(11) 𝑟3−2 𝑟4 −2 𝑥13 (𝑦14 +𝑦34 )
𝐻𝑥 (𝜉, 𝜂) = − ൬−𝑁𝑄4 (𝜉, 𝜂) + 𝑁𝑄7 (𝜉, 𝜂) + 𝑁𝑄8 (𝜉, 𝜂) + 𝑁𝑄9 (𝜉, 𝜂)൰
4 4 4∆
(12) 𝑞3 𝑞4 𝑥13 (𝑥14 +𝑥34 )
𝐻𝑥 (𝜉, 𝜂) = − 𝑁𝑄7 (𝜉, 𝜂) − 𝑁𝑄8 (𝜉, 𝜂) + 𝑁𝑄9 (𝜉, 𝜂)
4 4 4∆ (180)
(10) 𝑝3 𝑝4 3𝑦31
𝐻𝑦 (𝜉, 𝜂) = − 𝑁𝑄7 (𝜉, 𝜂) + 𝑁𝑄8 (𝜉, 𝜂) + 𝑁𝑄9 (𝜉, 𝜂)
4 4 2∆
(11) 𝑞3 𝑞4 𝑦31 (𝑦14 +𝑦34 )
𝐻𝑦 (𝜉, 𝜂) =− 4
𝑁𝑄7 (𝜉, 𝜂) − 4
𝑁𝑄8 (𝜉, 𝜂) + 4∆
𝑁𝑄9 (𝜉, 𝜂)
(12) 𝑠3−2 𝑠4−2 𝑦31 (𝑥14 +𝑥34 )
𝐻𝑦 (𝜉, 𝜂) = − ൬−𝑁𝑄4 (𝜉, 𝜂) + 𝑁𝑄7 (𝜉, 𝜂) + 𝑁𝑄8 (𝜉, 𝜂) + 𝑁𝑄9 (𝜉, 𝜂)൰
4 4 4∆

where 𝑁𝑄 is the least-squares polynomial shape functions for a quadrilateral plate element given in [42]
as follows:
1 1 1 1 1 1
𝑁𝑄1 (𝜉, 𝜂) = − 36 + 12 𝜉 2 + 4 𝜉𝜂 + 12 𝜂2 − 4 𝜉 2 𝜂 − 4 𝜉𝜂2
1 1 2 1 2 1 1 1
𝑁𝑄2 (𝜉, 𝜂) = − + 𝜉 − 𝜂 − 𝜉2 𝜂 + 𝜉𝜂 + 𝜉𝜂2
36 12 12 4 4 4
1 1 2 1 2 1 1 1
𝑁𝑄3 (𝜉, 𝜂) = 𝜉𝜂 + 12 𝜂 + 4 𝜉 2 𝜂 +
− 36 + 12 𝜉 +
4 4
𝜉𝜂2
1 11 1 1 1
𝑁𝑄4 (𝜉, 𝜂) = − + 𝜉2 − 𝜉𝜂 + 𝜂2 + 𝜉 2 𝜂 − 𝜉𝜂2
36 12
4 12 4 4
1 1 1 1 1
𝑁𝑄5 (𝜉, 𝜂) = − 𝜂 − 𝜉 2 + 𝜂2 + 𝜉 2 𝜂 (181)
18 2 6 3 2
1 1 1 2 1 2 1
𝑁𝑄6 𝜉, 𝜂 = + 𝜉 + 𝜉 − 𝜂 − 𝜉𝜂2
( )
18 2 3 6 2
1 1 1 2 1 2 1
𝑁𝑄7 (𝜉, 𝜂) = + 𝜂 − 𝜉 + 𝜂 − 𝜉 2 𝜂
18 2 6 3 2
1 1 1 2 1 2 1
𝑁𝑄8 (𝜉, 𝜂) = 18 − 2 𝜉 + 3 𝜉 − 6 𝜂 + 2 𝜉𝜂2
8 2 2
𝑁𝑄9 (𝜉, 𝜂) = 9 − 3 𝜉 2 − 3 𝜂2
where
𝑏1 = 𝑦12 , 𝑏2 = 𝑦23 , 𝑏3 = 𝑦34 , 𝑏4 = 𝑦41
𝑐1 = 𝑥21 , 𝑐2 = 𝑥32 , 𝑐3 = 𝑥43 , 𝑐4 = 𝑥14
∆= 𝑋1 𝑦24 + 𝑋2 𝑦31 + 𝑋3 𝑦42 +𝑋4 𝑦13
6𝑐𝑖 3𝑏𝑖 𝑐𝑖 3𝑏𝑖2 3𝑐𝑖2 6𝑏𝑖
(182)
𝑝𝑖 = 𝐿2𝑖
, 𝑞𝑖 = 𝐿2𝑖
, 𝑟𝑖 = 𝐿2𝑖
, 𝑠𝑖 = 𝐿2𝑖
, 𝑡𝑖 = 𝐿2𝑖
𝐿2𝑖 = 𝑏𝑖2 + 𝑐𝑖2

Replacing the above shape functions with the equations (157) and (158), the force vector and stiffness
matrix for a Kirchhoff quadrilateral plate element can be calculated.
6.4.2.4 Mindlin Quadrilateral Plate Element

The Mindlin quadrilateral plate element is based on a work presented in reference [43] in which the
(𝑗)
bending strain operator 𝐁𝑏 is similar to the Mindlin triangular plate element as given equation (169)
(𝑗)
with bilinear quadrilateral shape functions. The shear strain operator 𝐁𝑠 is defined such that shear-
locking is prevented (i.e. no additional correction factor required) [43]. The details are presented below.

213

SACS® Collapse Advanced Release 15.1

 (𝑗)
𝐁𝑠 (𝜉, 𝜂, 𝑧) = 𝑓(𝑧)[𝐛(𝑗)
𝑠1
(𝑗)
𝐛𝑠2
(𝑗)
𝐛𝑠3 ]
(𝑗) 𝐠 (𝑝)(𝜉,𝜂) 𝐠 (𝑗)(𝜉,𝜂)
𝐛𝑠1 (𝜉, 𝜂) = ℎ (𝑝)
− ℎ (𝑗)
(𝑗) 1 (𝑗) (𝑗) (183)
𝐛𝑠2 (𝜉, 𝜂) = ൬𝑏2 𝐠 (𝑝) (𝜉, 𝜂) − 𝑎2 𝐠 (𝑗) (𝜉, 𝜂)൰
2
(𝑗) 1 (𝑗) (𝑗)
𝒃𝑠3 (𝜉, 𝜂) = − ൬𝑏1 𝐠 (𝑝) (𝜉, 𝜂) − 𝑎1 𝐠 (𝑗) (𝜉, 𝜂)൰
2
where 𝐠 (𝑗) (𝜉, 𝜂) are
(𝑗) (𝑗) (𝑗) (𝑛) (𝑛)
𝑁𝐿 (𝜉,𝜂) 𝑎 𝑏 (𝑛)
𝑁𝐿 (𝜉,𝜂) 𝑏 𝑎
𝐠 𝜉, 𝜂) = 1−(𝛼(𝑗) )2 ({ 1(𝑗) } − 𝛼 (𝑗) { 1(𝑗) }) − 1−(𝛼
(𝑗) (
(𝑛))2
({ 1(𝑛) } − 𝛼 (𝑛) { 1(𝑛) }) (184)
𝑎2 𝑏2 𝑏2 𝑎2

with
𝑋(𝑛) −𝑋(𝑗) 𝑋(𝑝) −𝑋(𝑗)
(𝑗) { } { (𝑗) }
𝑎 𝑌 −𝑌(𝑗) 𝑏1 𝑌(𝑝) −𝑌(𝑗)
{ 1(𝑗) } = 𝑋(𝑛)
(𝑛) −𝑋(𝑗)
, { (𝑗)
} = 𝑋(𝑝) −𝑋(𝑗)
𝑎2 |{
𝑌(𝑛) −𝑌(𝑗)
}| 𝑏2 |{
𝑌(𝑝) −𝑌(𝑗)
}|
(185)
(𝑗) 2 2
ℎ = √(𝑋(𝑛) − 𝑋(𝑗) ) + (𝑌(𝑛) − 𝑌(𝑗) )
(𝑗) (𝑗) (𝑗) (𝑗)
𝛼 (𝑗) = 𝑎1 𝑏1 + 𝑎2 𝑏2

the indices (𝑗, 𝑝, 𝑛) are defined as follows:

𝑗 = 1, 𝑝 = 4, 𝑛 = 2
𝑗 = 2, 𝑝 = 1, 𝑛 = 3
(186)
𝑗 = 3, 𝑝 = 2, 𝑛 = 4
𝑗 = 4, 𝑝 = 3, 𝑛 = 1
As before, the force vector and stiffness matrix can be calculated using equations (157) and (158).

214

SACS® Collapse Advanced Release 15.1

6.4.3 Plate Internal Forces and Moments
The plate internal forces per unit width can be calculated at a given point (𝑥, 𝑦) within the plate element
by integrating the stress over the plate thickness (i.e. 𝑧) and can be expressed as:

Membrane force:
𝑡 ⁄2 𝑡 ⁄2 𝑡 ⁄2
𝑇𝑥 = ∫−𝑡 ⁄2 𝜎𝑥 𝑑𝑧 , 𝑇𝑦 = ∫−𝑡 ⁄2 𝜎𝑦 𝑑𝑧 , 𝑇𝑥𝑦 = ∫−𝑡 ⁄2 𝜏𝑥𝑦 𝑑𝑧 + 𝜏𝑑𝑟𝑖𝑙𝑙 𝑡 (187)

Bending internal moments:

𝑡 ⁄2 𝑡 ⁄2 𝑡 ⁄2
𝑀𝑥 = ∫−𝑡 ⁄2 −𝜎𝑥 𝑧𝑑𝑧 , 𝑀𝑦 = ∫−𝑡 ⁄2 −𝜎𝑦 𝑧𝑑𝑧 , 𝑀𝑥𝑦 = ∫−𝑡 ⁄2 −𝜏𝑥𝑦 𝑧𝑑𝑧 (188)

Mindlin out-of-plane shear forces:

𝑡 ⁄2 𝑡 ⁄2
𝑄𝑥 = ∫−𝑡 ⁄2 𝑓(𝑧)𝜏𝑥𝑧 𝑑𝑧 , 𝑄𝑦 = ∫−𝑡 ⁄2 𝑓 (𝑧)𝜏𝑦𝑧 𝑑𝑧

Note: The drilling stress is constant over plate thickness and does not produce any moment. (189)

6.4.4 Plate Element Numerical Integration

As mentioned in the previous section, the local force vector and stiffness matrix can be derived by
integrating over plate volume for an isoparametric system. The integral can be rewritten in general form
as:
𝑡 𝑀 𝑁 (𝜉,𝜂)
𝑝
𝐼 = ∭(𝜉,𝜂,𝑡) 𝑓(𝜉, 𝜂, 𝑧)|𝐉(𝜉, 𝜂)|𝑑𝜉𝑑𝜂𝑑𝑧 = 2 ∑𝑖=1 𝑝
∑𝑗=1 𝑓(𝜉𝑖 , 𝜂𝑖 , 𝑧𝑗 )|𝐉(𝜉𝑖 , 𝜂𝑖 )|𝑊𝑖 𝑊𝑗𝑧 (190)

Collapse Advanced utilizes a quadrature integration rule to calculate integrals. For triangular plate
elements, the coordinates of integration point within plate are given by the following expressions:

1 2 1 1 1 2 (𝜉,𝜂) 1 1 1
𝜉 = (6 , 3 , 6) , 𝜂 = (6 , 6 , 3) , 𝑊 = (6 , 6 , 6) (191)

For quadrilateral plate elements, 2×2 Gauss–Legendre points are used over the plate element as defined
below:

1 1 1 1 1 1 1 1
𝜉 = (− , , ,− ) , 𝜂 = (− ,− , , ) , 𝑊 (𝜉,𝜂) = (1,1,1,1) (192)
√3 √3 √3 √3 √3 √3 √3 √3

Both elements types utilize 6 Gauss–Legendre points to monitor elastic plastic stress variation over plate
thickness. Therefore, elastoplastic calculations are carried out at 18 and 24 points for a triangular plate
and quadrilateral plate elements, respectively.

215

SACS® Collapse Advanced Release 15.1

6.5 Elastoplastic Material
The Collapse Advanced elastoplastic analysis is based on a J2 yield surface and with constant isotropic
hardening. It utilizes backward Euler method to iteratively determine elastoplastic stress vector and
consistent constitutive matrix at all integration points. The formulation is presented in the current
section and is based on work presented in reference [1].

The following notation is used to express elastoplastic formulation:

f J2 yield surface.
𝐚 normal to the yield surface
𝛔0 , 𝛔 initial and current stress vectors
∆𝛆 strain vector increment
∆𝛔, ∆𝛔𝑒𝑙 , ∆𝛔𝑝 total stress increment, elastic portion of stress increment, plastic portion of stress
increment
𝛿𝛔 iterative change in stress increment
𝜀𝑝𝑠0 , 𝜀𝑝𝑠 , ∆𝜀𝑝𝑠 initial and current plastic strain, and plastic strain increment
𝛿𝜀𝑝𝑠 iterative change in plastic strain increment
𝐸, 𝐺, 𝜈 elastic modulus (Young’s modulus), shear modulus, and Poisson’s ratio
𝐹𝑦 yield stress
𝐸𝑇
𝐸𝑇 , 𝜌 slope of stress-strain curve after the yield stress and 𝜌 = 𝐸
𝜌𝐸
𝐻 strain hardening parameter 𝐻 = 1−𝜌
𝜎, 𝜎𝑥 , 𝜎𝑦 , 𝜎𝑧 normal stresses
𝜎𝑒𝑞 equivalent (von Misses) stress
𝐀 effective stress operator
𝜏, 𝜏𝑥𝑦 , 𝜏𝑥𝑧 , 𝜏𝑦𝑧 shear stresses
𝐂, 𝐂𝑡 elastic constitutive matrix and elastoplastic consistent constitutive matrix
𝐫𝟏 , 𝑟2 residual for backward Euler method
𝑇 matrix transpose operator
𝐈 identity matrix

To define the J2 yield surface, the stress-strain relationship is defined as:

𝑓(𝛔, 𝜀𝑝𝑠 ) = 𝜎𝑒𝑞 (𝛔) − (𝐹𝑦 + 𝐻𝜀𝑝𝑠 ) (193)

where the equivalent stress is given as:

𝜎𝑒𝑞 (𝛔) = √𝛔𝑇 𝐀𝛔 (194)

Where 𝛔 and 𝐀 are given based on element type as below:

Euler-Bernoulli element: 𝐀 = [1] , 𝐂 = [𝐸] , 𝛔 = {𝜎}

1 0 𝐸 0 𝜏}𝑇
Timoshenko element: 𝐀=[ ] ,𝐂 = [ ] , 𝛔 = {𝜎
0 3 0 𝐺

216

SACS® Collapse Advanced Release 15.1

 𝐸 𝐸𝜈
1 −0.5 0 0 𝜎𝑥
1−𝜈 2 1−𝜈 2
Kirchhoff plate: 𝐀 = [−0.5 0] , 𝐂 = [ 𝐸𝜈 𝜎
0] , 𝛔 = { 𝑦 }
1 𝐸
1
0 0 3 1−𝜈 2 1−𝜈 2 𝜏𝑥𝑦
0 0 𝐺

𝐸 𝐸𝜈
1 −0.5 0 0 0 1−𝜈 2 1−𝜈 2
0 0 0 𝜎𝑥
−0.5 1 0 0 0 𝐸𝜈 𝐸 𝜎𝑦
0 0 0
Mindlin plate: 𝐀= 0 0 3 0 0 ,𝐂 = 1−𝜈 2 1−𝜈 2 𝜏
, 𝛔 = 𝑥𝑦
0 0 0 3 0 0 0 𝐺 0 0 𝜏𝑥𝑧
[ 0 0 0 0 3] 0 0 0 𝐺 0 𝜏
{ 𝑦𝑧 }
[ 0 0 0 0 𝐺]

The backward Euler method requires the derivatives of the yield surface with respect to stress 𝛔 and
plastic strain 𝜀𝑝𝑠 as follows:
𝜕𝑓 1
𝐚 = 𝜕𝛔 = 𝜎 𝐀 (195)
𝑒 (𝛔)

𝜕𝐚 𝜕2 𝑓 1 1 𝜕𝑓 𝜕𝑓 𝑇
= =𝜎 𝐀 − ( ) ( ) (196)
𝜕𝛔 𝜕𝛔2 𝑒 (𝛔) 𝜎𝑒 (𝛔) 𝜕𝛔 𝜕𝛔

𝜕𝑓
= −𝐻 (197)
𝜕𝜀𝑝𝑠

If the stress state is still inside the yield surface (i.e. 𝑓(𝛔0 + 𝐂∆𝛆, 𝜀𝑝𝑠0 ) ≤ 0), the elastic stress
increments are ∆𝛔 = ∆𝛔𝑒𝑙 = 𝐂∆𝛆 and plastic strain remains unchanged. For the condition where the
stress state crosses the yield surface or 𝑓(𝛔0 + 𝐂∆𝛆, 𝜀𝑝𝑠0 ) > 0, the crossing point needs to be
determined. Following [1], ∆𝛔𝑒𝑙 can be obtained by solving the following quadratic equation:

(∆𝛆𝑇 𝐂𝑇 𝐀𝐂∆𝛆)𝛼 2 + (𝛔0 𝑇 𝐀𝐂∆𝛆 + ∆𝛆𝑇 𝐂𝑇 𝐀𝛔0 )𝛼 + 𝛔0 𝑇 𝐀𝛔0 − (𝐹𝑦 + 𝐻𝜀𝑝𝑠 ) = 0 (198)

Assuming 𝛼1 and 𝛼2 are the solutions to the above equation the following relationship can be expressed:

∆𝛔𝑒𝑙 = 𝑚𝑎𝑥(𝛼1 , 𝛼2 )𝐂∆𝛆 (199)

Note for an Euler-Bernoulli beam element, the above formulation can be simplified to scalar equations.

6.5.1 Elastoplastic for Beam Element

Collapse Advanced uses two different approaches to determine ∆𝛔, plastic strain 𝜀𝑝𝑠 and elastoplastic
consistent constitutive matrix 𝐂𝑡 for both beam and plate elements as described in this section.

The main objective of elastoplastic calculation is to determine plastic strain 𝜀𝑝𝑠 and stress increment ∆𝛔
at integration points while enforcing following conditions

i. The stress point should remain on the yield surface

ii. The plastic strain rate should be normal to the yield surface (Prandtl-Reuss flow rules)
iii. The stress increment should be related to the elastic portion of the strain increment via the
elastic constitutive matrix 𝐂.

Condition (i) may be represented as follows:

217

SACS® Collapse Advanced Release 15.1

 𝑓(𝛔0 + ∆𝛔, 𝜀𝑝𝑠0 + ∆𝜀𝑝𝑠 ) = 0 (200)

Combining conditions (ii) and (iii) leads to the following equation:

𝐫𝟏 = ∆𝛔 + 𝐂 ∆𝜀𝑝𝑠 𝐚 − ∆𝛔𝑝 = 𝟎 (201)

where the yield surface normal 𝐚 is already defined in equation (195) and the plastic stress increment
∆𝛔𝑝 is defined as:

∆𝛔𝑝 = 𝐂∆𝛆 − ∆𝛔𝑒𝑙 (202)

Equations (200) and (201) can be solved iteratively by following the steps outlined in reference [1].

To begin the procedure, assume ∆𝛔 = 𝟎 , ∆𝜀𝑝𝑠 = 0, followed by the steps outlined below:

1. Calculate the new values for the stress vector 𝛔 = 𝛔0 + ∆𝛔 and the plastic strain 𝜀𝑝𝑠 = 𝜀𝑝𝑠0 +
∆𝜀𝑝𝑠
𝜕𝐚
2. Calculate the new values for the yield surface 𝑓, the residual vector 𝐫𝟏 , derivative of 𝐚 , 𝜕𝛔
3. Determine the iterative change in the plastic strain 𝛿𝜀𝑝𝑠 and the iterative change in stress 𝛿𝛔
using the following equations [1]:
𝑓−𝐚𝑇 𝐐−𝟏 𝐫
𝛿𝜀𝑝𝑠 = 𝐚𝑇 𝐐−𝟏 𝐂𝐚+𝐻𝟏 , 𝛿𝛔 = −𝐐−𝟏 (𝐫𝟏 + 𝛿𝜀𝑝𝑠 𝐂𝐚) (203)

where
𝜕𝐚
𝐐 = 𝐈 + 𝐂 ∆𝜀𝑝𝑠 𝜕𝛔 (204)

Note: The above equations are derived by applying the Newton-Raphson method and
differentiating equation (201) – for detail see reference [1].
4. Update the stress and strain incremental values: ∆𝜀𝑝𝑠 = ∆𝜀𝑝𝑠 + 𝛿𝜀𝑝𝑠 and ∆𝛔 = ∆𝛔 + 𝛿𝛔
5. Repeat the above step until the convergence is achieved (𝐫𝟏 = 𝟎)

Last step is to determine the consistent constitutive matrix 𝐂𝑡 as [1]

𝐑 𝐚 𝐚𝑇 𝐑𝑇
𝐂𝑡 = 𝐑 − 𝐚𝑇 𝐑 𝐚+𝐻 (205)

with 𝐑 = 𝐐−𝟏 𝐂

To improve the convergence rate, the stress increment ∆𝛔𝑝 in equation (201) can be applied over
multiple steps by using a sub-incrementation scheme to improve the efficiency of elastoplastic
calculations. In addition, the equations for the Euler-Bernoulli element, above equations can be reduced
to a scalar expression to improve computational performance.

218

SACS® Collapse Advanced Release 15.1

6.5.2 Elastoplastic for Plate Element
The previous section discussed the backward Euler method for a beam element. For plane stress
conditions (i.e. plate element with 𝜎𝑧 = 0), it has been shown that the backward Euler method has
difficulty in achieving convergence [1]. To resolve this issue for plate elements, the residual function is
2
rewritten in form of (𝜎𝑒𝑞 (𝛔)) − (𝐹𝑦 + 𝐻𝜀𝑝𝑠 )2 and plane stress condition (𝜎𝑧 = 0) are enforced to get
the equivalent stress. The scalar residual function 𝑟2 [1] can be defined as:
1 𝐶 𝐶 2
′
𝑟2 = 4 (𝐴12 + 𝐴22 ) − (𝐹𝑦 + 𝐻𝜀𝑝𝑠 + 𝐻∆𝜀𝑝𝑠 𝜎𝑒𝑞 (𝛔)) = 0 (206)
1 2

′
∆𝜀𝑝𝑠
′
where ∆𝜀𝑝𝑠 is the normalized plastic strain increment 𝜎 and 𝐴1 , 𝐴2 are as below:
𝑒𝑞 (𝛔)
′ ′ 1+𝜈
𝐴1 = 1 + ∆𝜀𝑝𝑠 𝐺 1−𝜈 , 𝐴2 = 1 + 3∆𝜀𝑝𝑠 𝐺 (207)

To determine parameters 𝐶1 and 𝐶2 , define 𝛔𝐵 = 𝛔0 + 𝐂∆𝛆 which in turn leads to the following
expressions for 𝐶1 and 𝛔 for elements based upon Mindlin plate bending theory:
2 2 2 2 2
𝐶1 = (𝜎𝐵𝑥 + 𝜎𝐵𝑦 ) , 𝐶2 = (𝜎𝐵𝑥 − 𝜎𝐵𝑦 ) + 12(𝜏𝐵𝑥𝑦 + 𝜏𝐵𝑥𝑧 + 𝜏𝐵𝑦𝑧 ) (208)
1 1 1 1
2𝐴1
+ 2𝐴 2𝐴1
− 2𝐴 0 0 0
2 2
1 1 1 1
− + 0 0 0
2𝐴1 2𝐴2 2𝐴1 2𝐴2
1
𝛔= 0 0 2𝐴2
0 0 𝛔𝐵 (209)
1
0 0 0 2𝐴2
0
1
[ 0 0 0 0
2𝐴2 ]

For Kirchhoff thin plate bending theory assume transverse shears, 𝜏𝑥𝑧 , 𝜏𝑦𝑧 , and corresponding strains
are zero.
′
Equation (206) can be solved iteratively for ∆𝜀𝑝𝑠 using the Newton-Raphson method. It is worth noting
′
that the stress vector can be immediately calculated as a function of ∆𝜀𝑝𝑠 through equation (209). The
′
iterative change in ∆𝜀𝑝𝑠 is given by the following equation:
−1
′ 𝜕𝑟
δ𝜀𝑝𝑠 = − ൬𝜕∆𝜀2′ ൰ 𝑟2 (210)
𝑝𝑠

where
1+𝜈
𝜕𝑟2 𝐺 𝐶1 3𝐶2
1−𝜈 ′
′ =− ( + ) − 2 (𝐹𝑦 + 𝐻𝜀𝑝𝑠 + 𝐻∆𝜀𝑝𝑠 𝜎𝑒𝑞 (𝛔)) 𝐻𝜎𝑒𝑞 (𝛔) (211)
𝜕∆𝜀𝑝𝑠 2 𝐴31 𝐴32

Note: the elastoplastic consistent constitutive matrix 𝑪𝑡 is determined using same equations as per beam
element. Note also, Collapse Advanced uses a sub-incrementation method to improve convergence rate
by gradually applying 𝝈𝐵 .

219

SACS® Collapse Advanced Release 15.1

6.5.3 Numerical Samples for Elastoplastic Calculation
In this section, a few numerical samples for elastoplastic calculations are shown. Two strain time history
functions are considered, monotonic and harmonic, to determine the stress-strain response history. The
stress-strain curves are shown in Figure 194 to Figure 197.

Figure 194: Elastoplastic response for Euler-Bernoulli beam element

Figure 195: Elastoplastic response for Timoshenko beam element

220

SACS® Collapse Advanced Release 15.1

 Figure 196: Elastoplastic response for Kirchhoff (thin) plate element

221

SACS® Collapse Advanced Release 15.1

 Figure 197: Elastoplastic response for Mindlin (thick) plate element

222

SACS® Collapse Advanced Release 15.1

6.6 Arc-length Method
6.6.1 Introduction
The arc-length iterative approach implemented in Collapse Advanced is based on Section 9.3 of
reference [1]. The arc-length iterations are intended to guide a solution past unstable ‘limit points’
where the determinant of the tangent stiffness matrix becomes momentarily zero followed by the ‘post-
buckling’ region where the structure will continue to deform at decreased load levels.

Under normal load incrementation, a solution method will attempt to calculate the deflections
associated with a set of advancing (increasing or decreasing) load factors that are applied to a fixed or
reference load vector. For normal load incrementation, the load factor is essentially an input.

However, under arc-length control, the incremental load factor is now an output. More accurately, the
objective of the method is to calculate a load factor increment and an associated equilibrium state that
results in a force or deflection increment that is in some way constrained by a fixed ‘arc-length’. There
are many methods to stipulate how the arc-length constraint should be calculated. Collapse Advanced
uses two of them, namely the cylindrical and spherical methods. The arc-length itself is calculated
automatically but it can be scaled by a user-defined factor.
The following notation is used to describe the arc-length approach:
𝐝̇𝐫𝐞𝐟 The converged reference deflection increment calculated for a previous load increment. It is
calculated for the free DOF’s.
𝐝̇ The current load condition’s (un-factored) specified deflection increment. It is calculated for the
fixed DOF’s.
𝐟̇ The current load condition’s (un-factored) applied force increment. It is calculated for the free
DOF’s.
𝐝𝐫,𝟎 The initial deflection at the onset of the load increment, calculated for the fixed DOF’s.
𝐝𝐫 The current (iterative) deflection, calculated for the fixed DOF’s.
𝐝𝐟,𝟎 The initial deflection at the onset of the load increment, calculated for the free DOF’s.
𝐝𝐟 The current (iterative) deflection, calculated for the free DOF’s.
𝐝𝟎 The initial deflection at the onset of the load increment, calculated for all DOF’s.
𝐝 The current (iterative) deflection, calculated for all DOF’s.
𝐟𝟎 The initial applied load at the onset of the increment, calculated for the free DOF’s.
𝐟 The current (iterative) applied load, calculated for the free DOF’s.
Δλ Total, (incremental) load factor increment.
δλ Iterative change in load factor.
α arc-length parameter.

Quantities for free and fixed DOF’s are occasionally put together in vector form, for example,
𝐝
displacement vector can be written as 𝐝 = { 𝐟 }
𝐝𝐫

223

SACS® Collapse Advanced Release 15.1

6.6.2 The Cylindrical Arc-Length Method
The objective of the analysis is to establish a load factor increment, Δλ such that at the end of the load
increment three criteria involving the quantities defined in the previous section are met.

(i) On convergence, the specified deflections and applied loads are given by,

𝐝𝐫 = 𝐝𝐫,𝟎 + Δλ𝐝̇
(212)
𝐟 = 𝐟𝟎 + Δλ𝐟̇

(ii) The converged incremental deflections at the free DOF’s are equal to the pre-
calculated/specified arc-length, α:

‖𝚫𝐝𝐟 ‖ = α ⟹ 𝚫𝐝𝐟 . 𝚫𝐝𝐟 = 𝚫𝐝𝟐𝐟 = α2 (213)

(iii) The solution should go forwards (minimum angle criterion);

𝚫𝐝𝐟 (Δλ). 𝐝̇𝐫𝐞𝐟 > 𝚫𝐝𝐟 (Δλ̃). 𝐝̇𝐫𝐞𝐟 (214)

In which Δλ̃ is any other candidate load factor change that satisfies (i) and (ii).

The starting point is to evaluate the tangent stiffness matrix, 𝐊 at the initial deflection state, 𝐝𝟎 .

𝐊 = 𝐊|𝐝𝟎 (215)

The tangent stiffness matrix should be evaluated such that the plasticity state for each sub-area or plate
layer remains unchanged. In other words, an integration point that is plastic should remain plastic and
not unload, and an integration point that is elastic should not become plastic.

To develop the equations further, the tangent stiffness matrix can be partitioned as follows:
𝐊 𝐟𝐟 𝐊 𝐟𝐫
𝐊= [ ] (216)
𝐊 𝐭𝐟𝐫 𝐊 𝐫𝐫

The internal resisting load due to a small change in the free DOF deflections, 𝛅𝐝𝐟 together with a small
change in the specified deflections, 𝛅𝐝𝐫 should be balanced by a small change in the applied loads at the
free DOF’s, 𝛅𝐟:

𝛅𝐟 = 𝐊 𝐟𝐟 𝛅𝐝𝐟 + 𝐊 𝐟𝐫 𝛅𝐝𝐫 ⟹ 𝛅𝐝𝐟 = 𝐊 −𝟏

𝐟𝐟 (𝛅𝐟 − 𝐊 𝐟𝐫 𝛅𝐝𝐫 ) (217)

If intermediate loading on a member is present, the above expression can be written as;
n1 n2
𝛅𝐟 = 𝐊 𝐟𝐟 𝛅𝐝𝐟 + 𝐊 𝐟𝐫 𝛅𝐝𝐫 + ∑l=1 𝛅𝐟𝐥𝐢 + ∑m=1 𝛅𝐟𝐦𝐢 (218)

In which n1 and n2 denote the number of elements that connect at end-1 and end-2 of a given element
respectively, and the expressions 𝛅𝐟𝐥𝐢 and 𝛅𝐟𝐦𝐢 denote the un-factored equivalent nodal load
increment for each connecting element at end-1 and end-2 respectively.

The iterative changes in the applied loads and specified deflections can written in terms of the initial
load factor increment, δλ. Noting that (by definition),

224

SACS® Collapse Advanced Release 15.1

 𝛅𝐟 = δλ 𝐟̇
(219)
𝛅𝐝𝐫 = δλ 𝐝̇
Using the above two equations, 𝛅𝐝𝐟 can be written as:

𝛅𝐝𝐟 = δλ 𝐊 −𝟏 ̇ ̇ ̃
𝐟𝐟 (𝐟 − 𝐊 𝐟𝐫 𝐝) = δλ 𝐝𝐟 (220)

In which the free DOF predictor deflection, 𝐝̃𝐟 can be expressed in terms of the un-factored deflections
and applied loads as follows:

𝐝̃𝐟 = 𝐊 −𝟏 ̇ ̇
𝐟𝐟 (𝐟 − 𝐊 𝐟𝐫 𝐝) (221)

The cylindrical arc-length technique requires (initially) that the modulus of the change in free DOF’s is
equal to the arc-length, α.

𝛅𝐝𝟐𝐟 = δλ2 (𝐝̃𝐟 . 𝐝̃𝐟 ) = α2 (222)

There are two potential solutions for δλ, namely;

α2
δλ = ±√𝐝̃𝟐 (223)
𝐟

which can be denoted, δλ1 and δλ2 .

Choose δλ using the minimum angle criterion (criterion III), which entails maximizing the scalar product:

δλ (𝐝̃𝐟 . 𝐝̇𝐫𝐞𝐟 ) = max{ δλ1 𝐝̃𝐟 . 𝐝̇𝐫𝐞𝐟 , δλ2 𝐝̃𝐟 . 𝐝̇𝐫𝐞𝐟 } (224)

At the onset of the procedure, the total incremental load factor increment is initialized to the iterative
load factor increment as derived above – i.e. Δλ = δλ. Similarly, the iterative deflections are also
initialized:

𝛅𝐝 𝐝̃
𝐝 = 𝐝𝟎 + { 𝐟 } = 𝐝𝟎 + δλ { 𝐟 } (225)
𝛅𝐝𝐫 𝐝̇
The next phase of the procedure is aimed at finding a value for Δλ such that equilibrium is attained. This
will involve an iterative procedure. The deflection state defined by 𝐝 will rarely attain equilibrium.
Equilibrium must be attained whilst simultaneously ensuring that the deflections at the free DOF’s
adhere to the arc-length specification, which under cylindrical arc-length control amounts to:

‖𝐝𝐟 − 𝐝𝐟,𝟎 ‖ = α (226)

A residual, 𝐫 is defined to represent the degree to which the applied forces are not balanced with the
internal loads, 𝐪𝐢 associated with the current deflection state, 𝐝. Thus,

𝐫 = 𝐟𝟎 + Δλ𝐟̇ − 𝐪𝐢 |𝐝 (227)

Note: In the above expression, the meaning of ‘internal load’ is the load that needs to be applied to
produce a material deflection, as opposed to the equal and opposite material ‘resisting’ force.

The objective of the procedure is to minimize 𝐫, which varies with respect to 𝐝𝐟 and Δλ and therefore
the incremental form can be derived:

225

SACS® Collapse Advanced Release 15.1

 ∂𝐫 ∂𝐫
𝛅𝐫 = ∂λ δλ + ∂𝐝𝐟
𝛅𝐝𝐟 (228)

With the aid of the tangent stiffness matrix:

∂𝐫 ∂𝐪
∂𝐝𝐟
= − ∂𝐝 𝐢 = − 𝐊 𝐟𝐟 |𝐝 (229)
𝐟

∂𝐫 ∂𝐪𝐢
∂λ
= 𝐟̇ − ∂𝐝𝐫
𝐝̇ = 𝐟̇ − 𝐊 𝐟𝐫 |𝐝 𝐝̇ (230)

Using the above expressions;

𝛅𝐫 = δλ (𝐟̇ − 𝐊 𝐟𝐫 |𝐝 𝐝̇) − 𝐊 𝐟𝐟 |𝐝 𝛅𝐝𝐟 (231)

Following Newton iteration, the updated residual is set to zero. Therefore,

𝐫 + 𝛅𝐫 = 𝟎 ⟹ 𝐫 + (𝐟̇ − 𝐊 𝐟𝐫 |𝐝 𝐝̇) δλ − 𝐊 𝐟𝐟 |𝐝 𝛅𝐝𝐟 = 𝟎 (232)

solving for 𝛅𝐝𝐟 :

𝛅𝐝𝐟 = 𝐊 −𝟏 ̇ ̇
𝐟𝐟 {𝐫 + (𝐟 − 𝐊 𝐟𝐫 |𝐝 𝐝) δλ } (233)

This can be written in a simplified form as:

𝛅𝐝𝐟 = 𝐝̅ + δλ 𝐝̃𝐟 (234)

in which the stiffness terms are now assumed to be evaluated at the current (iterative) deflection and
where the following notations have been used;

𝐝̅ = 𝐊 −𝟏
𝐟𝐟 𝐫 (235)

𝐝̃𝐟 = 𝐊 −𝟏 ̇ ̇
𝐟𝐟 |𝐝 (𝐟 − 𝐊 𝐟𝐫 |𝐝 𝐝) (236)

For the free DOF’s, the arc-length should be maintained. Therefore:

‖∆𝐝𝐟 ‖ = ‖𝐝𝐟 − 𝐝𝐟,𝟎 + 𝛅𝐝𝐟 ‖ = α

(237)
‖𝐝𝐟 − 𝐝𝐟,𝟎 + 𝐝̅ + δλ 𝐝̃𝐟 ‖ = ‖ 𝐃 + δλ 𝐝̃𝐟 ‖ = α

where;

𝐃 ∶= 𝐝𝐟 − 𝐝𝐟,𝟎 + 𝐝̅ (238)

Thus:
𝟐
(𝐃 + δλ 𝐝̃𝐟 ) = α2 ⟹ 𝐃𝟐 + 2δλ 𝐝̃𝐟 . 𝐃 + δλ2 𝐝̃𝟐𝐟 = α2 (239)

Equation (239) is a quadratic equation in δλ;

a1 δλ2 + a2 δλ + a3 = 0 (240)

with;

226

SACS® Collapse Advanced Release 15.1

 a1 = 𝐝̃𝟐𝐟 , a2 = 2 𝐝̃𝐟 . 𝐃 , a3 = 𝐃𝟐 − α2 (241)

The above quadratic equation has two roots, denoted, δλ1 and δλ2 , each of which provides an
associated free DOF deflection increment:

𝛅𝐝𝐟 ( δλ1 ) = 𝐝̅ + δλ1 𝐝̃𝐟

(242)
𝛅𝐝𝐟 ( δλ2 ) = 𝐝̅ + δλ2 𝐝̃𝐟
The current deflection state for the free DOF’s is then updated by 𝛅𝐝𝐟 as defined by the following
equation:

𝐝𝐟 = 𝐝𝐟 + 𝛅𝐝𝐟 (243)

The updated incremental deflection states are represented by the following equations:

𝚫𝐝𝐟 ( δλ1 ) = 𝚫𝐝𝐟 + 𝐝̅ + δλ1 𝐝̃𝐟

(244)
𝚫𝐝𝐟 ( δλ2 ) = 𝚫𝐝𝐟 + 𝐝̅ + δλ2 𝐝̃𝐟
Applying the minimum angle criterion to δλ1 and δλ2 yields the following condition:

𝚫𝐝𝐟 . 𝐝̇𝐫𝐞𝐟 = max{ 𝚫𝐝𝐟 ( δλ1 ). 𝐝̇𝐫𝐞𝐟 , 𝚫𝐝𝐟 ( δλ2 ). 𝐝̇𝐫𝐞𝐟 } (245)

For a given value of δλ from equation (254) δλ, the current (iterative) updated update deflection state 𝐝
is given by:

𝐝𝐟 𝐝 + 𝛅𝐝𝐟 ( δλ)
𝐝= { } = { 𝐟 } (246)
𝐝𝐫 𝐝𝐫 + δλ𝐝̇
and the current (iterative) update incremental load factor Δλ given by the following equation:

Δλ = Δλ + δλ (247)

The next stage requires the determination of a new residual and a repeat of the above procedure until
the residual is bounded by an acceptable tolerance.

227

SACS® Collapse Advanced Release 15.1

6.6.3 The Spherical Arc-Length Method
The procedure for the spherical arc-length is almost identical to that for the cylindrical arc-length
method. However, the criterion (ii) for the user-specified arc-length becomes:

𝚫𝐝𝟐𝐟 + Δλ2 ψ2 𝐟̇ 𝟐 = α2 (248)

where the constant, ψ is also user-specified, but almost always set to unity. Noting that:

𝛅𝐝𝐟 = δλ 𝐊 −𝟏 ̇ ̇ ̃
𝐟𝐟 (𝐟 − 𝐊 𝐟𝐫 𝐝) = δλ 𝐝𝐟 (249)

As with the cylindrical arc-length method, the free DOF predictor deflection, 𝐝̃𝐟 has been written in
terms of the un-factored deflections and applied loads as follows:

𝐝̃𝐟 = 𝐊 −𝟏 ̇ ̇
𝐟𝐟 (𝐟 − 𝐊 𝐟𝐫 𝐝) (250)

Combining equations (249) and (250) gives the initial predictor of the load factor increment as follows:

δλ2 (𝐝̃𝟐𝐟 + ψ2 𝐟̇ 𝟐 ) = α2 (251)

where the initial statement for the spherical arc-length criterion is given by:

𝛅𝐝𝟐𝐟 + Δδ ψ2 𝐟̇ 𝟐 = α2 (252)

Solving for the predictor calculation for δλ gives:

α2
δλ = ±√𝐝̃𝟐 + ψ2 𝐟̇𝟐 (253)
𝐟

For the residual minimization iterations, the same incremental expression is used as that for the
cylindrical arc-length method except that the spherical arc-length criterion now requires:
𝟐
(𝐃 + δλ 𝐝̃𝐟 ) + ψ2 𝐟̇ 𝟐 ( Δλ + δλ)2 = α2 (254)

In which 𝐃 is defined as follows:

𝐃 ∶= 𝐝𝐟 − 𝐝𝐟,𝟎 + 𝐝̅ (255)

This requires the solution of the following quadratic equation for δλ:

a1 δλ2 + a2 δλ + a3 = 0 (256)

With;

a1 = 𝐝̃𝟐𝐟 + ψ2 𝐟̇ 𝟐 , a2 = 2 𝐝̃𝐟 . 𝐃 + 2 ψ2 𝐟̇ 𝟐 Δλ, a3 = 𝐃𝟐 + ψ2 𝐟̇ 𝟐 Δλ2 − α2 (257)

As before, the root δλ can be selected based on angle minimization.

228

SACS® Collapse Advanced Release 15.1

6.6.4 Arc-Length Parameter Calculations and User-defined factors
The Arc-length parameter α is automatically calculated on last converged Newton iteration and is given
by the following equations for the two arc length approaches discussed above:

cylindrical arc-length α = ‖𝚫𝐝𝐟 ‖ = √𝚫𝐝𝟐𝐟 (258)

spherical arc-length α = √𝚫𝐝𝟐𝐟 + ψ2 𝚫𝐟 𝟐 (259)

The parameter α may be scaled by a user-defined factor defined in columns 24-29 on ARCLEN input line
(the default value is 1.0). In the case of spherical arc-length method, the parameter ψ may be revised in
columns 31-36 on ARCLEN input line (the default is 1.0).

229

SACS® Collapse Advanced Release 15.1

6.7 Element Offsets
This section illustrates the application of element offsets for the large rotation and displacement
approach implemented in Collapse Advanced. The element offsets lead to additional terms in the global
tangent stiffness and geometric stiffness matrices.
The following notations are used in this section:
𝐟𝐽 , 𝐦𝐽 force and moment vector at joint
𝐟𝑂 , 𝐦𝑂 force and moment vector at offset
𝐝𝐽 , 𝛉𝐽 displacement and rotation vector at joint
𝐝𝑂 , 𝛉𝑂 displacement and rotation vector at offset
δ𝛉 change in orientation due to large rotation at joint
𝛚 offset vector
𝐒 spin operator defined in equation (16)
𝐑 rotation matrix
𝐈 identity matrix
δ variational operator

The rotation matrix associated with δ𝛉 can be expressed as:

𝐑(δ𝛉) = 𝐈 + 𝐒(δ𝛉) (260)

also,
𝛚 + δ𝛚 = 𝐑(δ𝛉)𝛚 ⟹ δ𝛚 = 𝐒(δ𝛉)𝛚 = −𝐒(𝛚)δ𝛉 (261)

The relationship between the internal forces offset (𝐟𝑂 , 𝐦𝑂 ) and the internal forces at the joint, (𝐟𝐽 , 𝐦𝐽 )
can be written matrix form as:
𝐟𝐽 𝐈 𝟎 𝐟𝑂
{ } [ ]{ }
𝐦𝐽 = 𝐒(𝛚) 𝐈 𝐦𝑂 (262)

where equation (17) is used to define the cross product between an offset and the force vector. Using
above equations leads to the following expression for the moment vector at joint (𝐟𝐽 , 𝐦𝐽 ):
δ𝐦𝐽 = 𝐒(𝐟𝑂 )𝐒(𝛚)δ𝛉 + 𝐒(𝛚) + 𝛅𝐟𝑂 + δ𝐦𝑂 (263)

Next step is to determine a similar relationship between the variation of displacements and rotations.
The displacement 𝐝𝑂 and rotation vector 𝛉𝑂 at offset can be expressed in terms of the displacement
𝐝𝑗 and rotation 𝛉𝑗 at the joint as:
𝛉𝑂 = 𝛉𝑗 , 𝐝𝑂 = 𝐝𝑗 + 𝛚 (264)

Hence,
δ𝛉𝑂 = δ𝛉𝑗 , δ𝐝𝑂 = δ𝐝𝑗 + 𝐒(δ𝛉)𝛚 (265)

using equation (261) for δ𝛚, equation (264) can be written in matrix form as:

230

SACS® Collapse Advanced Release 15.1

 δ𝐝𝑂 𝐈 −𝐒(𝛚) δ𝐝𝐽
{ }=[ ]{ } (266)
δ𝛉𝑂 𝟎 𝐈 δ𝛉𝐽

Combining above equation with (262):

δ𝐟𝐽 𝐈 𝟎 δ𝐟𝑂 𝟎 𝟎 δ𝐝
{ }=[ ]{ }+[ ] { 𝑂} (267)
δ𝐦𝐽 𝐒(𝛚) 𝐈 δ𝐦𝑂 𝟎 𝐒(𝐟𝑂 )𝐒(𝛚) δ𝛉𝑂

δ𝐟𝑂 δ𝐝
and considering { } = 𝐊 𝑂 { 𝑂 } and using equation (266) leads to:
δ𝐦𝑂 δ𝛉𝑂
δ𝐟𝐽 𝐈 𝟎 𝐈 −𝐒(𝛚) 𝟎 𝟎 δ𝐝𝐽
{ } = ൬[ ]𝐊 [ ]+[ ]൰ { } (268)
δ𝐦𝐽 𝐒(𝛚) 𝐈 𝑂 𝟎 𝐈 𝟎 𝐒(𝐟𝑂 )𝐒(𝛚) δ𝛉𝐽

Equations (262) and (268) are used to calculate effects of offsets on the element force and stiffness.

231

SACS® Collapse Advanced Release 15.1

6.8 Element End-Releases
The member end releases are used to define the connection between the member ends and model
joints. End-releases are defined in member local coordinate systems and for large deformation analysis,
the member ends may experience large deformation – leading to change in orientation of end releases
with respect to the global coordinate systems.

End-released can be modelled in local coordinate system as follow

• For released degrees of freedom, the stiffness is zero and no load is translated from model joint
to member end.
• For fixed degrees of freedom, the member ends and model joints are connected using rigid
links.

Following above assumptions, Collapse utilizes a special zero-length finite element (i.e. subsegment) to
model the releases. These subsegments is automatically inserted at member ends with releases in
addition to standard subsegments. The stiffness matrix of the special subsegment is given by
T
𝐊 𝑟𝑒𝑙 = 𝐓𝑟𝑒𝑙 𝐊 𝑟𝑒𝑙 𝐓𝑟𝑒𝑙 (269)

where 𝐊 𝑟𝑒𝑙 is local release stiffness which is determined based on zero-stiffness or rigid links connecting
member ends and model joints. 𝐓𝑟𝑒𝑙 is the transformation matrix computed using member ends
corotational coordinate systems in deformed configuration as shown in Figure 198. Thus,

• At End-A, 𝐓𝑟𝑒𝑙 is computed based on 𝐔 (1) (see section 6.3.1) of the first member standard
subsegment
• At End-B, 𝐓𝑟𝑒𝑙 is computed based on 𝐔(2) (see section 6.3.1) of the last member standard
subsegment

𝑧
𝑦

𝑥 𝐔(2)
𝐔(1)

Deformed Configuration

Figure 198: Collapse Advanced member end releases

232

SACS® Collapse Advanced Release 15.1

6.9 Joint Strength
This section presents the formulation for various joint strength check methods supported by the
program.

The following notation is used to describe the joint strength methods:

𝑃, 𝑀 brace axial force and moment, respectively
𝑃𝑢𝑗 joint ultimate capacity for axial load
𝑀𝑢𝑗 joint ultimate capacity for bending moment
𝜑𝑗 resistance factors (set by the method or user-defined factors given on RSFAC/RSFACO input line)
𝐴𝑋𝐿 axial tension or compression load
𝐼𝑃𝐵 in-plane bending moment
𝑂𝑃𝐵 out-of-plane bending moment
𝜃 brace angle
𝑔 gap for K connection
𝑒 eccentricity for K and X connections
𝑡 brace wall thickness
𝑇 chord wall thickness
𝑑 brace outer diameter
𝐷 chord outer diameter
𝑡
𝜏 brace wall thickness to chord wall thickness ratio,
𝑇
𝑑
𝛽 brace diameter to chord diameter ratio, 𝐷
𝐷
𝛾 chord radius to chord wall thickness ratio,
2𝑇
𝐹𝑦𝑏 brace yield stress
𝐹𝑦𝑐 chord yield stress
𝑃𝐶 , 𝑀𝐶 chord axial force and moment, respectively
𝑓𝑎𝑥𝑙 axial stress in chord
𝑓𝑖𝑝𝑏 in-plane bending stress in chord
𝑓𝑜𝑝𝑏 out-of-plane bending stress in chord
𝜎𝑡 , 𝜎𝑐 brace tensile and compression stress
𝜎𝑏𝑦 , 𝜎𝑏𝑧 brace bending stresses
𝑃𝑦 chord plastic axial strength
𝑀𝑝 chord plastic strength moment
𝑄𝑓 chord force factor
𝑄𝑢 joint strength factor
𝑄𝛽 factor for brace diameter to chord diameter ratio
𝑄𝑔 gap factor for K connection
𝑄𝑦𝑦 angle correction factor for K connection
𝐸𝑏 brace Young’s modulus
𝐸𝑐 chord Young’s modulus
𝑍𝑏 , 𝑍𝑐 brace and chord plastic section modulus
𝑆𝑏 , 𝑆𝑐 brace and chord elastic section modulus

233

SACS® Collapse Advanced Release 15.1

6.9.1 Applicability Ranges
The program enforces applicability ranges presented in Table 7 for joint strength checks.

Table 7: Joint Strength Applicability Ranges

Method Applicability Range Description

10 ≤ 𝛾 ≤ 50
0.2 ≤ 𝛽 ≤ 1.0
For all connection types (see the
30° ≤ 𝜃 ≤ 90°
notes below)
API LRFD 𝑘𝑁
𝐹𝑦𝑐 , 𝐹𝑦𝑏 ≤ 50
𝑐𝑚 2
𝑔≥0 For K connections (no overlap)
𝑒
| | ≤ 0.25 For K and X connections
𝐷
10 ≤ 𝛾 ≤ 50
0.2 ≤ 𝛽 ≤ 1.0
30° ≤ 𝜃 ≤ 90°
For all connection types
ISO 19902 𝜏 ≤ 1.0
𝑘𝑁
𝐹𝑦𝑐 , 𝐹𝑦𝑏 ≤ 50
𝑐𝑚 2
𝑔
−0.6 < ≤ 1.0 For K connections
𝐷
10 ≤ 𝛾 ≤ 50
0.2 ≤ 𝛽 ≤ 1.0
30° ≤ 𝜃 ≤ 90°
For all connection types
NORSOK 𝜏 ≤ 1.0
𝑘𝑁
𝐹𝑦𝑐 , 𝐹𝑦𝑏 ≤ 50
𝑐𝑚 2
𝑔
−0.6 ≤ ≤ 1.0 For K connections
𝐷
10 ≤ 𝛾 ≤ 50
0.2 ≤ 𝛽 ≤ 1.0
30° ≤ 𝜃 ≤ 90° For all connection types
MSL 𝑘𝑁
𝐹𝑦𝑐 , 𝐹𝑦𝑏 ≤ 50
𝑐𝑚 2
𝑔
−0.6 < For K connections
𝐷
𝑘𝑁
Additional 𝐸𝑐 , 𝐸𝑏 > 2000 For all connection types (see notes
Checks 𝑐𝑚 2 below)
Brace offsets

234

SACS® Collapse Advanced Release 15.1

The following comments should be noted regarding joint strength applicability ranges

✓ If a given K or X connection does not satisfy gap or eccentricity limits, the program assumes the
joint acts as a Y/T connection.

✓ If a given connection does not satisfy brace/chord applicability ranges, by default the program
will exclude the joint from strength check calculations.

✓ API LRFD does not explicitly provide brace/chord applicability range. The above limits are
reported by Bomel Limited in [44].

✓ Young’s modulus checks are enforced to ensure numerical stability.

✓ The program checks the braces offsets to ensure it is located outside of the chord. The default
tolerance for offset check is 0.1 cm or inch which can be revised on JSOPT input line.

✓ The joint applicability checks can be overridden by options available on CLPOP2 input line.

6.9.2 API RP 2A-LRFD

API RP 2A-LRFD states the joint capacity should satisfy following conditions:
For axial load (tension or compression):
𝑃
𝜑𝑗 𝑃𝑢𝑗
≤ 1.0 (270)

For in-plane bending (IPB) moment or out-of-plane bending (OPB) moment:

𝑀
𝜑𝑗 𝑀𝑢𝑗
≤ 1.0 (271)

and for the combined axial and bending:

2 2 1⁄2
𝜋 𝑃 𝑀 𝑀
1-cos[ ൬ ൰] + [൬ ൰ +൬ ൰ ] ≤ 1.0 (272)
2 𝜑𝑗 𝑃𝑢𝑗 𝜑𝑗𝑀𝑢𝑗 𝜑𝑗 𝑀𝑢𝑗
𝐼𝑃𝐵 𝑂𝑃𝐵

where the default values for the resistance factor 𝜑𝑗 are given in Table 8, and these can be revised using
RSFAC or RSFACO input lines. The program reports equations (270) and (272) as unity check in
accordance to API LRFD joint strength checks.

Table 8: API resistance factors based on API LRFD Commentary section E.1

Joint Type Axial Tension Axial Compression IPB OPB

T/Y 2.11 2.57 2.81 2.61
X 2.11 2.57 2.81 2.61
K 2.51 2.51 2.81 2.61

235

SACS® Collapse Advanced Release 15.1

In equation (272), 𝑃𝑢𝑗 and 𝑀𝑢𝑗 represents the ultimate capacities as below:

For axial load (tension or compression):

𝐹𝑦𝑐 𝑇2
𝑃𝑢𝑗 = sin (𝜃) 𝑄𝑢 𝑄𝑓 (273)

For in-plane bending (IPB) moment or out-of-plane bending (OPB) moment:

𝐹𝑦𝑐 𝑇2
𝑀𝑢𝑗 = sin(𝜃) (0.8𝑑)𝑄𝑢 𝑄𝑓 (274)

where, the strength factor 𝑄𝑢 are given in Table 9, and the chord force factor 𝑄𝑓 is determined as
follows:

2 2
𝑄𝑓 = { 1.0 𝑖𝑓 𝑓𝑎𝑥𝑙 ≥ √𝑓𝐼𝑃𝐵 + 𝑓𝑂𝑃𝐵 (𝑎𝑙𝑙 𝑐ℎ𝑜𝑟𝑑 𝑓𝑖𝑏𝑒𝑟𝑠 𝑎𝑡 𝑡𝑒𝑛𝑠𝑖𝑜𝑛) (275)
1.0 − 𝜆𝐴2 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
with

2 +𝑓 2 +𝑓 2
√𝑓𝐴𝑋𝐿 𝐼𝑃𝐵 𝑂𝑃𝐵
𝐴= 𝜑𝑞 𝐹𝑦𝑐
(276)

where 𝜆 is defined by API LRFD section E.3.1.1:

• brace axial force: 0.030

• brace in-plane bending moment: 0.045

• brace out-of-plane bending moment: 0.021

and 𝜑𝑞 is the yield resistance factor with default value of 1.0 which can be revised using the RSFAC or
RSFACO input lines.

Table 9: API LRFD strength factor 𝑄𝑢 based on API LRFD table E.3-2

Joint
Axial Tension Axial Compression IPB OPB
Type
T/Y 3.4 + 19𝛽 3.4 + 19𝛽 3.4 + 19𝛽 (3.4 + 7𝛽)𝑄𝛽
X 3.4 + 19𝛽 (3.4 + 13𝛽)𝑄𝛽 3.4 + 19𝛽 (3.4 + 7𝛽)𝑄𝛽
K (3.4 + 19𝛽)𝑄𝑔 (3.4 + 19𝛽)𝑄𝑔 3.4 + 19𝛽 (3.4 + 7𝛽)𝑄𝛽
0.3
𝑖𝑓 𝛽 > 0.6
𝑄𝛽 = {𝛽(1 − 0.833𝛽)
1.0 𝑖𝑓 𝛽 ≤ 0.6
𝑔
1.8 − 0.1 ≥ 1.0 𝑖𝑓 𝛾 ≤ 20
𝑄𝑔 = { 𝑇
𝑔
1.8 − 4 ≥ 1.0 𝑖𝑓 𝛾 > 20
𝐷

236

SACS® Collapse Advanced Release 15.1

The program determines the joint classification based on API LRFD Figure E.3-2. The classification
weighting factors are defined as 𝐶𝑌 , 𝐶𝑋 , and 𝐶𝐾 with 𝐶𝑌 + 𝐶𝑋 + 𝐶𝐾 = 1. The Collapse Advanced uses
the classification weighting factors and follows API LRFD recommendations to determine the joint
capacity with the mixed classification using following expressions:
𝜑𝑗 𝑚𝑖𝑥𝑒𝑑 = 𝐶𝑌 𝜑𝑗 𝑌 + 𝐶𝑋 𝜑𝑗 𝑋 + 𝐶𝐾 𝜑𝑗 𝐾 (277)

𝑄𝑢 𝑚𝑖𝑥𝑒𝑑 = 𝐶𝑌 𝑄𝑢 𝑌 + 𝐶𝑋 𝑄𝑢 𝑋 + 𝐶𝐾 𝑄𝑢 𝐾 (278)

𝑄𝑓 𝑚𝑖𝑥𝑒𝑑 = 𝐶𝑌 𝑄𝑓 𝑌 + 𝐶𝑋 𝑄𝑓 𝑋 + 𝐶𝐾 𝑄𝑓 𝐾 (279)

Above expressions are separately computed for axial tension, axial compression, in-plane bending, and
out-of-plane bending. Once the average quantities are computed, the program utilizes equations (273)
and (274) to obtain the joint capacities for axial tension, axial compression, in-plane bending, and out-
of-plane bending.
6.9.3 ISO 19902
ISO 19902 states the joint capacity should satisfy following conditions:
2
𝑃 𝑀 𝑀
| 𝑃𝑢𝑗 |+( 𝑀𝑢𝑗 ) +| 𝑀𝑢𝑗 | ≤ 1.0 (280)
𝜑𝑗 𝜑𝑗 𝜑𝑗
𝛾𝑅,𝑗 𝛾𝑅,𝑗 𝛾𝑅,𝑗
𝐼𝑃𝐵 𝑂𝑃𝐵

2
𝑃 𝑀 𝑀 𝑈
| 𝜑𝑗𝑃𝑢𝑗 | + ( 𝜑𝑗𝑀𝑢𝑗 ) + | 𝜑𝑗𝑀𝑢𝑗 | ≤𝛾𝑏 (281)
𝑧,𝑗
𝛾𝑅,𝑗 𝛾𝑅,𝑗 𝛾𝑅,𝑗
𝐼𝑃𝐵 𝑂𝑃𝐵

where
• 𝛾𝑅,𝑗 is partial resistance factor for joints with default value of 1.05 and it can be revised on
RSFAC input line.
• 𝛾𝑧,𝑗 is extra partial resistance factor to ensure that member fails before the joint yields. Its
default value is 1.17 and can be revised using the RSFAC input line.
• 𝜑𝑗 are the user-defined resistance factors with default values of 1.0. The defaults values can be
revised for different joint/load type using the RSFAC or RSFACO input lines.
• 𝑈𝑏 is brace utilization factor determine ISO 19902 section 13.
In addition to above unity check, the Collapse Advanced reports minimum tensile brace strain and
compares it with ISO 19902 ductility limit of 5.0%.
𝑃𝑢𝑗 and 𝑀𝑢𝑗 ultimate capacities are given as follow

For axial load (tension or compression):

𝐹𝑦𝑐 𝑇2
𝑃𝑢𝑗 = 𝑄𝑢 𝑄𝑓 (282)
sin (𝜃)

For in-plane bending (IPB) moment or out-of-plane bending (OPB) moment:

237

SACS® Collapse Advanced Release 15.1

 𝐹𝑦𝑐 𝑇2 𝑑
𝑀𝑢𝑗 = 𝑄𝑢 𝑄𝑓 (283)
sin(𝜃)

The strength factor 𝑄𝑢 is given in Table 10 and the chord force factor 𝑄𝑓 is determined as follow (based
ISO 19902 section 14.3.4):

𝑄𝑓 = 1.0 − 𝜆𝑞𝐴2 (284)

where 𝜆 is defined by ISO 19902

• brace axial force: 0.030

• brace in-plane bending moment: 0.045

• brace out-of-plane bending moment: 0.021

and

2 2 2
𝑃 𝑀 𝑀
𝑞𝐴 = 𝛾𝑅,𝑞 √𝐶1 ൬𝑃𝐶 ൰ + 𝐶2 ൬𝑀𝐶 ൰ + 𝐶2 ൬𝑀𝐶 ൰ (285)
𝑦 𝑝 𝐼𝑃𝐵 𝑝 𝑂𝑃𝐵

where 𝛾𝑅,𝑞 is the partial resistance factor for the yield with default value of 1.05 and can be revised
using the RSFAC input line. Coefficients 𝐶1 and 𝐶2 are given in Table 11.

The brace utilization factor 𝑈𝑏 is by default 1.0. If option ‘B’ for brace utilization factor is selected on
JSOPT input line, 𝑈𝑏 is calculated as follows (ISO 19902 equation 13.3-2 and 13.3-8):

For combined tension and bending

2 +𝜎2
𝛾𝑅,𝑏√𝜎𝑏,𝑦
𝛾𝑅,𝑡 𝜎𝑡 𝑏,𝑧
𝑈𝑏 = + (286)
𝑓𝑡 𝑓𝑏

For combined compression and bending

2 +𝜎2
𝛾𝑅,𝑏√𝜎𝑏,𝑦
𝛾𝑅,𝑐𝜎𝑐 𝑏,𝑧
𝑈𝑏 = 𝑓𝑦𝑐
+ 𝑓𝑏
(287)

in which (ISO 19902 sections 13.2.2 (tension), 13.2.3.3 (compression), and 13.2.4 (bending)):
𝑓𝑡 = 𝐹𝑦𝑏
𝐹𝑦𝑏
𝐹𝑦𝑏 𝑓𝑥𝑒
≤ 0.170
2𝐶𝑋 𝐸𝑏 𝑡
𝑓𝑦𝑐 = { 𝐹 𝐹𝑦𝑏 𝑤𝑖𝑡ℎ 𝑓𝑥𝑒 = 𝑑
𝑎𝑛𝑑 𝐶𝑋 = 0.3
(1.047 − 0.274 𝑦𝑏) 𝐹𝑦𝑏 > 0.170
𝑓𝑥𝑒 𝑓𝑥𝑒
𝑍𝑏 𝐹𝑦𝑏𝑑 (288)
𝐹𝑦𝑏 ≤ 0.0517
𝑆𝑏 𝐸𝑏 𝑡
𝐹𝑦𝑏𝑑 𝑍𝑏 𝐹𝑦𝑏𝑑
𝑓𝑏 = (1.13 − 2.58 𝐸𝑏 𝑡 𝑆𝑏
) 𝐹𝑦𝑏 0.0517 < 𝐸𝑏 𝑡
≤ 0.1034
𝐹𝑦𝑏𝑑 𝑍𝑏 𝐹𝑦𝑏𝑑 𝐹𝑦𝑏
{(0.94 − 0.76 𝐸 𝑡 ) 𝑆 𝐹𝑦𝑏
𝑏 𝑏
0.1034 < 𝐸𝑏 𝑡
≤ 120 𝐸𝑏

and 𝛾𝑅,𝑡 = 1.05, 𝛾𝑅,𝑐 = 1.18, 𝛾𝑅,𝑏 = 1.05.

238

SACS® Collapse Advanced Release 15.1

The program determines the joint classification based on ISO 19902 Figure 14.2-2. The classification
weighting factors are defined as 𝐶𝑌 , 𝐶𝑋 , and 𝐶𝐾 with 𝐶𝑌 + 𝐶𝑋 + 𝐶𝐾 = 1. The Collapse Advanced uses
the classification weighting factors and follows ISO 19902 recommendations to determine the joint
capacity with the mixed classification using following expressions:
(𝜑𝑗 𝑃𝑢𝑗 )𝑚𝑖𝑥𝑒𝑑 = 𝐶𝑌 (𝜑𝑗 𝑃𝑢𝑗 )𝑌 + 𝐶𝑋 (𝜑𝑗 𝑃𝑢𝑗 )𝑋 + 𝐶𝐾 (𝜑𝑗 𝑃𝑢𝑗 )𝐾 (289)

(𝜑𝑗 𝑀𝑢𝑗 ) = 𝐶𝑌 (𝜑𝑗 𝑀𝑢𝑗 )𝑌 + 𝐶𝑋 (𝜑𝑗 𝑀𝑢𝑗 )𝑋 + 𝐶𝐾 (𝜑𝑗 𝑀𝑢𝑗 )𝐾 (290)

𝑚𝑖𝑥𝑒𝑑

In the above equations, mixed ultimate capacities are separately computed for axial tension, axial
compression, in-plane bending, and out-of-plane bending components.

Table 10: ISO 19902 strength factor 𝑄𝑢 based on ISO 19902 section 14.3.3

Joint
Axial Tension Axial Compression IPB OPB
Type
T/Y 30𝛽 (1.9 + 19𝛽)𝑄𝛽0.5 4.5𝛽𝛾 0.5 3.2𝛾 (0.5𝛽
2)

20.7 + (𝛽 − 0.9)(17𝛾 − 220) 𝑖𝑓 𝛽 > 0.9 2)

X { [2.8 + (12 + 0.1𝛾)𝛽]𝑄𝛽 4.5𝛽𝛾 0.5 3.2𝛾 (0.5𝛽
23𝛽 𝑖𝑓 𝛽 ≤ 0.9
K (1.9 + 19𝛽)𝑄𝛽0.5 𝑄𝑔 (1.9 + 19𝛽)𝑄𝛽0.5 𝑄𝑔 4.5𝛽𝛾 0.5 3.2𝛾 (0.5𝛽
2)

0.3
𝑖𝑓 𝛽 > 0.6
𝑄𝛽 = {𝛽(1 − 0.833𝛽)
1.0 𝑖𝑓 𝛽 ≤ 0.6
𝑡𝐹𝑦𝑏 0.5 𝑔
0.13 + 0.65 𝛾 ≤ −2.0
𝑇𝐹𝑦𝑐 𝑇
0.5
𝑄𝑔 = 1.9 − 0.7𝛾 −0.5 (𝑔) ≥ 1.0 𝑔
≥ 2.0
𝑇 𝑇
𝑔
{ 𝑖𝑛𝑡𝑒𝑟𝑝𝑜𝑙𝑎𝑡𝑒 −2.0 < < 2.0
𝑇

Table 11: ISO 19902 𝐶1 and 𝐶2 coefficients for chord force factor 𝑄𝑓 based on ISO 19902 Table 14.3-2

Joint Type 𝐶1 𝐶2
T/Y Axial Tension/Compression 25 11
X Axial Tension/Compression 20 22
K Axial Tension/Compression 14 43
All joints IPB and OPB 25 43

239

SACS® Collapse Advanced Release 15.1

6.9.4 NORSOK
NORSOK states the joint capacity should satisfy following condition:
2
𝑃 𝑀 𝑀
| |+൬ ൰ +| | ≤ 1.0 (291)
𝜑𝑗 𝑃𝑢𝑗 𝜑𝑗 𝑀𝑢𝑗 𝜑𝑗 𝑀𝑢𝑗
𝐼𝑃𝐵 𝑂𝑃𝐵

where
• 𝜑𝑗 are the user-defined resistance factors with default value of 1.0. The defaults values can be
revised for different joint/load types via RSFAC input line.
In equation (291), 𝑃𝑢𝑗 and 𝑀𝑢𝑗 represent the ultimate capacities as follows:

For axial load (tension or compression):

𝐹𝑦𝑐 𝑇2
𝑃𝑢𝑗 = γ 𝑄𝑢 𝑄𝑓 (292)
𝑀 sin (𝜃)

For in-plane bending (IPB) moment or out-of-plane bending (OPB) moment:

𝐹𝑦𝑐 𝑇2 𝑑
𝑀𝑢𝑗 = 𝑄𝑢 𝑄𝑓 (293)
γ𝑀 sin(𝜃)

where γ𝑀 is NORSOK material factor with a default value of 1.15 and can be revised using the RSFAC
input line.

The strength factor 𝑄𝑢 and the chord force factor 𝑄𝑓 are discussed in revisions 1,2, and 3 of the NORSOK
code of practice.

The program determines the joint classification based on NORSOK Figure 6-2. The classification
weighting factors are defined as 𝐶𝑌 , 𝐶𝑋 , and 𝐶𝐾 with 𝐶𝑌 + 𝐶𝑋 + 𝐶𝐾 = 1. The program uses the
classification weighting factors and follows NORSOK recommendations to determine the average joint
capacity using following expressions:
(𝜑𝑗 𝑃𝑢𝑗 )𝑎𝑣𝑔 = 𝐶𝑌 (𝜑𝑗 𝑃𝑢𝑗 )𝑌 + 𝐶𝑋 (𝜑𝑗 𝑃𝑢𝑗 )𝑋 + 𝐶𝐾 (𝜑𝑗 𝑃𝑢𝑗 )𝐾 (294)

(𝜑𝑗 𝑀𝑢𝑗 )𝑎𝑣𝑔 = 𝐶𝑌 (𝜑𝑗 𝑀𝑢𝑗 )𝑌 + 𝐶𝑋 (𝜑𝑗 𝑀𝑢𝑗 )𝑋 + 𝐶𝐾 (𝜑𝑗 𝑀𝑢𝑗 )𝐾 (295)

In the above equations the average ultimate capacities are separately computed for axial tension, axial
compression, in-plane bending, and out-of-plane bending components.

6.9.4.1 NORSOK Revision 1

The strength factor 𝑄𝑢 is given in Table 12, and the chord force factor 𝑄𝑓 is determined as follows:

2 2
𝑄𝑓 = { 1.0 𝑖𝑓 𝑓𝑎𝑥𝑙 ≥ √𝑓𝐼𝑃𝐵 + 𝑓𝑂𝑃𝐵 (𝑒𝑥𝑐𝑒𝑝𝑡 𝑓𝑜𝑟 𝑋 𝑗𝑜𝑖𝑛𝑡 𝑤𝑖𝑡ℎ 𝛽 > 0.9) (296)
1.0 − 𝜆𝑐𝐴2 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
where 𝑐 is

• 14 for Y and K joints

240

SACS® Collapse Advanced Release 15.1

 • 25 for X joints

and 𝜆 is

• brace axial force: 0.030

• brace in-plane bending moment: 0.045

• brace out-of-plane bending moment: 0.021

NORSOK Revision 1 defines the parameter 𝐴 as follows:

2 2 2
𝑓 𝑓𝐼𝑃𝐵 +𝑓𝑂𝑃𝐵
𝐴2 = ൬ 𝐹𝐴𝑋𝐿 ൰ + ( 2
𝑓𝑚
) (297)
𝑦𝑐

with
𝑍𝑐 𝐹𝑦𝑐𝐷
𝐹
𝑆𝑐 𝑦𝑐 𝐸𝑐 𝑇
≤ 0.0517
𝐹𝑦𝑐 𝐷 𝑍𝑐 𝐹𝑦𝑐 𝐷
𝑓𝑚 = (1.13 − 2.58 ) 𝐹
𝐸𝑏 𝑇 𝑆𝑐 𝑦𝑐
0.0517 < 𝐸𝑐 𝑇
≤ 0.1034 (298)
𝐹𝑦𝑐 𝐷 𝑍 𝐹𝑦𝑐 𝐷 𝐹𝑦𝑐
{(0.94 − 0.76 𝐸 𝑇 ) 𝑆𝑐 𝐹𝑦𝑐 0.1034 < 𝐸𝑐 𝑇
≤ 120 𝐸𝑐
𝑏 𝑐

Table 12: NORSOK Revision 1 strength factor 𝑄𝑢 based on NORSOK Table 6-3

Joint
Axial Tension Axial Compression IPB OPB
Type
T/Y 30𝛽 (1.9 + 19𝛽)𝑄𝛽0.5 4.5𝛽𝛾 0.5 3.2𝛾 (0.5𝛽
2)

20.7 + (𝛽 − 0.9)(17𝛾 − 220) 𝑖𝑓 𝛽 > 0.9 2)

X { (2.8 + 14𝛽)𝑄𝛽 4.5𝛽𝛾 0.5 3.2𝛾 (0.5𝛽
23𝛽 𝑖𝑓 𝛽 ≤ 0.9
K (1.9 + 19𝛽)𝑄𝛽0.5 𝑄𝑔 𝑄𝑦𝑦 (1.9 + 19𝛽)𝑄𝛽0.5 𝑄𝑔 𝑄𝑦𝑦 4.5𝛽𝛾 0.5 3.2𝛾 (0.5𝛽
2)

0.3
𝑖𝑓 𝛽 > 0.6
𝑄𝛽 = {𝛽(1 − 0.833𝛽)
1.0 𝑖𝑓 𝛽 ≤ 0.6
𝑡𝐹𝑦𝑏 0.5 𝑔
0.13 + 0.65 𝛾 ≤ −2.0
𝑇𝐹𝑦𝑐 𝑇
0.5
𝑄𝑔 = 1.9 − ( 𝑔 ) ≥ 1.0 𝑔
≥ 2.0
𝐷 𝑇
𝑔
{ 𝑖𝑛𝑡𝑒𝑟𝑝𝑜𝑙𝑎𝑡𝑒 −2.0 < < 2.0
𝑇

1.0 𝑖𝑓 𝜃𝑡 ≤ 4𝜃𝑐 − 90°

𝑄𝑦𝑦 = {110°+4𝜃𝑐 −𝜃𝑡
200°
𝑖𝑓 𝜃𝑡 > 4𝜃𝑐 − 90°

𝜃𝑡 and 𝜃𝑐 are angles of the tension and compression braces.

241

SACS® Collapse Advanced Release 15.1

6.9.4.2 NORSOK Revision 2
The strength factor 𝑄𝑢 is given in Table 13, and the chord force factor 𝑄𝑓 is determined as follows:

2 2
𝑄𝑓 = { 1.0 𝑖𝑓 𝑓𝑎𝑥𝑙 ≥ √𝑓𝐼𝑃𝐵 + 𝑓𝑂𝑃𝐵 (𝑒𝑥𝑐𝑒𝑝𝑡 𝑓𝑜𝑟 𝑋 𝑗𝑜𝑖𝑛𝑡 𝑤𝑖𝑡ℎ 𝛽 > 0.9) (299)
1.0 − 𝜆𝐴2 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
and 𝜆 is

• brace axial force: 0.030

• brace in-plane bending moment: 0.045

• brace out-of-plane bending moment: 0.021

NORSOK Revision 2 defines the parameter 𝐴 as follows:

2 2 2
𝑓 𝑓𝐼𝑃𝐵 +𝑓𝑂𝑃𝐵
𝐴2 = 𝐶1 ൬ 𝐹𝐴𝑋𝐿 ൰ + 𝐶2 ൬ 2
1.62𝐹𝑦𝑐
൰ (300)
𝑦𝑐

in which coefficients 𝐶1 and 𝐶2 are given in Table 14.

Table 13: NORSOK Revision 2 strength factor 𝑄𝑢 based on NORSOK Table 6-3

Joint
Axial Tension Axial Compression IPB OPB
Type
T/Y 30𝛽 (1.9 + 19𝛽)𝑄𝛽0.5 4.5𝛽𝛾 0.5 3.2𝛾 (0.5𝛽
2)

20.7 + (𝛽 − 0.9)(17𝛾 − 220) 𝑖𝑓 𝛽 > 0.9 2)

X { (2.8 + 14𝛽)𝑄𝛽 4.5𝛽𝛾 0.5 3.2𝛾 (0.5𝛽
23𝛽 𝑖𝑓 𝛽 ≤ 0.9
K (1.9 + 19𝛽)𝑄𝛽0.5 𝑄𝑔 (1.9 + 19𝛽)𝑄𝛽0.5 𝑄𝑔 4.5𝛽𝛾 0.5 3.2𝛾 (0.5𝛽
2)

0.3
𝑖𝑓 𝛽 > 0.6
𝑄𝛽 = {𝛽(1 − 0.833𝛽)
1.0 𝑖𝑓 𝛽 ≤ 0.6
𝑡𝐹𝑦𝑏 0.5 𝑔
0.13 + 0.65 𝛾 ≤ −2.0
𝑇𝐹𝑦𝑐 𝑇
0.5
𝑄𝑔 = 1.9 − ( 𝑔 ) ≥ 1.0 𝑔
≥ 2.0
𝐷 𝑇
𝑔
{ 𝑖𝑛𝑡𝑒𝑟𝑝𝑜𝑙𝑎𝑡𝑒 −2.0 < < 2.0
𝑇

Table 14: NORSOK Revision 2 𝐶1 and 𝐶2 for chord force factor 𝑄𝑓 based on NOSROK Table 6-4

Joint Type 𝐶1 𝐶2
T/Y Axial Tension/Compression 25 11
X Axial Tension/Compression 20 22
K Axial Tension/Compression 20 22
All joints IPB and OPB 25 30

242

SACS® Collapse Advanced Release 15.1

 6.9.4.3 NORSOK Revision 3
The strength factor 𝑄𝑢 is given in Table 15, and the chord force factor 𝑄𝑓 is determined as follows:

𝑓𝐴𝑋𝐿 𝑓
𝑄𝑓 = 1.0 + 𝐶1 𝐹𝑦𝑐
𝐼𝑃𝐵
− 𝐶2 1.62𝐹 − 𝐶3 𝐴2 (301)
𝑦𝑐

NORSOK Revision 3 defines the parameter 𝐴 as follows

2 2 2
𝑓 𝑓𝐼𝑃𝐵 +𝑓𝑂𝑃𝐵
𝐴2 = ൬ 𝐹𝐴𝑋𝐿 ൰ + ൬ 2
1.62𝐹𝑦𝑐
൰ (302)
𝑦𝑐

in which coefficients 𝐶1 , 𝐶2 , and 𝐶3 are given in Table 16.

Table 15: NORSOK Revision 3 strength factor 𝑄𝑢 based on NORSOK Table 6-3

Joint
Axial Tension Axial Compression IPB OPB
Type
2.8 + (20 + 0.8𝛾)𝛽1.6
T/Y 30𝛽 𝑚𝑖𝑛 { (5 + 0.7𝛾)𝛽1.2 2.5 + (4.5 + 0.2𝛾)𝛽2.6
2.8 + 36𝛽1.6
2
X 6.4𝛾 (0.6𝛽 ) [2.8 + (12 + 0.1𝛾)𝛽]𝑄𝛽 (5 + 0.7𝛾)𝛽1.2 2.5 + (4.5 + 0.2𝛾)𝛽2.6
1.2
(16 + 1.2𝛾)𝛽 𝑄𝑔 (16 + 1.2𝛾)𝛽1.2 𝑄𝑔
K 𝑚𝑖𝑛 { 𝑚𝑖𝑛 { (5 + 0.7𝛾)𝛽1.2 2.5 + (4.5 + 0.2𝛾)𝛽2.6
40𝛽1.2 𝑄𝑔 40𝛽1.2 𝑄𝑔
0.3
𝑖𝑓 𝛽 > 0.6
𝑄𝛽 = {𝛽(1 − 0.833𝛽)
1.0 𝑖𝑓 𝛽 ≤ 0.6
𝑡𝐹𝑦𝑏 0.5 𝑔
0.13 + 0.65 𝛾 ≤ −0.05
𝑇𝐹𝑦𝑐 𝐷
3
𝑄𝑔 = 1.0 + 0.2 (1 − 2.8 𝑔 ) ≥ 1.0 𝑔
≥ 0.05
𝐷 𝐷
𝑔
{ 𝑖𝑛𝑡𝑒𝑟𝑝𝑜𝑙𝑎𝑡𝑒 −0.05 < < 0.05
𝐷

Table 16: NORSOK Revision 3 𝐶1 and 𝐶2 for chord force factor 𝑄𝑓 based on NOSROK Table 6-4

Joint Type 𝐶1 𝐶2 𝐶3
T/Y Axial Tension/Compression 0.3 0.0 0.8
𝛽 ≤ 0.9 0.0 0.0 0.4
X Axial Tension { 𝛽 = 1.0 0.2 0.0 0.2
𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒 𝑖𝑛𝑡𝑒𝑟𝑝𝑜𝑙𝑎𝑡𝑒 𝑖𝑛𝑡𝑒𝑟𝑝𝑜𝑙𝑎𝑡𝑒 𝑖𝑛𝑡𝑒𝑟𝑝𝑜𝑙𝑎𝑡𝑒

𝛽 ≤ 0.9 0.2 0.0 0.5

X Axial Compression { 𝛽 = 1.0 −0.2 0.0 0.2
𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒 𝑖𝑛𝑡𝑒𝑟𝑝𝑜𝑙𝑎𝑡𝑒 𝑖𝑛𝑡𝑒𝑟𝑝𝑜𝑙𝑎𝑡𝑒 𝑖𝑛𝑡𝑒𝑟𝑝𝑜𝑙𝑎𝑡𝑒

K Axial Tension/Compression 0.2 0.2 0.2

All joints IPB and OPB 0.2 0.0 0.4

243

SACS® Collapse Advanced Release 15.1

6.9.5 MSL
The MSL documentation states the joint capacity should satisfy following condition:
2
𝑃 𝑀 𝑀
| |+൬ ൰ +| | ≤ 1.0 (303)
𝜑𝑗 𝜑𝑐 𝑃𝑢𝑗 𝜑𝑗 𝜑𝑐 𝑀𝑢𝑗 𝜑𝑗 γ𝑐 𝑀𝑢𝑗
𝐼𝑃𝐵 𝑂𝑃𝐵

where
• 𝜑𝑐 is 1.0 for MSL Mean Level (‘MS’ option on MSLOPT input line) and represents the
characteristic bias factors given in Table 17 for MSL Characteristic Level (‘CS’ option on MSLOPT
input line).
• 𝜑𝑗 represents the user-defined resistance factors with default values of 1.0. The default values
can be revised for different joint/load types using the RSFAC or RSFACO input lines.

Table 17: MSL Characteristic bias factors based on MSL C20400R014 Table 2.2

Joint Type Axial Tension Axial Compression IPB OPB

T/Y 0.554 0.791 0.824 0.789
X 0.878 0.864 0.810 0.878
K Gapped 0.769 0.769 0.804 0.804
K Overlapped 0.867 0.867 0.840 0.822

𝑃𝑢𝑗 and 𝑀𝑢𝑗 ultimate capacities are given as follow:

For axial load (tension or compression):

𝐹𝑦𝑐 𝑇2
𝑃𝑢𝑗 = 𝑄𝑢 𝑄𝑓 (304)
sin (𝜃)

For in-plane bending (IPB) moment or out-of-plane bending (OPB) moment:

𝐹𝑦𝑐 𝑇2 𝑑
𝑀𝑢𝑗 = sin(𝜃)
𝑄𝑢 𝑄𝑓 (305)

The Mean Level (unbiased) strength factor 𝑄𝑢 is given in Table 18, and the chord force factor 𝑄𝑓 are
determined as follows:

2 2
𝑄𝑓 = { 1.0 𝑖𝑓 𝑓𝑎𝑥𝑙 ≥ √𝑓𝐼𝑃𝐵 + 𝑓𝑂𝑃𝐵 (𝑒𝑥𝑐𝑒𝑝𝑡 𝑓𝑜𝑟 𝑋 𝑗𝑜𝑖𝑛𝑡 𝑤𝑖𝑡ℎ 𝛽 > 0.9) (306)
1.0 − 𝜆𝑈 2 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
where 𝜆 is defined as

• brace axial force: 0.030

• brace in-plane bending moment: 0.045

• brace out-of-plane bending moment: 0.021

244

SACS® Collapse Advanced Release 15.1

 The parameter 𝑈 is given by the following expression:

2 2 2
1 𝑃 𝑀𝐶 𝑀𝐶
𝑈= √𝐶1 ൬ 𝐶 ൰ + 𝐶2 ൬ ൰ + 𝐶2 ൬ ൰ (307)
𝛾𝑞 𝑃𝑦 𝑀𝑝 𝑀𝑝
𝐼𝑃𝐵 𝑂𝑃𝐵

in which 𝛾𝑞 is the assessment factor of the safety with the default value of 1.0 which can be revised on
MSLOPT input line. Coefficients 𝐶1 and 𝐶2 are given in Table 19.

The classification weighting factors are defined as 𝐶𝑌 , 𝐶𝑋 , and 𝐶𝐾 with 𝐶𝑌 + 𝐶𝑋 + 𝐶𝐾 = 1. The program
uses the classification weighting factors and follows MSL recommendation to determine the joint
capacity with the mixed classification using the following expressions:
(𝜑𝑗 𝜑𝑐 𝑃𝑢𝑗 )𝑚𝑖𝑥𝑒𝑑 = 𝐶𝑌 (𝜑𝑗 𝜑𝑐 𝑃𝑢𝑗 )𝑌 + 𝐶𝑋 (𝜑𝑗 𝜑𝑐 𝑃𝑢𝑗 )𝑋 + 𝐶𝐾 (𝜑𝑗 𝜑𝑐 𝑃𝑢𝑗 )𝐾 (308)

(𝜑𝑗 𝜑𝑐 𝑀𝑢𝑗 )𝑚𝑖𝑥𝑒𝑑 = 𝐶𝑌 (𝜑𝑗 𝜑𝑐 𝑀𝑢𝑗 )𝑌 + 𝐶𝑋 (𝜑𝑗 𝜑𝑐 𝑀𝑢𝑗 )𝑋 + 𝐶𝐾 (𝜑𝑗 𝜑𝑐 𝑀𝑢𝑗 )𝐾 (309)

In the above expressions, the mixed ultimate capacities are separately computed for axial tension, axial
compression, in-plane bending, and out-of-plane bending components.

Table 18: Mean Level MSL strength factor 𝑄𝑢 based on MSL C20400R014 Table 2.1

Joint
Axial Tension Axial Compression IPB OPB
Type
T/Y 17.6 + 42.3𝛽 1.27(1.9 + 19𝛽)𝑄𝛽0.5 5.5𝛽𝛾 0.5 4.2𝛾 (0.5𝛽
2)

𝑒
40 + (𝛽 − 0.9)(37.6𝛾 − 364) 𝑖𝑓 𝛽 > 0.9 𝑎𝑛𝑑 ≤ 0.2
𝐷 1.16(2.8 + 14𝛽)𝑄𝛽 5.5𝛽𝛾 0.5
2)
X { 𝑒 4.2𝛾 (0.5𝛽
6.6 + 37.3𝛽 𝑖𝑓 𝛽 ≤ 0.9 𝑜𝑟 > 0.2
𝐷
K 1.1 × 1.3(1.9 + 19𝛽)𝑄𝛽0.5 𝑄𝑔 𝑄𝑦𝑦 1.3(1.9 + 19𝛽)𝑄𝛽0.5 𝑄𝑔 𝑄𝑦𝑦 5.5𝛽𝛾 0.5 4.2𝛾 (0.5𝛽
2)

0.3
𝑖𝑓 𝛽 > 0.6
𝑄𝛽 = {𝛽(1 − 0.833𝛽)
1.0 𝑖𝑓 𝛽 ≤ 0.6
𝑡𝐹𝑦𝑏 0.5 𝑔
0.13 + 0.65 𝛾 ≤ −2.0
𝑇𝐹𝑦𝑐 𝑇
0.5
𝑄𝑔 = 1.9 − ( 𝑔 ) ≥ 1.0 𝑔
≥ 2.0
𝐷 𝑇
𝑔
{ 𝑖𝑛𝑡𝑒𝑟𝑝𝑜𝑙𝑎𝑡𝑒 −2.0 < < 2.0
𝑇

1.0 𝑖𝑓 𝜃𝑡 ≤ 4𝜃𝑐 − 90°

𝑄𝑦𝑦 = {110°+4𝜃𝑐 −𝜃𝑡
200°
𝑖𝑓 𝜃𝑡 > 4𝜃𝑐 − 90°

𝜃𝑡 and 𝜃𝑐 are angles of the tension and compression braces.

245

SACS® Collapse Advanced Release 15.1

 Table 19: MSL 𝐶1 and 𝐶2 coefficients for chord force factor 𝑄𝑓 based on MSL C20400R014 Table 2.4

Joint Type 𝐶1 𝐶2
T/Y Axial Tension/Compression 25 11
X Axial Tension/Compression 25 43
K Axial Tension/Compression 14 43
All joints IPB and OPB 25 43

6.9.5.1 MSL Fracture

MSL joint calculation also performs ductility facture checks on joints in tension. For each joint in tension,
the joint axial deformation is compared with duality limits summarized in Table 20 for the Mean Level
and Characteristic Level.

Table 20: MSL duality limits based on MSL C20400R014 Section 6.5

Joint Type Mean Level Characteristic Level

T/Y 0.076𝐷 0.044𝐷
X (0.13 − 0.11𝛽)𝐷 (0.089 − 0.075𝛽)𝐷
K 0.026𝐷 0.015𝐷

The program uses the classification weighting factors and follows MSL recommendation to determine
the average joint ductility limit capacity using following expression:
(𝛿𝑙𝑖𝑚 )𝑎𝑣𝑔 = 𝐶𝑌 (𝛿𝑙𝑖𝑚 )𝑌 + 𝐶𝑋 (𝛿𝑙𝑖𝑚 )𝑋 + 𝐶𝐾 (𝛿𝑙𝑖𝑚 )𝐾 (310)

246

SACS® Collapse Advanced Release 15.1

6.10 Joint Flexibility
This section presents the implementation details of the joint flexibility in the program.

The following notation is used to describe the joint flexibility methods:

𝐴𝑋𝐿 axial tension or compression displacement or load
𝐼𝑃𝐵 in-plane bending rotation or moment
𝑂𝑃𝐵 out-of-plane bending rotation or moment
𝜃 brace angle
𝑔 gap for K connection
𝑒 eccentricity for K and X connections
𝑡 brace wall thickness
𝑇 chord wall thickness
𝑑 brace outer diameter
𝐷 chord outer diameter
𝑡
𝜏 brace wall thickness to chord wall thickness ratio, 𝑇
𝑑
𝛽 brace diameter to chord diameter ratio, 𝐷
𝐷
𝛾 chord radius to chord wall thickness ratio, 2𝑇
𝐹𝑦𝑏 brace yield stress
𝐹𝑦𝑐 chord yield stress
𝐸𝑏 brace Young’s modulus
𝐸𝑐 chord Young’s modulus
𝐪𝐛𝑖 brace end force vector at member joint i in global coordinates
̂𝐛𝑖
𝐪 brace end force vector at member joint i in the joint local coordinate system
𝐩𝐛𝑖 brace displacement vector at member joint i in global coordinates
̂𝐛𝑖
𝐩 brace displacement vector at member joint i in the joint local coordinate system
𝛅𝐩𝐛𝑖 brace displacement increment vector at member joint i in global coordinates
𝛅𝐩̂𝐛𝑖 brace displacement increment vector at member joint i in the joint local coordinate system
̂𝐣𝑖
𝐪 joint force vector at member joint i in the joint local coordinate system
̂ 𝐣𝑖
𝐮 joint distortion vector at member joint i in the joint local coordinate system
𝛅𝐮
̂ 𝐣𝑖 joint distortion increment vector at member joint i in the joint local coordinate system
𝐗𝑖 joint flexibility transformation matrix at member joint i

247

SACS® Collapse Advanced Release 15.1

6.10.1 Applicability Ranges
The program enforces applicability ranges presented in Table 21 below for joint flexibility checks.

Table 21: Joint Flexibility Applicability Ranges

Method Applicability Range Description

10 ≤ 𝛾 ≤ 20
For all connection types (see the
0.3 ≤ 𝛽 ≤ 0.8
Fessler notes below)
30° ≤ 𝜃 ≤ 90°
𝑔≥0 For K connections (no overlap)
10 ≤ 𝛾 ≤ 30
0.2 ≤ 𝛽 ≤ 1.0 𝑌, 𝑋 𝑜𝑟 𝐾 𝑤𝑖𝑡ℎ 𝑔 ≥ 0
{
0.3 ≤ 𝛽 ≤ 0.9 𝐾 𝑤𝑖𝑡ℎ 𝑔 < 0 For all connection types
Buitrago 30° ≤ 𝜃 ≤ 90°
0.25 ≤ 𝜏 ≤ 1.0
𝑔
0.02 < | | ≤ 0.5 For K connections
𝐷
10 ≤ 𝛾 ≤ 50
0.2 ≤ 𝛽 ≤ 1.0
30° ≤ 𝜃 ≤ 90° For all connection types
MSL 𝑘𝑁
𝐹𝑦𝑐 , 𝐹𝑦𝑏 ≤ 50
𝑐𝑚 2
𝑔
−0.6 < For K connections
𝐷
𝑘𝑁
Additional 𝐸𝑐 , 𝐸𝑏 > 2000 For all connection types (see notes
Checks 𝑐𝑚 2 below)
Brace offsets

The following comments should be noted regarding joint strength applicability ranges

✓ If a given K or X connection does not satisfy gap or eccentricity limits, then the program will
assume the joint acts as a Y/T connection.

✓ If a given connection does not satisfy brace/chord applicability ranges, then by default the
program excludes that joint from strength check calculation.

✓ Young’s modulus checks are enforced to ensure numerical stability.

✓ The program checks the braces offsets to ensure it is located outside of the chord. The default
tolerance for offset check is 0.1 cm or inch which can be revised on JSOPT input line.

✓ The joint applicability checks can be overridden by options available on CLPOP2 input line.

248

SACS® Collapse Advanced Release 15.1

6.10.2 Joint Flexibility Implementation
This section discusses the implementation of joint flexibility in the Collapse Advanced program. Joint
flexibility approach uses a transformation matrix mapping global quantities (like member end loads and
displacements) to local quantities associated with the joint flexibility formulation. The transformation
matrix can be expressed for member joints 1 and 2 as follows:
𝐓𝑖 𝟎
𝐗 𝑖 = 𝐄𝐫 [ ] 𝑤𝑖𝑡ℎ 𝑖 = 1,2 (311)
𝟎 𝐓𝑖
where 𝐄𝐫 is the permutation matrix that maps local DOF’s (translations and rotations) to the specific
joint flexibility DOF’s, namely axial (𝐴𝑋𝐿) displacement/force, in-plane (𝐼𝑃𝐵) rotation/bending and out-
of-plane (𝑂𝑃𝐵) rotation/bending.
1 0 0 0 0 0
0 0 0 0 1 0
𝐄𝐫 = 0 0 0 0 0 1
0 0 0 1 0 0
0 1 0 0 0 0
[0 0 1 0 0 0]
and 𝐓𝑖 is a 3x3 matrix that performs a global to joint-j brace transformation without mapping to the
joint-formulation system of DOF’s. The equations of transformation are summarized for both end-forces
as well as deflection increments below:
̂𝐛𝑖 = 𝐗𝑗 𝐪𝐛𝑖
𝐪
𝑤𝑖𝑡ℎ 𝑖 = 1,2 (312)
̂𝐛𝑖 = 𝐗𝑗 𝛅𝐩𝐛𝑖
𝛅𝐩

The next step to determine the relationship between the global brace end displacement vector 𝐩𝐛𝑖 with
the local joint distortion vector 𝐮
̂ 𝐣𝑖 . Utilizing equation (312) and assuming the chord deformation
remains unchanged,
∂𝐩𝐛𝑖 ∂𝐩𝐛𝑖 ∂𝐩
̂𝐛𝑖 𝐈
= = 𝐗 −𝟏
𝑖 [ ] 𝑤𝑖𝑡ℎ 𝑖 = 1,2 (313)
∂𝐮
̂ 𝐣𝑖 ∂𝐩
̂𝐛𝑖 ∂𝐮
̂𝐣𝑖 𝟎
̂𝐣𝑖 and
The following equation represents the constitutive relationship between local joint-brace forces 𝐪
joint-brace distortions 𝐮
̂ 𝐣𝑖 :
̂
𝐊
∂𝐪
̂ 𝐣𝑖
̂ 𝐣𝑖 [ 𝐈 ] 𝑤𝑖𝑡ℎ 𝑖 = 1,2
= [ 𝐣𝑖 ] = 𝐊 (314)
∂𝐮
̂ 𝐣𝑖 𝟎 𝟎

where 𝐊 ̂ 𝐣𝑖 is the 3x3 joint local stiffness matrix defined by the selected joint flexibility method. For
details, see following sections for Fessler (section 6.10.3), Buitrago (section 6.10.4) and MSL (section
6.11) joint flexibility methods.

The residual vector for the nonlinear analysis of the joint flexibility is expressed as the out-of-balance
local (joint-brace system) forces between brace and joint:
𝐫̂𝑖 = 𝐪
̂𝐣𝑖 + 𝐪
̂𝐛𝑖 𝑤𝑖𝑡ℎ 𝑖 = 1,2 (315)

A change in the residual owing to changes in the local joint distortions is derived by exploring the
differentials. For the joint-1, the change in residual can be formally written as:

249

SACS® Collapse Advanced Release 15.1

 ∂𝐪
̂𝐣𝟏 ∂𝐪
̂𝐛𝟏 ∂𝐪
̂𝐛𝟏
𝛅𝐫̂𝟏 = ∂𝐮
̂𝐣𝟏
𝛅𝐮
̂ 𝐣𝟏 + ∂𝐮
̂𝐣𝟏
𝛅𝐮
̂ 𝐣𝟏 + ∂𝐮
̂𝐣𝟐
𝛅𝐮
̂ 𝐣𝟐 (316)

A similar derivation is trivially available for joint-2.

∂𝐪
̂𝐣𝟐 ∂𝐪
̂𝐛𝟐 ∂𝐪
̂𝐛𝟐
𝛅𝐫̂𝟐 = ∂𝐮
̂𝐣𝟐
𝛅𝐮
̂ 𝐣𝟐 + ∂𝐮
̂𝐣𝟐
𝛅𝐮
̂ 𝐣𝟐 + ∂𝐮
̂𝐣𝟏
𝛅𝐮
̂ 𝐣𝟏 (317)

The brace’s global to global tangent stiffness matrix relates changes in the global brace end-forces to
changes in the global brace end-deflections, 𝐩𝐛𝟏 and 𝐩𝐛𝟐 .
𝛅𝐪𝐛𝟏 = 𝐊 𝐛𝟏𝟏 𝛅𝐩𝐛𝟏 + 𝐊 𝐛𝟏𝟐 𝛅𝐩𝐛𝟐
(318)
𝛅𝐪𝐛𝟐 = 𝐊 𝐛𝟐𝟏 𝛅𝐩𝐛𝟏 + 𝐊 𝐛𝟐𝟐 𝛅𝐩𝐛𝟐

The previously derived global to local transformation matrices and are employed to give:
∂𝐪
̂𝐛𝟏 ∂𝐩𝐛𝟏
= 𝐗 𝟏 𝐊 𝐛𝟏𝟏
∂𝐮
̂ 𝐣𝟏 ∂𝐮
̂𝐣𝟏
∂𝐪
̂𝐛𝟏 ∂𝐩𝐛𝟏
= 𝐗 𝟏 𝐊 𝐛𝟏𝟐
∂𝐮
̂ 𝐣𝟐 ∂𝐮
̂𝐣𝟐
∂𝐪
̂𝐛𝟐 ∂𝐩𝐛𝟐
(319)
∂𝐮
̂ 𝐣𝟏
= 𝐗 𝟐 𝐊 𝐛𝟐𝟏 ∂𝐮
̂𝐣𝟏
∂𝐪
̂𝐛𝟐 ∂𝐩𝐛𝟐
∂𝐮
̂ 𝐣𝟐
= 𝐗 𝟐 𝐊 𝐛𝟐𝟐 ∂𝐮
̂𝐣𝟐

For any iteration, a change in the residual should be aimed at reducing the out-of-balance forces to zero.
̃ = [𝐈 𝟎] (𝐫̂𝑖 + 𝛅𝐫̂𝑖 ) 𝑤𝑡𝑖ℎ 𝑖 = 1,2
𝟎 (320)

The tilde (~) has been used to denote quantities associated with DOF’s from the joint flexibility
formulation, namely axial (𝐴𝑋𝐿), in-plane (𝐼𝑃𝐵) and out-of-plane (𝑂𝑃𝐵). Substituting equations (316)-
(319) into equation (320),
̂ 𝐣𝟏 𝛅𝐮
𝛅𝐫̃𝟏 = −[𝐈 𝟎]𝐫̂𝟏 = 𝐊 ̂ 𝐣𝟏 + 𝐀𝟏𝟏 𝛅𝐮
̂ 𝐣𝟏 + 𝐀𝟏𝟐 𝛅𝐮
̂ 𝐣𝟐
(321)
̂ 𝐣𝟐 𝛅𝐮
𝛅𝐫̃𝟐 = −[𝐈 𝟎]𝐫̂𝟐 = 𝐊 ̂ 𝐣𝟐 + 𝐀𝟐𝟏 𝛅𝐮
̂ 𝐣𝟏 + 𝐀𝟐𝟐 𝛅𝐮
̂ 𝐣𝟐

where
𝐈
𝐀𝟏𝟏 = [𝐈 𝟎]𝐗 𝟏 𝐊 𝐛𝟏𝟏 𝐗 −𝟏
𝟏 [ ]
𝟎
𝐈
𝐀𝟏𝟐 = [𝐈 𝟎]𝐗 𝟏 𝐊 𝐛𝟏𝟐 𝐗 −𝟏
𝟐 [ ]
𝟎 (322)
𝐈
𝐀𝟐𝟏 = [𝐈 𝟎]𝐗 𝟐 𝐊 𝐛𝟐𝟏 𝐗 −𝟏
𝟏 [ ]
𝟎
𝐈
𝐀𝟐𝟐 = [𝐈 𝟎]𝐗 𝟐 𝐊 𝐛𝟐𝟐 𝐗 −𝟏
𝟐 [ ]
𝟎
The above equations can be summarized as follows:

𝛅𝐫̃𝟏 ̂ 𝐣𝟏
𝐀𝟏𝟏 + 𝐊 𝐀𝟏𝟐 𝛅𝐮
̂ 𝐣𝟏
൬ ൰=[ ]( ) (323)
𝛅𝐫̃𝟐 𝐀𝟐𝟏 ̂
𝐀𝟐𝟐 + 𝐊 𝐣𝟐 𝛅𝐮
̂ 𝐣𝟐

250

SACS® Collapse Advanced Release 15.1

The tangent stiffness matrix of the brace is given in the global co-ordinate system by 𝐊 𝐛
𝛅𝐪𝐛𝟏 𝛅𝐩
൬ ൰ = 𝐊 𝐛 ൬ 𝐛𝟏 ൰ (324)
𝛅𝐪𝐛𝟐 𝛅𝐩𝐛𝟐
̂ 𝐛 , is given by
The brace-joint local representation of the tangent stiffness matrix, 𝐊
̂ 𝐛 = 𝚽 𝐊 𝐛 𝚽−𝟏
𝐊 (325)

where
𝐗 𝟎
𝚽=[ 𝟏 ] (326)
𝟎 𝐗𝟐
Utilizing equations (324) and (325), the constitutive relationship between brace local end forces and
brace local end displacement is
̂𝐛𝟏
𝛅𝐪 ̂𝐛𝟏
̂ 𝐛 ൬ 𝛅𝐩
൬ ൰=𝐊 ൰ (327)
̂𝐛𝟐
𝛅𝐪 ̂𝐛𝟐
𝛅𝐩

The final step is to express constitutive relationship between combined displacement increment 𝛅𝐩 ̂𝐜𝑖
(brace displacement + the joint distortion) and combined end forces increment 𝛅𝒒 ̂𝐜𝑖 in local coordinate
system. The combine displacement increment vector is defined by
̂𝐛𝑖 = 𝛅𝐩
𝛅𝐩 ̂𝐜𝑖 + 𝛅𝐮
̂ 𝐣𝑖 𝑤𝑖𝑡ℎ 𝑖 = 1,2 (328)

Noting that
̂𝐛𝑖 = 𝛅𝐪
𝛅𝐪 ̂𝐜𝑖 = − 𝛅𝐪 ̂𝐣𝑖
̂ 𝐣𝑖 𝛅𝐮 𝑤𝑖𝑡ℎ 𝑖 = 1,2 (329)
̂𝐣𝑖 = 𝐊
𝛅𝐪 ̂ 𝐣𝑖

Substituting equation (329) in equation (328)

̂ −𝟏
𝐊 𝟎
𝛅𝐩 ̂𝐜𝑖 − [ 𝐣1
̂𝐛𝑖 = 𝛅𝐩 ] 𝛅𝐪
̂𝐜𝑖 𝑤𝑖𝑡ℎ 𝑖 = 1,2 (330)
𝟎 𝟎
Utilizing equation (327), the constitutive relationship for the combined system (brace + joint flexibility) is
given by
̂𝐜𝟏
𝛅𝐪 ̂𝐜𝟏
̂ 𝐛 ൬𝛅𝐩
̂ 𝐛 𝚿]−𝟏 𝐊
൬ ൰ = [𝐈 + 𝐊 ൰ (331)
̂𝐜𝟐
𝛅𝐪 ̂𝐜𝟐
𝛅𝐩
where
̂ −𝟏
𝐊 𝐣𝟏 𝟎 𝟎 𝟎
𝚿= 𝟎 𝟎 𝟎 𝟎 (332)
𝟎 ̂ −𝟏
𝟎 𝐊 𝐣𝟐 𝟎
[ 𝟎 𝟎 𝟎 𝟎]

251

SACS® Collapse Advanced Release 15.1

Equation (331) is for the brace for which the both ends have joint flexibility. For the cases where there is
only one end has the joint flexibility, equation (332) is modified as follows:
The joint flexibility only at joint 1 ( 𝛅𝐮 ̂𝐜𝟐 = 𝛅𝐩
̂ 𝐣𝟐 = 𝟎 𝑜𝑟 𝛅𝐩 ̂𝐛𝟐)
̂ −𝟏
𝐊 𝐣𝟏 𝟎 𝟎 𝟎
𝚿=[ 𝟎 𝟎 𝟎 𝟎] (333)
𝟎 𝟎 𝟎 𝟎
𝟎 𝟎 𝟎 𝟎
The joint flexibility only at joint 2 ( 𝛅𝐮 ̂𝐜𝟏 = 𝛅𝐩
̂ 𝐣𝟏 = 𝟎 𝑜𝑟 𝛅𝐪 ̂𝐛𝟏 )
𝟎 𝟎 𝟎 𝟎
𝟎 𝟎 𝟎 𝟎
𝚿=[ ̂ −𝟏 ] (334)
𝟎 𝟎 𝐊 𝐣𝟐 𝟎
𝟎 𝟎 𝟎 𝟎
6.10.3 Fessler Joint Flexibility
̂ 𝐣 in section 6.10.2 as a linear diagonal matrix
Fessler method [3] defines the local joint stiffness matrix 𝐊
in following form
𝐸𝑐 𝐷
0 0
𝑓𝐴𝑋𝐿
𝐸𝑐 𝐷 3
̂𝐣 =
𝐊 𝟎 0 (335)
𝑓𝐼𝑃𝐵
𝐸𝑐 𝐷3
[ 0 0 𝑓𝑂𝑃𝐵 ]

in which the joint flexibility parameters are defined below:

𝑓𝐴𝑋𝐿 = 1.95𝛾 2.15 𝑠𝑖𝑛(𝜃)2.19 (1 − 𝛽)1.3
𝑓𝐼𝑃𝐵 = 134𝛾1.73 𝑠𝑖𝑛(𝜃)1.22 𝑒 −4.52𝛽 (336)
𝑓𝑂𝑃𝐵 = 85.5𝛾 2.20 𝑠𝑖𝑛(𝜃)2.16 𝑒 −3.85𝛽
6.10.4 Buitrago Joint Flexibility
̂ 𝐣 in section 6.10.2 as a linear diagonal
Buitrago method [5] defines the local joint stiffness matrix 𝐊
matrix in following form
𝐸𝑐 𝐷
𝑓𝐴𝑋𝐿
0 0
𝐸𝑐 𝐷3
̂𝐣 =
𝐊 𝟎 0 (337)
𝑓𝐼𝑃𝐵
𝐸𝑐 𝐷3
[ 0 0 𝑓𝑂𝑃𝐵 ]

in which the joint flexibility parameters are defined below:

𝑓𝐴𝑋𝐿 = 5.69𝜏 −0.111 𝑒 −2.251𝛽 𝛾1.898 sin(𝜃)1.769
𝑓𝐼𝑃𝐵 = 1.39𝜏 −0.238 𝛽−2.245 𝛾1.898 sin(𝜃)1.240 (338)
𝑓𝑂𝑃𝐵 = 55𝜏 −0.220 𝑒 −4.076𝛽 𝛾 2.417 sin(𝜃)1.1883

252

SACS® Collapse Advanced Release 15.1

6.11 MSL Joint Flexibility
This section presents the MSL formulation for various joint strength check methods supported by the
program.

The following notation is used to describe the MSL method:

𝐴𝑋𝐿 axial tension or compression displacement or load
𝐼𝑃𝐵 in-plane bending rotation or moment
𝑂𝑃𝐵 out-of-plane bending rotation or moment
𝜃 brace angle
𝑔 gap for K connection
𝑒 eccentricity for K and X connections
𝑡 brace wall thickness
𝑇 chord wall thickness
𝑑 brace outer diameter
𝐷 chord outer diameter
𝑡
𝜏 brace wall thickness to chord wall thickness ratio, 𝑇
𝑑
𝛽 brace diameter to chord diameter ratio, 𝐷
𝐷
𝛾 chord radius to chord wall thickness ratio,
2𝑇
𝐹𝑦𝑏 brace yield stress
𝐹𝑦𝑐 chord yield stress
𝑄𝑓 chord force factor
𝑍𝑏 , 𝑍𝑐 brace and chord plastic section modulus
𝑆𝑏 , 𝑆𝑐 brace and chord elastic section modulus

𝑃, 𝑀 brace axial force and moment, respectively

𝑃𝑢𝑗 joint ultimate capacity for axial load
𝑀𝑢𝑗 joint ultimate capacity for bending moment
𝜑𝑗 user-defined factors given on RSFAC or RSFACO input lines
𝜑𝑐 MSL characteristic bias factors
𝛤 MSL plastic interaction function (𝛤 = 0 is the yield surface)
𝛿, 𝜃 joint axial and rotational distortion (deformation)
𝛎 joint distortion (deformation) vector (𝐴𝑋𝐿, 𝐼𝑃𝐵, 𝑂𝑃𝐵)
𝐒 joint internal force vector (𝐴𝑋𝐿, 𝐼𝑃𝐵, 𝑂𝑃𝐵)
𝑅 joint strain hardening function
𝐑 joint strain hardening vector (𝐴𝑋𝐿, 𝐼𝑃𝐵, 𝑂𝑃𝐵)
Δ incremental change to a quantity
d iterative change to a quantity

253

SACS® Collapse Advanced Release 15.1

6.11.1 MSL Plastic Interaction Function
MSL joint flexibility utilizes a plastic interaction function defined following equation to determine the
state of the joint:
𝛼1 𝛼3 𝛼3 𝛼5
𝑃 𝑀𝐼𝑃𝐵 𝑀𝑂𝑃𝐵
𝛤(𝑃, 𝑀𝐼𝑃𝐵 , 𝑀𝑂𝑃𝐵 ) = (𝑅 ) + ( (𝑅 ) + (𝑅 ) ) −1 (339)
𝑃 𝑃0 𝐼𝑃𝐵 𝑀0 𝑂𝑃𝐵 𝑀0

where 𝑃0 and 𝑀0 are brace elastic limit for axial force and bending moment (i.e. 𝐴𝑏 𝐹𝑏𝑦 and 𝑆𝑏 𝐹𝑏𝑦 ), 𝑅𝑃 ,
𝑅𝐼𝑃𝐵 , and 𝑅𝑂𝑃𝐵 are strain hardening functions (defined in section 6.11.4). Based on recommendation in
MSL documentation [6], the program uses following α-parameters:
𝛼1 = 1.0 𝛼3 = 2.0 𝛼4 = 2.0 𝛼5 = 1.0 (340)

The program computes the interaction function equation (339) at each load increment, and for the
elastic state 𝛤(𝑃, 𝑀𝐼𝑃𝐵 , 𝑀𝑂𝑃𝐵 ) < 0, the 𝐊̂ 𝐣 is determined using uncoupled nonlinear elastic
̂ 𝐣 is
formulation (discussed in section 6.11.2) and, for the plastic state 𝛤 (𝑃, 𝑀𝐼𝑃𝐵 , 𝑀𝑂𝑃𝐵 ) = 0, the 𝐊
determined using coupled nonlinear elastoplastic formulation (discussed in section 6.11.4).

6.11.2 MSL Uncoupled Joint Flexibility

̂ 𝐣 when
The section presents MSL formulation to determine the uncoupled local joint stiffness matrix 𝐊
𝛤(𝑃, 𝑀𝐼𝑃𝐵 , 𝑀𝑂𝑃𝐵 ) < 0. MSL uncoupled joint force-deflection can be written as follows
2
𝑃 (𝛿 ) = 𝑑 − 𝑎(1 − 𝑏𝑒 −𝑐𝛿 )
2 (341)
𝑀(𝜃) = 𝑑 − 𝑎(1 − 𝑏𝑒 −𝑐𝜃 )

where parameters, 𝑎, 𝑏, 𝑐, and 𝑑 are functions of joint material properties, geometry, classification
(Y/X/K), and the load type (axial tension or compression, in-plane bending, out-of-plane bending). The
tangent stiffness and the initial stiffness for force-deflection equation (341) may be written as follows
𝑑𝑃
𝐾𝑃 (𝛿 ) = 𝑑𝛿 = 2𝑎𝑏𝑐(𝑏𝑒 −𝑐𝛿 − 1)𝑒 −𝑐𝛿
𝑑𝑀 (342)
𝐾𝑀 (𝜃) = 𝑑𝜃
= 2𝑎𝑏𝑐(𝑏𝑒 −𝑐𝜃 − 1)𝑒 −𝑐𝜃

𝐾𝑖𝑛𝑡 = 2𝑎𝑏𝑐(𝑏 − 1) (343)

Utilizing above expressions for the tangent stiffness, the uncoupled MSL joint stiffness matrix is defined
as
𝐾𝐴𝑋𝐿 (𝛿 ) 0 0
̂𝐣 = [
𝐊 0 𝐾𝐼𝑃𝐵 (𝜃) 0 ] (344)
0 0 𝐾𝑂𝑃𝐵 (𝜃)

Next section provides details formulation for MSL joint flexibility coefficients.

254

SACS® Collapse Advanced Release 15.1

 6.11.3 MSL Joint Flexibility Coefficients
As stated in MSL documentation [6], only three independent parameters are possible. Therefore
(equation 3.2 in [6])
𝑑
𝑏 = 1 + √𝑎 (345)

Following MSL notation, the force-deflection parameters can be defined as follows

For axial force

𝐵
𝑎 = 𝜑𝑗 𝜑𝑐 𝑃𝑢𝑗 𝐴 𝑐 = 𝑑 = 𝜑𝑗 𝜑𝑐 𝑃𝑢𝑗 (346)
𝜑𝑗𝜑𝑐 𝑄𝑓𝐹𝑦𝑐 𝐷

For bending moment

𝐵
𝑎 = 𝜑𝑗 𝜑𝑐 𝑀𝑢𝑗 𝐴 𝑐 = 𝜑 𝑑 = 𝜑𝑗 𝜑𝑐 𝑀𝑢𝑗 (347)
𝑗𝜑𝑐 𝑄𝑓 𝐹𝑦𝑐

where 𝜑𝑐 are 1.0 for MSL Mean Level and they are given in Table 17 for MSL Characteristic Level. The
user-defined factors 𝜑𝑗 have default value of 1.0 and they can be revised on RSFAC or RSFACO input
lines. The ultimate capacities 𝑃𝑢𝑗 and 𝑀𝑢𝑗 and the chord factor 𝑄𝑓 are given in section 6.9.5. The MSL
parameters 𝐴 and 𝐵 are defined for various joint types and load conditions in Table 22.

Table 22: MSL coefficients 𝐴 and 𝐵 based on MSL C20400R014 Table 3.2

Joint
Axial Tension Axial Compression IPB OPB
Type
(𝛾 − 4)(sin(𝜃))3
𝐴 0.001 0.001 0.001
T/Y 62
𝐵 12000𝛽 + 1200 600𝛽 + 13500 9700𝛽 + 6700 8600𝛽 + 1200
𝛾 + 10
𝐴 0.001 0.001 0.001
100
X 3900𝛽 + 5000 𝛽 ≤ 0.9
𝐵 { 6000000 90000𝛽𝛾 −0.4 9700𝛽 + 6700 8600𝛽 + 1200
8510 + (𝛽 − 0.9) ൬ − 80000൰ 0.9 < 𝛽 ≤ 1.0
𝛾
𝜙(𝛾 − 7) 𝜙(𝛾 − 7)
18 18
𝐴 𝑔 𝑔 0.001 0.001
𝜙 = − 0.1 𝜙 = − 0.1
𝐷 𝐷
K 0.025 ≤ 𝜙 ≤ 0.25 0.025 ≤ 𝜙 ≤ 0.25
(13 + 4𝛾)𝜓 (13 + 4𝛾)𝜓
𝑔 𝑔
𝐵 𝜓 = 320 − 450 𝜓 = 320 − 450 9700𝛽 + 6700 8600𝛽 + 1200
𝐷 𝐷
170 ≤ 𝜓 ≤ 320 170 ≤ 𝜓 ≤ 320

255

SACS® Collapse Advanced Release 15.1

For a joint with mixed classification, the program follows MSL recommendations (MSL C20400R014
section 4.3.3) to determine the mixed flexibility parameters. The classification weighting factors are
defined as 𝐶𝑌 , 𝐶𝑋 , and 𝐶𝐾 with 𝐶𝑌 + 𝐶𝑋 + 𝐶𝐾 = 1 and the mixed coefficients are determined as follow:

𝑎𝑚𝑖𝑥𝑒𝑑 = √𝐶𝑌 𝑎𝑌2 + 𝐶𝑋 𝑎𝑋2 + 𝐶𝐾 𝑎𝐾2

𝑑𝑚𝑖𝑥𝑒𝑑 = 𝐶𝑌 𝑑𝑌 + 𝐶𝑋 𝑑𝑋 + 𝐶𝐾 𝑑𝐾
𝑑 (348)
𝑏𝑚𝑖𝑥𝑒𝑑 = 1 + √𝑎𝑚𝑖𝑥𝑒𝑑
𝑚𝑖𝑥𝑒𝑑
(𝐾𝑖𝑛𝑖𝑡 )𝑚𝑖𝑥𝑒𝑑
𝑐𝑚𝑖𝑥𝑒𝑑 = 2𝑎
𝑚𝑖𝑥𝑒𝑑 𝑏𝑚𝑖𝑥𝑒𝑑 (𝑏𝑚𝑖𝑥𝑒𝑑 −1)

where the initial mixed stiffness (𝐾𝑖𝑛𝑖𝑡 )𝑚𝑖𝑥𝑒𝑑 is defined as

1 𝐶𝑌 𝐶𝑋 𝐶𝐾
(𝐾𝑖𝑛𝑖𝑡 )𝑚𝑖𝑥𝑒𝑑
= (𝐾 + (𝐾 + (𝐾 (349)
𝑖𝑛𝑖𝑡 )𝑌 𝑖𝑛𝑖𝑡 )𝑋 𝑖𝑛𝑖𝑡 )𝐾

The mixed ultimate capacities given by the equation above are separately computed for axial tension,
axial compression, in-plane bending and out-of-plane bending components.

6.11.4 MSL Coupled Joint Flexibility

6.11.4.1 Strain Hardening
The first step of the coupled elastoplastic MSL joint flexibility is to define the strain hardening functions
as
𝑃(𝛿) 1 2
𝑅𝑃 (𝛿 ) = 𝑃 = 𝑃 (𝑑 − 𝑎(1 − 𝑏𝑒 −𝑐𝛿 ) )
𝑏0 𝑏0
𝑀(𝜃) 1 2 (350)
𝑅𝑀 (𝜃) = = (𝑑 − 𝑎(1 − 𝑏𝑒 −𝑐𝜃 ) )
𝑀𝑏0 𝑀𝑏0

where 𝑃0 and 𝑀0 are brace elastic limit for axial force and bending moment
𝑃𝑏0 = 𝐴𝑏 𝐹𝑏𝑦
(351)
𝑀𝑏0 = 𝑆𝑏 𝐹𝑏𝑦

and the flexibility coefficients are given in section 6.11.3.

The MSL elastoplastic model assumes the following behavior that is typical of all elastoplastic with strain
hardening models. Under plastic action, the total displacement increment, 𝚫𝛎, can be split into elastic
and plastic components.
𝚫𝛎 = 𝚫𝛎𝐞 + 𝚫𝛎𝐩 (352)

where 𝛎 is defined as follows:

𝛿
𝛎 = ( 𝜃𝐼𝑃𝐵 ) (353)
𝜃𝑂𝑃𝐵

256

SACS® Collapse Advanced Release 15.1

The elastoplastic formulation requires that under plastic action, the displacement increments satisfy the
following tangent stiffness behavior.
𝛥𝑃
𝚫𝐒 = ( 𝛥𝑀𝐼𝑃𝐵 ) = diag(𝑘𝐸,𝑃 , 𝑘𝐸,𝐼𝑃𝐵 , 𝑘𝐸,𝑂𝑃𝐵 ) 𝚫𝛎𝐞 = 𝐊 𝐄 𝚫𝛎𝐞
𝛥𝑀𝑂𝑃𝐵
(354)
𝛥𝑅𝑃
𝚫𝐑 = ( 𝛥𝑅𝐼𝑃𝐵 ) = diag(𝑘𝐻,𝑃 , 𝑘𝐻,𝐼𝑃𝐵 , 𝑘𝐻,𝑂𝑃𝐵 ) 𝚫𝛎𝐩 = 𝐊 𝐇 𝚫𝛎𝐩
𝛥𝑅𝑂𝑃𝐵

In the first expression, the elastic stiffness 𝐊 𝐄 components are defined as the secant stiffnesses of the
individual (uncoupled) modes of deformation at the onset of yielding:
𝑃
𝑘𝐸,𝑃 = 𝛿𝑏0
0
𝑀𝑏0
𝑘𝐸,𝐼𝑃𝐵 = (355)
𝜃𝐼𝑃𝐵,0
𝑀𝑏0
𝑘𝐸,𝑂𝑃𝐵 = 𝜃𝑂𝑃𝐵,0

where δ0 , θIPB,0 and θOPB,0 are joint distortion associated with the brace elastic limits 𝑃𝑏0 and 𝑀𝑏0 and
they can be determined by solving equation (341). The second expression is sometimes referred to as a
hardening rule and the component of the hardening stiffness 𝐊 𝐇 can be determined utilizing following
expression [6]
1 1 1
𝑑𝑅 = + (356)
𝑘𝐸 𝑘𝐻
𝑑𝛿

Thus,
−1
𝑃𝑏0 1
𝑘𝐻,𝑃 = ൬𝐾 −𝐾 ൰
𝑃 (𝛿) 𝐸,𝑃
−1
𝑀𝑏0 1
𝑘𝐻,𝐼𝑃𝐵 = ൬ − ൰ (357)
𝐾𝐼𝑃𝐵 (𝜃) 𝐾𝐸,𝐼𝑃𝐵
−1
𝑀𝑏0 1
𝑘𝐻,𝑂𝑃𝐵 = ൬𝐾 −𝐾 ൰
𝑂𝑃𝐵 (𝜃) 𝐸,𝑂𝑃𝐵

in which 𝐾𝑃 (𝛿 ), 𝐾𝐼𝑃𝐵 (𝜃), 𝐾𝑂𝑃𝐵 (𝜃) are given in equation (342).

6.11.4.2 Consistency Rule

The consistency rule is expressed such that the internal loads should remain on the interaction (yield)
surface. This is stated in differential form as:
∂Γ ∂Γ
ΔΓ = ∂𝐒
Δ𝐒 + ∂𝐑
Δ𝐑 = 𝟎 (358)

where
∂𝛤 𝜕𝛤 𝜕𝛤 𝜕𝛤 𝑇
∂𝐒
= (𝜕𝑂 , 𝜕𝑀 , 𝜕𝑀 )
𝐼𝑃𝐵 𝑂𝑃𝐵
𝑇 (359)
∂𝛤 𝜕𝛤 𝜕𝛤 𝜕𝛤
= ( , , )
∂𝐑 𝜕𝑅𝑃 𝜕𝑅𝐼𝑃𝐵 𝜕𝑅𝑂𝑃𝐵

257

SACS® Collapse Advanced Release 15.1

6.11.4.3 Flow Rule
The MSL plasticity formulation for joints employs the flow rule for associative flow, which states that the
plastic deformations are normal to the yield surface, multiplied by a factor, Δλ , called the plastic strain
rate multiplier.
∂𝛤
𝚫𝛎𝐩 = Δλ (360)
∂𝐒

6.11.4.4 Elastoplastic Stiffness Matrix

The joint elastoplastic stiffness matrix can now be formulated such that

̂ 𝐣 = 𝐝𝐒
𝐊 (361)
𝐝𝛎

The starting point is the variation in internal loads due to changes in the elastic deformations

𝚫𝐒 = 𝐊 𝐄 𝚫𝛎𝐞 (362)

where

𝚫𝛎𝐞 = 𝚫𝛎 − 𝚫𝛎𝐩 (363)

Combining this with the flow rule,

∂𝛤
𝚫𝐒 = 𝐊 𝐄 (𝚫𝛎 − 𝚫𝛎𝐩 ) = 𝐊 𝐄 𝚫𝛎 − Δλ𝐊 𝐄 (364)
∂𝐒

Invoking the hardening rule in equation (354), and inserting into the consistency rule in equation (358)
∂𝛤 𝑇 ∂𝛤 𝑇
( ) Δ𝐒 + ( ) 𝐊 𝐇 𝚫𝛎𝐩 = 0 (365)
∂𝐒 ∂𝐑

Substituting 𝚫𝛎𝐩 by the flow rule in equation (360) and 𝚫𝐒 from equation (364) in above equation, Δλ
can be determined in following steps
∂𝛤 𝑇 ∂𝛤 ∂𝛤 𝑇 ∂𝛤
( ) (𝐊 𝐄 𝚫𝛎 − Δλ𝐊 𝐄 ) + Δλ ( ) 𝐊 𝐇 =0
∂𝐒 ∂𝐒 ∂𝐑 ∂𝐒
∂𝛤 𝑇 ∂𝛤 ∂Γ 𝑇 ∂𝛤 ∂𝛤 𝑇
Δλ ൬(∂𝐑) 𝐊 𝐇 ∂𝐒 − ( ∂𝐒 ) 𝐊 𝐄 ∂𝐒 ൰ = − ( ∂𝐒 ) 𝐊 𝐄 𝚫𝛎
(366)
∂𝛤 𝑇
( ) 𝐊 𝐄 𝚫𝛎
∂𝐒
Δλ = ∂𝛤 𝑇 ∂𝛤 ∂𝛤 𝑇 ∂𝛤
( ) 𝐊𝐄 − ( ) 𝐊𝐇
∂𝐒 ∂𝐒 ∂𝐑 ∂𝐒

Substituting Δλ into equation (364), the elastoplastic joint stiffness becomes

∂𝛤 ∂𝛤 𝑇
( )( ) 𝐊 𝐄
𝚫𝐒 = 𝐊 𝐄 (𝐈− ∂𝐒 ∂𝐒 ̂ 𝐣 𝚫𝛎
) 𝚫𝛎 = 𝐊 (367)
∂𝛤 𝑇 ∂𝛤 ∂𝛤 𝑇 ∂𝛤
( ) 𝐊𝐄 − ( ) 𝐊𝐇
∂𝐒 ∂𝐒 ∂𝐑 ∂𝐒

258

SACS® Collapse Advanced Release 15.1

6.11.4.5 Integration
The work of this section assumes that we have an initial internal force state, 𝐒𝟎 , lies on the yield surface
such that

𝛤(𝐒𝟎 ) = 0 (368)

The internal force state is associated with an initial deformation state, 𝛎𝟎 . For an applied total
deformation increment, 𝚫𝛎, it is desirable to find a new internal force increment, 𝚫𝐒, such that

𝛤(𝐒𝐧 ) = 0 (369)

Together with
𝛎𝐧 = 𝛎𝟎 + 𝚫𝛎
(370)
𝐒𝐧 = 𝐒(𝛎𝐧 ) = 𝐒𝟎 + 𝚫𝐒
whilst maintaining the previously developed rules of the MSL strain hardening formulation. The starting
point is an elastic predictor that essentially keeps that hardening function stationary and sets the
incremental plastic deformation to zero.
𝚫𝐒 = 𝐊 𝐄 𝚫𝛎
(371)
𝚫𝐑 = 𝟎
The idea here is to make a sequence of changes to 𝛤 such that its final value is close to zero. This will
ensure that the final (converged) internal force state lies on the interaction surface.
Γ (𝑘) = Γ(𝐒(𝑘) ) → 0
(372)
𝐒 (𝑘+1) = 𝐒(𝑘) + 𝐝𝐒(𝑘)
where 𝐝 denotes the iterative change to the quantity. The first value of 𝛤 is given by the elastic
predictor
Γ (1) = Γ(𝐒𝟎 + 𝚫𝐒) = Γ(𝐒(0) + 𝐝𝐒 (0) )
(373)
𝐝𝐒 (0) = 𝐊 𝐄 𝚫𝛎

In general, given an internal force state 𝐒 (𝑘) at the kth iteration, we would like to find 𝐝𝐒(𝑘) such that
Γ(𝐒(𝑘+1) ) is zero, or close to zero. A basic gradient technique is applied. The derivatives of the
consistency criterion are invoked
𝑇 𝑇
(𝑘+1) (𝑘) ∂𝛤 (𝑘) (𝑘) ∂𝛤 (𝒌)
𝛤 = 𝛤 + ൬ ൰ 𝐝𝐒 + ൬ ൰ 𝐝𝐑(𝑘) = 0 (374)
∂𝐒 ∂𝐑

However, the work of the previous section has shown that changes in the internal force state as well as
the hardening functions are related to changes in the plastic strain rate multiplier as follows

∂𝛤 (𝑘)
𝐝𝐒 (𝑘) = 𝐊 𝐄 𝐝𝛎𝐞 = 𝐊 𝐄 (𝐝𝛎 − 𝐝𝛎𝐩 ) = − 𝐊 𝐄 𝐝𝛎𝐩 = − 𝐊 𝐄 ∂𝐒 d(Δλ ) (375)

which assumes that the total applied deformation 𝚫𝛎 remains constant (i.e. 𝐝𝛎 = 𝟎) over the
increment.

Furthermore, changes in the hardening functions are related to a change in the plastic strain rate
multiplier as follows
∂𝛤 (𝑘)
𝐝𝐑(𝑘) = 𝐊 𝐇 ∂𝐒
d(Δλ ) (376)

259

SACS® Collapse Advanced Release 15.1

Substituting the above expression in equation (374), and solving for d(Δλ )
− 𝛤 (𝑘)
d(Δλ ) = 𝑇 𝑇 (377)
∂𝛤(𝑘) ∂𝛤(𝑘) ∂𝛤(𝑘) ∂𝛤(𝑘)
( ) 𝐊𝐇 − ( ) 𝐊𝐄
∂𝐑 ∂𝐒 ∂𝐒 ∂𝐒

The internal force states and the hardening functions can now be updated using the freshly calculated
value of d(Δλ )
∂𝛤 (𝑘)
𝐒 (𝑘+1) = 𝐒(𝑘) − 𝐊 𝐄 d(Δλ )
∂𝐒
(378)
∂𝛤 (𝑘)
𝐑(𝑘+1) = 𝐑(𝑘) + 𝐊 𝐇 ∂𝐒
d(Δλ )

6.11.4.6 Unloading
During unloading of a joint, the program reverts to an uncoupled joint stiffness where the joint
distortion is now calculated with respect to the permanent plastic deformation computed in previous
section. Therefore, the joint stiffness during unloading pahse becomes:

𝐾𝐴𝑋𝐿 (𝛿 − 𝛿𝑝 ) 0 0
̂𝐣 = [
𝐊 0 𝐾𝐼𝑃𝐵 (𝜃 − 𝜃𝑝 ) 0 ] (379)
0 0 𝐾𝑂𝑃𝐵 (𝜃 − 𝜃_𝑝 )
where 𝛿𝑝 and 𝜃𝑝 represent joint plastic deformations.

260

SACS® Collapse Advanced Release 15.1

6.12 Local Buckling
This section summarizes the local buckling limits calculation for Marshall and Gates [9], API Bulleting 2U,
API RP 2A-LRFD, and ISO 19902.
The following notation is used to describe the local buckling method:
𝑇 wall thickness at the cross-section under consideration
𝐷 outer diameter at the cross-section under consideration
𝛼 half of the apex angle for a conical section
𝐹𝑦 yield stress
𝐸 Young’s modulus
𝐹𝑙𝑏 local buckling stress limit
𝐹𝑒𝑙𝑏 elastic local buckling stress limit
𝐹𝑖𝑒𝑙𝑏 inelastic local buckling stress limit
𝜎𝑎 axial stress at the cross-section under consideration
𝜎𝑏 bending stress at the cross-section under consideration
𝑃 compression axial load at the cross-section under consideration
𝑀 bending moment load at the cross-section under consideration

6.12.1 Marshall and Gates

The Marshall and Gates lower bound limit is defined:

16
𝐹𝑙𝑏 = E 𝐷 2 ≤ 𝐹𝑦 (380)
( )
𝑇

Note: For conical section, the larger diameter of the cone is used to calculate the local buckling stress
limit.

6.12.2 API Bulletin 2U

API Bulletin 2U elastic local buckling limit is defined by:
𝑇
𝐹𝑒𝑙𝑏 = 𝛼𝑥𝐿 𝐶𝑥 2𝐸 𝐷 (381)
169
where 𝐶𝑥 = 0.605 and 𝛼𝑥𝐿 = 𝐷 . The inelastic limit is given by:
195+0.5
𝑇

233 𝐷
𝐷 𝐹𝑦 ≤ 𝐹𝑦 𝑇
< 600
166+0.5
𝐹𝑖𝑒𝑙𝑏 = { 𝑇 (382)
𝐷
0.5𝐹𝑦 𝑇
≥ 600

and API Bulletin 2U local buckling stress limit is

𝐹𝑙𝑏 = min(𝐹𝑒𝑙𝑏 , 𝐹𝑖𝑒𝑙𝑏 ) (383)

Note: For conical section, the larger diameter of the cone is used to calculate the local buckling stress
limit.

261

SACS® Collapse Advanced Release 15.1

6.12.3 API RP 2A-LRFD
API RP 2A-LRFDelastic local buckling limit is defined by:
𝑇
𝐹𝑒𝑙𝑏 = 2𝐶𝑥 𝐸 𝐷 (384)

where 𝐶𝑥 = 0.3. The inelastic limit is given by:

𝐷
𝐹𝑦 ≤ 60
𝑇
1
𝐹𝑖𝑒𝑙𝑏 = { 𝐷 4 𝐷
(385)
[1.64 − 0.23 ( ) ] 𝐹𝑦 > 60
𝑇 𝑇

The API RP 2A-LRFD local buckling stress limit is

𝐹𝑙𝑏 = min(𝐹𝑒𝑙𝑏 , 𝐹𝑖𝑒𝑙𝑏 ) (386)

6.12.3.1 API RP 2A-LRFD: Conical Members

𝐷
For conical members, API RP 2A-LRFD recommends using an equivalent diameter of to obtain the
cos(𝛼)
local buckling limits from equations (384) to (386) (API RP 2A-LRFD section D.4.1.2). In addition, the
equivalent compression stress at cross section under consideration is therefore (API RP 2A-LRFD section
D.4.1.1):
σ +σ
𝜎= a b (387)
cos(𝛼)

API LRFD local buckling is only carried out for conical section with apex angle less than 60° (𝛼 < 30°)

6.12.4 ISO 19902

ISO 19902 local buckling formulation is valid within following ranges
𝑇 ≥ 6 𝑚𝑚
𝐷
𝑇
≤ 120
𝑘𝑁 (388)
𝐹𝑦 ≤ 50 𝑐𝑚 2
𝛼 < 30 𝑓𝑜𝑟 𝑐𝑜𝑛𝑖𝑐𝑎𝑙 𝑚𝑒𝑚𝑏𝑒𝑟𝑠
The program does not perform local bucking checks for sub-segments outside above validity ranges. ISO
19902 elastic local buckling limit is defined by:
𝑇
𝐹𝑒𝑙𝑏 = 2𝐶𝑥 𝐸 𝐷 (389)

where 𝐶𝑥 = 0.3. The inelastic limit is given by

𝐹𝑦
𝐹𝑦 𝐹𝑒𝑙𝑏
≤ 0.170
𝐹𝑖𝑒𝑙𝑏 = { 𝐹𝑦 𝐹𝑦 (390)
[1.047 − 0.274 𝐹 ] 𝐹𝑦 𝐹𝑒𝑙𝑏
> 0.170
𝑒𝑙𝑏

The ISO 19902 local buckling stress limit given by:

𝐹𝑙𝑏 = min(𝐹𝑒𝑙𝑏 , 𝐹𝑖𝑒𝑙𝑏 ) (391)

262

SACS® Collapse Advanced Release 15.1

6.12.4.1 ISO 19902: Conical Members
𝐷
For conical members, ISO 19902 recommends using the equivalent diameter of to obtain the local
cos(𝛼)
buckling limits from equations (389) to (390) (ISO 19902 section 13.6.3.2). Also, the local buckling stress
limit from equation (390) should be reduced by a factor 𝛾𝑅,𝑐 with default value of 1.18.
In addition, the equivalent compression stress at cross section under consideration is given by (ISO
19902 section 13.6.2.1):
σ +σ
𝜎= a b (392)
cos(𝛼)

where
𝑃 4𝑀
𝜎𝑎 = 𝜋(𝐷−𝑇𝑐𝑜𝑠(𝛼))𝑇 𝜎𝑏 = 2 (393)
𝜋(𝐷−𝑇𝑐𝑜𝑠(𝛼)) 𝑇

6.12.5 Implementation
The program checks for local buckling at mid-point of all subsegments associated with tubular and
conical sections. For example, for a member with eight subsegments, the local buckling will be checked
at eight cross-sections points along the member. The local buckling limit is based on the material
properties and the dimensions of a cross-section under consideration. It is worth noting that if the local
buckling critical stress is capped by the yield stress, the program will not check the subsegment against
local buckling since the member failure mechanism is governed by yielding.

Once local buckling occurs for a given subsegment, the subsegment loses its rotational stiffness while
the axial stiffness remains unchanged. To model the loss of the rotational stiffness, the program
combines the subsegment finite element stiffness with two rotational linear springs with a stiffness of:
1 𝐸𝐼𝑦𝑦 1 𝐸𝐼𝑧𝑧
𝐾𝜃𝑦 = 100 𝐿
𝐾𝜃𝑧 = 100 𝐿
(394)
𝐸𝐼
where 𝐿 is the subsegment length, 𝐼 is second moment of the area, and 𝐿
represents elastic rotational
stiffness of the subsegment. Equation (394) provides reasonable loss of the rotational stiffness due to
the local buckling while preventing numerical singularity. Figure 199 illustrates the effect of equation
(394) the cantilever beam in section 5.4.1.

Figure 199: Response of the cantilever beam under local buckling

263

SACS® Collapse Advanced Release 15.1

7 REFERENCES
[1] M. A. Crisfield, Non-Linear Finite Element Analysis of Solids and Structures. Wiley, 1996.
[2] M. A. Crisfield, Advanced Topics, Volume 2, Non-Linear Finite Element Analysis of Solids and
Structures, Volume 2 edition. Chichester ; New York: Wiley, 1997.
[3] H. Fessler, P. Mockford, and J. Webster, “Parametric equations for the flexibility matrices of single
brace tubu lar joints in offshore structures.,” Proc. Inst. Civ. Eng., vol. 81, no. 4, pp. 659–673, Dec.
1986.
[4] H. Fessler, P. Mockford, J. Webster, and W. Manners, “Parametric equations for the flexibility
matrices of multi-brace tubul ar joints in offshore structures. discussion on paper 9124,” Proc. Inst.
Civ. Eng., vol. 83, no. 4, pp. 887–888, Dec. 1987.
[5] J. Buitrago, B. E. Healy, and T. Y. Chang, “LOCAL JOINT FLEXIBILITY OF TUBULAR JOINTS,” 1993.
[6] MSL Engineering Limited, “JIP - Assessment Criteria, Reliability and Reserve Strength of Tubular
Joints (Phase II),” MSL Engineering Limited, MSL Document Reference C20400R014 Rev 0, Jul.
2000.
[7] American Petroleum Institute, API RP 2A-LRFD, 1st ed. 1993.
[8] Design of Steel Strucutres, Norsok Standard N-004, Rev 1. December 19998.
[9] P. W. Marshall, W. E. Gates, and S. W. Anagnostopoulos, “Inelastic Dynamic Analysis Of Tubular
Offshore Structures,” presented at the Offshore Technology Conference, 1977.
[10] S. P. T. and W. Krieger, Theory of Plates and Shells Second Edition, 2nd edition. McGraw-Hill
Publishing Co., 1959.
[11] F. W. Williams, “AN APPROACH TO THE NON-LINEAR BEHAVIOUR OF THE MEMBERS OF A RIGID
JOINTED PLANE FRAMEWORK WITH FINITE DEFLECTIONS,” Q. J. Mech. Appl. Math., vol. 17, no. 4,
pp. 451–469, Nov. 1964.
[12] “Theory of Elastic Stability - Stephen P. Timoshenko, James M. Gere - Google Books.” [Online].
Available:
https://books.google.com/books/about/Theory_of_Elastic_Stability.html?id=98B6JOW2HiUC.
[Accessed: 08-Aug-2017].
[13] L. L. Yaw, Co-rotational Meshfree Formulation for Large Deformation Inelastic Analysis of Two-
dimensional Structural Systems. University of California, Davis, 2008.
[14] K. Y. Sze, X. H. Liu, and S. H. Lo, “Popular benchmark problems for geometric nonlinear analysis of
shells,” Finite Elem. Anal. Des., vol. 40, no. 11, pp. 1551–1569, Jul. 2004.
[15] S. L. Chan, “Geometric and material non-linear analysis of beam-columns and frames using the
minimum residual displacement method,” Int. J. Numer. Methods Eng., vol. 26, no. 12, pp. 2657–
2669, Dec. 1988.
[16] R. M. D. Souza and R. M. D. Souza, “Force-based Finite Element for Large Displacement Inelastic
Analysis of Frames,” 2000.
[17] J. L. Meek and S. Loganathan, “Geometric and material non-linear behaviour of beam-columns,”
Comput. Struct., vol. 34, no. 1, pp. 87–100, Jan. 1990.
[18] M. S. Park and B. C. Lee, “Geometrically Non-Linear and Elastoplastic Three-Dimensional Shear
Flexible Beam Element of Von-Mises-Type Hardening Material,” Int. J. Numer. Methods Eng., vol.
39, no. 3, pp. 383–408, Feb. 1996.
[19] Z.G Azizian, Instability and nonlinear analysis of thin walled structures. University of Wales, Cardiff,
Wales, 1983.
[20] H. B. Harrison, The application of the principles of plastic analysis to three dimensional steel
structures. University of Sydney, Australia., 1964.

264

SACS® Collapse Advanced Release 15.1

[21] X.-M. Jiang, H. Chen, and J. Y. R. Liew, “Spread-of-plasticity analysis of three-dimensional steel
frames,” J. Constr. Steel Res., vol. 58, no. 2, pp. 193–212, Feb. 2002.
[22] Teh Lip H. and Clarke Murray J., “Plastic-Zone Analysis of 3D Steel Frames Using Beam Elements,” J.
Struct. Eng., vol. 125, no. 11, pp. 1328–1337, Nov. 1999.
[23] J. Y. R. Liew, H. Chen, N. E. Shanmugam, and W. F. Chen, “Improved nonlinear plastic hinge analysis
of space frame structures,” Eng. Struct., vol. 22, no. 10, pp. 1324–1338, Oct. 2000.
[24] N. Gebbeken, “A refined numerical approach for the ultimate‐load analysis of 3‐D steel rod
structures,” Eng. Comput., vol. 15, no. 3, pp. 312–344, May 1998.
[25] G. Turkalj, J. Brnic, and S. Kravanja, “A beam model for large displacement analysis of flexibly
connected thin-walled beam-type structures,” Thin-Walled Struct., vol. 49, no. 8, pp. 1007–1016,
Aug. 2011.
[26] B. Alemdar and D. Whie, “Displacement, Flexibility, and Mixed Beam–Column Finite Element
Formulations for Distributed Plasticity Analysis | Journal of Structural Engineering | Vol 131, No
12,” Dec. 2004.
[27] H. M. Bolt, C. J. Billington, and J. K. Ward, “Results From Large-Scale Ultimate Load Tests On
Tubular Jacket Frame Structures,” presented at the Offshore Technology Conference, 1994.
[28] Alf G. Engseth, “Finite element collapse analysis of tubular steel offshore structures,” Division of
Marine Structures, University of Trondheim, Norwegian Institute of Technology, Trondheim, 1984.
[29] “USFOS - Verification Manuals.” [Online]. Available:
http://www.usfos.no/manuals/usfos/verification/index.html. [Accessed: 28-Nov-2017].
[30] C. Polat, “Co-rotational formulation of a solid-shell element utilizing the ANS and EAS methods,” J.
Theor. Appl. Mech., vol. 48, Jan. 2010.
[31] F. Abed-Meraim and A. Combescure, “An improved assumed strain solid–shell element
formulation with physical stabilization for geometric non-linear applications and elastic–plastic
stability analysis,” Int. J. Numer. Methods Eng., vol. 80, no. 13, pp. 1640–1686, Dec. 2009.
[32] M. Mostafa, M. v. Sivaselvan, and C. a. Felippa, “A solid-shell corotational element based on
ANDES, ANS and EAS for geometrically nonlinear structural analysis,” Int. J. Numer. Methods Eng.,
vol. 95, no. 2, pp. 145–180, Jul. 2013.
[33] A. Masud and C. L. Tham, “Three-Dimensional Corotational Framework for Elasto-Plastic Analysis
of Multilayered Composite Shells,” AIAA J., vol. 38, no. 12, pp. 2320–2327, 2000.
[34] R. A. Fontes Valente, M. P. L. Parente, R. M. Natal Jorge, J. M. A. César de Sá, and J. J. Grácio,
“Enhanced transverse shear strain shell formulation applied to large elasto-plastic deformation
problems,” Int. J. Numer. Methods Eng., vol. 62, no. 10, pp. 1360–1398, Mar. 2005.
[35] J. H. Argyris, M. Papadrakakis, and L. Karapitta, “Elasto-plastic analysis of shells with the triangular
element TRIC,” Comput. Methods Appl. Mech. Eng., vol. 191, no. 33, pp. 3613–3636, Jun. 2002.
[36] U. Montag, W. B. Krätzig, and J. Soric, “Increasing solution stability for finite-element modeling of
elasto-plastic shell response,” Adv. Eng. Softw., vol. 30, no. 9, pp. 607–619, Sep. 1999.
[37] T. M. Roberts, “Second order strains and instability of thin walled bars of open cross-section,” Int.
J. Mech. Sci., vol. 23, no. 5, pp. 297–306, Jan. 1981.
[38] T. M. Roberts and Z. G. Azizian, “Influence of pre-buckling displacements on the elastic critical
loads of thin walled bars of open cross section,” Int. J. Mech. Sci., vol. 25, no. 2, pp. 93–104, Jan.
1983.
[39] W. Kanok-nukulchai, “A simple and efficient finite element for general shell analysis,” Int. J. Numer.
Methods Eng., vol. 14, no. 2, pp. 179–200, Jan. 1979.
[40] J.-L. Batoz, K.-Jür. Bathe, and L.-W. Ho, “A study of three-node triangular plate bending elements,”
Int. J. Numer. Methods Eng., vol. 15, no. 12, pp. 1771–1812, Dec. 1980.

265

SACS® Collapse Advanced Release 15.1

[41] A. Tessler and T. J. R. Hughes, “A three-node mindlin plate element with improved transverse
shear,” Comput. Methods Appl. Mech. Eng., vol. 50, no. 1, pp. 71–101, Jul. 1985.
[42] C. Jeyachandrabose, J. Kirkhope, and L. Meekisho, “An improved discrete Kirchhoff quadrilateral
thin-plate bending element,” Int. J. Numer. Methods Eng., vol. 24, no. 3, pp. 635–654, Mar. 1987.
[43] T. J. R. Hughes and T. E. Tezduyar, “Finite Elements Based Upon Mindlin Plate Theory With
Particular Reference to the Four-Node Bilinear Isoparametric Element,” J. Appl. Mech., vol. 48, no.
3, pp. 587–596, Sep. 1981.
[44] Bomel Limited, Comparison of tubular joint strength provisions in codes and standards. HSE Books,
2002.

266

SACS® Collapse Advanced Release 15.1

8 INPUT LINES

267

SACS® Collapse Advanced Release 15.1

 COLLAPSE ANALYSIS INPUT
COLUMNS COMMENTARY __________________________ COLUMNS COMMENTARY ___________________________
(42-43) ENHANCED COLLAPSE ONLY. CHOOSE YES TO INCLUDE ARC-LENGTH
GENERAL THIS LINE IS REQUIRED IN ANY COLLAPSE ANALYSIS RUN. IT IS SOLVER FOR POST-BUCKLING ANALYSIS
USED TO SPECIFY THE OVERALL ANALYSIS PARAMETERS.
(44-45) ENHANCED COLLAPSE ONLY. CHOOSE YES TO INCLUDE SUB-INCREMENTATION
(11-13) ENTER MAXIMUM NUMBER OF ITERATIONS FOR EACH LOADING SOLVER FOR MORE ROBUST NEWTON ITERATION.
INCREMENT.
(50-51) ENTER 'SF' TO CREATE SACS IV FILE WITH FINAL DEFLECTED SHAPE.
(14-16) ENTER THE MINIMUM NUMBER OF SEGMENTS THAT A MEMBER IS TO BE
DIVIDED FOR THE NONLINEAR MEMBER ANALYSIS. A NON-PRISMATIC (52-53) IF THE LOCAL BUCKLING OPTION 'LB' IS REQUESTED, ENTER THE
MEMBER MAY RESULT IN MORE THAN THIS NUMBER SINCE EACH SEGMENT METHOD USED TO DETERMINE LOCAL BUCKLING CRITERIA AS:
WILL BE LESS THAN OR EQUAL TO THE SECTION LENGTH. 'MG' - MARSHALL GATES LOWER LIMIT OF CRITICAL STRAIN.
'2U' - API BULLETIN 2U.
(17-19) ENTER THE MAXIMUM NUMBER OF ITERATIVE LOOPS FOR EACH MEMBER 'LR' - API LRFD.
ANALYSIS. 'IS' - ISO 19902 SECTION 13.2.3.3.

(26-41) SELECT FROM THE FOLLOWING ANALYSIS OPTIONS: (56-60) ENTER THE DEFLECTION TOLERANCE REQUIRED FOR CONVERGENCE OF
'LB' - LOCAL BUCKLING EFFECTS INCLUDED. ANY LOAD INCREMENT.
'JF' - JOINT FLEXIBILITY EFFECTS INCLUDED.
'FF' - JOINT FLEXIBILITY EFFECTS FROM SINGLE BRACE FORMULATION (61-65) ENTER THE ROTATION TOLERANCE REQUIRED FOR CONVERGENCE OF ANY
DUE TO FESSLER, MOCKFORD AND WEBSTER. LOAD INCREMENT.
'BF' - JOINT FLEXIBILITY EFFECTS FROM SINGLE BRACE FORMULATION
DUE TO BUITRAGO, HEALY AND CHANG. (66-70) ENTER THE CONVERGENCE TOLERANCE FOR THE MEMBERS.
'NS' - SKIPPED MEMBERS NOT TREATED AS LINEAR.
'PP' - INCLUDE PILE PLASTICITY. (71-75) ENTER THE MAXIMUM DEFLECTION ALLOWED BEFORE THE STRUCTURE IS
'CN' - CONTINUE IF MAXIMUM NUMBER OF ITERATIONS IS EXCEEDED. CONSIDERED COLLAPSED.
'JS' - JOINT STRENGTH CHECK (API-LRFD).
'N1' - JOINT STRENGTH CHECK (NORSOK N-004 REV. 1). (76-80) ENTER THE STRAIN HARDENING RATIO. THIS IS THE RATIO OF THE
'ND' - JOINT STRENGTH CHECK (NORSOK N-004 REV. 2). SLOPE OF THE PLASTIC PORTION OF THE STRESS-STRAIN CURVE TO
'N3' - JOINT STRENGTH CHECK (NORSOK N-004 REV. 3). THE SLOPE OF THE ELASTIC PORTION.
'IS' - JOINT STRENGTH CHECK (ISO 19902)
'DY' - DYNAMIC ANALYSIS OPTION.
'ME' - ALL MEMBERS ELASTIC.
'PE' - ALL PLATES ELASTIC.
'EB' - ELASTIC BUCKLING MONITOR

ANALYSIS PARAMETERS ANALYSIS OPTIONS CONVERGENCE CRITERIA

INLCUDE
CREATE LOCAL STRAIN
LINE MAXIMUM ARC- INCLUDE COLLAPSE
MAXIMUM MODEL BUCKLING HARDENING
LABEL NUMBER OF NUMBER OF LENGTH SUB-INC.? MEMBER DEFLECTION
NUMBER OF FILE METHOD DEFLECTION ROTATION RATIO
ITERATIONS MEMBER 1ST 2ND 3RD 4TH 5TH 6TH 7TH 8TH SOLVER? DEFLECTION
MEMBER TOLERANCE TOLERANCE
PER LOAD SEGMENTS TOLERANCE
ITERATIONS
INCREMENT

CLPOPT
1-- 6 11-->13 14-->16 17-->19 26--27 28--29 30--31 32--33 34--35 36--37 38--39 40--41 42--43 44--45 50--51 52--53 56<--60 61<--65 66<--70 71<--75 76<--80

DEFAULT 20 8 20 0.01 ENG 0.001 0.01 ENG 1000.0 ENG

ENGLISH IN RAD IN IN

METRIC CM RAD CM CM
 COLLAPSE ANALYSIS INPUT
COLUMNS COMMENTARY __________________________ COLUMNS COMMENTARY ___________________________
(55-56) ENTER "NR" TO IGNORE MEMBER END-RELEASES FOR COLLAPSE ANALYSIS.
GENERAL THIS LINE IS USED TO SPECIFY ADDITIONAL OVERALL ANALYSIS NOTE: THIS OPTION HAS NO EFFECT ON WISHBONES.
PARAMETERS. WISHBONES RELEASES ARE ALWAYS CONSIDERED IN THE ANALYSIS.

( 8-13) ENTER THE MEMBER ECCENTRICITY RATIO WHICH IS DEFINED AS: (57-58) ENTER IJ, IF, OR IS TO INCLUDE CONNECTIONS WITH PROPERTIES EXCEEDED
ER = E * C / R**2 WHERE THE APPLICABILITY RANGES FOR BOTH FLEXIBILITY AND STRENGTH, ONLY
E - ECCENTRICITY FOR FLEXIBILITY OR ONLY FOR STRENGTH, RESPECTIVELY.
C - DISTANCE FROM NEUTRAL AXIS TO EXTREME FIBER NOTE: INCLUDING OUT-OF-RANGE CONNECTIONS MAY LEAD TO
R - RADIUS OF GYRATION NON-CONVERGENCE OR MY SLOW DOWN THE CONVERGENCE.

(15-20) ENTER MAXIMUM DUCTILITY ALLOWED FOR ANY MEMBER. ANY MEMBER (59-59) SELECT DIVERGENCE RATIO EXPONENT. A GIVEN ITERATION WILL BE
THAT EXCEEDS THIS LIMIT WILL BE TREATED AS FRACTURED. ABORRTED IF CURRENT RESIDUAL TO INITIAL RESIDUAL BECOMES
LARGES THAN 10^(EXPONENT)
(22-24) ENTER "ITC" TO INCLUDE HIGHER ORDER CO-ROTATIONAL TERMS FOR
TUBULAR AND OTHER CLOSED SECTIONS IN ADDITION TO (61-61) SELECT DIVERGENCE MAXIMUM EXPONENT. A GIVEN ITERATION WILL BE
WIDE FLANGE OR OTHER OPEN SECTIONS INTERRUPTED IF CURRENT RESIDUAL VAULE BECOMES LARGES THAN
ENTER "EXC" TO EXCLUDE HIGHER ORDER CO-ROTATIONAL TERMS FOR 10^(EXPONENT)
ALL SECTION TYPES
(63-64) SELECT RELAXED CONVERGENCE EXPONENT. IF RELAXED CONVERGECE
(26-28) ENTER THE MEMBER STRAIN OPTION: CRETRION IS SELECTED, A GIVEN ITERATION WILL BE CONSIDERED
BLANK - INCLUDE AXIAL-BENDING SECOND-ORDER (LARGE) STRAINS CONVERGED IF CURRENT RESIDUAL TO INITIAL RESIDUAL
IWT - INCLUDE WAGNER STRAIN FOR THIN-WALLED OPEN SECTIONS BECOMES LESS THAN 10^(EXPONENT) .
IN ADDITION TO AXIAL-BENDING SECOND-ORDER TERMS
NLS - EXCLUDE ALL SECOND-ORDER (LARGE) STRAINSENTER (66-68) SELECT YES TO INCLUDE INTERATION RELAXATION FOR ENTIRE MODEL
TO INCREASE CHANCE OF CONVERGENCE.
(30-35) ENTER THE MEMBER OUT-OF-STRAIGHTNESS RATIO TO ADD NOTE: MAY INCREASE NUMBER OF ITERATIONS
IMPERFECTION IN THE FORM OF A HARMONIC FUNCTION.
(70-72) SELECT YES TO INCLUDE (RESIDUAL < 10^-5 INITIAL RESIDUAL) AS RELAXED
(37-39) SELECT YES TO USE MINDLIN PLATE THEORY (THICK PLATES) TO INCLUDE CRITERION FOR ENTIRE MODEL.
SHEAR DEFORMATION EFFECT IN PLATES. NOTE: IT MAY INCREASE NUMBER OF ITERATIONS OR CAUSE DIVERGENCE
IN LARGE NUMBER OF LOAD STEPS.
(41-46) ENTER THE PLATE DRILLING FACTOR FOR PLATE TWIST STIFFNESS
(74-76) SELECT YES TO INCLUDE ITERATION RELAXATION FOR MEMBERS
(48-53) ENTER THE SHEAR-LOCKING FACTOR FOR TRAINGULAR MINDLIN (THICK) TO INCREASE CHANCE OF CONVERGENCE.
PLATES. NOTE: MAY INCREASE NUMBER OF ITERATIONS.
NOTE: THICK QUAD. PLATES DO NOT NEED SHEAR-LOCKING FACTOR.
(78-80) SELECT YES TO INCLUDE A RELAXED CRITERION FOR MEMBERS.
(54-54) ENTER 'C' TO THE SECTION PROPERTIES ARE OVERRIDEN BY CALCULATED PROPERTIES NOTE: MAY INCREASE NUMBER OF ITERATIONS.
FOR MMEBERS FOR WHICH INPUT PROPERTIES ARE DIFFER FROM CALCULATED
PROPERTIES.

TRIANGULAR
INCLUDE JOINT
USE THICK THICK IGNORE MEMBER MEMBER
MEMBER MAXIMUM MEMBER OUT-OF PLATE OVERRIDE CONNECTIONS DIVERGENCE DIVERGENCE RELAXED MODEL MODEL MEMBER
LINE CO-ROTATIONAL MEMBER STRAIN (MINDLIN) (MINDLIN) END-RELEASES? RELAXED
ECCENTRICITY DUCTILITY STRAIGHTNESS DRILLING MEMBER EXCEEDED THE RATIO MAXIMUM CONVERGENCE RELAXED RELAXED RELAXED
LABEL 2ND-ORDER TERMS? OPTION PLATE PLATE SHEAR- (NO EFFECT ON CONVERGENCE
RATIO ALLOWED RATIO FACTOR PROPERTIES? APPLICABILITY EXPONENT EXPONENT EXPONENT ITERATION? ITERATION? ITERATION?
THEORY? LOCKING WISHBONES) CRITERION?
RANGES?
FACTOR

CLPOP2
1-- 6 8<--13 15<--20 22--24 26--28 30-35 30-35 41-46 48-53 54-54 55-56 57-58 59-59 61-61 63-64 66-68 70-72 74-76 78-80

DEFAULT 0.00 0.00 0.05 2.00 4 6 -5

ENGLISH PERCENT

METRIC PERCENT
 ARC LENGTH OPTIONS
COLUMNS COMMENTARY __________________________

GENERAL USE THIS LINE TO SPECIFY VARIOUS PARAMETERS FOR USE WITH
ARC-LENGTH ITERATIONS.

( 8-10) ENTER THE TYPE OF ARC-LENGTH ITERATIONS. THERE ARE TWO ARC-LENGTH
METHODS IN ENHANCED COLLAPSE: CYLINDRICAL AND SPHERICAL.
CYLINDRICAL METHOD ONLY USES DEFLECTION INCREMENTS TO DETERMINE THE ARC-LENGTH
WHILE SPHERICAL METHOD USES BOTH DEFLECTION AND FORCES INCREMENTS TO CALCULATE
ARC-LENGTH. BOTH METHODS SUPPORT SUB-INCREMENTATION.

BY DEFAULT, ENHANCED COLLAPSE USES FOLLOWING ORDER

1. CYLINDRICAL METHOD
2. IF 1 FAILS, CYLINDRICAL METHOD WITH SUB-INCREMENTATION
3. IF 2 FAILS, SPHERICAL METHOD
4. IF 3 FAILS, SPHERICAL METHOD WITH SUB-INCREMENTATION
5. IF 4 FAILS, CYLINDRICAL METHOD WITH INCREASING ARC-LENGTH

(12-18) MAXIMUM NUMBER OF ARC-LENGTH STEPS TO BE USED BEFORE

THE ANALYSIS IS TERMINATED.

(20-22) MAXIMUM NUMBER OF ARC-LENGTH SUB-INCREMENTATION LEVELS TO BE

USED BEFORE THE ANALYSIS IS TERMINATED. IF AN ARC-LENGTH
ITERATION FAILS TOP CONVERGE, THEN THE TARGET ARC-LENGTH IS
HALVED AND ANOTHER ATTEMPT IS MADE TO CALCULATE THE NEXT
LOAD FACTOR. THE MAXIMUM NUMBER OF SUB-INCREMENTATION LEVELS
DETERMINES HOW MANY TIMES THIS REFINEMENT TAKES PLACE BEFORE
THE ANALYSIS STOPS.

(24-29) ENTER THE SCALING FACTOR FOR AUTOMATICALLY COMPUTED ARC LENGTH
OR LEAVE BLANK FOR DEFAULT VALUE.

(31-36) ENTER SPHERICAL ARC-LENGTH PARAMETER OR LEAVE BLANK FOR

DEFAULT VALUE.

ARC-LENGTH MAXIMUM MAXIMUM

LINE SPHERICAL ARC-
FORMULATION NUMBER OF SUB-INCREMENTATION ARC-LENGTH FACTOR LEAVE BLANK
LABEL LENGTH PARAMETER
TYPE ITERATIONS LEVEL
ARCLEN
1-- 6 8<--10 12<--18 20<--22 24<--29 31<--36 37----80

DEFAULT 500 10 1.0 1.0

 AUTOMATIC SUB-INCREMENTATION OPTIONS
COLUMNS COMMENTARY __________________________

GENERAL USE THIS LINE TO SPECIFY VARIOUS PARAMETERS FOR USE WITH
THE OPTION TO APPLY AUTOMATIC SUB-INCREMENTATION.

( 8-9) MAXIMUM NUMBER OF SUB-INCREMENTATION LEVELS TO BE

USED BEFORE THE ANALYSIS IS TERMINATED. IF CONVERGENCE IS
NOT ACHIEVED FOR A GIVEN LOAD INCREMENT, THEN THE INCREMENTAL
LOAD FACTOR IS HALVED AND ANOTHER ATTEMPT IS MADE TO ANALYZE
THE SYSTEM AT A REDUCED LOAD FACTOR. THE MAXIMUM NUMBER OF
SUB-INCREMENTATION LEVELS DETERMINES HOW MANY TIMES THIS
REFINEMENT TAKES PLACE BEFORE
THE ANALYSIS STOPS.

(11-12) MAXIMUM ACCELERATION THRESHOLD. IF SUB-INCREMENTATION HAS

BEEN ACTIVATED AND CONVERGENCE HAS BEEN ACHIEVED FOR A
CERTAIN NUMBER OF LOAD INCREMENTS AFTER THE INITIAL
NON-CONVERGENCE, THEN THE INCREMENTAL LOAD FACTOR IS
DOUBLED. THE MAXIMUM ACCELERATION THRESHOLD SPECIFIES THE
NUMBER OF LOAD INCREMENTS THAT CONVERGE AT THE LOWER
LOAD FACTOR INCREMENT BEFORE THE ANALYSIS PROCEDES WITH
A LARGER LOAD FACTOR INCREMENT.

MAXIMUM MAXIMUM
LINE
SUB-INCREMENTATION ACCELERATION LEAVE BLANK
LABEL
LEVEL THRESHOLD
SUBINC
1-- 6 8<-- 9 11<--12 13------80

DEFAULT 10 4
 FORCE CONVERGENCE TOLERANCE
COLUMNS COMMENTARY __________________________

GENERAL USE THIS LINE TO SPECIFY FORCE AND MOMENT CONVERGENCE TOLERANCES.

( 8-13) ENTER THE TOLERANCE VALUE FOR FORCE CONVERGENCE

FOR ENHANCED COLLAPSE OR LEAVE BLANK FOR DEFAULT VALUE

(15-20) ENTER THE TOLERANCE VALUE FOR MOMENT CONVERGENCE

FOR ENHANCED COLLAPSE OR LEAVE BLANK FOR DEFAULT VALUE

LINE
FORCE TOLERANCE MOMENT TOLERANCE LEAVE BLANK
LABEL

FRCTOL
1-- 6 8<--13 15<--20 21---80

DEFAULT 0.001 0.001

 MSL JOINT FLEXIBILITY OPTIONS
COLUMNS COMMENTARY __________________________

GENERAL THIS LINE IS REQUIRED TO SPECIFY PARAMETERS WHEN USING JOINT

FLEXIBILITY OPTIONS FROM MSL.

( 8- 9) SELECT THE FLEXIBILITY OPTION:

'MF' - MEAN LEVEL
'CF' - CHARACTERISTIC LEVEL

(10-11) SELECT THE STRENGTH CHECK OPTION:

'MS' - MEAN LEVEL
'CS' - CHARACTERISTIC LEVEL

(12-13) SELECT THE FRACTURE CHECK OPTION:

'MT' - MEAN LEVEL
'CT' - CHARACTERISTIC LEVEL

(15-19) ENTER THE JOINT DISTORTION TOLERANCE.

(20-24) ENTER THE JOINT ROTATION TOLERANCE.

(25-29) ENTER MSL ASSESSMENT FACTOR FOR SAFETY

(78-80) ENTER 'NMP' TO IGNOTE ELASTOPLASTIC CALCULATION IN JOINT

FLEXIBILITY CALCULATION

TOLERANCES
LINE JOINT JOINT JOINT MSL ASSESSMENT IGNORE MSL
LEAVE BLANK
LABEL FLEXIBILITY STRENGTH FRACTURE DISTORTION ROTATION FACTOR PLASTICITY
TOLERANCE TOLERANCE
MSLOPT
1-- 6 8-- 9 10--11 12--13 15<--19 20<--24 25<--29 30------77 78---80

DEFAULT 0.001 ENGL 0.001 1.00

ENGLISH IN RAD

METRIC CM RAD
 COLLAPSE ANALYSIS REPORT SELECTION
COLUMNS COMMENTARY __________________________ COLUMNS COMMENTARY ___________________________
(32-36) ENTER THE PLASTICITY RATIO FOR THE MEMBER STRESS REPORT. ONLY
GENERAL THIS LINE IS USED TO SPECIFY THE COLLAPSE OUTPUT REPORT THOSE MEMBER SEGMENTS THAT EXCEED THIS LEVEL WILL BE INCLUDED
SELECTIONS. IN THIS REPORT.

( 8-29) SELECT FROM THE FOLLOWING OUTPUT REPORT CHOICES: (38-42) ENTER THE PLASTICITY RATIO FOR THE PILE DETAIL REPORT. ONLY
JOINT DISPLACEMENT REPORT OPTION: THOSE PILE INCREMENTS THAT EXCEED THIS LEVEL WILL BE INCLUDED
'P0' - PRINT FINAL DEFLECTIONS ONLY (DEFAULT) IN THIS REPORT.
'P1' - PRINT EVERY LOAD INCREMENT
'P2' - PRINT EVERY LOOP AND EVERY LOAD INCREMENT (44-48) ENTER THE PLASTICITY RATIO FOR THE PLATE STRESS DETAIL REPORT.
JOINT REACTION REPORT OPTION: ONLY THOSE MEMBER SEGMENTS THAT EXCEED THIS LEVEL WILL BE
'R0' - PRINT FINAL REACTIONS (DEFAULT) INCLUDED IN THIS REPORT.
'R1' - PRINT REACTIONS AT EACH LOAD INCREMENT
'R2' - PRINT REACTIONS AT EACH LOOP (58-59) ENTER 'EN' TO PRODUCE A MEMBER NORMAL STRAIN REPORT. THE REPORT IS
MEMBER STRESSES AND INTERNAL LOADS REPORT OPTION: GENERATED AT INTEGRATION POINTS FOR ALL MEMBER SUBSEGMENTS.
'M0' - PRINT FINAL MEMBER STRESSES (DEFAULT)
'M1' - PRINT MEMBER STRESSES AT EACH LOAD INCREMENT (62-63) ENTER 'EP' TO PRODUCE A MEMBER PLASTIC STRAIN REPORT. THE REPORT IS
'M2' - PRINT MEMBER STRESSES AT EACH LOOP GENERATED AT INTEGRATION POINTS FOR ALL MEMBER SUBSEGMENTS.
'MP' - OPTION TO ONLY INCLUDE PLASTIC MEMBERS/PLATES IN
MEMBER/PLATE INTERNAL LOADS REPORT (64-65) LEAVE BLANK OR ENTER 'EN' TO SAVE MAXIMUM NORMAL STRAIN INTO
'SP' - PRINT MEMBER STRESSES AT EACH SUBAREA AROUND COLLAPSE VIEW FOR REPORT GENERATION. ENTER 'EP' TO SAVE MAXIMUM PLASTIC
THE CROSS SECTION
STRAIN TO COLLAPSE VIEW.
JOINT STRENGTH REPORT OPTION:
'J1' - PRINT JOINT STRENGTH AT EACH LOAD INCREMENT
'J2' - PRINT JOINT STRENGTH AT EACH LOOP
SUMMARY REPORT OPTION:
'SM' - COLLAPSE SUMMARY REPORT
'MS' - MEMBER SUMMARY REPORT OPTION
'PW' - OPTION TO PRINT MEMBER WARNING MESSAGES

(26-27) PILEHEAD REACTION REPORT OPTION:

'F0' - PRINT FINAL REACTIONS
'F1' - PRINT REACTIONS AT EACH LOAD INCREMENT
'F2' - PRINT REACTIONS AT EACH LOOP

(30-31) ENTER 'VM' FOR A VON MISES STRESS CHECK FOR PLATES DESIGNATED
AS ELASTIC.

OUTPUT SELECTIONS MEMBER STRAIN REPORT

MEMBER PILE PLATE
VON
DETAIL DETAIL DETAIL
LINE MISES
REPORT REPORT REPORT LEAVE BLANK
LABEL PLATE
PLASTIC PLASTIC PLASTIC
CHECK
RATIO RATIO RATIO NORMAL PLASTIC MAX STRAIN FOR
1ST 2ND 3RD 4TH 5TH 6TH 7TH 8TH 9TH 10TH 11TH
STRAIN STRAIN COLLAPSE VIEW

CLPRPT

1-- 6 8-- 910--1112--1314--1516--1718--1920--2122--2324--2526--2728--29 30--31 32--36 38--42 44--48 58--59 62--63 64--65 66--80
 LOADING SEQUENCE INPUT
COLUMNS COMMENTARY __________________________

GENERAL THIS LINE IS REQUIRED IN ANY COLLAPSE ANALYSIS RUN. IT IS

USED TO SPECIFY THE LOAD SEQUENCE. IN A NONLINEAR ANALYSIS,
THE ORDER IN WHICH LOADS ARE APPLIED CAN BE SIGNIFICANT. FOR
EXAMPLE, THE DEAD LOAD SHOULD BE APPLIED BEFORE ANY
ENVIRONMENTAL LOADING. AS MANY AS SIX LOAD SEQUENCES CAN BE
DEFINED. EACH OF THESE WILL BE ANALYZED AS INDEPENDENT
NONLINEAR ANALYSES. A TOTAL OF 50 LOAD PATHS ARE ALLOWED.

( 7-10) ENTER THE IDENTIFICATION OF THIS LOAD SEQUENCE. EACH LOAD

SEQUENCE MUST HAVE A NON-BLANK LOAD SEQUENCE IDENTIFIER. IF
THIS FIELD IS LEFT BLANK, THE LOAD PATHS ARE APPLIED TO THE
PREVIOUS LOAD SEQUENCE.

(21-24) ENTER THE SACS IV LOAD CASE NAME FOR THE FIRST LOAD TO BE
APPLIED.

(25-28) ENTER THE NUMBER OF INCREMENTS FOR THIS LOAD STEP. THIS IS
THE NUMBER OF STEPS FROM THE STARTING LOAD FACTOR TO THE
ENDING LOAD FACTOR. IF THE STARTING FACTOR IS GREATER THAN
ZERO, THEN AN ADDITIONAL LOAD STEP IS CREATED TO REACH THE
STARTING LOAD FACTOR POSITION.

(29-34) ENTER THE STARTING LOAD FACTOR. THIS FACTOR CAN BE USED TO
SKIP THE LINEAR PORTION ON THE ANALYSIS AND SAVE UNNECESSARY
RUN TIME.

(35-40) ENTER THE ENDING LOAD FACTOR. THIS FACTOR MUST BE GREATER
THAN OR EQUAL TO THE STARTING FACTOR.

(41-60) ENTER THE PARAMETERS FOR THE SECOND LOAD STEP.

(61-80) ENTER THE PARAMETERS FOR THE THIRD LOAD STEP.

NOTE: IF MORE THAN 3 LOAD STEPS ARE DESIRED, THEY MAY BE

ENTERED ON ADDITIONAL LDSEQ INPUT LINES LEAVING THE LOAD
SEQUENCE IDENTIFIER BLANK.

FIRST LOAD STEP SECOND LOAD STEP THIRD LOAD STEP

LOAD
LINE
SEQUENCE
LABEL LOAD LOAD LOAD
ID NUMBER OF STARTING ENDING NUMBER OF STARTING ENDING NUMBER OF STARTING ENDING
CASE CASE CASE
INCREMENTS FACTOR FACTOR INCREMENTS FACTOR FACTOR INCREMENTS FACTOR FACTOR
NAME NAME NAME
LDSEQ

1-- 5 7--10 21-->24 25-->28 29<--34 35<--40 41-->44 45-->48 49<--54 55<--60 61-->64 65-->68 69<--74 75<--80
 LOADING SEQUENCE INPUT
COLUMNS COMMENTARY __________________________

GENERAL THIS LINE CAN BE USED IN ANY COLLAPSE ANALYSIS RUN. HOWEVER,
IT IS REQUIRED FOR ANY DYNAMIC COLLAPSE ANALYSIS. IT IS USED
TO SPECIFY THE LOAD STEPS IN A LOAD SEQUENCE AND ALSO THE
TIME DURATION FOR THE LOAD STEPS. IN A NONLINEAR ANALYSIS,
THE ORDER IN WHICH LOADS ARE APPLIED CAN BE SIGNIFICANT. FOR
EXAMPLE, THE DEAD LOAD SHOULD BE APPLIED BEFORE ANY
ENVIRONMENTAL LOADING. AS MANY AS SIX LOAD SEQUENCES CAN BE
DEFINED. EACH OF THESE WILL BE ANALYZED AS INDEPENDENT
NONLINEAR ANALYSES. A TOTAL OF 50 LOAD PATHS ARE ALLOWED.

( 7-10) ENTER THE IDENTIFICATION OF THIS LOAD SEQUENCE. EACH LOAD

SEQUENCE MUST HAVE A NON-BLANK LOAD SEQUENCE IDENTIFIER. IF
THIS FIELD IS LEFT BLANK, THE LOAD PATHS ARE APPLIED TO THE
PREVIOUS LOAD SEQUENCE.

(21-24) ENTER THE SACS IV LOAD CASE NAME FOR THE FIRST LOAD TO BE
APPLIED.

(25-29) ENTER THE NUMBER OF INCREMENTS FOR THIS LOAD STEP. THIS IS
THE NUMBER OF STEPS FROM THE STARTING LOAD FACTOR TO THE
ENDING LOAD FACTOR. IF THE STARTING FACTOR IS GREATER THAN
ZERO, THEN AN ADDITIONAL LOAD STEP IS CREATED TO REACH THE
STARTING LOAD FACTOR POSITION.

(30-36) ENTER THE STARTING LOAD FACTOR. THIS FACTOR CAN BE USED TO
SKIP THE LINEAR PORTION ON THE ANALYSIS AND SAVE UNNECESSARY
RUN TIME.

(37-43) ENTER THE ENDING LOAD FACTOR. THIS FACTOR MUST BE GREATER
THAN OR EQUAL TO THE STARTING FACTOR.

(44-50) ENTER THE TIME DURATION FOR THIS LOAD STEP. THIS IS REQUIRED
FOR A DYNAMIC ANALYSIS.

LOAD LOAD
LINE NUMBER OF STARTING ENDING
SEQUENCE CASE TIME DURATION LEAVE BLANK
LABEL INCREMENTS FACTOR FACTOR
ID NAME
LDAPL

1-- 5 7--10 21-->24 25-->29 30<--36 37<--43 44<--50 51------80

 SHIP IMPACT ENERGY
COLUMNS COMMENTARY __________________________

GENERAL THIS LINE SPECIFIES THE TOTAL IMPACT ENERGY IN A SHIP IMPACT.
IT PROVIDES A MEANS FOR THE USER TO SPECIFY SHIP VELOCITY AND
MASS WITH THE PROGRAM COMPUTING IMPACT ENERGY.

( 8-15) ENTER THE SHIP MASS M.

(17-22) ENTER THE ADDED MASS COEFFICIENT C. SHIP KINETIC ENERGY IS

CALCULATED WITH THE FORMULA Ek = 1/2*C*M*V^2.

(24-29) ENTER THE SHIP VELOCITY V.

ADDED
LINE SHIP SHIP
MASS LEAVE BLANK
LABEL MASS VELOCITY
COEFFICIENT
ENERGY
1-- 6 8<--15 17<--22 24<--29 30----80

DEFAULT

ENGLISH TON FT/S

METRIC TONNE M/S

 IMPACT LOAD CASE
COLUMNS COMMENTARY __________________________ COLUMNS COMMENTARY ___________________________
(38-38) DENT ENERGY FORMULA:
GENERAL THIS LINE SPECIFIES THE POSITION AND THE TOTAL IMPACT ENERGY 'BLANK' - NO DENTED MEMBER
TO BE ABSORBED IN AN IMPACT EVENT. ONE IMPACT LINE IS ANALYZED 'F' - FURNES FORMULA (API C18.9.2-2)
PER COLLAPSE EXECUTION. 'E' - ELLINAS FORMULA (API C18.9.2-7)

( 8-11) ENTER THE IMPACT LOAD CASE NAME IN THE MODEL USED TO DEFINE (40-43) ENTER JOINT 'A' OF DENTED MEMBER.
THE PERSETS FOR ENERGY CALCULATIONS.
(45-48) ENTER JOINT 'B' OF DENTED MEMBER.
(13-16) ENTER THE IMPACT JOINT NAME. ENERGY FOR THIS LOAD CASE WILL
BE TRANSFERRED TO THE STRUCTURE THROUGH THIS JOINT. LEAVE (50-52) ENTER 'ALL' TO SPECIFY THAT ALL LOADED JOINTS IN THE IMPACT
BLANK IF ALL LOADED JOINTS IN THE IMPACT LOAD CONDITION ARE LOAD CONDITION WILL BE USED FOR THE MONITORING OF STRUCTRUAL
TO BE USED FOR THE MONITORING OF STRUCTURAL DEFORMATION ENERGY. DEFORMATION ENERGY.

(56-61) ENTER A LIMIT FOR THE B RATIO, WHERE B = BRACE OD / DENT DEPTH.
(18-25) ENTER THE TOTAL IMPACT ENERGY TO BE ABSORBED. IF LEFT BLANK THE MEMBER INDENTATION ENERGY ABSORPTION WILL BE LIMITED BY
THE TOTAL IMPACT ENERGY WILL BE CALCULATED USING THE 'ENERGY' THE INDENTATION ENERGY CALCULATED FROM THIS VALUE OF B.
LINE.
(63-68) ENTER A LIMIT FOR THE PERCENTAGE OF THE KINETIC ENERGY OF
(27-30) ENTER THE SHIP INDENTATION CURVE NAME. THERE ARE FIVE IMPACT THAT IS TO BE ABSORBED AS MEMBER
STANDARD NAMES WHICH MAY BE ENTERED: INDENTATION ENERGY.
'DNV1' - BOW IMPACT FROM DNV TN A 202.
'DNV2' - BROAD SIDE IMPACT (D=1.5M) FROM DNV TN A 202. (79-80) ENTER 'PD' TO MODEL DENT/SHIP DENT AS PLASTIC
'DNV3' - BROAD SIDE IMPACT (D=10.M) FROM DNV TN A 202. (ENERGY/DENT IS UNRECOVERABLE).
'DNV4' - STERN IMPACT (D=1.5M) FROM DNV TN A 202.
'DNV5' - STERN IMPACT (D=10.M) FROM DNV TN A 202.

USER-SPECIFIED SHIP INDENTATION CURVES MAY BE SUPPLIED WITH

THE 'SHPIND' LINE SET. LEAVING THIS FIELD BLANK MEANS THAT
THE TOTAL ABSORBED ENERGY WILL BE DUE TO STRUCTURAL
DEFORMATION ALONE.

(32-33) ENTER 'EX' TO EXCLUDE AUTOMATIC UNLOADING AFTER IMPACT.

AUTOMATIC UNLOADING IS INCLUDED BY DEFAULT.

MEMBER DENT
IMPACT IMPACT IMPACT SHIP EXCLUDE MEMBER DENTED MEMBER ALL
LINE ENERGY LIMIT LEAVE PLASTIC
LOAD JOINT ENERGY INDENTATION AUTOMATIC DENT LOADS
LABEL BLANK DENT
CASE NAME ABSORBED CURVE NAME UNLOADING OPTION SPECIFIER
JOINT A JOINT B B %

IMPACT
1-- 6 8-->11 13-->16 18<--25 27-->30 32<--33 38 40-->43 45-->48 50--52 56<--61 63<--68 69--78 79--80

DEFAULT NONE NONE NONE

ENGLISH KIP-FT

METRIC MJ
 SHIP INDENTATION CURVE INPUT
COLUMNS COMMENTARY __________________________

GENERAL THIS OPTIONAL LINE SET IS USED TO DEFINE A SHIP INDENTATION

CURVE. THE SHIP INDENTATION CURVE IS REFERENCED BY THE
'IMPACT' LINE. FOR SHIP INDENTATION CURVES WITH MANY
FORCE/INDENTATION PAIRS, THIS LINE IS REPEATED WITHOUT THE
INDENTATION CURVE NAME UNTIL THE CURVE DEFINITION IS COMPLETE.

( 8-11) ENTER THE INDENTATION CURVE IDENTIFIER ON THE FIRST LINE OF A

'SHPIND' LINE SET.

(13-18) ENTER THE SHIP FORCE.

(20-24) ENTER THE SHIP INDENTATION. SUBSEQUENT FORCE/INDENTATION

PAIRS ARE ENTERED SIMILARLY. A BLANK FORCE/INDENTATION PAIR
IS SKIPPED.

INDENTATION
LINE SHIP SHIP SHIP SHIP SHIP
CURVE FORCE FORCE FORCE FORCE FORCE
LABEL INDENTATION INDENTATION INDENTATION INDENTATION INDENTATION
NAME
SHPIND
1-- 6 8-->11 13<--18 20<--24 26<--31 33<--37 39<--44 46<--50 52<--57 59<--63 65<--70 72<--77

DEFAULT

ENGLISH KIP FT KIP FT KIP FT KIP FT KIP FT

METRIC(KN) MN M MN M MN M MN M MN M

METRIC(KG) TONNE M TONNE M TONNE M TONNE M TONNE M

 JOINT SELECTION INPUT
COLUMNS COMMENTARY __________________________

GENERAL THIS LINE IS OPTIONAL IN ANY COLLAPSE ANALYSIS RUN. IT IS

USED TO SELECT JOINTS FOR DEFLECTION PRINT. IF OMITTED, THEN
ALL JOINTS ARE SELECTED BY DEFAULT.

(12-80) ENTER THE JOINT NAMES OF JOINTS SELECTED FOR OUTPUT

DEFLECTION PRINT. USE AS MANY OF THESE INPUT LINES AS DESIRED.

JOINT SELECTIONS FOR DEFLECTION PRINT

LINE
LABEL 1ST 2ND 3RD 4TH 5TH 6TH 7TH 8TH 9TH 10TH 11TH 12TH 13TH 14TH
JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT
JTSEL

1-- 5 12-->15 17-->20 22-->25 27-->30 32-->35 37-->40 42-->45 47-->50 52-->55 57-->60 62-->65 67-->70 72-->75 77-->80
 MEMBER SELECTION INPUT
COLUMNS COMMENTARY __________________________

GENERAL THIS LINE IS OPTIONAL IN ANY COLLAPSE ANALYSIS RUN. IT IS

USED TO SELECT MEMBERS FOR OUTPUT PRINT. IF OMITTED, THEN
ALL MEMBERS ARE SELECTED BY DEFAULT.

(12-80) ENTER THE CONNECTING JOINT NAMES OF MEMBERS SELECTED FOR

OUTPUT PRINT. USE AS MANY OF THESE INPUT LINES AS DESIRED.

MEMBER SELECTIONS FOR OUTPUT REPORTS

LINE
1ST MEMBER 2ND MEMBER 3RD MEMBER 4TH MEMBER 5TH MEMBER 6TH MEMBER 7TH MEMBER
LABEL
1ST 2ND 1ST 2ND 1ST 2ND 1ST 2ND 1ST 2ND 1ST 2ND 1ST 2ND
JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT
MEMSEL

1-- 6 12-->15 17-->20 22-->25 27-->30 32-->35 37-->40 42-->45 47-->50 52-->55 57-->60 62-->65 67-->70 72-->75 77-->80
 PLATE SELECTION INPUT
COLUMNS COMMENTARY __________________________

GENERAL THIS LINE IS OPTIONAL IN ANY COLLAPSE ANALYSIS RUN. IT IS

USED TO SELECT PLATES FOR OUTPUT PRINT. IF OMITTED, THEN ALL
PLATES ARE SELECTED BY DEFAULT.

(12-80) ENTER THE PLATE IDENTIFIERS OF PLATES SELECTED FOR OUTPUT

PRINT. USE AS MANY OF THESE INPUT LINES AS DESIRED.

PLATE SELECTIONS FOR OUTPUT REPORTS

LINE
LABEL 1ST 2ND 3RD 4TH 5TH 6TH 7TH 8TH 9TH 10TH 11TH 12TH 13TH 14TH
PLATE PLATE PLATE PLATE PLATE PLATE PLATE PLATE PLATE PLATE PLATE PLATE PLATE PLATE
PLTSEL

1-- 6 12-->15 17-->20 22-->25 27-->30 32-->35 37-->40 42-->45 47-->50 52-->55 57-->60 62-->65 67-->70 72-->75 77-->80
 GROUTED MEMBER YIELD STRESS MODIFICATION
COLUMNS COMMENTARY __________________________

GENERAL THIS LINE IS OPTIONAL IN ANY COLLAPSE ANALYSIS RUN. IT IS

USED TO SPECIFY THE YIELD STRESS FOR GROUTED TUBULAR MEMBERS.
THE LINE MAY BE APPLIED ONLY TO CROSS SECTIONS WITH THE 'TUB'
TYPE WHICH HAVE A SPECIFIED INNER TUBE. IF APPLIED TO MEMBERS
WITHOUT THE 'TUB' CROSS SECTION TYPE, OR TO MEMBERS WITH THE
'TUB' CROSS SECTION TYPE BUT WITHOUT AN INNER TUBE, THIS LINE
IS IGNORED.

( 8-12) ENTER THE YIELD STRESS. THIS YIELD STRESS IS APPLIED ONLY TO
THE INNER (GROUTED) TUBE OF THE 'TUB' CROSS SECTION.

(14-72) ENTER THE CONNECTING JOINT NAMES OF UP TO SIX GROUTED TUBULAR

MEMBERS SELECTED FOR GROUTED YIELD STRESS MODIFICATION. USE
AS MANY 'GRMSEL' INPUT LINES AS REQUIRED.

GROUTED MEMBER SELECTION FOR YIELD STRESS MODIFICATION

LINE YIELD
1ST MEMBER 2ND MEMBER 3RD MEMBER 4TH MEMBER 5TH MEMBER 6TH MEMBER
LABEL STRESS
1ST 2ND 1ST 2ND 1ST 2ND 1ST 2ND 1ST 2ND 1ST 2ND
JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT
GRMSEL
1-- 6 8<--12 14-->17 19-->22 24-->27 29-->32 34-->37 39-->42 44-->47 49-->52 54-->57 59-->62 64-->67 69-->72

DEFAULT

ENGLISH KSI

METRIC(KN) KN/SQ.CM

METRIC(KG) KG/SQ.CM
 ELASTIC PLATE GROUP SELECTION
COLUMNS COMMENTARY __________________________

GENERAL THIS LINE IS OPTIONAL IN ANY COLLAPSE ANALYSIS RUN. IT IS

USED TO DESIGNATE PLATES AS ELASTIC.

(12-79) ENTER THE PLATE GROUP IDENTIFIERS OF PLATES SELECTED AS

ELASTIC. USE AS MANY OF THESE INPUT LINES AS DESIRED.

ELASTIC PLATE GROUP SELECTIONS

LINE
LABEL 1ST 2ND 3RD 4TH 5TH 6TH 7TH 8TH 9TH 10TH 11TH 12TH 13TH 14TH
GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP
PGRELA

1-- 6 12-->14 17-->19 22-->24 27-->29 32-->34 37-->39 42-->44 47-->49 52-->54 57-->59 62-->64 67-->69 72-->74 77-->79
 ELASTIC PLATE SELECTION
COLUMNS COMMENTARY __________________________

GENERAL THIS LINE IS OPTIONAL IN ANY COLLAPSE ANALYSIS RUN. IT IS

USED TO DESIGNATE PLATES AS ELASTIC.

(12-80) ENTER THE PLATE IDENTIFIERS OF PLATES SELECTED AS ELASTIC.

USE AS MANY OF THESE INPUT LINES AS DESIRED.

ELASTIC PLATE SELECTIONS

LINE
LABEL 1ST 2ND 3RD 4TH 5TH 6TH 7TH 8TH 9TH 10TH 11TH 12TH 13TH 14TH
PLATE PLATE PLATE PLATE PLATE PLATE PLATE PLATE PLATE PLATE PLATE PLATE PLATE PLATE
PLTELA

1-- 6 12-->15 17-->20 22-->25 27-->30 32-->35 37-->40 42-->45 47-->50 52-->55 57-->60 62-->65 67-->70 72-->75 77-->80
 JOINT STRENGTH OPTION
COLUMNS COMMENTARY __________________________

GENERAL THIS LINE IS OPTIONAL IN ANY COLLAPSE ANALYSIS RUN. IT IS

USED TO SPECIFY JOINT STRENGTH OPTIONS. IF OMITTED, THEN
DEFAULT OPTIONS WILL BE USED.

( 11 ) IF THE BRACE LOADS ARE TO BE BACKED OFF TO THE CHORD OUTER

SURFACE, ENTER 'R' HERE. OTHERWISE, THE PROGRAM WILL USE THE
BRACE END FORCES.

(12-17) ENTER THE MINIMUM GAP ALLOWED FOR THE JOINT STRENGTH ANALYSIS.

(18-23) ENTER THE MAXIMUM GAP ALLOWED FOR THE JOINT STRENGTH ANALYSIS.

( 25 ) USE EFFECTIVE THICKNESS FOR GROUTED ELEMENTS. ENTER '1' FOR

EFF THICK BASED ON THE COMPOSITE SECTION MOMENT OF INERTIA OR
'2' FOR EFF THICK BASED ON MOMENT OF INERTIAS OF THE TWO WALLS OR
'3' FOR EFF THICK BASED ON SQUARE ROOT OF SUM OF SQUARES (SRSS) OF
WALL THICKNESSES.

(26-30) ENTER THE EFFECTIVE THICKNESS LIMIT EXPRESSED AS A FACTOR OF

THE WALL THICKNESS OF THE LARGER (OUTSIDE) TUBE.

(31-35) ENTER THE UNITY CHECK LOWER LIMIT. ONLY JOINTS WITH STRENGTH
UNITY CHECK RATIOS ABOVE THIS VALUE WILL BE REPORTED.

( 37 ) PERFORM THE ISO 19902 JOINT STRENGTH CHECK USING

BRACE UTILIZATION. ENTER
'U' TO ASSUME A BRACE UTILIZATION OF UNITY, ENTER A 'B' TO
CALCULATE THE BRACE-END UTILIZATION USING ISO 19902 EQUATIONS
13.3-2 AND 13.3-8.

(39-45) ENTER THE TOLERANCE TO CHECK OFFSET FOR BRACE-CHORD CONNECTIONS.

(47-48) ENTER 'PG' TO PRINT CONNECTION GEOMETRIC PARAMETERS (BETAM, GAMMA

, AND TAU) TO THE LISTING FILE

EFFECTIVE
EFFECTIVE PRINT BRACE PRINT
LINE RELIEF MINIMUM MAXIMUM THICKNESS BRACE OFFSET
THICKNESS UC UTILIZATION GEOMETRIC LEAVE BLANK
LABEL OPTION GAP GAP LIMIT TOLERANCE
OPTION LEVEL OPTION PARAMETERS
RATIO
JSOPT
1-- 5 11 12<--17 18<--23 25 26<--30 31--35 37 39<--45 47--48 49--80

DEFAULT -100.0 ENGL +1000.0 ENGL 1.75 0.1

ENGLISH IN IN IN

METRIC CM CM CM
 JOINT STRENGTH JOINT SELECTION LINE
COLUMNS COMMENTARY __________________________

GENERAL THIS LINE ALLOWS THE USER TO INCLUDE OR EXCLUDE JOINTS WHEN
CONSIDERING JOINT STRENGTH. ALL BRACES CONNECTED TO THE
SPECIFIED JOINTS ARE INCLUDED OR EXCLUDED PROVIDED ONE OF THE
JOINT STRENGTH OPTIONS IS SPECIFIED ON THE CLPOPT LINE. THIS
LINE IS IGNORED IF NO STRENGTH OPTION IS SPECIFIED ON THE
CLPOPT LINE.

( 1- 5) ENTER 'JSSEL'.

( 7 ) ENTER 'I' TO INCLUDE OR 'X' TO EXCLUDE THE JOINTS SPECIFIED.

NOTE: THE INCLUDE AND EXCLUDE OPTIONS ARE MUTUALLY EXCLUSIVE AND
CANNOT BE USED TOGETHER. ALL JOINTS SPECIFIED ON JSSEL LINES
MUST BE EITHER INCLUDED OR EXCLUDED.

( 9-12) ENTER THE NAME OF THE FIRST JOINT TO BE INCLUDED OR EXCLUDED.

(14-77) ENTER ADDITIONAL JOINTS TO BE INCLUDED OR EXCLUDED.

JOINT NAMES
INCLUDE
LINE
OR
LABEL
EXCLUDE 1ST 2ND 3RD 4TH 5TH 6TH 7TH 8TH 9TH 10TH 11TH 12TH 13TH 14TH
JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT

JSSEL

1-- 5 7 9-->12 14-->17 19-->22 24-->27 29-->32 34-->37 39-->42 44-->47 49-->52 54-->57 59-->62 64-->67 69-->72 74-->77
 JOINT STRENGTH BRACE SELECTION LINE
COLUMNS COMMENTARY __________________________

GENERAL THIS LINE ALLOWS THE USER TO INCLUDE OR EXCLUDE BRACES WHEN
CONSIDERING JOINT STRENGTH. ONLY THE BRACE AND BRACE SIDE
SPECIFIED ARE INCLUDED OR EXCLUDED PROVIDED ONE OF THE JOINT
STRENGTH OPTIONS IS SPECIFIED ON THE CLPOPT LINE. THIS LINE
IS IGNORED IF NO STRENGTH OPTION IS SPECIFIED.

( 1- 5) ENTER 'BSSEL'.

( 7 ) ENTER 'I' TO INCLUDE OR 'X' TO EXCLUDE THE BRACES SPECIFIED.

NOTE: THE INCLUDE AND EXCLUDE OPTIONS ARE MUTUALLY EXCLUSIVE AND
CANNOT BE USED TOGETHER. ALL BRACES SPECIFIED ON BSSEL LINES
MUST BE EITHER INCLUDED OR EXCLUDED.

( 9-12) ENTER THE BEGIN JOINT OF THE FIRST BRACE TO INCLUDE OR EXCLUDE.

(13-16) ENTER THE END JOINT OF THE FIRST BRACE TO INCLUDE OR EXCLUDE.

(17-20) ENTER THE JOINT DESIGNATING THE BRACE CONNECTION TO INCLUDE

OR EXCLUDE. ENTER THE BEGIN JOINT TO CONSIDER THE CONNECTION
AT THE START OF THE BRACE OR THE END JOINT TO CONSIDER THE
CONNECTION AT THE END OF THE BRACE.

(22-72) ENTER ADDITIONAL BRACES TO BE INCLUDED OR EXCLUDED.

1ST BRACE 2ND BRACE 3RD BRACE 4TH BRACE 5TH BRACE
INCLUDE
LINE
OR
LABEL
EXCLUDE BEGIN END BEGIN END BEGIN END BEGIN END BEGIN END
CONNECTION CONNECTION CONNECTION CONNECTION CONNECTION
JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT

BSSEL

1-- 5 7 9-->12 13-->16 17-->20 22-->25 26-->29 30-->33 35-->38 39-->42 43-->46 48-->51 52-->55 56-->59 61-->64 65-->68 69-->72
 JOINT FLEXIBILITY JOINT SELECTION LINE
COLUMNS COMMENTARY __________________________

GENERAL THIS LINE ALLOWS THE USER TO INCLUDE OR EXCLUDE JOINTS WHEN
CONSIDERING JOINT FLEXIBILITY. ALL BRACES CONNECTED TO THE
SPECIFIED JOINTS ARE INCLUDED OR EXCLUDED PROVIDED ONE OF THE
JOINT FLEXIBILITY OPTIONS IS SPECIFIED ON THE CLPOPT LINE.
THIS LINE IS IGNORED IF NO FLEXIBILITY OPTION IS SPECIFIED ON
THE CLPOPT LINE.

( 1- 5) ENTER 'JFSEL'.

( 7 ) ENTER 'I' TO INCLUDE OR 'X' TO EXCLUDE THE JOINTS SPECIFIED.

NOTE: THE INCLUDE AND EXCLUDE OPTIONS ARE MUTUALLY EXCLUSIVE AND
CANNOT BE USED TOGETHER. ALL JOINTS SPECIFIED ON JFSEL LINES
MUST BE EITHER INCLUDED OR EXCLUDED.

( 9-12) ENTER THE NAME OF THE FIRST JOINT TO BE INCLUDED OR EXCLUDED.

(14-77) ENTER ADDITIONAL JOINTS TO BE INCLUDED OR EXCLUDED.

JOINT NAMES
INCLUDE
LINE
OR
LABEL
EXCLUDE 1ST 2ND 3RD 4TH 5TH 6TH 7TH 8TH 9TH 10TH 11TH 12TH 13TH 14TH
JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT

JFSEL

1-- 5 7 9-->12 14-->17 19-->22 24-->27 29-->32 34-->37 39-->42 44-->47 49-->52 54-->57 59-->62 64-->67 69-->72 74-->77
 JOINT FLEXIBILITY BRACE SELECTION LINE
COLUMNS COMMENTARY __________________________

GENERAL THIS LINE ALLOWS THE USER TO INCLUDE OR EXCLUDE BRACES WHEN
CONSIDERING JOINT FLEXIBILITY. ONLY THE BRACE AND BRACE SIDE
SPECIFIED ARE INCLUDED OR EXCLUDED PROVIDED ONE OF THE JOINT
FLEXIBILITY OPTIONS IS SPECIFIED ON THE CLPOPT LINE. THIS
LINE IS IGNORED IF NO FLEXIBILITY OPTION IS SPECIFIED.

( 1- 5) ENTER 'BFSEL'.

( 7 ) ENTER 'I' TO INCLUDE OR 'X' TO EXCLUDE THE BRACES SPECIFIED.

NOTE: THE INCLUDE AND EXCLUDE OPTIONS ARE MUTUALLY EXCLUSIVE AND
CANNOT BE USED TOGETHER. ALL BRACES SPECIFIED ON BFSEL LINES
MUST BE EITHER INCLUDED OR EXCLUDED.

( 9-12) ENTER THE BEGIN JOINT OF THE FIRST BRACE TO INCLUDE OR EXCLUDE.

(13-16) ENTER THE END JOINT OF THE FIRST BRACE TO INCLUDE OR EXCLUDE.

(17-20) ENTER THE JOINT DESIGNATING THE BRACE CONNECTION TO INCLUDE

OR EXCLUDE. ENTER THE BEGIN JOINT TO CONSIDER THE CONNECTION
AT THE START OF THE BRACE OR THE END JOINT TO CONSIDER THE
CONNECTION AT THE END OF THE BRACE.

(22-72) ENTER ADDITIONAL BRACES TO BE INCLUDED OR EXCLUDED.

1ST BRACE 2ND BRACE 3RD BRACE 4TH BRACE 5TH BRACE
INCLUDE
LINE
OR
LABEL
EXCLUDE BEGIN END BEGIN END BEGIN END BEGIN END BEGIN END
CONNECTION CONNECTION CONNECTION CONNECTION CONNECTION
JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT

BFSEL

1-- 5 7 9-->12 13-->16 17-->20 22-->25 26-->29 30-->33 35-->38 39-->42 43-->46 48-->51 52-->55 56-->59 61-->64 65-->68 69-->72
 ISO 19902 RESISTANCE FACTOR DATA
COLUMNS COMMENTARY __________________________

GENERAL THIS DATA ENABLES THE USER TO OVERRIDE THE

RESISTANCE FACTORS SPECIFIED IN THE ISO 19902 STANDARD FOR
STRENGTH OF TUBULAR JOINTS, NAMELY THE TUBULAR JOINT PARTIAL
RESISTANCE FACTOR, THE YIELD STRENGH PARTIAL RESISANCE
FACTOR AND THE EXTRA PARTIAL RESISTANCE FACTOR.
FURTHERMORE, CONNECTION RESISTANCE FACTORS MAY BE SPECIFIED,
THE DEFAULT VALUES FOR WHICH ARE 1.0.

( 6-25) ENTER THE CONNECTION RESISTANCE FACTORS FOR T AND Y TYPE

JOINTS. IF ANY ITEM IS ENTERED AS ZERO OR LEFT BLANK, THEN
ITS DEFAULT VALUE WILL BE USED.

(26-45) ENTER THE CONNECTION RESISTANCE FACTORS FOR X TYPE JOINTS.

(46-65) ENTER THE CONNECTION RESISTANCE FACTORS FOR K TYPE JOINTS.

(66-70) ENTER THE PARTIAL RESISTANCE FACTOR FOR TUBULAR JOINTS.

(71-75) ENTER THE PARTIAL RESISTANCE FACTOR FOR YIELD STRENGTH.

(76-80) ENTER THE EXTRA PARTIAL RESISTANCE FACTOR.

T & Y JOINTS X JOINTS K JOINTS

TUBULAR YIELD EXTRA
LINE
JOINT STRENGTH PARTIAL
LABEL
AXIAL AXIAL IN-PL OUT-OF-PLANE AXIAL AXIAL IN-PL OUT-OF-PLANE AXIAL AXIAL IN-PL OUT-OF-PLANE RES. FAC RES. FAC RES. FAC
TENS. COMP. BEND. BEND. TENS. COMP. BEND. BEND. TENS. COMP. BEND. BEND.

RSFAC
1-- 5 6<--10 11<--15 16<--20 21<--25 26<--30 31<--35 36<--40 41<--45 46<--50 51<--55 56<--60 61<--65 66<--70 71<--75 76<--80

DEFAULT 1 1 1 1 1 1 1 1 1 1 1 1 1.05 1.05 1.17

 LRFD RESISTANCE FACTOR DATA
COLUMNS COMMENTARY __________________________

GENERAL THIS DATA ENABLES THE USER TO OVERRIDE THE LRFD RESISTANCE
FACTORS AS SPECIFIED IN THE API RP 2A-LRFD. THE DEFAULT
VALUES ARE AS SPECIFIED IN THE COMMENTARY SECTION AS BETA
FACTORS.

( 6-25) ENTER THE CONNECTION RESISTANCE FACTORS FOR T AND Y TYPE

JOINTS. IF ANY ITEM IS ENTERED AS ZERO OR LEFT BLANK, THEN
ITS DEFAULT VALUE WILL BE USED.

(26-45) ENTER THE CONNECTION RESISTANCE FACTORS FOR X TYPE JOINTS.

(46-65) ENTER THE CONNECTION RESISTANCE FACTORS FOR K TYPE JOINTS.

(66-70) ENTER THE YIELD STRESS RESISTANCE FACTOR.

YIELD
T & Y JOINTS X JOINTS K JOINTS
LINE STRESS LEAVE
LABEL AXIAL AXIAL IN-PL OUT-OF-PLANE AXIAL AXIAL IN-PL OUT-OF-PLANE AXIAL AXIAL IN-PL OUT-OF-PLANE RESISTANCE BLANK
TENS. COMP. BEND. BEND. TENS. COMP. BEND. BEND. TENS. COMP. BEND. BEND. FACTOR

RSFAC
1-- 5 6<--10 11<--15 16<--20 21<--25 26<--30 31<--35 36<--40 41<--45 46<--50 51<--55 56<--60 61<--65 66<--70 71--80

DEFAULT 2.11 2.57 2.81 2.61 2.11 2.57 2.81 2.61 2.51 2.51 2.81 2.61 1
 NORSOK RESISTANCE FACTOR DATA
COLUMNS COMMENTARY __________________________

GENERAL THIS DATA ENABLES THE USER TO OVERRIDE THE MATERIAL

RESISTANCE FACTOR AS SPECIFIED IN THE NORSOK JOINT STRENGTH
STANDARD. THE DEFAULT VALUE IS 1.15. FURTHERMORE, CONNECTION
RESISTANCE FACTORS MAY BE SPECIFIED, THE DEFAULT VALUES FOR
WHICH ARE 1.0.

( 6-25) ENTER THE CONNECTION RESISTANCE FACTORS FOR T AND Y TYPE

JOINTS. IF ANY ITEM IS ENTERED AS ZERO OR LEFT BLANK, THEN
ITS DEFAULT VALUE WILL BE USED.

(26-45) ENTER THE CONNECTION RESISTANCE FACTORS FOR X TYPE JOINTS.

(46-65) ENTER THE CONNECTION RESISTANCE FACTORS FOR K TYPE JOINTS.

(66-70) ENTER THE MATERIAL RESISTANCE FACTOR.

T & Y JOINTS X JOINTS K JOINTS MATERIAL

LINE LEAVE
RESISTANCE
LABEL AXIAL AXIAL IN-PL OUT-OF-PLANE AXIAL AXIAL IN-PL OUT-OF-PLANE AXIAL AXIAL IN-PL OUT-OF-PLANE BLANK
FACTOR
TENS. COMP. BEND. BEND. TENS. COMP. BEND. BEND. TENS. COMP. BEND. BEND.
RSFAC
1-- 5 6<--10 11<--15 16<--20 21<--25 26<--30 31<--35 36<--40 41<--45 46<--50 51<--55 56<--60 61<--65 66<--70 71--80

DEFAULT 1 1 1 1 1 1 1 1 1 1 1 1 1.15
 MSL RESISTANCE FACTOR DATA
COLUMNS COMMENTARY __________________________

GENERAL THIS DATA ENABLES THE USER TO DEFINE ADDITIONAL

CONNECTION RESISTANCE FACTORS FOR MSL JOINT CALCULATION.
THE DEFAULT VALUES FOR WHICH ARE 1.0.

( 6-25) ENTER THE CONNECTION RESISTANCE FACTORS FOR T AND Y TYPE

JOINTS. IF ANY ITEM IS ENTERED AS ZERO OR LEFT BLANK, THEN
ITS DEFAULT VALUE WILL BE USED.

(26-45) ENTER THE CONNECTION RESISTANCE FACTORS FOR X TYPE JOINTS.

(46-65) ENTER THE CONNECTION RESISTANCE FACTORS FOR K TYPE JOINTS.

T & Y JOINTS X JOINTS K JOINTS

LINE
LEAVE BLANK
LABEL
AXIAL AXIAL IN-PL OUT-OF-PLANE AXIAL AXIAL IN-PL OUT-OF-PLANE AXIAL AXIAL IN-PL OUT-OF-PLANE
TENS. COMP. BEND. BEND. TENS. COMP. BEND. BEND. TENS. COMP. BEND. BEND.

RSFAC
1-- 5 6<--10 11<--15 16<--20 21<--25 26<--30 31<--35 36<--40 41<--45 46<--50 51<--55 56<--60 61<--65 62--80

DEFAULT 1 1 1 1 1 1 1 1 1 1 1 1
 BRACE RESISTANCE FACTOR OVERRIDE
COLUMNS COMMENTARY __________________________

GENERAL THIS DATA ENABLES THE USER TO OVERRIDE THE LRFD RESISTANCE
FACTORS AS SPECIFIED IN THE API RP 2A-LRFD FOR AN INDIVIDUAL
BRACE. OVERRIDES SPECIFIED ON THE 'RSFACO' LINE REPLACE ANY
RESISTANCE FACTOR OVERRIDES SPECIFIED ON THE 'RSFAC' LINE FOR
THE BRACE END SPECIFIED. IF ANY FACTOR IS ENTERED AS ZERO OR
LEFT BLANK, THEN THE VALUE ENTERED ON THE 'RSFAC' LINE IS USED.

( 1- 6) ENTER 'RSFACO'

( 8-11) ENTER THE BEGIN JOINT OF THE BRACE TO WHICH THE OVERRIDES
APPLY.

(12-15) ENTER THE END JOINT OF THE BRACE TO WHICH THESE OVERRIDES
APPLY.

(16-19) ENTER THE JOINT DESIGNATING THE BRACE CONNECTION TO WHICH

THESE OVERRIDES APPLY. ENTER THE BEGIN JOINT TO CONSIDER THE
CONNECTION AT THE START OF THE BRACE OR THE END JOINT TO
CONSIDER THE CONNECTION AT THE END OF THE BRACE.

(21-25) ENTER THE AXIAL TENSION RESISTANCE FACTOR.

(26-30) ENTER THE AXIAL COMPRESSION RESISTANCE FACTOR.

(31-35) ENTER THE IN-PLANE BENDING RESISTANCE FACTOR.

(36-40) ENTER THE OUT-OF-PLANE BENDING RESISTANCE FACTOR.

(41-45) ENTER THE YIELD STRENGTH RESISTANCE FACTOR.

( 47 ) OPTIONALLY SPECIFY THE CONNECTION TYPE TO WHICH THE OVERRIDE

FACTORS APPLY AS FOLLOWS:
'X' - X OR CROSS CONNECTION
'Y' - T & Y CONNECTION
'K' - K BRACE CONNECTION

BRACE RESISTANCE FACTOR OVERRIDES

LINE CONNECTION LEAVE
LABEL BEGIN END AXIAL AXIAL IN-PLANE OUT-OF-PLANE YIELD TYPE BLANK
CONNECTION
JOINT JOINT TENS. COMP. BEND. BEND. STRESS
RSFACO

1-- 6 8-->11 12-->15 16-->19 21--25 26--30 31--35 36--40 41--45 47 48--80
 YIELD STRESS FACTOR
COLUMNS COMMENTARY __________________________

GENERAL THIS DATA ENABLES THE USER TO MODIFY THE YIELD STRESS OF ALL
ENTITIES BY SPECIFYING A YIELD STRESS FACTOR. WITH THIS LINE,
THE MODIFIED YIELD STRESS USED BY COLLAPSE IS THE YIELD
STRESS SPECIFIED IN THE SACS IV MODEL MULTIPLIED BY THIS
FACTOR. THIS IS PARTICULARLY USEFUL IN PROVIDING THE MEANS OF
SUPPLYING A DYNAMIC INCREASE FACTOR IN BLAST ANALYSIS.

( 8-13) ENTER THE YIELD STRESS FACTOR.

YIELD
LINE LEAVE
STRESS
LABEL BLANK
FACTOR
YSFACT
1-- 6 8<--13 14----------------80

DEFAULT 1
 YIELD STRESS UNIVERSAL OVERRIDE LINE
COLUMNS COMMENTARY __________________________

GENERAL THE 'YSUOVR' LINE IS USED TO REPLACE ALL INSTANCES OF A SACS

IV MODEL YIELD STRESS WITH THE SAME VALUE. THIS LINE IS
OVERRIDDEN BY THE 'YSMGOV' AND 'YSUMOD' LINES.

( 8-13) ENTER THE YIELD STRESS VALUE.

LINE REPLACEMENT LEAVE

LABEL YIELD STRESS BLANK
YSUOVR
1-- 6 8<--13 14----------------80

DEFAULT

ENGLISH KSI

METRIC(KN) KN/SQ.CM

METRIC(KG) KG/SQ.CM
 YIELD STRESS MODIFICATION LINE
COLUMNS COMMENTARY __________________________

GENERAL THE 'YSUMOD' LINE IS USED TO REPLACE A SINGLE SACS IV MODEL

YIELD STRESS VALUE WITH A NEW VALUE. THIS LINE IS OVERRIDDEN
BY THE 'YSMGOV' LINE.

( 8-13) ENTER THE FIRST YIELD STRESS VALUE TO BE REPLACED FROM THE
SACS IV MODEL.

(14-19) ENTER THE COLLAPSE YIELD STRESS VALUE TO REPLACE THE SACS IV
MODEL VALUE FROM COLUMNS 8-13.

(20-25) ENTER THE SECOND YIELD STRESS VALUE TO BE REPLACED FROM THE
SACS IV MODEL.

(26-31) ENTER THE COLLAPSE YIELD STRESS VALUE TO REPLACE THE SACS IV
MODEL VALUE FROM COLUMNS 20-25.

(32-37) ENTER THE THIRD YIELD STRESS VALUE TO BE REPLACED FROM THE
SACS IV MODEL.

(38-43) ENTER THE COLLAPSE YIELD STRESS VALUE TO REPLACE THE SACS IV
MODEL VALUE FROM COLUMNS 32-37.

(44-49) ENTER THE FOURTH YIELD STRESS VALUE TO BE REPLACED FROM THE
SACS IV MODEL.

(50-55) ENTER THE COLLAPSE YIELD STRESS VALUE TO REPLACE THE SACS IV
MODEL VALUE FROM COLUMNS 44-49.

(56-61) ENTER THE FIFTH YIELD STRESS VALUE TO BE REPLACED FROM THE
SACS IV MODEL.

(62-67) ENTER THE COLLAPSE YIELD STRESS VALUE TO REPLACE THE SACS IV
MODEL VALUE FROM COLUMNS 56-61.

FIRST NEW SECOND NEW THIRD NEW FOURTH NEW FIFTH NEW
LINE SACS COLLAPSE SACS COLLAPSE SACS COLLAPSE SACS COLLAPSE SACS COLLAPSE
LABEL YIELD YIELD YIELD YIELD YIELD YIELD YIELD YIELD YIELD YIELD
STRESS STRESS STRESS STRESS STRESS STRESS STRESS STRESS STRESS STRESS
YSUMOD
1-- 6 8<--13 14<--19 20<--25 26<--31 32<--37 38<--43 44<--49 50<--55 56<--61 62<--67

DEFAULT

ENGLISH KSI KSI KSI KSI KSI KSI KSI KSI KSI KSI

METRIC(KN) KN/SQ.CM KN/SQ.CM KN/SQ.CM KN/SQ.CM KN/SQ.CM KN/SQ.CM KN/SQ.CM KN/SQ.CM KN/SQ.CM KN/SQ.CM

METRIC(KG) KG/SQ.CM KG/SQ.CM KG/SQ.CM KG/SQ.CM KG/SQ.CM KG/SQ.CM KG/SQ.CM KG/SQ.CM KG/SQ.CM KG/SQ.CM
 YIELD STRESS MEMBER GROUP OVERRIDE LINE
COLUMNS COMMENTARY __________________________

GENERAL THE 'YSMGOV' LINE IS USED TO OVERRIDE THE SACS IV MODEL YIELD
STRESS FOR SPECIFIED MEMBER GROUPS. UP TO 15 MEMBER GROUPS
MAY BE OVERRIDDEN ON ONE LINE.

( 8-13) ENTER THE REPLACEMENT YIELD STRESS. ALL MEMBER GROUPS

SPECIFIED WILL HAVE THEIR YIELD STRESS VALUES IN COLLAPSE SET
EQUAL TO THIS VALUE.

(15-73) ENTER MEMBER GROUPS.

MEMBER GROUP IDENTIFIERS

REPLACE-
LINE MENT
LABEL YIELD
STRESS GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP
ID ID ID ID ID ID ID ID ID ID ID ID ID ID ID
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

YSMGOV
1-- 6 8<--13 15--17 19--21 23--25 27--29 31--33 35--37 39--41 43--45 47--49 51--53 55--57 59--61 63--65 67--69 71--73

DEFAULT

ENGLISH KSI

METRIC(KN) KN/SQ.CM

METRIC(KG) KG/SQ.CM
 YIELD STRESS PLATE GROUP OVERRIDE
COLUMNS COMMENTARY __________________________

GENERAL THIS LINE IS OPTIONAL IN ANY COLLAPSE ANALYSIS RUN. IT IS

USED TO OVERRIDE THE YIELD STRESS FOR SELECTED PLATE GROUPS.

(8-13) ENTER THE NEW YIELD STRESS.

(15-80) ENTER THE PLATE GROUP IDENTIFIERS SELECTED FOR

OVERRIDE. USE AS MANY OF THESE INPUT LINES AS DESIRED.

GROUP PLATE SELECTIONS FOR OUTPUT REPORTS

LINE YIELD
LABEL STRESS PLATE PLATE PLATE PLATE PLATE PLATE PLATE PLATE PLATE PLATE PLATE PLATE PLATE PLATE PLATE
OVERRIDE GROUP ID GROUP ID GROUP ID GROUP ID GROUP ID GROUP ID GROUP ID GROUP ID GROUP ID GROUP ID GROUP ID GROUP ID GROUP ID GROUP ID GROUP ID

YSPGOV

1-- 6 8-->13 15-->17 19-->21 23-->25 27-->29 31-->33 35-->37 39-->41 43-->45 47-->49 51-->53 55-->57 59-->61 63-->65 67-->69 71-->73

DEFAULT

ENGLISH KSI

METRIC KN/SQ.CM
 ELASTIC MEMBER GROUPS
COLUMNS COMMENTARY __________________________

GENERAL MEMBERS OF THESE GROUPS ARE TO BE CONSIDERED AS ELASTIC

DEFLECTION ELEMENTS WITH NO PLASTIC EFFECTS. THIS CAN SIGNIFICANTLY
REDUCE THE TIME REQUIRED TO PERFORM AN ANALYSIS.

NOTE: DESIGNATING ELEMENTS TO REMAIN ELASTIC DOES NOT IMPLY

THEY DO NOT EXPERIENCE LARGE DEFORMATION. FOR EXAMPLE,
IF THE MEMBER IS SUBJECTED TO A LOCALIZED POINT LOAD,
IT MAY BE ELASTICALLY BUCKLED DURING ANALYSIS. TO PREVENT
UNFAVORABLE BUCKLING FOR ELASTIC ELEMENTS, THE USER MAY
USE MEMSEG OR GRPSEG TO SET A SINGLE SUB-SEGMENT TO THE
ELASTIC MEMBER.

( 1- 6) ENTER 'GRPELA'.

(16-18) ENTER THE FIRST GROUP IDENTIFIER. THIS IDENTIFIER MUST

CORRESPOND TO A GROUP IDENTIFIER ON A SACS IV 'GRUP' LINE.

(20-74) THE REMAINING GROUP IDENTIFIER FIELDS ARE SIMILAR.

ANY NUMBER OF 'GRPELA' INPUT LINES CAN BE SPECIFIED.

MEMBER GROUP IDENTIFIERS

LINE
LABEL GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP
ID ID ID ID ID ID ID ID ID ID ID ID ID ID ID
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

GRPELA

1-- 6 16--18 20--22 24--26 28--30 32--34 36--38 40--42 44--46 48--50 52--54 56--58 60--62 64--66 68--70 72--74
 ELASTIC MEMBER INPUT
COLUMNS COMMENTARY __________________________

GENERAL THE FOLLOWING MEMBERS ARE TO BE CONSIDERED AS ELASTIC LARGE

DEFLECTION ELEMENTS WITH NO PLASTIC EFFECTS. THIS CAN SIGNIFICANTLY
REDUCE THE TIME REQUIRED TO PERFORM AN ANALYSIS.

NOTE: DESIGNATING ELEMENTS TO REMAIN ELASTIC DOES NOT IMPLY

THEY DO NOT EXPERIENCE LARGE DEFORMATION. FOR EXAMPLE,
IF THE MEMBER IS SUBJECTED TO A LOCALIZED POINT LOAD,
IT MAY BE ELASTICALLY BUCKLED DURING ANALYSIS. TO PREVENT
UNFAVORABLE BUCKLING FOR ELASTIC ELEMENTS, THE USER MAY
USE MEMSEG OR GRPSEG TO SET A SINGLE SUB-SEGMENT TO THE
ELASTIC MEMBER.

( 1- 6) ENTER 'MEMELA' ON ALL INPUT LINES IN THIS SET.

( 9-12) ENTER JOINT 1 FOR THE FIRST MEMBER.

(14-17) ENTER JOINT 2 FOR THE FIRST MEMBER.

(20-72) REPEAT FOR ADDITIONAL MEMBERS. SIX MEMBERS CAN BE INPUT PER
LINE. REPEAT AS REQUIRED FOR ADDITIONAL MEMBERS.

1ST MEMBER 2ND MEMBER 3RD MEMBER 4TH MEMBER 5TH MEMBER 6TH MEMBER
LINE
LABEL JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT
1 2 1 2 1 2 1 2 1 2 1 2
MEMELA

1-- 6 9-->12 14-->17 20-->23 25-->28 31-->34 36-->39 42-->45 47-->50 53-->56 58-->61 64-->67 69-->72
 SKIPPED MEMBER LOCAL BUCKLING INPUT
COLUMNS COMMENTARY __________________________

GENERAL THE FOLLOWING MEMBERS ARE TO BE CONSIDERED AS ELASTO-PLASTIC

LARGE DEFLECTION ELEMENTS WITH NO LOCAL BUCKLING EFFECTS
INCLUDED. ANY LOCAL BUCKLING CHECK SPECIFIED UNDER THE 'LB'
OPTION WILL BE SKIPPED FOR THESE MEMBERS.

( 1- 6) ENTER 'MEMSKP' ON ALL INPUT LINES IN THIS SET.

( 9-12) ENTER JOINT 1 FOR THE FIRST MEMBER.

(14-17) ENTER JOINT 2 FOR THE FIRST MEMBER.

(20-72) REPEAT FOR ADDITIONAL MEMBERS. SIX MEMBERS CAN BE INPUT PER
LINE. REPEAT AS REQUIRED FOR ADDITIONAL MEMBERS.

1ST MEMBER 2ND MEMBER 3RD MEMBER 4TH MEMBER 5TH MEMBER 6TH MEMBER
LINE
LABEL JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT
1 2 1 2 1 2 1 2 1 2 1 2
MEMSKP

1-- 6 9-->12 14-->17 20-->23 25-->28 31-->34 36-->39 42-->45 47-->50 53-->56 58-->61 64-->67 69-->72
 SKIPPED LOCAL BUCKLING FOR MEMBER GROUPS
COLUMNS COMMENTARY __________________________

GENERAL MEMBERS OF THESE GROUPS ARE TO BE CONSIDERED AS ELASTO-PLASTIC

LARGE DEFLECTION ELEMENTS WITH NO LOCAL BUCKLING EFFECTS
INCLUDED. ANY LOCAL BUCKLING CHECK SPECIFIED UNDER THE 'LB'
OPTION WILL BE SKIPPED FOR MEMBERS FROM THESE GROUPS.

( 1- 6) ENTER 'GRPSKP'.

(16-18) ENTER THE FIRST GROUP IDENTIFIER. THIS IDENTIFIER MUST

CORRESPOND TO A GROUP IDENTIFIER ON A SACS IV 'GRUP' LINE.

(20-74) THE REMAINING GROUP IDENTIFIER FIELDS ARE SIMILAR.

ANY NUMBER OF 'GRPSKP' INPUT LINES CAN BE SPECIFIED.

MEMBER GROUP IDENTIFIERS

LINE
LABEL GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP
ID ID ID ID ID ID ID ID ID ID ID ID ID ID ID
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

GRPSKP

1-- 6 16--18 20--22 24--26 28--30 32--34 36--38 40--42 44--46 48--50 52--54 56--58 60--62 64--66 68--70 72--74
 MEMBER REMOVAL
COLUMNS COMMENTARY __________________________

GENERAL THE FOLLOWING MEMBERS ARE TO BE REMOVED FROM THE ANALYSIS

ON COMPLETION OF A SPECIFIED LOAD INCREMENT. FOR ALL SUBSEQUENT
INCREMENTS, EACH SPECIFIED
MEMBER IS NO LONGER CONSIDERED TO BE PART OF THE STUCTURE.

( 1- 6) ENTER 'MEMREM' ON ALL INPUT LINES IN THIS SET.

( 8-12) LOAD INCREMENT AT WHICH THE MEMBER IS TO BE REMOVED.

(14-17) ENTER JOINT 1 FOR THE FIRST MEMBER.

(19-22) ENTER JOINT 2 FOR THE FIRST MEMBER.

(24-72) REPEAT FOR ADDITIONAL MEMBERS. SIX MEMBERS CAN BE INPUT PER
LINE. REPEAT AS REQUIRED FOR ADDITIONAL MEMBERS.

LOAD 1ST MEMBER 2ND MEMBER 3RD MEMBER 4TH MEMBER 5TH MEMBER 6TH MEMBER
LINE
LABEL JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT
INCREMENT
1 2 1 2 1 2 1 2 1 2 1 2
MEMREM

1-- 6 8-->12 14-->17 19-->22 24-->27 29-->32 34-->37 39-->42 44-->47 49-->52 54-->57 59-->62 64-->67 69-->72
 DUCTILITY LIMIT FOR AN INDIVIDUAL MEMBER
COLUMNS COMMENTARY __________________________

GENERAL EACH OF THE FOLLOWING MEMBERS ARE TO BE REMOVED FROM THE

ANALYSIS WHEN THE SPECIFIED TENSILE DUCTILITY LIMIT HAS
BEEN REACHED.
FOR ALL SUBSEQUENT INCREMENTS, EACH SPECIFIED
MEMBER IS NO LONGER CONSIDERED TO BE PART OF THE STUCTURE.
THE DUCILITY LIMIT IS EXPRESSED AS A PRECENTAGE. THE INDIVIDUAL
MEMBER DUCTILITY LIMIT OVERRIDES THE GLOBAL MEMBER DUCTILITY
LIMIT THAT IS SPECIFIED ON THE CLPOP2 LINE.

( 1- 6) ENTER 'MEMDUC' ON ALL INPUT LINES IN THIS SET.

( 8-12) PERCENTAGE OF DUCTILITY AT WHICH THE MEMBER IS TO BE REMOVED.

(14-17) ENTER JOINT 1 FOR THE FIRST MEMBER.

(19-22) ENTER JOINT 2 FOR THE FIRST MEMBER.

(24-72) REPEAT FOR ADDITIONAL MEMBERS. SIX MEMBERS CAN BE INPUT PER
LINE. REPEAT AS REQUIRED FOR ADDITIONAL MEMBERS.

DUCTILITY 1ST MEMBER 2ND MEMBER 3RD MEMBER 4TH MEMBER 5TH MEMBER 6TH MEMBER
LINE
LABEL LIMIT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT
(%) 1 2 1 2 1 2 1 2 1 2 1 2
MEMDUC

1-- 6 8-->12 14-->17 19-->22 24-->27 29-->32 34-->37 39-->42 44-->47 49-->52 54-->57 59-->62 64-->67 69-->72
 ELASTIC MEMBER INPUT
COLUMNS COMMENTARY __________________________

GENERAL THE FOLLOWING MEMBERS ARE TO BE CONSIDERED AS ELASTIC LARGE

DEFLECTION ELEMENTS WITH NO PLASTIC EFFECTS. THIS CAN SIGNIFICANTLY
REDUCE THE TIME REQUIRED TO PERFORM AN ANALYSIS.

NOTE: DESIGNATING ELEMENTS TO REMAIN ELASTIC DOES NOT IMPLY

THEY DO NOT EXPERIENCE LARGE DEFORMATION. FOR EXAMPLE,
IF THE MEMBER IS SUBJECTED TO A LOCALIZED POINT LOAD,
IT MAY BE ELASTICALLY BUCKLED DURING ANALYSIS. TO PREVENT
UNFAVORABLE BUCKLING FOR ELASTIC ELEMENTS, THE USER MAY
USE MEMSEG OR GRPSEG TO SET A SINGLE SUB-SEGMENT TO THE
ELASTIC MEMBER.

( 1- 6) ENTER 'MEMDEL' ON ALL INPUT LINES IN THIS SET.

( 9-12) ENTER JOINT 1 FOR THE FIRST MEMBER.

(14-17) ENTER JOINT 2 FOR THE FIRST MEMBER.

(20-72) REPEAT FOR ADDITIONAL MEMBERS. SIX MEMBERS CAN BE INPUT PER
LINE. REPEAT AS REQUIRED FOR ADDITIONAL MEMBERS.

1ST MEMBER 2ND MEMBER 3RD MEMBER 4TH MEMBER 5TH MEMBER 6TH MEMBER
LINE
LABEL JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT
1 2 1 2 1 2 1 2 1 2 1 2
MEMDEL

1-- 6 9-->12 14-->17 20-->23 25-->28 31-->34 36-->39 42-->45 47-->50 53-->56 58-->61 64-->67 69-->72
 ELASTIC MEMBER GROUPS
COLUMNS COMMENTARY __________________________

GENERAL MEMBERS OF THESE GROUPS ARE TO BE CONSIDERED AS ELASTIC

DEFLECTION ELEMENTS WITH NO PLASTIC EFFECTS. THIS CAN SIGNIFICANTLY
REDUCE THE TIME REQUIRED TO PERFORM AN ANALYSIS.

NOTE: DESIGNATING ELEMENTS TO REMAIN ELASTIC DOES NOT IMPLY

THEY DO NOT EXPERIENCE LARGE DEFORMATION. FOR EXAMPLE,
IF THE MEMBER IS SUBJECTED TO A LOCALIZED POINT LOAD,
IT MAY BE ELASTICALLY BUCKLED DURING ANALYSIS. TO PREVENT
UNFAVORABLE BUCKLING FOR ELASTIC ELEMENTS, THE USER MAY
USE MEMSEG OR GRPSEG TO SET A SINGLE SUB-SEGMENT TO THE
ELASTIC MEMBER.

( 1- 6) ENTER 'GRPDEL'.

(16-18) ENTER THE FIRST GROUP IDENTIFIER. THIS IDENTIFIER MUST

CORRESPOND TO A GROUP IDENTIFIER ON A SACS IV 'GRUP' LINE.

(20-74) THE REMAINING GROUP IDENTIFIER FIELDS ARE SIMILAR.

ANY NUMBER OF 'GRPDEL' INPUT LINES CAN BE SPECIFIED.

MEMBER GROUP IDENTIFIERS

LINE
LABEL GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP
ID ID ID ID ID ID ID ID ID ID ID ID ID ID ID
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

GRPDEL

1-- 6 16--18 20--22 24--26 28--30 32--34 36--38 40--42 44--46 48--50 52--54 56--58 60--62 64--66 68--70 72--74
 MEMBER SUBSEGMENT SPECIFICATION
COLUMNS COMMENTARY __________________________

GENERAL THE FOLLOWING MEMBERS ARE TO BE DIVIDED INTO A SPECIFIED

NUMBER OF SUBSEGMENTS FOR THE PURPOSE OF AN ELASTO-PLASTIC
ANALYSIS. THIS NUMBER OVERRIDES THE DEFAULT
SETTING FOR THE NUMBER OF SUBSEGMENTS FROM COLUMNS 14-16
OF THE CLPOPT LINE, AS WELL AS SUBSEGMENT OVERRIDES THAT
ARE SPECIFIED USING THE GRPSEG LINE.

( 1- 6) ENTER 'MEMSEG' ON ALL INPUT LINES IN THIS SET.

( 8- 9) THE NUMBER OF SUBSEGMENTS TO BE USED FOR EACH MEMBER ON THE

INPUT LINE.

(14-17) ENTER JOINT 1 FOR THE FIRST MEMBER.

(19-22) ENTER JOINT 2 FOR THE FIRST MEMBER.

(24-72) REPEAT FOR ADDITIONAL MEMBERS. SIX MEMBERS CAN BE INPUT PER
LINE. REPEAT AS REQUIRED FOR ADDITIONAL MEMBERS.

NUMBER OF 1ST MEMBER 2ND MEMBER 3RD MEMBER 4TH MEMBER 5TH MEMBER 6TH MEMBER
LINE
LABEL JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT JOINT
SUBSEGMENTS
1 2 1 2 1 2 1 2 1 2 1 2
MEMSEG

1-- 6 8--> 9 14-->17 19-->22 24-->27 29-->32 34-->37 39-->42 44-->47 49-->52 54-->57 59-->62 64-->67 69-->72
 MEMBER GROUP SUBSEGMENT SPECIFICATION
COLUMNS COMMENTARY __________________________

GENERAL THE FOLLOWING MEMBER GROUPS CONTAIN MEMBERS THAT ARE TO BE

DIVIDED INTO A SPECIFIED
NUMBER OF SUBSEGMENTS FOR THE PURPOSE OF AN ELASTO-PLASTIC
ANALYSIS. THIS NUMBER OVERRIDES THE DEFAULT
SETTING FOR THE NUMBER OF SUBSEGMENTS FROM COLUMNS 14-16
OF THE CLPOPT LINE.

( 1- 6) ENTER 'GRPSEG' ON ALL INPUT LINES IN THIS SET.

( 8- 9) THE NUMBER OF SUBSEGMENTS TO BE USED FOR EACH MEMBER THAT

IS CONTAINED BY A MEMBER GROUP ON THE INPUT LINE.

(14-16) ENTER MEMBER GROUP NAME.

(18-72) REPEAT FOR ADDITIONAL MEMBER GROUPS. FIFTEEN MEMBER GROUPS

CAN BE INPUT PER
LINE. REPEAT AS REQUIRED FOR ADDITIONAL MEMBER GROUPS.

NO. OF MEMBER GROUP IDENTIFIERS

LINE
LABEL GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP
SUBSEGS ID ID ID ID ID ID ID ID ID ID ID ID ID ID ID
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
GRPSEG

1-- 6 8--> 9 14--16 18--20 22--24 26--28 30--32 34--36 38--40 42--44 46--48 50--52 54--56 58--60 62--64 66--68 70--72
 NONLINEAR SPRING INPUT
COLUMNS COMMENTARY __________________________

GENERAL THIS LINE IS OPTIONAL IN ANY COLLAPSE ANALYSIS RUN. IT IS

USED TO SPECIFY NONLINEAR SUPPORTS. ANY NUMBER OF NONLINEAR
SUPPORTS CAN BE USED.

( 8-11) ENTER THE JOINT NAME WHERE THE SUPPORT IS LOCATED.

(13-14) ENTER THE DEGREE OF FREEDOM FOR THIS CONSTRAINT FROM THE
FOLLOWING:
'DX' - DEFLECTION IN GLOBAL X-DIRECTION
'DY' - DEFLECTION IN GLOBAL Y-DIRECTION
'DZ' - DEFLECTION IN GLOBAL Z-DIRECTION
'RX' - ROTATION ABOUT GLOBAL X-AXIS
'RY' - ROTATION ABOUT GLOBAL Y-AXIS
'RZ' - ROTATION ABOUT GLOBAL Z-AXIS

(15-78) ENTER THE FORCE-DEFLECTION OR MOMENT-ROTATION PAIRS OF VALUES

IN ORDER OF INCREASING DEFLECTION/ROTATION USING THE UNITS
BELOW:

* TYPE * ** ENGLISH ** ** METRIC-KN ** ** METRIC-KG **

FORCE KIP KN KG
MOMENT KIP-FT KN-M KG-M
DEFLECTION IN CM CM
ROTATION RADIANS RADIANS RADIANS

ANY NUMBER OF POINTS CAN BE ENTERED BY REPEATING THIS DATA,

LEAVING THE JOINT NAME BLANK ON SUBSEQUENT LINES.

1ST POINT 2ND POINT 3RD POINT 4TH POINT

DEGREE
LINE JOINT
OF
LABEL NAME FORCE DEFLECTION FORCE DEFLECTION FORCE DEFLECTION FORCE DEFLECTION
FREEDOM
OR OR OR OR OR OR OR OR
MOMENT ROTATION MOMENT ROTATION MOMENT ROTATION MOMENT ROTATION
NLSPRG

1-- 6 8-->11 13--14 15<--22 23<--30 31<--38 39<--46 47<--54 55<--62 63<--70 71<--78
 JOINT TO JOINT NONLINEAR SPRING INPUT
COLUMNS COMMENTARY __________________________

GENERAL THIS LINE IS OPTIONAL IN ANY COLLAPSE ANALYSIS RUN. IT IS

USED TO SPECIFY NONLINEAR SPRINGS BETWEEN JOINTS. ANY NUMBER
OF NONLINEAR SPRINGS CAN BE USED.

( 8-11) ENTER THE BEGIN JOINT OF THE SPRING.

(12-15) ENTER THE END JOINT OF THE SPRING.

(18-19) ENTER THE DEGREE OF FREEDOM FOR THIS CONSTRAINT FROM THE
FOLLOWING:
'DX' - DEFLECTION IN LOCAL X-DIRECTION
'DY' - DEFLECTION IN LOCAL Y-DIRECTION
'DZ' - DEFLECTION IN LOCAL Z-DIRECTION
'RX' - ROTATION ABOUT LOCAL X-AXIS
'RY' - ROTATION ABOUT LOCAL Y-AXIS
'RZ' - ROTATION ABOUT LOCAL Z-AXIS
NOTE THAT LOCAL COORDINATES ARE DEFINED IN THE SAME MANNER AS
FOR MEMBERS.

(25-80) ENTER THE FORCE-DEFLECTION OR MOMENT-ROTATION PAIRS OF VALUES

IN ORDER OF INCREASING DEFLECTION/ROTATION USING THE UNITS
BELOW:

* TYPE * ** ENGLISH ** ** METRIC-KN ** ** METRIC-KG **

FORCE KIP KN KG
MOMENT KIP-FT KN-M KG-M
DEFLECTION IN CM CM
ROTATION RADIANS RADIANS RADIANS

ANY NUMBER OF POINTS CAN BE ENTERED BY REPEATING THIS DATA,

LEAVING THE JOINTS BLANK ON SUBSEQUENT LINES.

1ST POINT 2ND POINT 3RD POINT 4TH POINT

DEGREE
LINE BEGIN END
OF
LABEL JOINT JOINT FORCE DEFLECTION FORCE DEFLECTION FORCE DEFLECTION FORCE DEFLECTION
FREEDOM
OR OR OR OR OR OR OR OR
MOMENT ROTATION MOMENT ROTATION MOMENT ROTATION MOMENT ROTATION
NLSPJJ

1-- 6 8-->11 12-->15 18--19 25<--31 32<--38 39<--45 46<--52 53<--59 60<--66 67<--73 74<--80
 COROTATIONAL NONLINEAR AXIAL SPRING INPUT
COLUMNS COMMENTARY __________________________

GENERAL THIS LINE IS OPTIONAL IN ANY COLLAPSE ANALYSIS RUN. IT IS

USED TO SPECIFY A COROTATIONAL AXIAL NONLINEAR SPRING
BETWEEN JOINTS. ANY NUMBER
OF NONLINEAR AXIAL SPRINGS CAN BE USED.

( 8-11) ENTER THE BEGIN JOINT OF THE SPRING.

(12-15) ENTER THE END JOINT OF THE SPRING.

NOTE THAT LOCAL COORDINATES ARE DEFINED

USING THE UPDATED CO-ORDINATES. THIS IS IN CONTRAST TO
TRADITIONAL NONLINEAR SPRINGS, WHICH HAVE LOCAL CO-ORDINATES
DEFINED USING THE UNDEFORMED STATE OF THE STRUCTURE.

(25-80) ENTER THE FORCE-DEFLECTION OR MOMENT-ROTATION PAIRS OF VALUES

IN ORDER OF INCREASING DEFLECTION/ROTATION USING THE UNITS
BELOW:

* TYPE * ** ENGLISH ** ** METRIC-KN ** ** METRIC-KG **

FORCE KIP KN KG
MOMENT KIP-FT KN-M KG-M
DEFLECTION IN CM CM
ROTATION RADIANS RADIANS RADIANS

ANY NUMBER OF POINTS CAN BE ENTERED BY REPEATING THIS DATA,

LEAVING THE JOINTS BLANK ON SUBSEQUENT LINES.

1ST POINT 2ND POINT 3RD POINT 4TH POINT

LINE BEGIN END
LABEL JOINT JOINT FORCE DEFLECTION FORCE DEFLECTION FORCE DEFLECTION FORCE DEFLECTION
OR OR OR OR OR OR OR OR
MOMENT ROTATION MOMENT ROTATION MOMENT ROTATION MOMENT ROTATION

NLSPST

1-- 6 8-->11 12-->15 25<--31 32<--38 39<--45 46<--52 53<--59 60<--66 67<--73 74<--80
 MATERIAL PROPERTY HEADER LINE
COLUMNS COMMENTARY __________________________

GENERAL USE THIS LINE TO START MATERIAL MODEL DEFINITION. THIS LINE MUST BE FOLLOWED BY 'MATPRP PLAS' LINE(S).

( 1- 6) ENTER 'MATPRP'.

( 8-11) ENTER 'HEAD'.

(13-16) ENTER THE NAME OF MATERIAL MODEL. IT SHOUDL BE SAME AS

THE ONE SPECIFIED IN THE PRECEDING 'MATGRP' LINE.

(17-80) SHOULD BE LEFT BLANK.

THIS LINE MUST BE FOLLOWED BY 'MATPRP PLAS' LINE(S) TO DEFINE

THE POST YIELD BEHAVIOR FOR THIS MATERIAL MODEL.

MATERIAL
LINE HEAD
MODEL
LABEL LABEL
NAME
MATPRP HEAD

1-- 6 8--11 13--16 17------------------80

 PLASTIC STRESS-STRAIN INPUT LINE
COLUMNS COMMENTARY __________________________

GENERAL THIS LINE IS USED TO INPUT THE POST YIELD PLASTIC STRAIN -
STRESS FACTOR DATA FOR MATERIAL MODELS SPECIFIED USING
'MATGRP' LINES. THE FIRST DATA POINT IN THE FIRST
'MATPRP PLAS' CARD REPRESENTS THE YIELD POINT AND MUST BE
ENTERED AS (0.0, 1.0). THE PLASTIC STRAIN AND THE STRESS
FACTOR VALUES MUST INCREASE MONOTONICALLY THEREAFTER.

FOR VALUES OF PLASTIC STRAIN GREATER THAN THE LARGEST

SPECIFIED VALUE, THE VALUE OF STRESS FACTOR IS CALCULATED
USING THE DEFAULT STRAIN HARDENING RATIO SPECIFIED IN THE
CLPOPT LINE.

( 1- 6) ENTER 'MATPRP'.

( 8-11) ENTER 'PLAS'.

(13-72) ENTER PLASTIC STRAIN AND STRESS FACTOR VALUES FOR EACH
DATA POINT.
PLASTIC STRAIN IS EQUAL TO (MECHANICAL STRAIN - YIELD STRAIN).
STRESS FACTOR IS EQUAL TO (TOTAL STRESS / YIELD STRESS).

THIS LINE MAY BE REPEATED AS NECESSARY. MAXIMUM 50 DATA POINTS

ARE ALLOWED FOR EACH MATERIAL MODEL.

PLASTIC STRAIN - STRESS CURVE

LINE LINE 1ST POINT

2ND POINT 3RD POINT 4TH POINT 5TH POINT
LABEL TYPE (YIELD)
PLASTIC STRESS PLASTIC STRESS PLASTIC STRESS PLASTIC STRESS PLASTIC STRESS
STRAIN FACTOR STRAIN FACTOR STRAIN FACTOR STRAIN FACTOR STRAIN FACTOR
MATPRP PLAS

1-- 6 8--11 13<--18 19<--24 25<--30 31<--36 37<--42 43<--48 49<--54 55<--60 61<--66 67<--72

DEFAULT
 MATERIAL MODEL ASSIGNMENT TO MEMBER GROUPS
COLUMNS COMMENTARY __________________________

GENERAL THIS LINE IS OPTIONAL IN ANY COLLAPSE ANALYSIS RUN. IT IS

USED TO ASSIGN A MATERIAL MODEL (POST-YILED STRESS-STRAIN
BEHAVIOR) TO DIFFERENT MEMBER GROUPS AVAILABLE IN SACS INPUT
FILE. IF A MEMBER IS NOT ASSIGNED ANY MATERIAL MODEL, IT
WILL HAVE A CONSTANT STRAIN-HARDEDNING RATIO AS DEFINED IN
THE 'CLPOPT' LINE.

( 1- 6) ENTER 'MATGRP'.

( 8-11) ENTER THE MATERIAL MODEL NAME.

(13-15) ENTER 'ALL' IF ALL THE MEMBERS ARE

TO BE ASSIGNED THIS MATERIAL MODEL. ONLY THE FIRST 'ALL'
ASSIGNMENT WILL BE CONSIDERED. ALL OTHER MATERIAL
MODEL ASSIGNMENTS WILL BE IGNORED.

(17-75) ENTER MEMBER GROUP IDENTIFIERS. MAXIMUM OF 15 MEMBER GROUPS

CAN BE SPECIFIED IN A LINE. REPEAT THE 'MATGRP' CARD
FOR ASSIGNING MORE MEMBER GROUPS TO THE SAME MATERIAL MODEL.

'MATGRP' CARDS SHOULD BE ENTERED "AFTER" MATPRP HEAD' AND 'MATPRP

PLAS' CARDS, WHICH ARE USED TO DEFINE THE MATERIAL MODEL.

MEMBER GROUP IDENTIFIERS

MATERIAL
LINE SELECT
MODEL GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP
LABEL 'ALL'
NAME ID ID ID ID ID ID ID ID ID ID ID ID ID ID ID
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

MATGRP

1-- 6 8-->11 13--15 17--19 21--23 25--27 29--31 33--35 37--39 41--43 45--47 49--51 53--55 57--59 61--63 65--67 69--71 73--75
 MATERIAL MODEL ASSIGNMENT TO PLATE GROUPS
COLUMNS COMMENTARY __________________________

GENERAL THIS LINE IS OPTIONAL IN ANY COLLAPSE ANALYSIS RUN. IT IS

USED TO ASSIGN A MATERIAL MODEL (POST-YILED STRESS-STRAIN
BEHAVIOR) TO DIFFERENT PLATE GROUPS AVAILABLE IN SACS INPUT
FILE. IF A PLATE IS NOT ASSIGNED ANY MATERIAL MODEL, IT
WILL HAVE A CONSTANT STRAIN-HARDEDNING RATIO AS DEFINED IN
THE 'CLPOPT' LINE.

( 1- 6) ENTER 'MATPGR'.

( 8-11) ENTER THE MATERIAL MODEL NAME.

(13-15) ENTER 'ALL' IF ALL THE PLATES ARE

TO BE ASSIGNED THIS MATERIAL MODEL. ONLY THE FIRST 'ALL'
ASSIGNMENT WILL BE CONSIDERED. ALL OTHER MATERIAL
MODEL ASSIGNMENTS WILL BE IGNORED.

(17-75) ENTER PLATE GROUP IDENTIFIERS. MAXIMUM OF 15 PLATE GROUPS

CAN BE SPECIFIED IN A LINE. REPEAT THE 'MATPGR' CARD
FOR ASSIGNING MORE PLATE GROUPS TO THE SAME MATERIAL MODEL.

'MATPGR' CARDS SHOULD BE ENTERED "AFTER" 'MATPRP HEAD' AND 'MATPRP

PLAS' CARDS, WHICH ARE USED TO DEFINE THE MATERIAL MODEL.

PLATE GROUP IDENTIFIERS

MATERIAL
LINE SELECT
MODEL GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP
LABEL 'ALL'
NAME ID ID ID ID ID ID ID ID ID ID ID ID ID ID ID
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

MATPGR

1-- 6 8-->11 13--15 17--19 21--23 25--27 29--31 33--35 37--39 41--43 45--47 49--51 53--55 57--59 61--63 65--67 69--71 73--75
 MATERIAL MODEL ASSIGNMENT TO PILE GROUPS
COLUMNS COMMENTARY __________________________

GENERAL THIS LINE IS OPTIONAL IN ANY COLLAPSE ANALYSIS RUN. IT IS

USED TO ASSIGN A MATERIAL MODEL (POST-YILED STRESS-STRAIN
BEHAVIOR) TO DIFFERENT PILE GROUPS AVAILABLE IN SACS INPUT
FILE. IF A PILE IS NOT ASSIGNED ANY MATERIAL MODEL, IT
WILL HAVE A CONSTANT STRAIN-HARDEDNING RATIO AS DEFINED IN
THE 'CLPOPT' LINE.

( 1- 6) ENTER 'MATPGL'.

( 8-11) ENTER THE MATERIAL MODEL NAME.

(13-15) ENTER 'ALL' IF ALL THE PILES ARE

TO BE ASSIGNED THIS MATERIAL MODEL. ONLY THE FIRST 'ALL'
ASSIGNMENT WILL BE CONSIDERED. ALL OTHER MATERIAL
MODEL ASSIGNMENTS WILL BE IGNORED.

(17-75) ENTER PILE GROUP IDENTIFIERS. MAXIMUM OF 15 PILE GROUPS

CAN BE SPECIFIED IN A LINE. REPEAT THE 'MATPLG' CARD
FOR ASSIGNING MORE PILE GROUPS TO THE SAME MATERIAL MODEL.

'MATPLG' CARDS SHOULD BE ENTERED "AFTER" 'MATPRP HEAD' AND 'MATPRP

PLAS' CARDS, WHICH ARE USED TO DEFINE THE MATERIAL MODEL.

PILE GROUP IDENTIFIERS

MATERIAL
LINE SELECT
MODEL GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP GROUP
LABEL 'ALL'
NAME ID ID ID ID ID ID ID ID ID ID ID ID ID ID ID
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

MATPLG

1-- 6 8-->11 13--15 17--19 21--23 25--27 29--31 33--35 37--39 41--43 45--47 49--51 53--55 57--59 61--63 65--67 69--71 73--75
 END LINE
COLUMNS COMMENTARY __________________________

GENERAL THIS LINE SIGNIFIES THE END OF THE COLLAPSE INPUT AND IS THE
LAST LINE OF THE COLLAPSE INPUT.

LINE
LEAVE BLANK
LABEL
END

1-- 3 4------------------------80