PDC - TR 06 02 PDF
PDC - TR 06 02 PDF
US Army Corps
of Engineers ®
FOREWORD
This document was prepared for the U.S. Army Corps of Engineers Protective Design
Center by Protection Engineering Consultants, LLC (PEC) and Baker Engineering and
Risk Consultants, Inc. (BakerRisk). PEC and BakerRisk made every reasonable effort to
perform the work contained herein in a manner consistent with high professional
standards.
The work was conducted on the basis of information made available to PEC and
BakerRisk. Neither PEC, BakerRisk, nor any person acting on its behalf makes any
warranty or representation, expressed or implied, with respect to the accuracy,
completeness, or usefulness of this information. All observations, conclusions and
recommendations contained herein are relevant only to the project, and should not be
applied to any other facility or operation.
Any third party use of this Report or any information or conclusions contained therein
shall be at the user's sole risk. Such use shall constitute an agreement by the user to
release, defend and indemnify BakerRisk from and against any and all liability in
connection therewith (including any liability for special, indirect, incidental or
consequential damages), regardless of how such liability may arise.
PEC and BakerRisk regard the work that it has done as being advisory in nature. The
responsibility for use and implementation of the conclusions and recommendations
contained herein rests entirely with the client.
This information is furnished by the United States Government and is accepted and
used by the recipient with the express understanding that the United States
Government makes no warranties, expressed or implied, concerning the accuracy,
completeness, reliability, usability, or suitability for any particular purpose of the
information and data contained in this document, and the United States Government
shall be under no liability whatsoever to any person by reason of any use made thereof.
CONTENTS
Page
CHAPTER 1 - INTRODUCTION .................................................................................. 1-1
1-1 Purpose and Scope.......................................................................................... 1-1
1-2 Applicability ...................................................................................................... 1-1
CHAPTER 2 - Free-Field Airblast Load........................................................................ 2-1
2-1 General Discussion of Explosions.................................................................... 2-1
2-1.1 Deflagration vs. Detonation ...................................................................... 2-1
2-2 Blast Load Parameters..................................................................................... 2-2
2-3 Cube Root Scaling ........................................................................................... 2-4
2-4 Prediction Equations for Blast Load parameters from a Free-Air Burst Of
Spherical TNT........................................................................................................... 2-5
2-5 Surface Burst for Hemispherical TNT............................................................... 2-8
2-5.1 Near Surface Burst................................................................................. 2-12
2-6 Charge mass Equivalency Factors................................................................. 2-14
2-6.1 TNT Equivalency Factor for High Explosive Type .................................. 2-14
2-6.2 Equivalency Factor for Cased Explosives .............................................. 2-17
2-6.3 Non-Hemispherical Charge Geometry ................................................... 2-18
2-7 Confined Detonations..................................................................................... 2-20
2-8 Blast Walls and Revetments .......................................................................... 2-22
CHAPTER 3 - Blast Loads on Structures ..................................................................... 3-1
3-1 Reflected Blast Loads ...................................................................................... 3-1
3-2 Clearing of Reflected Blast Loads .................................................................... 3-5
3-3 Consideration of Negative Phase Blast Load ................................................... 3-8
3-4 Structural Components with Non-Uniform Blast Loads .................................... 3-9
3-5 Blast Loads on Primary Framing Components............................................... 3-15
3-5.1 Loaded Area Factor, Af, for Components with Concentrated Loads ...... 3-17
CHAPTER 4 - Single-Degree-of-Freedom (SDOF) Analysis of Structural Component
Response to Blast Load ............................................................................................... 4-1
4-1 Dynamic Response of Elastic and Elastic-Plastic SDOF Systems................... 4-1
4-1.1 Transformation Factors ............................................................................ 4-2
4-1.2 Resistance-Deflection Relationships ........................................................ 4-7
4-1.3 Damping................................................................................................. 4-49
4-1.4 Initial Displacement and Velocity............................................................ 4-51
4-2 Methods for Solving SDOF Equation of Motion.............................................. 4-53
4-2.1 Time Step............................................................................................... 4-56
4-3 Shear Forces and End Reaction Forces ........................................................ 4-57
4-3.1 Dynamic Reaction Forces ...................................................................... 4-58
4-3.2 Equivalent Static Reaction Force ........................................................... 4-63
4-4 Combined Axial and Lateral Loads ................................................................ 4-65
4-4.1 P-delta Effects........................................................................................ 4-66
i
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
4-5 Maximum Response Parameters ................................................................... 4-70
4-5.1 Component Damage Levels and Building Levels of Protection.............. 4-71
4-5.2 Shear Controlled Response ................................................................... 4-74
4-6 Strain-rate to First Yield.................................................................................. 4-75
4-7 Connections ................................................................................................... 4-76
4-8 Limitations of SDOF Analysis......................................................................... 4-77
CHAPTER 5 - Reinforced Concrete Components........................................................ 5-1
5-1 Dynamic Material Properties ............................................................................ 5-1
5-1.1 Concrete Dynamic Strength ..................................................................... 5-1
5-1.2 Reinforcing Steel Dynamic Strength ........................................................ 5-3
5-2 Flexural Response of Reinforced Concrete Components ................................ 5-4
5-2.1 Ultimate Dynamic Moment Capacity of Reinforced Concrete Beams and
Slabs 5-5
5-2.2 Dynamic Moment Capacity of Prestressed Concrete Beams and Slabs .. 5-8
5-2.3 Moment of Inertia and Modulus of Elasticity........................................... 5-10
5-3 Shear Response of Reinforced Concrete Components ................................. 5-12
5-3.1 Dynamic Diagonal Shear Capacity of Concrete ..................................... 5-14
5-3.2 Dynamic Diagonal Shear Strength of Prestressed Concrete ................. 5-16
5-3.3 Required Steel Shear Reinforcement for Diagonal Shear...................... 5-16
5-3.4 Direct Shear ........................................................................................... 5-18
5-4 Combined Axial and Lateral Load .................................................................. 5-19
5-4.1 Reinforced Concrete Components With Combined Axial Tension and
Lateral Loads ...................................................................................................... 5-19
5-4.2 Reinforced Concrete Beam-Columns With Combined Compression and
Lateral Load ........................................................................................................ 5-20
5-5 Deep Beams .................................................................................................. 5-21
5-6 Tension and Compression Membrane Response .......................................... 5-22
5-7 Reinforcing Steel Development Lengths and Splices..................................... 5-22
5-8 Rebound......................................................................................................... 5-23
5-9 Connections ................................................................................................... 5-23
CHAPTER 6 - Steel and Aluminum Components....................................................... 6-27
6-1 Dynamic Material Properties .......................................................................... 6-27
6-2 Flexural Response of Structural Steel Components....................................... 6-29
6-2.1 Ultimate Dynamic Moment Capacity of Plates ....................................... 6-29
6-2.2 Ultimate Dynamic Moment Capacity of Hot Rolled Beams .................... 6-31
6-2.3 Ultimate Dynamic Moment Capacity of Cold-Formed Components ....... 6-32
6-3 Metal STud Walls ........................................................................................... 6-34
6-4 Steel Beam-Columns With Combined Flexure and Compressive Axial Load. 6-35
6-5 Shear Capacity of Plates, Panels, and Beams............................................... 6-36
6-6 Steel Components with Tension Membrane Response ................................. 6-38
6-7 Open Web Steel Joists................................................................................... 6-38
6-8 Rebound......................................................................................................... 6-41
ii
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
6-9 Connections ................................................................................................... 6-42
CHAPTER 7 - Masonry Components ......................................................................... 7-46
7-1 Types of Masonry Walls ................................................................................. 7-46
7-2 Dynamic Masonry Properties ......................................................................... 7-50
7-3 Masonry Reinforcement ................................................................................. 7-51
7-4 Flexural Response of Masonry....................................................................... 7-53
7-4.1 Flexural Response of Unreinforced Masonry ......................................... 7-53
7-4.2 Flexural Response of Reinforced Masonry Walls................................... 7-54
7-4.3 Moment of Inertia of Masonry Walls....................................................... 7-56
7-5 Shear Strength of Masonry ............................................................................ 7-57
7-6 Compression Membrane Response Between Rigid Supports and Axial Load
Arching in Masonry Walls ....................................................................................... 7-58
7-7 Double Wythe Unreinforced Masonry Walls................................................... 7-58
7-8 Rebound Response ....................................................................................... 7-62
CHAPTER 8 - Wood Components ............................................................................... 8-1
8-1 Wood component Types .................................................................................. 8-1
8-2 Wood Material Properties................................................................................. 8-1
8-3 SDOF Analysis of Wood Building Components................................................ 8-4
8-3.1 Blast Tests on Wood Buildings................................................................. 8-4
8-3.2 SDOF Analyses of Wood Components Subjected to Blast Loads............ 8-5
8-4 Dynamic Moment Capacity of Wood Components........................................... 8-6
8-5 Wood Connections ........................................................................................... 8-7
CHAPTER 9 - Pressure-Impulse and Charge Weight-Standoff Diagrams ................... 9-1
9-1 Pressure-Impulse Diagrams............................................................................. 9-1
9-1.1 P-i Diagrams for Blast Loads with Positive Phase Load Only .................. 9-2
9-1.2 P-i Diagrams for Blast Loads with Positive and Negative Phase Load..... 9-3
9-2 Charge Weight-Standoff Diagrams .................................................................. 9-6
9-3 Limitations of Pressure-Impulse and Charge-Weight Standoff Diagrams ........ 9-6
9-4 Scaled Pressure-Impulse Diagrams ................................................................. 9-9
CHAPTER 10 - REFERENCES ................................................................................. 10-1
iii
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
FIGURES
Figure Title Page
Figure 2-1 Typical Pressure-Time History of an Airblast in Free Air............................2-2
Figure 2-2 Pressure History Shape Used in SBEDS for Negative Phase of Load.......2-3
Figure 2-3 Simplified Right Triangular Blast Pressure History for Positive Phase Blast
Load ......................................................................................................................2-4
Figure 2-4 Positive Phase Shock Wave Parameters for a Spherical TNT Explosion in
Free Air at Sea Level (Metric) ...............................................................................2-6
Figure 2-5 Positive Phase Shock Wave Parameters for a Spherical TNT Explosion in
Free Air at Sea Level (English) .............................................................................2-6
Figure 2-6 Negative Phase Shock Wave Parameters for a Spherical TNT Explosion in
Free Air at Sea Level (Metric) ...............................................................................2-7
Figure 2-7 Negative Phase Shock Wave Parameters for a Spherical TNT Explosion in
Free Air at Sea Level (English) .............................................................................2-7
Figure 2-8 Typical Reflected and Side-on Pressure-Time Histories ............................2-8
Figure 2-9 Positive Phase Shock Wave Parameters for a Hemispherical Surface Burst
of TNT at Sea Level (Metric) .................................................................................2-9
Figure 2-10 Positive Phase Shock Wave Parameters for a Hemispherical Surface Burst
of TNT at Sea Level (English) ............................................................................. 2-10
Figure 2-11 Negative Phase Shock Wave Parameters for a Hemispherical Surface
Burst of TNT at Sea Level (Metric)...................................................................... 2-10
Figure 2-12 Negative Phase Shock Wave Parameters for a Hemispherical Surface
Burst of TNT at Sea Level (English).................................................................... 2-11
Figure 2-13 Peak Side-On Overpressure Versus Peak Dynamic Pressure, Density of
Air Behind Shock Front, and Particle Velocity..................................................... 2-11
Figure 2-14 Mach Stem Effect from Near Surface Burst ............................................ 2-12
Figure 2-15 Shock Profile Above and Below the Triple Point .................................... 2-13
Figure 2-16 Scaled Height of Triple Point.................................................................. 2-13
Figure 2-17 Equivalent Experimental Charge Masses for Various Casing Materials 2-18
Figure 2-18 Azimuth Angles, θ, For Blast Relative to Longitudinal Axis of Cylindrical
Shaped Charge (Initiation Point at θ= 0) ............................................................. 2-19
Figure 2-19 Peak Pressure at 0, 90, 180 Degree Azimuth Angles from Cylindrical
Charge with L/D=5 .............................................................................................. 2-19
Figure 2-20 Impulse at 0, 90, 180 Degree Azimuth Angles from Cylindrical Charge with
L/D=5 .................................................................................................................. 2-20
Figure 2-21 Blast Pressure From Confined Explosion .............................................. 2-21
Figure 2-22 Wave Diffraction Around Walls .............................................................. 2-22
Figure 2-23 High Calculated Shock Pressures (in Red) Within L-Shaped Enclosure In
Front of Door ....................................................................................................... 2-23
Figure 3-1 Blast Loads on Building Surfaces ..............................................................3-1
Figure 3-2 Reflected Pressure Coefficient Versus Angle of Incidence ........................3-2
Figure 3-3 Scaled Reflected Impulse Versus Angle of Incidence................................3-3
Figure 3-4 Plan View Showing Angle of Incidence of Shock Front Relative to Building
Wall .......................................................................................................................3-3
Figure 3-5 Shock Wave Loads on Building Surface ....................................................3-4
Figure 3-6 Numerical Calculations Showing Pressure Spike at Angle of Large Angle of
Incidence...............................................................................................................3-5
iv
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Figure 3-7 Reflected Pressure Coefficient Versus Angle of Incidence Without Hump
(from ASCE)..........................................................................................................3-5
Figure 3-8 Reflected Blast Load Affected by Clearing.................................................3-6
Figure 3-9 Sound Velocity in Reflected Overpressure Region ....................................3-8
Figure 3-10 Dimensions of Reflecting Surface ............................................................3-8
Figure 3-11 Equivalent Uniform Roof and Side Wall Loading ................................... 3-12
Figure 3-12 Equivalent Load Factor and Blast Wave Location Ratio ........................ 3-13
Figure 3-13 Back Wall Blast Loading ........................................................................ 3-14
Figure 3-14 Comparison of Blast Load and Dynamic Reaction for Secondary
Component ......................................................................................................... 3-15
Figure 3-15 Dynamic Reactions Over Tributary Area of Supporting Component ....... 3-17
Figure 3-16 Example of Roof Girder with Loaded Area Factor (AF) Equal to 0.5 ...... 3-18
Figure 4-1 Equivalent Spring-Mass SDOF System .....................................................4-1
Figure 4-2 Deflected Shape Functions for Simply Supported Beam ...........................4-5
Figure 4-3 Load-Mass Factors for Plastic Response of Two-Way Spanning
Components........................................................................................................ 4-10
Figure 4-4 Fixed-End Beam in Ductile, Flexural Response....................................... 4-12
Figure 4-5 Resistance-Deflection Curve For Flexural Response .............................. 4-12
Figure 4-6 Location of Symmetrical Yield Lines for Two-Way Component with Four
Edges Supported ................................................................................................ 4-18
Figure 4-7 Location of Symmetrical Yield Lines for Two-Way Component with Three
Edges Supported ................................................................................................ 4-19
Figure 4-8 Location of Yield Lines for Two-Way Component with Two Adjacent Edges
Supported ........................................................................................................... 4-19
Figure 4-9 Stiffness Coefficients for Uniformly-Loaded Two-Way Component with Three
Edges Simply-Supported and One Edge Free .................................................... 4-23
Figure 4-10 Illustration of Tension Membrane Response.......................................... 4-24
Figure 4-11 Resistance-Deflection Relationship for Steel Stud with Restrained
Boundaries (from Salim et al, 2003).................................................................... 4-25
Figure 4-12 Resistance Deflection Curve for Steel Components with Tension
Membrane ........................................................................................................... 4-26
Figure 4-13 Comparison of Measured Maximum Deflections of Lightweight Steel
Beams to Predicted Values with SBEDS Method................................................ 4-27
Figure 4-14. Resistance-Deflection Curve for Reinforced Concrete and Masonry
Components with Compression and Tension Membrane.................................... 4-28
Figure 4-15. Boundary Conditions for Tension and Compression Membrane Response
4-30
Figure 4-16 Compression Ring Effect Providing In-Plane Restraint for Tension
Membrane in Two-Way Component (from Park and Gamble, 2000) .................. 4-31
Figure 4-17 Forces Developing During Compression Membrane Response ............ 4-32
Figure 4-18 Response of Brittle Component Under Combined Lateral and Axial Load.4-
38
Figure 4-19 Resistance-Deflection Relationships for Unreinforced Masonry with Brittle
Flexural Response and Axial Load...................................................................... 4-40
Figure 4-20 General Resistance-Deflection Diagrams in SBEDS ............................. 4-44
Figure 4-21 Strain Energy for Component with Ductile Flexural and Tension Membrane
Response ............................................................................................................ 4-47
v
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Figure 4-22 Ratio of Maximum Rebound Resistance to Inbound Ultimate Resistance for
SDOF System ..................................................................................................... 4-48
Figure 4-23 Undamped Response Charts for SDOF Response to Right Triangular Blast
Load History ........................................................................................................ 4-54
Figure 4-24 Free Body Diagram Used to Determine Dynamic Reaction Force for
Uniformly Loaded Beam (From Biggs) ................................................................ 4-58
Figure 4-25 Connection Between Components Subject to Internal Blast Loads ....... 4-66
Figure 4-26 Support Rotation Angle .......................................................................... 4-70
Figure 5-1 Strain-Rate vs. Concrete Compression Strength .......................................5-2
Figure 5-2 Strain-Rate vs. Concrete Splitting Tensile Strength ...................................5-3
Figure 5-3 Strain-Rate vs. Reinforcing Steel Yield and Ultimate Strength...................5-4
Figure 5-4 Type I and II Cross Sections ......................................................................5-7
Figure 5-5 Coefficient for Moment of Inertia of Cracked Sections ................................... 5-11
Figure 5-6 Locations of Critical Sections for Diagonal Tension Shear ....................... 5-13
Figure 5-7 Shear Steel Reinforcement for Blast-Loaded Components ..................... 5-17
Figure 5-8 Diagonal Bars Resisting Direct Shear ...................................................... 5-19
Figure 5-9. Reinforced Concrete Beam-Column Axial Load Moment Capacity
Interaction Diagram............................................................................................. 5-21
Figure 5-10 Example Connections at Top and Bottom of Precast Wall Panel (from
ASCE) ................................................................................................................. 5-24
Figure 5-11 Example Connection Between Precast Concrete Wall Panels (from ASCE)
5-25
Figure 6-1 Strain-Rate vs. Structural Steel Yield Strength ........................................ 6-28
Figure 6-2 Typical Moment-Curvature Relationships for Steel Plates and Beams .... 6-30
Figure 6-3 Local Buckling in Maximum Moment Region of Cold-Formed Beams and
Panels ................................................................................................................. 6-33
Figure 6-4 Moment Resisting Steel Beam Connections Designed for Large Rotations
(from TM 5-1300) ................................................................................................ 6-44
Figure 7-1 Typical Masonry Unit Shapes .................................................................. 7-47
Figure 7-2 Percent Solid Thickness Through Web of CMU Cross Section................ 7-48
Figure 7-3 Double Wythe Cavity Wall with Brick and CMU ....................................... 7-50
Figure 7-4 Horizontal Joint Reinforcement ................................................................ 7-52
Figure 8-1 Typical Static Stress-Strain Relationship for Wood in Flexure ...................8-2
Figure 8-2 Wood Frame Building After Long Duration Blast Load with 34 KPa (5 psi)
Peak Overpressure ...............................................................................................8-5
Figure 8-3 Wood Stud Hanger Connections................................................................8-7
Figure 9-1 Illustration of P-i Diagram For Positive Phase Blast Load Showing Three
Regions of Response............................................................................................9-1
Figure 9-2 P-i Diagram for Positive Phase Blast Load Only ........................................9-3
Figure 9-3 P-i Diagram for Positive and Negative Phase Blast Load ..........................9-4
Figure 9-4 P-i Diagram with Positive and Negative Phase Blast Load with Points from
Maximum Deflection During Second Inbound Response ......................................9-5
Figure 9-5 Representative SDOF Output for Highlighted Cases in Figure 9-4 ............9-6
Figure 9-6 Charge Weight-Standoff Diagram for Positive and Negative Phase Blast
Load ......................................................................................................................9-7
Figure 9-7 Scaled P-i Curve-fits vs. Scaled SDOF Points in Terms of Support Rotation
for Flexural Response of Reinforced Concrete Slabs ......................................... 9-10
vi
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
TABLES
Table Title Page
Table 2-1 TNT Equivalency Factors for Free-Air Explosions..................................... 2-15
Table 2-2 TNT Equivalency Factors for Free-Air Explosions (Continued) .................. 2-16
Table 3-1 Side-on Element Dynamic Drag Coefficients ............................................ 3-13
Table 4-1 Load and Mass Factors for One-Way Components ....................................4-8
Table 4-2 Load-Mass Factors for Elastic Response of Two-Way Spanning Components
4-9
Table 4-3 Flexural Resistances, Stiffnesses, and Support Shears for One-Way
Members ............................................................................................................. 4-16
Table 4-4 Ultimate Flexural Resistances and Support Shears for Uniformly-Loaded
Two- Way Members with Uniform Moment Capacities........................................ 4-17
Table 4-5 Approximate Flexural Resistances, Stiffnesses, and Support Shears for Two-
Way Members ..................................................................................................... 4-20
Table 4-6 Consideration of Compression Membrane Between Rigid Supports for
SBEDS Component Types.................................................................................. 4-33
Table 4-7 Definition of Terms for Stiffness Region Criteria Tables............................ 4-45
Table 4-8 Criteria for Determining To Stay in Current Stiffness Region During Dynamic
Response ............................................................................................................ 4-45
Table 4-9 Rebound Flag Information......................................................................... 4-46
Table 4-10 Damping Ratio Information For Primarily Elastic Dynamic Response..... 4-51
Table 4-11 Dynamic Reaction Force Coefficients for One-Way Components........... 4-60
Table 4-12 Dynamic Reaction Coefficients for Two-Way Slabs with Four Sides Fixed
and Two Sides Fixed........................................................................................... 4-61
Table 4-13 Dynamic Reaction Coefficients for Two-Way Slabs with Four Sides Simply
Supported ........................................................................................................... 4-62
Table 4-14 Dynamic Reaction Coefficients for Two-Way Slabs with Opposite Edges
Fixed and Simply Supported ............................................................................... 4-62
Table 4-15 Dynamic Reaction Coefficients for Two-Way Slabs with Three Edges Fixed
4-63
Table 4-16 Ultimate Shear Stress at Distance d from Face of Support for Two-Way
Elements ............................................................................................................. 4-65
Table 4-17 SBEDS Component Types with P-Delta Effects...................................... 4-68
Table 4-18 Calculated Deflections from SBEDS for W12x40 Beam-Column Compared
to Theoretical Values (Moment Magnifier) .......................................................... 4-69
Table 4-19 Levels of Protection – New and Existing Buildings.................................. 4-72
Table 4-20 Response Limits for Reinforced Concrete (R/C) and Reinforced Masonry
(R/M) Components Controlled by Shear Response from ASCE ......................... 4-74
Table 4-21 ASCE Component Damage Level Descriptions ...................................... 4-75
Table 5-1 Dynamic Yield Strength Information for Conventional Reinforcing Steels ...5-4
Table 5-2 Cross Sectional Area Information for Welded Wire Fabric (WWF)..............5-5
Table 5-3 Cross Sectional Area Information for Steel Prestressing Strands ...............5-6
Table 5-4. Depth to Reinforcing Steel Used by SBEDS for Shear Strength Calculations
5-13
Table 5-5 Values for Constants in Equation 5-20...................................................... 5-19
Table 6-1 Material Strength Increase Factors for Structural Steel............................. 6-28
Table 6-2 Material Strength Increase Factors for Aluminum ..................................... 6-29
vii
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Table 6-3 Dynamic Shear Strengths for Cold-Formed Steel Panels at Fixed Interior
Supports (Combined Bending and Shear) .......................................................... 6-37
Table 6-4 Guidance on Including Tension Membrane Response in SBEDS Calculations
6-39
Table 7-1 Masonry Block Information........................................................................ 7-48
Table 7-2 Assumed Cross Section Parameters for CMU Blocks............................... 7-48
Table 7-3 Information on Mortar Used for Masonry.................................................... 7-49
Table 7-4. Recommended Conservative Values for Static Masonry Prism Compressive
Strength .............................................................................................................. 7-51
Table 8-1 Allowable Bending Stresses for Wood Species from NDS ..........................8-1
Table 8-2 Modulus of Elasticity for Wood Species from NDS......................................8-2
Table 8-3 House Damage from Very Long Duration Blast Loads................................8-4
viii
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
CHAPTER 1 - INTRODUCTION
This Technical Report explains the methodology for the SBEDS workbook. The
SBEDS workbook is an Excel-based tool for design of structural components subjected
to dynamic loads using single degree of freedom (SDOF) methodology. It was
developed by for the U.S. Army Corps of Engineers Protective Design Center as a tool
for designers to use in satisfying Department of Defense (DoD) antiterrorism standards.
The SBEDS workbook is intended for structural engineers with some experience in
structural dynamics and blast effects. It is not for the non-structural engineer. SBEDS is
suited for preliminary design or final design when used by a skilled engineer. SBEDS
will aid the engineer in design of the member, but the actual design of members and
connections is the full responsibility of the engineer.
1-2 APPLICABILITY
The SBEDS workbook and the methodology in this report applies to new construction,
major renovations, and leased buildings and must be utilized in accordance with the
applicability requirements of UFC 4-010-01 Minimum Antiterrorism Standards for
Buildings (UFC 4-010-01) or as directed by Service Guidance. See UFC 4-010-01 for
additional detail on the structures that must be considered.
1-1
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
1-2
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
CHAPTER 2 - FREE-FIELD AIRBLAST LOAD
In the far field, a process known as “shocking up” occurs to pressure waves, where the
central and latter parts of the pressure wave move faster through air that is densified by
the leading edge of the wave and “catch up” to the front of the wave to create a shock
wave, where the peak pressure is applied nearly instantaneously. It is possible that
blast pressures from the detonation and deflagration of explosive sources with
equivalent energies can be almost the same in the far field, particularly if the reaction
speed for the deflagration is near the speed of sound in the explosive material.
Shocking up is a gradual process and the distance over which it occurs depends on a
number of different variables (Baker et al 1983).
It is always conservative to treat an explosion as a detonation, particularly if it involves
the rapid reaction of a solid material. For that reason, this methodology manual will
focus on airblast pressures created by detonation of high explosives. Calculation of
2-1
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
airblast loads from deflagrations is a more complicated process that is outside the
scope of this manual (Baker et al 1983).
The shape of the positive phase blast load from an open air explosion in Figure 2-1 can
be represented mathematically with Equation 2-1, which is known as the modified
Friedlander equation. The decay coefficient θ can be determined from the integral of
Equation 2-1 if the positive phase peak pressure and impulse are known. Blast loads
inside confined areas, or near surrounding buildings that can reflect the blast load, will
cause blast load histories with different shapes as discussed later in this manual. Figure
2-2 shows the approximate equation used in SBEDS to model the shape of the negative
phase blast load (Granstrom, 1956), which has not been studied or understood as well
as the positive phase. The integral of the negative phase blast load equation in Figure
2-2 sometimes differs from the impulse determined from available empirical prediction
equations unless a correction term Cp- is calculated and applied to the time variable, as
stated in Equation 2-2. A number of trials have shown that Cp- is usually within the
range of 0.9 to 1.1.
2-2
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Equation 2-1
⎡ ⎛ t − t A ⎞⎤ −⎛⎜⎝ t −θt A ⎞⎟⎠
Ps (t ) = Pso ⎢1 − ⎜⎜ ⎟⎟⎥ e
⎣ ⎝ o ⎠⎦t
t A ≤ t ≤ t A + to
where: Ps(t)= shock overpressure as a function of time (MPa)
Pso = peak side-on overpressure (MPa)
t =detonation time (ms)
tA = arrival time of initial shock front (ms)
to = positive phase duration (ms)
θ = shape constant of pressure waveform
Figure 2-2 Pressure History Shape Used in SBEDS for Negative Phase of Load
0.00
2
0.20 P− 27 ⎛⎜ t − ⎞⎛
⎟⎜1 − t
− ⎞
⎟
=
0.40 Po − 4 ⎜⎝ to − ⎟⎜
⎠⎝ to − ⎟
⎠
P /P o
-
-
0.60
0.80
1.00
0 0.2 0.4 0.6 0.8 1
- -
t /t o
where: P-o = peak negative pressure
P
2-3
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
will cause similar or somewhat more structural response than the actual blast load,
depending on the ratio of the blast load duration to the natural period of the structural
component. Note that fictitious duration td in Figure 2-3 is less than the actual positive
blast load duration, to, in Figure 2-1.
Figure 2-3 Simplified Right Triangular Blast Pressure History for Positive Phase
Blast Load
Pressure
Impulse, (i)
Time
0 t
d
Note that the charge radius is proportional to the cube root of the charge mass. In this
case, the blast load parameters with dimensions of pressure and velocity will be equal
for both cases at the equal “scaled” times. The scaled time is time divided by the cube
root of the charge mass (i.e., t/W1/3). Therefore, the peak pressures at time zero will be
equal for two explosion scenario cases with the same scaled standoffs and the pressure
will decay to zero over different duration times t0 for the two cases, where t0/W1/3 is the
same for both cases. Blast parameters with units of time, such as duration, arrival time,
and impulse are equal for two geometrically similar explosion cases (i.e., having the
same scaled standoff) only in terms of their scaled values (i.e., t/W1/3, i/W1/3), rather than
in absolute values of i and t.
2-4
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Equation 2-3
1
R1 ⎛ W1 ⎞ 3 R
=⎜ ⎟ or Z1 = Z 2 where Z =
R2 ⎜⎝ W2 ⎟⎠ 1
W3
where: Wi = explosive mass (kg) for case i
Ri = standoff distance from detonation point of Wi (m)
Zi = scaled standoff for case i
Cube-root scaling relationships have been proven experimentally for explosive masses
ranging from several grams to tens of thousands of kilograms. Measured blast load
parameters from a given type of explosion scenario, such as hemispherical or spherical
TNT explosions, can be plotted in terms of the scaled standoff to measurement points
and measured pressures, velocities, scaled blast load durations, and scaled blast load
impulses at those points to create relatively simple plots in terms of scaled standoff
distance and scaled blast parameters. The scaling relations apply when there are: (1)
identical ambient conditions, (2) identical charge shapes, and (3) identical charge-to-
surface geometries to those in the experiments used to develop the plots. However,
reasonable values can be obtained using the scaling relations even when these
conditions are not exactly the same.
2-5
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Figure 2-4 Positive Phase Shock Wave Parameters for a Spherical TNT Explosion
in Free Air at Sea Level (Metric)
1000 1/3
Pr, MPa ta, ms/kg
500 Pso, MPa to, ms/kg1/3
Ir, MPa-ms/kg1/3 U, m/ms
200 Is, MPa-ms/kg
1/3
Lw, m/kg
1/3
100
50
20
10
5
2
1
0.5
0.2
0.1
0.05
0.02
0.01
0.005
0.002
0.001
0.05 0.1 0.2 0.3 0.50.7 1 2 3 4 5 678 10 20 30 50
1/3
Scaled Distance Z = R/W1/3, (m/kg )
Figure 2-5 Positive Phase Shock Wave Parameters for a Spherical TNT Explosion
in Free Air at Sea Level (English)
100000 1/3
Pr, psi ta, ms/lb
50000 1/3
Pso, psi to, ms/lb
20000 Ir, psi-ms/lb1/3 U, ft/ms
1/3 1/3
10000 Is, psi-ms/lb Lw, ft/lb
5000
2000
1000
500
200
100
50
20
10
5
2
1
0.5
0.2
0.1
0.05
0.02
0.01
0.005
0.1 0.2 0.3 0.50.7 1 2 3 4 5 67 10 20 30 50 70 100
1/3
Scaled Distance Z = R/W1/3,(ft/lb )
2-6
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Figure 2-6 Negative Phase Shock Wave Parameters for a Spherical TNT
Explosion in Free Air at Sea Level (Metric)
20
10
7
5 -Pr, MPa
-Pso, MPa
3
-Ir, MPa-ms/kg1/3
1/3
2 -Is, MPa-ms/kg
1/3
-to, ms/kg
1 -Lw, m/kg1/3
0.7
0.5
0.3
0.2
0.1
0.07
0.05
0.03
0.02
0.01
0.05 0.1 0.2 0.3 0.5 0.7 1 2 3 4 5 6 78 10 20
1/3
Scaled Distance Z = R/W1/3,(m/kg )
Figure 2-7 Negative Phase Shock Wave Parameters for a Spherical TNT
Explosion in Free Air at Sea Level (English)
100
-Pr, psi
70 -Pso, psi
50 -Ir, psi-ms/lb1/3
1/3
-Is, psi-ms/lb
1/3
30 -to, ms/lb
-Lw, ft/lb1/3
20
10
7
5
3
2
1
0.7
0.5
0.3
0.2
0.1 0.2 0.3 0.5 0.7 1 2 3 4 5 6 78 10 20 30 50
1/3
Scaled Distance Z = R/W1/3, (ft/lb )
2-7
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
The general trends in Figure 2-4 and Figure 2-6 are that the pressure and impulse of
the blast load decrease as the standoff increases and duration terms increase with
increasing standoff distance. This is consistent with the fact that the energy in the shock
wave dissipates as it propagates through a larger volume of air and is distributed over a
longer wavelength. These figures apply for a spherical free-air burst, which is the
simplest blast loading case and the basis of many tests. However, this case only applies
when both the charge and the point of interest are “far” from any large surface that
impedes blast wave propagation and prevent spherical expansion (other than a surface
containing the point of interest). From a practical perspective, this means that the point
of interest, which may be on a surface or in the free-field, must be significantly closer to
the charge than to the ground or to any other surrounding building or walls. An example
of this is an elevated explosion above a building roof with points of interest on the roof.
A more typical case is an explosion that occurs much nearer to the ground than to the
point of interest, which may be a distant building. In this case the blast load consists of
the pressure pulse for a free-air explosion of the charge mass plus the shock wave
reflection off the ground, which also propagates to the point of interest. This is
discussed further in the next section. Thus, the total blast load for most practical cases
includes load from the shock wave reflection off the ground, and is therefore greater
than the blast loads predicted for a free-air burst explosion.
2-8
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
used for charge masses with spherical shapes and other shapes at, or near the surface.
The effect of non-hemispherical charge shape is discussed in Section 2-6.3.
Blast loads from a surface burst of a given charge mass are more intense than from a
free-air explosion because the energy in the shock wave is confined by the ground
surface so that, at any standoff distance, it only expands into one-half the volume as a
free-air explosion (i.e., hemispherically instead of spherically). As a rough approximation
based on test data, the blast loads from a surface burst with a given charge mass W are
the same as those from a free-air explosion with a charge mass of 1.8W. The
multiplication factor would be closer to 2.0 if the ground were perfectly rigid.
Figure 2-13 shows relationships between several other blast load parameters and the
peak side-on blast pressure, which should be determined from Figure 2-4 or Figure 2-9
as applicable. Typically, it is conservative to use Figure 2-9. The dynamic pressure is
the “wind” pressure from air particles accelerated by the shock wave. It dynamic
pressure history can be approximated as having the same shape and duration as the
positive shock pressure history and a peak pressure, qs, from Figure 2-13. The dynamic
pressure is included in the predicted reflected blast loads and it is included in the blast
load after clearing of the reflected blast load, as discussed in Section 3-2. The dynamic
pressure creates suction on components not facing the explosive source, similar to wind
loads, and therefore reduces blast pressures on these components. This effect is
usually secondary and is often ignored, although it is included in some cases as
described in Section 3-4. The peak particle velocity and density of air particles caused
by the shock wave is also shown in Figure 2-13.
Figure 2-9 Positive Phase Shock Wave Parameters for a Hemispherical Surface
Burst of TNT at Sea Level (Metric)
1000 Pr, MPa ta, ms/kg
1/3
100
50
20
10
5
2
1
0.5
0.2
0.1
0.05
0.02
0.01
0.005
0.002
0.05 0.1 0.2 0.3 0.50.7 1 2 3 4 5 678 10 20 30 50
1/3
Scaled Distance Z = R/W1/3, (m/kg )
2-9
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Figure 2-10 Positive Phase Shock Wave Parameters for a Hemispherical Surface
Burst of TNT at Sea Level (English)
200000 Pr, psi ta, ms/lb
1/3
100000 Pso, psi to, ms/lb1/3
50000 Ir, psi-ms/lb1/3 U, ft/ms
1/3 1/3
20000 Is, psi-ms/lb Lw, ft/lb
10000
5000
2000
1000
500
200
100
50
20
10
5
2
1
0.5
0.2
0.1
0.05
0.02
0.01
0.005
0.1 0.2 0.3 0.50.7 1 2 3 4 5 67 10 20 30 50 70 100
1/3
Scaled Distance Z = R/W1/3, (ft/lb )
Figure 2-11 Negative Phase Shock Wave Parameters for a Hemispherical Surface
Burst of TNT at Sea Level (Metric)
20
10
7
5
3
2
-Pr, MPa
1 -Pso, MPa
0.7 -Ir, MPa-ms/kg1/3
1/3
0.5 -Is, MPa-ms/kg
1/3
-to, ms/kg
0.3
-Lw, m/kg1/3
0.2
0.1
0.07
0.05
0.03
0.02
0.01
0.007
0.005
0.2 0.3 0.5 0.7 1 2 3 4 5 6 7 8 10 20 30 4050
1/3
Scaled Distance Z = R/W1/3, (m/kg )
2-10
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Figure 2-12 Negative Phase Shock Wave Parameters for a Hemispherical Surface
Burst of TNT at Sea Level (English)
100
70
50
30
20
10
7
5
3
2
1
0.7
0.5
0.3
0.2 -Pr, psi
-Pso, psi
0.1 -Ir, psi-ms/lb1/3
1/3
0.07 -Is, psi-ms/lb
1/3
0.05 -to, ms/lb
1/3
-Lw, ft/lb
0.03
0.02
0.1 0.2 0.3 0.50.7 1 2 3 4 5 67 10 20 30 50 70 100
1/3
Scaled Distance Z = R/W1/3, (ft/lb )
Figure 2-13 Peak Side-On Overpressure Versus Peak Dynamic Pressure, Density
of Air Behind Shock Front, and Particle Velocity
2-11
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
2-5.1 NEAR SURFACE BURST
Figure 2-14 shows the case of a near surface explosion, where the height of burst H is
small relative to the standoff to the points of interest on the building, RG. In this case,
both the incident shock wave (solid lines) and the ground reflection of the shock wave
(dashed lines) propagate towards the building. It can be assumed simplistically that the
shock wave’s angle of reflection off the ground at each point is approximately equal to
its angle of incidence, as for an elastic wave. The ground reflection wave moves
through air that was densified by the incident shock wave, which allows it to travel faster
than the incident shock wave, catch up to it, and merge to form a Mach front.
The Mach front includes the energy of both the incident and ground reflection and has
essentially the same intensity as the shock wave from a surface burst, even though the
charge exploded some distance above the surface. The height of the Mach front
increases as the wave propagates away from the center of the detonation. The height of
the triple point at radial distances from the explosive source is referred to as the path of
the triple point and is formed by the intersection of the initial, reflected, and Mach waves
as shown in Figure 2-14. The blast load from the Mach front below the triple point has a
single pressure pulse, as shown in Figure 2-15. This figure also shows that the blast
load above the triple point consists of two pressure pulses, one from the incident shock
wave and one from the ground reflection, that superimpose in time. Figure 2-16 gives
the scaled height of the triple point as a function of the scaled horizontal distance from
the charge and the scaled charge height.
Figure 2-14 Mach Stem Effect from Near Surface Burst
The near surface explosion in Figure 2-14 is the most common case for terrorist threats
and accidental explosions. It is generally conservative to assume the blast load for this
case is caused by the Mach front and equals that from a surface burst of the given
charge mass. If the point of interest is above the triple point, the blast load will have less
peak pressure and about the same impulse, so that it is a less severe case. If a detailed
analysis of the blast load is merited, the BLASTX computer program (Britt et al, 2001)
can be used to explicitly account for the ground reflection and calculate a multi-pulse
2-12
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
blast load similar to Case b in Figure 2-15. Computational fluid dynamics (CFD)
computer programs can also be used for this purpose.
Figure 2-15 Shock Profile Above and Below the Triple Point
2-13
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
2-6 CHARGE MASS EQUIVALENCY FACTORS
The prediction methods to determine blast load parameters for TNT explosions
presented previously in this chapter can be extended to include other explosives,
charge shapes, and casing effects by multiplying the charge mass by an equivalency
factor. This factor is equal to ratio of mass of the given explosive necessary to produce
the same blast effects as a unit mass of bare spherical TNT for the same explosion
conditions (i.e., free-air, surface burst, etc.) In general, the equivalency factor is not a
constant for all blast parameters (such as peak pressure and impulse) and for all scaled
distances from the explosion. Ideally, equivalency factors are determined by testing or
based on factors determined from previous test results.
As shown in Table 2-1 and Table 2-2, the TNT equivalency values for peak pressure
and impulse are not always the same. This presents a potential problem since structural
component response to blast load is generally sensitive to both the peak pressure and
impulse and inaccurate response can be calculated if either value is incorrect.
Fortunately, TNT equivalency values for peak pressure and impulse are almost always
within 15%. In practice, the TNT equivalency factors for peak pressure and impulse are
typically averaged and the average value applied to the charge mass. The resulting
error is almost always within the range of other uncertainties associated with the blast
load prediction methods.
If a large discrepancy exists between the TNT equivalency factors for peak pressure
and impulse, an assessment can be made of whether the peak pressure or impulse has
more affect on the structural response for which the blast load is calculated, and the
TNT equivalency factor selected accordingly. In almost all cases of conventional, non-
hardened construction, structural component response is more sensitive the impulse of
the blast load than the peak pressure for loading by explosions outside the building. In
this case, the TNT equivalency factor for the impulse would be used. The sensitivity of
structural component response to the peak pressure and the impulse from a blast load
is discussed more in Section 9-1. The TNT equivalency factor can be estimated from its
heat of detonation relative to that of TNT, as shown in Equation 2-4, if the explosive is
not listed in Table 2-1 and Table 2-2 and there is no available blast measurements that
can be used to empirically determine the TNT equivalency factor.
2-14
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
2-15
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
2-16
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Equation 2-4
Figure 2-17 is a plot of Equation 2-5, with experimental data plotted as points. The plot
shows considerable scatter between Equation 2-5 and the experimental data, especially
for materials other than steel. The legend in Figure 2-17 indicates that a variety of
materials were used as casings, and W´ is generally greater than W. All casing
materials were highly brittle or were held together with a brittle matrix material.
Equation 2-5 is not recommended for GP bombs, since data shows that peak pressure
and impulse from these weapons correlates better when no casing effects are
considered.
2-17
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Figure 2-17 Equivalent Experimental Charge Masses for Various Casing Materials
2-18
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Figure 2-18 Azimuth Angles, θ, For Blast Relative to Longitudinal Axis of
Cylindrical Shaped Charge (Initiation Point at θ= 0)
D
L = Length of charge
L along longitudinal axis
D = Charge diameter
Figure 2-19 Peak Pressure at 0, 90, 180 Degree Azimuth Angles from Cylindrical
Charge with L/D=5
2-19
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Figure 2-20 Impulse at 0, 90, 180 Degree Azimuth Angles from Cylindrical Charge
with L/D=5
At scaled standoffs greater than 5.0 m/kg1/3, the blast loads from a cylindrical charge
are relatively equal at all azimuth angles to the blast loads from a spherical or
hemispherical shaped charge of equal mass. Figure 2-19 and Figure 2-20 show the
magnitudes of positive phase peak pressure and scaled impulse from the surface burst
of a long cylindrical charge with L/D=5 relative to those from a hemispherical surface
burst. The BLASTX computer program (Britt et al, 2001) can also be used to calculate
the blast loads from cylindrical high explosive charges.
The airblast pulses from a reverberating shock wave can be predicted using the image
charge method, where the pulse from the shock wave reflection off each surface of the
room is equal to the pressure pulse from a spherical free-air burst of an “image” charge
2-20
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
with the same weight as the actual charge, located at a standoff from the point of
interest equal to the full path length of the shock wave to the reflecting wall and then to
the point of interest (such as paths B and C in Figure 2-21), impacting the point of
interest with an angle of incidence equal to that of the reflected wave. The calculated
pressure pulses from all surfaces can be linearly superimposed in time as a first
approximation, or they can be combined more accurately using non-linear superposition
laws based on computational fluid dynamics principles such as the LAMB nonlinear
shock wave addition equations (Needam and Wittwer, 1977).
Figure 2-21 Blast Pressure From Confined Explosion
The peak quasistatic pressure is primarily a function of the charge weight relative to the
confined volume. Simplistically, the rise time to peak quasistatic pressure is often
assumed to coincide with the duration of the shock pulses, so that there is a direct
transition from shock loading to the peak quasistatic pressure. The duration of the gas
pressure load is a function of the vent area, including open vent areas and vent area
covered by frangible surfaces failed by the blast loading during the duration of the
quasistatic pressure. The decay time of the quasistatic pressure can be very long
relative to the blast load from the shock wave.
The relative importance of the shock and quasistatic parts of an internal blast load
depends on many factors and may not be apparent until a complete response analysis
is performed. For example, even though the peak gas pressure may be much lower
than the peak shock pressure, the gas pressure impulse could be many times greater
than the shock impulse and could control the response of a structural component of
interest. Therefore, both loadings must be considered when evaluating the effects of an
internal explosion. Detailed discussion of methods to determine confined blast loads is
outside the scope of this manual. This discussion is provided in UFC 3-340-01, TM 5-
1300, and DOE/TIC 11628.
2-21
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
2-8 BLAST WALLS AND REVETMENTS
In general, blast walls and revetments are only effective for reducing blast loads in the
immediate vicinity behind the wall or revetment (i.e., within one or two wall or revetment
heights). They are much more effective at stopping fragments. Figure 2-22 shows the
process by which a blast wave propagates above and around a wall, which involves a
turbulent area of vortices immediately behind the wall. The shock wave “fills in” the
area behind the wall in a relatively efficient manner so that pressure reduction only
occurs over a limited area behind the wall. More detailed methods to calculate the blast
load behind a wall are discussed in UFC 3-340-01.
Figure 2-22 Wave Diffraction Around Walls
Figure 2-23 shows how a shock wave fills in the area within an L-shaped shield wall
projecting out from a building wall in front of a door, and reflects off interior surfaces of
the shield wall, based on computational fluid dynamics (CFD) modeling results. The wall
is impacted by a shock wave propagating normal to the wall surface, which has
reflected off the wall and is propagating around the edges of the wall. This analysis
illustrates how a shock wave can fill into, and reverberate within, an enclosure to cause
enhanced blast loading. The relatively high pressures caused by multiple reflections of
the shock wave occurring within the L-shaped shield in the CFD model was validated
with shock tube testing (Thomas et al, 2004).
2-22
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Figure 2-23 High Calculated Shock Pressures (in Red) Within L-Shaped
Enclosure In Front of Door
2-23
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
2-24
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
CHAPTER 3 - BLAST LOADS ON STRUCTURES
When a blast wave propagates into a building surface that blocks the propagation and
redirects the blast wave, the pressure, density, and temperature near the surface are
increased to values greater than those for the incident, or incoming blast wave by the
reflection process. This causes the reflected blast load applied on any surface that
faces, or partially faces, the explosive source, and therefore reflects the blast wave, to
be greater than the side-on blast load at the same point. The reflected blast load has
maximum values based on the fully reflected blast pressure and impulse described in
Chapter 2 if the surface faces directly towards the explosive charge.
Figure 3-1 illustrates blast loads applied to different surfaces of a building. Each surface
can experience a different blast pressure based on its orientation relative to the
direction of shock wave propagation. A reflected blast pressure load is shown on the
side of the building facing the explosive source. The blast wave reflects off the building
face back towards the source. The shock wave also envelops the whole building. The
blast load on all sides of the building parallel to the direction of shock wave propagation,
such as the flat roof in Figure 3-1 and sidewalls of a building, is equal to the side-on
blast load. The blast load on the back or leeward surface of a building can
conservatively be assumed equal to the side-on blast load. However, it can be affected
significantly by vortices in the shock wave at the corners of the surface and is often less
than the side-on blast load.
Figure 3-1 Blast Loads on Building Surfaces
PRESSURE
TIME
AMBIENT
PRESSURE
PEAK
PRESSURE
Pso
Pr Building _ Pso
Pb <
Time = T 0 T1 T2 T3 T4 T5
3-1
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
phase of the blast load can be predicted based on the side-on blast load at the
reflecting surface and the angle of incidence using Figure 3-2 and Figure 3-3. Figure 3-2
shows values for the reflection factor, which is the ratio of the peak positive phase
reflected pressure to the peak side-on pressure predicted with Figure 2-4 and Figure
2-9. Therefore, the peak reflected pressure is equal to the reflection factor multiplied by
the peak side-on pressure. The reflected impulse is equal to the scaled reflected
impulse in Figure 3-3 multiplied by the cube root of the charge mass. The angle of
incidence is shown in Figure 3-4 for the case of a plane wave propagating into the wall
of a building, where all points on the wall have the same angle of incidence. This is
typically the case when the standoff is large compared to the building dimensions. If a
building surface is directly normal to the direction of blast wave propagation, which
corresponds a zero angle of incidence (α = 0 in Figure 3-2 and Figure 3-3), a fully
reflected blast load is applied to the surface as described in Chapter 2. The peak
pressure and impulse from a fully reflected blast load can be predicted directly from
Figure 2-4 or Figure 2-9, or from the side-on peak pressure from these two figures and
Figure 3-2 and Figure 3-3 using α = 0.
Figure 3-2 Reflected Pressure Coefficient Versus Angle of Incidence
3-2
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Figure 3-3 Scaled Reflected Impulse Versus Angle of Incidence
Figure 3-4 Plan View Showing Angle of Incidence of Shock Front Relative to
Building Wall
α = Angle of Incidence
α Components at Locations
a, b, or c (anywhere on
Blast wall) all have same α
Wave
Blast Source Front
3-3
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Figure 3-5 shows how different points on a building surface subject to a close-in
explosion have differing angles of incidence,α, as well as differing standoff distances, R.
The blast load at each point has a peak pressure equal to the reflection factor from
Figure 3-2 multiplied by Pso, as determined from Chapter 2 based on the TNT charge
weight and standoff distance (assuming a free-air burst in this case), and an impulse
equal to the scaled reflected impulse from Figure 3-3 multiplied by the cube root of the
charge weight. The value of Pso from Chapter 2 is needed for both Figure 3-2 and
Figure 3-3. SBEDS calculates the blast load at a given point on the building component
based on the user inputs of charge weight, range, component height and width and
angle of incidence.
Figure 3-5 Shock Wave Loads on Building Surface
There is a “hump” in the reflection factors in Figure 3-2 at angles between 45 and 70
degrees, depending on the peak side-on pressure, where the reflection factor increases
with the angle of incidence instead of decreasing as it generally does. Based on
analysis and testing, this is due to a very short duration pressure spike that occurs at
the very front of the pressure pulse for these cases (Ketchum et al, 1998). Figure 3-6
shows calculated pressure spikes with durations on the order of microseconds at the
front of the pressure pulse that cause the hump in the reflection factors in Figure 3-2.
These pressure spikes, which were not noted for angles of incidences and pressures
outside the hump region, were confirmed with shock tube testing. Numerous analyses
of structural components showed that this type of pressure spike with very short
durations does not significantly affect response of typical components in hardened or
conventional buildings (Ketchum et al, 1998). This is expected because the spike
applies a very small amount of impulse with a duration much less than the natural
period of almost any structural component. Based on this work, the curves defining
reflection factors in Figure 3-2 can be smoothed so that there is monotonic decay in the
reflection factor with increased angle of incidence, as shown in Figure 3-7 from ASCE
(1997). However, it is current practice to conservatively include the effect of the hump
3-4
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
in Figure 3-2 for blast loads used in most analysis and design of structural components
subject to blast loads.
Figure 3-8 shows a blast load affected by clearing if tc is less than the equivalent
triangular load duration of the reflected blast load, trf in Equation 3-1. In this case
clearing occurs prior to the end of the reflected blast load and the pressure after tc is
defined by the side-on blast pressure, Pso, and the dynamic pressure, q0, multiplied by a
drag coefficient, CD. The dynamic pressure is shown as qs in Figure 2-13. The drag
coefficient is equal to 1.0 for points on flat building surfaces that reflect the shock wave.
For other cases, such as blast loads on a cylindrical surface, it can be based on
available testing that preferably is done at particle velocities (see Figure 2-13)
representative of those caused by the shock wave. DOE-TIC 11268 (1992) has limited
information on drag coefficients for components with different shapes. The combined
side-on and dynamic pressure acts on the component between times tc and tof from
Equation 3-2. If the clearing time is greater than trf, there is no effect from clearing.
Figure 3-8 also shows a simplified representation of the negative blast load.
Equation 3-1
trf = 2ir / Pr
where:
trf = triangular loading duration
ir = reflected impulse
Pr = peak reflected pressure
3-6
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Equation 3-2
tof = 2is / Pso
where:
tof = positive phase duration
is = overpressure impulse
Pso = peak side-on overpressure
The clearing time, tc, can be determined using one of two equations depending on the
type of blast load that is calculated. The average clearing time needed to clear the
reflected pressure on an entire surface from free edges at the sides and roof is given by
Equation 3-3. The clearing time for a point on the wall is more closely defined by
Equation 3-4. Equation 3-3 is typically used when the calculated blast load represents
the average blast load over an entire structural component. This is the most typical case
for practical problems and the assumption in the SBEDS workbook.
Equation 3-3
4 HW
tc =
(W + 2 H )Cr
where:
tc = average reflected pressure clearing time on a building surface (ms)
H =structure height – see Figure 3-10 (m)
W =structure width – see Figure 3-10 (m)
Cr = sound velocity (see Figure 3-9) (m/ms)
Equation 3-4
3S p
t cp =
U
where:
tcp = average reflected pressure clearing time at a point (ms)
Sp = shortest distance from the point of interest to a free edge (m)
U = shock front velocity (m/ms) (see Figure 2-9)
3-7
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Figure 3-9 Sound Velocity in Reflected Overpressure Region
Blast
Wave
Direction
In almost all cases, excluding the negative phase blast load has either a conservative
effect on calculated maximum component response (i.e., increases response) or has
little effect. The negative phase blast load has maximum effect on the component
response when the whole negative phase duration occurs prior to maximum component
response. Conversely, it has no effect when the duration of the positive phase blast
load exceeds the component maximum response time. The latter case is most often
true when the blast load duration is longer than one or two times the natural period of
the component, while the former case is most often true with the blast load duration is
shorter than approximately one-third of the component natural period. In most practical
cases where non-confined high explosive loading is applied to building components
3-8
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
(i.e., the explosion is outside the building), the blast load duration is less than the
natural period of the component so that inclusion of negative phase blast load does
affect the calculated maximum component response, reducing it in the large majority of
cases. In a limited number of cases, the negative phase of the blast load can be in
phase with rebound of the structural component from positive phase blast load and a
resonance effect occurs, where the maximum component response occurs in rebound
or during the second inbound response phase.
An important simplification in SBEDS is that it uses the fully reflected negative phase
blast load from Figure 2-11 for all incidence angles less than, or equal to 45 degrees,
and the side-on negative phase blast load for all incidence angles greater than 45
degrees. This approach is exact compared to the available prediction method only when
the input angle of incidence is zero degrees (fully reflected) or 90 degrees (side-on) and
becomes less accurate as the angle of incidence approaches 45 degrees. It tends to
over predict the negative phase blast load for angles of incidence between 0 and 45
degrees and under predict the negative phase blast load for angles of incidence
between 45 and 90 degrees. Also, there never any clearing calculated for the negative
phase blast load in SBEDS, which may over predict the negative phase impulse when
for components on one or two story buildings. These simplified approaches are used
because so little is known about the negative phase blast load, including the effect of
angle of incidence and clearing. Generally speaking, the maximum component
response to blast load can be under predicted if the negative phase blast load is over
predicted. Therefore, the user may want to use positive phase blast load only, or at
least analyze this case also, if the angle of incidence is between 30 and 45 degrees.
3-9
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
when there is significant blast load variation over the component of interest. However,
component response is usually calculated with simplified methods based on a spatially
uniform “equivalent” blast load over the whole component at each time step. Ideally, the
equivalent uniform load will cause the same dynamic response as the actual non-
uniform load.
When the blast load variation on a reflecting surface is such that the midspan blast load
cannot be used as an average over the whole span, the variation in impulse over the
component span can be defined in terms of step function with small regions of constant
impulse or with a curve-fit mathematical function. Dynamic response is almost always
much more sensitive to the impulse of the blast load than any other parameter for the
relatively close-in standoff cases causing blast loads that vary significantly over the
span on a reflecting surface. The calculated distribution of the blast load impulse over
the component span can be normalized relative to the peak impulse to have the form
imaxΙ(x,y), where Ι(x,y) is the impulse at each point along the area of interest divided by
the peak impulse, imax. The equivalent uniform impulse ie can then be calculated from
Equation 3-5, where φ(x,y) is the component deflected shape function based on an
assumed modal response shape as explained in Section 4-1.1.1.
Equation 3-5
LH
imax ∫ ∫ Ι ( x, y )ϕ ( x, y )dxdy
i ( x, y )
ie = 0 0
LH
where Ι ( x, y ) =
imax
∫ ∫ ϕ ( x, y)dxdy
0 0
where:
ie = equivalent impulse of uniformly distributed load
L= length of loaded area
H= width of loaded area
φ(x,y) = deflected shape function of blast-loaded component
i(x,y) = Non-uniform impulse applied to loaded area
imax = maximum impulse over loaded area
Ι(x,y) = normalized impulse shape function over loaded area
If significant plastic response is expected, φ(x,y) can usually be based on the shape of
the component after it has yielded under uniform static pressure (i.e., the yield line
3-10
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
pattern). Otherwise, an elastic deflected shape can be used. For one-way components,
such as beams, all parameters are only a function of one direction.
The peak pressure of the equivalent uniform blast load, which will not affect the
calculated response very much, can be based on the peak pressure at midspan. The
equivalent uniform blast load therefore has this equivalent uniform peak pressure and
an impulse from Equation 3-5 with the simplified shape of a right triangle. This method
is discussed in Chapter 9 of UFC 3-340-01, along with several other methods for
determining an equivalent uniform blast load. UFC 3-340-01 recommends using a 1.25
safety factor if a highly non-uniform blast load is simplified as an equivalent uniform
impulse in a blast analysis.
The ConWep computer program distributed by the U.S government calculates the
average impulse over a given loaded area from a reflected free-air or surface burst
explosion with Equation 3-5 taking into account clearing and the different angles of
incidence of the blast load at each point on the surface (ConWep, Version 2.1.0.3).
This capability is included in the “Slab Loading” option in the program. It calculates the
average peak pressure over the loaded area in a similar manner. ConWep also
calculates all the properties of the blast wave from free-field and surface blast
explosions using the applicable figures in Chapter 2.
Simplified methods are also available for determining the equivalent uniform blast load
on components along non-reflecting surfaces with non-uniform blast load. This occurs if
the arrival time of the shock wave varies significantly along components that span in the
direction of the shock wave propagation, such as roof component or sidewall
component on a building that is parallel to the direction of shock wave propagation. In
this case, the loading on the surface is non-uniform at the points along the component
span at any given time, as illustrated in Figure 3-11, since the load at any time along the
span depends on how far the shock wave has traveled along the component length.
However, there is no variation in the incidence angle, which is always 90 degrees.
An equivalent uniform blast load can be calculated for this case, as shown in Figure
3-11, in terms of the blast wave parameters at Point b, the back-end of the component.
It is assumed in Figure 3-11 that the component spans the full dimension of the building,
but points “f” and “b” can be the front and back, respectively, of a shorter component
that spans parallel to the direction of shock wave propagation. The distance L will be the
projection of the span length along the direction of shock wave propagation if the
component spans at an angle relative to the direction of shock wave propagation, but
the procedure will otherwise be the unaffected.
3-11
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Figure 3-11 Equivalent Uniform Roof and Side Wall Loading
Equation 3-6
Por = C E Psob + C d qob
where:
Por = peak pressure of equivalent uniform blast load
Psob = peak incident pressure at Point b from Figure 2-10
CE = equivalent load factor = f (Lwb/L) (see Figure 3-12)
Lwb = wavelength of shock wave at Point b from Figure 2-10
L = length of component span parallel to direction of shock wave
propagation
qob = peak dynamic pressure at Point b calculated using Figure 2-13
Cd = side-on element dynamic drag coefficient (see Table 3-1)
tf = shock wave arrival time at point f, equal to Rf/U
tb = shock wave arrival time at point b, equal to Rb/U
Ri = standoff distance from explosive source to point i (representing either
point f or b)
td = time of shock wave propagation to peak stress point on span = D/U
D = distance to point on component with peak stress (see Figure 3-12)
U = shock wave velocity based on standoff to point of interest
(see Figure 2-9)
tofb = duration of the positive phase blast load at point b using Equation 3-2
3-12
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Figure 3-12 Equivalent Load Factor and Blast Wave Location Ratio
The peak pressure of the equivalent uniform blast load, Por, in Figure 3-11 is calculated
using Equation 3-6. The equivalent load factor, CE, and the ratio of the point of
maximum stress, d, to span length L (D/L), are given as functions of the wavelength-
span ratio, Lwb/L in Figure 3-12 (see definitions of Lwb and L in Equation 3-6). The
equivalent uniform blast load on the component is assumed to increase linearly from
zero at the time, tf, when the blast wave reaches the beginning of the component (Point
f) to the peak pressure, Por, at time, tf + td, when the blast wave reaches Point d. The
equivalent uniform pressure then decays linearly to zero at time, tb + tofb, where tofb is the
equivalent triangular pulse duration of the positive phase side-on blast load calculated
as shown in Equation 3-2 at Point b. Cd for components spanning parallel to the
direction of shock wave propagation is a function of the peak dynamic pressure as
provided in Table 3-1. A similar method can be used to determine the equivalent
uniform blast load along a component on the sidewall of a building that spans parallel to
the direction of blast wave propagation.
It is usually a simplified, conservative value for the equivalent uniform blast load on a
component spanning parallel to the direction of shock wave propagation is the side-on
blast load at midspan on the component, since it will have an immediate rise time to a
peak pressure that is greater than or equal to Por in Figure 3-11. The blast loads at any
specific points on the roof or sidewalls of buildings can be based on the side-on peak
pressure and impulse from Figure 2-9 with a modified standoff distance, Rm. Rm is the
3-13
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
“line of site” standoff distance, equal to the slant distance from the explosive source to
the nearest edge of the building side facing the explosive source to the point of interest
plus the distance from this edge to the point of interest. Blast loads calculated with this
approach have compared well to measured blast loads on the sides and roof of five
story building in repeated explosive tests (Coltharp and Mosher, 1999).
As the shock front passes over the rear of the structure, the pressure front expands
over the roof and around the sidewalls, and loads the rear or back wall. The applied
pressures from these waves are reinforced by reflections of the shock wave up off the
ground. However, vortices in the shock wave that form at the free edges of the back
wall may reduce the strength of the shock waves that propagate onto the back wall.
Limited experimental information is available on the blast load on the rear wall.
Conservatively, the equivalent uniform blast load for the back wall can be assumed
equal to the side-on blast load at Point b in Figure 3-13. The equivalent uniform blast
load may also be determined as shown in Figure 3-13. The peak value of the pressure,
Por, is the sum of the contributions of the peak side-on pressure and drag pressures
shown in Equation 3-7, where Point e occurs at a distance of Hs from the rear edge of
the roof slab. The distance Hs is the lesser of the back wall height or one-half the
building width (see H and W in Figure 3-10).
Equation 3-7
Por = C E Pe + Cd qoe
where:
Por = peak pressure
Pe = peak incident pressure at Point e
3-14
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
CE = equivalent load factor = f (Lwb/Hs) (see Figure 3-12 with L=Hs)
Lwb = wavelength of shock wave at Point b from Figure 2-9
Hs = height of back wall
qoe = peak dynamic pressure at Point e calculated using Figure 2-13
Cd = dynamic drag coefficient = f (qoe) (see Table 3-1)
td = time of shock wave propagation to peak stress point on span = D/U
D = distance to point on component with peak stress (see Figure 3-12 with
L=Hs) see Equation 3-6 for definitions of analogous parameters
Figure 3-14 shows that the peak pressure in the doubled dynamic reaction pressure (3
psi) is significantly less than the peak applied blast pressure (6 psi), since the dynamic
reaction pressure is affected by the relative low flexural strength of the component.
However, the impulse of the first positive pulse of the doubled dynamic reaction
pressure is slightly larger than the impulse of the blast load. The dynamic reaction
pressure on a primary component typically “smears out” the applied blast pressure, so
that it has a lower peak pressure than the applied blast pressure, a longer duration, and
approximately the same impulse. This is not always true, but it is very typical for
secondary components subject to blast loads large enough to cause permanent
deformation. Also, the dynamic pressure history includes multiple positive and negative
3-15
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
pressure oscillations based on the dynamic response of the supported secondary
component. Comparisons to blast tests show that dynamic response calculations are
not typically as accurate for the second, third, etc. oscillations of the response of blast-
loaded components as for initial response and therefore the calculated dynamic reaction
history is much less reliable after initial response. In particular, observed damping of the
response is usually not calculated very well.
As noted above, the dynamic reaction pressure must typically be multiplied by a factor
of two to compare to the blast load over the tributary area of a supporting component.
This is true because of a characteristic in the way the dynamic reaction pressure
histories are traditionally calculated in blast design. As shown in Equation 4-24, the
dynamic reaction pressure applied by a secondary component must be multiplied by the
full loaded area to get the reaction force, whereas the tributary area of the secondary
component that loads a primary component is only one-half of the full secondary
component area. This inconsistency requires that SBEDS multiplies the calculated
reaction pressure histories of a component (VR(t) in Equation 4-24) by a factor of 2
when it writes reaction pressure histories to save files that can be read as an applied
load file for a primary component. This is because the dynamic reaction history that is
read in will only be applied over the tributary area of the primary component. See
Equation 3-8 and Figure 3-15 for additional explanation. An exception is the nontypical
case where the secondary component cantilevers off the primary component. In this
case, SBEDS does not multiply the saved reaction pressure history by a factor of 2.
V (t ) = VR (t ) LG = 2VR (t ) B
Equation 3-8
where: P = total blast load on girder from dynamic reaction loads
V(t) = dynamic reaction load along primary component (girder)
VR(t) = dynamic reaction pressure from component spanning Lc
(see Equation 4-24)
3-16
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
B = width of tributary area of component supported by girder
LG = girder span
See Figure 3-15 for more information on parameters
Note: SBEDS saves 2VR(t) as the dynamic reaction pressure that can be
read in as a applied blast load pressure on tributary area of
supporting component (i.e., girder) unless input component is
cantilever and then it saves VR(t)
LG
B Lc
V(t)
Girders
3-5.1 LOADED AREA FACTOR, AF, FOR COMPONENTS WITH
CONCENTRATED LOADS
The SBEDS workbook analyzes components with concentrated blast loading at
midspan or at one-third points along the span. The concentrated blast loads at these
points are based on an input blast pressure history applied over the tributary area that
the secondary framing components loads the primary component. The input blast
pressure history can be the applied blast pressure history or to the dynamic reaction
history input as a pressure, as shown in Figure 3-14, increased by a factor of two as
applicable.
Roof and wall framing plans typically include some secondary components that transfer
blast load directly into columns, rather than into the primary framing component of
interest. This is illustrated in Figure 3-16. In this case, the blast load or representative
dynamic reaction pressure does not represent load along the entire span. This effect is
accounted for in SBEDS with a “Loaded Area Factor”, AF, input by the user. As shown
in Figure 3-16, AF is the ratio of the actual blast tributary area of secondary components
that frame into primary component of interest divided by the maximum tributary area,
which equals the primary component span multiplied by the primary component
spacing. In Figure 3-16, only one beam frames into the primary component to cause a
concentrated midspan load condition and this beam only transfers blast load into the
girder from one-half the default tributary area of the girder. The columns are indicated
by small triangles. If the blast load is uniformly applied along the entire length of the
primary component, AF =1. This is also applicable for cases where multiple closely
spaced secondary components frame into the primary component of interest. If more
3-17
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
than three secondary components are supported by a primary component, a simplified
conservative approach is to assume AF=1.
Figure 3-16 Example of Roof Girder with Loaded Area Factor (AF) Equal to 0.5
Beam
L Girder AF =
blast _ loaded _ area
BL
Blast Loaded Area
on Girder
.
Note that beams only transfer blast load from 50% of the area B*L into girder in this example where the beams
have a pinned connection to girders. The blast load from the rest of this area is transferred by beams directly
into columns supporting girder
All resistance and stiffness values in the equivalent SDOF system for a beam
component in SBEDS are divided by the loaded area factor. However, the loaded area
factor is not applied to the mass. It does apply to the component self weight mass and
the supported weight is input per unit loaded area. The mass factor and load factors
could both be increased when AF<1, but this would have partially offsetting effects on
the load-mass factor as described in the first sections of the next chapter so that it is
conservative to ignore it and it is considered acceptable to ignore it compared to other
approximations in the overall SDOF analysis approach.
3-18
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
4-1
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Equation 4-1
M e u ′′(t ) + C e u ′(t ) + Re (u (t )) = Fe (t )
where:
Me = effective mass of equivalent SDOF system
u′′(t) = acceleration of the mass
u′(t) = velocity of the mass
Ce = effective viscous damping constant of equivalent SDOF system
Re(u)= effective resistance of equivalent SDOF system
u(t) = displacement of the mass
Fe(t) = effective load history on the equivalent SDOF system
t = time
The equivalent mass, stiffness, and load of the SDOF system in Equation 4-1 are
obtained by multiplying the mass, stiffness, and load of the blast-loaded
component by transformation factors as shown in
Equation 4-3. As stated in
Equation 4-3, KR and Kd are assumed equal to KL, as discussed in Section 4-1.1.1. This
implies that Equation 4-1 can be written as Equation 4-4. This can be simplified further
and written as Equation 4-5.
4-2
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Equation 4-3
M e = KmM c
Fe (t ) = K L Fc (t )
Re (u (t )) = K R Rc (u (t ))
Ce = K d Cc
where:
Mc = mass of blast-loaded component
Fc(t) = load history on blast-loaded component
Rc(u(t))= resistance of blast-loaded component
Cc = viscous damping constant of blast-loaded component
Km = mass transformation factor
KL = load transformation factor
KR = resistance transformation factor, however KR=KL
Kd = viscous damping transformation factor, however, Kd=KL
Equation 4-4
K m M cu′′(t ) + K LCcu′(t ) + K L Rc (u (t )) = K L Fc (t )
Equation 4-5 is the basic equation of motion for an equivalent SDOF system. It is
solved to determine the response of the equivalent SDOF system, which is
directly related to the response of the blast-loaded component since it has a
deflection equal to the maximum deflection of the blast-loaded component.
Therefore, assuming that all the terms for the blast-loaded component in
Equation 4-3 can be defined as described in this manual, all the terms in Equation 4-5
can be defined if the transformation factors KL and Km are known. For some simplified
load vs. time relationships, a closed-form solution of Equation 4-5 is possible. However,
a time-stepping solution using approximate numerical methods is generally required and
numerous available computer programs have been developed for this purpose.
Equation 4-5
K Lm M c u ′′(t ) + C c u ′(t ) + Rc u (t ) = Fc (t )
where:
KLm = load-mass factor, equal to Km/KL
4-3
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
single mode shape) and constant spatial blast load distribution shape p(x) at all time
when KL applies. This allows the deflections and load along the component span to be
written in terms two uncoupled terms that are functions of time and span, respectively.
The dimensionless shape functions φ(x) and p(x) are usually normalized to have a peak
value of 1.0. The assumption of a constant deflected shape is analogous to assuming
response in a single, dominant mode shape. For the typical case of a uniformly
distributed blast load, p(x)=1.0 and Pmax(t) equals the uniform load magnitude at each
time t in Equation 4-6. For a concentrated blast load, p(x) only has non-zero values at
the loaded points. A spatially non-uniform blast load will have a more complex p(x)
function, but p(x) is still assumed constant with time when KL applies. Finally, it is also
assumed in Equation 4-6 that the SDOF system deflection and the maximum
component deflection are equal. This assumption is crucial for applying the SDOF
system analysis results to the blast-loaded component. A similar procedure to that
shown in Equation 4-6 is used to calculate KL of a two-way spanning component, except
the shape functions φ(x) and p(x) are functions of x and y and all integrals are double
integrals over the span lengths in each direction.
Equation 4-6
L
WE (t ) component = ∫ P( x, t )U ( x, t )dx let U ( x, t ) = U max (t )ϕ ( x) and P( x, t ) = p( x) Pmax (t )
0
L
Thus : WE (t )component = Pmax (t )U max (t ) ∫ p ( x )ϕ ( x)dx
U ( x, t ) P ( x, t )
where ϕ ( x) = p( x) =
0
U max (t ) Pmax (t )
WE (t ) SDOF = Fe (t )u (t )
L ∫ p(x )φ (x )dx L
Fe (t ) = Pmax (t ) ∫ p ( x )ϕ ( x)dx = K L Fc (t ) where K L = 0
L
Fc (t ) = Pmax (t ) ∫ p ( x )dx
0
∫ p(x )dx
0
0
where:
WE(t) = work energy on SDOF spring and blast loaded component at each time
U(x,t) = displacement of blast-loaded component
P(x,t) = dynamic load on blast-loaded component
Umax(t) = maximum displacement of blast-loaded component at constant x location
Pmax (t) = maximum dynamic load on blast-loaded component at constant x location
Fe(t) = dynamic load applied to SDOF system at each time
Fc(t) = total dynamic load on blast-loaded component at each time
u(t) = deflection of SDOF system at each time
L= length of blast-loaded component
φ(x) = deflected shape function of blast-loaded component
p(x) = load shape function of blast-loaded component
4-4
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
KL= load factor
Figure 4-2 shows an example of assumed deflected shape functions, φ(x), for a one-
way component (i.e., a beam) with a uniformly distributed load and simple supports
during elastic response and during plastic response, which occurs after the component
has yielded at all maximum moment regions and becomes a mechanism. Both
response realms are typically modeled during an analysis of a blast-loaded component
with the equivalent SDOF system. The deflected shape of the component changes
during plastic response because the maximum moment region(s) become hinges.
Therefore, KL has different values for elastic response, when u(t) in Equation 4-5 is less
than the deflection at which the component becomes a mechanism (i.e., the yield
deflection), and for plastic response, when u(t) is greater than this deflection.
prior to yield at maximum moment region after yield in maximum moment region
The resistance factor, KR, can be derived by setting the internal strain energy due to
stress and strain in the spring of the equivalent SDOF system equal to the internal strain
energy due to the stresses and strains in the blast-loaded component. The strain
energy of both the blast-loaded component and equivalent SDOF system can also be
defined in terms of an internal work energy equal to a resisting force multiplied by the
maximum deflection at each time, where the resisting force has the same spatial
distribution as the applied load. The component is also assumed to have the same
deflected shape used to define the load factor, KL. Therefore, all the shape functions in
4-5
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Equation 4-6 are also applicable to the internal work energy of the blast-loaded
component. Equation 4-6 also shows the equivalency of strain energies, or internal
work, for the equivalent SDOF system and blast loaded component except that all the
load terms are replaced with similar internal resisting force terms and the final equation
is Re = KLRc, where KL is defined as shown in Equation 4-6. Therefore the
transformation factors KR and KL are equal. The concept of the internal resisting force is
discussed in Section 4-1.2.
Also, as stated in
Equation 4-3, the transformation factor for damping, Kd, is assumed equal to KL. This is
not necessarily mathematically correct. It is a matter of convenience so that only a
single KLm factor is needed in Equation 4-5, which is justified by the fact that the
damping term only contributes a small amount of energy to the response of the
equivalent SDOF system during the response of typical blast-loaded components.
Damping is discussed further in Section 4-1.3.
Equation 4-7
L • 2 • •
1
2 ∫0
KE (t ) component = m ( x ) U ( x, t )dx let U ( x, t ) = U max (t )ϕ ( x) and thus U ( x, t ) = U max (t )ϕ ( x)
•
L •
1 U ( x, t ) U ( x, t )
= ∫ m( x) U max (t )ϕ 2 ( x)dx where ϕ ( x) = =
2
KE (t ) component
20 U max (t ) U• (t )
max
1
KE (t ) SDOF =
2
M e KE (t ) SDOF (t )
2
∫ m(x )φ (x )dx
2
L • L
M e = ∫ m( x) U max (t )ϕ 2 ( x)dx = K m M c where K m = M c = ∫ m( x )dx
2 0
L
and
∫ m(x )dx
0 0
0
where:
KE(t) = kinetic energy on SDOF spring and blast loaded component at each time
U(x,t)* = displacement of blast-loaded component
m(x) = mass per unit length of blast-loaded component
Umax(t)* = maximum displacement of blast-loaded component at constant x location
Me = equivalent mass of SDOF system
Mc = total mass of blast-loaded component
u(t)* = deflection of SDOF system at each time
L = length of blast-loaded component
φ(x) = deflected shape function of blast-loaded component
Km= mass factor
* A dot over displacement indicates velocity
The resisting force always has the same spatial distribution along the component span
as the applied load (e.g., uniformly distributed). It is a function of the maximum
component deflection, as mentioned previously in this paragraph, and the plot of
component resistance (or resisting force) vs. maximum deflection is called the
resistance-deflection relationship. This relationship allows the resisting force, Rc, in
Equation 4-5 to be defined at each time based on the deflection of the equivalent SDOF
system. Note that the deflection of the equivalent SDOF system equals the maximum
deflection of the blast-loaded component at each time step. The slope at each point of
the resistance-deflection relationship of a component represents the component
stiffness. The stiffness is the ratio of a delta static load increase to the corresponding
delta deflection of the component.
4-8
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
4-9
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
In the case shown in Figure 4-4 the beam has indeterminate boundary conditions and
therefore has two stiffness regions as described in the previous paragraph. However,
the corresponding resistance-deflection relationship for the equivalent SDOF spring is
simplified to have a single “equivalent” elastic region, which has the same area under its
resistance-deflection curve as the elastic and elastoplastic regions of the beam’s
resistance-deflection curve.
Figure 4-5. Similar formulas can be developed for more complicated resistance-
deflection relationships with yielding at multiple maximum moment regions that have
more than two non-zero slopes. SBEDS uses the formula in
Equation 4-8 to determine the equivalent elastic stiffness for two-way spanning
components.
4-11
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Equation 4-8
4-12
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
⎛ r ⎞
xE = xe + x p ⎜⎜1 − e ⎟⎟
⎝ ru ⎠
r
kE = u
xE
2. Determine the location along the beam with the highest ratio of applied moment to
dynamic moment capacity for a load with the same spatial distribution as the blast
load. In Figure 4-4, where the blast load is uniformly distributed, this point is at both
supports if the beam has a constant moment capacity along its length. Then, solve
for the load causing an applied moment at this location equal to the dynamic
moment capacity. This load is the initial elastic resistance, re, as shown in
Equation 4-9. If the component was determinate, it would be the ultimate resistance, ru.
The maximum deflection caused by this load is the initial yield deflection, xe, in
Equation 4-9
− wL2
M max =
12
− −
Let M du = M max and solve for w = Re
−
12M du 384re L4
re = and xe =
L2 EI
where:
Mmax-= moment at fixed supports from uniform load
w = uniform load per unit length on beam
L = span length
Mdu- = dynamic moment capacity of beam at supports
re = elastic resistance (uniform load per unit length causing yielding at
supports). This can be converted to a pressure by dividing by the
beam spacing.
xe = initial yield deflection of beam
E = Young’s modulus for beam
4-13
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
I = moment of inertia for beam
4. If the component is indeterminate, place a hinge at the location(s) of the initial yield
with an applied resisting moment equal to the dynamic moment capacity. Repeat
step 2 for the modified component with the hinge and applied resisting moment. If
the component becomes a mechanism after repeating step 2, this load and
corresponding maximum deflection are the ultimate resistance ru, and yield
deflection xp, respectively. This is shown in Equation 4-10. Yield line theory can also
be used. The stiffness after first yield is the elastoplastic stiffness, kep, equal to
(ru-re)/(xp-xe).
5. If the component is not a mechanism after step 3, repeat step 2 by adding additional
hinge(s) where yielding occurred during step 2 and applied resisting moments equal
to the dynamic moment capacity until it is a mechanism. In this case, the resistance-
deflection relationship will have more than two non-zero slopes. The final load
causing the component to become a mechanism is the ultimate resistance. The
ultimate resistance can also be determined using yield line theory as described in
this section.
7. It should be verified that a load on the component equal to the ultimate resistance
does not cause component failure in any response mode other than flexural
response. Therefore, the loads with the same spatial distribution as the blast load
causing shear failure, buckling failure, connection failure, and failure in any other
applicable response mode of the component should be greater than the ultimate
resistance. For blast resistant design, the component should be sized, if possible, so
that the ultimate resistance from flexural response of the component is less than the
loads causing failure in other response modes.
8. If the ultimate load capacity from any other response mode is less than the ultimate
resistance determined from flexural response as described in step 1-3, then the
ultimate resistance is set equal to the lower ultimate load capacity from the
controlling response mode. The resistance-deflection relationship based on flexure
is usually applicable up the new, lower ultimate resistance. The resistance is then
assumed to be constant at the ultimate resistance for greater deflections up to a
limited failure deflection. However, the failure deflection for a response mode other
than flexure may only be slightly greater than the deflection causing the ultimate
resistance.
Equation 4-10
4-14
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
+ wL2 −
M max = − M du
8
+ +
Let M max = M du and solve for w = ru
ru =
( +
8 M du + M du
−
) and x p = xe +
384(ru − re )L4
L2 5EI
where:
Mmax+= maximum moment in beam at midspan
Mu+ = dynamic moment capacity of beam at midspan
Mu- = dynamic moment capacity of beam at supports
ru = ultimate resistance (uniform load per unit length causing yielding at
midspan). This can be converted to a pressure by dividing by the
beam spacing.
xe = initial yield deflection of beam
xp = final yield deflection of beam
The approach used in SBEDS to calculate the terms that define the resistance-
deflection relationship of two-way components is more complicated. First, available
design procedures for defining these terms will be discussed, and then the approach in
SBEDS, which is based on these procedures, will be discussed.
Table 4-4 shows formulas for the ultimate resistances of two-way components based on
yield line theory, where interior yield lines were extended straight into the corners and
the moment capacities near the corners were conservatively limited to two-thirds of the
dynamic moment capacity. This accounts for the fact that rotation in the corner is limited
along the yield lines so that the full resisting moment capacity is not developed. The
yield line dimensions (i.e., x and y) are determined based on the boundary conditions
and moment capacities at yield locations as shown in Figure 4-6 through
Figure 4-8. Simple supports are accounted for with a zero negative moment capacity.
See other notes at the bottom of
Table 4-4.
Chapter 3 in TM 5-1300 and Appendix D.10.2 in UFC 3-340-01 have examples showing
the usage of these charts. Equivalent support shears, which equal the support reaction
forces, are discussed in Section 4-3. Chapter 3 in TM 5-1300 and Appendix D.10.2 in
UFC 3-340-01 have identical procedures that can be used to define the elastic and
4-15
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
elastoplastic stiffnesses and resistances of the full resistance-deflection relationship for
a variety of two-way slab components in a step-by-step method with simple equations
and look-up values from graphs. The resulting resistance-deflection relationship may
involve up to five stiffnesses.
Table 4-3 Flexural Resistances, Stiffnesses, and Support Shears for One-Way
Members
4-16
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Table 4-4 Ultimate Flexural Resistances and Support Shears for Uniformly-
Loaded Two- Way Members with Uniform Moment Capacities
4-17
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Figure 4-6 Location of Symmetrical Yield Lines for Two-Way Component with
Four Edges Supported
4-18
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Figure 4-7 Location of Symmetrical Yield Lines for Two-Way Component with
Three Edges Supported
Figure 4-8 Location of Yield Lines for Two-Way Component with Two Adjacent
Edges Supported
4-19
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
The ultimate resistances for two-way components in Table 4-5 are determined with yield
line theory of slabs using 0.785-rad (45-deg) yield line locations from all corners as
shown in the table and adding a 10 percent reduction to correct for corner effects for
cases with fixed supports (Morison, 2005). Although the resulting equations are not
exact, the error in ultimate resistance is less than 5 percent for plates with uniform
moment capacities in both directions. The previously described yield line method from
Chapter 3 in TM 5-1300 and Appendix D.10.2 in UFC 3-340-01 is generally more
accurate because it does not assume equal moment capacities in both directions.
However, Table 4-5 is a simpler procedure and its usage is often justified by the overall
approximate nature of the SDOF analysis procedure, as discussed in Section 4-8.
Table 4-5 Approximate Flexural Resistances, Stiffnesses, and Support Shears for
Two-Way Members
4-20
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
4-21
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Notes:
1
Equations for ru,Vx, and Vy in this table are based on approximate (0.785 rad [45 deg]) positive yield line positions.
Equations for re, ke, and kep in this table are based on elastic uniform plate theory, except that for fixed edge cases
re has been determined when most of the support length has yielded. Values of kep are given for the condition of
all supports fully yielded, effectively omitting one intermediate elastoplastic step for Cases 2 and 5.
2
Ultimate resistance, ru, values for members with fixed supports (Cases 2 through 5) have been reduced 10 percent to
account for a more detailed yield line configuration (fan pattern) near corners.
3
Stiffness coefficients listed are for a Poisson's ratio of 0.15 (concrete). For a Poisson's ratio of 0.30 (steel), stiffness
values are approximately 7 percent greater.
4
Definitions and Units.
Lx and Ly = x and y-span lengths (mm or inch)
Mnx and Mny = Ultimate negative (support) moment capacity per unit width in x- and y-span directions
(N-mm/mm or b-in/in)
Mpx and Mpy = Ultimate positive (interior) moment capacity per unit width in x- and y-span directions
(N-mm/mm or b-in/in)
E =Modulus of elasticity (MPa or psi)
I =Moment of inertia per unit width (mm4/mm of in4/in) (Ia for slabs, see 10.3.6)
r =Elastic unit resistance (MPa or psi)
ru = Ultimate unit resistance (MPa or psi)
ke = Elastic unit stiffness (MPa/mm or psi/in)
kep = Elastoplastic unit stiffness (MPa/mm or psi/in)
Vx and Vy = Equivalent static support shears in x- and y-span directions (N/mm or lb/in)
Table 4-4, for determining the ultimate resistance of slabs with openings, including
windows and doors. Elastic and elastoplastic resistances and stiffnesses slabs with
openings are typically assumed equal to those without openings. A static finite element
approach with material non-linearity can also be used to develop the full resistance-
deflection relationship for two-way components responding in flexure. This approach
can be particularly useful for cases with unusual geometry and loading conditions.
Table 4-4 with an 8% increase factor and curve-fits to Figure 4-6 through
Table 4-4 is based on the assumption that only two-thirds of the moment capacity acts
at corners, as discussed previously. The 8% increase factor is included because studies
by Morison (2005) show that this approach overcorrects for straight extension of yield
lines into corners by 6% to 10% compared to advanced analyses with yield lines that
transition to multiple fan-shaped and triangular shaped yield lines in the corners.
EI
k=
γ 1 (1 − υ 2 )L y 4
k = flexural stiffness
E= Young’s modulus
I = moment of inertia
ν = Poisson’s ratio
Ly = short span length
As discussed in Section 4-1.2.4, the strain energy for ductile flexural response, and
therefore the maximum response, is usually dependent primarily on plastic response.
Therefore, the ultimate resistance usually has a much larger affect on the maximum
dynamic response than the elastic and elastoplastic resistances and stiffnesses and a
simplified approach for determining these parameters is justified.
4-23
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
relationships, depending on the component type and boundary conditions. These
response modes can act together with flexural response, as discussed in the following
paragraphs.
Figure 4-11 shows the measured resistance-deflection relationship for a cold formed
steel beam with boundary conditions fully restrained against in-plane movement. In a
series of many such tests, it was only possible to bound the measured the resistance-
deflection relationships as indicated in Figure 4-11 (Salim et al, 2003). In an actual
building, there is typically more flexibility in the supports and local yielding around the
connections than for these tests. There is no well validated approach for including
tension membrane response in a resistance-deflection relationship due to a number of
variables affecting tension membrane that are difficult to fully understand and control,
even under test conditions.
Figure 4-10 Illustration of Tension Membrane Response
Resisting Moment = Hl x f
4-24
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Figure 4-12 shows how tension membrane response is simplistically included in the
resistance-deflection relationships for steel components in SBEDS. It is intended to be
used primarily with light, cold-formed steel components, as discussed more in Section
6-6. A constant tension membrane stiffness based on the maximum tension that can
develop in the component is assumed to add to the flexural response at deflection
greater than xTM, which is calculated as shown in Equation 4-11. The maximum tension
of lightweight steel components is usually limited by the shear capacity of the
connections rather than the cross sectional tension capacity of the component. xTM in
Equation 4-11 is the sum of the deflection at which the component becomes a
mechanism in flexure plus the theoretical deflection required to develop the maximum
tension in the component in pure, elastic catenary action with no support or connection
flexibility.
Figure 4-11 Resistance-Deflection Relationship for Steel Stud with Restrained
Boundaries (from Salim et al, 2003)
12
10 Component Experiment
Pressure per unit width of wall (psi/in)
Model (LE)
Model (HE)
0
0 2 4 6 8 10 12 14 16 18
Deflection (in)
The approximate approach in Figure 4-12 is an attempt to account for support and
connection flexibility in structures with typical lightweight steel components, such as
panels and cold-formed beams, and be generally consistent with static and blast test
data for components with combined flexure and tension membrane response. The
support and connection flexibility are assumed to delay significant tension buildup in the
component until after it fully yields, and then the supports are assumed to act rigidly
against further in-plane movement. The maximum deflections based on SDOF
calculations with the resistance-deflection relationship in Figure 4-12 compared
reasonably well to test data on cold-formed steel girts from one-half scale shock tube
tests, as indicated in Figure 4-13 (Oswald et al, 1998).
1.4
1.2
KTM
ru 1
Resistance
0.8
0.6 ke or KE
0
0
xe or 1xE x2TM 3 4 5 6
Deflection
Equation 4-11
4-26
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
xTM = x E +
4TL2
π 2 EA
where T = Minimum f d y A , Vc [( ) ]
8T
K TM _ 1 =
bL2
Tπ 3 1
K TM _ 2 = where A = 1− and Lx ≥ L y
⎡1 ( n −1) / 2 ⎤ nπL x
4bL y ∑ ⎢ 3 (− 1)
2
A⎥ cosh
n =1, 3, 5, 7 ⎣ n ⎦ 2Ly
0
4 4A 5 6 7 10 11 12 13 14 15 17 18 19
T e s t Nu m b e r
4-27
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Figure 4-15, a two-way component with an aspect ratio between 0.5 and 2 can be
assumed to develop a compression ring, as shown in Figure 4-16, so that tension
membrane resistance occurs even if the supports provide no in-plane restraint. As the
central portion of the slab reaches large deflections, compression struts that form along
the component edges parallel to the span direction keep the edges of the component
from deflecting inward near midspan. This occurs in both directions. Reinforced
concrete components must also have continuous steel through the span that is
unspliced or has splices that can develop the full tension strength at large deflections.
4-28
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Equation 4-12
rT 8T
= K Tm = 2t where Tt = 0.5 f u Atot
ΔT L
where: KTm = tension membrane stiffness
Tt = ultimate tensile force per unit width assuming 50% of total
reinforcement along span effectively resists tension at large
displacement and is fully strain hardened
Atot = total reinforcement per unit width continuous over span from both
faces (based on lesser of input negative and positive moment area
steel at each face in SBEDS)
L = clear span dimension
fu = ultimate strength of reinforcing steel
Equation 4-13
rT Tty π 3
= K Tm =
ΔT ∞ ⎡1 n −1
⎤
4 L2y ∑ ⎢ 3 ( −1) 2
( A)⎥
n =1, 3, 5... ⎣ n ⎦
1
A = 1−
⎡ nπL Tty ⎤
cosh ⎢ x
⎥
⎢⎣ 2 L y Ttx ⎥⎦
Tty = 0.5 f u Ay _ tot Ttx = 0.5 f u Ax _ tot
The compressive forces associated with the crushing at the loaded face near midspan
and at the unloaded faces at the supports, as shown in Figure 4-17 where the material
crushing strength is represented as f’m, form a resisting couple. This couple, or resisting
moment, resists lateral load in addition to any ductile flexural resisting moments. The
4-29
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
compression membrane forces are based on the material crushing strength acting over
an area that increases as the component deflects. The moment arm between the
compression forces at midspan and at the supports is a maximum value at small
deflections and decreases to zero at a deflection near the component thickness.
Figure 4-14 shows the assumed resistance deflection response for compression
membrane response in a component with ductile flexural response. The resistance
increases linearly above the flexural resistance due to compression membrane
response to a peak resistance, and then reduces to the flexural resistance at a
deflection that is a little less than the component thickness when the resisting moment
caused by compression membrane response becomes equal to zero. Therefore,
compression membrane response adds a large triangular area above the resistance-
deflection relationship of a component with ductile flexural response. The resistance-
deflection relationship of a component with brittle flexural response is similar except that
all resistance after peak flexural response is due to compression membrane response
alone. Therefore, the total resistance reduces to zero at a deflection that is a little less
than the component thickness.
4-30
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Figure 4-16 Compression Ring Effect Providing In-Plane Restraint for Tension
Membrane in Two-Way Component (from Park and Gamble, 2000)
4-31
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
The peak resistance from compression membrane response and ductile flexural yielding
at midspan and at the supports, ruc, for concrete and masonry components with a solid
cross section and a one-way span is calculated as shown in Equation 4-14. Equation
4-15 shows the similar equation for two-way components. Equation 4-16 defines the
deflections Δuc and Δst in Figure 4-14 causing peak compression membrane resistance
and zero compression membrane resistance, respectively, for one-way and two-way
components. All three equations are from Chapter 10 of UFC 3-340-01. They are
derived in Park and Gamble (2000), which discusses static reinforced concrete slab
tests that are the basis for the equations. They are used in SBEDS to calculate the
contribution of compression membrane to the resistance-deflection relationships of
reinforced concrete, reinforced masonry, and unreinforced masonry except for concrete
masonry units (CMU) that have large internal void areas.
ruc =
8
(M uc′ + M uc − N uc Δ uc )
L2
⎧h ⎛ β 1 ⎞ Δ uc βL2 β β ⎫
(β 1 − 1)⎜ ε + ⎟ + ⎜ 2 − 1 ⎞⎟ + βL ⎛⎜1 − 1 ⎞⎟ε '⎪
⎛ 2t ⎞ Δ2uc ⎛ 2
⎪ ⎜1 − ⎟ + (β 1 − 3) +
8(0.85 f ' dc β 1 h ) ⎪ 2 ⎝ 2 ⎠ 4 4Δ uc ⎝ L ⎠ 8h ⎝ 2 ⎠ 4h ⎝ 2 ⎠ ⎪
ruc = ⎨ ⎬
⎪− β 1 β L ε ' 2
2 2 4
L ⎪
⎪ 16hΔ2 ⎪
⎩ uc ⎭
8 ⎧ 1 Δ Δ ⎫
− 2⎨ (T ′ − T − C s′ + C s )2 + (C s′ + C s )⎛⎜ h − d ′ − uc ⎞⎟ + (T '+T )⎛⎜ d − h + uc ⎞⎟⎬
L ⎩ 3.4 f dc ⎝2 2 ⎠ ⎝ 2 2 ⎠⎭
⎡ ⎛ h Δ uc T ′ − T − C s′ + C s ⎞ ⎤
⎢.85 f dc β 1 ⎜⎜ − − ⎟⎟ + C s − T ⎥
2t ⎛ 1 2 ⎞ ⎣⎢ ⎝2 4 1.7 f dc β 1 ⎠ ⎦⎥
where ε ' = ε + = ⎜⎜ + ⎟⎟
L ⎝ hE c LS ⎠ f ' β βL ⎛ 12
2 ⎞
1 + 0.2125 dc 1 ⎜⎜ + ⎟⎟
Δ uc ⎝ hE c LS ⎠
It is assumed in SBEDS that the effects of small, uniformly distributed voids in European
insulated block are accounted for in the compressive prism strength, f’m, where f’m is
determined based on the compressive failure force divided by the gross area.
Therefore, these blocks can be treated as solid material with an input compressive
prism strength that accounts for the effects of voids.
Equation 4-18 show modified versions of Equation 4-14 and Equation 4-15 for
unreinforced masonry with large voids centered at midthickness, such as CMU.
These equations account for the reduced compression force in compression
membrane response caused by voids concentrated at the center of the cross
section. They assume a cross section typical of CMU block, where there is solid
exterior shell and interior rectangular voids centered at midthickness at a regular
spacing. These two equations also account for a gap between the top of the wall
and the rigid supporting component, which is common construction practice to
allow expansion at the top of in-fill masonry walls. The values of Δuc and Δst in
Equation 4-17 are modifications of Equation 4-16 that include the effect of a gap.
For two-way unreinforced CMU, see notes at end of
Equation 4-18.
4-34
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Equation 4-15
⎧ L x ⎛ 1 3Δ uc Δ2uc ⎛1 Δ Δ ⎞ ⎞ ⎛ 1 Δ2 ⎛ 1 Δ2 ⎞ ⎞⎫
⎪ ⎜⎜ − + 2 − β 1 ⎜ − uc + uc2 ⎟ ⎟⎟ + ⎜ − uc2 − β 1 ⎜⎜ − uc 2 ⎟⎟ ⎟⎪
⎝ 4 4h 16h ⎠ ⎠ ⎜⎝ 2 12h ⎟
⎪ Ly ⎝ 2 4h 4h ⎝ 4 48h ⎠ ⎠⎪
⎪ 2 ⎪
⎪ ε ′x ⎛ L y ⎞ L x ⎛ ⎛ 2h 1 ⎞ h ⎞ ⎪
⎪+ 16 ⎜⎜ h ⎟⎟ L ⎜⎜ β 1 ⎜⎜ Δ − 2 ⎟⎟ + 1 − Δ ⎟⎟ ⎪
ruc =
(
24 0.85 f ' dc β 1 h 2 ) ⎪ ⎝ ⎠ y ⎝ ⎝ uc ⎠ uc ⎠ ⎪
⎨ ⎬
⎛ L ⎞ ⎪ ε ′y ⎛ L y ⎞ ⎛⎜ L x ⎛ β 1 h β 1
2
2h ⎞ ⎞ ⎪
L2y ⎜ 3 x − 1⎟
⎜ L ⎟ ⎪+ 16 ⎜⎜ h ⎟⎟ ⎜ 2 L ⎜⎜ Δ − 2 + 1 − Δ ⎟⎟ + (.5β 1 − 1)⎟⎟ ⎪
⎝ y ⎠ ⎪ ⎝ ⎠ ⎝ y ⎝ uc uc ⎠ ⎠ ⎪
⎪ ⎪
⎪− β 1 h L x ⎛⎜ L y ⎞⎟ ⎛⎜ (ε ′ )2 L x + (ε ′ )2 ⎞⎟
2 4
⎪
⎪ 64Δ2 L y ⎜ h ⎟ ⎜ x L y y ⎟ ⎪
⎩ uc ⎝ ⎠ ⎝ ⎠ ⎭
⎧ 1 ⎛⎜ ⎞ ⎫
Tx′ − Tx − C sx′ + C sx ) + x (T y′ − T y − C sy′ + C sy ) ⎟
L
⎪− ( 2 2
⎪
⎪ 3.4 f dc ⎜⎝ Ly ⎟
⎠ ⎪
⎪ ⎪
⎪ ⎛ h Δ uc ⎞ ⎛ h Δ uc ⎞ ⎪
⎨+ (C sx′ + C sx )⎜ − − d x ⎟ + (Tx′ + Tx )⎜ d x − +
24
+ ⎟ ⎬
⎛
2 ⎜ Lx
⎞⎪ ⎝2 2 ⎠ ⎝ 2 4 ⎠ ⎪
Ly 3 − 1⎟ ⎪ ⎪
⎜ L ⎟
⎝ y ⎠ ⎪+ (C ' +C )⎡ L x ⎛⎜ h − Δ uc − d ⎞⎟ + Δ uc ⎤ + (T ' +T )⎡ L x ⎛⎜ d − h + Δ uc ⎞⎟ − Δ uc ⎤ ⎪
sy ⎢ ⎥ y ⎢ ⎥
⎪ 4 ⎥⎦ ⎪⎭
sy y y y
⎩ ⎢⎣ L y ⎝ 2 2 ⎠ 4 ⎥⎦ ⎢⎣ L y ⎝ 2 2 ⎠
2t 2t y
where ε x′ = ε x + x and ε ′y = ε y +
Lx Ly
Note: subscript x or y refers to x or y span direction
where: ruc = peak resistance from compression membrane and ductile flexural
response
Δuc = midspan deflection at which ruc is reached
Muc= ultimate flexural moment capacity at midspan
M’uc= ultimate flexural moment capacity at supports
N = axial force from compression membrane forces
L = clear span length of the member
h = overall thickness of the member
β1 = ratio of the concrete compression block depth to the distance from
the compression face to the neutral axis = 0.85 for f'c =30 MPa and
decreases by 0.008 for each 1 MPa that f'c exceeds 30 MPa; Β1≥0.65
β = ratio of the distance from the support to the interior yield line, to the
member length L
Cs, C's = compression force per unit width in the reinforcing steel at interior
yield lines and supports, respectively, assuming yielding of steel
T, T’= tension force per unit width in the reinforcing steel at the interior
yield lines and at the supports, respectively, assuming yielding of steel
d = distance from the compression face to the centroid of the tension
4-35
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
reinforcement; assumed to be the same at interior yield lines and
supports
d’= distance from compression face to the centroid of the compression
reinforcement; assumed to be the same at interior yield lines and
supports
f’dc = dynamic ultimate compressive strength of the concrete
Ec = concrete modulus of elasticity
S =lateral stiffness of the supports (assumed equal to 0.2 Ec in SBEDS)
ε= member axial strain
ε'= combined member strain and support strain (definition for ε' in
Equation 4-14 applies in each direction in Equation 4-15)
t = lateral displacement of each support corresponding to ruc
Equation 4-16
Δ uc = 0.033L ≤ 0.5h
Δ st = 0.08L ≤ 0.8h
where:
Δuc = midspan deflection at ultimate compression membrane capacity
Δst = midspan deflection for snap-through of compression membrane
response
L = clear span length (minimum span of two-way components)
h = overall thickness of the component
Note: Equation 4-16 recommended primarily for 2 < L/h < 30, but used for
all cases in SBEDS
Equation 4-17
ruc =
8 ⎡ ⎛ h ts ⎞ ⎛h (0.85c − t s ) ⎞ − (C + C )Δ ⎤
⎢2 C1 ⎜ 2 − 2 ⎟ + 2C 2 ⎜ 2 − t s − ⎟ 1 2 uc ⎥
L2 ⎣ ⎝ ⎠ ⎝ 2 ⎠ ⎦
h Δm ⎛ δg ⎞ 2t
c= − − βL2 ⎜⎜ ε '+ ⎟⎟ where ε ' = ε +
2 2 ⎝ L ⎠ L
C1 = 0.85 f ' m t s
C 2 = 0.85 f ' m k (0.85c − t s )
⎛ L h⎞
Δ uc = MIN ⎜ , ⎟ + Δ g
⎝ 30 2 ⎠
⎛ L +δg
2
⎞
2
⎛ L⎞
Δ g = h − ⎜ ⎟ + h 2 − ⎜⎜ ⎟⎟
⎝2⎠ ⎝ 2 ⎠
[
Δ st = MIN (0.08L,0.8h ) + Δ g ] ≤h
4-36
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Equation 4-18
⎡ Lx − L y Δ uc ⎛ ⎞⎤
⎢m' x + m x + m' y + + m y + (m y + m' y ) ⎜ n x − n y + 2n y Lx ⎟ ⎥
24
ruc = −
L y (3Lx − L y ) ⎣⎢ Ly 2 ⎜⎝ L y ⎟⎠⎥⎦
⎛h t ⎞ ⎛h
m x = m' x = C1 x ⎜ − s ⎟ + C 2 x ⎜⎜ − t s −
( )
0.85c x − t s ⎞
⎟⎟
⎝2 2⎠ ⎝2 2 ⎠
⎛h t ⎞ ⎛h (0.85c y − t s )⎞⎟
m y = m' y = C1 y ⎜ − s ⎟ + C 2 y ⎜⎜ − t s − ⎟
⎝2 2⎠ ⎝2 2 ⎠
C1x = C 2 x = 0.85 f ' m t s
C 2 x = 0.85 f ' m k x (.85c x − t s )
C 2 y = 0.85 f ' m k y (.85c y − t s )
h Δ uc 2⎛ δg ⎞ 2t
cx = − − βLx ⎜⎜ ε '+ ⎟⎟ where ε ' = ε +
2 2 ⎝ Lx ⎠ Lx
h Δ uc ⎛ δg ⎞ 2t
cy = − − βL y ⎜ ε '+ ⎟ where ε ' = ε +
2
2 2 ⎜ Ly ⎟ Ly
⎝ ⎠
n x = C1x + C 2 x
n y = C1 y + C 2 y
Note: subscript x or y refers to x or y span direction
4-37
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
β = distance from support to yield location divided by span length = 0.5
mi = resisting positive moment from arching per unit width in the i direction
m’i = resisting moment at supports from arching per unit width in the i
direction
ni = compression force from arching per unit width in the i direction
Notes: 1) β(x,y)<=0.5 for two-way component depending on yield line pattern,
however, conservatively β = 0.5 in SBEDS
3) δg conservatively assumed equal in both directions for two-way spans
4-38
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
P axial P axial
v v
h
Length, L
r1 & r2
x
r3
x
v v
W + Paxial W + Paxial
Response prior to ultimate flexural resistance, r2 Response after r2 where x= arching moment arm
Equation 4-19
r3 =
8
(h − x 2 )⎛⎜ P + WL ⎞⎟
L2
⎝ 2 ⎠
where:
r3 = maximum resistance from axial load effects
x3 = flexural deflection at r2+ (r3 – r2) /Kep
Kep = elastic-plastic stiffness for indeterminate components, otherwise
equal to elastic stiffness
h = overall wall thickness
P = input axial load per unit width along wall, Paxial
W = areal self-weight and supported weight of wall
L = span length equal to wall height
Resistance from axial load arching, as illustrated in Figure 4-18, is only significant for
thick components with a low post-yield flexural capacity such as unreinforced masonry.
Unreinforced masonry, which is the only component type where axial load arching
response can be considered in SBEDS, responds brittlely in flexure, essentially loosing
all flexural resistance after flexural yielding in the maximum moment regions. Axial load
arching is more significant for load bearing walls, but even the self-weight above
midspan of an unreinforced masonry wall may cause enough resistance and strain
energy through arching action to be significant compared to the resistance and strain
energy from brittle flexural response.
r3 Decaying
Phase r3
Decaying
Phase
Resistance
Resistance
r2
r1 = r2
k ep
r3 k ep
ke 1 1
ke 1
r3 1 r1
Elastic &
ke Elastic-Plastic
Elastic 1
x1= x2 x3 h x1 x2 x3 h
Deflection Deflection
Table 4-4
= V/(ruL) from Table 4-3 for one-way components
= Vx/(rux), Vy/(ruy), Vx/(ruLx), or Vy/(ruLy) from
If the ultimate shear resistance of the component is less than its ultimate flexural
resistance, all flexural resistances in the resistance-deflection relationship that are
greater than the ultimate shear resistance (i.e., this may or may not involve the initial
elastic flexural resistance) are set equal to the ultimate shear resistance if they are
larger. Parts of the resistance-deflection relationship based on compression membrane,
tension membrane, or axial load arching response are unaffected. Also, all flexural
stiffness values corresponding to resistances less the ultimate shear resistance are
unaffected. Therefore, the ultimate shear resistance is used only to limit the maximum
flexural response in the component at the ultimate shear resistance. This assumes that
component response after the resistance equals the ultimate shear resistance (i.e.,
shear yielding) involves ductile yielding in a similar manner as flexural yielding. Shear
yielding is not well understood, so the shape of the resistance-deflection curve after the
component deflects past the ultimate shear resistance is not known. However, as
discussed in Section 0, the allowable deflections in a component after shear yielding are
usually very limited.
For example, if a simply supported beam has an ultimate dynamic shear capacity of
40,000 lbs near each support, and shear span of 100 inches, and a loaded width of 100
inches, the ultimate shear resistance is 4 psi. If the ultimate flexural resistance is 5 psi,
then the ultimate shear resistance of 4 psi is used as the ultimate resistance for the
component’s resistance-deflection relationship because it is lower than the ultimate
flexural resistance. All stiffness values in the resistance-deflection relationship are
4-41
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
unchanged and any initial elastic yield resistance is unchanged if it is less than 4 psi.
The component is assumed to yield ductilely at 4 psi, where the resistance remains
constant at this value with increasing deflection. The plastic yield deflection is the
deflection in the resistance-deflection relationship where the resistance is equal to 4 psi.
SBEDS will automatically prompt the user to set a “shear flag” if the calculated ultimate
shear resistance is lower than the ultimate flexural resistance. After setting the shear
flag to 1 or 2, corresponding to a k value of 0 or 1, respectively in Equation 4-20,
SBEDS will automatically adjust the resistance-deflection relationship as stated above
and the user can click the SDOF button to recalculate the SDOF equation of motion
based on the modified resistance-deflection relationship where all flexural response is
capped by the ultimate shear resistance. Also, the user may decide to ignore the shear
flag if the blast load causes limited response and the shear loads are less than the
shear capacity of the component. However, this should only be considered when the
threat is well known and is unlikely to ever be exceeded.
Softening, which is shown in Figure 4-20b, occurs for components with compression
membrane response and axial load arching. For components with compression
membrane response, SBEDS automatically arranges the rebound resistance-deflection
curve to prohibit compression membrane during rebound until the deflection is negative,
since this response mode depends on the location of the midspan deflection relative to
the supports. Similarly, it automatically arranges the resistance-deflection curve for
subsequent inbound response to prohibit compression membrane until the deflection is
positive. It is arguable if previously crushed material can still provide the same stiffness
and strength during subsequent response cycles, but SBEDS simplistically assumes no
degradation of compressive strength or stiffness due to multiple response cycles.
4-42
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
For components with axial load arching (only unreinforced masonry), SBEDS
automatically arranges the rebound resistance-deflection curve to prohibit flexural
response during rebound if the ultimate flexural resistance has occurred during initial
inbound response, implying that the component has cracked in a brittle fashion through
the thickness. Otherwise, this case is identical to the case with compression membrane
from rigid supports. Therefore, this case is identical to the case with compression
membrane from rigid supports except that the maximum negative flexural resistance is
zero. If the ultimate flexural resistance does not occur during inbound response, then no
axial load arching occurs and the resistance-deflection relationship is symmetric for
inbound and rebound response. More information on this case is provided in Section 7-
8. These rules used in SBEDS to automatically modify the resistance-deflection
relationship during rebound as summarized in Table 4-8. The user can specify these
rules for a General SDOF component with the input Rebound Flag value in Table 4-9.
4-43
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Figure 4-20 General Resistance-Deflection Diagrams in SBEDS
R max
R3 = R4 k5
R2 k4 = 0
k3
Resistance
k2 k1_rb
R1
k1
x 4_reb x4 x max
R 1_rb
Deflection
k2_rb
R 2_rb
R max
R2
Resistance
Softening
(k 3 < 0)
k2_rb = 0 k2 k3 k5
R1
R 3 =R 4 k1_rb
k1 k4 = 0
X 4_rb X 2_rb = 0
k1 R 1_rb = R 2_rb = R 4_rb x4 xmax
4-44
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Table 4-7 Definition of Terms for Stiffness Region Criteria Tables
Term Definition
Ri where i=1..4 Inbound yield resistance for ith stiffness region (Ri ≥ 0)
Ri_rb where i=1..4 Rebound yield resistance for ith stiffness region (Ri_rb≤ 0)
R_max Maximum resistance from previous cycle
ki Stiffness value for ith stiffness region in inbound response
ki_rb Stiffness value for ith stiffness region in rebound.
xi and xi_reb Yield displacements for ith stiffness region in inbound and rebound. SBEDS uses
these values as a basis for changing stiffness region when the current stiffness
equals zero. Otherwise, SBEDS compares the resistance at the current time
step to the yield resistance for the current stiffness region to change stiffness
regions,. see Table 4-8 for more information.
Table 4-8 Criteria for Determining To Stay in Current Stiffness Region During
Dynamic Response
Case Criteria for Staying in ith Region 1,2
th
Response in i region if region has no SBEDS stays in current ith stiffness region for ki >0 when R<Ri
softening in resistance-deflection during inbound and R>Ri_rb during rebound and for ki <0 when
relationship (i.e., no negative R>Ri during inbound and R<Ri_rb during rebound. See next two
stiffnesses) rows for cases with ki = 0 and ki_rb = 0.
Response in ith inbound region when ki SBEDS stays in current ith stiffness region for x<xi if ki+1 ≠ 0,
= 0 and no softening otherwise ki = 0 until rebound response.
Response in ith rebound region when SBEDS stays in current ith stiffness region for x>xi_rb if ki+1_rb ≠ 0,
ki_rb = 0 and no softening otherwise ki_rb = 0 until next inbound response.
Response in ith region for components Same as all three cases above for first inbound and rebound
with softening from compression response except SBEDS creates rebound resistance-deflection
membrane response between rigid relationships where no rebound compression membrane occurs
supports until x< 0, using a zero stiffness as required (see Figure 4-20b),
and Ri=-Ri_reb during subsequent inbound (i.e., rebound
resistances and stiffnesses used for inbound response after
first inbound cycle) – see discussion in paragraphs below.
Response in ith region for components Same as first three cases above if no cracking (i.e., no
with softening from brittle flexural softening) has occurred during inbound response. Otherwise,
response and axial load arching (only same as softening case above except maximum resistance
unreinforced masonry components) during rebound = 0 until x< 0 (i.e., no negative flexural
resistance assumed during rebound, negative axial load
arching resistance occurs for x<0) see discussion in
paragraphs below.
Note 1: If velocity changes sign, stiffness is changed to input k1 value for opposite response direction
Note 2: When criteria for staying in region is not met, stiffness is changed to the input stiffness value for
next highest region. When i=5, the stiffness is only changed by rebound (i.e., velocity sign change).
Finally, when any inbound stiffness K1 through K5 exceeds K1_rb, calculations are
stopped at the maximum deflection in SBEDS and no rebound response is calculated
because the area under the resistance-deflection curve could become negative during
rebound. This is only of practical concern for unusual cases, such as where the tension
membrane stiffness (KTm) of a component is unusually large and exceeds the flexural
stiffness (KE) in Figure 4-12. Usually K1_rb is assumed equal to K1 in SBEDS and K1 is
larger than any other inbound stiffness.
4-45
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Table 4-9 Rebound Flag Information
Rebound Rules Used by SBEDS for Rebound Response Based on Rebound Flag
Flag
1 Rebound stiffnesses and resistances input into the General SDOF input sheet are always
used for rebound response. This the default value on the SDOF input sheet.
0 Rules for stiffness applicable for components with compression membrane response are
used in SBEDS. These rules cause the negative value of the input rebound resistance
values and the input rebound stiffness values to be used as the inbound resistance and
stiffness inbound values after the first inbound cycle. The rebound stiffness and resistance
should be input by the user to prevent compression membrane stiffness from occurring until
the deflection is less than zero during rebound as shown in Figure 4-20b.
-1 Rules for stiffness applicable for unreinforced masonry components with brittle flexural
response and arching from axial load are used in SBEDS. These rules for based on loss of
all flexural stiffness and resistance after cracking (i.e., after first negative stiffness value in
calculated response). User should input rebound assuming cracking occurs first inbound
response with initial rebound yielding at zero resistance and no negative phase axial load
arching until x<0, similar to Figure 4-20b with R1_rb= R2_rb= R4_rb=0.
As shown in Figure 4-21, the strain energy is a function of both the component ultimate
resistance and maximum deflection. It is generally much better to absorb strain energy
with a large maximum deflection than a large ultimate resistance. The component shear
and connection forces increase directly with the resistance, so that a large ultimate
resistance causes larger component shear and connection forces that are transferred
into the building lateral or gravity resisting system - all of which must be accounted for in
design. On the other hand, large component displacements are only a problem in rare
cases where the component may impact nearby personnel or sensitive equipment.
Therefore, it is better, when possible, to absorb sufficient strain energy to resist the blast
load by increasing component ductility or using a ductile component, so that it can
undergo large deflection without failing, rather than increased component strength and
ultimate resistance.
Also, Figure 4-21 demonstrates how the strain energy is most dependent on plastic, or
post-yield, flexural response occurring between deflections xy and xm, as opposed to
elastic response before the component becomes a mechanism (i.e., for deflections
when x<xy). The strain energy absorbed during elastic flexural response is usually a low
4-46
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
percentage of the total strain energy, except for components with very low ductility.
Consequently, the accuracy of the calculated elastic and elastoplastic stiffnesses is less
important than the accuracy of the ultimate flexural resistance, ruf. Components typically
exhibit low ductility when the ultimate resistance is controlled by brittle response modes
such as shear, buckling, rebar slip, etc. Usually components can be designed to avoid
these conditions.
Figure 4-21 Strain Energy for Component with Ductile Flexural and Tension
Membrane Response
Resistance
rm
KTm
ruf 1
Ke
xy xTm xm
Maximum Deflection
4-1.2.5 REBOUND
Rebound response occurs after the component reaches maximum deflection and
begins moving back in the direction of the applied blast load. Peak rebound stresses
and connection forces are directly related to the maximum rebound resistance. Figure
4-22 from TM 5-1300 shows ratios of maximum rebound resistance to the inbound
ultimate resistance of a SDOF spring-mass system, where all contours are upper
bounds for the contour values shown in the figure. This diagram is provided mainly to
help provide an understanding of rebound response since SBEDS will directly calculate
the rebound response of input components as part of its solution of the equation of
motion of the equivalent SDOF system representing the component.
All cases lying above and to the left of the -1.0 contour have a rebound resistance equal
to the inbound ultimate resistance. Tn in the figure is the natural period of the SDOF
system (see Equation 4-22). If xm/xe is less than 1.0 (i.e. elastic response), the rebound
resistance can be assumed equal to the peak inbound resistance. Long duration blast
load durations, relative to the component natural period, and low inbound maximum
deflections, xm, relative to the component yield deflection, xe, both tend to decrease the
rebound resistance. An equivalent static design load for rebound equal to one-half the
inbound ultimate resistance is often used as a rule of thumb for blast design. Figure
4-22 shows that this assumption is applicable when T/Tn >1.5 to 2 and xm/xe < 5 (i.e.,
low plasticity during inbound response) and when T/Tn>3 to 4 for larger xm/xe values.
Typically, a component will have some reserve capacity and ductility that will help
4-47
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
prevent failure if the rebound resistance from Figure 4-22 is a little larger than the
ultimate rebound resistance calculated for the component.
Figure 4-22 Ratio of Maximum Rebound Resistance to Inbound Ultimate
Resistance for SDOF System
100
10
7
6
5
4
1
0.2 0.3 0.5 0.7 1 2 3 4 5 6 7 8 910 20 30 40 50 70 100
T / Tn
Usually, maximum response of blast-loaded components occurs during the first inbound
response cycle. However, this is not always the case. Components with nonsymmetrical
section properties or nonsymmetrical bracing can be much weaker during rebound than
during inbound response because the cross section or bracing is designed to be
effective against inbound response stresses rather than the reversed stresses that
occur during rebound (i.e., reversal of locations within cross section where tension and
compression stresses occur). Also, any damage to the component during inbound
response can reduce its properties during rebound, although such damage is usually
very difficult to quantify. Finally, in some relatively unusual cases the negative phase
blast load can coincide with rebound response, causing the rebound response to be
larger than inbound response.
In spite of its possible importance, component response to blast loads during rebound is
not well understood. This is due in large part to a scarcity of observed rebound phase
damage to components in blast tests and accidental or terrorist explosions. A blast load
must be small enough not to cause significant damage during inbound motion, but large
enough to cause observable damage during rebound. Also, components may have
4-48
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
redundant response modes, such as limited tension membrane capacity, that are large
enough to compensate for lower flexural resistance during rebound for the brief period
before a second inbound response occurs.
The SDOF analyses in SBEDS account for nonsymmetrical conditions that reduce the
ultimate flexural resistance in rebound based on user input properties for some
component types, as discussed in Chapter 5 through 8. These inputs are only provided
for cases where a parameter of the resistance-deflection relationship can be different
during rebound response based on typical construction practice. However, the
resistance-deflection relationships for most component types are assumed to be
symmetric for inbound and rebound response in SBEDS, No degradation of component
properties from inbound response. The only exception for this is unreinforced masonry
with brittle flexural response and axial load arching, as explained in Section 4-1.2.3.
4-1.3 DAMPING
Damping is a resisting force that develops during dynamic component response due to
energy dissipation that is not included in the strain energy, work energy, or kinetic
energy. It is only well understood for simple systems, such as mass oscillation in a
viscous fluid (i.e., viscous damping) or along a frictional surface (i.e., Coulomb
damping). In practice, damping is a catch-all category that approximately accounts for
all energy dissipated during dynamic response that is not accounted for with other terms
in the equation of motion. In blast-loaded components, it is used to account for
numerous complex factors that cause energy losses, such as slip and frictional action at
joints and supports, material cracking, and concrete reinforcement bond slip.
Damping can be measured in a given dynamic system based on the decay in maximum
response during successive response cycles of unforced elastic oscillation (i.e., free
vibration). However, there are not any available methods to predict damping force
coefficients for typical blast-loaded components other than approximate viscous
damping coefficients recommended for structural response to earthquake motion. The
damping values recommended for earthquake response are intended for cyclic
response of larger component systems, such as reinforced concrete and steel frames in
sway response, that is mainly elastic response. There is no available discussion on the
validity of these methods for large plastic component response, which is usually the
case for blast-loaded structures. Thus, there is very little available guidance on damping
that is directly applicable for blast-loaded components.
In spite of this, a viscous damping term is generally included in the equivalent SDOF
system for a blast-loaded component, as shown in Equation 4-5, so that some damping
can be accounted for if desired. The damping force is equal to a damping constant
multiplied by the velocity of the SDOF system at each time step. The damping constant
is usually defined in terms of a damping ratio, ζ, which is multiplied by the critical
damping constant for the system, Ccr, to calculate the damping constant as shown in
Equation 4-21. Ccr is a property of the SDOF system, which is usually based on the
initial elastic stiffness and load-mass factor of the system. In SBEDS, the user inputs a
damping ratio, ζ, as a percentage, which is multiplied by Ccr within SBEDS to get the
damping constant Cc used in the equation of motion in Equation 4-5 for the equivalent
SDOF system representing the input component.
4-49
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Equation 4-21
C cr = 2 kMK Lm
ζ
Cc = C cr
100
where:
Ccr = critical damping constant
Cc = damping constant used in equation of motion (See Equation 4-5)
ζ = input damping ratio (as a percentage)
k = elastic stiffness of component
M = mass of component
KLm = load-mass factor
• Damping has very little effect on the first peak of response, which is usually the only
cycle of response that is of interest.
• The strain energy dissipated through plastic deformation is much greater than the
energy dissipated by normal structural damping.
• Ignoring damping is a conservative approach.
UFC 3-340-01 recommends that damping values commonly used for earthquake
analysis and design in Table 4-10 can be used for elastic and elastoplastic component
response only. Therefore, the damping term is set to zero for all plastic response. This
is recommended because the damping forces for structure response to earthquake
loading were developed primarily for elastic component response, which involves
deformation over the whole volume of the component. After ultimate resistance,
deformation consists primarily of localized hinging at maximum moment regions, which
occurs over a much smaller volume, so that no significant additional damping occurs.
Also, damping is not specifically defined for plastic response in available blast design
and analysis references, including TM 855-1, TM 5-1300, or in a number of structural
dynamics textbooks that were reviewed.
4-50
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
This response consists is elastic flexural response followed by axial load arching, which
inherently causes some frictional response. The use of a 2% damping ratio causes the
calculated maximum deflection and damage to match data much better in limited
comparisons than the more conservative case of zero damping after brittle flexural
response (Oswald, 2005).
Table 4-10 Damping Ratio Information For Primarily Elastic Dynamic Response
The impulse i’ from the initial segment of the blast load history can be input into the
SBEDS as an equivalent initial velocity, vo, based on Equation 4-22 if the duration of
this segment, td’, is less than one-third of the natural period of the equivalent SDOF
system. The user should calculate and input v0 based on Equation 4-22 and input the
remainder of the blast load not including the impulse i’, as discussed in the previous
4-51
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
paragraph, since it is accounted for with v0. This will cause SBEDS to recommend a
larger input time step based on a longer segment of the blast load history, so that
response can be calculated over a longer response time.
Equation 4-22
i' 1 2i ' K LM 1m
v0 = where t d ' < Tn and t d ' = , Tn = 2π
mK LM 1 3 P' k1
where:
i’ = impulse from segment of blast load with duration td’
vo = equivalent initial velocity representing i’
KLM1 = load mass factor for initial elastic response
k1 = stiffness for initial elastic response
Tn = natural period of component
m = mass of component
The effect of fragments on the global response of a component can also be included in
SBEDS by inputting the impulse applied by the fragments as an initial velocity using
Equation 4-22. The fragment impulse, equal to the summed momentum of all
fragments impacting the component area divided by the component area, is equal to i’
Equation 4-22. It can be assumed for this case that td’, representing the duration of the
load pulse from fragment impact, is less than one-third of the natural period of the
component. The effect of fragment localized penetration and breaching should be
investigated separately using applicable methods in TM 5-1300 or UFC 3-340-01.
• If the component is parallel to the ground surface and deflects under gravity in the
same direction as the applied blast load, such as a roof component loaded by
exterior blast load, the initial displacement is calculated as the sum of the
component self-weight and supported weight divided by the initial component
stiffness. SDOF calculations start at time zero with this initial displacement and a
resistance equal to the self-weight and supported weight. This condition causes no
initial movement. Both the initial displacement and resistance have positive sign in
this case.
• If the component is parallel to the ground surface and deflects under gravity in the
opposite direction as the applied blast load, such as a roof component under internal
blast load, an identical approach is used except the initial displacement and
resistance have negative sign.
4-52
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
The SBEDS user must define one of these three cases in a dropdown box on the Input
Sheet for the component orientation relative to the applied blast load. Initial
displacement not caused by static conditions (e.g., when a spring is pulled and then
suddenly released) can only be input as an initial deflection in the General SDOF input
in SBEDS. This displacement adds to any displacement from gravity effects. It is rarely
applicable in practical blast design problems..
4-53
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Figure 4-23 only considers a single load-mass factor within Tn and no elastoplastic
response.
Figure 4-23 also shows the time to maximum displacement, tm. Graphical
solutions such as
Figure 4-23 are very useful when the blast load history has a given simplified shape and
the simplifying assumptions noted above for the equivalent SDOF system are
acceptable. Similar graphs for some other simplified blast-load history shapes are
available in Appendix D of UFC 3-340-01 and Chapter 3 of TM 5-1300.
Figure 4-23 Undamped Response Charts for SDOF Response to Right Triangular
Blast Load History
4-54
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
SBEDS uses a numerical method to solve the equation of motion for the equivalent
SDOF system representing the input component at each time step. The Newmark-Beta
method is most commonly used in available computer programs that solve the SDOF
equation of motion in this manner. The constant velocity method is probably the
simplest numerical method recommended for solving the equation of motion for blast
4-55
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
design and analysis problems (Biggs, 1964). The choice of a numerical solution method
is a tradeoff between the complexity of terms used to determine the approximate
acceleration, velocity, and displacement at each time step and the number of time steps
required to achieve a given level of accuracy. For example, Biggs found that
approximately 1.5 times more time steps were necessary to achieve the same accuracy
with the constant velocity method compared to the more complex linear acceleration
method for a given SDOF system. Numerical solution methods are also unconditionally
stable (i.e., they will always tend to converge to the correct solution as the time step is
reduced) or conditionally stable (i.e., will converge when the time step is less than a limit
value). However, any time step that is small enough to cause the numerical solution to
meet accuracy requirements will almost always ensure stability.
The SBEDS workbook uses the constant velocity method, as described by Biggs
(1964), to perform the time-stepping solution of the equation of motion. It was noted in
trial analyses with several numerical solution methods that were initially considered that
“overshoot” of the resistance at a time step where a yield occurred in the resistance-
deflection relationship was the primary cause of error. In a computer program, this
problem can be corrected by backtracking in time and subdividing the time step based
on an interpolation so that the stiffness prior to yield is used only for that portion of the
time step prior to yield and the new stiffness is used for the remainder of the time step.
In a spreadsheet such as SBEDS, however, backtracking is very difficult. Therefore,
SBEDS recommends a very small time step to the user that practically eliminates the
overshoot. Since a small time step is required for this reason, the simple constant
velocity method is used in SBEDS as the numerical solution method. Based on
comparisons to computer programs using the Newmark-Beta method, very accurate
solutions (i.e., typically within 2%) are obtained using the small recommended time step
in SBEDS.
The bulleted list below shows upper bound limits on the recommended time step in
SBEDS used to determine SDOF response for an input blast-loaded component. As
shown in the list, a smaller time step is recommended for cases where blast loads
histories are calculated from charge weight-standoff input, which have an exponential
decay (see Figure 2-1) where the load changes more than proportionally with time near
the peak pressure. The last bulleted requirement helps ensure that a small time step
will be used even when the natural period and time segments associated with the blast
load history are large. This is necessary in SBEDS to help minimize overshoot of the
resistance during time steps where yielding occurs. SBEDS displays a recommended
4-56
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
time step based on the smallest time step determined from the bulleted list below, but
not less than 0.01 millisecond.
A large number of time steps are hardcoded into SBEDS (i.e., 2900), but SBEDS will
not calculate response all the way out to the peak deflection if the time step is too small
compared to the component response time. In this case, which is rarely occurs, SBEDS
displays a warning message. If no peak deflection is calculated, but the maximum
deflection is very large, this may indicate that the component is failed and no further
analysis is required. If the recommended time step is controlled by a short duration for
the initial segment of the blast load history, the impulse in this segment can be replaced
by an equivalent initial velocity as described in Section 4-1.4.1. Then the time step can
be increased. Otherwise, the user can usually increase the time step to approximately
1.5 times the recommended time step without causing an unacceptable decrease in the
accuracy of the calculated solution.
The dynamic reaction forces from a supported component, such as a beam, can be
used as the applied load on a supporting component, such as a girder. Although it is
more complicated, this approach is generally more accurate than applying the blast load
directly to the supporting component because it accounts for the significant change that
can occur in the load as it is transmitted through the supported component. This is
particularly true if the supported component has a low ultimate resistance compared to
the peak blast load. Dynamic reaction forces are not usually used to determine the
maximum shear stress in a component or to design connections since the design
strengths used for these cases are based on dynamic strength increase factors for
overall component response rather than dramatic fluctuations that can occur in the first
milliseconds of dynamic reaction force histories.
In-plane reactions are not considered in either the equivalent static, or dynamic reaction
forces when tension or compression membrane response occurs. Out-of-plane
4-57
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
reactions from tension or compression membrane response can be considered since
these response modes increase the resistance, and therefore the calculated reaction
forces. However, only the reactions due to flexural response are typically considered in
blast design because tension and compression membrane response is based on
available capacity at the supports (i.e., compression membrane is only considered if the
supports are rigid and therefore resist the maximum connection forces associated with
crushing of masonry/concrete, tension membrane response is calculated based on input
connection force capacity). Also, diagonal shear stresses are only compared to the
reaction force caused by flexural response since they are caused by flexure.
The assumption that the acceleration distribution along the span equals the deflected
shape is correct only if the shape function does not vary with time. However, this is an
approximate approach because dynamic finite element calculations show that very early
in time the deflected shape of blast-loaded components subject to blast loads is flatter
than the static deflected shape, with almost all deflection occurring very close to the
supports. At later times, when significant deflections occur, the shape changes to more
closely approximate the first mode, or static flexural response shape that is typically
assumed in SDOF analyses. Therefore, the dynamic reaction is typically least accurate
during very early time response.
4-58
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Equation 4-23
61 ⎛ 61 ⎞
V (t ) L − M c (t ) − 0.5F (t )⎜ L − 0.25L ⎟ = 0
192 ⎝ 192 ⎠
R(t ) L
M c (t ) =
8
V (t ) = 0.39 R(t ) + 0.11F (t )
where:
V(t) = dynamic reaction force
R(t) = resisting force
Mc = midspan resisting moment
F(t) = applied blast load
L = span length
Similar free body diagrams and equations can be developed for one-way components
with other boundary conditions and load distributions, and for two-way components with
uniform blast load. The dynamic reaction can be solved in the form shown in Equation
4-24 for each of these cases. Since different shape factors are assumed for elastic and
plastic response, the factors CR and CF in Equation 4-24 are a function of the response
realm, as well as the boundary conditions and blast load distribution. Static deflection
shape functions are assumed for elastic response and linear deflection shape functions
based on the yield line pattern are assumed for plastic response. See the discussion in
Section 3-5 for use of VR(t) as a load on a primary component in SBEDS.
Equation 4-24
VR (t ) = CR R(t ) + CF F (t )
where:
VR(t) = dynamic reaction force of component1,2
R(t) = resistance of equivalent SDOF system1
F(t) = blast load on equivalent SDOF system1
CR = coefficient applied to resistance
CF = coefficient of applied to blast load
t = time
Note 1: These parameters have same units as applied blast load in SBEDS
Note 2: VR(t) must be multiplied by the full blast loaded area of the component to
get the reaction force
SBEDS calculates V(t) for each support for each SDOF analysis based on Equation
4-24 using CR and CF values for elastic response at all time steps when the resistance
is between the inbound and rebound ultimate flexural resistances and CR and CF
values for plastic response at all other time steps. CR and CF for many different one-way
and two-way cases with differing boundary conditions and loading distributions are
published in Chapter 11 of UFC 3-340-01. Also, CR and CF values are provided at, or
along each support for components with nonsymmetrical support conditions. Morison
(2005) also provides CR and CF values derived from dynamic force equilibrium
equations using MathCad calculations for one-way spans and static finite element
analyses for two-way spans. Unlike the CR and CF values in UFC 3-340-01, Morison’s
4-59
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
values are based on documented calculation procedures and present day
computational methods. There are many cases of “small” disagreement between the
two sets of CR and CF, values, which are within 20%, and a limited number of cases
where the disagreement is more significant. SBEDS predominantly uses CR and CF,
values in UFC 3-340-01 because it is based primarily on DoD procedures. However,
Morison’s values are used in some cases where there was significant disagreement
between the two sets of values.
Table 4-11 shows the CR and CF coefficients from Equation 4-24 used in SBEDS for
one-way spanning components. The elastic values are also used for elastoplastic
response. These values are taken from UFC 3-340-01, except shaded coefficient values
are taken from Morison (2005) because they differed from UFC values by more than
20%. Table 4-12 through Table 4-15 show equations for the dynamic reactions of two-
way components with uniformly applied blast load from UFC 3-340-01. The CR and CF
values for two-way components in SBEDS are based on curve-fits to the values in the
equations in these tables. The curve-fits values are within 0.02 of the coefficients in the
tables for the large majority of cases, with a maximum difference of 0.04. The CR and CF
values in Table 4-12 through Table 4-15 are based on yield line patterns for two-way
components with equal moment capacities in each direction. The elastoplastic values in
the tables are not used in SBEDS since it does not consider dynamic reaction in this
response region.
There are some very significant differences between the CR and CF coefficients in Table
4-12 through Table 4-15 and similar coefficients from Morison in the elastic response
range, and lesser differences in the plastic realm. Morison’s values are not used in
4-60
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
SBEDS for two-way response because there are few cases of significant difference in
the plastic response realm where most response to blast loads typically occurs. Also,
there are more inherent assumptions in the calculation of the dynamic reaction forces of
two-way components for an equivalent SDOF system that cause both procedures to be
very approximate. In particular, Morison’s finite element analyses show that the dynamic
reaction forces of two-way components vary significantly along the supports, which is
not addressed at all in the SBEDS approach. Based on Morison’s calculations, the
equations in Table 4-12 through Table 4-15 from the UFC can be considered to
represent approximate average dynamic reaction force values. Significantly higher
localized dynamic reaction forces occur along the supports. VA is the dynamic reaction
along the short side of the two-way spanning component in Table 4-12 through Table
4-15, and VB is the dynamic reaction along the long side.
Table 4-12 Dynamic Reaction Coefficients for Two-Way Slabs with Four Sides
Fixed and Two Sides Fixed
Four Sides Fixed
Two Adjacent Sides Fixed
4-61
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Table 4-13 Dynamic Reaction Coefficients for Two-Way Slabs with Four Sides
Simply Supported
Four Sides Simply Supported
Table 4-14 Dynamic Reaction Coefficients for Two-Way Slabs with Opposite
Edges Fixed and Simply Supported
Simply-Supported Short Edges, Simply-Supported Long Edges,
Fixed Long Edges Fixed Short Edges
4-62
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Table 4-15 Dynamic Reaction Coefficients for Two-Way Slabs with Three Edges
Fixed
Three Sides Fixed
Table 4-4, which are based on yield line theory. The equivalent support shears in Table
4-3 and
The equivalent static reaction forces in SBEDS are never based on a static load greater
than the ultimate flexural resistance, ru. This is true even when tension or compression
membrane causes the overall resistance to exceed ru, because the equivalent static
reaction force is typically not used to calculate shear or connection forces for these
response modes. SBEDS always calculates the maximum equivalent static reaction
loads based on ru, even when the calculated maximum component resistance is less
than ru, because this ensures that the connections will not be the weak point in the
component design. Connections typically have significantly less ductility than the
connected components and therefore the ultimate strength of the component should not
be controlled by the connections. The user can redefine the maximum equivalent static
shear forces calculated by SBEDS based on the ratio of rmax/ru for cases where the
component does not yield at all maximum moment regions if they are confident that the
calculated blast load will not be exceeded and the maximum equivalent static shear
force calculated by SBEDS causes the connection force or maximum shear stress to be
too large.
For reinforced concrete and masonry components, SBEDS also calculates the
equivalent static reaction force at a distance “d” from the support using modifications of
the formulas in Table 4-3 and
Table 4-4 as shown in the bulleted list below. The shear at a distance “d” from the
support (where d is the distance from the loaded face of the component to the
reinforcing steel nearest the opposite face) is the critical shear section for reinforced
concrete and masonry components loaded on the opposite face from the support, as
discussed in Section 5-3 and 7-5. More accurate, less conservative, formulas for
determining the equivalent support shears at a distance d from the supports of two-way
spanning reinforced concrete and masonry components from Chapter 4 of TM 5-1300
are shown in Table 4-16. However, these formulas are significantly more complicated
and are not programmed into SBEDS. To apply these formulas use the following
procedure:
• Use (L-2d) in place of L in formulas for Equivalent Support Shears in Table 4-3 for
cases with two supports and uniform load
• Use (L-d) in place of L in formulas for Equivalent Support Shears in Table 4-3 for
cases with one support and uniform load
• No change in formulas for Equivalent Support Shears in Table 4-3 for cases with
concentrated loads
Use (x-d) in place of x and (y-d) in place of y in formulas for Equivalent Support Shears
in
4-64
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
• Table 4-4
Table 4-16 Ultimate Shear Stress at Distance d from Face of Support for Two-Way
Elements
Edge conditions Yield line location Limits Ultimate horizontal shear stress Limits Ultimate vertical shear stress
3ru (1 − d e / x)
2
3ru (1 − d e / H )(2 − x / L − d e x / HL)
0 < de/x < ½ 0 < de/H < ½
d e / x(5 − 4d e / x) d e / H(6 − x / L − 4d e x / HL)
3ru (1 − d e / x)
2
3ru (1 / 2 − d e / H )(1 − x / L − 2d e x / HL)
0 < de/x < ½ 0 < de/H < ¼
d e / x(5 − 4d e / x) d e / H(3 − x / L − 8d e x / HL)
In an SDOF analysis, the ultimate flexural resistance for components with combined
axial and lateral load is calculated based on a reduced ultimate dynamic moment
capacity that accounts for simultaneous stress from the applied dynamic axial load. The
reduced moment capacity is based on an applicable static interaction formula for
combined flexure and axial tension or compression load, where the axial load is equal to
the equivalent static reaction force from the supported component. The interaction
diagrams used in SBEDS are discussed for reinforced concrete and steel components
in Section 5-4 and Section 6-4, respectively. This procedure is conservative because it
treats the peak dynamic axial load, which actually has a short duration, as a static load.
A more accurate determination of the response of components with combined lateral
and axial load from blast can be obtained by using a multi-degree of freedom (MDOF)
model (e.g., finite element), which can directly account for the simultaneous interaction
of all affected elements.
The equivalent lateral load method is incorporated into SBEDS using Equation 4-25 for
the component types shown in Table 4-17. These component types are the most likely
to have significant axial loads. The user can always use the General SDOF component
type to include P-delta effects in the SDOF analysis for any component, where the user
calculates and inputs WF from Equation 4-25.The equivalent lateral load method was
tested against the moment magnifier method in the elastic load range in SBEDS by
applying uniform lateral load equal to 50% of the load causing first yield with a very long
rise time compared to the component natural period (i.e., essentially statically) and axial
load equal to 50% of the component axial load capacity to a W12x40 beam-column,
where the weak axis had continuous lateral support. The maximum deflection was
calculated in SBEDS with and without axial load and the deflection magnification was
compared to the moment magnifier.
Table 4-18 shows a comparison of results for a uniform load distribution, where the
moment magnifier method included Cm values as shown in the notes. SBEDS matched
the moment magnifier method well for all boundary conditions, particularly for simple
supports. The SBEDS calculations were repeated with a concentrated lateral load at
4-68
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
midspan and the calculated ratios of SBEDS magnification to the theoretical moment
magnifier were generally within 5% of those shown Table 4-18. This comparison
showed that the method in SBEDS is generally more approximate for point loads, but
not by a large amount. All moment magnifiers in both series of SBEDS calculations
were less than 1.8.
Comparisons were also made between the theoretical buckling load and the limit axial
load in SBEDS, where increasing values of axial load were input until a very small
lateral “blast” load with a very long rise time (i.e., a very small, essentially static lateral
load) caused lateral instability. These comparison ratios between buckling loads based
on theory and SBEDS analyses were 5% to 10% higher than the ratios in Table 4-18.
This is probably because the comparison in Table 4-18 involved axial loads
considerably less than the limit buckling load. Therefore, the SBEDS method may be
more conservative at higher axial loads. Theoretical buckling loads for two-way slabs
(Timoshenko and Gere, 1961) were compared to the limit axial load determined in
SBEDS as described previously in this paragraph. This comparison showed wide
results depending on the boundary conditions and aspect ratio, where SBEDS was
conservative in all cases by factors ranging from 1.16 to 3.0. This is due, at least in part,
to the simplified assumption that C in Equation 4-25 is equal to 0.64 for all two-way
components. This assumption can be improved with more study.
While these limited elastic response comparisons indicate that SBEDS is conservative,
it would be beneficial to perform additional comparisons of P-delta effects from dynamic
finite element analyses considering non-linear material and geometry effects to
comparable SBEDS calculations for both elastic and plastic response. This comparison
would possibly lead to an improved methodology in SBEDS to account for P-delta
effects, especially for two-way spanning components. It would also provide information
on the accuracy of the SBEDS approach during plastic response for one-way and two-
way components.
4-69
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
4-5 MAXIMUM RESPONSE PARAMETERS
The calculated maximum dynamic deflection of the equivalent SDOF system is used to
calculate two maximum response parameters in SBEDS, the maximum support rotation
and the ductility ratio. These two parameters are defined in Figure 4-26 and Equation
4-26, respectively. The portion of the span used to determine the support rotation for
one-way components in SBEDS is always one-half the span length, except for
cantilevers when it is the full span length. This is true for fixed-simple beams even
though the maximum deflection point for elastic response is not at midspan. For this
case, the maximum deflection is within approximately 15% of midspan and moves
closer to midspan during elastoplastic response. Since it is relatively complicated to
determine the exact location, the maximum deflection point is assumed at midspan
when determining the support rotation. The support rotation of two-way components in
SBEDS is based on the shortest distance from the supports to the yield line area with
maximum deflection, as determined from the calculated yield line pattern.
Figure 4-26 Support Rotation Angle
⎛ 2x ⎞
where θ = tan −1 ⎜ m ⎟
⎝ L ⎠
Equation 4-26
xm
μ=
xe
where: μ = ductility ratio
xm = maximum component deflection
xe = deflection causing yield of determinate component or yield of
equivalent elastic slope of indeterminate component (see
Figure 4-5)
SBEDS also compares the calculated maximum support rotation and the ductility ratio
to maximum support rotation and ductility criteria for a user defined building level of
protection (LOP) from the DoD and a user defined component framing type (i.e.,
secondary or primary component). The calculated response is acceptable only if both
the calculated maximum support rotation and ductility ratio are less than corresponding
maximum criteria values. The component framing type and building LOP are discussed
more in the next section. For some component types, only the calculated support
rotation or ductility ratio is compared to maximum support rotation or ductility criteria
from the DoD because only one of the two parameters correlates well to blast damage.
For non-DoD analyses, the user can directly input a limit support rotation and/or ductility
4-70
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
ratio. As stated in Equation 4-26, the ductility ratio for indeterminate components is
calculated in SBEDS based on the equivalent yield deflection, xE in
Figure 4-5. This is an average ductility ratio, which is less the ductility ratio at the initial
yield areas and greater than the ductility ratio at the final yield location.
Separate maximum allowable support rotations and/or ductility ratios are defined in
SBEDS for up to eight sub-categories for each component type (i.e., metal stud walls,
reinforced concrete slab, etc.). The sub-categories depend on specific factors for each
component type such as reinforcing steel ratio, tension membrane response, etc. The
user selects the most appropriate sub-category for the input component from a drop-
down list in SBEDS. They also select a primary or secondary component, as applicable
for the input component based on definitions in the notes of Table 4-19, and the
required building LOP. The required LOP for the building with the input component
must be determined separately using an approved security engineering approach.
SBEDS displays the maximum allowable support rotations and/or ductility ratios based
on this selections and criteria in PDC-TR 06-08 Revision 1.
The maximum acceptable ductility ratios and/or support rotations defining each
damage/performance level for all available component types and sub-categories are
maintained by the Protective Design Center (PDC) at the Omaha District, U.S. Army
Corps of Engineers, in the Response_limits.xls workbook. These maximum acceptable
response criteria are applicable for flexural, compression, and/or tension membrane
response, as noted in the sub-category definitions for each set of criteria. They are not
applicable for shear controlled response.
4-71
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Severe Damage
Below AT Greater than Greater than Majority of personnel in collapse
B3 to B4 Doors and windows will fail
standards B4 B4 Progressive collapse region suffer fatalities. Potential
6 catastrophically and result in lethal
1 Hazardous 5 5 likely. Space in and fatalities in areas outside of
Blowout Blowout hazards. (High hazard rating)
around damaged area is collapsed area likely.
unusable.
Heavy Damage Glazing will fracture, come out of
Majority of personnel in damaged
the frame, and is likely to be
area suffer serious injuries with a
B2 to B3 B3 to B4 B3 to B4 Onset of structural propelled into the building, with the
potential for fatalities. Personnel
Very Low 7 6 6 collapse: Progressive potential to cause serious injuries.
Heavy Hazardous Hazardous in areas outside damaged area
collapse is unlikely. Space (Low hazard rating). Doors may be
will experience minor to moderate
in and around damaged propelled into rooms, presenting
injuries.
area is unusable. serious hazards.
Superficial Damage
Less than B1 Less than B1 Less than B1 Glazing will not break. (No hazard
High 9 9 9 No permanent Only superficial injuries are likely.
Superficial Superficial Superficial rating) Doors will be reusable.
deformations. The facility
is immediately operable.
4-72
PDC TR-06-01 Rev 1
September 2008
Notes:
1. This is not a level of protection, and should never be a design goal. It only defines a realm of more severe structural response, and may provide useful information in some
cases.
2. Primary structural component can be identified as a member whose loss would affect a number of other components supported by that member and whose loss could
potentially affect the overall structural stability of the building in the area of loss. Examples include: columns, girders, and other primary framing components directly or in-
directly supporting other structural or non-structural members, and any load-bearing structural components such as walls are included.
3. Secondary structural component is a Non-load bearing infill wall components and any other structural component supported by a primary framing component.
4. Non-structural component can be identified as a member whose loss would have little effect on the overall structural stability of the building in the area of loss. Examples
include interior non-load bearing walls, overhead lights, heaters, and other mechanical or architectural items attached to building structural components.
5. Blowout: The component is overwhelmed by the blast load causing debris with significant velocities.
6. Hazardous failure: The component has failed, and debris velocities range from insignificant to very significant.
7. Heavy damage: The component has not failed, but it has significant permanent deflections causing it to be un-repairable. The component is not expected to withstand the
same blast load again without failing.
8. Moderate damage: The component has some permanent deflection. It is generally repairable, if necessary, although replacement may be more economical and aesthetic.
The component is expected to withstand the same blast load again without failing but the end state may be a lower level of protection.
9. Superficial damage: No visible permanent damage. The component is expected to withstand the same blast load and maintain the level of protection.
4-73
PDC TR-06-01 Rev 1
September 2008
SBEDS displays the maximum allowable support rotations based on this file located in
the same directory as the SBEDS.xls file for the user input building LOP, component
framing type (i.e., primary or secondary/non-structural), and component sub-category.
For a building not subject to DoD AT/FP design criteria, the user can refer to other blast
design and analysis references to determine a limit support rotation and ductility ratio for
given damage levels or blast design for given component types, including ASCE (1997),
TM 5-1300, and UFC 3-340-01. The documents are used primarily for design against
explosions in petrochemical facilities, explosive safety design, and design of military
hardened structures, respectively.
4-74
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Table 4-21 ASCE Component Damage Level Descriptions
Damage Level Description
High Component has not failed, but it has significant permanent
deflections causing it to be unrepairable
Medium Component has some permanent deflection. It is generally
repairable, if necessary, although replacement may be more
economical and aesthetic
Low Component has none to slight visible permanent damage
where:
ε'y = average strain rate for component material
εy = dynamic yield strain of component material
DIF = dynamic increase factor, which is a function of ε'y
fy = estimated actual static yield strength
E= Young’s modulus
time required for component deflection to equal yield deflection based
tE =
on calculated component response to applied blast load
Therefore, an iterative process is necessary to determine an accurate value for the DIF.
The strain-rate is recalculated based on tE from the second SDOF analysis and used in
4-75
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
the applicable DIF vs. strain-rate relationship to recalculate the DIF. If this recalculated
DIF is not equal to the DIF used in the second SDOF analysis within a tolerance of
approximately 5% to 10%, the whole process is repeated during a third SDOF analysis
using the recalculated DIF. Fortunately, the DIF is a relatively weak function of strain-
rate and usually either the default DIF value is acceptable, or only one iteration is
required to define an acceptable DIF. The equivalent static reaction force used to check
for shear failure and connection design increases with the component moment capacity,
and therefore with the DIF, so an unacceptably low DIF value is not conservative for
these parts of a blast design. If a component has more than one material, such as
reinforced concrete, separate strain rates and DIF values are determined for each
material based on their different dynamic yield strains and the same value of tE, which is
dependent on the overall component response.
SBEDS simplifies the process of determining or checking the DIF by calculating the
strain-rate for each component material, if there is more than one, for most component
types. The user can keep the default DIF recommended by SBEDS, or calculate a
more accurate DIF based on the output strain-rate and the applicable DIF vs. strain-rate
relationship. If the component is indeterminate, the time to yield, tE, in Equation 4-27
based on the equivalent yield deflection, xE,(see
Figure 4-5) in SBEDS. No DIF is calculated for open-web steel joists, since a default
DIF for this component is embedded in the methodology used to calculate the ultimate
flexural resistance. The user must input the strain to yield for the General SDOF
component.
4-7 CONNECTIONS
Typically, equivalent static reaction loads are used for design or analysis of connections
of blast-loaded members (see Section 4-3.2). In this approach, the ultimate flexural
resistance of the component is applied as an equivalent static load to determine the
maximum end reaction loads on the connections. The ultimate dynamic capacity of the
connection should be equal to, or greater than the equivalent static reaction load. The
ultimate dynamic connection capacity can be based on Load and Resistance Factor
Design (LRFD), using the applicable dynamic yield strength of the connection material
in place of the static yield strength. The applicable strength reduction factor (i.e., the ϕ
factor) from LRFD design should typically be included in the dynamic connection
strength. The DIF for structural bolts and welds, which typically have a high yield
strength greater than approximately 620 MPa (90,000 psi), is only on the order of 1.05.
The connection strength of components that are bolted or screwed to supports, such as
cold-formed steel, is often controlled by failure of the connected material around the bolt
or screw. Therefore, the overall connection strength calculations should also consider
the strength of the connected material in the same manner as static LRFD connection
design, where the dynamic yield strength of the connected material is used in place of
the static yield strength. Also, LRFD design methods are not available for some
connections, such as drilled anchors and screws, where manufacturers provide
allowable connection capacity information. In these cases, the dynamic connection
strength can be assumed equal to 1.7 times the allowable connection strength.
4-76
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Most of the observed failures of components subject to blast loads have occurred at
connections. When components are overloaded by blast loads, they often transition
from flexural response to tension membrane, or catenary, response even if this
response was not assumed for their design. The large reaction forces from tension
membrane response overload the connections and cause failure. Therefore, the most
cost-effective manner to increase the blast capacity of components is usually to design
extra strength into the connections, which can usually be done at a relatively low
incremental cost.
Combined axial and lateral load from component interaction is also modeled in an
equivalent SDOF system in an approximate, conservative manner, as described in
Section 4-4. Components that intersect at right angles and support each other can
have dynamic reaction forces that cause axial load in the supporting component, which
is also directly loaded by blast. Secondary P-delta moments can also occur in
components subject to simultaneous axial compression load and lateral blast load due
to the combination of axial load and lateral deflection. This can be modeled
approximately in a SDOF analysis as described in Section 4-4.1. Global structural
response that involves more than one blast-loaded component, such as frame sway
response, can also be analyzed in an approximate manner as an equivalent SDOF
system if a mode shape is assumed. The response of components that make up the
frame responding between their supports to direct blast load and their response as part
as frame sway must be combined in a conservative manner. Therefore, many types of
component interaction can be modeled in an approximate, conservative manner in
SDOF analysis if the dynamic interactions are recognized and understood by the user.
Limitations Due to Blast Load Assumptions Finally, it can be very difficult to model
component response to blast loads with complex spatial load distributions using a
SDOF analysis. Blast loads that vary significantly over the area of a structural
component, such as close-in explosions and confined explosions with several nearby
reflecting surfaces, must be converted into an “equivalent” blast load that is spatially
uniform over the whole area of the component at each time step because the load factor
in SDOF analyses almost always assumes spatially uniform loading. The process of
calculating equivalent uniform loads usually involves simplifying assumptions regarding
the deflected shape of the component. Additionally, blast loads with very non-uniform
spatial distributions can excite higher mode shapes that are not accounted for in the
SDOF analysis. Blast loads with significant spatially non-uniformity can be modeled
directly with dynamic finite element analyses.
4-78
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
In summary, equivalent SDOF systems can be used to model most blast-loaded
components if complexities associated with the limitations of the SDOF approach are
accounted for in a conservative manner. Alternative approaches include the use of
equivalent multi-degree-of-freedom spring-mass system analyses and dynamic finite
element analyses. Many commercially available finite element programs have upgraded
dynamic analysis capabilities to include modeling geometric and material non-linearity
with user-friendly input schemes. Engineering judgment and experience is usually
necessary, however, to interpret the results in a meaningful manner for a particular
problem and verify that all the relevant blast load and response issues for the given
problem are included in the analysis. Also, analyses with these more complex analysis
tools are more time consuming. Often a SDOF based analysis can provide an engineer
with an understanding of the most relevant response issues involved in a complex
problem, which may help to set up a finite element analysis, and can provide an upper
bound on the dynamic response. A finite element analysis, on the other hand, can
provide more precise results that do not involve nearly as many user assumptions
regarding the component response.
4-79
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Figure 5-1 from UFC 3-340-01 shows the empirical relationship between strain-rate and
the dynamic increase factor (DIF) in Equation 5-1 for the ultimate unconfined concrete
compressive strength. The strain-rate in Figure 5-1 should be based on a concrete
5-1
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
strain to yield of 0.002 and the time to yield of the concrete component caused by the
blast load of interest as described in Section 4-6.
Figure 5-1 Strain-Rate vs. Concrete Compression Strength
Figure 5-2 from UFC 3-340-01 shows the effect of strain-rate on the concrete splitting
tensile strength. This shows very little increase in tensile strength at the recommended
default strain-rates of 0.1 to 0.3 sec-1. However, the tensile strength DIF increases
rapidly with strain-rate at higher strain-rates. The tensile strength DIF in Figure 5-2 can
be used to calculate concrete dynamic shear according to UFC 3-340-01. TM 5-1300
more conservatively recommends that the no DIF be used to calculate the concrete
dynamic shear strength.
5-2
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Figure 5-2 Strain-Rate vs. Concrete Splitting Tensile Strength
Figure 5-3 shows the empirical relationship between strain-rate and the DIF for yield
stress (fy) and ultimate stress of steel (fu) conforming to ASTM A615 Grades 40 and 60.
Default DIF from Figure 5-3 based on strain-rates of 0.1 sec-1 and 0.3 sec-1 for far-range
and close-in blast loading discussed in Section 5-1 can be used in Equation 5-2. These
default DIF values are 1.22 and 1.3 for far-range and close-in blast loading,
respectively, for Grade 60 reinforcing steel. Based on a slightly different relationship
between strain-rate and DIF, TM 5-1300 recommends a default DIF value of 1.17 for
far-range loading that is also used often for blast design. Larger default DIF values
based on these two strain-rates can be determined from Figure 5-3 for Grade 40
reinforcing steel.
5-3
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Almost all reinforcing steel placed in the United States after approximately 1970 is
Grade 60. Prior to this, Grade 40 was used extensively. Welded wire fabric (WWF) is
used extensively in reinforced concrete slabs. Dynamic yield strength information for
conventional reinforcing steel used in SBEDS is summarized in Table 5-1. WWF has a
low DIF value per TM 5-1300. Table 5-2 has cross sectional area information for WWF
in English units from ACI 318-95.
Prestressed concrete is typically reinforced with conventional reinforcing steel and high
strength prestressing cables that are Grade 250 (1723 MPa or 250,000 psi) or Grade
270 (1860 MPa or 270,000 psi). Ke and DIF values equal to 1.0 are typically assumed
for these steels. Table 5-3 contains cross sectional area information for steel
prestressing strands from ACI 318-95. Most concrete is prestressed with seven-wire
strand.
Figure 5-3 Strain-Rate vs. Reinforcing Steel Yield and Ultimate Strength
Table 5-2 Cross Sectional Area Information for Welded Wire Fabric (WWF)
5-5
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
shown in Equation 5-4 based on a resisting moment arm dc between the compression
and tension face steel and a resisting force equal to the yield force of the lesser of the
steel areas at each face.
Table 5-3 Cross Sectional Area Information for Steel Prestressing Strands
Equation 5-3
As f dy ⎛ A f ⎞
M du = ⎜⎜ d − s dy ⎟⎟
b ⎝ 1.7bf ' dc ⎠
where:
Mdu = dynamic ultimate moment capacity per unit width
As = area of tension reinforcement in width, b
fdy = dynamic design stress
d =effective depth of Type I cross section (distance from the extreme
compression fiber to the centroid of the tension reinforcement – see
Figure 5-4)
b = width of rectangular section
f'dc = concrete dynamic compressive strength
5-6
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Equation 5-4
As ' f dy
M du _ II = (h − d ct − d cb )
b
where:
Mdu_II = dynamic ultimate moment capacity per unit width for Type II cross
section
As’ = lesser of the top or bottom face reinforcing steel areas in width, b
fdy = dynamic design stress
h = cross section thickness
dcb = cover depth to bottom face reinforcing steel
dct = cover depth to top face reinforcing steel
Type II cross sections must be used for blast design of DoD buildings according to UFC
3-340-01 and TM 5-1300 for response with a support rotation greater than 2 degrees.
The assumption that a Type II cross section develops at 2 degrees of support rotation is
based primarily on static tests and is intended to be a design conservative approach for
hardened blast resistant structures with relatively high reinforcing ratios. It implies
immediate failure of singly reinforced concrete components at 2 degrees of support
rotation. This does not occur in blast tests on singly reinforced concrete slabs and
reinforced masonry walls, especially for lower reinforcing ratios used in typical
conventional design (Oswald, 2005).
The ultimate dynamic moment capacity for reinforced concrete slabs and beams in
SBEDS can be based on Equation 5-3 without any consideration of a Type II cross
section, even at support rotations greater than 2 degrees. This is considered the most
realistic approach for blast analysis of conventionally constructed concrete components.
Alternatively, SBEDS has a user option so that it will to transition to a Type II cross
section at support rotations greater than 2 degrees. In this case, the ultimate resistance
in the resistance-deflection relationship transitions in SBEDS from a resistance based
on the Type I moment capacity to a resistance based on the Type II moment capacity
5-7
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
over a deflection range from 2.0 to 2.2 degrees of support rotation. There is no
theoretical basis for this 10% transition range. It is simply used because it is realistic to
assume some transition range and a gradual transition is preferable for computer
programming purposes. The resistance-deflection relationship for a component with
compression membrane response transitions directly to an ultimate flexural resistance
based on a Type II moment capacity at deflections greater than those causing
compression membrane response.
Equation 5-6
⎧⎪ γ p ⎡ ⎤⎫
( p − p')⎥ ⎪⎬
f pu df dy
f ps = f pu ⎨1 − ⎢pp +
⎪⎩ β 1 ⎢⎣ f ' dc d p f ' dc ⎥⎦ ⎪⎭
A A' A ps
where p = s p' = s p p =
bd bd bd p
where:
effective prestressing tendon strength at maximum moment capacity
fps =
specified ultimate tensile strength of prestressing tendon
fpu =
γp =
factor for type of prestressing tendon
0.40 for fpy /fpu ≥ 0.80
=
0.28 for fpy /fpu ≥ 0.90
=
ß1 =
0.85 for f'dc up to 4000 psi and is reduced 0.05 for each 1000 psi in excess of
4000 psi
ρp = Prestressed reinforcement ratio
ρ = reinforcing ratio of non-prestressed tension reinforcement
ρ' = reinforcing ratio of non-prestressed compression reinforcement
A’s = total area of non-prestressed compression reinforcement within width b
d’ = distance from extreme tension fiber to the centroid of the non-prestressed
compression reinforcement
See Equation 5-5 for additional parameter definitions
Equation 5-7
f pu df dy
pp + ( p − p') ≥ 017
.
f ' dc d p f ' dc
where:
fse = effective stress in prestressed reinforcement after allowances for all
prestress losses (assumed equal to 60% of fpu in SBEDS)
L = span length
Note: See Equation 5-5 and Equation 5-6 for additional parameter definitions
5-9
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Equation 5-9 must be satisfied to insure against sudden compression failure of
rectangular beams and slabs and of flanged beams where the thickness of the
compression flange is greater than or equal to the depth of the equivalent rectangular
stress block.
Equation 5-9
p p f ps df dy
+ ( p − p') ≤ 0.36β1
f ' dc d p f ' dc
Equation 5-5 through Equation 5-8 are used in SBEDS to determine the ultimate
dynamic moment capacity of prestressed beams and slabs. Equation 5-9 is used in
SBEDS to check that the ultimate moment capacity is controlled by yielding of the
prestressing steel rather than crushing of concrete. It is always assumed in SBEDS that
the thickness of the compression flange is greater than or equal to the depth of the
equivalent rectangular stress block, as is almost always the case in conventional
design. This assumption is checked and the ratio of compression block thickness to
flange thickness is printed out near the bottom of the left side of the input form for
flanged beam input. Ideally this ratio should not be greater than 1.0 for Equation 5-5 to
be applicable. However, as long as this ratio is no more than 1.2, the assumptions in
SBEDS do not typically cause significant error in the ultimate moment dynamic capacity.
2
where: Ia = average moment of inertia per unit width of reinforced concrete cross
section
Ig = moment of inertia per unit width of uncracked gross concrete section
(neglect reinforcement)
Ic = moment of inertia per unit width of cracked reinforced concrete section
The moment of inertia of the cracked concrete section, Ic, is based on the concrete
compression area and the transformed steel areas computed about the centroid of the
5-10
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
transformed section. The cracked moment of inertia per unit width of a slab or a beam
with a rectangular cross section is given by Equation 5-11.
Equation 5-11
I c = Fd 3
where:
Ic = moment of inertia per unit width of cracked reinforced concrete section
F = coefficient for cracked moment of inertia
d =distance from the extreme compression fiber to the centroid of the
tension reinforcement (use average value for multiple maximum
moment regions)
The coefficient, F, is can be calculated with Figure 5-5 for singly reinforced concrete
cross sections, with reinforcement on one face only, and for doubly reinforced cross
sections with equal reinforcement in both faces. The coefficient depends on the modular
ratio as given by Equation 5-12.
Figure 5-5 Coefficient for Moment of Inertia of Cracked Sections
For prestressed elements the moment of inertia of the cracked section may be
approximated by Equation 5-13, where n is given in Equation 5-12. This cracked
moment of inertia is used in Equation 5-10 with the gross moment of inertia to
determine the average moment of inertia. The average moment of inertia is used to
5-11
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
calculate the spring stiffness for the equivalent SDOF system representing the
component. See Equation 5-5 and Equation 5-6 for definitions of the terms in Equation
5-13.
Equation 5-13
[
I c = nA ps d p 1 − p p
2
( )
1/ 2
]
Equation 5-10 is used in SBEDS to determine the moment of inertia of conventionally
reinforced concrete components. Ic for conventionally reinforced concrete components
is always based on F in Figure 5-5 for singly reinforced cross sections. This approach is
conservatively used for all cross sections, regardless of the amount of compression
reinforcement. Ic for prestressed reinforced concrete components is calculated with
Equation 5-13. In both cases the gross moment of inertia is based only on the concrete
cross section, not including the transformed steel area. This is simple, conservative
approach that typically underestimates the actual gross moment of inertia by a small
amount. The gross moment of inertia of a prestressed I-beam is calculated based only
on the assumption the section is a T-beam with a compression flange because SBEDS
does not consider an I-beam.
The modulus of elasticity for concrete in dynamic response to blast loads is assumed
identical to that for static loads. The formula for the concrete modulus of elasticity in
SBEDS is shown in Equation 5-14.
Equation 5-14
E c = 0.043w 1.5 f' c Metric
E c = 33w 1.5 f' c English
where:
Ec = concrete modulus of elasticity, MPa (psi)
w = concrete unit weight, kg/m3 (lb/ft3)
f’c = concrete compression strength, MPa (psi)
The depth, d, used for shear calculations in SBEDS, which equals the thickness minus
the cover distance to the tensile reinforcing steel, is illustrated in Figure 5-4 for a Type I
cross section. It is based on the concrete cover depth to the top layer of reinforcing steel
at the supports of one-way spans, except for a simply support span where it is based on
the concrete cover depth to bottom steel at midspan. For two way spans, the above
approach is used except the depths in both span directions are averaged. If the two-way
span is fixed on two opposite sides and simply supported on two opposite sides, then
the depth is based on the cover distance to the top steel only in the direction of the fixed
span. This is summarized in Table 5-4. The intent is to use the depth at the maximum
moment region nearest the highest shear area.
Figure 5-6 Locations of Critical Sections for Diagonal Tension Shear
Table 5-4. Depth to Reinforcing Steel Used by SBEDS for Shear Strength
Calculations
Boundary Condition Depth (d) Used for Shear Strength Calculations
One-way, Cantilever d at support
One-way, Fixed-Fixed d at supports
One-way, Fixed-Simple d at fixed support*
One-way, Simple-Simple d at midspan
Two-Way, All Edges Simply Average d at midspan in both directions
Supported
Two-Way, All Edges Fixed Average d at supports in both directions
Two-Way, Opposite Edges d at fixed supports *
Fixed and Simply Supported
*Highest shear load is at fixed support
If the diagonal shear capacity, as discussed in the next sections of this chapter, is less
than the equivalent static reaction force at the critical location, SBEDS calculates the
required area of shear reinforcing steel so that the sum of the concrete and steel shear
strengths is equal to the equivalent reaction force. If this amount of reinforcing steel is
not provided (which may occur in an existing component), the component is subject to
shear yielding when resistance equals the ultimate shear resistance, as discussed in
5-13
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Section 4-1.2.2.4, which is less than the ultimate flexural resistance used in the SDOF
analysis of the component. The can account for shear yielding using the “shear flag” in
SBEDS as discussed in Section 4-1.2.2.4.
SBEDS also calculates the direct shear capacity of reinforced concrete components and
compares it to the equivalent static reaction at the support. The user should refer to
Chapter 4 of TM 5-1300 (1990) for calculating the required area of diagonal reinforcing
bars that can be used at the supports if the equivalent static reaction force exceeds the
direct shear capacity of the concrete. SBEDS does not calculate this steel area, which is
rarely needed for components subject to external explosions.
The equivalent static reaction force in SBEDS is always based on the ultimate flexural
capacity, which may be greater than the maximum calculated resistance for a given
blast load. In this case, the direct and diagonal shear capacity checks in SBEDS, which
are based on the ultimate flexural capacity, are conservative. However, it always good
design practice to provide shear capacities that are at least equal to the equivalent
static reaction force based on the ultimate flexural capacity of a blast-loaded component
to prevent less ductile, shear controlled response if the blast load exceeds the design
blast load.
vc =
DIF
(K e K a f ' c )0.5 for shear and flexure
6
DIF ⎛⎜ N u ⎞⎟
vc =
⎜
1+
⎟
(K e K a f ' c )0.5 for shear , flexure and axial load
6 ⎝ 14 Ag ⎠
5-14
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
where:
vc = shear stress capacity provided by the concrete
DIF = dynamic increase factor for concrete in splitting tension
(see Figure 5-2)
Ke = static strength increase factor (see Equation 5-1)
Note: Ke also referred to as an average strength factor (ASF)
Ka = concrete age increase factor (see Equation 5-1)
f'c = concrete static compressive strength
Nu = axial load normal to the cross section (tension negative)
Ag = gross area of the cross section
The concrete tensile DIF inEquation 5-15 is a function of strain-rate, and therefore must
be defined based on the default strain-rates or the iteration process described for the
concrete compression DIF in Section 5-1.1. The concrete tensile DIF is very sensitive to
strain-rate at higher strain-rates, as shown in Figure 5-2, and it may be worth the effort
to perform the iterative process for some cases. TM 5-1300, which is an older and more
conservatively based blast design document, recommends a DIF of 1.0 and no increase
in the static strength when calculating concrete shear strength. Therefore, the dynamic
concrete shear strength equals the static shear strength. This is a much more
conservative approach than Equation 5-15, which can result in 10% to 30% less
calculated concrete strength, or more for very high strain-rates.
where:
Vc = concrete dynamic diagonal shear force capacity (per unit supported
width for slabs)
vc = diagonal shear stress capacity provided by the concrete (MPa)
f’dc= dynamic concrete compressive strength from Equation 5-1
b = beam width
d = effective depth for Type I cross section in Figure 5-4 at supports or
midspan based on boundary conditions (see Table 5-4)
5-15
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
5-3.2 DYNAMIC DIAGONAL SHEAR STRENGTH OF PRESTRESSED
CONCRETE
Under conventional service loads, prestressed elements remain almost entirely in
compression, and hence are permitted a higher concrete shear stress than non-
prestressed elements. However at ultimate loads the effect of prestress is lost and
therefore concrete shear capacity should be calculated in the same manner as normally
reinforced concrete. Therefore, the shear strength of prestressed concrete is calculated
in SBEDS using Equation 5-16.
In TM 5-1300, it is recommended that the loss of the effect of prestress also means that
depth, d, is the actual distance to the prestressing tendon and is not limited to 0.8h as it
is in the ACI 318 code. The distance d is greatly reduced at supports of a prestressed
component with draped tendons. However, draped tendons also make it difficult to
properly anchor shear reinforcement at the supports. Thus it is recommended that only
straight tendons be used for blast design. The depth in Equation 5-16 is conservatively
based on the greater of the depth to any conventional reinforcing steel or the depth to
the prestressing steel.
Equation 5-17
V s = V max − Vc
where:
Vs = required shear force capacity provided by shear reinforcement
Vmax = equivalent static reaction forces at critical span locations
Vc = concrete dynamic shear force capacity from Equation 5-16
5-16
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Equation 5-18
Av V
= s for beams
s f dy d
Av V
= s for slabs
bs f dy d
where:
Vs = reinforcing steel shear force capacity from Equation 5-17 in units of
force/width along support for slabs and total force for beams
Av = area of required shear reinforcement within reinforcement spacing, s,
for beams and within reinforcement spacings, s and b, for slabs
b = spacing of slab shear reinforcing in direction along support
s = spacing of shear reinforcing along span direction (maximum of d/2)
d = effective depth of flexural reinforcing steel for Type I cross section in
Figure 5-4
fdy = shear reinforcement dynamic yield stress from Equation 5-2
(Note: Interior stirrup and slab shear reinforcement only require 1350 bend on tension side and
900 bend on compression side as a minimum)
The maximum spacing along the span direction, s, for shear reinforcement to be
considered effective is d/2, which guarantees that at least one stirrup crosses a
hypothetical shear crack at a 450 angle anywhere along the span. In many cases
column tie stirrups spaced only to prevent buckling of longitudinal steel are not effective
as shear reinforcement because of this requirement. Also, the shear reinforcement in
slabs is always placed at a spacing, b, along the supports equal to the flexural
reinforcing steel spacing. In most cases, this means that any shear reinforcement in
two-way slabs will only be effective if the flexural reinforcement spacing is limited to d/2.
This may require using more flexural bars with a smaller area in a blast design than
would otherwise be used.
5-17
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
5-3.4 DIRECT SHEAR
Direct shear failure of concrete is characterized by the rapid propagation of a crack
through the depth of the component perpendicular to the span direction. The crack
usually occurs at the peak dynamic shear location, which is at the face of a support,
very early in the component response before significant flexural motion occurs.
Therefore, a Type I cross section is assumed for direct shear stress. The direct shear
capacity at the support face of a monolithic concrete joint is calculated as shown in
Equation 5-19 from UFC 3-340-01. This is a slightly unconservative (by about 15%) for
cleaned and roughened joints. The concrete direct shear capacity for the unusual case
of an unroughened joint is conservatively assumed equal to zero. SBEDS always uses
Equation 5-19 to calculate the direct shear capacity, since the roughness of the joint is
not an input. If necessary, the user must modify the value calculated by SBEDS.
Equation 5-19
Vd = vdc d for slabs, Vd = vdc bd for beams
v dc = 0.16 f ' dc
where:
Vd = concrete dynamic direct shear force capacity (per unit supported
width for slabs)
vdc = direct shear stress capacity provided by the concrete
f’dc= dynamic concrete compressive strength from Equation 5-1
b = beam width
d = effective depth for Type I cross section in Figure 5-4
If Vd > Vmax (i.e., the equivalent static shear force), shear friction reinforcement is
required such that the combined direct shear capacity of concrete and shear
reinforcement at least equals Vmax. The direct shear capacity of concrete with shear
friction reinforcement, Vdu, is given by Equation 5-20 from UFC 3-340-01, where the
contribution of the concrete depends on the type of joint between the component and
the support, as indicated in Table 5-5. The shear friction reinforcement area is in the
span direction perpendicular to the plane of the face of the support (i.e., the same as
the flexural reinforcement). All reinforcement crossing the crack plane, except that
which is required to resist net tension in the component, is allowed to serve as shear
friction reinforcement. This includes flexural bars. Usually the flexural steel and concrete
provide sufficient direct shear resistance. In the cases where they do not, diagonal bars
may also be used as shown in Figure 5-8. Only the area of these bars acting
perpendicular to the direct shear crack, equal to (Assinα), is included as Avf in Equation
5-20. Equation 5-20 is not calculated in SBEDS. The user must perform this check in
the unusual case when SBEDS output indicates that Vd from Equation 5-19 is less than
Vmax.
Equation 5-20
Vdu = v du d for slabs , Vdu = v du bd for beams
K2
(N c + A vf f dy ) ≤ K 3 f ' dc
v du = K 1 f ' dc +
d
where: Vdu = concrete dynamic direct shear force capacity (per unit supported
width for slabs) including effects of reinforcing steel
5-18
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
where:
At = area of axial tension reinforcement per unit width
Nt = axial tension force per unit width, taken as the maximum equivalent
support reaction force of the intersecting member
fds = dynamic design stress for the member carrying axial tension, based
on its flexural response
For blast resistant design, At can be incorporated by adding it to the quantity of flexural
bars equally in each face or by placing separate longitudinal bars at the middepth of the
component. It is more conservative to place separate tension steel at middepth since
this will ensure that tensile steel does not contribute to the dynamic moment capacity of
the component and cause higher than expected equivalent static reactions. If the area
of flexural steel is fixed, as in an existing component, At from Equation 5-21 can be
subtracted from the flexural steel that is continuous across the component and the
ultimate dynamic moment capacity can be determined based on the remaining steel
area. At can be subtracted from the continuous flexural steel at each face of doubly
reinforced components in proportion to the steel areas at each face.
SBEDS does not directly consider the case where applied axial tension forces act with
lateral blast loads. The user can subtract the area At from the flexural steel of an
existing component to determine a reduced flexural steel input into SBEDS or
separately calculate At for a designed component. The shear strength in SBEDS is
calculated without accounting for any tension forces.
As shown in Figure 5-9, the available moment capacity with axial load increases initially
at lower axial loads up to Pb, the balanced axial load, and then decreases as the axial
5-20
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
load approaches the ultimate axial load capacity, Pdo. Even though more moment
capacity is available at axial load up to Pb, the axial load causes a reduction in the
maximum support rotation causing failure. Therefore, axial load may reduce the
available strain energy to failure, and therefore the maximum blast load that can be
resisted compared to the same component without axial load.
Figure 5-9. Reinforced Concrete Beam-Column Axial Load Moment Capacity
Interaction Diagram
5-21
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
nonlinear strain distribution develops through the cross section. A laterally unrestrained
beam or slab can be considered deep when its effective span-to-thickness ratio, Le/h, is
less than 2 for simple supports, or less than 2.5 for continuous support conditions. The
effective span length, Le, is taken as the smaller of the center-to-center distance
between supports, or 1.15 times the clear span. In the case of two-way slabs, Le is the
short span direction. A laterally restrained concrete member is considered a deep beam
when its clear span-to-thickness ratio, L/h, is less than five.
The lateral load capacity of deep beams is usually controlled by the shear capacity of
the member. The shear capacity can be designed to exceed the flexural capacity, but
this will not necessarily improve the ductility of a deep beam significantly because
crushing of the concrete after yielding of the reinforcement in flexure occurs at much
lower support rotations in deep beams than for components with conventional span-to-
height ratios. This is due to enhanced compression strains from arching action. The
moment capacity and shear capacity of deep beams can be calculated as shown in
Chapter 10 of UFC 3-340-01. SBEDS does not consider deep beams, which are not
commonly used in conventional construction.
Mechanical splices that have test data showing they will develop the full ultimate
strength of the reinforcing bars without reducing their ductility may be used according to
TM 5-1300. Splices that develop the full ultimate strength of reinforcing steel have been
used for precast concrete frame buildings in earthquake zones and are commercially
available. Use of such splices can greatly reduce congestion at corners of reinforced
5-22
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
concrete components, which can be a very serious problem for components that are
designed to resist high blast loads. This is especially true for internal explosions, which
cause combined tension and flexure in the wall and roof components. Also, the
concrete plasticity during placement and maximum aggregate size must also be
considered in such cases with congested reinforcement. The use of a maximum
aggregate size of 6 mm (0.25 inches) is recommended.
5-8 REBOUND
Rebound causes a reversal of the component stresses (i.e., tension stress regions
become compression regions and vice versa) and a reversal in the direction of
connection forces (i.e., into the building and outward from the building) as discussed in
Section 4-1.2.5. Stress reversals are a concern for conventionally reinforced concrete
components because little or no steel is often designed for areas that are in
compression under gravity loading. Therefore, rebound can cause tension in areas with
insufficient steel reinforcement or no steel reinforcement. Also, it can cause tension in
steel that is spliced assuming the surrounding concrete is in compression rather than
tension. The SDOF response of reinforced concrete components in rebound is
calculated in SBEDS with a similar approach as inbound response except that the
flexural resistance is based on the input steel area for rebound, assuming this steel can
develop its full dynamic tensile yield strength. Reinforced concrete components
designed to resist blast load are typically symmetrically reinforced at both faces with all
steel reinforcement splices designed assuming the splice is in tension.
The moment of inertia of reinforced concrete during rebound is not well understood, as
discussed in Section 5-2.3. For lack of a better, validated approach, the moment inertia
during rebound is assumed equal to that during inbound response in SBEDS.
5-9 CONNECTIONS
Reinforced concrete components can be cast-in-place or precast at an off-site location
and then connected together at the building site. Cast-in-place reinforced concrete
components have no exterior connections between intersecting components. The cross
sections at the connections must have adequate strength to resist direct and diagonal
tension shear forces. The connections must develop the forces from the ultimate
dynamic moment capacity, if it is a moment connection, and the tension forces in
tension steel used to resist any axial tension forces from connected components.
Section 10.7 of UFC 3-340-01 has typical details for connections in monolithic blast-
resistant concrete connections.
Bearing connection (tension during rebound) Connection in shear and tension during rebound
The connection in Figure 5-10a may be required to take upward shear in the plane of
the wall against the rebound reaction force from a supported roof, and out-of-plane
shear if the wall acts as a shear wall, in addition to tension during rebound. Shear forces
acting in different directions can be combined into a resultant shear force. In a more
detailed design, the relative timing of the different forces on a connection may be
considered so that maximum values are not simultaneously assumed for all the forces.
Precast walls are also connected together along their sides. If the walls are designed to
act as together as a monolithic shear wall, heavy structural connections are normally
required, as shown in Figure 5-11.
5-24
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Figure 5-11 Example Connection Between Precast Concrete Wall Panels (from
ASCE)
Moment-resisting connections have been used for precast reinforced concrete framing
members used in high earthquake zones and therefore can be used between blast-
loaded precast concrete members, if necessary. The connection must transfer a tension
load based on the full dynamic yield strength of the flexural reinforcing steel in the
precast component, as well as a shear reaction force based on the component’s
maximum equivalent static reaction force. The fact that full stress reversal may occur
during rebound must also be considered. Ideally, any moment-resisting components
between precast components should be placed at low stress locations. Moment-
resisting precast concrete frames in high earthquake areas have been connected in this
manner, where precast columns are cast with protruding tees that extend out to the
inflection point in the moment distribution of the beam span. If this is not the case, the
limited ductility of the connection may control the allowable support rotation for the
component, rather than the ductility of the reinforced concrete cross section at
maximum moment regions.
5-25
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
5-26
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Figure 6-1 shows the empirical relationship between strain-rate and the dynamic
increase factor (DIF) in Equation 6-1 for several structural steels. The strain-rate in
Figure 6-1 should be based on a steel dynamic yield strain and the time to yield of the
steel component caused by the blast load of interest. This involves an iterative process
described in Section 4-6, because the time to yield and dynamic yield strain are based,
in part, on an assumed value for the DIF. However, according to guidance in TM 5-
1300 strain-rates of 0.01 sec-1 and 0.03 sec-1 can be assumed for secondary
components (i.e., those directly subject to blast load) that are subject to far-range blast
loads (i.e., scaled standoff > 1.2 m/kg1/3) and close-in blast loads, respectively. A
smaller strain-rate should be assumed in each case for components subject to a filtered
dynamic load from the reactions of supported components (e.g., columns), as discussed
in Section 5-1.1.
Table 6-1 shows recommended default values for the DIF based on an assumed strain-
rate of 0.01 sec-1, and the average static strength increase factor (SIF), for different
types of structural steel components commonly used in building construction from TM 5-
1300. As shown in Table 6-1, DIF and SIF values both reduce with increasing yield
strength. This occurs because steel manufacturers minimize the additional strength
provided over the minimum specified strength for more expensive high yield strength
steels and testing shows that the DIF decreases with increasing yield strength, as
shown in Figure 6-1. The use of cold-formed steel with a yield strength greater than
400 MPa (60,000 psi) is not recommended for blast design because the higher yield
strength is usually attained by additional rolling and therefore more strain hardening of
the steel, which significantly reduces its ductility.
6-27
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Figure 6-1 Strain-Rate vs. Structural Steel Yield Strength
Table 6-2 shows recommended values for the SIF and DIF for different types of
aluminum components commonly used in window frame construction. A great deal of
research conducted on the strain-rate sensitivity of aluminum for aircraft and spacecraft
research in the range of 1e-4 sec-1 up to 100 sec-1indicates that the aluminum yield
strength is essentially independent of strain-rate (DOE/TIC 11268, 1992). This is
reflected in the very low value for the DIF in Table 6-2 from ASCE (1997). The values
for the SIF in Table 6-2 were tabulated by dividing typical tensile yield strengths at 750F
provided by the Aluminum Association (1993) by corresponding minimum specified yield
strengths.
6-28
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Table 6-2 Material Strength Increase Factors for Aluminum
Type of Aluminum Yield Strength MPa SIF (a)* DIF (c)
6061 T6 241 (35,000 psi) 1.07 1.02
6063 T5 110 (16,000 psi) 1.16 1.02
6063 T6 172 (25,000 psi) 1.12 1.02
*Also referred to as an average strength factor (ASF)
High shears can have a significant effect on moment capacity of components with
rectangular cross sections. When the applied shear does not exceed 55% of the
ultimate dynamic shear capacity (i.e., 0.55Vdu), the effect of shear on moment capacity
is less than 10 percent and can be ignored. The ultimate dynamic moment capacity for
rectangular plate sections with an applied shear greater than 0.55Vdu should be
multiplied by the reduction factor Kmv in Equation 6-3. Equation 6-3 is not accounted for
in SBEDS because it rarely applies for typical structural plates. For a plate with fixed
ends (i.e., the most conservative case), the span-to-thickness ratio must be less than
6.0 for Kmv in Equation 6-3 to cause more than a 10% reduction in moment capacity.
This is an unusual case for conventionally designed building components.
6-29
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Figure 6-2 Typical Moment-Curvature Relationships for Steel Plates and Beams
Equation 6-2
⎧ ⎛S+Z⎞ ⎫
⎪ f dy ⎜ ⎟ μ ≤ 7⎪
M du =⎨ ⎝ 2 ⎠ ⎬
⎪ f dy Z μ > 7⎪⎭
⎩
where:
Mdu = ultimate dynamic moment capacity per unit width
fdy = dynamic yield strength of steel or aluminum plate
S =elastic section modulus per unit width
Z =plastic section modulus per unit width
μ= ductility ratio
Equation 6-3
4
⎛ V ⎞
K mv = 1 − ⎜⎜ ⎟⎟
⎝ Vdu ⎠
where:
Kmv = moment reduction factor for steel plates with high shear
V =applied shear force per unit width at the section in question (See
equivalent static reaction load in Section 4-3.2)
Vdu = dynamic ultimate shear capacity per unit width for steel plates (See
6-5)
6-30
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
6-2.2 ULTIMATE DYNAMIC MOMENT CAPACITY OF HOT ROLLED BEAMS
The AISC Manual of Steel Construction divides structural steel shapes into non-
compact and compact sections based on the width-to-thickness ratios of the beam
flanges and the height-to-thickness ratios of the beam web. Only compact sections are
generally recommended for blast design. Almost all structural steel tube sections and I-
beams with depths greater than 200 mm (8 inches) are compact members. Compact
beams with well-braced compression flanges reach fully plastic moment capacity at
lower ductilities than plates, as shown in Figure 6-2. Little error (< 3 percent) is obtained
by assuming that the resisting moment increases linearly with curvature up to the full
plastic moment and then stays constant with increasing curvature for I-beams,
channels, and tube sections responding with a ductility ratio of least three. Thus, the
ultimate dynamic moment capacity for well-braced, compact beams is given by
Equation 6-4. A ductility ratio greater than 3 is almost always acceptable for design and
typically implies limited blast damage. Therefore, the moment capacity for beams in
SBEDS is always calculated with the plastic section modulus in Equation 6-4 assuming
μ>3. The ratio between the two formulas for moment capacity in Equation 6-4 is less
than 10% for most beam cross sections, based on a typical ratio of Z/S equal to 1.15.
Equation 6-4
⎧ ⎛S+Z⎞ ⎫
⎪ f dy ⎜ ⎟ μ ≤ 3⎪
M du =⎨ ⎝ 2 ⎠ ⎬
⎪ f Z μ > 3⎪
⎩ dy ⎭
where:
Mdu = ultimate dynamic moment capacity per unit width
fdy = dynamic yield strength or steel or aluminum beam
S =elastic section modulus per unit width
Z =plastic section modulus per unit width
μ = ductility ratio
The compression flanges of beams must be braced adequately to resist lateral and
torsional displacements at plastic hinge locations. Such lateral bracing should be
provided at expected plastic hinge locations (i.e., maximum moment locations) and at
adjacent points spaced at no more than the minimum unbraced length for the beam.
According to UFC 3-340-01, points of zero moment (inflection points) may be
considered as braced points for blast design. Spacing between lateral bracing points for
beams bent about their strong axis should not exceed that given by Equation 6-5 in
order for the full plastic moment capacity Mdu to develop at the maximum moment
region. There is no limit on distance between braced points for members with circular
or tubular cross sections or for any beam bent about its weak axis. Equation 6-6 from
TM 5-1300 calculates the reduced ultimate dynamic moment capacity, Mm, for beams
with lateral bracing spaced at more than lb in Equation 6-5.
Equation 6-5
17200ry
lb =
f dy
where:
lb = distance between cross sections braced against twist or lateral
6-31
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
displacement of the compression flange, equal to the actual length for
cantilevers braced only at the support (mm)
ry = radius of gyration of the member about its weak axis (mm)
fdy = steel dynamic yield stress (MPa)
SBEDS requires an input distance between laterally braced points of steel beams for
inbound and rebound response, lbp, which is used in Equation 6-6 to calculate Mm for
inbound and for rebound response. Mm is used in SBEDS to calculate the ultimate
flexural resistance of all I-beams and channels. SBEDS always calculates an ultimate
dynamic moment capacity based on the plastic section modulus in Equation 6-4 for
tubular cross sections and beams responding in the weak bending axis direction.
Equation 6-6
⎛ l bp ⎞
M m = ⎜1.07 − f dy ⎟ M du ≤ M du
⎜ 8300ry ⎟
⎝ ⎠
where:
Mm = ultimate dynamic moment capacity accounting for effects of lateral
bracing
lbp = distance between lateral bracing points for compression flange
provided to beam input separately in SBEDS for inbound and
rebound response(mm)
ry = radius of gyration of the member about its weak axis (mm)
fdy = steel dynamic yield stress (MPa)
Ultimate design procedures, combined with the effective flange width concept to get
cross sectional properties to account for some local buckling in the maximum moment
region, are used to determine the ultimate moment capacity of cold-formed steel
components. The effective section modulus at flexural yielding of the cross section and
the value of moment of inertia at a service stress level of 140 MPa (20,000 psi) are
provided for standard cold-formed steel component shapes. These properties can also
be calculated based upon the relationships in the AISI Design Specification.
Equation 6-7 shows the ultimate dynamic moment capacity calculated in SBEDS for
cold-formed steel beams, corrugated steel panels, and metal studs. These components
typically do not maintain the ultimate moment capacity at deflections greater than the
6-32
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
yield deflection because of increased localized buckling over the whole compression
area of the cross section with increasing post-yield deflection. This is characterized by
“kinking” that occurs in the maximum moment regions on the compression face, as
shown in Figure 6-3. However, typical connections and in-plane restraint for these
components can allow a limited amount of tension membrane to develop that at least
replaces the loss of flexural moment capacity caused by buckling in the compression
flange out to post-yield deflections corresponding to high damage levels. This is more
true for beams than panels, which is the reason for the 0.9 reduction factor on the
ultimate moment capacity for panels in Equation 6-7.
Equation 6-7
M du = S e f dy for beams
M du = 0.9 S e f dy for panels
where:
Mdu = ultimate dynamic moment capacity
Se = effective section modulus accounting for localized buckling at from
high stress in compressive face
fdy = dynamic yield strength of cold-formed steel
Figure 6-3 Local Buckling in Maximum Moment Region of Cold-Formed Beams
and Panels
6-33
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
6-3 METAL STUD WALLS
Metal stud walls consist of cold-formed steel sections, typically spaced at 400 mm (16
inch) centers, spanning vertically between floors,. Designated structural metal stud
shapes are provided by the Steel Stud Manufacturer’s Association (SSMA). The studs
typically have web punch-outs to facilitate electrical conduit placement. Standard factory
punch-outs have a minimum center-to-center spacing of 600 mm (24 inch), a maximum
width equal to the lesser of one-half the member depth or 65 mm (2.5 inch), and a
maximum length of 115 mm (4.5 inch). The minimum distance to the supports is 255
mm (10 inch). Punch-outs are assumed to reduce the critical axial section of the stud for
compression and tension in SBEDS, but not the flexural or shear capacity. Therefore,
the effective cross sectional area, Ae in Equation 6-8, is used in Equation 6-10 and
Equation 4-11 to help determine the axial load capacity and controlling tension
membrane force, respectively, in metal studs with web punch-outs.
Equation 6-8
Ae = A − w po t
where:
Ae = effective cross sectional area
A = full member cross sectional area away from punch-out
wpo = width of punch-out (lesser of 65 mm of one-half member depth)
t = member thickness
The dynamic response of metal studs is calculated in a similar manner as any cold-
formed steel beam in SBEDS. However, the combined response of a metal stud wall
and a masonry veneer wall can also be considered in SBEDS. In this case, the veneer
wall is assumed to deflect with the metal studs and to have the same span as the studs
with simple boundary conditions. There is not available test data to show that the
veneer wall deflects with the studs during blast loading, but significant ties are required
between the veneer wall and studs for conventional design and very little deflection
occurs during blast loading before the low-strength veneer wall yields in a brittle manner
and the studs provide almost all additional strength. The resistance-deflection
relationship for the veneer wall and metal stud wall are superimposed to get the
relationship for the combined wall system. The veneer wall is assumed to have brittle
flexural response and axial load arching from self-weight, as explained in Section 4-
1.2.2.3.
The metal studs are assumed to respond in ductile flexural response. The first part of
the combined resistance-deflection relationship is typically dominated by the veneer wall
response, since it is much stiffer than the metal studs. After the veneer wall reaches
maximum flexural resistance and responds in axial load arching from self-weight, the
combined resistance-deflection relationship is usually dominated by the metal stud
response. Typically, veneer walls are relatively thin, weak, and brittle and therefore do
not add much blast capacity to the combined wall system (less than 20%). If the
masonry veneer wall fully yields in flexure (i.e., cracks) during inbound response, only
the metal stud wall response is considered by SBEDS during rebound.
6-34
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
6-4 STEEL BEAM-COLUMNS WITH COMBINED FLEXURE AND
COMPRESSIVE AXIAL LOAD
The moment capacity for beam-columns is calculated in SBEDS based on the
interaction formula for combined lateral and axial load from the AISC ASD Design
Specification, Chapter N, Plastic Design, with the dynamic yield strength substituted for
the static yield strength. The same interaction equations are in Chapter 5 of TM 5-1300.
These interaction equations are set equal to 1.0 and solved to determine the available
moment capacity considering simultaneous axial compressive load in Equation 6-9. The
ultimate dynamic moment including axial effects, MduA, in Equation 6-9 is used in
SBEDS to determine the ultimate flexural resistance for all beams with input axial load.
The term (1-P/Pe)/Cm in Equation 6-9 is assumed equal to 1.0 in SBEDS because
second order P-delta effects are accounted for directly in SBEDS with an equivalent
lateral load, as described in Section 4-4.1. The moment capacity for metal studs in
SBEDS is only based on the interaction equation that considers column slenderness,
which is consistent with static design for these members.
Equation 6-9
⎛ P ⎞⎡ 1 ⎛ P ⎞⎤ 1 ⎛ P⎞
M duA1 = M m ⎜⎜1 − ⎟⎟ ⎢ ⎜⎜1 − ⎟⎟⎥ where ⎜⎜1 − ⎟⎟ = 1 in SBEDS
⎝ Pcr ⎠ ⎣ C m ⎝ Pe ⎠⎦ Cm ⎝ Pe ⎠
⎛ P⎞
M duA2 = 1.18M du ⎜1 − ⎟ ≤ M du
⎜ P ⎟
⎝ y ⎠
Equation 6-10
KL f dy
Pcr = 0.658 λc f dy A λc ≤ 1.5 where λc =
2
if
rπ E
0.877
Pcr = f dy A if λc > 1.5
λc 2
where:
MduA= ultimate dynamic moment capacity accounting for axial compressive
load
P = applied axial compression load including input static load and peak
dynamic axial load
Pe = Euler buckling load
Pcr = axial load capacity accounting for column slenderness (see Equation
6-10)
Py = axial load capacity for short column based on beam-column cross
section area = fdyA
A = cross sectional area
Mm= moment capacity accounting for bracing to compression flange in
weak axis direction (see Equation 6-6)
Mdu= moment capacity accounting for fully braced component =fdyZ (see
6-35
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Equation 6-4)
Cm = coefficient applied to bending term in moment-axial load interaction
formula and dependent upon column curvature caused by applied
moments (see AISC Design Specification)
E = Young’s modulus
KL/r = larger slenderness ratio considering strong and weak bending axes
where L = unbraced length, K= effective length factor based on
boundary conditions, and r = radius of gyration about each axis. K
assumed =1.0 in weak bending axis. Critical KL/r is calculated for
inbound response and also used for rebound response in SBEDS.
where:
fdv = dynamic shear strength of steel or aluminum
fdy = dynamic yield strength of steel or aluminum
Av = shear area of component (i.e., beam web area)
h = height of beam web
t = web thickness
E = Young’s modulus
Equation 6-12
For fy= 227 MPa (33,000 psi)
⎛h⎞ 8.7x103
If ⎜ ⎟ ≤ 83, then f dv = MPa
⎝t⎠ (h/t)
⎛h⎞ 7.4x105
If 83 < ⎜ ⎟ ≤ 150, then f dv = MPa
⎝t⎠ (h/t) 2
Equation 6-13
For fy= 345 MPa (50,000 psi)
6-36
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
⎛h⎞ 10.6x103
If ⎜ ⎟ ≤ 67, then f dv = MPa
⎝t⎠ (h/t)
⎛h⎞ 7.4x105
If 67 < ⎜ ⎟ ≤ 150, then f dv = MPa
⎝t⎠ (h/t) 2
Equation 6-14
For fy = 550 MPa (80,000 psi)
⎛h⎞ 12.3x10 3
If ⎜ ⎟ ≤ 58, then f dv = MPa
⎝t⎠ (h/t)
⎛h⎞ 7.4x105
If 58 < ⎜ ⎟ ≤ 150, then f dv = MPa
⎝t⎠ (h/t) 2
Table 6-3 Dynamic Shear Strengths for Cold-Formed Steel Panels at Fixed
Interior Supports (Combined Bending and Shear)
Web Height to Thickness fy = 227 MPa fy = 345 MPa fy = 550 MPa
Ratio (h/t)
20 75 113 149
30 75 112 145
40 74 110 138
50 73 109 130
60 72 103 121
70 70 98 110
80 68 90 99
90 66 81 86
100 62 72 74
110 57 60 61
120 51 51 51
The shear capacity of components not meeting the h/t criterion in Equation 6-11 is
dictated by web instability due to shear stresses, including elastic buckling for large
values of h/t and inelastic buckling for intermediate h/t values. Equation 6-12 through
Equation 6-14 from TM 5-1300 account for the effects of web instability on shear
strength based on criteria from AISI (American Iron and Steel Institute) LRFD design,
where the dynamic yield strength is substituted for the minimum static yield strength. If
the input yield strength does not exactly match any of the minimum specified static yield
strengths shown for each of these equations, the equation for the closest matching yield
strength is used.
Standard load tables are available from the SJI for simply supported, uniformly loaded
joists with the top chord continuously braced by steel deck. The load capacity of a
particular joist may be governed by either flexure or shear (maximum end reaction or
buckling of compression web member), where flexure controls for longer spans and
shear controls for shorter spans. At typical spans, the load capacity of most joists is
controlled by flexural response. The governing response mode is not explicitly stated in
current load tables, but it can be inferred if the allowable load changes with the square
of span length (flexure control) or linearly with span length (shear control). At very short
spans, the allowable load capacity becomes constant with span. It is preferable in blast
6-38
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
applications to select a member whose capacity is controlled by flexure rather than
shear, since shear is a less ductile response mode.
The ultimate flexural resistance of an OWSJ with a span L may be calculated from the
published allowable load capacity for the joist type and span length of interest using
Equation 6-15 from TM 5-1300. This formula does not differentiate between a load
capacity based on shear or flexure because both the joist shear and flexural capacity
are assumed to increase by the same static and dynamic strength increase factors. The
values for the SIF and DIF in
6-39
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Equation 6-15 are conservatively based on the values in Table 6-1 for the highest
possible steel yield strength in the joist of 345 MPa (50,000 psi). After shear failure or
during flexural response with large deflections, the top chord of the joist will act as a
tension element between supports and some tension membrane resistance will
develop.
Equation 6-15
⎛w ⎞ w
ruf = 1.7ac⎜ dL ⎟ = 2.12 dL
⎝ b ⎠ b
where:
ruf = ultimate flexural resistance per unit area of joist for span L
wdL = allowable load per unit length for joist with span L from SJI load
tables
b = joist spacing
a = applicable static increase factor (SIF) from Table 6-1 (assumed 1.05)
c = applicable dynamic increase factor (DIF) from Table 6-1 (assumed
1.19)
Significant tension membrane resistance was noted in blast tests sponsored by the U.S.
government on OSWJ designed to represent typical industrial building construction, with
standard welded connections to base plates embedded in walls of a reinforced concrete
test building (Coltharp et al, 1999). SBEDS conservatively does not account for any
tension membrane response in OWSJ. However, tension membrane response provides
an additional level of redundancy to help maintain the resistance of OWSJ governed by
brittle response at larger deflections.
The flexural stiffness of an OWSJ can be calculated from the published load causing
deflection equal to L/360, where L is the span length, for the joist type and span length
of interest, as shown in
Equation 6-16. Values for wLL in
Equation 6-16 are published for each joist type and span length in the standard SJI load
tables.
Equation 6-16
360w LL
K=
bL
where:
K = flexural stiffness in units of pressure/length
wLL = load per unit length for joist with span L causing L/360 deflection
from Manufacturer’s load tables or SJI load tables
Values for wdL and wLL in the standard SJI load tables for each K series and LH series
joist type at various span lengths were used to back-calculate an average allowable
moment capacity and flexural stiffness (i.e., EI) for each joist type assuming uniformly
6-40
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
loaded, simply supported spans. These moment capacities and flexural stiffnesses are
stored in a data table in SBEDS, along with the maximum allowable load for each joist
type that controls at short span lengths. The allowable moment capacity for the input
joist type and the input span length are used to calculate wdL from
Equation 6-15 in SBEDS, which is not allowed to exceed the maximum allowable load
from the SJI load tables for the joist type. The flexural stiffness of the input joist and the
input span length are used to calculate wLL from
Equation 6-16 in SBEDS, which is used to calculate the spring stiffness K of the
equivalent SDOF system. The use of back-calculated moment capacity and bending
stiffness values for each joist type in SBEDS precludes the need to enter the entire load
tables with data for all spans of interest for each joist type into a look-up data table.
6-8 REBOUND
Rebound causes a reversal of connection forces and component stresses (i.e., tension
regions become compression regions and vice versa), as discussed in Section 4-1.2.5.
The tension flange or tension chord of steel components is not normally provided with
sufficient lateral bracing to develop full compression yield strength during rebound. This
can be accounted for in SBEDS for steel beams by inputting separate unbraced lengths
for inbound and rebound response. In each case, the ultimate moment capacity
considering the unbraced length, Mm, is calculated with Equation 6-6.
Tension membrane response provides resistance at large rebound deflections for
beams that loose flexural capacity because they undergo lateral torsional buckling in
rebound due to insufficient lateral bracing of the tension flange. This is particularly true
for cold-formed beams because the tension membrane resistance is a larger
percentage of the low flexural resistances of these components and these components
customarily have relatively light bracing on the tension flange and an unsymmetrical
cross section that is more susceptible to twisting. There have been many observed
cases of severely twisted cold-formed beams that are still attached to industrial
buildings after accidental explosions. Therefore, total failure of steel beams in rebound
is very uncommon in blast tests and explosion events. It is possible to see some beams
that apparently were damaged more during rebound than during inbound response.
Unlike steel beams, there is no separate input of rebound properties in SBEDS for open
web steel joists. Joists are usually only designed with lateral bracing on the bottom
chord to develop 20 psf to 40 psf in uplift pressure, depending on the design wind load,
which is less than the rebound resistance necessary to respond to most blast loads.
However, limited blast testing has shown that the top chord can develop sufficient
tension membrane response after the bottom chord buckles during rebound response to
prevent failure, even when bottom chord buckling during rebound was severe enough to
cause plastic deformation (Coltharp et al, 1999). The limited duration of rebound
response also helps limit damage.
These observations, combined with the fact that cross sectional properties and bracing
for the lower chord of joists are often difficult to obtain without a site inspection, are
reasons why no rebound input for open web steel joists is included in SBEDS. The
inbound and rebound ultimate flexural resistances of open web steel joists are always
simplistically assumed equal, where top chord tension membrane is implicitly assumed
to make up for any lower flexural resistance available during rebound. Open web steel
6-41
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
joists that are designed to resist blast load should ideally have adequate lower chord
bracing to develop the full compression yield strength of the bottom flange, or at least to
develop the bottom flange stress based on the maximum calculated rebound resistance
of component.
Plates, corrugated steel panels, and beams with closed cross sections or are in bending
about the weak axis typically do not require lateral bracing to develop full compressive
yield stress in flexure. Therefore, these components have equal inbound and rebound
ultimate flexural strength.
6-9 CONNECTIONS
Steel connections are designed to resist the equivalent static reaction force from the
component, as described in Section 4-7. The strengths of simple shear or bearing
connections can be determined based on LRFD connection strengths or manufacturers’
allowable connection strengths, as described in Section 4-7. Any bearing connections
normally must also be designed for tension from reaction forces occurring during
component rebound, unless outward failure of the component is acceptable (e.g., cold-
formed steel panels). The connection strength of cold-formed steel components is often
controlled by material bearing failure or tear-out around bolts and screws. The
connection strength can be improved for these cases by using washers to distribute
material stresses around screws and screwing additional gage thickness steel plate or
beam sections to beams near bolted connections. Conservatively, the maximum
rebound connection force is assumed equal to the maximum inbound connection force,
or it may be based on the equivalent static reaction force from the peak rebound
resistance. Rebound is discussed in more detail in Section 4-1.2.5.
Moment resisting steel connections require special attention to detail to help ensure
they can rotate adequately without a reduction in moment capacity at large component
deflections. Ideally, the connection should have a greater capacity than the maximum
dynamic flexural and shear loads that are delivered by the connected components. This
can be accomplished by using a connection with a reduced beam flange or by using
other proprietary and non-proprietary connections that have been developed and tested
by the researchers after the Northridge earthquake for seismic steel beam-column
connections. As a minimum, connections should include stiffeners that will help transfer
tension and compression forces from the flanges of connected components through the
connection. Bolted connections and connections where fillet welds are kept in shear are
more ductile, and therefore preferable, to welded connections that are in tension or
compression.
The web area within the boundaries a moment resisting connection should meet the
shear stress requirements of Section 6-5. If the web area is unsatisfactory, diagonal
stiffeners or web doubler plates should be provided. Web crippling must also be
checked at points of load application, such as beam-girder intersections. In these cases,
the requirements of AISC LRFD (2001) should be considered.
6-42
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Figure 6-4 shows an example of a moment resisting beam connection designed to
undergo large rotations from blast loading adapted from TM 5-1300.
6-43
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Figure 6-4 Moment Resisting Steel Beam Connections Designed for Large
Rotations (from TM 5-1300)
Much research on steel frames response to earthquakes has been performed because
of many cases of observed damage to steel frames from earthquakes. There have not
been known cases where moment resisting connections in steel frames have failed
during blast loading, but there has been very few of these types of structures that have
been subject to blast tests or severe loads from accidental explosions. There have been
cases where significant local buckling of flanges on built-up girders in pre-engineered
buildings occurred without causing overall frame failure. Seismic criteria for
connections of moment resisting steel frames subject to blast loads is probably
conservative due to the much shorter duration of blast induced stresses. However,
these criteria are the best available design criteria for steel frames.
6-44
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
almost always control. These criteria can conservatively be used for blast design. Less
conservative support rotation and ductility ratio criteria for blast design and analysis of
steel frames are given in ASCE (1997) that are intended primarily for one or two-story
steel frame buildings.
6-45
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Table 7-1 provides information for the most common types of masonry units that are
available in SBEDS. European insulated block has variable properties; the values
shown in Table 7-1are average values based on limited available information. Table 7-1
also includes several weights of CMU. CMU manufacturing depends in part on locally
available materials, but most CMU used in modern U.S. construction is lightweight
CMU. This has become increasingly true over the last 20 to 30 years due to advances
in CMU materials that has allowed lightweight CMU to meet ASTM strength
requirements. Heavier weight CMU has been used more in older construction and cold
weather climates. Unfortunately, it is very hard to determine the weight of CMU block,
without sampling some block and measuring the specific gravity, if it is not noted in
construction drawings or specifications. It is usually conservative to assume lightweight
CMU for blast design or analysis because this will result in the least calculated mass
and corresponding inertial resisting force.
Table 7-2 shows cross sectional property information used in SBEDS for each masonry
type. This table includes a parameter for the percentage of solid thickness through the
webs, as illustrated in Figure 7-2. The parameter is used to determine the shear
strength, as discussed in 7-5. SBEDS also considers grouted CMU block, where the
areal wall weight, moment of inertia area, and percentage of solid thickness through the
webs of grouted CMU block are calculated as shown in Equation 7-1. The uncracked
section modulus of grouted CMU is equal to the moment of inertia divided by one-half
the thickness.
7-46
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Figure 7-1 Typical Masonry Unit Shapes
Various Brick Shapes
7-47
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Table 7-1 Masonry Block Information
Type Density Percentage Solid Moment Section
Solid Cross Thickness of Inertia Modulus
kg/m3 (pcf) Section (%) Through (per unit (per unit
Webs1 (%) width)2 width)2
Solid Brick 1925 (120) 100 100 t3/12 t2/6
European Insulated 962-1925 50 50 0.3(t3/12) 0.3(t2/6)
Block3,4 (60-120)
Heavy Weight CMU 2165 (135)
Light Weight CMU 1524 (95) See Table 7-2
Medium Weight CMU 1925 (120)
Note 1: See Figure 7-2.
Note 2: “t” is block or brick thickness.
Note 3: A density of 1395 kg/m3 (87 pcf) is used in SBEDS. Based on comparisons of available
manufacturers’ values to solid block values, the area is assumed equal to 50% of the solid area and the
section modulus and moment of inertia is assumed equal to 30% of the solid section values for insulated
blocks in Figure 7-1.
Figure 7-2 Percent Solid Thickness Through Web of CMU Cross Section
Shell Thickness ( t s )
A A
Percent Solid Thickness Through Web =100 x (width of solid space through block along A-A)/(block
width)
Equation 7-1
⎛ V ⎞
w = ⎜ A fs ρ m + (1 − A fs ) ρ g ⎟t
⎝ 100 ⎠
(t − 2t s ) 3 V ⎛ ψ m ⎞
I = Im + ⎜1 − ⎟
12 100 ⎝ 100 ⎠
ψ =ψ m +
V
(100 − ψ m )
100
where:
7-48
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
w = areal weight of grouted cross section
I = moment of inertia of grouted CMU
Ψ = percentage of solid thickness through webs and grout of CMU
Afs = fraction of cross sectional area that is solid CMU (see Table 7-2)
rm = density of CMU (see Table 7-2)
rg = density of grout (equal to 120 pcf or 1922 kg/m3 in SBEDS)
ts =CMU shell thickness (see Table 7-2)
t = masonry thickness
V = percent of voids that are grouted
Im = moment of inertia of CMU not considering grout (see Table 7-2)
Ψm = percentage of solid thickness through webs of CMU
Note: Brick and European Block do not include consideration of any grouted area
Table 7-3 shows information on mortars used for masonry construction. The voids of
CMU walls are filled with grout if they have vertical reinforcement. Usually, unreinforced
voids are not grouted. Grout, which is very similar to concrete without any large
aggregate, generally has compression strengths near 20 MPA (3000 psi). It must have
sufficient slump, or plasticity, and must be vibrated during placement in the CMU voids
to avoid voids, or honeycombing in the grout. Excessive mortar squeezed into the void
space of CMU block, or falling off into the void space, inhibits good placement of grout
and may compromise composite action of the vertical reinforcement with the
surrounding CMU wall.
Masonry walls can be constructed as a single, monolithic wall with one or more units
through the thickness, or as multi-wythe cavity walls consisting of multiple, closely
spaced walls with an air gap or insulation between walls. Almost all modern masonry
walls in the northern part of the United States and Canada are multi-wythe cavity walls,
which are necessary to provide required insulation. They are also becoming more
common in the other parts of the U.S. for buildings where insulation is a primary
concern. Figure 7-3 shows a cavity wall constructed with brick and CMU. Cavity walls
7-49
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
can also be constructed with other types of masonry. All cavity walls are typically
connected together across the gap between walls with joint reinforcement, as shown in
Figure 7-3, or with steel ties or tabs that cause both walls to deflect together when
lateral loads are applied to the outer wall.
Figure 7-3 Double Wythe Cavity Wall with Brick and CMU
7-50
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Table 7-4. Recommended Conservative Values for Static Masonry Prism
Compressive Strength
Type of Masonry Unit Compressive Strength1 (f'm)
Ungrouted CMU 1350 psi
Fully Grouted CMU 1500 psi
Brick or Solid Masonry or European Insulated Block 1800 psi
Note 1: SBEDS assumes that f’m is based on the prism crushing load divided by gross
cross sectional area of block with small uniformly distributed voids, such as European
insulated block. However, f’m is based on the crushing load divided by net area of
masonry with large voids, such as CMU or clay tile.
Equation 7-2
f ' dm = f ' m (DIF )
where:
f’dm = masonry dynamic compression stress
f’m = static masonry prism compression strength
DIF = dynamic increase factor (equal to 1.19)
Equation 7-3
E m = 1000 f ' m
where:
Em = masonry modulus of elasticity
f’m = masonry prism compression strength
The flexural tensile strength of masonry walls is controlled by the adhesion between the
mortar and masonry units. It can be measured with wallettes, which are short
unreinforced masonry wall sections that are tested in three-point bending. The tension
strength is determined from the breaking load and the section modulus of the masonry
unit. Measured static flexural strengths vary significantly because of variables that affect
the mortar adhesion to the masonry, including mix proportions of the mortar and
absorption of moisture from the mortar by the masonry block. The static flexural tension
strength for a vertically spanning ungrouted CMU wall, which cracks in the maximum
moment region within a single horizontal mortar joint, is in the range of 0.2 MPa (30 psi)
to 0.35 MPa (50 psi). The flexural tension strength for a horizontally spanning ungrouted
CMU wall with a running bond masonry placement, which cracks in the maximum
moment region through the mortar in a stair-stepping pattern around the masonry units
or through the masonry units, is in the range of 0.69 MPa (100 psi) to 1.0 MPa (150 psi)
(Hamid and Drysdale, 1988). Grouted block and solid block have somewhat higher
tensile strengths.
Usually, masonry walls are built continuous in front of steel columns, with a gap
between the wall and column, or between concrete columns, with an expansion gap
between the wall and columns. In both cases there are usually only light metal ties
attaching the walls to the columns that cannot transfer the equivalent static reaction
load from the wall spanning horizontally between columns. If the gap between the wall
7-52
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
and adjacent concrete columns is small enough, the wall can develop some bearing
support in the horizontal direction as it deflects or rotates into the column.
Another issue affecting flexural response of masonry walls is the assumed support
condition between the bottom of the wall and the foundation. It seems reasonable to
assume that unreinforced walls with dowels extending from the foundation are fixed at
the bottom of the wall. However, these dowels almost never extend into the wall
sufficiently to provide a full tension splice as required in ASCE 530-02 for reinforced
masonry. In this case, the wall will only act fixed at the base for a short time until the
splice fails and it is conservative to analyze the wall assuming a simple support at the
base. If the dowels do have a necessary splice length, the wall can be assumed fixed at
the foundation if the rigidity of the foundation against rotation is adequate. This is often
the case considering the very short duration of blast load response. Unreinforced walls
without dowels should typically be assumed pinned at the base because of the unknown
adhesion between the mortar and the foundation material.
Brittle flexural response of an unreinforced masonry wall is based on the wall moment
capacity, which is controlled by the flexural tensile strength between masonry units. As
discussed in Section 7-2, there is no available test data on the dynamic flexural tensile
7-53
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
strength of masonry walls. If no better information is available, a value of 1.38 MPa (200
psi) is recommended based on use of this assumed value in SDOF analyses that
approximately matched measured unreinforced masonry wall response from a number
of high explosive tests (Oswald, 2005). If f’dm exceeds 1380 MPa, a value of 10% of f’dm
may also be used provided that it does not exceed 1.75 MPa (250 psi). Equation 7-4
shows the ultimate dynamic moment capacity of unreinforced masonry walls based on
the dynamic tensile strength of the wall and the elastic section modulus of the masonry
unit. Values for the elastic section modulus of solid blocks and European insulated
blocks that are assumed in SBEDS are shown in Table 7-1. Values for CMU that are
assumed in SBEDS are shown in Table 7-2.
Equation 7-4
M du = Sf dt
where:
Mdu = ultimate dynamic moment capacity
fdt = dynamic tensile strength of masonry wall
S = elastic section modulus of masonry unit
The dynamic moment capacity in Equation 7-4 is used to determine the ultimate flexural
resistance of an unreinforced masonry wall based on the boundary conditions as
described in Section 4-1.2.1. In the case of ductile flexural response, this resistance is
assumed to remain constant with increased deflection beyond the yield deflection.
However, the recommended approach is to assume brittle flexural response, where the
flexural resistance is zero at deflections beyond the yield deflection and axial load
arching resistance occurs between the flexural yield deflection and the wall thickness.
Axial load arching is discussed in Section 4-1.2.2.3.
Ductile flexural response is always assumed in SBEDS for steel reinforced masonry
walls. The ultimate dynamic moment capacity is calculated using Equation 7-5 from TM
5-1300. This equation assumes that reinforced masonry will respond in essentially the
same manner to blast load as reinforced concrete. The applicability of this assumption
has been demonstrated in numerous static tests (Heeringa et al, 1989) (Hamid et al,
1992). It is also consistent with data from a series of twenty-one shock tube tests on
reinforced masonry CMU walls that were subject to inspection and material testing
during construction, as required in the International Building Code and ASCE 530-02
(Oswald and Holgado, 2004).
7-54
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
However, the deflections calculated with SDOF analyses based on a moment capacity
from Equation 7-5 were approximately 30% to 40% lower than measured deflections in
twenty-two reinforced CMU walls from three different test series in the CMUDS
database (Wesevich et al, 2002) where no quality assurance (QA) inspection or
materials testing was performed during test wall construction (Polcyn et al, 2004).
Therefore, Equation 7-5 was non-conservative for these cases assuming that the
difference between measured and calculated deflections was due to a reduced moment
capacity in the test walls caused by poor composite action of the masonry and
reinforcing steel as it was stressed near its dynamic yield strength. Poor composite
action in reinforced CMU walls is caused by grout that is not well consolidated during
placement due to insufficient vibrating or rodding or mortar after it is placed in the cells,
mortar protrusion and drippings in the cells, or insufficient plasticity due to incorrect
proportioning and mixing of materials used to make the grout.
Equation 7-5
As f dy ⎛ A f ⎞
M du = ⎜⎜ d − s dy ⎟⎟
b ⎝ 1.7bf ' dm ⎠
where:
Mdu = dynamic ultimate moment capacity per unit width
As = area of tension reinforcement in width, b
fdy = dynamic design stress
d =effective depth (distance from the extreme compression fiber to the
centroid of the tension reinforcement)
b = spacing between reinforcement
f'dm = masonry prism dynamic compressive strength
This indicates that QA during construction as required in the IBC and ACI 350-02 is very
important for ensuring the applicability of Equation 7-5 for reinforced masonry walls
subject to blast loads. If no such QA is performed, the ultimate dynamic moment
capacity in Equation 7-5 should be reduced by a factor of approximately 60%, which
would cause the previously sited twenty-two walls from the CMUDS database to have
calculated deflections approximately equal, on the average, to measured values. This is
consistent with similar guidance for static design of reinforced masonry walls from the
Uniform Building Code (UBC, 1997), which recommended that only one-half the
theoretical wall design strength should be assumed for uninspected reinforced masonry
walls. The applicability of Equation 7-5 also depends on proper splicing of reinforcing
steel in accordance with ACI 530-02 and splicing only small bars in maximum moment
regions. Based on recent shock tube testing, only small spliced bars with a diameter no
larger than 12 mm (0.5 in) maintained their full dynamic yield strength in the maximum
moment region without slipping at large rotations (i.e., greater than 2 degrees), when
spliced according to the ACI requirements (Oswald and Holgado, 2004).
7-55
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
used for blast design of DoD buildings according to UFC 3-340-01 and TM 5-1300 for
response with a support rotation greater than 2 degrees.
Masonry can also be reinforced with fiber reinforced plastics (FRP) that are attached
with epoxy on the exterior surfaces of the walls. The FRP can be placed as a large
sheet or in separate strips that are continuous over the wall span. The tension strength
provided by the fibers significantly increases the flexural strength of masonry, but do not
add very much ductility. The fibers do not significantly increase the strength of the
masonry compression block in the maximum moment region or the masonry shear
strength. Therefore, the strength of masonry walls reinforced with FRP must be checked
for possible failure in these response modes.
FRP reinforcement must be placed on both sides of walls to avoid failure during
rebound response. Regardless of the controlling response mode, the flexural response
of walls reinforced with FRP is usually not very ductile. Because steel reinforcement
provides both increased ductility and strength, it is usually used to reinforce new
masonry walls designed to resist blast loads. However, FRP may be a practical option
for reinforcing existing unreinforced masonry walls, as discussed in UFC 04-023-02. No
options are provided for input of FRP reinforcement in the current version of SBEDS
(SBEDS V3.0), although FRP reinforcement can be input for a Reinforced Masonry
component with the user input option for material properties.
The moment of inertia for reinforced masonry walls in the elastic and elastoplastic range
is complicated by increasing cracking along the masonry wall span as the wall deflects.
It is recommended in TM 5-1300 that the flexural stiffness in the elastic and
elastoplastic range be based on an average moment of inertia calculated with Equation
7-6. This formula is based on a singly reinforced wall with a 0.1% reinforcement ratio,
which is a lower bound on commonly used reinforcement ratios. However, even tripling
the value of Ic in Equation 7-6 typically has little effect on Ia, which is dominated by the
gross moment of inertia term. SBEDS calculates the moment of inertia for reinforced
masonry walls as Ia in Equation 7-6.
Equation 7-6
In + Ic
Ia =
2
I c = 0.005d 3
where:
Ia = average moment of inertia per unit width used to determine flexural
7-56
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
stiffness for equivalent SDOF system of a reinforced masonry wall
In = gross moment of inertia per unit width based on net section of
masonry units in wall
Ic = cracked moment of inertia per unit width
d = effective depth (distance from the extreme compression fiber to the
centroid of the tension reinforcement)
Closed stirrups similar to those used in reinforced concrete beams can be placed along
the horizontal bed joints near the support where the equivalent static reaction load
exceeds Vm from Equation 7-7. In this case, the shear strength is the sum of Vm from
Equation 7-7, which represents the shear force resisted by the masonry, and Vs from
Equation 7-8, which represents the shear force resisted by shear reinforcing steel
placed perpendicular to the flexural steel. However, the shear reinforcement must be
placed at a maximum spacing along the span equal to the depth of flexural reinforcing
steel from the compression face, d, divided by 2 (i.e., d/2) to be effective. This is almost
impossible to do for most reinforced masonry walls, since the reinforcement is often
7-57
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
placed at midthickness and the bed joints are always 200 mm (8 inches) apart. For this
reason, stirrups are only practical for a pilaster. If necessary, steel plates can be
through-bolted on both sides of a wall in the high shear area where the bolts act as
shear reinforcement. Fortunately, the equivalent static reaction force is less than Vm for
most practical cases.
Equation 7-8
Av f dy d
Vs =
bs
where:
Vs = dynamic shear force per unit width along wall resisted by reinforcing
steel
Av = area of shear reinforcement placed perpendicular to the flexural steel
b = spacing along wall between stirrups with area Av
s = spacing along span length between stirrups with area Av
d = effective depth of flexural reinforcement (distance from the extreme
compression fiber to the centroid of the tension reinforcement)
fdy = dynamic yield stress of the shear reinforcement
7-58
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Composite multi-wythe walls should be input in SBEDS as monolithic single walls so
that composite action will be considered.
Non-composite double wythe walls are assumed in SBEDS to respond to lateral blast
load as two springs in parallel, so that they deflect equally, have a flexural stiffness
equal to the summed stiffness of both walls, and resist the applied blast load in
proportion to their relative flexural stiffnesses. This is consistent with the approach
recommended in ACI 530-02. The two walls are assumed to resist lateral load in flexure
together in this manner until one of the walls yields at its dynamic tensile stress at all
maximum moment locations. The combined resistance of the two wall system when
this occurs is the ultimate flexural resistance.
If brittle flexural response with axial load arching is input for the walls, all resistance is
assumed to occur only due to axial load arching at deflections greater than the ultimate
flexural yield deflection of the wall yielding first. The peak resistance from axial load
arching is calculated separately for each wall and summed to get the peak axial load
arching resistance for the double wall system. The axial load arching resistance is
assumed to decay linearly to zero resistance at a deflection equal to the thickness of the
wall providing the larger axial load arching resistance. If ductile flexural response is
selected, the ultimate flexural resistance of the two-wall system is assumed to remain
constant with increasing deflections.
Equation 7-9 based on the assumptions that the walls have equal midspan
deflection, Δ, due to ties between the walls, and equal span (L) and boundary
conditions. For indeterminate walls,
Equation 7-9 is used to define both the maximum elastic resistance and ultimate
resistance of the two-wall system in terms of the corresponding resistances of
each wall. It is illustrative to consider a few examples. For the case where Wall 2
is identical to Wall 1 except that it has one-half the thickness, Wall 1 will have a
moment of inertia eight times greater than Wall 2 and a section modulus four
times greater. Based on
7-59
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Equation 7-9, Wall 1 will yield first and, since K1/K2 equals eight, the value of Ru for the
two-wall system will only be 1.125 times greater than Ru of the thicker wall, R1. This
contrasts sharply with the case where both walls are equal thickness, where the value
of Ru for the combined wall system is 2.0 times R1. The stiffness of the double wythe
wall system is shown in Equation 7-10. This equation is assumed to apply to both the
elastic and elastic-plastic stiffnesses of indeterminate multi-wythe walls.
The shear capacity of the two-wall system, Vu, is based on the combined
maximum resisting shear forces of the two walls when the first wall yields in
shear, as shown in
Equation 7-11. Again, it is assumed that both walls have the same boundary
conditions and span. The shear capacity is compared to equivalent static reaction
forces from the double wythe wall system that are calculated based on the
ultimate flexural resistance of the two-wall system (Ru from
Equation 7-9) and the span and boundary conditions of the two-wall system using the
procedures in Section 4-3.2.
Equation 7-9
R1 K C K Ei I i R1 K 1 EI
R1 = K 1 Δ R2 = K 2 Δ therefore = 1 and Ki ∝ so = = 1 1
R2 K 2 L4 R2 K 2 E 2 I 2
Cm M i R1 M 1 σ 1 S1 σ 1 R1 S 2 E1 I 1 S 2 σ 1y
Ri ∝ therefore = = and = = = σ r , Let σ r y =
L2 R2 M 2 σ 2 S 2 σ 2 R2 S1 E 2 I 2 S1 σ 2y
σr ⎛ E I ⎞
If ≥ 1 , Wall 1 yields first and Ru = R1u + R2 = R1u ⎜⎜1 + 2 2 ⎟⎟ ,
σ ry ⎝ E1 I 1 ⎠
7-60
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
⎛ EI ⎞
Else, Ru = R2u + R1 = R2u ⎜⎜1 + 1 1 ⎟⎟
⎝ E2 I 2 ⎠
Note: Ru is also bounded where: Max(R1u, R2u) ≤ Ru ≤ (R1u+R2u)
Equation 7-10
K = K1 + K 2
where:
i =1 for inner wall and 2 for outer wall
Ri = resistance ith wall
Riu = ultimate or maximum resistance of ith wall
Ru = ultimate or maximum resistance of two-wall system
Ki = flexural stiffness of ith wall
K = flexural stiffness of two-wall system
Ei = Young’s modulus of walls
Ii = moment of inertia of ith wall
Si = section modulus of ith wall
Δ = equal midpan deflections of both walls
L = equal spans of both walls
σi = maximum flexural stress in walls
Mi = maximum moment capacity of ith wall
σiy = yield stress of ith wall
Cm = moment constants dependent on boundary conditions - same for
each wall
CK = stiffness constants dependent on boundary conditions - same for
each wall
Equation 7-11
C v Ri V1 R1 K 1 EI
Vi ∝ therefore = = = 1 1
L V2 R2 K 2 E 2 I 2
V1 v1 A1 v1 A2 R1 A2 K 1 v y1
= let v r = = = and v yr =
V2 v 2 A2 v 2 A1 R2 A1 K 2 vy2
7-61
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
vr ⎛ E I ⎞
If ≥ 1 , then Wall 1 yields first and Vu = V1u + V 2 = V1u ⎜⎜1 + 2 2 ⎟⎟
v ry ⎝ E1 I 1 ⎠
⎛ EI ⎞
Else, Wall 2 yields first and Vu = V2u + V1 = V2u ⎜⎜1 + 1 1 ⎟⎟
⎝ E2 I 2 ⎠
where:
i =1 for inner wall and 2 for outer wall
Vi = shear force on of ith wall
Viu = ultimate resisting shear force of ith wall
Vu = ultimate resisting shear force of two wall system
Ki = flexural stiffness of walls
Ai = shear area of ith wall
L = equal spans of both walls
vi = maximum shear stress in ith wall
vyi = yield shear stress of ith wall from Equation 7-7
Cv = shear force constants dependent on boundary conditions - same for
each wall
Connections between masonry walls and steel or reinforced concrete framing are
sometimes weak in rebound, especially for double wythe walls. These connections may
only be sized for suction wind pressures, which can be much less than the ultimate
7-62
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
flexural resistance of the wall during rebound. In this case, the walls can collapse
outward during rebound response to blast loads. This is especially hazardous for load
bearing walls.
7-63
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
7-64
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Wood properties are affected by many factors including the wood type, grade,
size, load duration, and moisture content. The grade is usually based on visual
inspection and accounts for the effects of knots, checks, shakes, splits, and
slope of grain, which can all decrease strength properties.
Figure 8-1 shows a typical flexural tension stress-strain relationship for wood, which
indicates that wood can have a limited amount of plasticity. Numerous national design
code bodies and manufacturers associations, including the National Design
Specification (NDS) and the International Building Code (IBC), issue tables of wood
properties with the flexural tensile strength and modulus of elasticity for various wood
types, grades, and common sizes. Tables are available showing both allowable
strengths and ultimate strengths, although allowable strength values are more
commonly available. The published strength values are intended for loads with a given
duration and factors are applied for design loads with shorter or longer durations.
8-1
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Published allowable design flexural strengths for wood are intended for loads with
approximately 10 year durations. A strength increase factor of up to 2.0 can be applied
to these strengths for impact loads. The ultimate strength of wood can be assumed
equal to 2.5 times the allowable strength, so the available strength to resist stresses
from short duration loads is approximately five times greater than the allowable design
stress. For example, the measured static failure load was 5 to 10 times greater than the
allowable load based on NDS flexural strength information in recent static tests with
durations of several minutes on wood stud wall systems at Penn State University (Starr
and Krauthammer, 2005). The walls were constructed with 2”x6” No. 2 grade Southern
Pine.
The only available testing on the relationship between wood strength and strain-rate is
for the crushing strength of wood (DOE/TIC 11268. 1992). This testing indicates that
between a slow loading rate, where yielding occurs in approximately 7 minutes (strain-
rate of 1e-5 sec-1), and a very fast loading rate, where yielding occurs in approximately
50 ms (strain-rate of 1e-1 sec-1), the crushing strength increases by a factor of 1.3 to 1.5.
There was no significant change in the modulus of elasticity. Even though this
information is for compression, it indicates that high strain-rates caused by blast loads
may cause some increase in wood flexural strength. However, no additional strain-rate
effect on yield strength, other than the impact factor of 2.0, is typically used for blast
analysis.
Table 8-1 and Table 8-2 show allowable static design strengths and moduli of elasticity
for various wood species and grades in SBEDS from the NDS. SBEDS calculates a
dynamic flexural strength for a user selected wood species and grade by multiplying the
allowable bending strength in Table 8-1 by a factor of 5.0 to represent the ultimate
strength under blast loading. SBEDS uses the static moduli of elasticity in Table 8-2
without any modifications to calculate the wood beam stiffness.
8-2
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Table 8-1 Allowable Bending Stresses for Wood Species from NDS
Allowable Bending Strength - psi (MPa)
Species Select
No. 1 No. 2 No. 3 Stud Construction Standard Utility
Structural
ALASKA CEDAR 1150 (7.93) 975 (6.72) 800 (5.52) 450 (3.1) 625 (4.31) 900 (6.21) 500 (3.45) 250 (1.72)
ALASKA HEMLOCK 1300 (8.96) 900 (6.21) 825 (5.69) 475 (3.28) 650 (4.48) 950 (6.55) 525 (3.62) 250 (1.72)
ALASKA SPRUCE 1400 (9.65) 950 (6.55) 875 (6.03) 500 (3.45) 675 (4.65) 1000 (6.89) 550 (3.79) 275 (1.9)
ALASKA YELLOW CEDAR 1350 (9.31) 900 (6.21) 800 (5.52) 475 (3.28) 625 (4.31) 925 (6.38) 500 (3.45) 250 (1.72)
ASPEN 875 (6.03) 625 (4.31) 600 (4.14) 350 (2.41) 475 (3.28) 700 (4.83) 375 (2.59) 175 (1.21)
BALDCYPRESS 1200 (8.27) 1000 (6.89) 825 (5.69) 475 (3.28) 650 (4.48) 925 (6.38) 525 (3.62) 250 (1.72)
BEECH-BIRCH-HICKORY 1450 (10) 1050 (7.24) 1000 (6.89) 575 (3.96) 775 (5.34) 1150 (7.93) 650 (4.48) 300 (2.07)
COAST SITKA SPRUCE 1300 (8.96) 925 (6.38) 925 (6.38) 525 (3.62) 725 (5) 1050 (7.24) 600 (4.14) 275 (1.9)
COTTONWOOD 875 (6.03) 625 (4.31) 625 (4.31) 350 (2.41) 475 (3.28) 700 (4.83) 400 (2.76) 175 (1.21)
DOUGLAS FIR-LARCH 1500 (10.34) 1000 (6.89) 900 (6.21) 525 (3.62) 700 (4.83) 1000 (6.89) 575 (3.96) 275 (1.9)
DOUG. FIR-LARCH (NORTH) 1350 (9.31) 850 (5.86) 850 (5.86) 475 (3.28) 650 (4.48) 950 (6.55) 525 (3.62) 250 (1.72)
DOUGLAS FIR-SOUTH 1350 (9.31) 925 (6.38) 850 (5.86) 500 (3.45) 675 (4.65) 975 (6.72) 550 (3.79) 250 (1.72)
EAST. HEMLOCK-BALSAM FIR 1250 (8.62) 775 (5.34) 575 (3.96) 350 (2.41) 450 (3.1) 675 (4.65) 375 (2.59) 175 (1.21)
EAST. HEMLOCK-TAMARACK 1250 (8.62) 775 (5.34) 575 (3.96) 350 (2.41) 450 (3.1) 675 (4.65) 375 (2.59) 175 (1.21)
EASTERN SOFTWOODS 1250 (8.62) 775 (5.34) 575 (3.96) 350 (2.41) 450 (3.1) 675 (4.65) 375 (2.59) 175 (1.21)
EASTERN WHITE PINE 1250 (8.62) 775 (5.34) 575 (3.96) 350 (2.41) 450 (3.1) 675 (4.65) 375 (2.59) 175 (1.21)
HEM-FIR 1400 (9.65) 975 (6.72) 850 (5.86) 500 (3.45) 675 (4.65) 975 (6.72) 550 (3.79) 250 (1.72)
HEM-FIR (NORTH) 1300 (8.96) 1000 (6.89) 1000 (6.89) 575 (3.96) 775 (5.34) 1150 (7.93) 650 (4.48) 300 (2.07)
MIXED MAPLE 1000 (6.89) 725 (5) 700 (4.83) 400 (2.76) 550 (3.79) 800 (5.52) 450 (3.1) 225 (1.55)
MIXED OAK 1150 (7.93) 825 (5.69) 800 (5.52) 475 (3.28) 625 (4.31) 925 (6.38) 525 (3.62) 250 (1.72)
NORTHERN RED OAK 1400 (9.65) 1000 (6.89) 975 (6.72) 550 (3.79) 750 (5.17) 1100 (7.58) 625 (4.31) 300 (2.07)
NORTHERN SPECIES 975 (6.72) 625 (4.31) 625 (4.31) 350 (2.41) 475 (3.28) 700 (4.83) 400 (2.76) 175 (1.21)
NORTHERN WHITE CEDAR 775 (5.34) 575 (3.96) 550 (3.79) 325 (2.24) 425 (2.93) 625 (4.31) 350 (2.41) 175 (1.21)
RED MAPLE 1300 (8.96) 925 (6.38) 900 (6.21) 525 (3.62) 700 (4.83) 1050 (7.24) 575 (3.96) 275 (1.9)
RED OAK 1150 (7.93) 825 (5.69) 800 (5.52) 475 (3.28) 625 (4.31) 925 (6.38) 525 (3.62) 250 (1.72)
Species Allowable Bending Strength psi (MPa)
8-1
PDC TR-06-01 Rev 1
September 2008
Select
No. 1 No. 2 No. 3 Stud Construction Standard Utility
Structural
REDWOOD 1350 (9.31) 975 (6.72) 925 (6.38) 525 (3.62) 575 (3.96) 825 (5.69) 450 (3.1) 225 (1.55)
SOUTHERN PINE 2850 (19.65) 1850 (12.76) 1500 (10.34) 850 (5.86) 850 (5.86) 1100 (7.58) 625 (4.31) 300 (2.07)
SOUTHERN PINE (MIXED) 2050 (14.13) 1450 (10) 1300 (8.96) 750 (5.17) 750 (5.17) 1000 (6.89) 550 (3.79) 275 (1.9)
SPRUCE-PINE-FIR 1250 (8.62) 875 (6.03) 875 (6.03) 500 (3.45) 675 (4.65) 1000 (6.89) 550 (3.79) 275 (1.9)
SPRUCE-PINE-FIR (SOUTH) 1300 (8.96) 875 (6.03) 775 (5.34) 450 (3.1) 600 (4.14) 875 (6.03) 500 (3.45) 225 (1.55)
WESTERN CEDARS 1000 (6.89) 725 (5) 700 (4.83) 400 (2.76) 550 (3.79) 800 (5.52) 450 (3.1) 225 (1.55)
WESTERN WOODS 900 (6.21) 675 (4.65) 675 (4.65) 375 (2.59) 525 (3.62) 775 (5.34) 425 (2.93) 200 (1.38)
WHITE OAK 1200 (8.27) 875 (6.03) 850 (5.86) 475 (3.28) 650 (4.48) 950 (6.55) 525 (3.62) 250 (1.72)
YELLOW CEDAR 1200 (8.27) 800 (5.52) 800 (5.52) 475 (3.28) 625 (4.31) 925 (6.38) 525 (3.62) 250 (1.72)
YELLOW POPLAR 1000 (6.89) 725 (5) 700 (4.83) 400 (2.76) 550 (3.79) 800 (5.52) 450 (3.1) 200 (1.38)
8-2
PDC TR-06-01 Rev 1
September 2008
Select
No. 1 No. 2 No. 3 Stud Construction Standard Utility
Structural
EAST. HEMLOCK-BALSAM
1.2x106 (8274) 1.1x106 (7584) 1.1x106 (7584) 0.9x106 (6205) 0.9x106 (6205) 1.0x106 (6895) 0.9x106 (6205) 0.8x106 (5516)
FIR
EAST. HEMLOCK-TAMARACK 1.2x106 (8274) 1.1x106 (7584) 1.1x106 (7584) 0.9x106 (6205) 0.9x106 (6205) 1.0x106 (6895) 0.9x106 (6205) 0.8x106 (5516)
EASTERN SOFTWOODS 1.2x106 (8274) 1.1x106 (7584) 1.1x106 (7584) 0.9x106 (6205) 0.9x106 (6205) 1.0x106 (6895) 0.9x106 (6205) 0.8x106 (5516)
EASTERN WHITE PINE 1.2x106 (8274) 1.1x106 (7584) 1.1x106 (7584) 0.9x106 (6205) 0.9x106 (6205) 1.0x106 (6895) 0.9x106 (6205) 0.8x106 (5516)
HEM-FIR 1.6x106 (11032) 1.5x106 (10342) 1.3x106 (8963) 1.2x106 (8274) 1.2x106 (8274) 1.3x106 (8963) 1.2x106 (8274) 1.1x106 (7584)
HEM-FIR (NORTH) 1.7x106 (11721) 1.6x106 (11032) 1.6x106 (11032) 1.4x106 (9653) 1.4x106 (9653) 1.5x106 (10342) 1.4x106 (9653) 1.3x106 (8963)
MIXED MAPLE 1.3x106 (8963) 1.2x106 (8274) 1.1x106 (7584) 1.0x106 (6895) 1.0x106 (6895) 1.1x106 (7584) 1.0x106 (6895) 0.9x106 (6205)
MIXED OAK 1.1x106 (7584) 1.0x106 (6895) 0.9x106 (6205) 0.8x106 (5516) 0.8x106 (5516) 0.9x106 (6205) 0.8x106 (5516) 0.8x106 (5516)
NORTHERN RED OAK 1.4x106 (9653) 1.4x106 (9653) 1.3x106 (8963) 1.2x106 (8274) 1.2x106 (8274) 1.2x106 (8274) 1.1x106 (7584) 1.0x106 (6895)
NORTHERN SPECIES 1.1x106 (7584) 1.1x106 (7584) 1.1x106 (7584) 1.0x106 (6895) 1.0x106 (6895) 1.0x106 (6895) 0.9x106 (6205) 0.9x106 (6205)
NORTHERN WHITE CEDAR 0.8x106 (5516) 0.7x106 (4826) 0.7x106 (4826) 0.6x106 (4137) 0.6x106 (4137) 0.7x106 (4826) 0.6x106 (4137) 0.6x106 (4137)
RED MAPLE 1.7x106 (11721) 1.6x106 (11032) 1.5x106 (10342) 1.3x106 (8963) 1.3x106 (8963) 1.4x106 (9653) 1.3x106 (8963) 1.2x106 (8274)
RED OAK 1.4x106 (9653) 1.3x106 (8963) 1.2x106 (8274) 1.1x106 (7584) 1.1x106 (7584) 1.2x106 (8274) 1.1x106 (7584) 1.0x106 (6895)
REDWOOD 1.4x106 (9653) 1.3x106 (8963) 1.2x106 (8274) 1.1x106 (7584) 0.9x106 (6205) 0.9x106 (6205) 0.9x106 (6205) 0.8x106 (5516)
SOUTHERN PINE 1.8x106 (12411) 1.7x106 (11721) 1.6x106 (11032) 1.4x106 (9653) 1.4x106 (9653) 1.5x106(10342) 1.3x106 (8963) 1.3x106(8963)
SOUTHERN PINE (MIXED) 1.6x106 (11032) 1.5x106 (10342) 1.4x106 (9653) 1.2 x106 (8274) 1.2x106 (8274) 1.3x106(8963) 1.2x106 (8274) 1.1x106(7584)
SPRUCE-PINE-FIR 1.5x106 (10342) 1.4x106 (9653) 1.4x106 (9653) 1.2x106 (8274) 1.2x106 (8274) 1.3x106 (8963) 1.2x106 (8274) 1.1x106 (7584)
SPRUCE-PINE-FIR (SOUTH) 1.3x106 (8963) 1.2x106 (8274) 1.1x106 (7584) 1.0x106 (6895) 1.0x106 (6895) 1.0x106 (6895) 0.9x106 (6205) 0.9x106 (6205)
WESTERN CEDARS 1.1x106 (7584) 1.0x106 (6895) 1.0x106 (6895) 0.9x106 (6205) 0.9x106 (6205) 0.9x106 (6205) 0.8x106 (5516) 0.8x106 (5516)
WESTERN WOODS 1.2x106 (8274) 1.1x106 (7584) 1.0x106 (6895) 0.9x106 (6205) 0.9x106 (6205) 1.0x106 (6895) 0.9x106 (6205) 0.8x106 (5516)
WHITE OAK 1.1x106 (7584) 1.0x106 (6895) 0.9x106 (6205) 0.8x106 (5516) 0.8x106 (5516) 0.9x106 (6205) 0.8x106 (5516) 0.8x106 (5516)
YELLOW CEDAR 1.6x106 (11032) 1.4x106 (9653) 1.4x106 (9653) 1.2x106 (8274) 1.2x106 (8274) 1.3x106 (8963) 1.2x106 (8274) 1.1x106 (7584)
YELLOW POPLAR 1.5x106 (10342) 1.4x106 (9653) 1.3x106 (8963) 1.2x106 (8274) 1.2x106 (8274) 1.3x106 (8963) 1.1x106 (7584) 1.1x106 (7584)
8-3
PDC TR-06-01 Rev 1
September 2008
Table 8-3 House Damage from Very Long Duration Blast Loads
House Type Peak Free-Field Observed Damage
Pressure
MPa (psi)
Two-story house and 11.7 to 13(1.7 to 1.9) Minimal damage except failed windows and
one-story house doors
One-story house 18.5 (2.7) 60% damage to roof framing and surface,
20% damage to exterior sheathing and
framing
The BAIT (Blast and Airman Injury Tests) test series was conducted in 2001 on SEA
Huts, which are small wood buildings with light wood construction and no windows
(Marchand, 2002). These huts had 50 mm x 100 mm (2 in x 4 in) wood wall studs and
roof purlins supporting 9.5 mm (5/8 in) plywood. The purlins were supported by wood
trusses, which were supported by the walls. Blast loads were applied from high
explosive charge weights modeling large vehicle bomb threats at typical distances
between military outpost boundaries and base housing. These blast loads had much
shorter durations than those in the 1960s test series modeling nuclear explosions.
Damage to the SEA Huts was characterized by damage and failure of the reflected wall
studs at lower blast loads followed by damage and failure of side-on wall studs and roof
8-4
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
purlins at higher blast loads. The wood roof truss systems on the SEA Hut tests were
not seriously damaged in the tests. Also, there were no overall failures of building lateral
load resisting systems, indicating that the wood stud shear walls were able to resist the
applied side-on blast loads and shear loads transferred from blastward wood stud walls
through the roof diaphragm system.
Figure 8-2 Wood Frame Building After Long Duration Blast Load with 34 KPa (5
psi) Peak Overpressure
The results from these test series on two different types of wood construction with very
long and short duration blast loads both indicate that the wall studs and roof purlins
were the weak part of the overall wood building system against blast load. These
components typically did not transfer large enough dynamic reaction loads into
supporting roof trusses, framing members, or building lateral load resisting system to
cause failure of these systems. Overall roof failure occurred when the blast loads were
large enough to catastrophically fail load-bearing walls. This information is helpful for
SDOF analysis of wood buildings because it indicates that typically such analyses can
focus on the wood studs and roof purlins, since damage to these components will tend
to control the overall building damage.
8-5
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
calculated ductility ratios from the SDOF analysis between 1 and 2. Wood studs that
failed had calculated ductility ratios of 3 and greater.
Therefore, the calculated ductility ratios are consistent with the observed damage for a
relatively brittle material like wood. This consistency gives credence to the assumed
dynamic flexural tensile strength used in the SDOF analyses. A modulus of elasticity of
1.4e6 psi and density of 35 lb/ft2 were also assumed. More details of these analyses are
provided in the CEDAW Methodology Report (Oswald, 2005). No SDOF analyses were
performed of the wood houses from the first test series discussed in the previous
section because fill pressures inside the building complicate the applied net load on the
walls.
Based on available static and shock tube test data, the section modulus should not be
increased to account for composite action with any attached material. Strain
measurements in static tests on wood studs in wood wall systems and on 12.5 mm (0.5
in) plywood connected to the studs with nails showed that only limited composite action
occurred between the plywood and studs (Starr and Krauthammer, 2005). Much higher
strains were measured when the plywood was also attached with adhesive, indicating
much more composite response prior to peak load capacity. However, no difference in
load capacity was measured between walls where the plywood was only nailed to the
studs and where it was nailed and attached with adhesive.
The wood stud walls used for the shock tube tests described in Section 8-3.2 had 12.5
mm (0.5 in) thick plywood cladding nailed on one face or two faces with nails as close
as 3 inches on center. The measured peak dynamic reactions were consistent with an
ultimate resistance calculated with a moment capacity equal to the section modulus of
only the studs and a dynamic yield strength equal to five times the allowable design
flexural strength. Also, similar peak dynamic reaction forces were measured when the
plywood was cut horizontally at midspan so that it could not contribute to the overall
flexural capacity of the wall system and when the cladding was very well nailed to one
and two sides of the studs (Oswald, 2005)
8-6
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
8-5 WOOD CONNECTIONS
Wood components can be directly nailed together or attached with many different types
of metal straps and plates. Figure 8-3 shows hanger straps that are used with nails to
attach wood studs to sill and header boards that form the top and bottom framing for a
wood stud wall. The studs can also be directly nailed to the sill and header boards. Main
framing members and truss members are almost always attached with metal plates that
span across the connected members and are nailed into each member. Allowable load
capacity information for connections with hanger straps and metal plates are provided
by manufacturers.
Figure 8-3 Wood Stud Hanger Connections
8-7
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
8-8
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Dynamic Region
Pressure Region
Line of Constant
Line Deflection
of Constant Damage
Impulse
P-i diagrams with up to four P-i curves can be generated in SBEDS for blast loads with
positive phase blast load only and blast loads with positive and negative phase blast
load. In the case of positive phase blast load only, the blast load histories have a
simplified right triangular shape as shown in Figure 2-3. In the case of positive and
negative phase blast loads, the blast load histories are calculated with charge weight-
standoff combinations and have a shape similar to Equation 2-1. In some cases
involving negative phase blast load, the absolute value of the maximum deflection
9-1
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
during rebound response can exceed that during inbound response. However, P-i
diagrams in SBEDS are always based the inbound maximum deflection.
SBEDS creates the P-i curves by performing repeated SDOF analyses, iterating on the
peak pressure or charge weight depending if the P-i diagram blast loads include
negative phase blast load or not, until the trial SDOF analysis has a maximum deflection
equal to the target maximum deflection within a small tolerance. The peak pressure and
impulse of the blast load from this trial SDOF analysis defines a point on the P-i
diagram. This process is repeated with a different blast load duration or standoff and the
new peak pressure or charge weight causing the target maximum deflection in trial
SDOF analyses is determined until sufficient blast load points are created to form a full
P-i curve. All maximum deflections calculated within this iteration routine in SBEDS use
the same SDOF solver routine used for the Run SDOF function on the Input worksheet.
The only exception is that all cases involving charge weight-standoff combinations do
not consider any clearing of the blast wave and the blast load is either side-on or fully
reflected, depending on user input.
The P-i curve in Figure 9-1 is for the case of positive phase blast load only. It is
characterized by three response regions, known as the impulsive, dynamic, and
pressure sensitive regions. In the impulsive region, the applied impulse causing the
target deflection is constant regardless of the peak pressure. This implies that
component response to blast loads in the impulsive region is only sensitive to the
impulse in the blast loads. In the pressure region, the applied peak pressure causing
the target deflection is constant regardless of the impulse. Component response to
blast loads in the pressure region is only sensitive to the peak pressure in the blast
loads. In the dynamic region, the target deflection is sensitive to both the peak pressure
and the impulse in the applied blast load. As a rule of thumb, a structural component is
sensitive only to the impulse in the blast load if its natural period is at least three times
greater than the blast load duration (td) and a structural component is sensitive only to
the peak pressure of the blast load when td is at least 10 times greater than the natural
period (Tn). The structural component is sensitive to both the peak pressure and
impulse when 1/3 < td/Tn < 10. The natural period is defined in Equation 4-22.
9-1.1 P-I DIAGRAMS FOR BLAST LOADS WITH POSITIVE PHASE LOAD ONLY
The P-i diagram in Figure 9-2 from SBEDS shows all the peak pressures and impulses
of blast load histories with only positive phase load that cause a given target deflection
(i.e., 3.49 inches) for a given component on the Input worksheet. The positive phase
blast loads histories are all assumed to have a simplified right triangular shape shown in
Figure 2-3. The P-i curves are based on 15 points where the blast load duration ranges
from 0.05 to 60 times the natural period of the component. SBEDS iterates on the peak
pressure for each duration, calculating the maximum deflection for each trial peak
pressure, until the maximum deflection for a given trial peak pressure equals the target
maximum deflection within a small tolerance. This peak pressure and the impulse from
this peak pressure and the current blast load duration define a point on the P-i diagram.
Some error may occur in the pressure region shown in Figure 9-1, which is
characterized by long duration blast loads, because small changes in peak pressure
can cause relatively large changes in maximum deflection and this can cause
convergence problems for the iteration algorithm in SBEDS.
9-2
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Figure 9-2 P-i Diagram for Positive Phase Blast Load Only
Max. Defl.=3.49 in
1000
100
Pressure (psi)
10
1
1 10 100 1000 10000
Impulse (psi-ms)
9-1.2 P-I DIAGRAMS FOR BLAST LOADS WITH POSITIVE AND NEGATIVE
PHASE LOAD
P-i diagrams in SBEDS can also be based on blast load histories with positive and
negative phase blast loads, as shown in Figure 9-3. In this case, the blast load points
represent fully reflected blast loads determined for up to twenty-five charge weight-
standoff combinations with standoffs starting at 1.5 m (5 ft). For these cases, SBEDS
determines each blast load point on the P-i curve based on a standoff and for each
standoff iterates on the charge weight, calculating the maximum deflection for the full
positive and negative phase blast load from each trial charge weight until the maximum
deflection equals the target maximum deflection within a small tolerance. The positive
phase peak pressure and positive phase impulse from this trial charge weight-standoff
combination define a point on the P-i diagram. The blast load history calculated for all
charge weight-standoff combinations has the shape shown in Figure 2-1 and Figure 2-2
if negative phase is included.
9-3
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Figure 9-3 P-i Diagram for Positive and Negative Phase Blast Load
Deflections Calculated with Positive & Negative Phase Reflected Loads from Charge Weight-Standoff
C bi ti
Max. Defl.=3.49 in
10000
1000
Pressure (psi)
100
10
1
1 10 100 1000 10000
Impulse (psi-ms)
Comparison of Figure 9-2 and Figure 9-3, which were both generated for the same
target deflection in the same structural component with positive phase blast load only
and with positive and negative blast load, shows that they have the same pressure
asymptote, but different blast loads in the dynamic and impulsive region. Blast loads
with significantly greater positive phase impulse are required in Figure 9-3 to cause the
target deflection in these two regions of the P-i curve when negative phase blast load is
included. The negative phase blast load reduces the maximum deflection from a given
positive phase blast load if it occurs before the component reaches its maximum
deflection, as is the case in the dynamic and impulsive regions of Figure 9-3. Therefore,
more positive phase impulse is needed to cause the same deflection when the negative
phase blast load is included in the SDOF analysis.
Blast load durations are not shown on P-i curves, but the positive phase duration for any
blast load point can be approximated as twice the impulse divided by the peak pressure
(this is the exact duration if the blast load has the shape of a right triangle). The points
nearer the top of the impulsive region of P-i curves have the shortest durations, so that
negative phase blast load affects the component response most for these cases. This
causes this P-i curve in Figure 9-3 to curve to the right in the impulsive region. This
does not occur in Figure 9-2, which is asymptotic in the impulsive region, because no
negative phase blast load is considered. At a short enough duration, P-i curves for blast
loads with positive and negative phase blast load will also become asymptotic in the
impulsive region because both the positive and negative phase blast load will occur
within a time less than one-third of the natural period of the component. However, blast
9-4
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
loads with these durations are outside the realm of cases considered in SBEDS if they
correspond to standoffs less than 1.5 m.
The negative phase blast load can also occur in-phase with the rebound response of
the component for some blast load points on a P-i curve, so that the maximum inbound
deflection occurs during the second inbound phase rather than during initial inbound
response. In this case, the negative phase blast load increases the component
maximum deflection, rather than decreasing it. An example of this is shown in Figure
9-4, where the highlighted blast loads cause the maximum inbound deflections to occur
during the second inbound response cycle as shown in Figure 9-5. This case can occur
in the dynamic region of P-i diagrams with positive and negative phase blast load,
although it is relatively unusual.
Figure 9-4 P-i Diagram with Positive and Negative Phase Blast Load with Points
from Maximum Deflection During Second Inbound Response
Deflections Calculated with Positive & Negative Phase Reflected Loads from Charge Weight-Standoff
C bi ti
Max. Defl.=5.25 in
10000
1000
positive deflection
100 occurs after initial
response cycle
10
1
1 10 100 1000 10000 100000
Impulse (psi-ms)
9-5
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Figure 9-5 Representative SDOF Output for Highlighted Cases in Figure 9-4
Displacement History
10
Displacement (inch)
-5
-10
0 100 200 300 400 500 600 700
Time (ms)
10
Force (psi)
-5
0 100 200 300 400 500 600 700
Time (ms)
The blast load histories for charge weight-standoff combinations with positive phase
blast load only have an exponential decay shape, as shown in Figure 2-1, rather than a
right triangular shape. Otherwise, SBEDS uses the same iteration process and blast
load shapes described previously for P-i diagrams to generate charge weight-standoff
diagrams. SBEDS iterates to find charge weights causing the target inbound maximum
deflection for standoffs starting at 1.5 m for up to twenty-five charge weight-standoff
combinations or until the duration of the equivalent triangular positive phase blast load
is greater than 60 times the natural period of the component. Charge weight-standoff
curves for positive and negative phase blast load where some charge weight-standoff
combinations cause the target maximum inbound deflection during the second response
cycle will have a “hump” in the overall shape of the curve, similar to the hump in the P-i
curve in Figure 9-4.
Deflections Calculated with Positive & Negative Phase Reflected Loads from Charge Weight-Standoff
C bi ti
Max. Defl.=3.49 in
10000
1000
Charge Weight (lb)
100
10
1
1 10 100 1000
Standoff (ft)
P-i diagrams for positive phase blast load only, which are generated assuming the blast
load has the shape of a right triangle, are conservative for positive phase blast loads
with an exponential decay to ambient pressures. P-i diagrams for positive and negative
phase blast load are only valid for cases where the blast load is calculated from high
explosives as described in Chapter 2. They are generated in SBEDS assuming fully
reflected blast loads and therefore there may be some inaccuracy if they are applied for
cases with side-on blast loads. Charge weight-standoff diagrams can be determined for
side-on blast loads, including negative phase load if designated. All charge weight-
standoff diagrams for reflected blast loads are fully reflected blast loads without any
clearing effects.
Additionally, P-i diagrams and charge-weight standoff diagrams are subject to error from
any incorrect convergence in the iteration method in SBEDS to determine the peak
pressure or charge weight causing the target deflection. Errors occur when a time step
9-7
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
is too small for the SDOF analyses used in the iteration process to reach maximum
deflection and when the time step is too large compared to the duration of a positive
phase blast load. SBEDS uses logic that is intended to prevent both types of time step
problems, but it is not always possible for some blast load points and input components.
An error message will be displayed if the time step used in the SDOF analysis to
determine a point on the P-i or Charge Weight-Standoff diagrams is less than 10% of
the equivalent triangular duration of the blast load. An error message will also be
displayed if there is no calculated rebound deflections in the SDOF analysis used to
determine a point on the diagrams, indicating that the time step was too small to allow
the maximum deflection to be calculated. Convergence error can also occur
occasionally when the iteration routine gets stuck oscillating about the target maximum
deflection, rather than converging on it. A maximum of twelve iteration cycles is
performed for any blast load point, after which iteration is terminated and the last peak
pressure or charge weight from the final iteration is used to determine the point on the
diagram. A ratio of the actual calculated deflection in the iteration routine to the target
deflection is displayed for each blast load point in SBEDS to the right of the plots on the
P-i_Diagram worksheet.
Points on a P-i diagram or charge weight-standoff diagram that do not lie along a
smooth curve defined by surrounding points may be in error. In these cases, the curve
can be smoothed to obtain a corrected curve by ignoring outlying points. Any point on a
P-i diagram or charge weight-standoff diagram can be verified by inputting the load
corresponding to that point on the Input worksheet for exactly the same component that
the diagram was developed for, using the Runs SDOF button to determine the
maximum inbound deflection, and comparing it to the target deflection. It is important to
use the same type of load (i.e., with or without negative phase load and fully reflected
without clearing or side-on for a charge-weight standoff diagram) as specified for the P-i
or charge weight-standoff diagram when it was generated.
Charge weight-standoff input should be used on the Input worksheet if the point in
question was generated for a charge weight-standoff diagram or for a P-i diagram that
included negative phase load. The charge weight and standoff corresponding to all blast
load points that include negative phase blast load are shown to the right of the P-i
diagram on the P-i_diagram worksheet. A manually input pressure-time history with a
right triangular shape and a peak pressure and impulse corresponding to the point in
question should be used on the Input worksheet if the point in question was generated
for a P-i diagram with positive phase blast load only. In all cases, the calculated
maximum positive deflection for the blast load on the Input worksheet corresponding to
the point of interest should be within approximately five percent of the target deflection
for the P-i or charge weight-standoff curve.
Also, charge weight-standoff diagrams for side-on blast load can have an unusual
shape at small standoffs caused by the shape of the side-on blast in Figure 2-9 at small
scaled standoffs. The side-on impulse remains constant with decreasing scaled
standoff, and even decreases somewhat, at scaled standoffs between approximately
0.2 and 1.0 m/kg0.33 (0.5 and 1.5 ft/lb0.33) rather than increasing monotonically with
decreasing standoff as it does for all other scaled standoffs, as shown in Figure 2-9.
9-8
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
This can cause a relatively flat region in the charge weight-standoff diagram at small
scaled standoffs.
Figure 9-7 shows a scaled P-i diagram for reinforced concrete slabs, where the
dimensionless Pbar and Ibar terms are defined in Equation 9-1. It has P-i curves
showing all Pbar and Ibar values causing support rotations of 2, 5, and 10 degrees in a
reinforced concrete slab. The Y factor in Equation 9-1 accounts for the effect of
negative phase loading. If the peak positive phase pressure and impulse for a blast load
and all the properties of the equivalent SDOF system for the component in Equation 9-1
are known, then Pbar and Ibar terms for the blast loaded component can be calculated
and the support rotation of the component can be determined based on the location of
the (Pbar, Ibar) point in Figure 9-7 relative the P-i curves for the three support rotations.
For example, if the (Pbar, Ibar) point lies halfway between the P-i curves for 2 and 5
degrees of support rotation, then the deflection of the component from the blast load
has a 3.5 degree support rotation. This can be verified by a SDOF analysis.
Scaled P-i diagrams, such as that in Figure 9-7, have been developed for many
common building types in the CEDAW (Component Explosive Damage Assessment
Workbook) methodology developed for the U.S. Army Corps of Engineers (Oswald,
2005). In the first step of the development process, conservation of energy equations
that transform the peak pressure and positive phase impulse into the scaled blast load
terms Pbar and Ibar, respectively, were developed. The energy equations were based
on strain energy from the dominant response modes that affects given component
types, including flexure, tension membrane, and arching from axial load, and the non-
dimensional response terms that best correlates component response to damage (i.e.,
ductility ratio or support rotation). Essentially, the Pbar and Ibar terms normalize the
applied blast load on a component by properties of that component so that two different
components with the same non-dimensional response to separate blast loads lie along
the same scaled P-i curve.
9-9
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Figure 9-7 Scaled P-i Curve-fits vs. Scaled SDOF Points in Terms of Support
Rotation for Flexural Response of Reinforced Concrete Slabs
1000
VLLOP
θ=2
100
θ=5
Pbar
θ=1 0
10
0.1
0.01 0.1 1 10
Ibar
Equation 9-1
Pbar (0 .5−0 .39 Rbar )
−0 .127
P 1
Pbar = Ibar = i Y Y=
Ru K LM mRu L 2.59 Rbar 0 .067
where:
P = peak positive phase pressure
i = applied positive phase impulse
m = mass of equivalent SDOF system for component
KLM = load-mass factor of equivalent SDOF system for component
Ru = component ultimate flexural resistance
L = component span length
Rbar = Ru/Pa
Pa = atmospheric pressure
After the equations for Pbar and Ibar terms were developed, these terms were used to
scale the blast loads with a wide range of durations that all caused a given value of a
non-dimensional response parameter in SDOF analyses of a representative component
of the given component type. The positive phase peak pressure and impulse from each
of these blast loads were divided by properties of the component used in the SDOF
analyses to calculate Pbar and Ibar values, which defined points on a P-i diagram that
were connected to form a P-i curve for the given value of the nondimensional response
parameter. For example, each point along a curve in Figure 9-7 represents a scaled
blast load causing the given support rotation (θ) to any reinforced concrete component
with a blast load and component properties such that Pbar and Ibar for the blast-loaded
component lie on the curve. If the Pbar and Ibar terms are calculated correctly, peak
pressures and impulses acting on many different components of the given component
type that all cause component response with the same value of the given non-
dimensional response parameter for which the Pbar and Ibar terms were developed,
9-10
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
can be scaled using the Pbar and Ibar terms and these Pbar and Ibar points will all lie
along a single response curve on a scaled P-i diagram.
A scaled P-I diagram can be developed using blast loads either with, or without negative
phase blast load. The P-I diagram in Figure 9-7 was developed for blast loads including
negative phase. Scaled P-i curves that include the effect of the negative phase blast
load have the same basic shape as dimensional P-i curves that include the effect of the
negative phase blast load, as discussed in Section 9-1.2 . The inclusion of negative
phase must be considered in the development of the Pbar and/or Ibar terms. This effect
was included in the Ibar value for CEDAW using a curve-fitting approach to develop a
factor, Y, which is applied to the Ibar term. Y is a function of Pbar, and causes the Ibar
term to increase at high Pbar values relative to the ultimate resistance of the
component, which is the region of the P-i curve affected by negative phase blast load.
The scaled P-i diagrams in CEDAW have P-i curves that define upper and lower bounds
to given damage levels. Test data was used to establish the support rotation and
ductility ratio values for scaled upper and lower bound P-i curves for each damage level.
More information on the development of the scaled P-i curves and the use of test data
to establish scaled P-i curves bounding different damage levels for CEDAW is
described elsewhere (Oswald, 2005), (Oswald and Nebuda, 2006). Scaled P-i
diagrams were also developed to predict blast damage to conventional building
components for the Facility and Component Explosive Damage Assessment Program
(FACEDAP) computer program, which was developed for the U.S. Army Corps of
Engineers in the early 1990’s (Oswald, 1993). However, these scaled P-i diagrams
involved empirical modifications to match data, where whole P-i curves were moved
from their theoretically derived locations to match available scaled data points.
Fortunately, this was not necessary for the CEDAW P-i diagrams due to more carefully
developed Pbar and Ibar terms that include strain energy from varying response modes,
the inclusion of negative phase loading effects, and the consideration of response in
terms of support rotation where applicable.
9-11
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
9-12
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
CHAPTER 10 - REFERENCES
GOVERNMENT PUBLICATIONS:
NON-GOVERNMENT PUBLICATIONS:
10-1
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Farmington Hills, MI 48333-9094 2002
Baker, W. E., Cox, P. A., Westine, P. S., Kulesz, J. J., and Strehlow, R. A.,
Explosion Hazards and Evaluation, Elsevier Scientific Publishing Company, New
York, NY, ISBN 0-444-42094-0, 1983.
Britt, J. R., Ranta, D.E., and Joachim, C.E., "BLASTX Code, Version 4.2, User's
Manual", Report ERDC/GSL TR-01-2 Distributed by U.S Army Engineer
Research and Development Center, Vicksburg, MS, March, 2001.
Coltharp, D.R., Simmons, L., and Bogosian, D.D., “Blast Response Evaluation of
a Lightweight Steel Roof,” Proceedings of the 9th International Symposium on
Interaction of the Effects of Munitions with Structures, Berlin-Strausberg, Federal
Republic of Germany, May, 1999.
10-2
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Granstrom, S.A., “Loading Characteristics of Air Blasts from Detonating
Charges,” Report No. 100, Transactions of the Royal Institute of Technology,
Stockholm, Sweden, 1956.
Hamid, A.A. and Drysdale, R.G., “Flexural Tension Strength of Concrete Block
Masonry,” Journal of Structural Engineering, Vol. 144, No. 1, January, 1988.
Hamid, A.A. Chia-Calabria, C., and Harris, H.G., “Flexural Behaviour of Joint
Reinforced Block Masonry Wall,” ACI Structural Journal, January-February 1992.
Ketchum, D.E., Oswald, C.J., and Clutter, J.K, “The Applicability of Crα Blast
Pressure Estimates to Structural Analysis,” Proceedings from the 28th DDESB
Seminar, Orlando, FL, August, 1998.
Thomas, J.T., Geng, J., and Hallgarth, A.D. “Evaluation of Door Blast Shields,”
Proceedings from the 31rst DDESB Seminar, San Antonio, TX, August, 2004.
Marchand, K. A., “BAIT, BASS & RODS Testing Results,” Prepared for the USAF
Force Protection Battlelab by Applied Research Associates, Report within 12-CD
set issued by The Technical Support Working Group (TSWG) and Defense
Threat Reduction Agency (DTRA), April, 2002.
Needam, C.E., and Witttwer, L.A., “Low Altitude Multiple Burst Model, S-Cubed
Report No. 84-6402, 1977.
10-3
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
Army Corps of Engineers Protective Design Center, Contract No. DACA45-01-D-
0007-0013, May, 2005.
Oswald, C.J., Holgado, D., “Shock Tube Testing on Reinforced Masonry Walls,”
Prepared for Protective Design Center, U.S. Army Corps of Engineers, Omaha
District by Wilfred Baker Engineering, Inc., WBE Report No. 0724-001, August,
2004.
Oswald, C.J. and Nebuda, D, ”SBEDS (Single degree of freedom Blast Effects
Design Spreadsheets),” Proceedings from the 31st DoD Explosives Safety
Seminar, San Antonio, Texas, August, 2004.
Park and Gamble, Reinforced Concrete Slabs, John Wiley and Sons, New York,
NY, 2000.
Polcyn, Michael A., Oswald, Charles J., and Conrath, Edward J., "Response
Limits for Masonry and Cold-Formed Steel Structural Components," presented at
the Thirty-First United States Department of Defense Explosives Safety Seminar,
August, 2004.
Salim, H.A., Dinan, R., Kiger, S.A., Townsend, P.T., and Shull, J., “Blast-Retrofit
Wall Systems Using Cold-Formed Steel Studs,” 16th ASCE Engineering
Mechanics Conference, University of Washington, Seattle, July 16-18, 2003.
Wesevich, J.W., Oswald, C. J., Edel, M.T., Lowak, M. J., Alaoui, S. S., ”Compile
and Enhance Blast Related CMU Wall Test Database (CMUDS),” Prepared for
the Office of Special Technology by Baker Engineering and Risk Consultants,
Inc., Contract No. 41756-00-C-0900, September 6, 2002.
10-4
Methodology Manual TOC
PDC TR-06-01 Rev 1
September 2008
\\