Sie sind auf Seite 1von 285

The Design of Civilian Structures Subject to External Blast Loading

An Undergraduate Honours Thesis Submitted for the Bachelor Degree in Civil Engineering

Asher Gehl 9913569 October 2004

Abstract

Abstract
The paper is concerned with the prediction of dynamic effects of unconfined explosions needed for the structural analysis of blast-loaded aboveground structures. A number of case histories are presented, which demonstrate the devastating effects of an explosion on a structure. The basic features of the explosion and blast wave phenomena are presented along with a discussion of TNT equivalency and blast scaling laws. The characteristics of incident overpressure loading due to conventional high explosives are addressed as well as the description of the other blast loading components associated with air flow and reflection process. This is followed by the derivation of external blast loads on the different faces of aboveground rigid structures. The determination of the response of reinforced concrete and structural steel elements is then presented by converting the structure into an equivalent Single Degree of Freedom system and taking into account the dynamic effects of the load upon the material properties. A preliminary design is then presented for columns of varying cross-sectional shape under the impact of a conventional explosive load. Finally the overall global assessment of the building is presented with an outline of a number of Progressive Collapse design guidelines.

Asher Gehl 9913569

Acknowledgements

Acknowledgments
Firstly I would like to thank my supervisor Associate Professor Kim Rasmussen for his guidance, patience, advice and encouragement, throughout the duration of this study. Most of all I would like to express my appreciation for the permission to undertake this thesis topic, given that there was no clearly defined scope at the onset of the study. My sincere thanks also go to Dr Andy Davids and Jonathan Wongso of Hyder Consulting for their generosity in giving of their time to discuss the scope of work and for providing me with both a greater understanding of the topic and the core material utilised in this thesis. Their knowledge was invaluable and of great assistance, without which this thesis would not have been possible. I would also like to thank my thesis partner Adam Ritchie, for joining me in this endeavour, knowing full well the difficulties that lay ahead in acquiring the research material and grasping the fundamental concepts involved. Finally I would like to express my appreciation to my girlfriend for her support and understanding this year and in particular the many hours spent reading and editing this thesis. I am forever grateful for this selfless effort on her behalf.

Asher Gehl 9913569

Table of Contents

Table of Contents
ABSTRACT................................................................................................... 2 ACKNOWLEDGMENTS ............................................................................ 3 TABLE OF CONTENTS ............................................................................. 4 TABLE OF FIGURES AND TABLES ..................................................... 12 NOTATIONS .............................................................................................. 16
ROMAN UPPER CASE LETTERS ....................................................................................... 16 ROMAN LOWER CASE LETTERS ..................................................................................... 19 GREEK LETTERS............................................................................................................. 21

CHAPTER 1 INTRODUCTION............................................................... 22
1.1 KEY CONCEPTS AND AIMS ....................................................................................... 22 1.2 RESEARCH AREA/LIMITATIONS ............................................................................... 23 1.3 OUTLINE OF CONTENTS............................................................................................ 23

CHAPTER 2 BACKGROUND ................................................................. 25


2.1 BACKGROUND AND AIM .......................................................................................... 25 2.2 RISK ANALYSIS AND THREAT DETERMINATION ....................................................... 26 2.3 PAST ATTACKS ........................................................................................................ 31 2.3.1 Ronan Point London 1968 ............................................................................... 32 2.3.2 Exchequer Court, St Marys Axe, London 1992 ............................................. 34 2.3.3 AMIA Building: Buenos Aires, Argentina 18th July 1994 .............................. 37 2.3.4 A.P Murrah Federal Building Oklahoma 1995................................................ 39 2.3.5 Khobar Towers Saudi Arabia 1996 ................................................................. 41 2.3.6 Australian Embassy Jakarta 9 September 2004 ............................................... 42

Asher Gehl 9913569

Table of Contents

CHAPTER 3 UNIQUE CHARACTERISTICS....................................... 43


3.1 DURATION ............................................................................................................... 43 3.2 PLASTIC BEHAVIOUR ............................................................................................... 44 3.3 MAGNITUDE ............................................................................................................. 44 3.4 STRUCTURAL DAMAGE ............................................................................................ 45 3.5 DESIGN PROCEDURE ................................................................................................ 45

CHAPTER 4 BLAST, STRESS AND SHOCK WAVE THEORY........ 48


4.1 INTRODUCTION ........................................................................................................ 48 4.2 FORMATION OF BLAST WAVES ................................................................................ 49 4.3 DAMAGING PROPERTIES OF BLAST WAVE ............................................................... 50 4.4 BLAST SCALING ....................................................................................................... 53 4.5 TNT EQUIVALENCE ................................................................................................. 54 4.6 PRESSURE TIME HISTORY ........................................................................................ 55 4.7 BLAST WAVE PRESSURE PROFILES .......................................................................... 56 4.7.1 Theoretical Analysis ........................................................................................ 57 4.7.2 Friedlander Wave Equation ............................................................................. 58 4.7.3 Idealised Pressure-Time History...................................................................... 58 4.8 BLAST WAVEFRONT PARAMETERS .......................................................................... 60 4.8.1 Techniques Available ...................................................................................... 60 4.8.2 Predicting Blast Parameters: Rankine Hugoniot relations............................... 63 4.8.3 Peak Incident Overpressure ............................................................................. 65 4.8.4 Positive Phase Duration ................................................................................... 68 4.8.5 Blast Wave Impulse ......................................................................................... 69 4.9 TYPE OF EXPLOSION ................................................................................................ 70 4.9.1 Air Burst .......................................................................................................... 70 4.9.2 Surface Burst.................................................................................................... 71 4.9.3 Confined Explosion ......................................................................................... 71

Asher Gehl 9913569

Table of Contents 4.10 BLAST WAVE REFLECTIONS .................................................................................. 72 4.10.1 Normal Reflection.......................................................................................... 72 4.10.2 Regular Reflection ......................................................................................... 76 4.10.3 Mach Stem Formation ................................................................................... 77 4.11 BLAST WAVE DIFFRACTION................................................................................... 80

CHAPTER 5 BLAST WAVE LOADING ................................................ 82


5.1 INTRODUCTION ........................................................................................................ 82 5.2 DYNAMIC SHOCK WAVE EFFECTS AND PRINCIPLES................................................. 85 5.2.1 Stagnation Pressure.......................................................................................... 86 5.2.2 Dynamic Pressure ............................................................................................ 87 5.2.3 Drag Coefficients............................................................................................. 89 5.3 UNIFORM BLAST PRESSURES ................................................................................... 91 5.3.1 Front Face Loads ............................................................................................. 93 5.3.2 Rear Face Loads............................................................................................... 96 5.3.3 Side and Roof Loads........................................................................................ 97 5.3.4 Overall Loading on Structure/Frame ............................................................... 99 5.4 BLAST LOADING ON INDIVIDUAL COLUMNS .......................................................... 102

CHAPTER 6 STRUCTURAL RESPONSE ........................................... 104


6.1 INTRODUCTION ...................................................................................................... 104 6.2 PRINCIPLES OF DYNAMIC ANALYSIS ...................................................................... 105 6.2.1 Fundamental Characteristics of Vibration ..................................................... 105 6.2.2 Fundamental Properties ................................................................................. 106 6.3 SINGLE DEGREE OF FREEDOM SYSTEM .................................................................. 107 6.3.1 Equivalent Systems........................................................................................ 107 6.3.2 SDOF System ................................................................................................ 108 6.3.2.1 Derivation ............................................................................................... 110

Asher Gehl 9913569

Table of Contents 6.3.3 Dynamic Design Factors................................................................................ 113 6.3.3.1 Introduction............................................................................................. 113 6.3.3.2 Load Factor ............................................................................................. 114 6.3.3.3 Mass Factor............................................................................................. 115 6.3.3.4 Resistance Function ................................................................................ 116 6.3.3.5 Load-Mass Factor ................................................................................... 117 6.3.3.6 Natural Period of Vibration .................................................................... 117 6.4 QUASI-STATIC VS. IMPULSIVE LOADING ................................................................ 118 6.4.1 Quasi-Static Loading ..................................................................................... 119 6.4.2 Impulsive Loading ......................................................................................... 120 6.5 PRESSURE IMPULSE DIAGRAM ............................................................................ 123 6.6 RESISTANCE DEFLECTION FUNCTION .................................................................. 125 6.6.1 Ultimate Resistance ....................................................................................... 127 6.6.2 Ultimate Deflection........................................................................................ 128 6.6.3. Stiffness ........................................................................................................ 128 6.7 DEFORMATION LIMITS ........................................................................................... 130 6.7.1 Resistance Deflection Functions for Limited Deflections ............................. 133 6.7.2 The Resistance Deflection Function for Large Deflections........................... 134 6.8 MATERIAL PROPERTIES.......................................................................................... 135 6.8.1 Structural Behaviour of Reinforced Concrete ............................................... 136 6.8.1.1 General Behaviour .................................................................................. 136 6.8.1.2 Types of Blast Links ............................................................................... 142 6.8.1.2.1 Lacing Reinforcement...................................................................... 142 6.8.1.2.2Single Leg Stirrups ........................................................................... 143 6.8.1.3 Modes of Structural Behaviour............................................................... 144 6.8.1.3.1 Ductile Mode of Behaviour ............................................................. 144 6.8.1.3.2 Brittle Failure Modes ....................................................................... 145

Asher Gehl 9913569

Table of Contents 6.8.1.4 Strain Rate Effects on Concrete.............................................................. 146 6.8.1.4.1 Stress-Strain Curve .......................................................................... 147 6.8.1.4.2 Allowable Material Strengths .......................................................... 148 6.8.1.4.3 Dynamic Design Stresses for Reinforced Concrete......................... 149 6.8.1.4.4 Dynamic Design Stresses................................................................. 154 6.8.1.4.5 Moment of Inertia ............................................................................ 155 6.8.1.4.6 Changes in the Failure Mode due to Dynamic Loading .................. 157 6.8.2 Structural Steel Behaviour ............................................................................. 158 6.8.2.1 General Characteristics ........................................................................... 158 6.8.2.2 Development of Plastic Design Moment ................................................ 158 6.8.2.3 Strain Rate Effects .................................................................................. 161 6.8.2.4 Dynamic Design Stresses........................................................................ 163 6.8.2.4.1 Dynamic Design Stress for Protection Category 1 .......................... 163 6.8.2.4.2 Dynamic Design Stress for Protection Category 2 .......................... 164 6.8.2.4.3 Dynamic Design Stress for Shear .................................................... 164 6.8.3 Flexural Design Procedure............................................................................. 164 6.8.3.1 Reinforced Concrete Elements ............................................................... 165 6.8.3.2 Structural Steel Elements........................................................................ 169 6.8.4 Differences between Steel and Concrete Structures ...................................... 171

CHAPTER 7 LOCAL COLUMN DESIGN PROCEDURE ................ 173


7.1 AIM ........................................................................................................................ 173 7.2 DESIGN LOADS AND SAFETY FACTORS .................................................................. 173 7.3 DRAG COEFFICIENTS.............................................................................................. 174 7.4 REINFORCED CONCRETE COLUMN PROTECTION CATEGORY 1............................. 178 7.5 REINFORCED CONCRETE COLUMN PROTECTION CATEGORY 2............................. 192 7.6 STRUCTURAL STEEL PROTECTION CATEGORY 1 ................................................. 207 7.7 RESULTS ................................................................................................................ 219 7.8 DISCUSSION ........................................................................................................... 219

Asher Gehl 9913569

Table of Contents

CHAPTER 8 PROGRESSIVE COLLAPSE ......................................... 220


8.1 INTRODUCTION ...................................................................................................... 220 8.2 HISTORY OF PROGRESSIVE COLLAPSE ................................................................... 221 8.2.1 Ronan Point, London 1968 ............................................................................ 221 8.2.2 Murrah Building Oklahoma 1995.................................................................. 222 8.2.3 World Trade Center Tower 1 and 2, New York 2001 ................................... 223 8.3 DESIGN METHODS TO REDUCE PROGRESSIVE COLLAPSE....................................... 224 8.3.1 Redundancy or Alternate Load Paths ............................................................ 224 8.3.2 Local Resistance ............................................................................................ 225 8.3.3 Interconnection or Continuity ........................................................................ 226 8.4 CURRENT CODES AND GUIDELINES ........................................................................ 226 8.4.1 Introduction.................................................................................................... 227 8.4.2 Minimum Design Loads for Buildings and Other Structures, ASCE 7-02 by the American Society of Civil Engineers ..................................................................... 227 8.4.3 Building Code Requirements for Structural Concrete (ACI 318-02) and Commentary (ACI 318R-02) by the American Concrete Institute......................... 230 8.4.4 GSA PBS Facilities Standards 2000 .............................................................. 231 8.4.5 GSA PBS Facilities Standards 2003 .............................................................. 232 8.4.6 GSA Progressive Collapse Guidelines 2003.................................................. 232 8.4.7 The UK Building Regulations 1991 (PART A)............................................. 237 8.4.8 National Building Code Of Canada ............................................................... 240 8.4.9 Swedish Building Code ................................................................................. 242 8.5 FURTHER ANALYSIS ............................................................................................... 243

CHAPTER 9 CONCLUSIONS ............................................................... 245


9.1 GENERAL CONCLUSIONS ........................................................................................ 245 9.2 FUTURE RESEARCH ................................................................................................ 248

BIBLIOGRAPHY..................................................................................... 253

Asher Gehl 9913569

Table of Contents

APPENDIX A ............................................................................................ 263


TRANSFORMATION FACTORS FOR ONE WAY ELEMENTS ............................................. 263 TABLE A1: Transformation factors for One Way Elements ................................. 263 TABLE A2: General and Ultimate Deflections for One Way Elements ................ 264 TABLE A3: Elastic, Elasto-Plastic and Equivalent Elastic Stiffness for One Way Elements.................................................................................................................. 265 TABLE A5: Ultimate unit resistance for One Way Elements................................ 267

APPENDIX B ............................................................................................ 268


DYNAMIC DESIGN FACTORS............................................................................... 268 TABLE B1: Dynamic Increase Factor (DIF) for Reinforced Concrete Elements.. 268 TABLE B2: Dynamic Design Stresses for Design of Reinforced Concrete Elements ................................................................................................................................ 269 TABLE B3: Dynamic Increase Factor, c, for Yield Stress of Structural Steels ..... 270 TABLE B4: Dynamic Increase Factor, c, for Ultimate Stress of Structural Steels 270

APPENDIX C ............................................................................................ 271


MAXIMUM SINGLE DEGREE OF FREEDOM RESPONSE CHARTS .................................... 271 TABLE C1: Maximum response of elastic Single-Degree-of-Freedom system for triangular load ......................................................................................................... 271 TABLE C2: Maximum deflection of Elasto-Plastic, Single-Degree-of-Freedom system for triangular load ....................................................................................... 272 TABLE C3: Maximum response time of Elasto-Plastic, Single-Degree-of-Freedom system for triangular load ....................................................................................... 273

APPENDIX D ............................................................................................ 274


DRAG COEFFICIENT TABLES ........................................................................................ 274 TABLE D1: Drag Coefficient, CD, for various shapes........................................... 274 TABLE D2: Drag Coefficient, CD, for various shapes........................................... 275

Asher Gehl 9913569

10

Table of Contents

APPENDIX E ............................................................................................ 276


RESULTS FROM CONWEP .............................................................................................. 276 TABLE E1: Output Data from CONWEP for a 100kg weight of ANFO, Hemispherical Surface Burst at a Standoff Distance of 5m ................................... 276 TABLE E2: Conwep Screen Display ..................................................................... 277

APPENDIX F ............................................................................................ 280


DESIGN FLOW CHART ........................................................................................... 280 Figure F1: Design Process Flow Chart ................................................................... 280 (G.C Mays and P.D Smith 1995, Blast Effects on Buildings)................................ 280

APPENDIX G............................................................................................ 281


RESULTS FOR DESIGN OF COLUMNS OF VARYING CROSS SECTIONAL SHAPE SUBJECT TO BLAST LOAD .......................................................................................................... 281 TABLE G1: Output Data from MathCAD Worksheet 1 ........................................ 281 TABLE G2: Output Data from MathCAD Worksheet 2 ........................................ 282 TABLE G3: Output Data from MathCAD Worksheet 3 ........................................ 282 CHART G1: Results from MathCAD Worksheet 1 ............................................... 283 REINFORCED CONCRETE PROTECTION CATEGORY 1.............................. 283 REINFORCED CONCRETE PROTECTION CATEGORY 2.............................. 284 CHART G3: Results from MathCAD Worksheet 3 ............................................... 285 STRUCTURAL STEEL PROTECTION CATEGORY 1...................................... 285

Asher Gehl 9913569

11

Table of Figures and Tables

Table of Figures and Tables


Figure 2.1 Damage level curves for given Stand-off distance and Explosive weight (Federal Emergency Management Agency (FEMA), 2003)............................................. 29 Figure 2.2 Ronan Point Apartment Complex Collapse (Moore, 2002) ............................ 32 Figure 2.3 Damage of the Surrounding Buildings (Krauthammer, 2004) ........................ 34 Figure 2.4 Damage to ground floor steel columns Exchequer Court, St. Marys Axe (Moore, 2002) ................................................................................................................... 35 Figure 2.5 Damage to composite floors Exchequer Court, St. Marys Axe (Moore, 2002) ................................................................................................................................. 36 Figure 2.9 Aerial View of damage to the Murrah Building (Krauthammer, 2004).......... 39 Figure 2.10 Extent of Structural Collapse of Murrah Building (Krauthammer, 2004) .... 40 Figure 2.11 Damage to Khobar towers (Krauthammer, 2004) ......................................... 41 Figure 2.12 Crater SizeKhobar towers (Krauthammer, 2004)........................................ 41 Figure 4.1 Shockwave of an explosion ............................................................................. 49 Figure 4.2 Variations of Overpressure and Dynamic Pressure with Time (NATO FM 8-9, 1996) ................................................................................................................................. 51 Figure 4.3 Variations of Blast Effects Associated with Positive and Negative Phase Pressures with Time (NATO FM 8-9, 1996) .................................................................... 52 Figure 4.4 Pressure Time History (TM5-1300, 1991) ...................................................... 55 Figure 4.5 Idealised Pressure time Variation (TM5-1300, 1991)..................................... 59 Figure 4.6 Positive Phase shock wave parameters for a hemispherical TNT explosion on the surface at sea level. (TM5-1300, 1991) ...................................................................... 62 Figure 4.7 Model for one-dimensional shock waves (Leppanen, 2002)........................... 63 Table 4.1 Summary of Blast Wave investigations (Teo 2003)......................................... 67 Figure 4.11 Blast Wave overpressure of Nuclear and Chemical Explosions of equal duration (Kinney Graham)................................................................................................ 69 Figure 4.12 Air Burst Blast environment (TM5-1300, 1991)........................................... 70 Figure 4.13 Surface Burst Blast Environment (TM5-1300, 1991) ................................... 71 Figure 4.14 Normal reflections in air from a rigid surface (Baker, 1983)........................ 72 Figure 4.15 Peak Incident Pressure vs. Ratio of Normally Reflected Pressure/Incident Pressure for a Free Air Burst (TM5-1300, 1991) ............................................................. 73

Asher Gehl 9913569

12

Table of Figures and Tables Figure 4.16 Reflection Coefficient v.s Angle of incident shock (Kinney and Graham,
TU

1985) ................................................................................................................................. 75
UT

Figure 4.17 Reflected Pressure Coefficient vs. Angle of incidence (TM5-1300, 1991) .. 75
TU UT

Figure 4.18 Oblique reflection (Baker 1983).................................................................... 76


TU UT

Figure 4.19 Variation of reflected pressure as a function of angle of incidence (TM5TU

1300, 1991) ....................................................................................................................... 77


UT

Figure 4.20 Transition angle from regular reflection to Mach stem formation (Kinney and
TU

Graham, 1985) .................................................................................................................. 78


UT

Figure 4.21 Mach stem formation (Baker 1973) .............................................................. 79


TU UT

Figure 4.22 Reflection of strong Shock Waves from an air burst (Baker 1983) .............. 79
TU UT

Figure 4.23 Blast Wave Diffracting Around a Roof (Armstrong, Rickman, Baylot, and
TU

Bevins, 2002) .................................................................................................................... 81


UT

Figure 5.1 Blast pressure effects on a building (FEMA 426, 2003) ................................. 82
TU UT

Figure 5.2 Stagnation effects (www.princeton.edu, 2004) ............................................... 86


TU UT

Figure 5.3 Peak incident pressure (Pso) versus peak dynamic pressure (qs), density of air
TU UB UB UB UB

behind the shock front(s), and particle velocity(Us) (TM5-1300, 1991) ......................... 88
UB UB UB UB UT

Figure 5.6 Flow at low Reynolds Number (a) and High Reynolds Number (b) (Smits,
TU

2004) ................................................................................................................................. 91
UT

Figure 5.8 Plane Wave front of incident wave (TM5-1300, 1991) .................................. 92
TU UT

Figure 5.9 Loading on a front wall (TM5-1300, 1991) .................................................... 93


TU UT

Figure 5.10 Rear Wall Loading (TM5-1300, 1991) ......................................................... 96


TU UT

Figure 5.11 Roof and Side wall loading (TM5-1300, 1991) ............................................ 98
TU UT

Table 5.1 Drag coefficient Roof and Side Walls (TM5-1300, 1991) ............................... 99
TU UT

Figure 5.12 Load curves for the front and rear faces (a) and overall effect on the frame (b)
TU

(Kinney Graham, 1985) .................................................................................................. 100


UT

Figure 5.13 External force-time profile for a structure (Mays and Smith, 1995) ........... 100
TU UT

Figure 5.14 Loading of individual element..................................................................... 103


TU UT

Figure 6.1: Real and Equivalent SDOF system (Mays and Smith, 1995) ...................... 108
TU UT

Figure 6.2 Idealized triangular Blast Load ..................................................................... 110


TU UT

Figure 6.3: Single Degree of Freedom model (Conrath et al, 1999) .............................. 111
TU UT

Figure 6.4 Quasi-static loading .................................................................................... 119


TU UT

Asher Gehl 9913569

13

Table of Figures and Tables Figure 6.5 Impulsive loading .......................................................................................... 120 Figure 6.8: Typical Concrete Resistance Deflection function (Mays and Smith, 1995) 126 Figure 6.9 Resistance Deflection Functions for I, II and III step systems (TM5-1300, 1991) ............................................................................................................................... 129 Figure 6.10: Member end Rotations for beams and frames (TM5-1300, 1991)............. 131 Figure 6.11 Relationship between design parameters and Protection category.............. 132 (TM5-1300, 1991)........................................................................................................... 132 Figure 6.12 The Idealised Resistance Deflection function for limited deflections......... 134 Figure 6.13: The Resistance Deflection function for large deflections .......................... 135 Figure 6.14 Resistance deflection curve for flexural response of Concrete (TM5-1300, 1991) ............................................................................................................................... 137 Figure 6.16 Idealized resistance deflection function (TM5-1300, 1991) ....................... 140 Figure 6.18 Single Leg Stirrups (TM5-1300, 1991) ....................................................... 143 Figure 6.19 Post Failure Fragments in Laced and Unlaced Sections (TM5-1300, 1991) ......................................................................................................................................... 146 Figure 6.21 Stress Strain curves for standard and rapid strain rates of concrete (a) and steel (b) (TM5-1300, 1991)............................................................................................. 147 Figure 6.22 DIF vs. Strain Rate for concrete and Reinforcement Steel (TM5-1300, 1991) ......................................................................................................................................... 150 Figure 6.23 Relative increases in the ultimate uniaxial compressive strength as a function of the strain rate-(Bischoff and Perry.1991) ................................................................... 151 Figure 6.24: Strain rate dependency for concrete in tension (Ross et al. 1996) ............ 152 Figure 6.25 Proposed modified CEB curves in tension (Malvar and Crawford, 1998).. 153 Table 6.2 Design Increase Factor (DIF) for Design of Reinforced Concrete Elements (TM5-1300, 1991)........................................................................................................... 154 Figure 6.26 Coefficient for Determination Of the Moment of Inertia for Concrete Elements with tensile reinforcement only (TM5-1300, 1991)) ...................................... 156 Figure 6.27 Coefficient for Determination of the Moment of Inertia for Concrete Elements with symmetric reinforcement (TM5-1300, 1991))........................................ 156 Figure 6.29 Moment curvature diagram for simply supported dynamically loaded I beam (TM5-1300, 1991)................................................................................................. 160

Asher Gehl 9913569

14

Table of Figures and Tables Figure 6.30 DIF for Strain rate of different steels (TM5-1300, 1991) ........................... 162 Figure 6.31 Rectangular Section with Equal reinforcement (TM5-1300, 1991)............ 165 Figure 6.32 Circular Column Section with Uniformly distributed reinforcement (TM51300, 1991) ..................................................................................................................... 165 Figure 6.33 Column interaction diagram (TM5-1300, 1991) ......................................... 166 Figure 7.1 Separation point for higher velocity flow...................................................... 176 Figure 7.2Drag Coefficient Vs Mach number for a typical bullet (Bray, 2004) ............ 176 Figure 8.1 Sequence of events during structural collapse............................................... 220 (Smith, Brokaw, Swatzell 1998)..................................................................................... 220 Figure 8.2 Murrah Federal Office Building after 1995 Attack (Mlakar et al, 1997)...... 222 Figure 8.3 World Trade Center during progressive collapse (http://www.civil.usyd.edu.au/latest/wtc.php#why2001) ............................................... 223 Figure 8.4 Illustration of alternate load path distribution (Whittaker and Hamburger, 2004) ............................................................................................................................... 224 Figure 8.5 Overall flowchart for Progressive Collapse consideration (GSA, 2000) ...... 233 Figure 8.6 Definition of Allowable Collapse Area and Progressive Collapse ............... 236 (GSA 2000)..................................................................................................................... 236 Figure 8.7 Allowable damage of Building (Smith, Brokaw, Swatzell, 1998) ................ 239 Figure 8.8 Damage Limitations (Smith et al 1998) ....................................................... 242 Figure 9.1 Energy Absorbing Catch System (Smilowitz, 2003) .................................... 249

Asher Gehl 9913569

15

Notations

Notations
Roman Upper Case Letters
A Ac
B B

Area of Steel reinforcement Area of cross section in compression Area of cross section in tension Damping Coefficient Equivalent load factor
B

At
B B

C CE
B

Cr
B B

Reflected pressure coefficient Sound velocity in the reflected region


B

Cr
B B

CD
B

Drag Coefficient Depth of object normal to direction of fluid flow Youngs Modulus of Concrete Youngs Modulus of Steel Energy per unit mass and time; initial and time unit one Load Coefficient for Moment of Inertia Tables Drag Force
B

D Ec
B B

Es
B B

E0, E1 F F FD
B

FE
B B

Equivalent Load Distance Height of structure Heat of detonation


P B

G H Hd
P B

HT I Ic
B B

Height of Mach front Moment of Inertia Moment of Inertia of a Cracked Section Stiffness Elastic Stiffness
B

K Ke
B B

Kep
B

Elasto-Plastic Stiffness Equivalent Spring Constant, Equivalent Stiffness


B

KE
B

KL
B B

Load Factor Load-Mass Factor


B

KLM
B

Asher Gehl 9913569

16

Notations KM Kr L M M ME MN MP Mx , Mr My Mr P0, P1 Po Pr Ps Pu Pmax P(t) Ps0, PsPdrag Pstag Ru R, RG Re S S T Tx, Ty, Tr U U0,U1 Us Mass Factor Resistance Factor Length Mass Mach number Equivalent Mass Ultimate negative moment capacity Ultimate positive moment capacity Mach number for incident shock and reflected shock Moment at yield Mass of element r Pressure; initial and time unit one Atmospheric pressure Reflected pressure Static Overpressure, Side on Overpressure Axial Load Capacity Peak Pressure Overpressure at time t Peak pressures Drag Pressure Stagnation pressure Total ultimate resistance Distance from explosion Reynolds number Clearing distance Elastic Section Modulus Natural period of an element Temperature at location x, y, and reflected temperature Particle velocity, shock velocity Particle velocity; initial and time unit one Blast Wavefront Velocity, Particle velocity

Asher Gehl 9913569

17

Notations V W WEXP Ws Xe Xm Xp Xst Xu Z Z Velocity of flow Equivalent weight of TNT Weight of explosive material Width of structure Deflection at the Elastic limit Maximum Deformation Deflection at plastic hinge Deflection under equivalent static load Ultimate Deflection Scaling parameter Plastic Section Modulus

Asher Gehl 9913569

18

Notations

Roman Lower Case Letters


a a a0 b b c d dc f fy fu fdy fds fdu fc fdc fcu fdu k i s+ is ir m1 m2 m(x) me n p0 Depth of equivalent rectangular stress block (concrete) Average yield strength increase factor (steel) Speed of sound in air at ambient pressure Width of beam Constant for describe pressuretime history Dynamic Increase Factor for Steel Distance from extreme compression fibre to centroid of tension reinforcement Distance between centroids of compression and tension reinforcement Shape factor Static Yield stress (steel) Static Ultimate Stress (steel) Dynamic Yield Stress (steel) Dynamic design Stress (steel) Dynamic Ultimate Stress (steel) Static Ultimate Compressive strength (concrete) Dynamic ultimate compressive strength (concrete) Ultimate Compressive Strength (concrete) Dynamic Compressive Strength (concrete) Slope of Rayleigh line Positive impulse Negative impulse Reflected impulse Distance from neutral axis to the centroid of area in compression Distance from neutral axis to the centroid of area in tension Distributed Mass per unit length Effective unit mass Modular ratio Atmospheric pressure 19

Asher Gehl 9913569

Notations p(x) qs
B B

Distributed load per unit length Dynamic pressure Peak Dynamic pressure
B

q0
B

r rE
B

Total resistance of structural element Equivalent total resistance


B

ru
B B

Ultimate unit resistance Reflected stress and velocity Time Arrival time

r , ru t ta
B B

tc
B B

Clearing time Positive phase duration


B

td
B

tE
B B

Time to reach yield (steel reinforcement)


B

tm
B

Structural response time Fictitious positive phase duration Initial time, time unit one Specific volume Specific volume; initial and time unit one Deflection Depth of neutral axis below the extreme compression fibre

trf
B B

t0, t1 v v0 , v1 x x=

Asher Gehl 9913569

20

Notations

Greek Letters
crit
B B

Reflection coefficient for normal blast wave reflection Angle of incidence Angle of incident shock Critical angle for Mach reflection Reflection angles
B

, max
B

Maximum deflection of a structure Deflection as any point of a structure Deflection of mass r


B

(x) r
B

c u
B B

Ductility ratio Strain Concrete strain Ultimate concrete strain (rupture strain) in tension Strain in x-direction Strain rate Strain rate for static loading Shape function Maximum rotation about the supports
B

xx _ 0 _ (x) max
B

0, 1 0, s 1, 2 s
B B

Kinematic viscosity Density Density; initial and time unit one Initial density, solid density Density; medium one and medium two Density of air behind the shock front Stress Stress in x-direction Natural circular frequency of vibration Function of td
B B B B

xx

Function of td/T

Asher Gehl 9913569

21

Introduction

Chapter 1 Introduction
1.1 Key Concepts and Aims
The analysis of structures subjected to the effects of an explosion is very complex and requires an understanding of structural engineering, dynamics, strengths of materials, and explosive effects. The complex analysis is both time-dependent and accounts for non-linear behaviour. Engineers have mastered the design and analysis of structures to resist static loads. Analysis and design of structures to withstand the dynamic loading conditions of earthquakes and severe winds have also been thoroughly researched. For these loading conditions there are mandatory standards that structures must be adhere to, in order to be classified as suitable and safe for construction. However no such standard exists for Engineers designing buildings to withstand the effects of an explosive blast. The blast load is different to the above normal loading conditions essentially because of its high magnitude and transient dynamic nature. Damage from an explosion is essentially divided into direct local and indirect global effects. The local damage is as a direct consequence of the high intensity blast wave pressures acting on the element. The indirect global damage is a result of progressive collapse of a structure due to the local failure of a key structural element. The material presented in this thesis aims to present a strong background of knowledge into the dynamic loading from an external explosion and the response conditions of reinforced concrete and structural steel elements, in order to perform preliminary design calculations on these elements. Columns of varying cross-sectional shape will be used in these designs in order to present the design results for different shaped columns. Finally the global analysis of progressive collapse mitigation is presented in order to emphasise the importance of the overall robustness of building in order to minimise the overall damage from an explosive event.

Asher Gehl 9913569

22

Introduction

1.2 Research Area/Limitations


The research area in this thesis is to study structures subjected to an explosive loading from an external high-explosive bomb. The primary aim of this thesis is to achieve a thorough understanding into the key concepts associated with blast engineering. In order to demonstrate the achievement of this aim a wide survey of the literature has been presented. It is important to state that the vast majority of this material presented in this thesis is a representation of a number of previous works on the subject and the concepts are presented in a logical manner that enables the reader to achieve an understanding of blast engineering. The original content of this thesis is limited to the preliminary flexural design of reinforced concrete and steel columns of varying cross sectional shape subjected to an external surface burst, which is presented in Chapter 7. It is important to note however that the design of these columns completely ignores the effects of other load combinations and simply examines the effects of blast loads alone. Therefore secondary non-linear P- effects due to compressive loads are not considered.

1.3 Outline of Contents


In Chapter 2 the background required to determine the scope of threats to be considered as well as a number of the more significant case studies is presented. In Chapter 3 the general characteristics of the loading environment are briefly discussed to separate and emphasis the differences between regular loads and that of an explosion. In Chapter 4 the physics of the detonation process is presented, from its initiation, to the formation and propagation of the shock front of the blast wave. Also presented are the parameters involved in describing the blast wave that are significant in order to describe the expansion of the blast wave and the load it imparts on a surface upon impact with a structure in its path.

Asher Gehl 9913569

23

Introduction Chapter 5 describes the derivations of the loads acting on the surface of a structure, and introduces the loading that is applied on the columns for the preliminary flexural design of Chapter 7. In Chapter 6 the response of the structure is determined according to the required level of protection for the building. The structures conversion into an equivalent Single Degree Of Freedom is discussed as well as the relationship between the duration of the blast load and the fundamental period of the element. The characteristic properties of concrete and steel under the dynamic loading environment are presented. The important parameter is the strain rate and the increase in strength associated with the dynamic load. The content of the chapter is then summarised with the presentation of the procedure used for the design of concrete and steel columns. In Chapter 7 MathCAD worksheets present the preliminary flexural design of individual columns for both reinforced concrete and steel sections of varying cross sectional shape. The worksheets utilise the methods and principles of design that were outlined in previous chapters. Chapter 8 introduces the concept of Progressive Collapse mitigation techniques. A survey of the current guidelines of the phenomenon is presented and a number of key issues are introduced, outlining the limitations of current analysis techniques and presenting a selection of more advanced and accurate methods. In Chapter 9, conclusions are drawn from the thesis and recommendations presented for future research.

Asher Gehl 9913569

24

Background

Chapter 2 Background
2.1 Background and Aim
Due to the disastrous and horrific onset of terrorism over the last half century and in particular its sudden rise in the last decade, it is becoming crucial that engineers around the world start taking proactive steps to protect buildings against such catastrophic events. The world we exist in today is an environment with an ever-increasing threat. The threat of terrorism over the past decade has become a frequent and familiar occurrence. The most common and deadly act of terrorism recently has become the detonation of a car/truck bomb outside a target building. This action has the capacity to inflict massive damage to the target as well as the surrounding buildings and has the ability to kill and injure hundreds of people. In the past, the task of designing buildings to withstand the effect of an explosion was left to specialist companies, and limited to key government buildings and embassies. The world was given a lethal shock at 9 am on Wednesday, April 12, 1995 when a truck bomb exploded just three metres from the Murrah building in Oklahoma City, USA, causing 167 deaths and partial collapse of the structure. This catastrophe and many others since have created a new concern and a new loading phenomenon that must be understood by Structural engineers in order to adequately protect structures and the people within, from further tragedies. The objective of the engineer is first and foremost to save as many lives as possible. This does not necessarily imply complete protection of the structure from any damage, but rather aims to mitigate the extensive damage that may be caused and prevent progressive collapse of the structure in order to allow for emergency personnel to tend to the injured and evacuation of the building without further fatalities.

Asher Gehl 9913569

25

Background In order to achieve this, as many of the soft targets as possible must be hardened to ultimately dissuade the terrorist from the building and move them elsewhere. There are a number of steps required to achieve this. The aim of the engineer is not to reduce risk to zero, as this is unrealistic given the lay out of the majority of civilian structures and the high cost that would be associated with complete protection. The intention is to significantly minimise the damage in the event of an explosion.

2.2 Risk Analysis and Threat Determination


Historically, blast resistant design has been an avenue of research geared primarily towards military installations and petrochemical facilities. In principle, much of the theory and techniques used for the design of these structures is also applicable for civilian targets of terrorist attacks. There is of course a fundamental difference between the final design of the structures. This difference stems from the original design objective. The primary objective of military facilities is to maintain its function after any attack where as the primary design objective of civilian facilities is to ultimately save the lives contained within. Due to this fundamental difference, the design principles and procedures must be re-evaluated. Before any design stage, there a number of crucial significant decisions that are required to be made. This decision making process is not part of the design engineers responsibility, but the procedure must be understood and its implications realised. The prime objective of the risk analysis is to determine how any attack may be carried out and with which weapons and then assign a certain probability associated with each threat and thus the level of protection required. This assessment is generally the role of the contracted security professionals. However the design engineer must realise the importance of this assessment on the influence of the clients eventual requirements. These requirements are reflected in the ultimate design and level of protection, and therefore wherever possible the engineer must have as much input and knowledge on the threats and risks involved.

Asher Gehl 9913569

26

Background The Protective Design Center of the U.S Army Corps of Engineers has a detailed and in depth threat determination procedure outlining all design criteria and threats to assets within a building or infact the building itself (Conrath, Krauthammer, Marchand and Mlakar, 1999). The guide uses a series of worksheets to lead the design criteria developers to ascertain decision variables necessary to establish the asset to be protected, the actual threat level to the asset and the protection level required. Of considerable interest to the engineer is the level of protection required. This is the degree to which an asset is protected against a particular threat. This protection level will determine the allowable degree of damage to the building and ultimately the design process that will be implemented. The risk assessment process is often one of the most difficult decisions the client must make. The petrochemical industry is a good indication for comparison as to what issues must be addressed for the process to be more effective. The danger of accidental explosions is such that these decisions must be made for every facility. Many of the large engineering firms issue their own design standards for blast resistant design. Of more relevant interest however are the industry standards that are available (Bounds, Nene, Ko, 1998): Safety Guide SG-22: Siting and Construction of New Control Houses for Chemical Manufacturing Plants (The Manufacturing Chemists Association, 1978). 7-45S: Process Control Houses and Other Structures Subject to External Explosion Damage (Factory Mutual Engineering Association, 1980). The fundamental issue with these standards is that without full industry review and acceptance the ultimate decision of what to make resistant will not be standardised across the industry. Consequently even an industry where it is readily accepted that blast resistant design is a crucial requirement, there is still a hesitation and often reluctance on the clients part to the amount of resistance required with respect to the expense incurred.

Asher Gehl 9913569

27

Background Thus with regard to civilian structures there is a long and arduous path that must be paved to make the client aware and willing. There are a variety of ways to assist this process. A number of professionals must be brought into the team to help combat and defend against this threat. In correspondence with Dr Andy Davids (BE, MEng Sc, PHD), the Director of the Building Structures unit of Hyder Consulting (personal communication, June 2004), a number of possible solutions were discussed. Business leaders and building owners must be convinced that security is not a cost. One method in which to achieve this goal in the near future is for the concept of security star ratings to be promoted. This rating system is such that a 5 Star building will be more secure than a 2 Star building and so on. This will help encourage the leaders and building owners to spend little at the early stages of concept design to achieve a great deal. A cooperative body must be created between Commonwealth, State, Territory and Local governments, industry and all professionals. The ultimate of this body, is concerned with the reduction of the complaint culture that is held by the majority. The cooperative body must strive to achieve an intuitive and proactive security conscious culture. To achieve this aim, engineers must not sit around and wait for the required codes and standards to eventually be written but to apply fundamental and leading edge engineering practices to achieve the best defensive and robust designs. There are a number of design manuals available for protective design against explosions, for example TM5-1300 (1991) and TM5-855-1 (Hyde, 1991), however as yet there is no comprehensive design standard for a civilian structure. Of most importance in this regard is a consensus on the engineering calculations and construction details in order to standardize and unify the design approaches. For instance in the petrochemical industry guide SG-22 (1978) there is a standard loading that is suggested to be used of 1 Tonne TNT burst at 100 feet standoff. The risk with this standard loading is of course to be unnecessarily conservative or to dangerously under design the structure (Bounds et al, 1998).

Asher Gehl 9913569

28

Background The threat assessment must be determined based on actual site conditions and made specific to each method of attack possible as per the Protective Design Center of the U.S Army Corps of Engineers guidelines (Conrath et al, 1999). For civilian targets there are of course many options of attack with varying responses and design solutions. Historically however there has been one attack method that has proved to be of most detriment to large structures and the people within. This of course is the detonation of a large explosive device, generally transported by a vehicle, and exploded outside the target. The following Figure outlines the damage associated with a number of potential vehicular explosions.

Figure 2.1 Damage level curves for given Stand-off distance and Explosive weight (Federal Emergency Management Agency (FEMA), 2003)

Asher Gehl 9913569

29

Background For the analysis of this thesis a medium to high threat from a stationary vehicle bomb of 100kg was used for the blast environment acting on the columns. For complete protection of course, a more thorough and inclusive assessment must be obtained examining the effects of each individual threat for each worse case scenario. Protective design for the hardening of a structure is the least effective form of countermeasure against a terrorist attack. In order to achieve complete protection a number of issues must be addressed. Deterrence: The terrorists are looking for easy and attractive targets. If a facility has a certain level of protective security, the terrorist is likely to move on to the next softer target, as there are too many obstacles to overcome for a successful attack. This theory has been successful in a number of installations around the world. Stopping the attack: By preventing the attack via appropriate physical security measures like an external fence or bollards, the attack cannot get close enough to the target to inflict any substantial damage. If the above techniques have failed then the building must be hardened to absorb the energy from the blast and protect the lives held within. The threat assessment will combine with the ultimate design objective to determine the required amount of resistance the structure is to contain. It is ultimately the clients decisions as to how much damage is acceptable after any attack, and the design is a function of that decision. If the objective is to limit damage to minor repairable issues to allow the building to be functional after any attack, then all deformations must be within the elastic region and any plastic deformations must be restricted. If the objective is to save lives and for the building to remain standing so that emergency personnel may have access after the attack, then large plastic deformations are tolerated.

Asher Gehl 9913569

30

Background The former is as previously mentioned, is the prerequisite for military and critical infrastructure. Such limitations on deformation require much stronger structural elements and a much more expensive design. The latter is a more common and likely objective for the majority of civilian clients. Due to the small chance of their building being the target there will be obvious resistance on their behalf for paying the large expense involved with a more resilient structure. Thus large plastic deformations will be tolerated and in fact used to absorb the blast energy as a trade off for the stronger structural elements.

2.3 Past Attacks


The past has been filled with many horrific examples from which the world should learn from in order to effectively protect structures from such extreme events. There have been over one hundred car bomb attacks in the past few decades with nearly 5000 fatalities from these events. Some of the more significant attacks are presented below as case studies.

Asher Gehl 9913569

31

Background 2.3.1 Ronan Point London 1968

Figure 2.2 Ronan Point Apartment Complex Collapse (Moore, 2002) This incident, although not demonstrating an intentional external explosion, is an infamous example illustrating the devastating effects of a local failure resulting in significant progressive collapse. It is included at this point as it is one of the recent key historical events that illustrate the devastating effects of progressive collapse. It was influential in the development of a number of guidelines that will be presented in Chapter 8. On 16 May 1968 an explosion blew out a primary supporting exterior bearing panel which formed the side of the 24-story Ronan Point apartment building in east London, leading to a loss of support for the floors above. This in turn began a chain reaction of collapse, upwards, to the roof level. As a result of the loading generated by falling debris, collapse of the structure propagated from the 18th floor, downward, almost to the ground (HMSO, 1968).

Asher Gehl 9913569

32

Background Construction of the Ronan Point complex primarily consisted of precast concrete panels. While this type of construction can be designed to avoid progressive collapse from abnormal loading conditions, the Ronan Point complex lacked the connection details necessary to effectively redistribute load. The essential missing detail from this apartment building was reinforcement continuity between panels. Because of this, there was no mechanism in place for achieving effective alternate load paths once failure began to propagate. The spectacular nature of the collapse created an enormous impact on the philosophy of structural design and resulted in important revisions of design codes. The Ronan Point report of the Court of Inquiry (HMSO, 1968) stated: "It is the common aim of structural engineers so to design their structures that if one or two component parts or members fail due to any cause, the remaining structure shall be able to provide alternative paths to resist the loads previously borne by the failed parts." After the Ronan Point failure, many building regulation codes and standards were introduced to address this form of collapse. This will be considered further in Chapter 8.

Asher Gehl 9913569

33

Background 2.3.2 Exchequer Court, St Marys Axe, London 1992 The information from this event was primarily taken from The UK and European regulations for accidental actions (Moore, 2002). In April 1992 a bomb exploded in St. Marys Axe, London. The explosion caused damage to a number of buildings including Exchequer Court (25-51 St. Mary's Axe), the Baltic Exchange (24-28 St. Marys Axe) and the Chamber of Shipping (30-32 St. Marys Axe). Away from the immediate site of the bomb the damage was mainly non-structural with a large number of windows lost. The non-structural damage occurred for several hundred metres from the bombsite. Although none of the buildings close to the explosion collapsed, all suffered considerable damage. The most modern of these was Exchequer Court which at the time was a newly completed steel framed structure with composite floors and as such is worthy of further consideration. The details are presented according to The UK and European regulations for accidental actions (Moore, 2002).

Figure 2.3 Damage of the Surrounding Buildings (Krauthammer, 2004) Asher Gehl 9913569 34

Background Building Description The building was of modern construction and consisted of a steel frame with in-situ concrete floors acting compositely with a steel profile metal decking. To conform to steel frame construction standards in the UK, the building was designed to resist lateral wind loads by a system of braced steel bays. Consequently the main steel frame was designed to support gravity loads only. The beam to column and beam-tobeam connections were both of the flush end-plate type and were designed to transmit vertical shear only. However, because the structure was over five storeys high, the connections would have been designed to carry a horizontal tying force of not less than 75kN to preserve the integrity of the structure in the event of accidental damage (Moore, 2002) Summary of Damage The bomb was placed approximately 6 metres from the face of the building and consequently the cladding, columns, beams and floors close to the blast were badly damaged. Most of the cladding was removed at ground floor level and two of the ground to first floor columns that were close to the blast were bent 125mm and 406mm respectively as shown in Figure 2.4.

Figure 2.4 Damage to ground floor steel columns Exchequer Court, St. Marys Axe (Moore, 2002) Asher Gehl 9913569 35

Background The edge beams connected to these columns were also badly damaged and a detailed inspection of the connection between the column and the edge beam showed that the connection had been displaced approximately 200mm vertically. None of the bolts could be found and no distortion of the end plate was visible suggesting that the connection failed in shear (Moore 2002). The first and ground floors were also badly damaged. The first floor was bent upwards in a series of arches with the maximum displacement occurring between the secondary beams. In a number of cases the composite floor had come away from its supporting beam and shear studs were clearly visible as shown in Figure 2.5.

Figure 2.5 Damage to composite floors Exchequer Court, St. Marys Axe (Moore, 2002) The ground floor was bent downwards although the displacement was less pronounced that that of the first floor. A connection between a secondary beam and an internal column had moved approximately 150mm vertically (Moore, 2002).

Asher Gehl 9913569

36

Background Performance Observations Although the building suffered considerable damage to both its non-structural and structural members the building remained intact. Furthermore the type of explosion at St. Mary's Axe was of a completely different nature to the internal gas explosions which were the principal cause for the disproportionate collapse rules given in the Building Regulations, Approved Document A and the material Codes. The explosion that occurred at St. Mary's Axe was produced by a large amount of SEMTEX being detonated. This type of material produces a very rapid pressure rise often followed by a negative pressure. In contrast, gas explosions are relatively slow and can be reasonably modelled by a uniform static load. 2.3.3 AMIA Building: Buenos Aires, Argentina 18th July 1994

Figure 2.6 Location of the Blast on the AMIA Building (Rothschild, 2004) A car bomb exploded outside the six-storey Argentine Mutual Aid Association (AMIA) Jewish community building in Buenos Aires causing the building to collapse, killing eighty-five and injuring two hundred others. Subsequent investigations showed the instigators to have been the Iranian government using Hezbollah operatives, together with local neo-Nazi sympathisers within the police (www.ict.org.il). The focal point of the explosion was close to the edge of the pavement, 3 metres in front of the buildings entrance. Preliminary measurements indicated that the device consisted of 400kg of high explosive comprising mostly of fuel oil mixtures. The blast formed a 70 cm deep crater, 5.3 metres in diameter and its edge was 1.5 metres from the building. Examination of the scene discovered that the car that exploded opposite the entrance to the building was a 1990 model Renault Furgon Traffic (K. Rothschild 2004, personal communication, 30 August 2004).

Asher Gehl 9913569

37

Background The blast caused significant structural damage that ripped apart the building resulting in significant portion of the building to collapse.

Figure 2.7 Crater and Damage from Explosion (Rothschild, 2004)

Figure 2.8 Structural Damage to AMIA Building (Rothschild, 2004) Asher Gehl 9913569 38

Background 2.3.4 A.P Murrah Federal Building Oklahoma 1995 On the morning of April 19 1995, a truck containing approximately 2 tonnes of TNT equivalent explosives consisting of Ammonium Nitrate fertilizer doused with fuel oil (ANFO) exploded 5 metres from the north face of the 9 storey Alfred P Murrah Federal building in Oklahoma City. This attack resulted in over 167 fatalities and inflicted over 500 injuries to the occupants and damaged over 300 buildings. The cost of replacement is estimated to be 28 million dollars. The vast majority of the deaths were attributed to falling debris generated by the progressive collapse that the structure underwent. When the bomb detonated, primary support columns along the perimeter were locally damaged. The localized damage of these bearing members initiated a series of failure that extended across much of the north face and through the width of the building in the eastern part of the structure (Mlakar, Corley and Thornton, 1997).

Figure 2.9 Aerial View of damage to the Murrah Building (Krauthammer, 2004)

Asher Gehl 9913569

39

Background The building was essentially a reinforced concrete, ordinary moment frame with typical bay sizes of 6m by 10m. An essential structural feature of the building was the use of a transfer girder at the 3rd floor level. This girder provided support for additional intermediate columns that extended from the 3rd floor to the roof level. However, from the 1st floor to the 3rd floor, there were columns at every second position spaced at every 12m. Hence, the loss of a column along the ground level created an 24m unsupported length. It is believed that this was the primary mechanism that caused the progressive collapse of the building (Mlakar, Corley and Thornton, 1997). The findings from the post explosion analysis recommended provisions and guidelines for future buildings. Structural features such as transfer girders can potentially limit the building's ability to effectively redistribute load and are criticized by the reports, as are other features that inhibit the structures general robustness under abnormal loading situations (Prendergast, 1995).

Figure 2.10 Extent of Structural Collapse of Murrah Building (Krauthammer, 2004)

Asher Gehl 9913569

40

Background 2.3.5 Khobar Towers Saudi Arabia 1996

Figure 2.11 Damage to Khobar towers (Krauthammer, 2004) More than 90 tonnes of high explosive detonated outside a precast concrete building, which housed U.S. military personnel and served as the headquarters for the U.S. Air Forces 4404th Wing. Over 19 people were killed and scores injured in this destructive event. The explosive charge was estimated through glass breakage, crater dimensions and advanced simulations using the manual TM5-855-1 (Hyde,1991) (Krauthammer, 2002).

Figure 2.12 Crater SizeKhobar towers (Krauthammer, 2004)

Asher Gehl 9913569

41

Background 2.3.6 Australian Embassy Jakarta 9 September 2004 A car bomb was detonated outside the Australian embassy in Jakarta killing 9 and injuring over 180. The blast detonated outside the embassys gates, giving the building a substantial standoff distance.

Figure 2.13 Plume of smoke from Jakarta blast (www.bbc.com, 9 September 2004) The blast blew a huge crater in the road and shattered the glass off the surrounding buildings in its vicinity. The building had previously undertaken a vast and effective security upgrade, resulting in minimal damage to the structural elements. As seen in Figure 2.14 below, the majority of the damage caused was to the glass faade of the building.

Figure 2.14 Damage to Australian Embassy Jakarta( www.bbc.com, 9 September 2004)

Asher Gehl 9913569

42

Unique Characteristics

Chapter 3 Unique Characteristics


The aim of this Chapter is to emphasize a number of key properties of an explosive load that separate it from other loads, generally considered in the design process. The most fundamental issues are: Duration Plastic Behaviour Magnitude Structural Damage

Following this, the unique design process is outlined for structures subjected to an explosive load.

3.1 Duration
The duration of the load is significantly different to other dynamic loading phenomena. Earthquakes and wind gusts are generally measured in seconds or even hours for the case of sustained winds or floods. From an explosive event the load is transient and its duration is measured in the order of microseconds. It takes a certain amount of time for the mass of the structure to respond to the blast. For explosive loads at short standoff distance this response time is longer than the loading duration. Thus the mass of the structure acts to diminish the response. For elements with this type of response the impulse is the significant factor determining the resulting damage levels. This is different for other dynamic loads, such as earthquakes where the loading duration is similar to the response time that causes resonance effects, resulting in an enhanced level of damage.

Asher Gehl 9913569

43

Unique Characteristics

3.2 Plastic Behaviour


Although plastic behaviour is not generally permissible under service loading conditions, it is quite appropriate for design, when the structure is subjected to a severe blast loading only once or at most a few times during its existence. Under blast pressures, it will usually be uneconomical to design a structure to remain elastic and, as a result, plastic behaviour is normally anticipated and accepted, in order to better utilize the energy-absorbing capacity of blast-resistant structures. Plastic design for flexure is based on the assumption that the structure or members resistance is fully developed, with the formation of near total plastic sections at the most highly stressed locations. For economical design, the structure should be proportioned to assure its ductile behaviour up to the limit of its load-carrying capacity. The structure or structural element can attain its full plastic capacity provided that premature reduction of strength due to secondary effects, such as brittle fracture or instability, does not occur.

3.3 Magnitude
The intensity of the load from an explosive event is generally several orders of magnitude higher than other hazardous loads such as earthquakes, severe winds or floods. For an explosive attack from a vehicle a few meters from a building, the resultant load from the pressure is in the order of MPa. At such high pressure values significant damage is expected.

Asher Gehl 9913569

44

Unique Characteristics

3.4 Structural Damage


The loading from an explosive event imposes forces on a structure that standard designs do not cater for. When the blast wave hits a structure the faade is initially smashed and the pressure has access to the internal structural elements. This phenomenon is subsequently illustrated in Figure 5.1. A standard design criterion for floors is to resist gravity loads from both imposed and permanent actions. The blast wave however will impose loads on all internal surfaces in all directions. Thus the floors of a structure will be pushed up as well as down. The large surface area and small thickness amplifies the damage potential of the blast. Due to this vulnerability, as seen in the previous Chapter, a number of failures from an explosion are associated with floor slabs.

3.5 Design Procedure


The damage from an explosive event is essentially attributed to two main factors: Local damage resultant from direct effects of the high intensity blast Global progressive collapse via indirect means.

The former is thoroughly addressed within the subsequent Chapters of this thesis. Chapter 7 presents the preliminary designs methods required for column sections of various cross sectional shape subject to an external explosive load. Progressive collapse is associated with the local damage from the explosion inducing further collapses that progress into significant collapse of the structure. The techniques and guidelines are presented and reviewed in Chapter 8. It is considered important however to link these two design criteria and present the overall process required to design a structure to resist explosive loading. 1. Threat assessment The threats against the building must be determined and each worst-case considered separately.

Asher Gehl 9913569

45

Unique Characteristics 2. Load Combinations The likelihood of an explosive attack on a structure is very low. Therefore the other ultimate wind and earthquake loading events can be ignored and a partial safety factor of unity given to Dead and Live loads. With respect to the explosive load, each specific threat must be considered individually in its own analysis and the corresponding loads imparted on the structure must be determined. 3. Local Design Individual elements must then be analysed to determine their response to the explosive load. The analysis procedure essentially utilizes an equivalent Single Degree of Freedom system, taking into account the dynamic increase factors from the dynamic transient load. There is a significant inelastic deformation to the load. In combination with the gravity loads on the element there will be considerable second order P- effects. It is important to remind the reader that throughout the course of this thesis and especially in the Chapter 7 designs, other load combinations are neglected and the blast load is the only load considered. Thus, P effects are not considered. There are a number of conditions to determine whether the load causes the failure of the element. Generally however, the condition is a function of ductility and rotations about the support. If the response of the element is not within these restrictions, then the element has essentially failed. If possible the structural elements should be redesigned to resist the load.

Asher Gehl 9913569

46

Unique Characteristics 4. Global Design Under the impact of large explosions at a small standoff distance, local design methods are insufficient. The magnitude of the load, and the sheer expense involved with the redesign, will result in the collapse of a number of key elements. Often however restricting the collapse area associated with this event is an acceptable design criterion. The exact areas that are acceptable in many of the guidelines, is presented in Chapter 8. The general design global design process is to consider the building in it entirety and to subsequently remove elements that have failed under the explosive load. Upon removal, the response of the structure is to be analysed, to examine the structures alternative load paths. To ensure adequate strength and robustness in design, adequate steel continuity must be provided in all elements to withstand the inelastic deformations reached in redistributing the loads. Significantly advanced 3D Finite Element Modelling (FEM) techniques are required to model the structures response, and to provide solutions to minimize the risk of progressive collapse.

Asher Gehl 9913569

47

Blast, Stress and Shock Wave Theory

Chapter 4 Blast, Stress and Shock Wave Theory


4.1 Introduction
To understand the behaviour of structures subjected to severe loading from a blast, the nature of the fundamental principles and physics of explosions and the creation of a blast wave and reflections from a bomb must be understood. The three fundamental consequences of an explosion are: The Blast overpressures Fragments Shock loads

Blast overpressure is the main parameter in determining the loads on a structure. The magnitude and distribution of the blast loads is discussed in the next chapter. However for a thorough assessment of the damage upon a structure, the other effects and parameters of the blast must be determined. A significant amount of energy from an explosion is used in the shock wave as it transmits loads through air and ground inducing vibrations analogous to a high intensity short duration earthquake. Fragments from an explosion can cause significant damage to a structure upon impact (Medearis, 1975). An explosion is principally a chemical reaction in which an extremely large and rapid amount of energy is released. This energy consists of a variety of forms including sound, heat, light and a shockwave. It is this shockwave that is of primary concern for the building and causes significant damage. The formation of the shockwave and its characteristics is presented in this chapter.

Asher Gehl 9913569

48

Blast, Stress and Shock Wave Theory

4.2 Formation of Blast Waves


Upon detonation of an explosion a very rapid and stable exothermic chemical reaction releases copious amounts of energy. This reaction proceeds through the explosive material at the supersonic detonation velocity. The explosive material is rapidly converted into a gas with very high temperature density and pressure. This highpressure gas then expands radially as a supersonic shockwave into the lower pressure surroundings. The magnitude of the shockwave pressure is proportional to the weight of the explosive material used (W) and is several orders of magnitude higher than conventional loading situations, with resultant peak stresses often exceeding yield the stress of the material.

Figure 4.1 Shockwave of an explosion As the shockwave expands over a larger area, it decays in strength, lengthens in duration, and decreases in velocity. The following relation describes the pressure decrease over distance (R): P R1/3
P P

(4.1)

Over time the pressure is decaying exponentially over a period of milliseconds (Conrath et al, 1999). The decay phenomenon is caused by spherical divergence as well as by the fact that the chemical reaction is completed.

Asher Gehl 9913569

49

Blast, Stress and Shock Wave Theory

4.3 Damaging Properties of Blast Wave


To understand the damage caused by the blast wave a number of fundamental characteristics must be established. Consider the following sequence of events that occur in the blast wave formation (Mays and Smith, 1995). Detonation releases hot gases at pressures of 1-300000 bar and temperatures of 3-4000 degrees Celsius. These gases due to high temperature and pressure violently expand Upon expansion, the surrounding air gets displaced from the volume it occupies. This surrounding air forms a layer of compressed air in front of the gases. The majority of the energy from the detonation is in the compressed air layer The blast Wave. The rate of gas expansion slows over time. The back/tail of the blast pressure then falls below the ambient level- the negative phase of the blast wave suction. Flow reversal occurs to return to equilibrium and initial ambient pressure and temperature. There are two fundamental phenomena associated with the blast wave in air: 1. Static overpressure The static overpressure is the sharp increase in pressure that is exerted by the compression of the atmosphere by the dense front of air that comprises the shockwave. The magnitude of the overpressure at any given point is directly proportional to the density of the air in the wave. The static overpressure rises instantaneously from normal atmospheric pressure of the unaffected air in front of the blast wave to a peak value. The overpressure then decreases behind the front of the blast wave as it expands. After a certain distance the pressure behind the front drops below atmospheric pressure. This is the negative phase of the blast wave.

Asher Gehl 9913569

50

Blast, Stress and Shock Wave Theory 2. Dynamic pressure The Dynamic pressure is the drag force exerted by the strong transient blast wind associated with the movement of air at the shock front of the blast wave. As the blast wave travels through the atmosphere it impacts with the air molecules in front of the shockwave imparting energy that causes the air to move in the direction of the shock front. Over time there is a considerable build up of compressed air travelling at the face of the shock front. This motion of air is a severe transient wind, known as the blast wind". The destructive force associated with these winds is proportional to the square of the velocity and is measured in terms of dynamic pressure. These dynamic forces are highly destructive.

Figure 4.2 Variations of Overpressure and Dynamic Pressure with Time (NATO FM 8-9, 1996) A combination of the high static overpressures and the dynamic or blast wind pressures causes the majority of the material damage caused by an explosive blast. The impact over time of the positive and negative phase of blast wave pressure with a structure is shown below.

Asher Gehl 9913569

51

Blast, Stress and Shock Wave Theory

Figure 4.3 Variations of Blast Effects Associated with Positive and Negative Phase Pressures with Time (NATO FM 8-9, 1996) As illustrated in Figure 4.3 above, upon arrival of the shock front the pressure immediately increases from zero to its maximum value and then begins to decay over time. The majority of the damage occurs during the positive phase of the blast wave. The duration of the positive phase increases with explosive yield and standoff distance to the order of a few hundredths of a second from conventional highexplosive detonation. During the negative phase, the blast pressure will drop below normal atmospheric pressure and the blast winds will actually reverse direction and blow back towards the explosive source. Damage sustained during the negative phase is generally minor, however, because the peak values of the pressure and wind velocity are relatively low.

Asher Gehl 9913569

52

Blast, Stress and Shock Wave Theory

4.4 Blast Scaling


Due to the large cost involved and difficulty in modelling large-scale explosions, it is common to scale the properties from a smaller test explosion. A scaling parameter (Z) is introduced, first noted by Hopkinson (1915), (Bulson, 1997). With the parameter Z, it is possible to calculate the effect of a detonated explosion, of any material, as long as the equivalent weight of charge in TNT is known. This relation is commonly known as the Hopkinson-Cranz Scaling law. Z=R/W1/3
P P

(4.2)

Where: R is the distance from the detonation W is the equivalent weight of TNT. Z is the Scaled range

The peak pressure, the positive duration time and the positive impulse can be expressed as functions of Z, and the pressuretime history (Figure 4.4) can be described independently of the explosive material. In the literature there are several empirical formulas for the expressions (Bulson, 1997). In TM 5-855-1 Fundamentals of Protective Design for Conventional Weapons (Hyde, 1991), there are a number of tables and diagrams for different types of explosive materials. The damage from an explosion is often expressed in terms of the scaled range and the equivalent TNT weight of explosive.

Asher Gehl 9913569

53

Blast, Stress and Shock Wave Theory

4.5 TNT Equivalence


Blast waves exhibit the same character independently of the explosive source. However various high explosives tend to produce different magnitudes of peak pressure. All laws and curves used are expressed in terms of TNT weight. In order to establish a basis for comparison, an equivalent charge weight of explosive material is calculated to produce the same effect as a TNT explosive charge. TM5-1300 (1991) expresses the effective charge weight WE of an explosive material, in terms of the heat
B B

of detonation H of TNT and the explosive.


P P

WE =

d H EXP WEXP d H TNT

(4.3)

WEXP = explosive material weight


B B

A variety of sources present a wide range of commonly used explosive materials with their TNT equivalence (Conrath et al, 1999; Mays and Smith, 1995).

Asher Gehl 9913569

54

Blast, Stress and Shock Wave Theory

4.6 Pressure Time History


The pressuretime history of a blast wave can be illustrated with a general shape as shown below in Figure 4.4. The illustration is an idealisation for an explosion in free air.

Figure 4.4 Pressure Time History (TM5-1300, 1991) The pressuretime history is divided into a positive and a negative phase. In the positive phase, at arrival time tA, the pressure instantaneously jumps to a peak value of
B B

overpressure, Pso, and immediately decays in a quasi-exponential fashion to atmospheric pressure, P0, in the time t0. An object at this location will be subjected to
B B

an instantaneous lateral force equal to the product of this overpressure and the projected area in the plane of the blast wave. The positive specific impulse is, is the area under the positive phase of the pressure
B B

time curve.

is =

t A +t0

tA

( Pso P 0 )dt

(4.4)

Asher Gehl 9913569

55

Blast, Stress and Shock Wave Theory For the negative phase, the maximum negative pressure, Ps0--, has a much lower
P P

amplitude than the maximum overpressure. The duration of the negative phase, t0-, is
B PB P

much longer compared to the positive duration t0. The negative specific impulse, is-, is
B B B PB P

the area under the negative phase of the pressuretime curve. i s =


t A + t0 + t0

t A+t0

( Ps P 0 )dt 0

(4.5)

4.7 Blast Wave Pressure Profiles


In order to completely describe the behaviour of an ideal blast wave, its form must be defined as a function of time. It is extremely arduous to derive the pressure-time history of the blast wave profile. However a number of experimental, computational and theoretical techniques have been utilized to obtain the characteristics and profile of a blast wave. The experimental and computational techniques are not considered within this thesis. A number of computer programs use the theoretical equations to determine the blast wave parameters. Commonly available programs include: ATBlast Version 2.0 (Applied Research Associates (ARA) 2000 CONWEP (Hyde, 1991) AirBlast (Dewey McMillin and Associates LTD) www.blastanalysis.com
HTU UTH

The theoretical techniques are considered below.

Asher Gehl 9913569

56

Blast, Stress and Shock Wave Theory


4.7.1 Theoretical Analysis

The equations required to describe the blast wave are extremely complicated. Brode was the first to solve the equations numerically in 1955. There are a variety of Eulerian and Lagrangian approaches to methods that are utilized in a number of references that describe a wide number of equations. Generally however the theoretical analysis procedure utilises the mathematical derivations for normal shock in gas dynamic theory. There is a wide body of knowledge outlining the principles of gas dynamics and shock wave theory. According to Baker (1983), the principal emphasis of all these functional forms of the blast wave is to empirically fit the function to measured or theoretically predicted positive phase histories. The negative phase was considered to be of lesser concern given the difficulty accurately measuring its characteristics and its lesser effect for design. Since WWII many researchers have developed equations based on gas dynamics theory. Both Baker (1983) and Kinney and Graham (1985) outline the derivations and theory of fundamental gas dynamics in a methodical and precise manner. Baker (1983) gives an excellent comparison between numerous equations from the simple linear decay model from Flynns basic model to Friedlanders three-parameter model and Brodes five-parameter model. The pressing concern is of course which description is the most accurate according to recorded experimental data. Baker (1983) concludes that a more complex equation yields a higher accuracy result. Generally however, a compromise is reached between accuracy and complexity and most texts and references seem to agree on the Friedlander wave equation.

Asher Gehl 9913569

57

Blast, Stress and Shock Wave Theory


4.7.2 Friedlander Wave Equation

In 1939, Friedlander was credited with being one of the first to mathematically approximate the pressuretime history of a blast wave. According to Baker (1983)
P (t ) = P0 + Ps 0 (1
bt t )e t A + t o t A + t0

(4.6)

Where (as shown in Figure 4.4) P(t) is the overpressure at time t, tA + t0 (the positive duration) is the time for the pressure to return to
B B B B

atmospheric pressure, P0. By selecting a value for the constant b, various pressuretime histories can be described. The peak pressure, Ps0, is dependent on the distance from the charge and the weight of the explosives. In addition, if the peak pressure (Ps0), the positive impulse (is) and the positive time duration (t0) are known, the constant b can be
B B B B

calculated, and then the pressuretime history is known. The computer program CONWEP (Hyde, 1991) used for the design of subsequent Chapters utilises the Friedlander wave equation in its calculations of the blast wave. The assumptions presented in using this equation, are presented in Appendix E3.
4.7.3 Idealised Pressure-Time History

For design purposes, it is necessary to establish the decay of both the incident and dynamic pressures with time. The response of the structure is essentially dependant upon the impulse as well as the peak intensity. This is discussed further in Chapter 6. The decay rate is a function of the peak pressure and the explosive size. In order to determine the decay time an idealized form of the wave is used as shown below as the dotted line in Figure 4.5.

Asher Gehl 9913569

58

Blast, Stress and Shock Wave Theory

Figure 4.5 Idealised Pressure time Variation (TM5-1300, 1991) Two main approximations are utilised in the idealised equivalent triangular pressure pulse. 1. Positive Phase history The most important approximation is ignoring the effect of the negative phase of the pressure time history. The design manual TM5-855-1 (Hyde, 1991) suggests that approximations of the loading history are considered satisfactory when the peak pressure of the loading and the total positive phase impulse is preserved between the approximations and the actual history. 2. Linear Decay The approximate pressure time histories are generally presented as a triangular pulse with zero rise time preserving the magnitudes of the peak pressure and total impulse.

Asher Gehl 9913569

59

Blast, Stress and Shock Wave Theory The Friedlander equation (Equation 4.6) is often simplified with a triangular pressuretime curve (Bulson, 1997) P (t ) = P0 + Ps 0 (1 t ) t A + t0
(4.7)

Kinney and Graham (1985) conclude that a fundamental difference exists between normal shockwave theory, and the blast wave theory. Most significant of the differences is the spherical divergence of the blast wave and its inherent transient nature. These properties further complicate the purely mathematical approaches above. These fundamental gas dynamics equations can be related to air blast theory by utilising the Rankine Hugoniot relations as presented below in section 4.8.2.

4.8 Blast Wavefront Parameters


4.8.1 Techniques Available

These parameters from the pressure-time history are directly related to the weight of explosive source and standoff distance. These parameters are the main determinants of structural response. The most crucial and significant of the parameters are the following: Peak static overpressure Dynamic pressure Impulse Positive Phase duration

As previously mentioned, in addition to the wide variety of equations to be presented below by various references and the compiled air blast curves by Kingery and Bulmash (1984) there are also a number of computer programs available to calculate the magnitudes of these blast wave parameters. For the preliminary flexural design of Chapter 7, the CONWEP (Hyde, 1991) program was used to determine the blast wave parameters. Baker (1983) outlines the variations of experimental results to computational and theoretical solutions.

Asher Gehl 9913569

60

Blast, Stress and Shock Wave Theory There are a variety of key texts outlining the derivations of these key parameters. The parameters are based on computations on sets of experimental results that were primarily performed after WWII. A convenient method of representing these parameters is to plot them against scaled distance. Initially presented in Kingery and Bulmash (1984), these curves utilise the Hopkinson-Cranz scaling law (Equation 4.2) for TNT charges at atmospheric pressure and are presented in the majority of the referenced texts. Hemispherical air blast parameters are presented below. TM5-1300 (1991) presents the equivalent air burst parameters. Where: Pr Pso ir is tA t0 U Lw Peak reflected Pressure Peak Static Overpressure Reflected Impulse Specific impulse Arrival time Positive phase Duration Blast wave front Velocity Scaled Distance (Z in Section 4.4)

Asher Gehl 9913569

61

Blast, Stress and Shock Wave Theory

Figure 4.6 Positive Phase shock wave parameters for a hemispherical TNT explosion on the surface at sea level. (TM5-1300, 1991) Baker (1983) thoroughly discusses the limitations of compiled air blast parameters. The main issue with these curves is that due to the complication of the derivations, the values of the parameters, although based on a scaled distance, differ from one set of experimental or computational results to another. Hence to some extent, one can obtain a different prediction of blast overpressure, impulse and duration depending on the source data. The equations required to determine air blast parameters are presented below.

Asher Gehl 9913569

62

Blast, Stress and Shock Wave Theory


4.8.2 Predicting Blast Parameters: Rankine Hugoniot relations

Once the profile of the blast wave can be understood and applied the next issue is to determine the value of the fundamental blast parameters. There are a variety of techniques available in which this can be achieved. Rankine and Hugoniot first analytically presented the blast parameters in 1870. Known today as the Rankine Hugoniot relations, they describe the behaviour of normal shock in ideal gases. They essentially allow the fundamental gas dynamics equations to be transformed for air blast theory. They take into account the added complexity due to the viscosity and heat conduction effects that are associated with the steep gradients of shock fronts from blast waves. Both Baker (1983) and Kinney and Graham (1985) present these Rankine-Hugoniot conditions and outline their derivation and application. Bakers (1983) terminology is presented below. By choosing a fixed reference in space (the shock front), where the material motions are determined with respect to that region, the fundamental shock wave equations, known as the Rankine-Hugoniot equations, are derived from the equations for conservation of mass, momentum, and energy in the medium. Consider the onedimensional model (Leppanen, 2002) in Figure 4.7, where the material is moving with a velocity of U0 against the shock front, and the material velocity is U1 after passing the shock front. The pressure is P0 and the density 0 before the material reaches the shock front, and the pressure is P1 and density 1 after passage.

Figure 4.7 Model for one-dimensional shock waves (Leppanen, 2002)

Asher Gehl 9913569

63

Blast, Stress and Shock Wave Theory Consider the conservation of mass flow per unit time and area in the model of Figure 4.7. It can be expressed for times t0 and t1 as m = 0U 0 At = 1U 1 At 0U 0 = 1U 1
(4.8)

Considering the conservation of momentum, mass times the change in velocity, is equal to the impulse of external forces. By using the above equation, the conservation of momentum can be derived, where m is the flow of mass per time unit and area. m(U 0 U 1 ) = P1 P0 The change in internal energy and kinetic energy is equal to the work done by external forces. It can be shown, as in Baker (1983) that the conservation of energy per unit mass can be expressed as: E1 E 0 = 1 1 1 ( P1 + P0 )( ) 2 0 1
(4.10) (4.9)

The three equations presented above (4.8, 4.9, 4.10) are the Rankine-Hugoniot equations. The Hugoniot curve expresses the relationship for pressure and specific volume as shown below in Figure 4.8. However, the material state is described by a discontinuous jump from one state to another, known as the Rayleigh line (Baker, 1983).

Asher Gehl 9913569

64

Blast, Stress and Shock Wave Theory

Figure 4.8 Hugoniot curve and Rayleigh line of slope k (Leppanen2002)


4.8.3 Peak Incident Overpressure

The maximum overpressure (Pso) acting upon the face of a wall can be determined
B B

from a variety of experimental results and numerical methods. It essentially is the normal pressure upon the face of a surface before any reflections take place. Mays and Smith (1995) outline Brodes (1955) solution for peak static overpressure (Ps0)
B B

from the equations of motion of a blast wave. Brode presented two solutions for a surface close to and further away from the source, in terms of Z When: (Ps0>10bar)
B B

Ps 0 = (0.1<Ps0<10bar)
B B

6.7 + 1bar Z3

(4.11)

Ps 0 =

0.975 1.455 5.85 + + 3 0.019bar Z Z2 Z

(4.12)

Asher Gehl 9913569

65

Blast, Stress and Shock Wave Theory As described previously: Z is the scaled distance R is the standoff distance W is the equivalent TNT explosive weight Kinney and Graham (1985) present another accurate solution where nuclear and chemical charges are compared. The chemical solution is presented here in analytical form and in a more presentable form as overpressure vs. distance.
ps0 808[1 + ( 4Z.5 ) 2 ]
Z 1 + ( 0.048 ) 2 1 + ( 0.Z ) 2 1 + ( 1.Z ) 2 32 35

Z= R

Pa

(4.13)

Where: ps0 Pa is the ratio of the blast overpressure to ambient atmospheric pressure

Z is the scaled distance as above. The above equation is especially useful for computer computations and is also presented below (Figure 4.9) in graphical form. Its corresponding nuclear equation can be found in Kinney and Graham (1985).

Asher Gehl 9913569

66

Blast, Stress and Shock Wave Theory

Figure 4.9 Peak Overpressure ratio (Nuclear and Chemical) vs. Scaled Distance (Z) (Kinney and Graham, 1985) A less accurate solution is approximated by a simpler solution presented in Goschy (1990).
Ps 0 = 6784 / R 3 + 93(W / R 3 )
1 2

(4.14)

Teo (2003) summarises the investigations into blast wave analysis by various researchers as follows.
RESEARCHERS Newmann and Bethe Brode Newmark and Hansen Henrych Kingery and Bulmash Mills YEAR 1944 1955 1961 1979 1984 1987 TYPE OF BURST Spherical free-air Spherical free-air Hemispherical Surface Spherical free-air Spherical free-air and Hemispherical Surface Hemispherical Surface APPROACH Theoretical Theoretical Emperical Semi-Emperical Emperical Emperical

Table 4.1 Summary of Blast Wave investigations (Teo 2003)

Asher Gehl 9913569

67

Blast, Stress and Shock Wave Theory He concludes with presenting the variation of peak static overpressures as predicted by Kingery and Bulmash(1984) Henrych(1979) and Brode(1955). He explains that the lower accuracy of Brode and Henrych is a result of the complicated flows.

Figure 4.10 Comparison of Peak incident Overpressure vs. Scaled distance (Teo 2003)
4.8.4 Positive Phase Duration

The duration of the blast wave is one of the key aspects that determine the level of damage inflicted. The positive phase duration (td) is a fundamental parameter that determines the response of a structure and the resultant procedure required for design. As previously outlined the positive phase duration is more damaging and can be measured and predicted more precisely. Thus the majority of sources agree that the positive phase duration can be taken as an indication of the overall duration (Mays and Smith, 1995; Goschy, 1990; Kinney and Graham, 1985; Baker, 1983). Kinney and Graham (1985) present similar computational equations for the values of positive phase duration for both nuclear and chemical explosives for comparison.

Asher Gehl 9913569

68

Blast, Stress and Shock Wave Theory


4.8.5 Blast Wave Impulse

Impulse (Equation 4.4, 4.5) is considered to be the controlling parameter for the damage caused by a blast (Kinney and Graham, 1985). The area under the pressure time curve indicates the impulse, and thus it is generally the positive phase impulse that is of consequence in design. The impulse is not only dependent upon the peak overpressure and time duration of the blast wave, but also significantly upon the rate of decay of the overpressure. To illustrate this, blast wave of a nuclear and chemical explosions with equal peak overpressure and time duration are compared (Figure 4.11) as in Kinney and Graham (1985).

Figure 4.11 Blast Wave overpressure of Nuclear and Chemical Explosions of equal duration (Kinney Graham) It can be seen that the nuclear explosion (Figure 4.11(A)) decays faster and therefore has a smaller impulse than the chemical explosion (Figure 4.11(B)) even though the peak overpressure (Pso) and the durations (td) are the same magnitude.

Asher Gehl 9913569

69

Blast, Stress and Shock Wave Theory

4.9 Type of Explosion


There is a significant difference in the design method ultimately used, depending on the explosion type. The threat assessment will determine whether any or all of the following attacks are of feasible consequence. As discussed earlier this thesis is limited to the study and effects of external surface bursts, however the other types of explosions are presented briefly here for thoroughness and perspective. For further analysis into these concepts, the majority of texts and the technical manual TM5-1300 (1991) go into greater detail.
4.9.1 Air Burst

A shock wave resulting from an explosive detonation in free air above the ground is termed an air-blast shock wave, or simply a blast wave. The blast environment will differ depending on where the explosion takes place. In the case of an airburst, when the blast wave hits the ground surface, it will be reflected. The reflected wave will interact with the incident wave and a Mach front is created, as shown in Figure 4.12. The point where the three shock fronts meet incident wave, reflected wave and the Mach front is termed the triple point. For the purposes of design, the Mach front is considered as a plane wave over its height such that it imparts a uniform pressure when it interacts with a structure.

Figure 4.12 Air Burst Blast environment (TM5-1300, 1991)

Asher Gehl 9913569

70

Blast, Stress and Shock Wave Theory


4.9.2 Surface Burst

In the case of a surface burst, the reflection happens instantaneously against the ground surface and a single hemispherical shock wave is created. This wave is much like the Mach front of an airburst and is termed a ground-reflected wave, as shown in Figure 4.13. In comparing the plots of positive phase parameters versus scaled distance for surface bursts (Figure 4.6) and airbursts in TM5-1300 (1991), at a given distance the surface burst parameters are higher. At a short distance from the burst, the wave front can be approximated by a plane wave and therefore the blast loads on a structure can be determined with the same methods as an airburst by substituting the surface burst blast wave parameters.

Figure 4.13 Surface Burst Blast Environment (TM5-1300, 1991)


4.9.3 Confined Explosion

When an explosion occurs within the interior of a structure, the large number of reflections will considerably amplify the blast wave. Additional gas pressures are also contributed by the gaseous confinement at high temperature from the detonation. TM5-1300 (1991) presents the analysis and design methods for confined explosions. For complicated geometries however the use of Computational Fluid Dynamics (CFD) is recommended.

Asher Gehl 9913569

71

Blast, Stress and Shock Wave Theory

4.10 Blast Wave Reflections


When a blast wave strikes a surface, which is not parallel to its direction of propagation, a reflection of the blast wave takes place. The reflection can be either normal reflection or an oblique reflection. There are two types of oblique reflection, either regular or Mach reflection; the type of reflection depends on the incident angle and shock strength. Of considerable interest are the reflections that occur when the explosion is in contact with the ground, as with many car bomb explosions. A significant amount of the energy is imparted in the creation of ground vibrations and the formation of a crater, as seen in Figure 2.12. In addition to the crater, the magnitude of the reflected pressure is considerably higher than the magnitude of the incident pressure. For an efficient reflective surface the increase is double, however it has been determined that the actual value is closer to a 1.8 fold increase, due to the energy involved with the crater formation and ground shock (Medearis, 1975).
4.10.1 Normal Reflection

Normal reflection takes place when the blast wave hits perpendicular to a surface, as shown in Figure 4.14. As the incident blast wave expands and hits the surface, the movement of the air molecules in the plane of the shock front are abruptly terminated. This further compresses the molecules at the surface and induces an associated reflected overpressure on the wall. This reflected overpressure is of considerably higher magnitude than the incident overpressure of the surfaces immediate surroundings.

Figure 4.14 Normal reflections in air from a rigid surface (Baker, 1983)

Asher Gehl 9913569

72

Blast, Stress and Shock Wave Theory TM5-1300 (1991) presents the comparison between peak incident pressures to normally reflected pressure, shown below in Figure 4.15.

Figure 4.15 Peak Incident Pressure vs. Ratio of Normally Reflected Pressure/Incident Pressure for a Free Air Burst (TM5-1300, 1991) The properties of the reflected blast wave can be described in terms of a reflection coefficient, defined as the ratio of reflected overpressure to the incident overpressure in the blast wave. It can be shown that for an ideal gas with a specific gas constant ratio of 1:4, the reflection coefficient is, according to Kinney and Graham (1985) and Baker (1983).
Pr Px 8M x = Py Px M x 2 + 5
2

(4.15)

Asher Gehl 9913569

73

Blast, Stress and Shock Wave Theory This equation expresses the reflection coefficient in terms of the Mach number Mx for the incident shock wave. The Mach number is a dimensionless index, defining the ratio of a velocity to the speed of sound (Kinney and Graham, 1985). Limiting values of the reflection coefficient can thus be simply found. According to Kinney Graham (1985), at the lower limit, for a very weak shock front moving with Mx approaching one, (i.e. at sonic speed), the reflection coefficient will approach 2. This means that the overpressure is twice in the reflected blast wave. For stronger incident shocks, with increasing speed for the shock front, Mx, the reflection coefficient is increased. An upper limit is stated in Kinney and Graham (1985) as 8. This limit is based on the assumption that air behaves as an ideal gas with the heat capacity ratio (specific gas constant ratio) of 1.4. Baker (1983) cites references where greater limits, up to 20 have been deduced taking into account real gas effects such as dissociation and ionization of the air molecules. In a real blast wave, the specific gas constant ratio is not constant, and the coefficient is pressuredependent. The reflection coefficient increases with increasing pressure. Kinney and Graham (1985) plot reflection coefficients for selected Mach numbers over a range of incidence angles as shown in Figure 4.16.

Asher Gehl 9913569

74

Blast, Stress and Shock Wave Theory

Figure 4.16 Reflection Coefficient vs. Angle of incident shock (Kinney and Graham, 1985) TM5-1300 (1991) presents the Figure in terms of peak incident overpressure as shown below in Figure 4.17.

Figure 4.17 Reflected Pressure Coefficient vs. Angle of incidence (TM5-1300, 1991)

Asher Gehl 9913569

75

Blast, Stress and Shock Wave Theory It can be seen in the above Figures (4.16, 4.17) that as incident angles increase the coefficient passes through a minimum value before rising at around 40o. This is a
P P

critical angle that is discussed further with respect to the Mach stem formation.
4.10.2 Regular Reflection

In presenting the conditions of regular reflection, the same terminology is used as normal reflection above. In a regular reflection the blast wave shock Mx is incident upon a surface with an
B B

angle of and reflection takes place. The reflected shock at Mr has an angle of as shown in Figure 4.18. The angle of reflection is not generally equal to the angle of incidence. Regular reflection exhibits similar reflection coefficients to normal reflection discussed above. As the angle of incidence decreases to zero, regular reflection converges into normal reflection. The air conditions in front of the incident shock (region 1) are still at pressure Px and temperature Tx. Behind the incident shock (region 2), the air conditions are the same as for free air shock, with pressure Py and temperature Ty. The air conditions from the reflected shock (region 3) have the pressure Pr and temperature Tr.

Figure 4.18 Oblique reflection (Baker 1983)


B

Asher Gehl 9913569

76

Blast, Stress and Shock Wave Theory TM5-1300 (1991) presents the magnitude of reflected overpressure as it varies with angle of incidence (Figure 4.19).

Figure 4.19 Variation of reflected pressure as a function of angle of incidence (TM51300, 1991)
4.10.3 Mach Stem Formation

When the angle of incidence in Figure 4.18, exceeds crit, regular reflection will no
B B

longer occur and a new phenomena known as Mach reflection emerges. This critical angle depends on the shock strength where an oblique reflection cannot occur, and has been determined to be 40o (Kinney Graham, 1985). This value has been determined
P P

through a series of computations and is presented here as Figure 4.20.

Asher Gehl 9913569

77

Blast, Stress and Shock Wave Theory

Figure 4.20 Transition angle from regular reflection to Mach stem formation (Kinney and Graham, 1985) Figure 4.20 above shows that very weak incident shocks (Sound waves - Mach number (M) close to unity), are reflected at all angles up to 90o. As shock intensity
P P

increases, the limiting angle decreases rapidly, such that for strong explosive shocks with an angle of incidence greater that 39.97o, Mach stems will always be generated
P P

rather than shocks with oblique/regular reflection. This data has been modelled as an empirical formula for comparison purposes with high precision.

max =

1.75 + 39 ( M 1)

(4.16)

According to Baker (1983), Ernst Mach (Mach and Sommer, 1877) showed that the incident and reflected shock coalesce to form a third shock front at a point above the reflecting surface. The created shock front is termed the Mach stem or Mach front, which is moving approximately parallel to the ground surface, as shown in Figure 4.21, with increasing height of the shock front. The point where the three-shock fronts meet is termed the triple point. Behind the Mach stem is the slipstream region where pressures are the same, but there are different densities and particle velocities. The Mach front and the path of the triple point are also shown in Figure 4.12.

Asher Gehl 9913569

78

Blast, Stress and Shock Wave Theory

Figure 4.21 Mach stem formation (Baker 1973)

The importance of a Mach stem is evident for a number of instances (Baker, 1983) When a device is detonated inside a structure in a confined explosion, angles of incidence of the blast wave vary over a wide range and thus Mach fronts will occur. If the charge explodes close to the surface the Mach stem forms directly under the charge. If the charge explodes in contact with the ground then no separate reflection wave occurs and the entire shock front is considered as a Mach wave. If the charge explodes at a certain height above the ground the Mach front forms at a certain horizontal distance away from the charge shown below in Figure 4.22.

Figure 4.22 Reflection of strong Shock Waves from an air burst (Baker 1983)

Asher Gehl 9913569

79

Blast, Stress and Shock Wave Theory It is the latter that has attracted a great deal of military interest over the years. The pressure and positive impulse in the Mach stem and triple point region are substantially higher than from the other two areas mentioned above. Such that a bomb exploding above the ground will have a greater Mach front intensity at a certain horizontal distance, than if the bomb had exploded in contact with the ground. A large amount of testing has been performed whereby the exact height for maximum intensity Mach stem formation has been ascertained. This is especially useful for military tactical bombing situations. In Chapters 1 and 5 of Baker (1983), further information regarding the strength of the Mach stem and experimental techniques is presented.

4.11 Blast Wave Diffraction


As a blast wave passes over the corner of a surface, the wave diffracts around the surface. A large variety of research has been conducted on diffraction effects from blast waves on buildings. Baker presents the diffraction effects of a blast as it wraps around both a cylinder and a rectangular surface and is recommended for further analysis. The following figure demonstrates the blast pressure as it wraps around the roof of a building.

Asher Gehl 9913569

80

Blast, Stress and Shock Wave Theory

Figure 4.23 Blast Wave Diffracting Around a Roof (Armstrong, Rickman, Baylot, and Bevins, 2002) The forces on the surface of the building as the blast reaches them are presented in the following Chapter.

Asher Gehl 9913569

81

Blast Wave Loading

Chapter 5 Blast Wave Loading


5.1 Introduction
The high intensity pressures, of the air-blast, primarily cause the damage from an explosion. When the shockwave reaches a building it will induce localised failure of exterior windows, walls, floor systems, columns, and girders. The general damage of a building as the blast wave hits is summarised by the following figure. This chapter however examines the external loading on a structure and does not consider the effect upon structural elements as shown below in Figure 5.1.

Figure 5.1 Blast pressure effects on a building (FEMA 426, 2003)

Asher Gehl 9913569

82

Blast Wave Loading The discussions of the last chapter have so far only considered the blast wave reflecting of an infinitely large surface. If the blast wave is big enough with respect to the face of the structure, it will diffract around the face. When this phenomenon occurs, all the surfaces of the structure will be exposed to the blast not just the front face as previously considered. The process of determining the final magnitude of the blast wave for the design
T

process is linked directly to the results of the threat assessment for the particular structure. The owners requirements and needs for the facility should be clearly defined. Based on this requirement the level of protection required should be determined. The details of the assessment should provide the designer with the explosive weight and standoff distance that is of highest risk for the facility. Due to the impossibility of predicting the ultimate location of the explosion, the building should be designed for each worst case situation. Due to the fact that the blast wave decays over time and distance the designer should have separate analysis cases for each face of the structure. For the purposes of structural analysis, the resultant blast loading can be simplified, despite its complex nature. The blast loading on a building component depends on the magnitudes of the overpressure, impulse or duration and is dependant upon the orientation of the structure relative to the explosive source. The net load on the frame of the building is taken by taking into account the relative time phases of the load combination on each face of the frame. With respect to the overall structure, there are three specific interaction options that are possible. Large Blast Large Structure

The structure will be engulfed and compressed by the wave with unlikely probability of overall movement due to the blast loading. This option is generally termed: Diffraction Loading

Asher Gehl 9913569

83

Blast Wave Loading Large Blast Smaller Structure

The structure will be completely engulfed as above; however the crushing will be more severe, resulting in the movement of the structure. This option is generally termed: Drag Loading Small Blast Large Structure

The impact of Diffraction and Drag loading on the overall structure can be seen in Figure 5.13 in Section 5.3.4 where the force acting on a structure over time is illustrated. TM5-1300 (1991) divides the forces imparted upon a structure by the incident and dynamic pressures of a blast wave into the following four generalised components: 1. The force resulting from the incident pressure 2. The force associated with the dynamic pressures 3. The force resulting from the reflection of the incident pressure impinging upon an interfering surface 4. The pressures associated with the negative phase of the shock wave. The relative significance of each of these components is dependent upon the geometrical configuration and size of the structure, the orientation of the structure relative to the shock front and the design purpose of the blast loads. As a result there are different methods used to determine the loads acting on a structure. The uniformity of the blast pressure as it reaches the structure generally determines the approach taken. Blasts at a considerable distance from the target surface exert a uniform pressure on the surface. Close range explosions apply a non-uniform pressure over the target surface. A third approach is required when the geometry of the structure is non-uniform and the pressure has considerably complicated effects on the structures surface.

Asher Gehl 9913569

84

Blast Wave Loading Uniform Blast Pressures

The forces acting on a structure associated with a plane shock wave are dependent upon both the peak pressure and the impulse of the incident and dynamic pressures acting in the free field. The peak pressures and impulses for undisturbed shock waves are presented for various explosives in TM5-1300 (1991). Non-Uniform Blast Pressure. The forces acting on a structure associated with a non-planar shock front are not as simply determined. There are pressure variations on the target surface and the surface must be divided accordingly. Various programs are available however to calculate these pressure increments over a surface. CONWEP (Hyde, 1991) and AT Blast (ARA, 2000) are examples of such programs. Non-uniform surface. As the shock front (planar or non-planar) strikes a structure with complex surface geometry, there are considerable number of reflections and other interactions that complicate the resultant pressures on each surface. In order to determine these pressures advanced Computational Fluid Dynamics (CFD) is required. The subsequent discussions in this chapter begin with a brief look into the principles involved with dynamic shock waves. Following this, the loading effects from a planar blast wave on each surface of an above ground structure with no openings are considered. Finally the overall effect on the structure will be determined. The effects of non-uniform pressure and complex geometries will not be considered further. The last section of this chapter discusses the determination of the loading on the individual columns used in the subsequent designs in Chapter 7

5.2 Dynamic Shock Wave Effects and Principles


The blast wave imposes a dynamic load on objects within its field. There is an instantaneous rise time to a peak load value that decreases as the blast wave decays over time. The net load depends both on the nature of the blast wave as outlined in the preceding chapters and on the projected area of the structure.

Asher Gehl 9913569

85

Blast Wave Loading Once the blast hits the structure and diffracts around it there are a number of key principles that involve shock wave theory that must be presented before proceeding further into the description of the loading on separate faces.
5.2.1 Stagnation Pressure

Consider the flow patterns as a steady stream hits a perpendicular surface.

Figure 5.2 Stagnation effects (www.princeton.edu, 2004)


HTU UTH

As the stream flows towards the surface, all the flow above the centre streamline moves over the surface and the flow below the centre streamline goes below the surface. The fluid along the dividing, or ``stagnation streamline slows down and eventually comes to rest (stagnates) without deflection at the stagnation point.
Ps + 1 Vs2 = Pstag 2

(5.1)

The stagnation or total pressure, Pstag, is the pressure measured at the point where the
B B

fluid comes to rest. It is the highest pressure found anywhere in the flow field, and it occurs at the stagnation point. It is the sum of the static pressure (Ps), and the dynamic
B B

pressure (qs) measured far upstream. It is called the dynamic pressure because it arises
B B

from the motion of the fluid (Kinney and Graham, 1985).

Asher Gehl 9913569

86

Blast Wave Loading


5.2.2 Dynamic Pressure

The dynamic pressure of the blast is proportional to the velocity of the fluid. The dynamic pressure is related to the difference between the stream stagnation pressure Pstag and the local pressure P0.
B B B B

qs=Pstag-P0
B B B B B B

(5.2)

There is a particle or wind velocity associated with the blast wave that causes a dynamic pressure on objects in its path. In air, these dynamic pressures are essentially functions of the air density and the blast wind velocity (Kinney and Graham, 1985). Equations have been determined (Equations 5.3, 5.4) for the blast wavefront velocity Us and the maximum dynamic pressure qs based on typical values for the peak
B B B B

incident pressure (Pso), and the speed of sound (ao) at the ambient air pressure (Po)
B B B B B B

(Mays and Smith, 1995).


6 Pso + 7 Po ao 7 P0

Us =

(5.3)

qs =

5Ps2 2( Ps + 7 P0 )

(5.4)

Figure 5.3 (TM5-1300, 1991) gives the values of peak dynamic velocity, air density behind the wavefront and the particle velocity, versus the peak incident pressure. Of the three parameters, the dynamic pressure is the most important for determining the loads on structures.

Asher Gehl 9913569

87

Blast Wave Loading

Figure 5.3 Peak incident pressure (Pso) versus peak dynamic pressure (qs), density of
B B B B

air behind the shock front (s), and particle velocity (Us) (TM5-1300, 1991)
B B B B

Asher Gehl 9913569

88

Blast Wave Loading


5.2.3 Drag Coefficients

Drag forces are normally studied for moving bodies through fluids; however the principles associated with the drag from the blast wind on an object are essentially the same. The main differences stem from the transient nature and considerably higher magnitude of the blast force, and the relationship between Drag Coefficients and high Reynolds number. The Reynolds number is a dimensionless parameter that is dependant upon the velocity of flow (V), the depth of the object in the face of flow (D) and the kinematic viscosity (). Where:
Re = VD

(5.5)

The drag coefficient is a dimensionless ratio of the energy required to overcome drag forces of the fluid at a defined velocity over the kinetic energy of the fluid at that velocity such that:
d drag CD = Energy = kinetic 0.5V 2

(5.6)

Experimental values for drag coefficients have been determined for a wide variety of objects and situations. The value is essentially a function of the geometry of an object and its influence on flow patterns, such that flows around rounded objects stay attached for longer, permit less turbulence at the rear and thus reduces the pressure drop and the drag coefficient. With respect to an explosion however, the velocity of the air is considerably higher than standard winds. As such the Reynolds numbers are considerably higher. At large Reynolds numbers the drag is essentially dominated by the pressure losses, associated with the wake of the flow around an object. (Dr Graeme Wood 2004, Personal Communication, 21 October 2004). The variation of the drag coefficient with Reynolds number is shown in Figure 5.4, and the corresponding flow patterns are shown in Figure 5.5. We see that as the Reynolds number increases the drag coefficient decreases.

Asher Gehl 9913569

89

Blast Wave Loading

Figure 5.4 Drag Coefficient as a function of Reynolds number for smooth circular cylinders and spheres. (Smits, 2004)

Figure 5.5 Flow patterns for flow over a cylinder corresponding to regions in Figure 5.4: (A) Reynolds number = 0.2; (B) 12; (C) 120; (D) 30,000; (E) 500,000 (Smits, 2004)

Asher Gehl 9913569

90

Blast Wave Loading As illustrated in Figure 5.4, for Reynolds number between 105 and 106, the drag coefficient is suddenly reduced. This decrease is consistent with the changes between Figure 5.5(D) and 5.5(E). For low Reynolds numbers (Re < 105) there is a laminar boundary layer and the separation point of the flow occurs close to the front face, resulting in a very large wake and therefore a high drag. Whereas for high Reynolds numbers (Re > 106), there is a turbulent boundary layer and the flow stays attached for longer, resulting in a considerably smaller wake. This phenomenon is better illustrated via the following wind tunnel test at Princeton University.

Figure 5.6 Flow at low Reynolds Number (a) and High Reynolds Number (b) (Smits, 2004) Chapter 7 will continue the analysis with respect to differences between the available drag coefficient magnitudes for sections with varying cross sectional shape, that are presented in TM5-1300 (1991) and Kinney and Graham (1985) and in Appendix D.

5.3 Uniform Blast Pressures


The procedure for the determination of the loads on individual faces is presented in a number of sources (Forbes, 1999; Mays and Smith, 1995; TM5-1300, 1991; Goschy, 1991; Kinney and Graham, 1985). Kinney and Graham provide the most comprehensive description of the procedure and it is presented here for a thorough understanding. The process utilises the geometric dimensions of the following structural model throughout the subsequent sections.

Asher Gehl 9913569

91

Blast Wave Loading

Figure 5.7 Typical Above ground rectangular structure with individual panel distance (d) from the edge To reduce the complications involved with the interaction between the blast wave and the structure and to ensure uniform blast pressures on the structure a number of assumptions are made: The structure is essentially rectangular in shape with no openings The incident pressure of interest is 1400kPa or less The structure being loaded is in the region of the Mach stem The Mach stem extends above the height of the building.

Figure 5.8 Plane Wave front of incident wave (TM5-1300, 1991)

Asher Gehl 9913569

92

Blast Wave Loading


5.3.1 Front Face Loads

As a vertical blast wave directly strikes the front face of an above ground structure the pressure immediately increases as normal reflection occurs and the front face is subjected to reflected overpressures as illustrated below.

Figure 5.9 Loading on a front wall (TM5-1300, 1991) As outlined in section 4.10.2, the reflected overpressure is greater than the surrounding medium, resulting in a flow from high to low pressure regions. This flow travels over the face, progressing inwards from the edges. The procedure in effectively outlining the overall loading on the face begins by considering a panel at some distance from the edge of the incident face. Initially the entire face and the panel, is subjected to the peak reflected overpressure. The airflow immediately begins to diffract around the edges of the face (at t=0). The flow from the Blast wave proceeds from the edge of the face to the area of the individual panel, giving pressure relief to the panel. Consider the time at which the relief begins for an individual panel at distance d from the edge to be t1.

Asher Gehl 9913569

93

Blast Wave Loading Where in Figure 5.9 t1 = d Us


(5.7)

There are a number of different approximations for the clearing time (tc) which is
B B

the time whereby the overpressure on the panel is reduced from peak reflected overpressure to the stagnation overpressure at that instant in time. The Earthquake Engineering Research Institute (1952) cited in Kinney and Graham (1985) found that the relief process requires a time about twice that required for a Sound wave to travel across the entire front face, such that: t c t1 = 2 S
(5.8)

Us

Us is the velocity of the Blast wavefront.


B B

S is the lesser of the height (H) or half breadth of the face (Ws/2).
B B

The load on the entire front face is essentially the summation of the individual panels. Since the panels at the centre maintain the peak reflection overpressures for longer than panels at the edge of the face that are cleared first, the overall time for the entire face to reduce to stagnation overpressure is essentially the summation of each individual panel, resulting in the following equation: t c t1 = 3S
(5.9)

Asher Gehl 9913569

94

Blast Wave Loading Another more accurate equation for the clearing time is presented in TM5-1300 (1991) whereby the clearing time is expressed in milliseconds according to: t c t1 = 4S (1 + R)C r
(5.10)

Where: R is the ratio S/G Where G is the larger of H and Ws/2


B B

and S is the Clearing distance: the lesser of H and Ws /2


B B

Cr is the sound velocity in the reflected region.


B B

The pressure acting on the front wall after time tc is the sum of the incident pressure
B B

Ps and the drag pressure CDqs or:


B B B B B B

P = Ps+ CDqs
B B B B B B

(5.11)

The drag coefficient CD varies with the geometry of the structure, Reynolds Number
B B

and the Mach number of the high-pressure blast. As discussed it relates the dynamic pressure to the total translational pressure in the direction of the wind from the dynamic pressure. TM5-1300 (1991) assumes a value of CD=1 for the front wall is adequate for the
B B

pressure ranges. At higher-pressure ranges, a fictitious pressure-time curve is preferred because of the extremely short pressure pulse durations involved. The fictitious curve is seen in Figure 5.9 above. The fictitious duration trf for the normal reflected wave is calculated from:
B B

trf = 2ir / Pr
B B B B B B

(5.12)

ir is the total reflected pressure impulse


B B

Pr is the peak normal reflected pressure.


B B

Whichever curve (Figure 5.9) gives the smallest value of the impulse (area under curve), that curve should be used in calculating the wall loading.

Asher Gehl 9913569

95

Blast Wave Loading Chapter 4 previously discussed that generally only the positive phase is required for design purposes. However TM5-1300 (1991) as indicated in the above figure presents the effect of the negative phase. Its major impact is in determining the overall motion of the structure. In addition to the negative phase the manual also determines the front face loading at various angles of incidence.
5.3.2 Rear Face Loads

In most design cases, the primary reason for determining the blast loads acting on the rear wall is to determine the overall drag effects (both front and rear wall loadings) on the building. To achieve this, the same equivalent uniform pressure method for the front wall is applied to the rear wall (Kinney and Graham, 1985).

Figure 5.10 Rear Wall Loading (TM5-1300, 1991) In order for the blast wave to induce a load upon the rear face of the structure it must first travel the entire length of the structure L. Sticking with the concept of first considering the load upon an individual panel as outlined for the front face, the compression wave must also therefore travel the distance d from the edge of the rear face to the panel. Thus before the blast wave reaches the rear face, it is subjected to no loading effects from the blast wave.

Asher Gehl 9913569

96

Blast Wave Loading If the shock front is travelling at a velocity Us, then the time taken to reach the rear face is L/Us and the time taken to reach the panel is d/Us. After time L/Us+d/Us (denoted tb in Figure 5.10) the overpressure reaches the panel and rises from zero to a peak overpressure value of: Pstagnation-Pdrag
(5.13)

The pressure build up on the face occurs over a similar time duration to the front face clearing time tc. There is of course an additional time associated with the shock front reaching the face and the panel, such that for the individual panel the associated pressure build up time is 2S/Us. The entire load on the rear face is a summation of the load on the individual panels and is a function of the drag pressures in addition to the incident pressure. This results in an average time for the overpressure to reach the rear face of L/Us rising from zero to the peak value of Pstagnation-Pdrag over a time period of 3S/Us. The overpressure then decays over the duration of the blast wave td. The dynamic pressure of the drag corresponds to that associated with the equivalent pressure CE Psob, while the recommended drag coefficients are the same as used for the roof and side walls that are presented in Table 5.1 (TM5-1300, 1991).
5.3.3 Side and Roof Loads

As the shock front of the blast wave crosses a structure, the resultant dynamic loads on the side faces and top of the structure are taken from the values for the side on overpressure. For these faces reflection and stagnation effects are irrelevant. There are however, minor vortex effects near the edges of the face that are associated with the flow separation as the blast wave diffracts around the edge as seen in Figure 4.23. These vortex effects are associated with drag coefficients and act as drag pressure that reduces the pressure on the face.

Asher Gehl 9913569

97

Blast Wave Loading The methods used to determine the overall loading on the side or top face of the structure, is the same as previously considered for the front face. However the stepby-step integration of the pressure over time at individual panels is achieved by simply taking an average of the front and rear edges as the two limiting panels (Kinney and Graham, 1985). For the side and top faces there is an instantaneous rise time to side on overpressure once the shock front has reached the panel. Thus the front and rear edges are separated by the time taken to travel the length of the structure L/Us as described for the rear face. Thus for the overall pressure on the side or top face, the pressure rises linearly from the time at which the blast reaches the front edge (tf in Figure 5.11) to the peak value of the average uniform pressure of the two limiting edges at the time L/Us when the blast reaches the far edge (denoted as tf +td in Figure 5.11)). The pressure then follows an intermediate decay curve to a final time of L/Us + td (denoted as tf +tof in Figure 5.11).

Figure 5.11 Roof and Side wall loading (TM5-1300, 1991)

Asher Gehl 9913569

98

Blast Wave Loading To simplify the peak value of the pressure acting on the roof PR, TM5-1300 (1991) assumes the blast wavefront parameters across the face to be the same as the front edge. Accordingly PR is the sum of contribution of the equivalent uniform pressure and drag pressure: PR = CEPsof + CDqof CE is the equivalent load factor (Figure 2.196 of TM5-1300, 1991) Psof is the incident pressure occurring at point f qof is the dynamic pressure corresponding to CE Psof. The drag coefficient CD for the roof and side walls is a function of the peak dynamic pressure. Recommended values are as follows:
Peak dynamic pressure (kPa) Drag coefficient (5.14)

0-175 175-350 350-900

-0.4 -0.3 -0.2

Table 5.1 Drag coefficient Roof and Side Walls (TM5-1300, 1991)
5.3.4 Overall Loading on Structure/Frame

The resultant load on the structure as a whole is taken as the instantaneous summation of the load upon the front and rear face of the structure (accounting for the shock front travel time over the length of the structure L). It is important to note however that when the shock front arrives at the front face of the structure, there is significant amplification of the incident front resulting in the immediate failure of windows and doors, unless they are appropriately designed to withstand the load. The openings allow the blast pressures to flow into the interior of the structure. The shocks continue to reflect and combine throughout the buildings interior. TM5-1300 (1991) presents the determination of the loads for structures with openings and for a thorough design, the process is recommended. However for this thesis the confinement and interior loading effects are not considered.

Asher Gehl 9913569

99

Blast Wave Loading

Figure 5.12 Load curves for the front and rear faces (a) and overall effect on the frame (b) (Kinney Graham, 1985) The summation curves of Figure 5.12, taken from Kinney & Graham (1985) above, show how the overall load on the structure from the front and the rear faces are combined to provide the overall load on the frame. The dotted curves represent the side on overpressure values from the incident blast wave. There is an overall negative portion of the loading, towards the explosive source. This effect is independent of the negative phase of the pressure history and is related to the length of the structure and the blast duration, such that for larger L and td, there would be no negative loading phase (Kinney and Graham, 1985).

Figure 5.13 External force-time profile for a structure (Mays and Smith, 1995) Asher Gehl 9913569 100

Blast Wave Loading Similarly Mays and Smith (1995) present the overall loading of a structure as presented in Smith and Hetherington (1994) as shown above as Figure 5.13. This is essentially the same curve as Figure 5.12. From both these figures it can be seen that there is an initial rise in pressure as the blast wave hits the front face of the structure (t2 in Figure 5.13). The increase is due not only to reflection effects for large areas but also associated with the dynamic pressure and the drag loads from the blast wind. For smaller areas in the face of the blast pressure the time taken to reach the other side of the structure is small and thus in Figure 5.13 the time t4 is moved to t4 closer to the time t2 and thus the area in region I is reduced. The same argument decreases the area of the reduction that results in the void area. The broken red line in Figure 5.13 above essentially indicates the change for small elements. The loads on the sides of the structure effectively cancel each other out with a mild compressive net effect. The compressive loading on the top of the structure is of larger magnitude than on the sides, however generally members in this area are traditionally designed to resist such vertical compressive loads. A major source of the damage on the structure is generally a result of the loading imbalance caused by the overall difference between the front face and the rear face of the structure, which is a rare design standard for the majority of structures (Mays and Smith, 1995).

Asher Gehl 9913569

101

Blast Wave Loading

5.4 Blast Loading on Individual Columns


For the subsequent designs of individual columns in Chapter 7 several assumptions are made to simplify the analysis: Explosive size

In order to appropriately design an individual column to withstand the blast pressure, considerable limitations have been put on the size of the bomb. Conventional large-scale terrorist attacks as seen in Chapter 2 utilise large amounts of explosives in the order of 1000kg to destroy the target building. The pressures on an individual column however are extreme and sections required are too massive. Thus generally the procedure to mitigate the damage is to design against progressive collapse as outlined in Chapter 8. A smaller scale attack involves explosives in the order of 100kg detonated at the front of a building. The standoff distance is taken as 5m from the target element to be designed. Uniform Pressure Distribution

Although technically the blast pressures from a close-in explosion are considered non- uniform over the face of a structure, it is assumed that the pressures on an individual element can be determined using a normal incident blast wave. The blast wave parameters were determined using the CONWEP (Hyde, 1991) program for the aforementioned charge size and standoff distance. Effect of Clearing time

The clearing time associated with an individual element is essentially instantaneous. Due to this the value for tc is taken to be zero and thus the value for
B B

reflected pressure Pr is ignored. Accordingly the peak pressure taken for the
B B

determination of the load on the column is taken to be equal to P=Ps0+CDq0


B B B B B B

(5.15)

Asher Gehl 9913569

102

Blast Wave Loading

Figure 5.14 Loading of individual element Where


2 5Pso q0 = 2( Ps 0 + 7 P0 )

(5.16)

Ps is the peak static overpressure determined by CONWEP (Hyde, 1991)


B B

Po is the ambient pressure


B B

The impulse is equal to the area under the Figure as indicated. i= Pt of 2

(5.17)

Where td is the duration of the positive phase determined by CONWEP (Hyde, 1991)
B B

Asher Gehl 9913569

103

Structural Response

Chapter 6 Structural Response


6.1 Introduction
When subjected to an extreme loading event, structures may fail in a number of possible ways. Based on load characteristics, proximity and intensity of the blast, the response of the structure determines the mode of failure. The blast wave damages a structure in its field by causing it to deform. There is a vast range of possible deformation, from trivial damage through to total destruction. For design purposes it is widely agreed that the structures final state of deformation (Xm)
B B

that is caused by the blast is of foremost concern, as opposed to establishing the entire time-displacement history of the deformation (Conrath et al, 1999; Mays and Smith, 1995; TM5-1300, 1991; Kinney and Graham, 1985). The final state of deformation is determined through an analysis process that examines the dynamic response to the transient load. The analysis method presented involves the idealization of the load and resistance functions, as well as converting the real system to an equivalent system so that the displacement-time response of the equivalent system is exactly the same as that of a particular point of interest in the real system. This is achieved through the use of transformation factors that equate the energies of both the real and equivalent systems. In establishing the equivalent systems, considerations are given to the following key principles: Equivalent Single Degree of Freedom Models Relationship between the duration of the blast load and the natural period of the structure Based on the above relationship various classifications of Pressure Impulse diagrams and resistance functions will be introduced in order to determine the final response of the system as a deflection.

Asher Gehl 9913569

104

Structural Response The characteristic behaviour of steel and concrete elements under the dynamic load will be addressed separately outlining the increases in design strength under the transient dynamic load of a blast wave. The design processes performed for steel and concrete columns will then be presented utilising the material presented in TM5-1300 (1991) that is contained throughout the course of this chapter. The design process is essentially according to the summarised flowchart presented by Mays and Smith (1995) that is shown in Appendix F. Finally a variety of differences between reinforced concrete and steel protective structures are presented.

6.2 Principles of Dynamic Analysis


6.2.1 Fundamental Characteristics of Vibration

Vibration is a time-varying response to a change in motion of a structure in such a manner that inertia forces are imposed that resists the change. Only when inertia forces arise and affect the system, is the phenomenon considered a phenomenon of dynamics (Thomson, 1983). In the presence of inertia forces, Newtons Second Law governs the system such that it is in a state of dynamic equilibrium (dAlemberts Principle Equation 6.4). If the system is disturbed and the disturbance removed, the system tries to return to its former state and vibrates on its own until it does so. This is called free vibration. Vibration due to an external time-varying disturbance is called forced vibration. During free vibration, the structure moves from side to side and the time taken to move from one position, back to that original position, is called the natural period. Moving from one position, back to that original position, is called making a cycle, so the period, is the time required to make one cycle. Also, the number of cycles made in one second (Hertz, Hz), is the frequency (Thomson, 1983). The nature of dynamic equilibrium is that the system or structure always vibrates as a combination of basic deflected shapes of the structure. These basic shapes are called the modes of vibration. This tendency is evident mathematically, in the solution of the dynamic equilibrium equations that govern the motion.

Asher Gehl 9913569

105

Structural Response To complete the review of basic vibration phenomena, it must be mentioned that free vibration is always accompanied by damping. This is the decay in the free vibration over time due to friction. If there is too much damping, the structure will not oscillate and is called over-damped and the limiting value of damping required to achieve over-damping is called the critical damping (Thomson, 1983). The negligible effect of damping as a result of blast loading will be discussed further in subsequent paragraphs.
6.2.2 Fundamental Properties

Dynamic analysis typically involves the consideration of three fundamental properties 1. The work done The evaluation of the work done is dependant upon knowing the displacement at a point on the structure under load. 2. The strain energy The strain energy of a system is equal to the summation of the strain energies in all the structural elements, which may be in bending, tension, compression, shear or torsion. 3. The kinetic energy The kinetic energy of a system consists of the energy of translation and rotation of all the masses of the structure. To determine these quantities for a structure under a transient dynamic load is a complicated process. Various assumptions are presented that appropriately reduce the complications involved in the analysis and allows for the solution of practical problems.

Asher Gehl 9913569

106

Structural Response

6.3 Single Degree of Freedom System


A Single Degree of Freedom (SDOF) system is essentially the representation of a real structure as an equivalent single lumped-mass that is restricted to movement in one direction only and as such only on coordinate is required to define its motion. Hence it is called a single degree of freedom system. Under the action of short duration blast loading the fundamental mode of vibration primarily defines the behaviour of the structure, such that the somewhat otherwise complex analysis, may be simplified by an equivalent single degree of freedom system (TM5-1300, 1991).
6.3.1 Equivalent Systems

The technical manual TM5-1300 (1991) outlines the method by which a real system may be replaced by an equivalent system. The controlling criterion is for the energies of the two systems to remain equal to each other. To arrive at a simplified system the following assumptions and approximations are made: The distributed masses of the given structure are lumped together into a number of concentrated masses; The strain energy is assumed to be stored in several weightless springs; The distributed load is replaced by a number of concentrated loads acting on the concentrated masses. Therefore, the equivalent system simply consists of a number of concentrated masses joined together by weightless springs and subjected to concentrated loads, which vary with time (TM5-1300, 1991). This concentrated mass-spring-load system is defined as an equivalent dynamic system and is illustrated via the following real and equivalent systems. The parameters for transforming a real structure to an equivalent system will be subsequently shown in section 6.3.3.

Asher Gehl 9913569

107

Structural Response

Figure 6.1: Real and Equivalent SDOF system (Mays and Smith, 1995)
6.3.2 SDOF System

It is possible to establish a Multi Degree Of Freedom (MDOF) finite element representation of the structure in order to determine its response to an explosive load. This is the generally the analysis procedure for oscillatory loading situations such as earthquakes. However in a blast loading event there is generally a non-oscillatory response and only the peak response is required. It is widely agreed that a MDOF system is usually considered to be inefficient considering a SDOF model affectively accounts for the transient nature of the load and the dynamic structural response (Conrath et al, 1999; Mays and Smith, 1995; Goschy, 1990; TM5-1300, 1991). Real structures of course, have MDOF systems where each mass particle has its own equation of motion, independent of the rest. Many of these structures cannot be appropriately represented by the SDOF system and require a MDOF system due to the consideration of dynamic strain. The analysis of these systems is considered in Section 3.19.2 of TM5-1300 (1991) and is not considered further in this thesis.

Asher Gehl 9913569

108

Structural Response It is important to note that more advanced computational approaches have been developed to improve the accuracy of response calculations. Many people are improving the research and analysis in this field (Krauthammer, Frye Schoedel and Seltzer, 2003). The equivalent SDOF system consists of a concentrated lumped mass under the action of an equivalent concentrated load, resisted by a weightless spring and dampener system. The transformation process from a real system to the lumped mass system relies on kinematic similarity, whereby the energies of both systems are equivalent (Mays and Smith, 1995). The transformation factors are presented in Section 6.3.3. In the process various assumptions are made to considerably simplify the analysis. This simplified analysis is unable to provide the complete history of the structures response. This of course as previously mentioned is not the requirement for blast analysis and the conservative assessment of the maximum deformation is beneficial for design purposes. TM5-1300 (1991) presents three methods appropriate for the solution of simple systems under the impact of dynamic loads. 1. The first of these methods involves the solution of differential equations by classical, numerical or graphical means. 2. The second method utilizes various charts presented in the manual, that present the solutions of the previous method and whereby approximate solutions to the problem may be determined. The SDOF response chart for a triangular load (Figure 6.2) is presented in Appendix C. These charts are utilized particularly for Quasi-static elements that respond to pressure as seen in the design procedure of Chapter 7. 3. The third method involves the solution of the energy equation. This utilizes the fundamental dynamic Rayleigh Ritz method. The derivation of this method is not presented in this thesis but is widely available in a number of the texts. (TM5-1300, 1991)

Asher Gehl 9913569

109

Structural Response
6.3.2.1 Derivation

The key aspect to be determined through the analysis of the SDOF model is the relationship between the loading from the blast and the response due to that load. There are four key parameters to be determined for the utilization of the SDOF approximation: Load F(t) Mass M Stiffness K Damping C

These parameters are detailed below. To determine the fundamental equations, an idealized triangular pulse blast wave is assumed according to the discussions of Section 4.7.3, ensuring the peak pressure and total impulse of the loading history is preserved. F (t ) = F (1 t ) td
(6.1)

Figure 6.2 Idealized triangular Blast Load The impulse delivered by the wave is simply the area under the triangular pulse. i= 1 Ft d 2
(6.2)

Asher Gehl 9913569

110

Structural Response The structure is to be idealized as a typical elastic single degree of freedom model according to Figure 6.3 below.

Figure 6.3: Single Degree of Freedom model (Conrath et al, 1999) The simplest relationship between the above load and the SDOF models response is to consider the static equilibrium equation, described by Hookes law. F=Kx
(6.3)

Where F is the applied load, K is the structural stiffness and x is the deflection. Of course, the load is not a simple static load, but is within the dynamic domain. From the impulse delivered by the blast, the structures dynamic equation of motion is described by dAlemberts principle.

& F (t ) = M&& + Cx + Kx x The major differences between the static and dynamic response is due to inertial
& effects M&& and damping effects Cx . x

(6.4)

Inertia effects are very significant and dominate the response, when the loading duration (td) is considerably smaller than the structures response time (tm).
B B B B

Asher Gehl 9913569

111

Structural Response C is the damping coefficient and is quite insignificant for blast loading situations. The majority of design manuals conservatively ignore the effects of damping for the dynamic system. The primary interest in the analysis of structures under blast loads is the determination of the peak response. This of course occurs before the response cycle has begun and therefore before the damping forces may have any effect (TM51300, 1991; Goschy 1990). Thus ignoring the damping coefficient: M&& + Kx = F (1 x t ) td
(6.5)

Limiting the time to t<td


B B

x(t ) =

t F F (1 cos t ) + (sin t ) K Kt d
K M

(6.6)

As previously mentioned, the aim of the analysis is simply to determine the maximum dynamic displacement Xm. As with all oscillatory motions this occurs when the
B B

velocity is zero. dx(t ) =0 dt The time taken to reach this maximum position at Xm is tm.
B B B B

(6.7)

Solutions of the differential equation are of the form

t m = f (t d )
A similar equation can be determined for Xmax
B B

(6.8)

xmax
F K

t = (t d ) = ' ( d ) T

(6.9)

Asher Gehl 9913569

112

Structural Response Where and are functions of td and td/T respectively and T is the natural period
B B B B

of response. Thus according to the above equation, the maximum deflection (Xm) of a
B B

SDOF element is a function of the dynamic characteristics of the force of the blast wave represented by the decay time to ambient pressure (td), and the frequency
B B

characteristics of the response, represented by the natural period of the structure (T). TM5-1300 (1991) presents a number of response curves for different load rise times that demonstrate the interaction between these three key parameters and the structures ultimate resistance. As previously mentioned, the SDOF response chart for a triangular load is included in Appendix C. Thus in order to analyse a SDOF structure these parameters must be determined according to the subsequent discussions as shown in TM5-1300 (1991).
6.3.3 Dynamic Design Factors 6.3.3.1 Introduction

It has been determined that the parameters that appropriately define an equivalent SDOF system are: The equivalent mass ME,
B B

The equivalent spring constant KE


B B

The equivalent load FE.


B B

These factors define the conversion from the real system to the equivalent system and are derived by equating the kinetic energy, strain and work done by real system to the equivalent SDOF mass spring system. The criteria presented in TM5-1300 (1991) for the equivalent system other that equating energies is for the deflection of the concentrated mass to be the same as the majority of the structure. The deflected shape under the dynamic loads is obtained via assuming a shape obtained by the deflection under an equivalent static load. The transformation factors presented in TM5-1300 (1991) to determine the equivalent system are discussed below.

Asher Gehl 9913569

113

Structural Response
6.3.3.2 Load Factor

The load factor is the design or transformation factor by which the total load applied on the structural element is multiplied, in order to obtain the equivalent concentrated load on the equivalent single-degree-of-freedom system. If the actual total load on the structure is F and the equivalent load is FE, the load factor KL is defined by the
B B B B

equation: KL = FE /F
B B B B

(6.10)

The load factor is derived through equating the work done on each system by their respective loads in deflecting to the elements assumed deflected shape. For a structure with distributed loads:
WD = F max = p ( x) ( x)dx
0

(6.11)

Where max = maximum deflection of actual structure


B B

p(x) = distributed load per unit length (x) = deflection at any point of actual structure L= length of structure

Rearranging the terms results in


FE = p ( x) ( x)dx
0

(6.12)

Where (x) is the shape function.

( x) = ( x)

(6.13)
max

Asher Gehl 9913569

114

Structural Response The shape function (x) and therefore the load factor KL are different in the plastic
B B

and the elastic regions. Values for KL for one-way elements are presented in Appendix
B B

A1. The equations for a concentrated load system can be found in the reference TM51300 (1991).
6.3.3.3 Mass Factor

The mass factor is the design or transformation factor by which the total distributed mass of an element is multiplied, to obtain the equivalent lumped mass of the equivalent single-degree-of-freedom system. If the total mass of the actual element is M and the mass of the equivalent system is ME, the mass factor KM is defined by the
B B B B

equation KM = M E
(6.14)

KM can be obtained by equating the kinetic energies of each system based on the
B B

assumed deflected shape. For a structure with continuous mass:


K E = 1 M E ( max ) 2 = 1 m( x)[ ( x)] dx 2 2
0 L 2

(6.15)

Where: = natural circular frequency

m(x) = distributed mass per unit length The above equation can be rearranged for ME
B

M E = m( x) 2 ( x)dx
0

(6.16)

Asher Gehl 9913569

115

Structural Response Note that the shape function (x) is based on the deflected shape of the element due to the applied loading and not to the distribution of the mass. Since the deflected shape of the element is different for the elastic and plastic ranges, (x), and therefore KM, will also be different.
B B

For a concentrated mass system, the following equations derive the equivalent mass: K E = 1 M E ( max ) 2 = 1 M r ( r ) 2 2 2 r =1 Where Mr =
B B

(6.17)

rth mass deflection of mass r number of lumped masses

r =
B B

n=

Rearranging the equation to determine ME;


B B

M E = M r r
r =1

(6.18)

Using the above equations, values for KM have been calculated for one-way elements
B B

with constant mass. These tables taken from TM5-1300 (1991) are presented in Appendix A1.
6.3.3.4 Resistance Function

The resistance factor is the design factor by which the resistance of the actual structural element must be multiplied to obtain the resistance of the equivalent singledegree-of-freedom system. To obtain the resistance factor, the strain energy of each system is equated from the assumed deflection shape, and the strain energy of the equivalent single-degree-of-freedom system. If the computed total resistance of the structural element is r and the equivalent total resistance of the equivalent system is rE, then the resistance factor is defined by the equation.
B B

Kr = rE /r
B B B B

(6.19)

Asher Gehl 9913569

116

Structural Response As subsequently discussed in section 6.6, since the resistance of an element is the internal force tending to restore the element to its unloaded static position and is equal in magnitude to the load, according to Newtons laws it stands that the resistance factor Kr must always equal the load factor KL.
B B B B

6.3.3.5 Load-Mass Factor

Combining the two basic transformation factors, KL and KM, forms the load-mass
B B B B

factor (KLM). It is simply the ratio of the mass factor to the load factor, and it is
B B

convenient for obvious reason of simplicity. KLM = KM /KL


B B B B B B

(6.20)

Values of load, mass and load-mass factors from TM5-1300 (1991) are presented in Appendix A1.
6.3.3.6 Natural Period of Vibration

To utilise the available methods presented in TM5-1300 (1991) to determine the maximum response of a system, the effective natural period of vibration is required. This effective natural period of vibration, when related to the duration of a blast loading of given intensity and a given structural resistance, determines the maximum transient deflection Xm of the structural element.
B B

The effective natural period of vibration is: Tn = 2 ( Where me =


B B

me

KE

= 2 ( K LM m

KE

(6.21)

the effective unit mass the equivalent unit stiffness of system

KE =
B B

Asher Gehl 9913569

117

Structural Response The values used for the effective mass and stiffness for a particular element depends on the allowable maximum deflection according to the protection category. When designing for completely elastic behaviour, the elastic stiffness and the elastic value of the effective mass is used. In all other cases the equivalent elasto-plastic stiffness KE, is used. This equivalent value is explained further in Section 6.6.3. The effective mass is the average of the plastic and elastic equivalent value for small plastic deformations, whilst the plastic effective mass is required for larger plastic deformations.

6.4 Quasi-Static vs. Impulsive loading


The method of analysis for the SDOF system is dependant upon the relationship between the duration of the positive phase loading (td) and the response time of the structure (tm) determined by its fundamental period of vibration T. Elements that have a shorter response time with respect to the load duration typically respond to pressure and are termed Quasi-static. Elements with longer response times respond to impulse and are termed Impulsive. Quasi-static elements that respond to pressure may be analysed using either numerical methods or SDOF response design charts, while elements that respond to impulse are analysed using either SDOF response design charts for large deflection or an impulse method for designs with limited deflections. The technical manual TM5-1300 (1991) presents the numerical techniques analysis methods as well as the design charts derived from the systematic analysis of SDOF systems for several loading scenarios and is recommended for further review. The derivations of the results and methods are not presented here but are utilized in the design procedure of Chapter 7.

Asher Gehl 9913569

118

Structural Response
6.4.1 Quasi-Static Loading

When the duration of the positive phase of the blast load (td) is greater than the
B B

response time of the equivalent SDOF system (tm), the structure will reach its
B B

maximum displacement Xm before the load has significantly decayed.


B B

Figure 6.4 Quasi-static loading This type of explosion is typical of gas explosions where there is a long decay time for the reduction of load. Structural steel is also generally more ductile and hence more likely to respond to the pressure-time regime. For the Quasi-static case, the displacement reached is a function of the peak blast load F and the stiffness of the structure K. The basic principle of the dynamic analysis is energy conservation. The work done by the transient load upon the structure is equated to the strain energy produced from the deformation. Initially the work done by the blast on the structure is determined. This work will deform the structure, and as deformation ensues, the structure will acquire strain energy with increasing deformation. The key of the following equations is concerned with balancing the conversion of the work done by the blast (Equation 6.22), to the strain energy acquired by the structure (Equation 6.23). WD = FX m U = 1 KX m 2
= X max =2 X st

(6.22) (6.23) (6.24)

X max
F K

Where Xst is the equivalent displacement should the load be applied statically.
B B

Asher Gehl 9913569

119

Structural Response Equation 6.24 represents the Dynamic load factor (DLF), which is a non-dimensional ratio of the dynamic to static deflection. Here the DLF=2 and states that the limiting peak dynamic displacement is twice the equivalent static displacement. This gives the limiting response region and is represented in Figure 6.7 as the QuasiStatic asymptote.
6.4.2 Impulsive Loading

When the duration of the positive phase of the blast load (td) is less than the response time of the equivalent SDOF system (tm), there is significant decay of the blast load before the structure has had time to respond to the blast.

Figure 6.5 Impulsive loading For the impulsive regime, the displacement achieved by the structure is dependant upon the Impulse i, the Stiffness K, and the mass M. The analysis procedure again utilises the principle of energy conservation.

Asher Gehl 9913569

120

Structural Response When an impulse is delivered to a structure initially at rest a number of fundamental principles occur: Velocity will change
& x0 = I M (6.25)

Momentum is acquired Kinetic energy obtained


KE =
1

&2 Mx0 = I

2M

(6.26)

The kinetic energy is converted to strain energy within the structure, as it deforms
X max
F K

X max = X st

t d

(6.27)

Equation 6.27 above is considered the limiting displacement ratio or the impulsive asymptote of the response, as in Figure 6.7 (Mays and Smith, 1995). Since the maximum response Xm as presented in the SDOF response charts is
B B

dependent upon the ratio of the loading duration to the natural period (td/T), it is of
B B

considerable interest to plot dynamic magnification factors vs. td/T as shown below in
B B

Figure 6.6.

Asher Gehl 9913569

121

Structural Response

Figure 6.6 Dynamic response for various impulsive loads (Goschy, 1990) Goschy (1990) states the structural response is a function of the rise time such that if: td >3 T The load is considered as a Quasi-static load. Thus it is evident that the dynamic design process is generally based upon an energy solution whereby the kinetic energy delivered by the blast is equated to the strain energy produced in deforming to a limited deflection. This is determined from the area under resistance function that is presented in Chapter 6.6.
(6.28)

Asher Gehl 9913569

122

Structural Response

6.5 Pressure Impulse Diagram


The importance of the above derivations is to determine the relationships between T
B

the natural period of the structure and td the positive phase duration of the blast wave.
B B B

The response of an equivalent SDOF element is now compared based on the ratio of td/T. This is needed in order to distinguish between elements that are sensitive to
B B

pressure and those that are sensitive to Impulse. Later the analysis and design approaches are selected based on these ratios. It has been determined that the response of a structural system subjected to a dynamic load is defined in terms of its maximum deflection Xm and the time tm to reach this maximum deflection. The dynamic load is
B B B B

defined by its peak value P and duration td, whilst the single-degree-of-freedom
B B

system response is defined in terms of its ultimate resistance ru, elastic deflection XE
B B B B

and natural period T. SDOF response charts relate the dynamic properties of the blast load (P and td) to those of the element (ru, Xmax, XE, T, tm), that is, Xm/XE and tm/T are
B B B B B B B B B B B B B B B B

plotted in Appendix C as a function of ru/P and T/td.


B B B B

The asymptotes derived above for Impulsive (Equation 6.27) and Quasi-static (Equation 6.24) loading conditions can be rearranged and thus converted as the axes of a Pressure-Impulse diagram allowing the simple assessment of damage level caused by a load-impulse combination. The Q-S asymptote can be rewritten as: 2F/KXmax=1
B B

(6.29)

The impulsive asymptote can be rewritten as:

X max

KM = 1
B

(6.30)

Asher Gehl 9913569

123

Structural Response

Figure 6.7: Normalized Pressure Impulse Diagram (Krauthammer, 2004) This normalized Pressure-Impulse (P-I) diagram is used to define the expected response of elements based on specified blast loads. Once the maximum displacement is defined, the P-I curve specifies the load-impulse combination that will cause failure. Combinations of P-I to the right of the curve will fail due to damage in excess of the allowable limit. With combinations of P-I diagrams and blast pressure vs. scaled distance graphs; it is possible to specify critical damage levels of structural elements for certain charge weights at a given distance, as shown below in Table 6.1.

Asher Gehl 9913569

124

Structural Response

Table 6.1 Overpressure and scaled distance (Z) for certain types of Damage (Mays and Smith, 1995) Similar tables are also available to determine the damage levels upon humans from a certain weight of explosion at a given distance (Mays and Smith, 1995)

6.6 Resistance Deflection Function


It has already been established that the time taken to reach maximum displacement is of primary concern for establishing the response of the structure subjected to blast loads. The maximum displacement achieved is dependant upon the design criteria and the inherent characteristics of the structure, which is evident in its resistancedeflection function. Under the applied blast loads a structural element is deformed and internal forces set up. The sum of these internal forces aiming to restore the element to its unloaded static position is defined as the resistance. The resistance of a structural element is a reactive force associated with the deflection of the element produced by the applied load. It is convenient to consider the resistance as an equivalent load in the same manner as the applied load, but opposite in direction. The resistance of an element is essentially dependant upon its geometric and material properties. The variation of the resistance vs. displacement is expressed by a resistance-deflection function and may be represented graphically (TM5-1300, 1991).

Asher Gehl 9913569

125

Structural Response An idealized resistance-deflection function for a Reinforced Concrete element covering the complete flexural range to failure is shown in Figure 6.8.

Figure 6.8: Typical Concrete Resistance Deflection function (Mays and Smith, 1995) Upon the continual application of load the element deflects and at each instant in time the element exerts a resistance to further deformation. The resistance is a function of the elements unit Stiffness K, until yield when the elements ultimate unit resistance ru is reached at its deflection XE. The initial portion of the figure is comprised of the linear elastic region. The transition from elastic to the plastic region for the case of a beam with simple supports occurs with the formation of a plastic hinge. Additional elasto-plastic regions occur at every additional plastic hinge for elements with different support conditions. In the technical manual TM5-1300 (1991) these are called two-three and four step system, and their resistance deflection functions are illustrated in Figure 6.9. Following the loss of flexural resistance as indicated upon the crushing of the concrete in Figure 6.8, adequate restraint permits the development of the tensile reinforcement within the element. With this development the element has sufficient resistance to permit deflection well into the strain-hardening region.

Asher Gehl 9913569

126

Structural Response
6.6.1 Ultimate Resistance

The ultimate resistance of an element depends upon the protection design category and the magnitude of the explosive event as well as the following key aspects: 1. The distribution of the applied loads. 2. The geometry of the element (length and width). 3. The number and type of supports. 4. The distribution of the moment capacity or reinforcement in the case of reinforced concrete elements. The technical manual TM5-1300 (1991) considers the transient load distribution from intermediate to low-pressure ranges to be uniform. At high-pressure ranges however, the blast loads are variable across the surface of the element unless laced and shear reinforcement is used to allow only relatively small deflections. In this case it is sufficient to approximate the loading as uniform across the element. For reinforced concrete one-way elements with an elastic distribution of steel reinforcement all critical sections yield simultaneously and the ultimate resistance is based upon the moment capacity at first yield. Where there is a non-elastic distribution and structural steel elements, the ultimate resistance is a function of the moment capacity at the first yield plus the added moment capacity due to subsequent yielding at other critical sections. Values of the ultimate resistance for one-way elements are included in Appendix A as presented in TM5-1300 (1991) and reproduced Mays and Smith (1995). The notation used in the table is: Mn = ultimate negative unit moment capacity at the support Mp = ultimate positive unit moment capacity at midspan. L= ru = Ru = length ultimate unit resistance total ultimate resistance

Asher Gehl 9913569

127

Structural Response
6.6.2 Ultimate Deflection

The ultimate deflection Xm is essentially a function of the angle of rotation about the
B B

elements supports. With the attainment of ultimate resistance ru, the element
B B

essentially becomes a mechanism which rotates with no further increase in either the moment or curvature between the plastic hinges. For one-way elements, the rotation continues and the deflection increases until either the maximum deflection Xm is reached or failure occurs at max.
B B B B

Equations for the maximum deflection Xm in the range 0 < X < Xu for several oneB B B B

way elements as a function of the rotation angle and the ultimate deflection Xu are
B B

given in Table 3.5 of TM5-1300 (1991) and reproduced in Appendix A2, where the values for Xu are based on the development of a maximum support rotation, max,
B B B B

prior to failure. (TM5-1300, 1991) The maximum support rotation will vary with the material type and geometry of the element. The criteria for partial and developing failure for concrete and structural steel elements can be found in Section 6.7 in their respective sections.
6.6.3. Stiffness

The slopes of the elastic and elasto-plastic ranges of the resistance function are defined by the stiffness K of the element: K = ru/X
B B

(6.31)

Where ru is the unit resistance


B B

X is the deflection corresponding to the value of r. As seen in Figure 6.9 the stiffness within the elastic range is denoted as Ke, the elastoB B

plastic range as Kep, while in the plastic range the stiffness is zero.
B B

Asher Gehl 9913569

128

Structural Response

Figure 6.9 Resistance Deflection Functions for I, II and III step systems (TM5-1300, 1991) The stiffness within the elastic range is seen in the one step elasto-plastic system above. Ke = re /Xe For a two-step system there is a single elasto-plastic region, whereby the corresponding elasto-plastic stiffness is: Kep = (ru re) /(Xp Xe)
(6.33) (6.32)

The elastic and elasto-plastic stiffness of one-way elements are given in Appendix A3 from TM5-1300 (1991), as a function of the modulus of elasticity E, moment of inertia I, and span length. Knowing the resistance and stiffness, the corresponding elastic and elasto-plastic deflections can be computed from the above equations.

Asher Gehl 9913569

129

Structural Response

6.7 Deformation Limits


As previously discussed, the allowable degree of damage sustained by the element must be controlled with respect to the elements required level of protection. This thesis is concerned with the design of civilian facilities, not hardened military installations, and thus the deformation criteria is consistent with designing for only one accident, and generally will involve significant plastic deformations. Limitations are imposed in order to appropriately restrict the ultimate damage to a structure or element, which is subjected to the effects of an explosion. The criterion that determines the allowable damage is the imposed limit on the elements deflection. The technical manual TM5-1300 (1991) categorizes the maximum deflection Xm as
B B

either limited or large. This corresponds to the protection categories of 1 and 2 respectively. Protection category 1

This level is required for the protection of personnel through the reduction of blast pressure, Shielding from primary and secondary fragments and falling parts of the structure. Protection Category 2

This level is required for the preventing the structural elements from collapse under the blast load. The maximum deflection that is achieved within the limited range is generally kept within the elastic or elasto-plastic regions of the resistance function. For maximum deflections within the large range the system generally responds within the plastic range. There is a negligible error associated with the omission of the elastic and elasto-plastic ranges (TM5-1300, 1991).

Asher Gehl 9913569

130

Structural Response Single-degree-of-freedom elements such as beams, floor and wall panels, open-web joists, and plates, are restricted via the maximum ductility ratio and the maximum rotation at an end support. In the design of these elements the ductility ratio must be checked to determine whether the specified rotation is attainable without premature buckling. For framed structures limitations are imposed on the sideways deflection and individual frame member rotations as seen below in Figure 6.10. TM5-1300 (1991) specifies a general requirement of the overall structures stability such that Maximum end rotation 2 o Maximum side deflection H 25

Figure 6.10: Member end Rotations for beams and frames (TM5-1300, 1991) For concrete elements, deformations are expressed in terms of rotations () about the support. Structural steel deformations are in terms of the ductility ratio (). Ductility refers to the elements ability to absorb energy without failure.

Asher Gehl 9913569

131

Structural Response

Xm XE

(6.34)

Xm refers to the final deflection obtained under the imposed load


B B

XE refers to the deflection reached at the elastic limit.


B B

Thus, a ductility ratio of 3 corresponds to a maximum dynamic response three times the equivalent elastic response. Sufficient bracing can enable ductility ratios as high as 20 and Chapter 5 of TM51300 (1991) provides appropriate bracing requirements. The relationship between the various parameters involved in the design of beams and the protection category is readily described in the idealized resistance-deflection curve shown in TM5-1300 (1991) and reproduced below in Figure 6.11.

Figure 6.11 Relationship between design parameters and Protection category (TM5-1300, 1991) Asher Gehl 9913569 132

Structural Response Both of the above protection categories allow for significant levels of plastic deformation and thus are completely unusable after the event or require significant repair. For reusable designs, the deformation must be limited to the elastic region such that 1. This limitation will result in massive and costly construction and is considered unnecessary requirement for majority of civilian structures.
6.7.1 Resistance Deflection Functions for Limited Deflections

The resistance function for the case of limited deflections takes the form as shown in Figure 6.9 for a one-step system, a two-step system, and a three-step system, respectively. As shown in these Figures the dotted lines represent an equivalent elastic resistancedeflection function for two and three step systems. The equivalent elastic stiffness KE and the equivalent maximum elastic deflection XE
B B B B

are calculated by equating the potential energies in each system. This is achieved by matching the area under the dotted curve to the area under the solid curve in the resistance functions of Figure 6.9. The equivalent maximum elastic deflection XE for the two-step and three-step systems
B B

is expressed in the following equations.


X E = X e + X p (1 re ru )

(6.35) (6.36)

X E = xe ( rep ru ) + X ep (1 re ru ) + X p (1 rep ru ) In each case the equivalent elastic stiffness KE is equal to


B B

KE =

ru XE

(6.37)

Asher Gehl 9913569

133

Structural Response

Figure 6.12 The Idealised Resistance Deflection function for limited deflections The resistance deflection function for one-way elements is dependant upon the support conditions. The equivalent elastic stiffness KE for various support and loading conditions is presented in TM5-1300 (1991) and reproduced in many design manuals. From this and the relationship defined by the above equations, the equivalent elastic deflection (XE) can be determined. Two-way elements generally exhibit two- and three-step resistance-deflection curves, which are a function of not only the type of supports but also of the aspect ratio L/H of the element. The equivalent elastic deflection XE of the element under consideration must be evaluated from the above equations for two- and three-step systems. Two-way elements are not considered further in thesis. Further analysis and the value of KE can be found in TM5-1300 (1991).
6.7.2 The Resistance Deflection Function for Large Deflections

With respect to large deflections TM5-1300 (1991) neglects the elastic and elastoplastic regions of the resistance function and assumes an instantaneous rise in the resistance to ru from zero. A constant plastic range resistance as below in Figure 6.13 approximates the design resistance function for one-way elements.

Asher Gehl 9913569

134

Structural Response

Figure 6.13: The Resistance Deflection function for large deflections It is assumed that there is no change of the resistance function of an element when dynamically or quasi-statically loaded (Mays and Smith, 1995). Variations only arise due to the improvement in ultimate resistance from dynamic loading, which is further discussed below in Section 6.8.

6.8 Material Properties


The materials design strength although based upon the materials characteristic strength will be slightly improved under the application of a blast load. The load from a blast upon an element is essentially a rapidly applied load. The rate at which the load is applied induces certain dynamic effects to the element. Under the application of rapidly applied loads, the strain rate increases. This rise has a decisive influence upon the mechanical properties of structural materials. The design procedure for blast loading effects takes into account the variation of the mechanical properties from static to dynamic loading. The increases due to the dynamic nature of the load are summarised as follows:
Structural steel and Reinforcement Bars

Yield stress fy improves to the dynamic yield stress fdy


B B B

Ultimate tensile strength fu improves to dynamic ultimate strength fdu


B B B B

Youngs modulus E remains unchanged


B

The elongation at failure remains unaffected

Asher Gehl 9913569

135

Structural Response
Concrete

Compressive strength fcu significantly improves to dynamic compressive


B B

strength fdu
B

Youngs modulus E remains unchanged


B

The amount of improvement from static to dynamic strengths is described by the dynamic increase factor (DIF). For blast loading design, the dynamic stresses to be used are based upon the protection category required for the structure and the ultimate deformation or damage limits imposed by that chosen category. The properties of both reinforced concrete and steel are discussed separately below.
6.8.1 Structural Behaviour of Reinforced Concrete 6.8.1.1 General Behaviour

The behaviour of concrete differs in dynamic loading compared to static loading. The initial stiffness, as well the ultimate strength, increases in both compression and tension. Furthermore, the concrete strain capacity is extended in dynamic loading. Under the application of a dynamic load, the element deflects until either the strain energy of the element is sufficient to balance the kinetic energy delivered by the blast load or fragmentation of the concrete occurs and the element fails. Hence the ultimate capacity of the section is dependant upon its resistance whilst obtaining large inelastic deformations to absorb the copious amounts of energy before failure ensues. The actual deformation obtained is dependant upon the dimensions of the element and the reinforcement details used for the elements design (TM5-1300, 1991). The concrete flexural behaviour is best illustrated as a typical resistance deflection curve as in Figure 6.14 below.

Asher Gehl 9913569

136

Structural Response

Figure 6.14 Resistance deflection curve for flexural response of Concrete (TM5-1300, 1991) Upon initial loading, the resistance increases linearly with deflection until the steel reinforcement begins to yield. As deflection increases the resistance is constant until all the reinforcement has yielded. At the deflection limit corresponding to support rotation of = 2 o the concrete crushes in compression. In limiting the support rotations to 0 o 2 o the concrete cover remains intact on both sides and the concrete is effective in resisting the applied moment from the load. This defines a Type 1 section (Figure 6.15).

Asher Gehl 9913569

137

Structural Response For rotations 2 o the compression face of the concrete has already been crushed. Without sufficient shear reinforcement the crushing of the compression concrete results in failure of the element. However with sufficient ties (Section 6.8.1.2) around the flexural reinforcement the compression forces are transferred from the concrete to the compression reinforcement in the section. This transfer results in a loss of capacity for the section as seen in the resistance drop at this point. The compression reinforcement is essential to ensure the concrete crushing, does not result in the members complete failure at this point. Symmetrical reinforcement is required to sufficiently develop the tensile steel of the section. To effectively tie the flexural reinforcement blast links are required as either single leg stirrups or, lacing. This is defined as a type 2 section (Figure 6.15).

Figure 6.15: Typical Reinforced concrete cross sections (TM5-1300, 1991)

Asher Gehl 9913569

138

Structural Response From the notation in the above figure and using conventional plastic theory: The ultimate moment capacity of type 2 sections of width b is

= As

f ds dc b

(6.38)

The ultimate moment capacity of type 1 sections is M p = As f ds (d 0.45 x) b

A f x = s ds 0.6bf dc Where: Mp =
B B

(6.39)

ultimate moment capacity total area of reinforcement within the beam dynamic design stress of reinforcement (Section 6.8.1.4) distance from extreme compression fibre to centroid of tension depth of neutral axis below the extreme compression fibre width of beam dynamic ultimate compressive strength of concrete

As =
B B

fds =
B B

d= x= b= f'dc =
B B

reinforcement

As deflection increases, the steel reinforcement enters the strain-hardening region of the curve and the resistance increases with rising deflection. Without any tensile action the simple blast links (single leg stirrups (Section 6.8.1.2)) are initially sufficient to prevent the compression reinforcement from buckling. At rotation values of = 4 o the section will fail unless further lacing reinforcement is provided. This is a complex procedure and generally limited to explosive storage facilities. With sufficient lacing reinforcement however the element will be restrained until the tensile failure of the reinforcement at =12o.
P P

Asher Gehl 9913569

139

Structural Response For the purposes of design, the design Manual TM5-1300 (1991) approximates the response of a reinforced concrete element by the representation of the following idealized resistance-deflection function (Figure 6.16). This model underestimates the compression membrane capacity.

Figure 6.16 Idealized resistance deflection function (TM5-1300, 1991) In this Figure the structure deflects until its elastic limit with elastic properties within this zone. XE the deflection at the elastic limit ru the unit dynamic resistance modified for static loads present using dynamic theory KE the elastic stiffness Xm the maximum deflection permitted with respect to support rotation and protection category. The flexural design of concrete elements follows a different process according to the limitations imposed on the support rotations.

Asher Gehl 9913569

140

Structural Response For Protection Category 1 the element is expected to respond in the Quasi-static regime as defined in section 6.4.1. Therefore the load is assumed to be uniformly distributed and represented by the peak pressure Pmax. For this idealisation TM5-1300
B B

(1991) presents a response chart based on a SDOF analysis. This chart is presented in Appendix C, and is utilised to finalise the flexural design and ensure the limitations on ductility and support rotations have been met. For protection Category 2, the element is expected to respond in the impulsive regime as defined in Section 6.4.2. The blast load is again assumed to be uniformly distributed and is represented by the specific impulse, i. For this category the distance between the centroids of the compression fibre to the tension fibre of the flexural reinforcement (dc), is determined through the solution of the basic impulse equation:
B B

r X i2 = u E + ru ( X m X E ) 2 K LM m 2 Where: KLM is the load-mass factor (Section 6.3.3.5) obtained via Appendix A1
B B

(6.40)

m is the unit mass of the element. The flexural design is finalised by determining the ratio of tm/td to check whether the
B B B B

appropriate protection category has been used. After the flexural design is successful the shear reinforcement required must be determined. TM5-1300 (1991) presents the equations required for a thorough shear design, however they are not considered in the subsequent designs of Chapter 7. The different types of shear reinforcement however are considered below.

Asher Gehl 9913569

141

Structural Response
6.8.1.2 Types of Blast Links

The element will only achieve such large deformations above, providing there is not a premature failure caused by inadequate shear reinforcement in the section.
6.8.1.2.1 Lacing Reinforcement

Lacing reinforcement is a key method to allow concrete elements to obtain the large deformations (up to 12o) required to withstand the high intensity non-uniform blast loads from an explosive charge at close range. The use of lacing reinforcement tends to spread out the effects of the non-uniformity of a large explosion at a close standoff distance and permits the use of an average blast load over the entire surface area of the element. For large deflections, the lacing not only resists the high shears produced but also maintains the integrity of the severely cracked concrete between the tension and compression reinforcement during the latter stages of deflection. Lacing reinforcement is shown below in Figure 6.17.

Figure 6.17 Lacing reinforcement (TM5-1300, 1991) Essential to a laced element is symmetric reinforcement whereby the compression and tension reinforcement are equal. Bent diagonal bars tie together the flexural reinforcement within the concrete. The lacing reinforcement has the following effect on the concrete element according to TM5-1300 (1991).

Asher Gehl 9913569

142

Structural Response Fully developed ductility of flexural reinforcement well into the strain hardening region The extreme loads upon the element will produce substantial cracking. The concrete between the flexural reinforcement however will be effective. The compression reinforcement will not buckle Resistance to high shear stresses at the supports Local shear failure from the intensive pressures is prevented Reduction of the post failure fragments produced in a brittle failure.

6.8.1.2.2Single Leg Stirrups

Single leg stirrups (Figure 6.18) exhibit the same effects on a concrete element as outlined above, however they are less efficient at closer ranges than an element with lacing reinforcement. The maximum deflection obtainable from an element with single leg stirrups is limited to 4o under flexural action.
P P

Figure 6.18 Single Leg Stirrups (TM5-1300, 1991)

Asher Gehl 9913569

143

Structural Response
6.8.1.3 Modes of Structural Behaviour

A reinforced concrete element has 2 distinctive modes of structural response, namely ductile or brittle. Elements that exhibit a ductile response can maintain its peak strength to achieve large inelastic deformations before the onset of any failure mechanisms. Whereas for elements that demonstrate brittle characteristics, complete failure will occur suddenly with little inelastic elongation. Reinforced concrete with well-tied and anchored ductile reinforcement can be classified as a ductile material. Ductility influences the dynamic response and behaviour of elements subjected to blast loads. Its significance is inversely proportional to the duration of load tm and the
B B

required ultimate resistance such that ru is less for brittle elements when compared to
B B

ductile elements. For the design of concrete elements the behaviour mode is determined by the following parameters: Protection Category and Function of the element Characteristics of Explosive threat Occurrence of Primary Fragments

6.8.1.3.1 Ductile Mode of Behaviour

The ductile mode of behaviour for an element is dependant upon the location of the explosive event, namely whether it is close in or far away from the target structure. The applied load on an element from an event far away is of uniform distribution. To absorb the loading from this event the deflections required are small. Standard nonlaced sections are completely capable of resisting these loads. Essentially the response of these elements is demonstrated via Figure 6.15. Elements without shear protection will fail at 2o rotation as characteristic of type 1 sections
P P

whilst adequate shear protection and a type 2 section, will enable a further rotation to 4o.
P P

Asher Gehl 9913569

144

Structural Response The applied load from an explosive event close in is non-uniform and has a considerably higher magnitude. As mentioned, to withstand the pressures from this event elements require lacing reinforcement. The behaviour of such an element is exhibited from the entire length of Figure 6.14.
6.8.1.3.2 Brittle Failure Modes

The brittle behaviour of reinforced concrete is composed of three types of concrete failure: Direct spalling Scabbing Post-failure fragments.

Spalling occurs due to the extreme levels of blast pressure affecting the element. As the high pressure hits the element there is dynamic disengagement of the concrete cover over the flexural reinforcement. Spalling is illustrated in the sections of Figure 6.15. Scabbing is an effect that is a result of the elements large deformations. Scabbing also consists of the disengagement of the concrete cover over the flexural reinforcement. Both types of spalls affect the capacity of the element to resist the applied blast load. Spalling and scabbing are usually only of concern in those protective structures where personnel, equipment, or sensitive explosives require protection, typically this corresponds to protection category 1. Post-failure fragments are the result of the collapse of an element and are usually the more serious. The fragments are a result of the ultimate flexural failure of an element. Lacing reinforcement in the element has a significant effect on the size of the post failure fragments. TM5-1300 (1991) states that the failure occurs at the yield lines and the section between the yield lines remain intact. Laced elements fail in fewer larger sections whilst unlaced elements fail as rubble as shown in Figure 6.19 below.

Asher Gehl 9913569

145

Structural Response

Figure 6.19 Post Failure Fragments in Laced and Unlaced Sections (TM5-1300, 1991)
6.8.1.4 Strain Rate Effects on Concrete

The behaviour of concrete depends on the loading rate; this is called the strain rate effect. The strain rate in the material depends on the loading case, as shown in Bischoff and Perry (1991) for different loading cases such as creep, static, earthquake, and hard impact and blast effects.

Figure 6.20: Strain Rates On Different Loading Cases (Bischoff and Perry, 1991) Elements under the impact of blast loading exhibit a higher strength than under a statically applied load. This strength increase is due to the rapid strain rates that occur. A higher strain rate results in a higher compressive strength of concrete and thus a higher yield and ultimate strength of the reinforcement.

Asher Gehl 9913569

146

Structural Response To determine the elements resistance under the dynamic load, the increased material properties called dynamic stresses and strengths must be used in the elements final design.
6.8.1.4.1 Stress-Strain Curve

The increases are illustrated in Figure 6.21 below in the typical stress strain curves for both concrete and the steel reinforcement. In the figures below the dashed curves represent loading at higher strain rates and the significant increase in strength is evident.

Figure 6.21 Stress Strain curves for standard and rapid strain rates of concrete (a) and steel (b) (TM5-1300, 1991)

Asher Gehl 9913569

147

Structural Response Definitions of the symbols used are as follows: f'c =


B B

static ultimate compressive strength of concrete dynamic ultimate compressive strength of concrete static yield stress of reinforcing steel dynamic yield stress of reinforcing steel static ultimate stress of reinforcing steel dynamic ultimate stress of reinforcing steel modulus of elasticity for reinforcing steel modulus of elasticity of concrete rupture strain

f'dc =
B B

fy =
B B

fdy =
B B

fu =
B B

fdu =
B B

Es =
B B

Ec =
B B

u =
B B

The most important property is the increased yield and ultimate strengths of both materials.
6.8.1.4.2 Allowable Material Strengths

The response of an element is determined by its ultimate strength and ductility. Both the strengths of steel reinforcement and concrete will be examined separately.
Steel Reinforcement

The technical manual TM5-1300 (1991) outlines the steel reinforcement that is designated by the American Society for Testing and Materials (ASTM) to have adequate ductility appropriate for the large deflections associate with resisting blast loads. For all the reinforcement bars it is crucial to note that the ductility is reduced at bends, lapped splices, mechanical splices, etc., and location of these anchorages near points of maximum stress is undesirable and should be avoided. Steel similar to these standards must have a minimum yield of 75,000 psi or 520MPa. The ASTM Standards for the minimum values of the yield stress fy and the ultimate
B B

tensile stress fu are 410MPa and 620MPa respectively. Mill test reports for this steel
B B

indicate 10% increase from the ASTM minimum yield stress and marginal increase in the minimum ultimate stress.

Asher Gehl 9913569

148

Structural Response TM5-1300 (1991) recommends the following design values for the specified ASTM A615 Grade 60 reinforcement: fy = 450MPa fu = 680MPa
Concrete

The ultimate strength of the concrete should be selected to suit the design requirements of the element. In the design calculations it is only required for purely elastic deflections. However it is an important aspect of the materials behaviour. Higher strength concrete can eliminate or reduce the shear reinforcement required. With larger deformations higher strength concrete can reduce the cracking and crushing that is associated with large support rotations at high inelastic deflections. TM5-1300 (1991) specifies a minimum strength for blast resistant structures regardless of the magnitude of the blast load and deflection criteria. f'c = 28MPa.
6.8.1.4.3 Dynamic Design Stresses for Reinforced Concrete Dynamic Increase Factor

The dynamic increase factor, DlF, is equal to the ratio of the dynamic stress to the static stress. fdy/fy fdu/fu f'dc/f'c.

The DIF depends upon the rate of strain of the element, increasing as the strain rate increases. TM5-1300 (1991) presents the conservative DIF design curves for the unconfined compressive strength of concrete and for the yield stress of ASTM A 615, Grade 60, reinforcing steel. (Figure 6.22) The curves were derived from test data having a maximum strain rate of 10 x 10-3 in./in./msec. for concrete and 2.1 x 10-3 in./in./msec. for steel.

Asher Gehl 9913569

149

Structural Response

Figure 6.22 DIF vs. Strain Rate for concrete and Reinforcement Steel (TM5-1300, 1991) It has been widely accepted for many years that at these at these high strain rates, the apparent strength of concrete can increase significantly.

Asher Gehl 9913569

150

Structural Response Many researchers, such as, Ross et al. (1995); Isenberg et al. (1992), Bishoff and Perry (1991); Soroushian et al. (1986) have studied the effect of the strain-rate on the ultimate strength of concrete. Both the ultimate compressive and tensile strengths have been found to increase with increased strain rate. The ultimate strain at compression and tension has also been studied and is found to increase with increased strain rate (Bishoff and Perry, 1991; Soroushian et al, 1986; Zielinski, 1984). Bischoff and Perry (1991) performed analysis upon the compressive strength of concrete under the effect of strain rate. As seen in Figure 6.23, for concrete, the DIF can be more than 2 in compression.

Figure 6.23 Relative increases in the ultimate uniaxial compressive strength as a function of the strain rate-(Bischoff and Perry.1991) According to Ross, Tedesco and Kuennen (1995), the concrete ultimate uniaxial strength in tension increases by multiples of 5 to 7 at very high strain rates.

Asher Gehl 9913569

151

Structural Response

Figure 6.24: Strain rate dependency for concrete in tension (Ross et al. 1996) Whilst with respect to the static equivalent, the dynamic stiffness does not vary; the stresses sustained under dynamic conditions are substantially higher than the static compressive strength. Malvar and Crawford (1998) presented their findings in determining the then existing data to characterize the effects of strain rate on the compressive and tensile strengths of concrete. The findings (Figure 6.25) propose a modified curve based on experimental results to explain and correct the variations of the recommended European code CEB and is recommended for further analysis and comparison between results.

Asher Gehl 9913569

152

Structural Response

Figure 6.25 Proposed modified CEB curves in tension (Malvar and Crawford, 1998) Knowledge of the DIF is of significant importance in the design and analysis of structures for explosives safety. DIF curves for concrete have been published in manuals by the Tri-Services, the Defence Special Weapons Agency, the Air Force, and the Department of Energy. However, these curves are typically based on limited data.
Values of DIF have been established for design of members in the far design range as

well as for members in the close-in design range. These design values of DIF are given below in Table 6-2 from TM5-1300 (1991).

Asher Gehl 9913569

153

Structural Response

Table 6.2 Design Increase Factor (DIF) for Design of Reinforced Concrete Elements (TM5-1300, 1991) Due to the increased magnitude of the blast loads and subsequent increase in the strain rate, the dynamic increase factors for elements subjected to a close-in detonation are higher than those for elements subjected to an explosion located far from the element. The increase in capacity is primarily dependant upon the time to reach yield, tE, of the steel reinforcement. Section 4.13.2 of TM5-1300 (1991) presents the derivations of more accurate values for the DIF based on tE. For the subsequent designs in Chapter 7 the values of the table were utilized, however due to the significant effect of the DIF on the elements design, for a more accurate design the process is recommended.
6.8.1.4.4 Dynamic Design Stresses

As mentioned previously the dynamic design stress incorporates the dynamic effects of the load and is used for the design of elements under a dynamic load. The average dynamic stress is expressed as a function of the dynamic yield stress fdy and the dynamic ultimate stress fdu. Criteria for the dynamic stresses to be used in the plastic design of ductile reinforced concrete elements are presented in Appendix B2 from TM5-1300 (1991). The dynamic design stress is expressed in terms of fdy, fdu, and f'dc. The value of these terms is determined by multiplying the appropriate static design stress by the appropriate value of the DIF so that: f(dynamic) = DIF x f(static)
(6.41)

Asher Gehl 9913569

154

Structural Response
6.8.1.4.5 Moment of Inertia

The moment of inertia for a concrete element is complicated by resultant cracking with increasing deflections. As cracks in the element develop, the effective section changes and thus the moment of inertia changes. The moment of inertia of a cracked section (Ic) considers the concrete area in
B B

compression and the area of steel as an equivalent transformed concrete area, and is calculated about the centroid of the transformed section. (TM5-1300, 1991) Ic is calculated by the following equation, given in TM5-1300 (1991).
B B

I c = Fbd 3

(6.42)
P

Where b=
F=

width of beam coefficient given in Figure 6.27 and 6.28 distance from extreme compression fibre to the centroid of tension

d=

reinforcement The coefficient F varies as the modular ratio n and the amount of reinforcement in the section.
n = E s Ec

(6.43)

Where Es=200000Mpa
Ec = 1.5 0.043 f 'c
(6.44)

=2400kg/m3
P

fc=28Mpa (minimum)
P B B

For sections with tension reinforcement only, the coefficient F is given in Figure 6.26 while for sections with equal reinforcement on opposite faces, the coefficient F is given in Figure 6.27.

Asher Gehl 9913569

155

Structural Response

Figure 6.26 Coefficient for Determination Of the Moment of Inertia for Concrete Elements with tensile reinforcement only (TM5-1300, 1991))

Figure 6.27 Coefficient for Determination of the Moment of Inertia for Concrete Elements with symmetric reinforcement (TM5-1300, 1991))

Asher Gehl 9913569

156

Structural Response
6.8.1.4.6 Changes in the Failure Mode due to Dynamic Loading

Studies of reinforced concrete structures subjected to static and impact loading have been conducted in order to examine the failure behaviour at different loading rates (Kulkarni and Shah, 1998; Rossi, 1994; Mutsuyoshi and Machida, 1984). The general phenomenon reported is that reinforced concrete structures have a tendency to change their failure mechanism, from bending to shear or punching failure, as the loading rate is increased. Wang (1996) indicated an explanation of this phenomenon for the case of a simply supported beam of prestressed reinforced concrete loaded by a point load in the mid span:
A significant portion of the applied load is balanced by the distributed inertial load along the beam, whose resultant force has a much shorter moment arm to the mid span of the beam than the support load has, leading to a reduced bending moment at mid span. Therefore, the applied load can go too much higher levels than the static flexural cracking load without any flexural cracks in the beam. When this load reaches the shear cracking load, which is the same under impact as under static loading, a shear crack may appear first.

The parameters influencing this phenomenon are thus the density of the concrete, the geometry of the structure and the configuration of the supports. Johansson (1999) conducted finite element analyses of concrete members subjected to impulse loading, and found that a possible reason for an increased load capacity of concrete members at high-rate loading could be time-dependent boundary conditions. Hence a concrete member that is regarded as simply supported at static load conditions could be regarded as fixed in the initial stage at high-rate loading before the stress wave within the concrete member has reached the support. Fu et al, (1992) found that in many tests involving concrete members, the flexural mode of failure was observed at relatively slow loading rates, while at high loading rates the shear failure or mixed mode failure became the main dominant failure mode.

Asher Gehl 9913569

157

Structural Response Therefore for elements designed for blast loading, it is imperative that adequate shear strength is developed to enable elements to deform to the prescribed levels according to the chosen protection category.
6.8.2 Structural Steel Behaviour 6.8.2.1 General Characteristics

The response of steel can is generally illustrated via the stress strain relationship as taken from TM5-1300 (1991), also illustrating the effect of strain rate. This is presented previously as Figure 6.21. As illustrated by this figure there is a considerably more ductile response from steel elements, illustrating its ability to absorb energy from the blast before failure. The response is linear until the yield point, beyond which there is the yield plateau region in which there is a considerable strain increase for a minimal increase in stress. This plateau exists until a ductility ratio of 10-15. Beyond this point, the section enters the strain-hardening region. The stress grows until such time as a maximum stress level is achieved- the tensile strength. Beyond this region there is a drop in strength with further elongation until fracture at strain levels of 20-30% further than the original length. Generally blast resistant design utilizes the elements response well into the plastic regions to minimise economic concerns with the design.
6.8.2.2 Development of Plastic Design Moment

Figure 6.28 below shows the stress distribution at various stages of deformation for a plastic hinge section.

Figure 6.28 Theoretical stress distributions for pure bending at various stage of dynamic loading (TM5-1300, 1991)

Asher Gehl 9913569

158

Structural Response As the beam bends elastically, the outer fibres of the section reach fds and the yield moment My is attained (6.28(a)). With increasing moment to M1, the yield stress progresses inward from the outer fibres towards the neutral axis as shown in Figure 6.28(b). Finally the rectangular stress distribution is developed as in Figure 6.28(c). The ratio between the fully plastic moment to the yield moment is the shape factor, f, for the section, and is essentially the ratio between the plastic and elastic section moduli. The plastic section modulus Z can be calculated as the sum of the static moments of the fully yielded elements of the equal cross section areas above and below the neutral axis as in equation 6.45.
Z = Acm1 + Atm2
(6.45)

Where Ac = At = m1 = m2 = area of cross section in compression area of cross section in tension distance from neutral axis to the centroid of the area in compression distance from neutral axis to the centroid of the area in tension

Note: Acm1 = Atm2for a doubly-symmetric section TM5-1300 (1991) assumes that a fully plastic section offers no additional resistance to load. However, additional resistance due to strain hardening of the material is present as the deformation continues beyond the yield level of the beam as indicated in Figure 6.28. The extent of the plastic hinge has no substantial influence on the ultimate capacity; it has, however, an influence on the final magnitude of the deflection. For all practical purposes, the assumption of a concentrated plastic hinge is adequate. In blast design, although strains well into the strain-hardening range may be tolerated, the corresponding additional resistance is generally not considered in the analysis, since excessive support rotation and/or ductility ratios of beams are susceptible to local flange or lateral torsional buckling.

Asher Gehl 9913569

159

Structural Response According to TM5-1300 (1991), for standard I-shaped sections, the plastic section modulus is approximately 1.15 times the elastic modulus for strong axis bending and may be obtained from standard manuals on structural steel design. A moment-curvature diagram for a simply supported steel beam, as shown Figure 6.29 below, can also represent the above developments.

Figure 6.29 Moment curvature diagram for simply supported dynamically loaded I beam (TM5-1300, 1991)

The behaviour is elastic until the yield moment My is reached, and a corresponding stress distribution of Figure 6.28(a). Upon further load application, the curvature increases at a greater rate as the fully plastic moment value, M2, is approached, after which, there is considerable curvature increase for only a small increase in moment capacity. The bilinear representation of the moment-curvature relationship is utilised for design purposes as represented in Figure 6.29 above by the dashed lines.

Asher Gehl 9913569

160

Structural Response As indicated in Figure 6.29 above: For elements where: 3 ; The design plastic moment is
Mp = f ds ( Z + S ) 2
(6.46)

Mp=M1
B B B B

For elements where: 3 ; The design plastic moment is

Mp=M2
B B B B

M p = f ds S

(6.47)

Where fds is the dynamic design stress (as described in Section 6.8.2.4)
B B

Z is the Plastic section modulus S is the Elastic section modulus It is important to note that the above equations are for beams, which are supported against buckling. Design provisions for guarding against local and overall buckling of beams during plastic deformation are discussed in Sections 5-24, 5-25, and 5-26 of TM5-1300 (1991).
6.8.2.3 Strain Rate Effects

The effects of rapid loading on the mechanical behaviour of structural steel have been observed and measured in uniaxial tensile stress tests for many years. Under rapidly applied loads, the rate of strain increases and this has a large influence on the mechanical properties of structural steel. As previously introduced the strain rate effects for steel members are: (1) The yield point increases substantially to the dynamic yield stress value. This effect is termed the dynamic increase factor for yield stress.

Asher Gehl 9913569

161

Structural Response (2) The modulus of elasticity (Es) in general will remain insensitive to the rate of loading. (3) The ultimate tensile strength increases slightly. However, the percentage increase is less than that for the yield stress. This effect is termed the dynamic increase factor for ultimate stress. (4) The elongation at rupture either remains unchanged or is slightly reduced due to increased strain rate. TM5-1300 (1991) presents the design increase factors for ASTM A36 and A514 steels. Even though ASTM A514 is not recommended for plastic design, the curve in Figure 6.30 may be used for dynamic elastic design.

Figure 6.30 DIF for Strain rate of different steels (TM5-1300, 1991)

Dynamic increase factors for yield stresses in various pressure levels in the bending; tension, and compression modes are listed in Appendix Table B3.

Asher Gehl 9913569

162

Structural Response
6.8.2.4 Dynamic Design Stresses

Structural resistance is determined on the basis of plastic design concepts, taking into account dynamic yield strength values. The design procedures are based upon the required levels of protection to the anticipated blast loading and the limitations imposed by those protection levels. To determine the plastic strength of a section under dynamic loading, the appropriate dynamic yield stress, fdy, must be used to account for the appropriate strain rate
B B

enhancements.
6.8.2.4.1 Dynamic Design Stress for Protection Category 1

For a ductility ratio 10, The dynamic design stress, fds, is equal to the dynamic yield stress, fdy.
B B B B

fds = fdy
B B B B

(6.48)

The dynamic design stress, fds, for bending, tension, and compression is equal to the
B B

product of the dynamic increase factor, c, the average yield strength increase factor, a (a= 1.1 for steels with a specified minimum yield stress of 345MPa or less; otherwise a= 1.0 ), and the specified minimum yield stress of the steel.
fds = fdy = c.a. fy
B B B B B B

(6.49)

Asher Gehl 9913569

163

Structural Response
6.8.2.4.2 Dynamic Design Stress for Protection Category 2

Where excessive deflections or ductility ratios may be tolerated, the dynamic design stress can be increased to account for deformations in the strain-hardening region. In this case, for > 10, the dynamic design stress, fds, becomes
fds = fdy + (fdu fdy)/4
(6.50)

Where
fdy = fdu =

dynamic yield stress dynamic ultimate stress equal to the product of fu from Appendix B2 Dynamic Increase factor for steel (Appendix B3 or Figure 6.30)

It should be noted that the average strength increase factor, a, does not apply to fdu.
6.8.2.4.3 Dynamic Design Stress for Shear

The dynamic design stress for shear shall be: fdv = 0.55 fds
6.8.3 Flexural Design Procedure (6.51)

The design procedures are essentially a summation of the material presented throughout Chapter 6 in order to summarise the material and introduce the reader to the column designs presented in Chapter 7. The process follows the material presented in the design manual TM5-1300 (1991) and is appropriately illustrated in the design flowchart by Mays and Smith (1995), which is included in Appendix F.

Asher Gehl 9913569

164

Structural Response
6.8.3.1 Reinforced Concrete Elements Concrete Column Dynamic Design

Exterior columns are required to withstand severe loading conditions. Examples of typical cross sections are shown below.

Figure 6.31 Rectangular Section with Equal reinforcement (TM5-1300, 1991)

Figure 6.32 Circular Column Section with Uniformly distributed reinforcement (TM5-1300, 1991)

Asher Gehl 9913569

165

Structural Response These columns are subjected to both axial and transverse loading. The axial load results from the direct transfer of floor and roof beam reactions while the transverse load is due to the direct impact of the blast load. For the purposes of this thesis, only the transverse load is considered. This is consistent with TM5-1300 (1991) which states that exterior columns are generally designed as beam elements. The axial load on these columns may be significant, but usually the effect of the transverse load is greater. This is represented in the following Figure that plots the column axial load capacity (Pu) versus the moment it can simultaneously withstand (M).

Figure 6.33 Column interaction diagram (TM5-1300, 1991)

When the ultimate eccentric load Pu applied to a column, is less than the balanced value Pb, the member acts more as a beam than as a column. In this region of Figure 6.33, the addition of axial load increases the moment capacity of the member. Consequently, the design of an exterior column as a beam, where the axial load is neglected, is conservative. The failure of the section is initiated by yielding of the tension steel.

Asher Gehl 9913569

166

Structural Response TM5-1300 (1991) presents the equations for the ultimate load capacity for both rectangular and circular columns with equal reinforcement, for all regions of the P-M curve of Figure 6.33.
Flexural Design process

The design of concrete elements subjected to blast loads involves an iterative (trial and error) design procedure in which the element is assumed and then its adequacy is verified through a dynamic analysis. The beams are initially transformed into singledegree-of-freedom systems completely independent of connected elements. The design range is generally divided into two regions, namely beams with support rotations less than 2 degrees (limited deflections) and support rotations between 2 and 4 degrees. Except for the type of cross-section available to resist moment, the design procedure is the same. Initially a concrete section and reinforcement are assumed. Examples of the types of cross sections are shown in Figure 6.15. Using equations 6.39 and 6.38 for type I and II cross-sections respectively, the moment capacities of the trial section are determined. The capacity of the section is defined in terms of its dynamic design stress fds, which is discussed in Section 6.8.1.4. The moment capacities are required to calculate the ultimate unit resistance ru and the equivalent elastic deflection XE. These parameters, along with the natural period of vibration T, define the equivalent single-degree-of-freedom system of the beam. Single Degree of Freedom response charts are presented in Appendix C. The ultimate unit resistance is obtained from the table for one-way elements in Appendix A5, in terms of the moment capacity of the particular section. The procedures and parameters necessary to obtain the equivalent elastic deflection and natural period are presented in Section 6.7.1 and 6.3.3.6 respectively.

Asher Gehl 9913569

167

Structural Response Chapter 5 describes procedures for determining the dynamic load, which is defined by its peak value P, and duration td. For elements of protection category 1 the ratios P/ru and td/T the ductility ratio Xm/XE and tm/td can be obtained from the response charts of Appendix C. These values Xm, which is the maximum deflection, and tm, the time to reach the maximum deflection, define the dynamic response of the beam. For elements of protection category 2, large deflections are permissible and the element is assumed to be close to the explosion and as such is subjected to very short duration blast loads. The load is taken as an impulse (area under the pressure-time curve). This assumption results in an insignificant error since the time for the element to reach the maximum deflection (td) is large in comparison to the actual duration of the load tm. A preliminary estimate of the size of the element can be obtained using the impulse equation (Equation 6.40). It should be noted that this preliminary design will underestimate the required element. According to TM5-1300 (1991) the use of these procedures for the design of nonlaced elements subjected to impulse load will result in a variety of errors depending upon support conditions, thickness of the concrete section, quantity and distribution of the flexural reinforcement, etc. However, these procedures may be used to obtain a trial section, which then may be analysed according to the above procedure. The dynamic analysis performed must check if the beam meets the allowable deflection criteria. It should be noted that in the calculation of T the values used for the effective mass and stiffness of the beam depends upon the allowable maximum deflection. When designing for completely elastic behaviour, the elastic stiffness is used while, in other cases, the equivalent elasto-plastic stiffness KE is used. The elastic value of the effective mass is used for the elastic range while, in the elastoplastic range; the effective mass is the average of the elastic and elasto-plastic values. For small plastic deformations, the value of the effective mass is equal to the average of the equivalent elastic value and the plastic value while for large plastic deformations; the effective mass is equal to the plastic value. Finally, the assumed section is designed for shear and torsion. If the beam does not meet the allowable response criteria, the required shear reinforcement is excessive, or the beam is over designed, a new concrete section is selected and the entire design procedure is repeated. Asher Gehl 9913569 168

Structural Response
6.8.3.2 Structural Steel Elements

As discussed, to design elements to resist an explosive event, the structural resistance required to limit the deflections to within the prescribed limiting values (according to Section 6.7) must be determined. In general, the resistance and deflection may be determined according to flexural capacity, on the basis that the shear capacity of the web is not exceeded. According to TM5-1300 (1991) the elastic shearing deformations of steel members are negligible as long as the depth to span ratio is less than about 0.2 and hence, a flexural analysis is normally sufficient for establishing maximum deflections. As such a flexural design only is considered in the column designs of Chapter 7, and shear design is not included. TM5-1300 (1991) and Mays and Smith (1995) present the procedure for shear design.
Resistance and Stiffness Functions

TM5-1300 presents the formulas for determining the stiffness and resistance for one way steel beams. The tables are included in this thesis as Appendix A. These tables are utilised in the design of a fixed column in Chapter 7. The values of M in Appendix A represent the plastic design moment, Mp according to the discussions of 6.8.2.2 For example, the value of ru used for the fixed support column with a uniformly distributed blast load is:
ru = 16 Mp/L2
6.52)

Asher Gehl 9913569

169

Structural Response
SDOF Response Charts

Dynamic response charts for one-degree-of-freedom systems in the elastic or elastoplastic range under various dynamic loads are given in Appendix C. To use the charts, the effective natural period of vibration of a structural steel beam must be determined. The procedures for determining the natural period of vibration for one-way elements are outlined in Section 6.3.6. The equation is repeated below.
Tn = 2 ( me

KE

= 2 ( K LM m

KE

(6.53)

This equation can be used to determine the natural period of vibration for any system for which the total effective mass, Me (Section 6.3.3.3), and equivalent elastic
B B

stiffness, KE (Appendix A3) are known.


B B

Preliminary Dynamic Load Factors

For preliminary flexural design of beams situated in a low pressure range, it is suggested that an equivalent static ultimate resistance equal to the peak blast pressure be used for those beams designed for 2 degrees support rotation.
ru=1.0Pmax
B B B B

(6.54)

For large support rotations, a preliminary dynamic load factor of 0.5 is recommended.
ru=0.5Pmax
B B B B

(6.55)

The determination of peak blast pressure that is utilised for the preliminary designs of Chapter 7 is discussed in Section 5.4 According to TM5-1300 (1991), since the load duration for a low pressure ranges is generally the same or longer than the period of vibration of steel elements (QuasiStatic), revisions to this preliminary design from a dynamic analysis will usually not be substantial. However, for larger pressures, generally the load duration is short compared with the period of vibration of the steel element (Impulsive). As such, this procedure may result in a substantial overestimate of the required resistance.

Asher Gehl 9913569

170

Structural Response
Final Check of Flexural Design

Once a dynamic analysis is performed on the element, the deformations must be checked with the limitations set in the criteria (Section 6.7). The provisions for local buckling, web crippling and lateral bracing must be met according to sections 5.24-26 of TM5-1300 (1991). The deformation criteria for beam elements including purlins, spandrels and girts are summarized in Section 5-35 of TM5-1300 (1991).
6.8.4 Differences between Steel and Concrete Structures

Qualitative differences between steel and concrete protective structures are summarized by TM5-1300 (1991) and presented below: 1. For the protection against impact from fragments as a result of an explosion and for restricting large deflections, a massive reinforced concrete structure, rather than a steel structure, is generally preferred. 2. Elements of explosive containment structures such as blast doors, ventilation closures, etc., are generally designed using structural steel. Fragment protection is usually accomplished by increasing the element thickness to resist fragment penetration or by providing supplementary fragment protection. 3. Structural steel elements are considerably more slender, both in terms of the overall structure and the components of a typical member cross-section. As a result, the effect of overall and local instability upon the ultimate capacity is an important consideration in steel design. Moreover, in most cases, plate elements and structures will sustain large deformations in comparison to those of more rigid concrete elements. 4. Concrete structures essentially have internal damping, provided by the cracking of the section that significantly reduces rebound effects. The rebound is essentially eliminated for larger deformations. In structural steel, however, a larger response in rebound, up to 100 percent, can be obtained for a combination of short duration load and a relatively flexible element. As a result, steel structures require that special provisions be made to account for extreme responses of comparable magnitude in both directions. TM5-1300 (1991) presents a thorough insight for providing resistance to rebound effects.

Asher Gehl 9913569

171

Structural Response 5. The treatment of stress interaction is more of a concern for structural steel, since each element of the cross section must be considered subject to a state of combined stresses. In reinforced concrete, the provision of separate steel reinforcement for flexure, shear and torsion enables the designer to consider these stresses as being carried by more or less independent systems. 6. For steel, special care must be taken, to provide for connection integrity up to the point of maximum response. For example, in order to avoid premature brittle fracture in welded connections, the welding characteristics of the particular grade of steel must be considered and the introduction of any stress concentrations at joints and notches in main elements must be avoided.

Asher Gehl 9913569

172

Local Column Design Procedure

Chapter 7 Local Column Design Procedure


7.1 Aim
The primary aim of this chapter is to demonstrate the theory presented in previous chapters in a practical manner. The design procedure is summarised in section 6.8.3 and closely follows the flowchart presented in Mays and Smith (1995) that is based on the procedures within TM5-1300 (1991). This flowchart is included as Appendix F. Reinforced concrete and structural steel columns of various cross sectional shape are designed according to the aforementioned chart. It is intended to examine the relationship that the Drag Coefficient exhibits on the final design. The design is intended to simply examine this effect and demonstrate the theory and is not intended to be a thorough analysis. The design is simply a preliminary flexural design and does not look into shear design or the supporting elements and connections. Overall the design procedure must establish the level of permissible damage for the element and the corresponding allowable amount of deflection related to that damage.

7.2 Design Loads and Safety Factors


The likelihood of a terrorist attack against a specific target although being an extremely devastating event is very low. Due to this low probability of occurrence there is a safety factor of 1 associated with the event. Permanent actions also require a partial safety factor of unity. Imposed actions, wind and earthquake loads require low combination factors due to the extremely low likelihood of the two events occurring simultaneously. For this design however, combination loads are ignored and the only load considered is the Blast load. According to TM5-1300 (1991) exterior columns may be designed as a beam element. Despite having significant axial load, the transverse load from the blast is greater. Thus the manual states that neglecting the axial load for these elements is conservative.

Asher Gehl 9913569

173

Local Column Design Procedure TM5-1300 (1991) however limits the allowable deformation for external columns due to the axial load. Essentially large plastic deformations are restricted by limiting the ductility ratio, Xm/XE to three.
B B B B

7.3 Drag Coefficients


As introduced in Chapter 5, there are a number of parameters that influence the drag on an object

Area Velocity of flow Mach Number Reynolds number Air density

Mays and Smith (1995) present the drag force is given by


FD = C D q s A
(7.1)

At lower speeds the drag due to friction is dominant force on an object. At higher speeds however the pressure difference is a significant parameter of the resultant force on an object. The Drag coefficient is essentially a measure of aerodynamic efficiency and is highly dependant upon viscosity (Reynolds number) and compressibility (Mach number) (Bray, 2004). As shown in Section 5.2.3 the Reynolds number is:
Re = VD

(7.2)

Where D is the characteristic length of the object (essentially the length perpendicular to the incident flow is the kinematic viscosity V is the velocity of the air behind the blast wave

Asher Gehl 9913569

174

Local Column Design Procedure The velocity at which the Blast wave acts is extremely high.
6 Pso a0 7 Po

U = 1+

(7.3)

For a 100kg Blast at 5m Standoff Peak static overpressure Ambient pressure Speed of Sound Blast Wave front Velocity Ps0= 1000kPa
B B

Po= 101Kpa
B B

a0=340m/s
B B

U=1020m/s

(Mach Number =3) The air density behind the wave front is given by:

s =

6 Pso + 7 Po 0 Ps + 7 Po

(7.4)

Where 0=the density of air at ambient temperature


B B

s=4 0
B B B B

(7.5)

At such high values of velocity and density the Reynolds number is in the order of 5 10 5 and as shown in Figure 5.4, the Drag Coefficients will significantly decrease. The laminar separation point occurs when the flow separates from the object and turns turbulent, as illustrated in Figures 5.5 and 7.1. At high velocities the separation point on an object moves further back such that there is significantly less wake and correspondingly less drag on the object.

Asher Gehl 9913569

175

Local Column Design Procedure

Figure 7.1 Separation point for higher velocity flow

According to the Figure 5.4 the actual decrease in Drag Coefficient can be as high as a factor of 4. For example a section with Reynolds number of 1 x 105 has a corresponding drag coefficient of 2, whereas a higher Reynolds number of 5 x 105 results in a drag coefficient of 0.5. With respect to the Mach number Bray (2004) presented the following variation to determine Drag Coefficient for a typical bullet.

Figure 7.2Drag Coefficient Vs Mach number for a typical bullet (Bray, 2004)

As seen at particularly high Mach numbers the Drag coefficient begins to decrease The Drag coefficient values presented by TM5-1300 (1991), are included in Appendix F and seem to correspond with Reynolds numbers of 1 x 105, which is inconsistent with the velocity at which the flow hits an object.

Asher Gehl 9913569

176

Local Column Design Procedure Kinney Graham (1985) present drag coefficients that are significantly lower and seem to be a more accurate representation of the high velocities of blast waves and their associated higher Reynolds numbers. This table is included as Appendix D. .Kinney and Graham (1985) specifically state that:
these tabulated values pertain to flows with high Reynolds numbers (to turbulent flows), as are to be expected for blast winds from explosions

For entire structures however Kinney and Graham (1985) recommend a larger value of 2 is used for the coefficient of Drag. For the subsequent designs of individual columns, the Drag Coefficients will be taken from both Kinney and Grahams (1985) table for Drag coefficients and also from TM5-1300 (1991). The sections chosen with their corresponding Drag coefficients are:
DRAG COEFFICIENT 2.05 1.2 2 1.55 1.55 2 0.1

SECTION I SECTION CIRCULAR SQUARE (FACE ON) SQUARE (EDGE ON) TRIANGULAR (EDGE ON) TRIANGULAR (FACE ON) ELLIPSE

However it must be stressed that for a thorough design, the effects of high Reynolds number on Drag coefficient should be inspected and any standard Drag Coefficients used cautiously.

Asher Gehl 9913569

177

Local Column Design Procedure

7.4 Reinforced concrete column Protection Category 1

DESIGN OF REINFORCED CONCRETE COLUMN SUBJECT TO BLAST LOADING ACCORDING TO PROTECTION CATEGORY 1

The primary aim is to provide the column with sufficient ductility to enable deflection consistent with the allowable degree of damage

The design is based on the ultimate limit state

The procedure follows the steps outlined in the design flow chart presented in the text: G.C Mays and P.D Smith: Blast Effects on Buildings (1995) according to the Tri-Services Design Manual TM5-1300

INITIAL VARIABLES

0.5

CD

0.5

0.5 m

________________________________________________________________________

STEP 1:

ESTABLISH DESIGN LOADS APPLY SAFETY FACTORS

________________________________________________________________________

Due to the low probability of the blast event to occur with other ultimate load case combinations (e.g Ultimate Wind or Earthquake, Imposed actions), safety factors of 0.33 are normally applied to minimise their affect according to BS( 5950 & 8110).This acts to minimise the effective resistance of the column. The permanent actions upon the element normally have a safety factor of unity

For this design however, all other loading effects are ignored and only the blast load acting alone on the column is considered with a partial safety factor of unity

Asher Gehl 9913569

178

Local Column Design Procedure


____________________________________________________________________________ STEP 2: SELECTION OF STRUCTURAL MATERIAL ESTABLISH CHARACTERISTIC STRENGTH APPLY DYNAMIC INCREASE FACTOR (DIF)

____________________________________________________________________________ The column is to be designed from Reinforced concrete. The ultimate static concrete compressive strength f cu : Yield strength of steel reinforcement fsy: Ultimate strength of steel reinforcement fu : f cu f sy fu fu = Blast load is essentially a rapidly applied load and thus the concrete's characteristic strength will increase due to strain rate effects. The dynamic increase factor is the ratio that converts the static strength under static loads to the equivalent dynamic stress value. Values for the Dynamic Increase factor are taken from the manual TM5-1300 DIF (CONCRETE): 1.25 DIF (STEEL REO): 1.2 (yield) 1.05(ultimate) 40 410 1.05.f sy Units are in MPa

Asher Gehl 9913569

179

Local Column Design Procedure


___________________________________________________________________________ STEP 3: SELECTION OF PROTECTION CATEGORY ESTABLISH LIMITING SUPPORT RATIOS STEEL: DUCTILITY RATIO CONCRETE: SUPPORT ROTATIONS

____________________________________________________________________________ The protection category is directly related to the allowable degree of damage after a blast event. The tolerated damage level is dependant upon the risk assesment for the structure and is controlled via the specification of a deflection limit.

Structural Steel Elements are controlled via the ductility ratio : Where: Xm is the total is the total deflection XE is the deflection at the elastic limit

X m dc X E dc

Reinforced Concrete elements are contolled via the support rotation :

The limitation in terms of support rotation stems from the elements resistance deflection function

For the above column, high level protection is required and the deformation limit will be in accordance with Protection category 1 such that the allowable rotation =2o

Asher Gehl 9913569

180

Local Column Design Procedure


________________________________________________________________________ STEP 4: ESTABLISH DYNAMIC DESIGN STRESSES FOR CONCRETE f AND STEEL REINFORCEMENT f dc ds

________________________________________________________________________

The dynamic ultimate concrete compressive strength f

du: f du

1.25.f cu

f du =

MPa

According to Appendix Table B2: Dynamic design stress for Concrete fdc : f dc Converted to units of N and m f dc =
dy:

f du .10

N .m

The dynamic yield design stress for steel reinforcement f

f dy f dy =

1.2 .f sy .10

N .m

The dynamic ultimate design stress for steel reinforcement f

du:

f du f du =

1.05.f u .10

N .m

According to Appendix Table B2: Dynamic design stress for Steel Reinforcement fds : For Protection category 1 f ds f ds = f dy N .m
2

Standard Units of m and N

Asher Gehl 9913569

181

Local Column Design Procedure


________________________________________________________________________ QUESTION 1: IS THE ELEMENT EXPECTED TO RESPOND TO IMPULSE? ________________________________________________________________________

The response time of the element must be approximated to either greater than or less than the duration of the blast load. Essentially Structural steel systems and Reinforced Concrete systems that are designed to respond elastically according to Protection category 1 (<2o ) generally respond to the load within the pressure-time (dynamic) regime.

Reinforced concrete elements that are designed for protection category 2 (>2o ) will generally respond witin the impulsive regime.

For the above Reinforced Concrete column designed forProtection Category 1 , the response is not expected to be impulsive, but rather to respond to pressure

Assumption 1 :

tm td

Asher Gehl 9913569

182

Local Column Design Procedure


__________________________________________________________________________ STEP 5: CONSIDER THE IDEALIZED EXPLOSIVE LOAD FROM CONWEP TM855-1: CALCULATE REFLECTED IMPULSE DURATION OF BLAST LOAD (td)

__________________________________________________________________________ The explosive threat considered consists of a 100kg charge of TNT equivalent high explosive The standoff distance is considered to be 5m away from the structural element DEFINE THE SHAPE OF THE COLUMN AND ASSOCIATED DRAG COEFFICIENT

CD =

Blast load idealization Appendix E For small single element

Output results from Conwep: Impulse: Positive phase duration: Peak Incident Overpressure: i td Ps 944 9.3 1002 kPa .msec msec kPa

Asher Gehl 9913569

183

Local Column Design Procedure


Calculations for peak pressure and impulse of Blast load idealisation

Ambient pressure :

Po

101.3 5 .P s 2 Ps
2

kPa

Maximum dynamic Pressure:

qs

7 .P o

qs =
3 C D .q s .10

kPa

Total peak pressure on element

Ps

note converted to Standard units

P=

N .m

I Total peak Impulse on element note converted to Standard units I=

P .t d .10 2

N .s .m

Asher Gehl 9913569

184

Local Column Design Procedure


__________________________________________________________________________ STEP 6: DEFINE THE RESISTANCE DEFLECTION FUNCTION IN TERMS OF d

__________________________________________________________________________ The structural response of the Reinforced concrete column is represented by the idealized resistance deflection function.

Idealized resistance - deflection function The Ultimate unit resistance values have been determined assuming a plastic response to the applied transient load Refer to Appendix Table A5 for the appropriate Ultimate unit resistance r
u

Mn is the negative plastic moment Mp is the positive plastic moment For a fully fixed column: H=3m H 3 m r u( d)

8 . Mn H
2

Mp

For a type 1 section <2o the concrete is effective in resisting the moment caused by the deformation. For this type of section the concrete cover on both sides will remain intact, however there will be substantial cracking on the compression face. Generally tensile reinforcement is sufficient, however it is desirible for the section to also contain compression reinforcement to cater for rebound effects. Therefore symmetrical reinforcement is provided for type 1 sections Mn Mp

Asher Gehl 9913569

185

Local Column Design Procedure


Based on the assumed value of D and ignoring the contribution of the compression reinforcement: With 25mm cover to Y12 stirrups d d= Consider the steel reinforcement ratio s : As is the area of reinforcement on each face d is the effective depth of the tensile reinforcement Assume a 5% reinforcement ratio As As = For such type 1 sections the ultimate positive moment capacity M
p

0.037 m

As b .d 0.005 s .b .d m :
2

Let (x) be the depth from the compression face to the neutral axis:

x x=

A s .f ds 0.6 .b .f dc m

This value of x justifies ignoring the effect of the compression reinforcement Mp The ultimate positive moment capacity is then: Mp = Mn The ultimate negative moment capacity is then: Mn = : Mp As. f ds b .( d 0.45.x )

Therefore the resistance function can be expressed

ru

8 . Mn H
2

Mp

ru =

N .m

Asher Gehl 9913569

186

Local Column Design Procedure


Consider the maximum permitted deflection achieved X Plastic hinge assumed :

Type 1 section: Allowable rotation =2o

2 deg Xm 0.5 .H 0.5 .H .( tan( ) ) m

Assuming appropriate shear reinforcement to withstand plastic deflections

tan( )

Xm Initial Assumption for maximum deflection Consider the elastic stiffness KE : Xm =

Appendix Table A3 defines the equivalent elastic stiffness as: Ke Concrete Properties: Concrete density: Concrete characteristic strength: Concrete-Youngs modulus: fc Ec Ec = MPa Steel-Youngs modulus:
6 Es MPa 200000.10

307 .E c .I c H
4

2400 32
1.5

kg .m MPa

.0.043. f .106 c N .m
2

N .m

modular ratio

n n=

Es Ec

Asher Gehl 9913569

187

Local Column Design Procedure


Determine F from Figure 6.27 in the Thesis using the value of n above and the initial reinforcement ratio

F Ic Ke

0.025 F .b .d
3

307 .E c .I c H
4

Ke =

Determine the deformation at the elastic limit X ru Ke

XE

XE=

___________________________________________________________________________ STEP 7: CALCULATE THE NATURAL PERIOD OF THE ELEMENT

___________________________________________________________________________ Table A1 in the Appendix determines the Load Mass Factor K K LM m m= The natural period of the column T : T 2 . . K LM.m Ke 0.66 :

LM

Based on the initial estimate of D Mass per unit length : m

.D kgm
2

T=

Asher Gehl 9913569

188

Local Column Design Procedure


______________________________________________________________________ STEP 8: REFER TO SDOF RESPONSE CHARTS DETERMINE DUCTILITY RATIO SUPPORT ROTATION RESPONSE TIME

______________________________________________________________________ Ultimate Resistance ru = P= kPa

The load: P

Thus the values as required in the SDOF response charts (Appendix C) ru P

for ru /P

Note

conversion for td /T

t d .10 T

to seconds for td

The value for X m/X E can now be read from the SDOF response chart 1 Xm XE 5 XE=

Taken From the charts:

Now the support rotation must be checked: XE The Elastic deformation: XE= ru Ke m

Asher Gehl 9913569

189

Local Column Design Procedure


Given the ductility ratio; now determine the maximum deflection achieved X :

max

X max X max =

.X E m

To determine :

atan

X max 0.5 .H atan

= deg

X max 0.5 .H

= deg The value for t m/td can now be read from the SDOF response chart 2

tm td

1.1

Asher Gehl 9913569

190

Local Column Design Procedure


___________________________________________________________________________ CHECKPOINT 1: IS THE DUCTILITY RATIO AND THE SUPPORT ROTATION WITHIN THE PRESCRIBED LIMITS

<2o <10 ____________________________________________________________________________

IF <10 Deformation has been limited <2o = deg CHOSEN SECTION IS SATISFACTORY 5 PROCEED TO CHECKPOINT 2 to 2o Protection Category 1 is satifsfactory

OTHERWISE A LARGER DEPTH MUST BE CHOSEN RETURN TO STEP 6 (end) ___________________________________________________________________________ CHECKPOINT 2: IS THE INITIAL ASSUMPTION JUSTIFIED Is tm.td<3

__________________________________________________________________________ IF tm td Note

3 Assumption 1 is justified Quasi-Static loading design is valid for the element and the flexural capacity is sufficient for the blast load.

tm td

1.1

PROCEED TO DESIGN FOR SHEAR DYNAMIC REACTIONS SUPPORTING ELEMENTS OTHERWISE Section must be designed according to Protection Category 1 Repeat design procedure for Impulsive Design See Mathcad Worksheet 1

Asher Gehl 9913569

191

Local Column Design Procedure

7.5 Reinforced concrete column Protection Category 2

DESIGN OF REINFORCED CONCRETE COLUMN SUBJECT TO BLAST LOADING ACCORDING TO PROTECTION CATEGORY 2

The primary aim is to provide the column with sufficient ductility to enable deflection consistent with the allowable degree of damage The design is based on the ultimate limit state The procedure follows the steps outlined in the design flow chart presented in the text: G.C Mays and P.D Smith: Blast Effects on Buildings (1995) that follows the regulations of the Tri-Services Design Manual TM5-1300 INITIAL VARIABLES b

0.5

CD

1.2

0.01

_________________________________________________________________________ STEP 1: ESTABLISH DESIGN LOADS APPLY SAFETY FACTORS

_________________________________________________________________________

Due to the low probability of the blast event to occur with other ultimate load case combinations (e.g Ultimate Wind or Earthquake, Imposed actions), safety factors of 0.33 are normally applied to minimise their affect according to BS( 5950 & 8110).This acts to minimise the effective resistance of the column. The permanent actions upon the element normally have a safety factor of unity For this design however, all other loading effects are ignored and only the blast load acting alone on the column is considered with a partial safety factor of unity

Asher Gehl 9913569

192

Local Column Design Procedure


__________________________________________________________________________ STEP 2: SELECTION OF STRUCTURAL MATERIAL ESTABLISH CHARACTERISTIC STRENGTH APPLY DYNAMIC INCREASE FACTOR (DIF)

__________________________________________________________________________ The column is to be designed from Reinforced concrete. The ultimate static concrete compressive strength f cu : Yield strength of steel reinforcement fsy: Ultimate strength of steel reinforcement fu : f cu f sy fu fu =

40
410

MPa MPa

1.05 .f sy
MPa

Blast load is essentially a rapidly applied load and thus the concrete's characteristic strength will increase due to strain rate effects. The dynamic increase factor is the ratio that converts the static strength under static loads to the equivalent dynamic stress value. Values for the Dynamic Increase factor are taken from the manual TM5-1300 DIF (CONCRETE): 1.25 DIF (STEEL REO): 1.2 (yield) 1.05(ultimate)

Asher Gehl 9913569

193

Local Column Design Procedure


________________________________________________________________________ STEP 3: SELECTION OF PROTECTION CATEGORY ESTABLISH LIMITING SUPPORT RATIOS STEEL: DUCTILITY RATIO CONCRETE: SUPPORT ROTATIONS

_________________________________________________________________________ The protection category is directly related to the allowable degree of damage after a blast event. The tolerated damage level is dependant upon the risk assesment for the structure and is controlled via the specification of a deflection limit.

Structural Steel Elements are controlled via the ductility ratio Where: Xm is the total is the total deflection XE is the deflection at the elastic limit

X m dc X E dc

Reinforced Concrete elements are contolled via the support rotation

The limitation in terms of support rotation stems from the elements resistance deflection function

For the above column, significant plastic deformation will be tolerated and the deformation limit will be in accordance with Protection category 2 such that the allowable rotation =4o

Asher Gehl 9913569

194

Local Column Design Procedure


________________________________________________________________________ STEP 4: ESTABLISH DYNAMIC DESIGN STRESSES FOR CONCRETE f AND STEEL REINFORCEMENT f dc ds

________________________________________________________________________

The dynamic ultimate concrete compressive strength f

du:

f dcu f dcu =

1.25 .f cu
MPa

According to Appendix Table B2: Dynamic design stress for Concrete fdc : f dc f dc = The dynamic yield design stress for steel reinforcement f
dy:

f dcu MPa

f dy f dy = The dynamic ultimate design stress for steel reinforcement f


du:

1.2 .f sy
N .m 2

f du f du =

1.05 .f u
MPa

According to Appendix Table B2: Dynamic design stress for Steel Reinforcement fds : For Protection category 2 f ds Converting to standard UNITS of m and N f ds = Nm 2 f dy f du f dy 6 .10

Asher Gehl 9913569

195

Local Column Design Procedure

QUESTION 1: IS THE ELEMENT EXPECTED TO RESPOND TO IMPULSE?

The response time of the element must be approximated to either greater than or less than the duration of the blast load. Essentially Structural steel systems and Reinforced Concrete systems that are designed to respond elastically according to Protection category 1 (<2o ) generally respond to the load within the pressure-time (dynamic) regime.

Reinforced concrete elements that are designed for protection category 2 (>2o ) will generally respond witin the impulsive regime.

For the above Reinforced Concrete column designed forProtection Category 2 , the response is expected to be impulsive

Assumption 1 :

tm td

Asher Gehl 9913569

196

Local Column Design Procedure


________________________________________________________________________ STEP 5: CONSIDER THE IDEALIZED EXPLOSIVE LOAD FROM CONWEP TM855-1: CALCULATE REFLECTED IMPULSE DURATION OF BLAST LOAD (td)

_________________________________________________________________________ The explosive threat considered consists of a 100kg charge of TNT equivalent high explosive The standoff distance is considered to be 5m away from the structural element DEFINE THE SHAPE OF THE COLUMN AND ASSOCIATED DRAG COEFFICIENT CD =

Blast load idealization Appendix E For small single element Output results from Conwep: Impulse: i

944

kPa .msec

Positive phase duration:

td

9.3 msec 1002

Peak Incident Overpressure:

Ps

kPa

Asher Gehl 9913569

197

Local Column Design Procedure


Calculations for peak pressure and impulse of Blast load idealisation

Ambient pressure :

Po qs

101.3 5 .P s 2 2 Ps

kPa

Maximum dynamic Pressure:

7 .P o
kPa

qs = Total peak pressure on element P P= Total peak Impulse on element note converted to SI units I I= P .t d Ps C D .q s kPa

2
N .s .m 2

Asher Gehl 9913569

198

Local Column Design Procedure


__________________________________________________________________________ STEP 6: DEFINE THE RESISTANCE DEFLECTION FUNCTION IN TERMS OF d c

__________________________________________________________________________

The structural response of the Reinforced concrete column is represented by the idealized resistance deflection function.

Idealized resistance - deflection function

The Ultimate unit resistance values have been determined assuming a plastic response to the applied transient load Refer to table B3 for the appropriate Ultimate unit resistance r
u

Mn For a fully fixed column: H=3m Mp

is the negative plastic moment capacity

is the positive plastic moment capacity 8 . Mn dc 2 H Mp dc

r u dc

Asher Gehl 9913569

199

Local Column Design Procedure


Note that in order to sufficiently acheive deflection over =2o for a type 2 section there must be sufficient compression reinforcement to : i) sustain the section after the concrete in compression crushes at =2o ; and ii) allow the full development of the tensile steel Therefore symmetrical reinforcement must be provided for type 2 sections Mn dc Mp dc
p

For such type 2 sections the ultimate positive moment capacity M f ds b

As is the area of reinforcement on each face dc is the distance between the centroid of tensile and compression reinforcement

Mp dc

As.

.d c

Consider the steel reinforcement ratio s :

As b .d c

Assume a 1% reinforcement ratio s = The ultimate positive moment capacity is then: Mp dc The ultimate negative moment capacity is then: Mn dc s .f ds .d c 2 s .f ds .d c 2

Therefore the resistance function can be expressed in terms of dc : 8 . Mn dc 2 H

r u dc

Mp dc

r u dc

Asher Gehl 9913569

200

Local Column Design Procedure


Consider the maximum permitted deflection achieved X Plastic hinge assumed :

Type 2 section: Allowable rotation =4o

4 deg
Xm

tan( )

0.5 .H 0.5 .H .( tan( ) )


m

Xm Xm = Consider the elastic stiffness KE :

Table A1 from the Appendix defines the equivalent elastic stiffness as: Ke dc Material Properties: concrete density:

307 .E c .I c d c
4 H 3

2400 32

kg .m

concrete characteristic strength:

fc MPa

MPa

Concrete-Youngs modulus:

Ec Ec =

1.5 .

6 0.043 . f c .10

N .m

Steel-Youngs modulus:

Es

200000 .10
Es Ec

N .m

modular ratio

n n=

Asher Gehl 9913569

201

Local Column Design Procedure


Determine F from Figure 6.27 in the Thesis using the value of n above and the initial reinforcement ratio F

0.043
F .b .d c 3

Ic dc

Ke dc

307 .E c .I c d c
4 H

Ke dc

307 .E c . F .b .d c 3
4 H

Ke dc

Determine the deformation at the elastic limit X

: r u dc Ke dc

X E dc

X E dc

Asher Gehl 9913569

202

Local Column Design Procedure


_______________________________________________________________________ STEP 7: CALCULATE THE LOAD MASS FACTOR CALCULATE THE UNIT MASS OF ELEMENT IN TERMS OF d c

________________________________________________________________________ Table A1 in the Appendix determines the Load Mass Factor K :

LM

K lm

0.66

The unit mass of concrete (m) is determined from the following equation: m dc .d c

2400

kg .m

3 kg .m 2

m dc

2400 .d c

_______________________________________________________________________ STEP 8: SOLVE THE BASIC IMPULSE EQUATION FOR d


c

_______________________________________________________________________ Below is the equation for the impulse loading regeme, this equation is solved for the distance between the centroid of the compression and tension reinforcement in the section d c : 2 I

Consider the Basic Impulse Equation

2 .K lm .m d c

r u dc

Xm

X E dc

Rearranging the Impulse equation with each variable in terms of input parameters and d c:

16 . s .f ds .d c2
f dc

16 . s .f ds .d c2 . X m 2
H

2 H

2 I

2.

307 .E c . F .b .d c3
4 H

2 .K lm . .d c

Asher Gehl 9913569

203

Local Column Design Procedure


In order to solve the above equation fordc Mathcad requires an initial guess for the depth: dc

0.3

mm

The equation can now be solved using Mathcads root function to find the root of the function

root f d c , d c = dc The distance between the centroids of the compression and tension reinforcement: root f d c , d c

dc =

Consider the steel reinforcement ratio s :

As b .d c

Using the initial assumption of s Calculate the a rea of Steel in the section:

As As =

s .b .d c

Total depth of cross section: Total recommmended cover to 12mm stirups is 20mm Overall mimimum cover is 40mm Let D=Diameter of steel reinforcement bar: Define function h(D) for Area of bar: D

12 , 14 .. 32
. 2 D

h( D )

Asher Gehl 9913569

204

Local Column Design Procedure


The following table enables the area of steel and number of Bars to be chosen it is purely for presneting design process and has no effect on the overall results D= h( D ) =

2 .h ( D ) =

3 .h ( D ) =

4 .h ( D ) =

Ensuring 40mm cover and 2x12mm steel stirrups The overall thickness of the section is:

Dc Dc =

d c .1000

24
mm

80

28

Rounding up to nearest...

Dc

910 mm

Asher Gehl 9913569

205

Local Column Design Procedure


_______________________________________________________________________ STEP 9: CALCULATE THE RESPONSE TIME OF THE ELEMENT (tm)

_______________________________________________________________________ The elements response time is determined from the impulse and ultimate resistance: Impulse Ultimate resistance: I= r u dc = tm tm = tm Note converting td to seconds 3 = d c .1000 = mm I r u dc

t d .10 ________________________________________________________________________ CHECKPOINT 2: IS THE INITIAL ASSUMPTION JUSTIFIED Is tm/td>3

________________________________________________________________________ IF tm td

3
Assumption 1 is justified Impulsive loading design is valid for the element and the flexural capacity is sufficient for the blast load.

tm td

1.044
Design for Shear Calcuate support reactions Design supporting elements

OTHERWISE

Section has insufficient capacity to withstand the imposed blast load. Section must be redesigned to respond to pressure according to Worksheet 2. Return to Worksheet 2

Asher Gehl 9913569

206

Local Column Design Procedure

7.6 Structural Steel Protection Category 1

DESIGN OF STRUCTURAL STEEL COLUMN SUBJECT TO BLAST LOADING

The primary aim is to provide the column with sufficient ductility to enable deflection consistent with the allowable degree of damage. The design is based on deformations well within the plastic domain. The design is based on the ultimate limit state The procedure follows the steps outlined in the design flow chart presented in the text: G.C Mays and P.D Smith: Blast Effects on Buildings (1995) Parameters of chosen Sections were taken from BHP Steel tables

___________________________________________________________________________ STEP 1: ESTABLISH DESIGN LOADS APPLY SAFETY FACTORS

___________________________________________________________________________

Due to the low probability of the blast event to occur with other ultimate load case combinations (e.g Ultimate Wind or Earthquake, Imposed actions), safety factors of 0.33 are normally applied to minimise their affect according to BS( 5950 & 8110). This acts to minimise the effective resistance of the column. The permanent actions upon the element normally have a safety factor of unity

For this design however, all other loading effects are ignored and only the blast load acting alone on the column is considered with a partial safety factor of unity

Asher Gehl 9913569

207

Local Column Design Procedure


_________________________________________________________________________ STEP 2: SELECTION OF STRUCTURAL MATERIAL ESTABLISH CHARACTERISTIC STRENGTH APPLY DYNAMIC INCREASE FACTOR (DIF)

_________________________________________________________________________ The column is to be designed from Structural steel. Elastic modulus of Steel E: E
6 2 200000.10 Nm

Yield strength of steel f y:

fy

410 .10

Nm

The ultimate strength f u:

fu

440 .10

Nm

TM5-1300 applies a 10% increase factor for steels less than Grade 50

1.1

Blast load is essentially a rapidly applied load and thus the yield stress of steel will increase due to strain rate effects. The dynamic increase factor is the ratio that converts the static strength under static loads to the equivalent dynamic stress value.

Values for the Dynamic Increase factor are taken from the manual TM5-1300 For bending: Yield stress DIF : c = 1.20 Ultimate stress DIF : c =1.05

Asher Gehl 9913569

208

Local Column Design Procedure


__________________________________________________________________________ STEP 3: SELECTION OF PROTECTION CATEGORY ESTABLISH LIMITING SUPPORT RATIOS STEEL: DUCTILITY RATIO CONCRETE: SUPPORT ROTATIONS

___________________________________________________________________________ The protection category is directly related to the allowable degree of damage after a blast event. The tolerated damage level is dependant upon the risk assesment for the structure and is controlled via the specification of a deflection limit.

Structural Steel Elements are controlled via the ductility ratio : Where: Xm is the total is the total deflection XE is the deflection at the elastic limit

Xm XE

Reinforced Concrete elements are contolled via the support rotation :

The limitation in terms of support rotation stems from the elements resistance deflection function

For the above column, high level protection is required and the deformation limit will be in accordance with Protection category 1 such that the ductility ratio <10 and the support rotation <2o

Asher Gehl 9913569

209

Local Column Design Procedure


____________________________________________________________________________ STEP 4: ESTABLISH DYNAMIC DESIGN STRESSES FOR STRUCTURALSTEEL fds

____________________________________________________________________________

STRESS TYPE PROTECTION CATEGORY 1 BENDING SHEAR 2 1 AND 2

EQUIVALENT STATIC ULTIMATE RESISTANCE fds = fdy fds = fdy+ (fdu - fdy )/4 fdy = 0.55 fds

Average Yield stress increase factor

a=

Yield stress DIF:

1.2 c .a .f y Nm
2

The dynamic yield stress f dy :

f dy f dy = f ds

f dy Nm
2

The dynamic Design Stress fds :

f ds =

Asher Gehl 9913569

210

Local Column Design Procedure


__________________________________________________________________________ STEP 5: CONSIDER THE IDEALIZED EXPLOSIVE LOAD FROM CONWEP TM855-1: CALCULATE REFLECTED IMPULSE DURATION OF BLAST LOAD (td)

__________________________________________________________________________ The explosive threat considered consists of a 100kg charge of TNT equivalent high explosive The standoff distance is considered to be 5m away from the structural element DEFINE THE SHAPE OF THE COLUMN AND ASSOCIATED DRAG COEFFICIENT Take values from Appendix CD 0.1

Blast load idealization Appendix E Output results from Conwep: Impulse: i 944 For small single element

Positive phase duration:

td Ps

9.3

msec

Peak Incident Overpressure:

1002

kPa

Asher Gehl 9913569

211

Local Column Design Procedure


Calculations for peak pressure and impulse of Blast load idealisation

Ambient pressure :

Po

101.3 5 .P s 2 Ps
2

kPa

Maximum dynamic Pressure:

qs

7 .P o

qs =
3 C D .q s .10

kPa

Total peak pressure on element Note conversion of units

Ps

P=

Nm

I Total peak Impulse on element Note conversion of units I=

P .t d 2 .1000 Nsm
2

Asher Gehl 9913569

212

Local Column Design Procedure


____________________________________________________________________________ STEP 6: PRELIMINAR Y DESIGN USING EQUIVALENT STATIC RESISTANCE SELECTION OF STEEL SECTION

___________________________________________________________________________ Initially an equivalent static ultimate resistance is assumed according to the required level of protection: The Ultimate unit resistance values have been determined assuming a plastic response to the applied transient load

PROTECTION CATEGORY 1 2

EQUIVALENT STATIC ULTIMATE RESISTANCE rus =1.0pmax rus =0.5pmax

Refer to Appendix Table A5 for the appropriate Ultimate unit resistance ru :

For a fully fixed column: H=3m H 3 m

ru

8 . Mn H
2

Mp

Mn For Category 1: The ultimate equivalent static ultimate resistance r us: r us r us = M pi M pi = Since .>3 : Mp

Mp 1 .P

The Approximate Plastic moment of resistance M pi :

r us .H 16

Nm

M p f ds .Z f ds = Nm
2

Asher Gehl 9913569

213

Local Column Design Procedure


M pi f ds

Z Therefore : Minimum Plastic Section modulus

Z min

M pi f ds
3

Z min = Z min .10


9

Z The minimum plastic section modulus is: Z=

mm

THE FOLLOWING PROCEDURE IS PRESENTED PURELY FOR METHOD ONLY THE SECTIONS CHOSEN DO NOT HAVE THE ABOVE VALUE SPECIFIED FOR THE DRAG COEFFICIENNTS

Select an appropriate Steel Section: Select a 310UC158 Find actual Moment Capacity of section: Mp Z y .f ds N .m
2

Zy

1.23.10

Mp =

Asher Gehl 9913569

214

Local Column Design Procedure


________________________________________________________________________ STEP 7: CALCULATE THE NATURAL PERIOD OF THE ELEMENT

_______________________________________________________________________ Table A1 in the Appendix determines the Load Mass Factor K LM : For the chosen section: Mass per unit length : m Second Moment of Area Elastic modulus of Steel E: Equivalent elastic resistance: K E K LM m I E= 307 .E.I H
4

0.66

52.2 3.88.10
4

4 2

N .m

KE

KE = The natural period of the column T : T T= 2 . . K LM.m KE msec

Asher Gehl 9913569

215

Local Column Design Procedure


______________________________________________________________________ STEP 8: REFER TO SDOF RESPONSE CHARTS DETERMINE DUCTILITY RATIO SUPPORT ROTATION RESPONSE TIME

______________________________________________________________________ From the SDOF response charts: For the Steel section chosen : Mp = ru Ultimate Resistance The load: P P= Thus the values as required in the SDOF response charts ru P td T 8 .M p H
2

N .m

ru = N .m
2

for ru /P

for td /T

The value for X /XE can now be read from the appropriate SDOF response chart m NOTE VALUES ARE OFF THE CHARTS Xm XE 9

assume

Taken From the charts :

Now the support rotation must be checked: XE The Elastic deformation: XE= ru KE m

Asher Gehl 9913569

216

Local Column Design Procedure


Consider the maximum permitted deflection achieved X : Plastic hinge assumed

Xm Xm =

.X E

To determine :

atan

Xm 0.5 .H atan

= deg Xm

0.5 .H

= deg

Asher Gehl 9913569

217

Local Column Design Procedure


___________________________________________________________________________ CHECKPOINT 1: IS THE DUCTILITY RATIO AND THE SUPPORT ROTATION WITHIN THE PRESCRIBED LIMITS <2o <10

________________________________________________________________________

IF <10 <2

CHOSEN SECTION IS SATISFACTORY PROCEED TO DESIGN FOR: SHEAR LOCAL BUCKLING LATERALBUCKLING

OTHERWISE A LARGER SECTION MUST BE CHOSEN RETURN TO STEP 6 (end)

Asher Gehl 9913569

218

Local Column Design Procedure

7.7 Results
The results of the analysis are presented in Appendix G.

7.8 Discussion
For the reinforced concrete column, the support rotation must essentially be limited to one degree in order to survive the magnitude of the blast load. The maximum Drag coefficient of the section that enables a successful design according to the deformation criteria of Protection category 1 is 0.5. Therefore the column sections in both Kinney and Graham (1985) and TM5-1300 (1991) are insufficient. However due to the high Reynolds numbers it is expected that the actual value of the Drag coefficient is lower. The element may not be successfully designed according to protection category 2 due to not satisfying the critical value of the limiting ratio of tm/td>3. It is important however to note that the supports chosen for the column, were fixed ends. By providing different support conditions, it is expected that the response duration (tm) would be larger and thus satisfy the critical limiting ratio of tm/td>3. This seems to be in accordance with the requirement for exterior columns according to TM5-1300 (1991) as mentioned in section 7.2. For the exterior columns significant plastic deformation is not recommended and as such must be designed according to Protection category 1. For the structural Steel columns it was seen that a reduction in the Drag Coefficient resulted in a lower minimum Plastic section modulus. As such sections the shape of the column can greatly reduce the overall size of the required section due to the corresponding reduction in the applied load. The Drag Coefficient has an influence on the ultimate behaviour of the section under the impact of a blast load. As such it is recommended that ground floor columns, where the magnitude of the pressures is the highest, be designed with sections of low Drag coefficients, in order to improve the overall response under the load.

Asher Gehl 9913569

219

Progressive Collapse

Chapter 8 Progressive Collapse


8.1 Introduction
Progressive collapse can be considered as a chain reaction of failures resulting from an initial localised failure. It essentially occurs when a structural element is loaded beyond its capacity and subsequently fails. The residual structure is forced to seek alternative load paths to redistribute the load applied to it. As a result, other elements may fail, causing further load redistribution. The process will continue until the structure can find equilibrium either by shedding load, as a by-product of other elements failing, or by finding stable alternative load paths. It has been broadly defined as a situation where local failure of a primary structural element leads to the collapse of adjoining members which in turn leads to additional collapse (Allen and Schriever, 1972). Thus the final global damage is substantially greater than the local damage initially caused. The sequence of events that occur during a collapse scenario is appropriately illustrated in Figure 8.1.

Figure 8.1 Sequence of events during structural collapse (Smith, Brokaw, Swatzell 1998)

Asher Gehl 9913569

220

Progressive Collapse Throughout the course of this chapter a number of the more famous examples of progressive collapse will be presented. Following this brief history, the general design methods that are commonly used to reduce the risk of progressive collapse will be presented. This chapter attempts to present an insight into the phenomenon of progressive collapse. It is important to note, that following the catastrophic collapses of recent history, there has been a dedication to research and many analytical reports on this topic. The chapter continues with a comparison and review of the current guidelines available on progressive collapse in a number of countries to discuss their various attributes. The chapter concludes with the development of future research that is fundamental in order to appropriately standardise the necessary design procedures.

8.2 History of Progressive Collapse


Past events have forced the re-evaluation of existing engineering design approaches and procedures. Generally, design codes are developed for likely loading scenarios based on structures size, intended purpose and location. Upon the application of an abnormal loading situation like a blast wave, there have been many instances where catastrophic failures have illustrated the inadequacies of standard design procedures to redistribute the load after a local failure. These examples are examined below.
8.2.1 Ronan Point, London 1968

The collapse of the Ronan Point apartment complex in 1968 was primarily due to the lack of reinforcement continuity between the precast reinforced concrete panels (Moore, 2002). Due to this lack of connection detail the structure was unable to seek alternate load paths to redistribute the load.

Asher Gehl 9913569

221

Progressive Collapse After this event, prevention of progressive collapse became one of the most crucial issues in engineering. A large-scale effort began for code-writing bodies and governmental agencies to develop design guidelines and criteria that would effectively reduce or eliminate the susceptibility of buildings to this form of failure. The general focus of these guidelines was aimed towards improving redundancy and providing alternate load paths, to ensure that loss the loss of any single component would not lead to a total collapse. Through an appropriate combination of improved redundancy, local resistance and interconnection, it should be possible to greatly reduce the susceptibility of buildings to disproportionate collapse.
8.2.2 Murrah Building Oklahoma 1995

The effort to develop an effective system of codes and guidelines was given a further shake-up following the catastrophic collapse of the Murrah building in Oklahoma as a result of a large truck bomb exploding at the front of the building. The truck bomb at the base of the building destroyed three columns. Loss of support from these columns led to the failure of a transfer girder. Failure of the transfer girder caused the collapse of columns supported by the girder and floor areas supported by those columns.

Figure 8.2 Murrah Federal Office Building after 1995 Attack (Mlakar et al, 1997)

Asher Gehl 9913569

222

Progressive Collapse
8.2.3 World Trade Center Tower 1 and 2, New York 2001

The importance of mitigating progressive collapse was again illustrated in September 2001 when each of the twin towers of World Trade Center 1 and 2 collapsed upon the impact of a Boeing 767 jetliner crashing into each tower at high speed. The crash caused structural damage upon impact and most critically set off an intense fire within the building (see Figure. 8.3); then its structure near the impact zone lost ability to support the load above it as a result of the impact and fire damage combination. Having lost its strength for support, the structure above collapsed. The weight and impact of the upper part of the tower collapsing caused a progression of failures extending downward all the way to the ground (Wilkinson, 2004).

Figure 8.3 World Trade Center during progressive collapse (http://www.civil.usyd.edu.au/latest/wtc.php#why2001)

Asher Gehl 9913569

223

Progressive Collapse

8.3 Design Methods to Reduce Progressive Collapse


There are, in general, three alternative approaches to designing structures in order to reduce their susceptibility to disproportionate collapse: 1. Redundancy or alternate load paths; 2. Local resistance method and 3. Interconnection or continuity
8.3.1 Redundancy or Alternate Load Paths

The key behind the design of buildings to withstand the effects of a blast is to ensure the building has sufficient robustness. This is essentially the ability to redistribute the load when an element has been taken away or its capacity severely impacted. In this approach, the local failure of a structural element is tolerated providing the structure is designed so that alternate paths are available for the load from the failed structural element to be redistributed and prevent any further collapse. This approach has the benefit of simplicity and directness. In its most common application, designing for increased redundancy requires that a building be able to tolerate loss of any one column without collapse. This is an objective, easily understood performance requirement. This is illustrated by the following Figure in Whittaker and Hamburgers finite element analysis (2004).

Figure 8.4 Illustration of alternate load path distribution (Whittaker and Hamburger, 2004)

Asher Gehl 9913569

224

Progressive Collapse The primary advantage of this approach is that is promotes structural systems with ductility, continuity, and energy absorbing properties (Smilowitz 2002). Essentially it prevents structures like the Murrah Building from using large transfer girders, which prevents all of the columns from extending to the ground floor. Shankar Nair (2004) outlines the problem with the redundancy approach in the paper Progressive Collapse Basics. The main issue with the practice is that there is no variation with respect to degree of vulnerability from different sized columns. Nair (2004) states:

There is a much higher level of safety for larger built up sections over thinner smaller sections, such that an explosion of a given magnitude could easily take out several smaller columns as easily as a single larger section making the one column redundancy classification completely inadequate to prevent collapse for the smaller column case.
The codes and standards that utilise the requirement of redundancy within their guidelines do not distinguish between the two situations of varying vulnerability. They treat every column with the same likelihood of being destroyed regardless of their susceptibility to an abnormal loading event.
8.3.2 Local Resistance

This design approach attempts to reduce susceptibility to progressive collapse by providing critical structural elements with increased resistance to abnormal loading events in order to altogether prevent the collapse. In order to achieve this reduction, the nature of potential attack is required in order to specify the load that must be resisted. This is of course is a very difficult overall process to codify in a simple and objective way, given the high variety and magnitude of risks for a particular structure. A number of guidelines however specify a minimal load that must be resisted in order to comply with the criteria.

Asher Gehl 9913569

225

Progressive Collapse This method is commonly applied to the retrofit of buildings. Substantial improvement to the overall structures resistance can be attributed to the direct redesign of selected critical local elements such as exterior columns (Smilowitz, 2002). Alternative approaches such as the alternative load path method are generally impractical for retrofit of tall buildings due to the high costs involved with providing compliance (Burns et al, 2002).
8.3.3 Interconnection or Continuity

This is an indirect design approach that is a means of improving either redundancy or local resistance (or both). The process requires the consideration of minimum strength reinforcement details, ductility of structural elements and their connections. Studies of many recent building collapses, in particular the Ronan Point apartment complex, have shown that the failure could have been avoided or at least reduced in scale, at a fairly small additional cost, if structural components had been interconnected more effectively (HMSO, 1968). This is the basis of the structural integrity requirements in the ACI 318 specification (ACI, 2002). If the minimum values are provided in building regulations and guidelines, the structural element may resist the local failure otherwise the appropriate connection detail or reinforcement continuity will provide adequate ability for the structure to redistribute its load via alternate load paths.

8.4 Current Codes and Guidelines


Today, a number of these codes and guidelines exist for various buildings under specific government supervision, which address the requirement for the mitigation of progressive collapse. According to Smith et al (1998) however, there exists no explicit engineering design method available pertaining to this potential structural problem.

Asher Gehl 9913569

226

Progressive Collapse
8.4.1 Introduction

It is not possible to design structures for absolute safety, nor is it economical to design for abnormal events unless they have a reasonable chance of occurrence. These abnormal events are not usually considered in the standard structural design process. Designing for other extreme events however, such as earthquakes, fires and high winds are part of standard code requirements and it is important to note that these events can also cause local failures. Therefore it is necessary in all loading cases to design buildings to mitigate progressive collapse. At present, provisions for mitigating progressive collapse in most building codes and standards around the world are based either on explicit design requirements or on general structural integrity requirements. Most European, including the U.K., building codes follow the former approach and U.S. codes and standards the latter approach. It should also be noted that there are no provisions or recommendations in the current Australian standards with regard to progressive collapse (Ngo et al 2003).
8.4.2 Minimum Design Loads for Buildings and Other Structures, ASCE 7-02 by the American Society of Civil Engineers

The American Society of Civil Engineers Minimum Design Loads for Buildings and Other Structures (ASCE, 2002) has a section (1.4) on general structural integrity. The section states that:

Buildings and other structures shall be designed to sustain local damage with the structural system as a whole remaining stable and not being damaged to an extent disproportionate to the original local damage. This shall be achieved through an arrangement of the structural elements that provides stability to the entire structural system by transferring loads from any locally damaged region to adjacent regions capable of resisting those loads without collapse. This shall be accomplished by providing sufficient continuity, redundancy, or energy-dissipating capacity (ductility), or a combination thereof, in the members of the structure.
Clearly, the focus in this ASCE standard is concerned with redundancy and alternate load paths over all other means of avoiding a progressive collapse failure.

Asher Gehl 9913569

227

Progressive Collapse In the 1995 version of the code there are however no specifications with respect to the design criteria to ascertain the required degree of redundancy, or any requirements in regard to specific threats (Smith, et al 1998). The current version was revised to reflect the increase of progressive collapse from terrorism, (especially Oklahoma 1995) and to include more specific suggestions for enhancement of general structural integrity (Dusenberry, 2002). The following suggestions are included here and are taken from the code. 1. Good Plan Layout An important factor in achieving integrity is the proper plan layout of walls (and columns). In bearing-wall structures there should be an arrangement of interior longitudinal walls to support and reduce the span of long sections of crosswall, thus enhancing the stability of individual walls and of the structures as a whole. In the case of local failure this will also decrease the length of wall likely to be affected. 2. Sufficient continuity In order to achieve an appropriate level of continuity, an integrated system of ties among the principal elements of the structural system must be provided. These ties may be designed specifically as components of secondary loadcarrying systems, which often must sustain very large deformations during catastrophic events. 3. Changing Span Directions of the Floor Slab This refers to where a floor slab is reinforced in order that it can, with a low safety factor, span in another direction. If a load-bearing wall is removed, the collapse of the slab will be prevented and the debris loading of other parts of the structure will be minimized. Often, shrinkage and temperature steel will be enough to enable the slab to span in a new direction.

Asher Gehl 9913569

228

Progressive Collapse 4. Load-Bearing Interior Partitions. The interior walls must be capable of carrying enough load, in order to achieve the change of span direction in the floor slabs. 5. Catenary Action of Floor Slab Where the slab cannot change span direction, the span will increase if an intermediate supporting wall is removed. In this case, if there is enough reinforcement throughout the slab and enough continuity and restraint, the slab may be capable of carrying the loads by catenary action, though very large deflections will result. 6. Beam Action of Walls Walls may be assumed to be capable of spanning an opening if sufficient tying steel at the top and bottom of the walls allows them to act as the web of a beam with the slabs above and below acting as flanges. 7. Redundant Structural Systems The purpose is primarily to provide a secondary load path in the case of collapse. An example provided is an upper-level truss or transfer girder system that allows the lower floors of a multistorey building to hang from the upper floors in an emergency that allows framing to survive the removal of key support elements. 8. Ductile Detailing The recommendation is to avoid low-ductility detailing in elements that might be subject to dynamic loads or very large distortions during localized failures. For instance, consider the implications of shear failures in beams or supported slabs under the influence of building weights falling from above.

Asher Gehl 9913569

229

Progressive Collapse 9. Load Reversal Provide additional reinforcement to resist blast and load reversal when blast loads are considered in design. 10. Specialised Construction Consider the use of compartmentalised construction in combination with special moment resisting frames in the design of new buildings when considering blast protection.
8.4.3 Building Code Requirements for Structural Concrete (ACI 318-02) and Commentary (ACI 318R-02) by the American Concrete Institute

The American Concrete Institute code for design of reinforced concrete structures (American Concrete Institute, 2002) includes extensive Requirements for structural integrity in the Chapter on reinforcing steel details. The commentary to section 7.13 states:

Experience has shown that the overall integrity of a structure can be substantially enhanced by minor changes in detailing of reinforcement. It is the intent of this section [of the code] to improve the redundancy and ductility in structures so that in the event of damage to a major supporting element or an abnormal loading event, the resulting damage may be confined to a relatively small area and the structure will have a better chance to maintain overall stability.
Although the Commentary states that it is the intent of this section to improve

redundancy there is no explicit mention of redundancy or alternate load paths in


the code. The code does however address the detailing of reinforcement and connections, to effectively tie together the members of a structure to improve

integrity of the overall structure. The requirements address continuation of


reinforcement through supports, the location and nature of splicing, and requirements for hooks at terminations (Dusenberry, 2002).

Asher Gehl 9913569

230

Progressive Collapse The Code provisions include a general statement that:

In the detailing of reinforcement and connections, members of a structure shall be effectively tied together to improve integrity of the overall structure.
There are many specific prescriptive requirements for continuity of reinforcing steel and interconnection of components. There are additional requirements for the tying together of precast structural components. In section 7.13, the code requires transverse, longitudinal, and vertical tension ties around the perimeter of the structure. Section 16 describes details for the required ties, and prohibits use of connections that rely sole on friction from gravity load. None of the ACI provisions are threat-specific in any way (Dusenberry, 2002).
8.4.4 GSA PBS Facilities Standards 2000

The 2000 edition of the GSAs Facilities Standards for the Public Buildings Service (GSA, 2000) included the following statement under the Progressive Collapse heading in the Structural Considerations section:

The structure must be able to sustain local damage without destabilizing the whole structure. The failure of a beam, slab, or column shall not result in failure of the structural system below, above, or in adjacent bays. In the case of column failure, damage in the beams and girders above the column shall be limited to large deflections. Collapse of floors or roofs must not be permitted.
This is an absolute and obvious requirement for one-member (beam, slab, or column) redundancy, unrelated to the overall degree of vulnerability of the member or the level of threat to the structure.

Asher Gehl 9913569

231

Progressive Collapse
8.4.5 GSA PBS Facilities Standards 2003

The 2003 edition of the GSAs Facilities Standards for the Public Buildings Service (GSA, 2003a) retained the Progressive Collapse heading from the 2000 edition, but replaced all of the words reproduced above with this short statement: Refer to Chapter 8: Security Design. The structural provisions in Chapter 8 apply only to buildings deemed to be at risk of blast attack. For such buildings, the Chapter provides general performance guidelines and references to various technical manuals for study of blast effects. This represents a complete change of approach from the 2000 version of the same document.
8.4.6 GSA Progressive Collapse Guidelines 2003

The GSA Progressive Collapse Analysis and Design Guidelines for New Federal Office Buildings and Major Modernization Projects (GSA, 2003b) aim to: reduce the potential for progressive collapse regardless of the required level of

protection determined in the facility specific risk assessment


The guidelines begin with determining whether a building is exempt from progressive collapse considerations. Exemption is based on the occupancy, type and size of the structure (for instance, any building of over ten stories is non-exempt) and is unrelated to the level of threat. A flow-chart methodology is provided for this assessment.

Asher Gehl 9913569

232

Progressive Collapse

Figure 8.5 Overall flowchart for Progressive Collapse consideration (GSA, 2000)
Further, the approach to achieving general structural integrity is through a performance-based design.

This guideline presents the methodology and performance criteria for these determinations without prescribing the exact manner of design or analyses. As such, the architect/engineer may apply methods appropriate to the facility at hand.

Asher Gehl 9913569

233

Progressive Collapse This guideline suggests that sophisticated analyses (e.g., non-linear dynamic finite element analyses, linear dynamic finite element analyses, etc.) may be used to determine the potential for progressive collapse, but states that such analyses are complex, costly, and sensitive to small changes in assumptions. However, to facilitate decisions about survivability, this reference reproduces a table of maximum allowable ductility and/or rotation limits for many structural components of various construction types. This table is originally published in a reference by the Department of Defence (2002). The design approach is a simplified analysis that intends to determine the potential for progressive collapse and is not capable of predicting the detailed response of the building under a sudden removal of a primary structural member. For Buildings that are deemed to be non-exempt in section 3, the analysis and design for resistance to progressive collapse is separated in section 4, according to the two major subsections: (4.1) New construction Section 4.1 provides an analysis/redesign approach for reducing the potential of progressive collapse in newly constructed facilities. In the initial design phase for a building, structural features that are encouraged include the use of redundant lateral and vertical force resisting systems, ductile structural elements and detailing, as well as designing to resist load reversals, and prevention of shear failure. Guidance regarding material properties, modelling, and redesign of structural elements is also included. (4.2) Existing construction Section 4.2 provides an outline for assessing the potential of progressive collapse in existing structures and utilising the results into project-specific risk assessment. Guidelines for analysis techniques, criteria and considerations, material properties, and modelling are provided.

Asher Gehl 9913569

234

Progressive Collapse For both new and existing construction, guidance is given for the analyses of typical and atypical structural systems. Considerable detail is provided regarding the features of the analysis and the acceptance criteria. For typical systems, this reference recommends linear elastic, static analyses for tolerating the instantaneous removal of the following first-floor structural elements without collapse: 1. An exterior column near the centre of the short side of the structure, 2. An exterior column near the centre of the long side of the structure, 3. An exterior column at a corner, 4. An interior column, 5. An exterior bearing wall near the centre of the short side of the structure, 6. An exterior bearing wall near the centre of the long side of the structure, 7. An exterior bearing wall that wraps around a corner, and 8. An interior-bearing wall. In these analyses, the length of wall assumed to be removed in each case is the width of a bay or 30 feet whichever is smaller. According to this reference, atypical structures have features such as: 1. Combination structural systems 2. Vertical discontinuities 3. Variations in bay size 4. Extreme bay sizes 5. Plan irregularities 6. Closely spaced columns.

Asher Gehl 9913569

235

Progressive Collapse The 2003 GSA guidelines suggest the smaller of the following options as acceptable criteria of limited collapse for atypical structures: 1. The structural bays directly associated with the instantaneous removal of a vertical support member 2. In the case of an exterior element, 1,800 ft2 of floor area on the floor directly above the removed element 3. In the case of an interior element, 3,600 ft2 of floor area on the floor directly above the removed element. The general GSA security criterion defines an appropriate limit to the level of allowed failure. Essentially the structural collapse must be limited to the two bays that are associated with the failed structural member. This is illustrated in Figure 8.6 below.

Figure 8.6 Definition of Allowable Collapse Area and Progressive Collapse (GSA 2000)
The GSA defines Progressive collapse as any further collapse of the structure past this specific limitation. Asher Gehl 9913569 236

Progressive Collapse The actual potential for progressive collapse is determined by the calculation of a Demand-Capacity Ratio (DCR) for each primary and secondary structural element. The DCR for each primary and secondary structural element is determined as:

DCR =QUD / QCE


Where

(8-1)

QUD = acting force determined by linear elastic, static analysis in the element or
connection (moment, axial force, shear, and possible combined forces) and

QUC = expected ultimate, unfactored capacity of the component and/or connection.


In the calculation of QCE, the engineer may include strength increase factors to account for the rapid application of load when the engineer has confidence about the material strengths. The allowable strength increase factors are 1.05 for structural steel and 1.25 for reinforced concrete, concrete or clay tile masonry, and wood and light metal framing. According to this reference, facilities will be judged to have high potential for progressive collapse if any primary or secondary structural element outside the allowed collapse area has the following: DCR > 2.0 for typical structures or DCR > 1.5 for atypical structures. In some ways, these guidelines appear to be similar to the GSAs PBS Facilities Standards of 2000 in that their central provision is a requirement for one-member redundancy, unrelated to the degree of vulnerability of the member or the level of threat to the structure
8.4.7 The UK Building Regulations 1991 (PART A)

Following the partial collapse of Ronan Point in 1968, the Building Regulations were changed to deal specifically with damage due to an accident such as a gas explosion for buildings of five storeys or more.

Asher Gehl 9913569

237

Progressive Collapse The supplement to the UK building regulations recognised the need to reduce the potential for progressive collapse. The building regulations state:

The Building shall be constructed so that in the event of an accident, the building will not suffer collapse to an extent disproportionate to the cause.
The main aim behind the code is for designs to be of a more robust construction. There are two options offered in order to achieve a more robust design. (Moore 2002)

Local resistance:

The code specifies a required static pressure of 34kN/m2 that key structural
P P

elements must be capable of withstanding. The requirement provides for effective horizontal and vertical reinforcement continuity. This increases structural continuity creating a structure with a high degree of redundancy and provides the building with alternative load paths should part of the structure be removed by an accidental action. Subsequent quarter-scale testing has been carried out to determine the effectiveness of this approach in order to avoid a chain of failures.

Alternate load path method:

The code limits the collapse area caused by the removal of an individual supporting structural member as either: 1. 15% of the story area or 2. 70m2
P P

Asher Gehl 9913569

238

Progressive Collapse

Figure 8.7 Allowable damage of Building (Smith, Brokaw, Swatzell, 1998)


The above requirements are considered to produce more robust structures, which are more resistant to disproportionate failure due to various causes, such as impact as well as to gas explosions.

Asher Gehl 9913569

239

Progressive Collapse The accidental loading requirement has now been included in the British Standard, BS6399: Loading for buildings: Part 1: Code of practice for dead and imposed loads' (BSI, 1996) with the following text: -

'When an accidental load is required for a key or protected element approach to design (see appropriate material design Code), that load shall be taken as 34kN/m2.'
The methods for avoiding disproportionate collapse (tying, bridging and key element design) have been including in the various British Standard Material Codes. These include the British Standard for the Structural use of Steelwork in building (BSI, 2000), the British Standard for the Structural use of concrete (BSI, 1985) and the British Standard Structural work of masonry (BSI, 1978). There is however no guidance with respect to the choice of approach and the danger is of an unconservative solution to the design (Moore, 2002).
8.4.8 National Building Code Of Canada

The National Building Code of Canada (National Research Council of Canada, 1995) is one of the codes that have addressed progressive collapse in some form for decades. This code aims to reduce the risk of progressive collapse through the requirement of adequate structural integrity, whereby local failure is allowed without the onset of further collapse (Smith, Brokaw, Swatzell, 1998). The 1995 (National Research Council of Canada 1995) edition of the National Building Code of Canada is less specific, and more general, in its approach to regulating design to prevent progressive collapse when compared to earlier versions since 1975. The methods presented to reduce the risk of progressive collapse include the following:

control of accidental events, designing key elements to resist accidental events, designing adequate ties, providing alternate paths of support, and compartmentalizing the structure to limit the spread of a collapse.

Asher Gehl 9913569

240

Progressive Collapse The specific suggestions and guidelines within these methods however are brief and do not provide guidance or quantification of damage levels in order to appropriately verify the effectiveness of a design that aims to seek alternate load paths to reduce the potential for further progressive collapse (Dusenberry, 2002). The current edition (1995) defines structural integrity as:

the ability of the structure to absorb local failure without widespread collapse.
The following commentary addresses the link between integrity and adequate connection detailing:

Building structures designed in accordance with the CSA design standards will usually possess an adequate degree of structural integrity, generally through detailing requirements for connections between components.
However, the Commentary acknowledges that there are circumstances when additional attention is required:

Situations where structural integrity may require special attention include medium/high rise building systems made of components of different materials, whose interconnection is not covered by existing CSA design standards, buildings outside the scope of existing CSA design standards, and buildings exposed to severe accidental loads such as vehicle impact or explosion.
Section 6.1.2, Structural Integrity, from CAN/CSA-S16-01, Limit States Design of Steel Structures, states:

The general arrangement of the structural system and the connection of its members shall be designed to provide resistance to widespread collapse as a consequence of local failure. The requirements of this Standard generally provide a satisfactory level of structural integrity for steel structures. Supplementary provisions may be required for structures where accidental loads such as vehicle impact or explosion are likely to occur (see Clause 1.3). (Further guidance can be found in Chapter 4, Commentary C, Users Guide NBC 1995 Structural Commentaries (Part 4).
Asher Gehl 9913569 241

Progressive Collapse
8.4.9 Swedish Building Code

The Swedish Building Code is one of the most specific of all the guidelines ensuring that all buildings except single family homes be designed and built in order to reduce the likelihood of progressive collapse from local damage due to an abnormal loading event (Smith et al 1998). The Swedish Building Code presents two options for design in a similar fashion to the UK building regulations. 1. Alternate load path: The requirement is for local damage to be contained and not extended into a further disproportionate collapse. The code specifies a cubic volume to be removed with a cubic length equal to the storey height inclusive of the above and below floor thickness. This volume removes both the primary structural element and all the connected joints. Adequate reinforcement requirements are provided in this approach.

Figure 8.8 Damage Limitations (Smith et al 1998)


2. Local resistance: This approach designs specific structural elements and connections to withstand the abnormal event and thus reduce the risk of progressive collapse.

Asher Gehl 9913569

242

Progressive Collapse The SBC despite being stricter with respect to Alternate load path requirements than the UK building regulations still fails to provide any guidance to verify that adequate alternate load path designs are sufficient to redistribute the load and thus effectively minimise progressive collapse (Smith et al, 1998).

8.5 Further Analysis


Several approaches have been proposed for including progressive collapse resistance in building design. The alternative load path method is one of the more fundamental approaches in the guidelines that aim to mitigate progressive collapse. Essentially the only criterion for failure of an element is the stress or strain limit criteria. These methods are based on static considerations only. The results obtained from a static analysis using the Alternate Load Path method can be unconservative as a result of neglecting inertial effects that play a dominant role following the sudden failure of one or more structural members (Pretlove, Ramsden, Atkins, 1991). According to Hamburger and Whittaker (2004), one of the fundamental flaws in progressive collapse analysis is in effective connections. Further research is needed to develop new connection technologies, for make blast resistant design to become more efficient and accurate in order to minimize the risks of progressive collapse. A paper presented by Kaewkulchai and Williamson (2004) is focused upon developing an analysis tool for studying the dynamic response of planar frame structures subjected to an initiating localised failure. Dynamic effects appear to have significant impact on the response behaviour of the frames. Future research will focus on conducting parametric studies to identify key factors that contribute to the progressive collapse of planar frame structures. Choi and Krauthammer (2003) presented a paper in which they developed an external criteria-screening (ECS) technique to analyse progressive collapse taking into account material and geometric nonlinear effects. The primary purpose of which was to describe a new algorithm applicable to progressive collapse that considers buckling, in addition to the stress/strain failure criteria. The paper concludes that the collapse is initiated earlier when taking into account buckling as an additional failure mode criterion.

Asher Gehl 9913569

243

Progressive Collapse There is an extremely large amount of research that is directed towards improving the analysis methods for buildings subject to high intensity normal loads. The requirement for the future is to successfully relate this research into a series of specific code requirements. This specific code must be substantiated both by the more accurate analytical tools as well as the actual performance testing of buildings constructed to conform to the improved specific standards.

Asher Gehl 9913569

244

Conclusions

Chapter 9 Conclusions
9.1 General conclusions
The main aim of this thesis was to present sufficient knowledge of the behaviour of an external explosion and the induced response of reinforced concrete and structural steel structures subjected to explosive loading. Further to this it was intended to present the design methods that are generally utilised to mitigate the risk of progressive collapse from abnormal loadings upon a structure. The primary aims have been demonstrated through the utilisation of the existent knowledge in the field of blast engineering in order to complete the preliminary flexural design for columns of varying cross sectional shape that are subject to a typical external blast loading environment. In Chapter 2 the design objectives were presented, outlining the difficulties involved with achieving a thoroughly resistant structure and the necessity to accept considerable plastic damage in order to balance the associated costs against the risks. The methods used in defining the threats were presented and the complications involved with achieving appropriate design standards. The importance of standoff distance and fundamental physical security measures were introduced. A number of the more significant case studies were then given to outline the immense power and devastation associated with explosions. Chapter 3 emphasised the unique attributes of the explosive load, and outlined the overall design process that is required to encompass the unique characteristics associated with an explosive event. The Chapter effectively differentiates between the localised design method that aims to increase the resistance of the local element, and the global design procedures that aim to reduce the risk of progressive collapse and improve the redundancy of a structure. Chapter 4 the theory of blast waves and their associated parameters was presented to describe the nature of the detonation process and the resultant explosion.

Asher Gehl 9913569

245

Conclusions Chapter 5 presented the derivations of the explosive load that is applied upon each face of the structure and the summation of the load on the entire building. The loading utilised for the column designs of Chapter 7 was derived according to the theory presented in previous chapters. In Chapter 6 it was shown that the resultant load from the blast wave can be applied to structures in a practical manner for design purposes. The analysis utilises the fact that only the elements peak deformation is required and thus may be represented by an equivalent Single Degree of Freedom system. For the design, large inelastic deformations are associated with the structures response to the transient dynamic load. To determine the required procedure it first must be established whether the element is expected to respond to pressure or to impulse based on the relationship between the load duration and the fundamental period of the SDOF equivalent system. Limitations are placed upon the final deformation of the element based on the desired degree of damage. The fundamental response calculations are based on the equations for conservation of mass, momentum and energy and takes into account the fact that the concrete and steel material behaviour changes under dynamic loading as a result of strain rate effects. For concrete the ultimate strength in tension and compression will increase and the corresponding strain capacity is extended. Different failures occur with dynamic loading when compared to a statically applied load. Steel elements exhibit significantly higher stresses with increasing strain rate.

Asher Gehl 9913569

246

Conclusions In Chapter 7 the direct design process for local elements was presented for both reinforced concrete and structural steel columns. The aim of the procedure was essentially to demonstrate an understanding of the process and to examine the relationship between cross sectional shape and the elements response. It was stated that for such high velocities the drag coefficients are a function of Reynolds number and the coefficients magnitude decreases with higher velocities. As expected it was seen that the overall load upon the element reduces with smaller drag coefficients. The concrete section was successfully designed according to protection category 1, whereas for the steel section minimum plastic section modulus was presented for various Drag coefficient values, although the overall design was unsuccessful. The overall effect however, is such that a considerably lower load will affect a column section with a significantly lower Drag coefficient and thus smaller sections are possible. It is therefore recommended that column shape be a controlling factor in the design process. In particular, given that the magnitude of the load is considerably higher for columns at lower levels, it is highly recommended wherever possible to choose sections with lower drag coefficients, in order to reduce risks of failure and damage as well as minimising the required size of the section. Chapter 8 presented the principal design techniques that are used by the major design guidelines to mitigate the risk of a Progressive collapse. The benefits of the Alternate load path system were compared to the local resistance increase method and the overall importance of continuity and redundancy was emphasised. The major Progressive collapse guidelines were discussed and the methods used compared. The Chapter concluded with presenting areas of future research emphasising the overall importance of successfully limiting the collapse area in the event of a local failure. It was shown that there is a great deal of work required by code writing bodies to effectively detail the required design and analysis procedures to mitigate the risks of progressive collapse. It is however a critical aspect of engineering that requires a great deal of commitment in the future.

Asher Gehl 9913569

247

Conclusions The design of buildings against an explosive load is essentially a delicate balancing act that must weigh up the costs involved in protecting the structure against the devastating risks of attack. The message must be imparted that a considerable improvement in the buildings resistance is possible for minimal cost by taking into account fundamental physical security concepts to increase standoff distance. In addition to a secure site, utilising shapes with low Drag coefficients can significantly reduce the load on key structural elements. The overall response of a structure can be significantly improved by ensuring a robust design that provides sufficient load transfer capabilities in the event of a localised collapse, in order to mitigate the risk of progressive collapse from local damage.

9.2 Future Research


There is a vast amount of information in the field of blast engineering that has not been presented in this thesis. The objective of this section is not to list the different areas that have been omitted in this work, but rather to describe the key areas that have been described somewhat within the thesis that require further improvement in order to increase the levels of protection for the many civilian structures that are at risk.

Asher Gehl 9913569

248

Conclusions
Faade Protection

Although technically the faade of a structure is not a major structural component, in an explosion it is one of the most likely areas to fail and cause widespread damage. The high pressures in an explosion will shatter the glass and if not properly protected, this glass can be lethal over a wide area and the resultant pressure leakage into the interior of the building can cause further structural damage. Also of significance is the large cost involved in repairing the faade of a building after an explosive event. There are a number of companies throughout the world (Weidlinger Associates, MYY Limited) that specialise in faade protection, from basic glazing and blast curtains to high strength window frames that are connected to floor slabs to absorb the shock from the blast and mitigate the damage associated with faade damage. One such concept makes use out of the flexibility and capacity of an advanced window material to absorb and dissipate large amounts of blast energy, whilst preventing any debris from entering the confinements of the building. The catch system Cable Protected Window Systems (CPWS)) work in such a way that as the glass is damaged it bears against a cable catch system, which in turn deforms the window frames. Extensive explosive testing, as well as sophisticated computer simulations, has demonstrated the effectiveness of these systems.

Figure 9.1 Energy Absorbing Catch System (Smilowitz, 2003)

Asher Gehl 9913569

249

Conclusions
National Progressive Collapse Standards

Of utmost importance is the development of National Standards to describe the fundamental design principles required to improve a structures overall ability to withstand the loading effects of an explosion. These standards must address both local design and global design concerns in a thorough and specific manner giving the designer explicit steps that must be performed to check and ensure the design is satisfactory for the particular building in the blast loading environment. There is considerable effort overseas aimed at achieving this goal. On July 10-12, 2002 the Multi hazard Mitigation Council (MMC) held a National Workshop on Prevention of Progressive Collapse in Illinois, United States. A number of papers reviewed the various codes and standards, structural systems and analytical tools related to progressive collapse. A number of these papers have been referenced throughout Chapter 8. The workshop aimed to serve as the basis for an action plan to integrate progressive collapse prevention into standard design practice and relevant building codes and standards. However, A number of fundamental changes need to be made however in order to be successful in achieving this goal. There needs to be an educational change within the entire industry. No longer is it acceptable for designs to meet the minimum requirements in codes and standards. In the event of a collapse however, in order to escape the issue of negligence in-tort, professional engineers are required to demonstrate that at the conceptual design stage all reasonably foreseeable threats were considered. With the current global climate it is reasonable to conclude that the risk of explosions is greater than ever before. However it has been shown that the elimination of the risk of progressive collapse in its entirety is not possible. It is therefore essential for all engineers as part of the required design documentation to demonstrate that appropriate steps have been taken to minimise the risk of progressive collapse to an acceptable level. Further to this requirement it is essential for engineers as professionals to communicate the concerns over the devastating effects of the threats to building developers, architects, owners and occupants.

Asher Gehl 9913569

250

Conclusions
Temperature effects

It was determined that the load was significantly reduced when Drag Coefficients and Reynolds number was taken into account. One aspect of the blast wave that was not considered within this thesis is the large amount of energy released as high temperatures. To take into account temperature effects in a thorough analytical study would be of great interest.
Advanced Computational approaches

As discussed briefly in Chapter 6, there are a variety of alternative methods to calculate the response of a structure to an explosive load. Significant advances in computational power have enabled sophisticated software to utilise highly advanced Computational Fluid dynamics methods to describe complicated blast wave reflections in a complex structure. Further developments have also allowed the use of Finite Element Modelling for global analyses determining structural integrity with respect to Progressive Collapse. Terrorist attacks although generally directed against a single target building, have other damaging consequences. In a built up city environment, the proximity of buildings, act to amplify the pressures from the blast wave, and the damaging effects can be seen for a large distance. There are a number of concerns regarding the analysis techniques utilised to examine the complex reflections and interactions associated with the propagation of the Blast wave. Fairlie (1997) presented a paper comparing the numerical methods available for assessing the interaction of high explosive air blast within the complex geometries of a typical congested urban environment. The conventional analytical used by Conwep (Hyde, 1991) and Kinney and Graham (1985), were compared with a more advanced analysis utilising 3D geometries from the Finite Element program AUTODYN3D (Century dynamics, 1997). It was shown that for simple geometries a 2D analysis is sufficiently accurate; however for complex geometries a full 3D analysis is required in order not to dangerously underestimate the prediction of blast wave parameters.

Asher Gehl 9913569

251

Conclusions Thus in order to achieve a more thorough and accurate analysis of the blast wave and the loadings imparted on structures, considerably higher accuracy programs are required to predict its effects.

Asher Gehl 9913569

252

Bibliography

Bibliography
1. 7-45S: Process Control Houses and Other Structures Subject to External Explosion Damage Factory Mutual Engineering Corporation Boston MA, 1980. 2. ACI 318-02/318R-02 Building Code Requirements for Structural Concrete and Commentary. 3. Allen D.E and Schriever W.R. (1972). Progressive collapse, Abnormal loads and Building codes. Division of Building Research, National Research council, 4. Armstrong, B.J, Rickman, D.D, Baylot, J.T. and Bevins, T.L. (2002). Code Validation Studies for Blast in Urban Terrain. U.S. Army Engineer Research and Development Centre. Presented at the 12th Annual DoD High Performance Computing Modernization Program Users Group Conference, Austin, Texas. 5. AS 3600 (2001). Concrete Structures. Sydney, Australia, Australian Standards. 6. AS/NZS 1170.1 (2002). Structural Design Actions Permanent, imposed and other actions. Sydney, Australia, Australian Standards. 7. ASCE 7-02. American Society of Civil Engineers Minimum Design Loads for Buildings and Other Structures. American Society of Civil Engineers, New York. 8. ASCE 7-95. American Society of Civil Engineers Minimum Design Loads for Buildings and Other Structures. American Society of Civil Engineers, New York. 9. At Blast v 2.0. Applied Research Associates Inc. 10. Baker, W. E. (1983) Explosions in Air, 2nd edition. Univ. of Texas Press, Austin, pp 268. 11. Bangash, M. (1993). Impact and Explosion: Analysis and Design. CRC Press, Florida. 12. Beshara, F.B.A. (1994). Modelling of blast loading on aboveground structures I. General phenomenology and external blast. Computers & Structures 51 (5) 585 596. 13. Beshara, F.B.A. (1994). Modelling of blast loading on aboveground structures II. Internal blast and ground shock. Computers & Structures 51 (5) 597 606. 14. Biggs, J. M. (1964). Introduction to Structural Dynamics. New York McGrawHill.

Asher Gehl 9913569

253

Bibliography 15. Bischoff, P. H. and Perry S. H. (1991). Compressive behaviour of concrete at high strain rates Materials and Structures. 24, 425-450. 16. Bounds, W, Nene M, Ko H. (1998). Industrial Blast Resistant Facilities: An Engineers Constructors Viewpoint. Farmington Hills, Michigan. 17. Bounds, W. (ed) (1998). Concrete and Blast Effects. SP-175 American Concrete Institute, Farmington Hills, Michigan. 18. British Standards Institute (1978). BS5628: Code of Practice for use of Masonry; Part 1: Structural use of unreinforced masonry. London 1978. 19. British Standards Institute (1985). BS8110: Structural use of Concrete: Part 1; Code of Practice for design and construction. London, 1985. 20. British Standards Institute (1996). BS6399: Loading for Buildings; Part 1: Code of Practice for dead and imposed loads. London, 1996. 21. British Standards Institute (2000). BS5950: Structural use of Steelwork in Building; Part 1: Code of Practice for Design Rolled and Welded Sections. London 2001. 22. Bulson P. S. (1997). Explosive loading of engineering structures. E & FN Spon, London, pp 233. 23. Burns, J., Abruzzo, J., Tamaro, M. (2002). Structural Systems for Progressive Collapse Prevention. Multi Hazard Mitigation Council: Workshop on Prevention of Progressive Collapse, Illinois, U.S. Thornton-Thomasetti Engineers. 24. Choi H. J and Krauthammer T. (2003). Investigation of Progressive Collapse Phenomena in a Multi Story Building. Protective Technology Center, The Pennsylvania State University, USA. Presented at 11th International Symposium on the Interaction of the Effects of Munitions Effects with Structures. Mannheim, Germany. 25. Conrath, E.J.; Krauthammer, T.; Marchand, K.A.; and Mlakar, P. F., (1999) 2nd Edition. Structural Design for Physical Security: State of the Practice, Structural Engineering Institute of American Society of Civil Engineers, Reston, VA. 26. Corley Dr W. G, Mlakar P, Sozen, M, Thornton, C. (1998). The Oklahoma City Bombing: Summary and Recommendations for Multihazard Mitigation. Journal of Performance Constructed Facilities. 12, 100-112.

Asher Gehl 9913569

254

Bibliography 27. Corley DR W. G. (1998). Testimony of Dr W Gene Corley, P.E., S.E. on Security in Federal Buildings on Behalf of the American Society of Civil Engineers, before the Public Buildings and Economic Development Subcommittee of the House Committee on transportation and infrastructure. United States House of Representatives. 28. Department of Civil and Environmental Engineering (2004). Louca Dr L, Boh, J. Analysis and Design of Profiled Blast Walls. London, Health and Safety Executive. 29. Department of Defence (2002). DoD Minimum Antiterrorism Standards for Buildings. Distribution Statement A: Approved for public release. 30. Designers Notebook Blast Consideration at http://www.pci.org/pdf/DN/DN_blast1.pdf. 31. Dewey, J. and van Netten, A. Height-of-Burst Curves Revisited. Dewey McMillin & Associates Ltd, Canada. 32. Domire, J. (2003). Silver Spring District Courthouse, Structural Option. Senior Thesis, Architectural Engineering, Penn State University. 33. Dusenberry, D, O. (2002). Review of Existing Guidelines and Provisions Related to Progressive Collapse. Multi Hazard Mitigation Council: Workshop on Prevention of Progressive Collapse, Illinois, U.S. Simpson Gumpertz & Heger Inc. Arlington. 34. Ellingwood, Prof, B. (2002). Load and Resistance Factor Criteria for Progressive Collapse Design. School of Civil and Environmental Engineering. Georgia Institute of Technology. Multi Hazard Mitigation Council of Progressive Collapse, Illinois, U.S. 35. Ettouney, M., Smilowitz, R., Rittenhouse, T. (1996). Blast Resistant Design of Commercial Buildings. Practice Periodical on Structural Design and Construction. 1 (1). 36. Fairlie, G. (1997). Efficient Analysis of High Explosive Air Blast in Complex Urban Geometries Using the AUTODYN-2D & 3D Hydrocodes, Analytical and Experimental Methods. 15th International Symposium on Military Aspects of Blast and Shock, Banff Canada.

Asher Gehl 9913569

255

Bibliography 37. FEMA 426, (2003). Reference Manual to Mitigate Potential Terrorist Attacks Against Buildings. Washington, Federal Emergency Management Agency, US Department of Homeland Security. 38. FEMA 428, (2003). Primer to design safe school projects in case of terrorist attacks. Washington, Federal Emergency Management Agency, US Department of Homeland Security. 39. Forbes, D. (1999). Blast Loading on Petrochemical Buildings. Journal of Energy Engineering. December 125 (3). 40. Fu, H.C., Erki, M.A. & Seckin, M. (1992). Review of Effects of Loading Rate on Reinforced Concrete. Journal of the Structural Engineering ASCE, 117 (12) 3660-3679. 41. Gilmour, J and Virdi, K. (1998). Numerical Modelling of the Progressive Collapse of Framed Structures as a Result of Impact or Explosion. 2nd Int. PhD Symposium in Civil Engineering, Budapest. 42. Goschy, Dr B. (1990). Design of Buildings to Withstand Abnormal Loading. London, Butterworths. 43. Gould, P and Abu-Sitta, S. (1980). Dynamic Response of structures to wind and earthquake loading. Devon, Pentech Press. 44. Gustavon, R. (2000). Static and Dynamic Finite Element Analyses of Concrete Sleepers. Thesis for the Degree of Licentiate of Engineering. Division of Concrete Structures, Department of Engineering, Chalmers University of Technology, Gteborg, Sweden. 45. Hamburger, R.O., Whittaker, A., Comartin, C., Bachman, R., Rojahn, C., and Mahoney, M. (2003). Performance-Based Engineering of Buildings and Infrastructure for Extreme Loadings, Proceedings of the 74th Annual Shock and Vibration Symposium, San Diego, CA, Oct. 46. Harris, E (1956). Exact and Approximate Treatments of the One-Dimensional Blast Wave. Department of the Navy, Office of Naval Research. 47. Hjelmgren J. (2002). Dynamic Measurement of Pressure - A Literature Survey. SP Swedish National Testing and Research Institute. SP REPORT 2002 34.

Asher Gehl 9913569

256

Bibliography 48. HMSO (1968) Report of the inquiry into the Collapse of Flats at Ronan Point, Canning Town, London. 49. Hogan, M - OneSteel (1996). Hot Rolled and Structural Steel Products. (1998 edition). Taken from http://www.civil.usyd.edu.au/courses/civl3206/OneSteel_hot_rolled_properties.xl s, DyMaK, Australia. 50. Hyde, D. (1991). ConWep Application of TM 5-855-1. U.S. Army Engineers Waterways Experiment Station, Vicksburg. 51. Johansson, M. (1999). Non-Linear Finite Element analyses of civil defence shelter subjected to explosion load or collapse load. Division of Concrete Structures. Chalmers Institute of Technology, Gtenburg, Sweden, Report 99:8. 52. Kaewkulchai, G and Williamson, E. (2003). Dynamic Behavior of Planar Frames during Progressive Collapse. 16th ASCE Engineering Mechanics Conference, July 2003. University of Washington, Seattle. 53. Kingery, C and Bulmash, G. (1984). Airblast Parameters from TNT Spherical Air Burst and Hemispherical Surface Burst. Technical Report ARBRL-TR-02555, U.S Army Armament Research and Development Centre. 54. Kinney, G and Graham, K. (1985). Explosive Shocks in Air 2nd Edition. New York, Springer Verlag. 55. Krauthammer, Prof. T. (1999). Explosion Damage Assessment. First Structural Forensic Engineering Seminar on Structural Failure Investigations. Penn State University, Presentation at University of Toronto, January 1999. 56. Krauthammer, Prof. T. (2000). Modern Protective Structures, Design, Analysis and Evaluation. Course notes, The Pennsylvania State University. 358 pp. 57. Krauthammer, Prof. T. (2004). A Technical Overview of Protective Design Protecting Structures from Blast, Shock, and Impact Loads. Unpublished Protective Technology Centre, Penn State University. 58. Krauthammer, Prof. T., Frye, M., Schoedel, T.R and Seltzer, M. (2003). Advanced SDOF Approach for Structural Concrete Systems under Blast and Impact Loads. Presented at 11th International Symposium on the Interaction of the Effects of Munitions Effects with Structures, Mannheim, Germany.

Asher Gehl 9913569

257

Bibliography 59. Kulkarni, S.M and Shah, S.P (1998). Response of reinforced concrete beams at high strain rates. ACI Structural Journal November December, 1998, Title 95S64, No. 6. 60. Leppanen J (2002). Dynamic Behaviour of Concrete Structures subjected to Blast and Fragment Impacts, Department of Structural Engineering Concrete Structures, Chalmers University of Technology, Goteborg, Sweden. 61. Longinow, A and Alfawakhiri, F. (2003). Blast Resistant Design with Structural Steel Common Questions Answered. Modern Steel Construction, October. 62. Lucioni, B.M., Ambrosini, R. D, Danesi, R.F. (2004). Analysis of building collapse under blast loads. Engineering Structures 26 63 71. 63. Malvar L. and Crawford J. (1998). Dynamic Increase Factors for Concrete. 28th DDESB Seminar, Orlando Florida. 64. Malvar L. and Crawford J. (1998). Dynamic Increase Factors Steel Reinforcing Bars. 28th DDESB Seminar, Orlando Florida. 65. Malvar, L., Ross, C.A., (1997). Review of Static and Dynamic Properties of Concrete in Tension, Accepted for publication, ACI Materials Journal, December. 66. Mathcad 8 (1998). Mathcad 8 Users Guide. Cambridge, MathSoft Inc. 67. Mays, G.S. and Smith, P.D (eds) (1995). Blast Effects on Buildings: Design of Buildings to Optimize Resistance to Blast Loading. London, Thomas Telford Publications. 68. Meadearis, K. (1975). A comparative study of Structural Response to ExplosionInduced Ground Motions. Prepared for the ASCE Research Council on Performance of Structures. New York, American Society of Civil Engineers. 69. Mendis, P. and Ngo, T. (2003). Vulnerability Assessment of Concrete Tall Buildings Subjected to Extreme Loading Conditions. Proceedings of the CIBCTBUH International Conference on Tall Buildings, Malaysia. 70. Miyamoto, H.K. and Taylor, D. (1999). Structural Control of Dynamic Blast Loading using Passive Energy and Dissipators. SEAOC Convention 1999.

Asher Gehl 9913569

258

Bibliography 71. Mlakar, P., Corley, W. G., Sozen, M., Thornton, C. (1997). Blast Loading and response of Murrah Building. Forensic Engineering, Proceedings of the Congress. p. 36-43. 72. Moore, Dr, D. (2002). The UK and European Regulations for Accidental Actions. 16th ASCE Engineering Mechanics Conference, July 2003. University of Washington, Seattle. 73. Mutsuyoshi, H. and Machida, A. (1984). Properties and failure of reinforced concrete members subject to dynamic loading. Trans., Japanese Concrete Institute. No 6, 1984. 74. National Building Code of Canada (1975). National Research Council of Canada, Ottawa, Ontario. 75. National Building Code of Canada (1995). Part 4 and Commentary C, 1995 edition. National Research Council of Canada, Ottawa, Ontario. 76. NATO (FM 8-9) Handbook on the Medical Aspects of NBC Defensive Operations (1996). Departments of the Army, the Navy and the Air Force. Army Field Manual 8-9, AMEDp-6(B) Navy Medical Publication 5059 Air Force Joint Manual 44-151, Washington DC, 1 February 1996. 77. Ngo T.D, Mendis P.A., Teo D. Kusuma G. (2003) Behavior of high-strength concrete columns subjected to blast loading, University of Melbourne Australia. 78. Norris, C., Hansen, R., Holley, M Jr., Biggs, J., Namyet, S., Minami, J. (1959). Structural Design for Dynamic Loads. New York, McGraw Hill. 79. Prendergast, J. (1995). Oklahoma City Aftermath. Civil Engineering, October 1995. 80. Pretlove, A.J., Ramsden, M and Atkins, A. G. (1991). Dynamic Effects in Progressive Failure of Structures, International Journal of Impact Engineering. 11 (4), 539-546. 81. Remnikov, Dr A. (2003). The HSBC Bank Building Bombing: Analysis of Blast Loading. School of Civil, Mining and Environmental Engineering, University of Wollongong.

Asher Gehl 9913569

259

Bibliography 82. Remnikov, Dr A. (2004). Evaluation of Blast Loads on Commercial Buildings: From Hand Calculations to GIS-Based Numerical Simulations. University of Wollongong, Research Network for Engineering a Secure Australia - Second National Engineering & Security Research Forum, Melbourne. 83. Ross C. A., Tedesco J. W., Kuennen S. T. (1995). Effects of strain rate on concrete strength. ACI Materials Journal, January-February, Title: 92-M5, 3747. 84. Ross, C. A., Jerome D. M., Tedesco J. W., Hughes M. L. (1996). Moisture and strain rate effects on concrete strength. ACI Materials Journal, May-June, Title 93-M33, No. 3, pp. 293-300. 85. Rossi, P. (1994). Dynamic behaviour of concrete: from the material to the structure. Materials and Structures, 27, 319-323. 86. Safety Guide SG-22 (1978). Siting and Construction of New Control Houses for Chemical Manufacturing Plants. The Manufacturing Chemists Association, Washington DC. 87. Schmidt, J. (2003). Structural Design for External Terrorist Bomb Attacks. Structure, March pp. 1 5. 88. Shankar Nair, R. (Ph.D., P.E, S.E.) (2004) Progressive Collapse Basics. Modern Steel Construction. March 2004. 89. Shipe, J. A. and Carter, C.J. (2003). Defensive Design Modern Steel Construction, November. 90. Smilowitz, R. (2002). Analytical Tools for Progressive Collapse Analysis. Multi Hazard Mitigation Council: Workshop on Prevention of Progressive Collapse, Illinois, U.S. 91. Smilowitz, R. (2003). Designing Buildings to Resist Explosive Threats. Weidlinger Associates, Whole Building Design Guide. Multi Hazard Mitigation Council: Workshop on Prevention of Progressive Collapse, Illinois, U.S. 92. Smilowitz, R. (2003). Retrofitting Existing Buildings to Resist Explosive Threats. Weidlinger Associates, Whole Building Design Guide. Multi Hazard Mitigation Council: Workshop on Prevention of Progressive Collapse, Illinois, U.S.

Asher Gehl 9913569

260

Bibliography 93. Smith, J, Brokaw, J and Swatzell, S. (1998). Progressive Collapse Analysis and Design Guidance. Washington, General Services Administration. 94. Smith, P. (2003). An Introduction to Blast Loading, Structural Response and Design. One day course. University of Melbourne, Department of Civil and Environmental Engineering. 95. Smith, P.D. and Hetherington, J.G. (1994). Blast and Ballistic Loading of Structures. Butterworth-Heinemann. 96. Smits, A.J. (2004). Aerodynamics of Bicycles. Lecture Notes, School of Engineering and Applied Science, Aerospace Engineering, Princeton University, New Jersey. 97. Soroushian, P., Choi, K. and Alhamad, A. (1986). Dynamic constitutive behaviour of concrete. ACI Structural Journal. 83 (2) pp. 251-258. 98. Takeda, J, Tachikawa H. and Fujimoto K. (1974). Mechanical behavior of concrete under higher rate loading than in static test. Proc. Mechanical Behaviour of Materials, p. 479-486. 99. Teo, D. (2003). Assessment of Tall Concrete Buildings Subjected to Impact and Blast Loading. Unpublished Master Thesis, Department of Civil and Environmental Engineering, University of Melbourne. 100. The Institute of Structural Engineers (1995). The Structural Engineers Response to Explosive Damage London ETO, Ltd. 101. The Multihazard Mitigation Council of the National Institute if Building Sciences. Prevention of Progressive Collapse: Report on the July 2002 National Workshop and Recommendations for Future Efforts. 102. Thomson, W. (1983). Theory of Vibration with Applications, 4th Ed, California, Nelson Thornes. 103. TM 5-1300 (1991). Structures to Resist the Effects of Accidental Explosions U.S. Army Technical Manual. US Department of Army and Air Force, USA. 104. TM 5-855-1 (ConWep) (1992). Collection of conventional weapons effects calculations based on, Fundamentals of Protective Design for Conventional Weapons, U.S. Army Engineer Waterways Experiment Station, Vicksburg, USA.

Asher Gehl 9913569

261

Bibliography 105. TM5-855-1 (1991). Fundamentals of Protective Design (Non-nuclear). US Army Technical Manual. U.S Department the Army, USA. 106. U.S Air Force (1964). Project 1080 Research Summary. Kirtland Air Force Base, New Mexico. 107. U.S General Services Administration (2000). Facilities Standards for the Public Building Services. 108. U.S General Services Administration (2000). Progressive Collapse Analysis and Design Guidelines for New Federal Office Buildings and Major Modernization Projects. 109. U.S General Services Administration (2001). GSA Security Reference Manual: Part 3 Blast Design and Assessment Guidelines. 110. U.S General Services Administration (2003a). Facilities Standards for the Public Building Services. 111. U.S General Services Administration (2003b). Progressive Collapse Analysis and Design Guidelines for New Federal Office Buildings and Major Modernization Projects. 112. Wang, N. (1996). Resistance of concrete railroad ties to impact Loading. Department of Civil Engineering, University of British Columbia, Vancouver, Canada. 113. Wilkinson, T. (2004). World Trade Center Some Engineering Aspects. From www.civil.usyd.edu.au/latest/wtc.php#why 2001. 114. www.bbc.com 115. www.blastanalysis.com 116. www.civil.usyd.edu.au/latest/wtc.php#why 2001. 117. www.ict.org.il 118. www.princeton.edu 119. Yerushalmi, P., More, U., Reizes, A. (2002). Design Techniques to Strengthen Soft Buildings Against Acts of Terror and Car Bombs. Session V: Security Technology Focus Group II Security Related Research and Methodology.

Asher Gehl 9913569

262

Appendix A

APPENDIX A
Transformation Factors for One Way Elements
TABLE A1: Transformation factors for One Way Elements

(US Department of the Army Technical Manual, TM5-1300. Design of structures to resist the effect of accidental explosions)

Asher Gehl 9913569

263

Appendix A
TABLE A2: General and Ultimate Deflections for One Way Elements

(US Department of the Army Technical Manual, TM5-1300. Design of structures to resist the effect of accidental explosions)

Asher Gehl 9913569

264

Appendix A

TABLE A3: Elastic, Elasto-Plastic and Equivalent Elastic Stiffness for One Way Elements

(US Department of the Army Technical Manual, TM5-1300. Design of structures to resist the effect of accidental explosions)

Asher Gehl 9913569

265

Appendix A
TABLE A4: Elastic and Elasto-Plastic Unit Resistances for One Way Elements

(US Department of the Army Technical Manual, TM5-1300. Design of structures to resist the effect of accidental explosions)

Asher Gehl 9913569

266

Appendix A

TABLE A5: Ultimate unit resistance for One Way Elements

(US Department of the Army Technical Manual, TM5-1300. Design of structures to resist the effect of accidental explosions)

Asher Gehl 9913569

267

Appendix B

APPENDIX B
DYNAMIC DESIGN FACTORS
TABLE B1: Dynamic Increase Factor (DIF) for Reinforced Concrete Elements

(US Department of the Army Technical Manual, TM5-1300. Design of structures to resist the effect of accidental explosions)

Asher Gehl 9913569

268

Appendix B

TABLE B2: Dynamic Design Stresses for Design of Reinforced Concrete Elements

(US Department of the Army Technical Manual, TM5-1300. Design of structures to resist the effect of accidental explosions)

Asher Gehl 9913569

269

Appendix B

TABLE B3: Dynamic Increase Factor, c, for Yield Stress of Structural Steels

(US Department of the Army Technical Manual, TM5-1300. Design of structures to resist the effect of accidental explosions)

TABLE B4: Dynamic Increase Factor, c, for Ultimate Stress of Structural Steels

(US Department of the Army Technical Manual, TM5-1300. Design of structures to resist the effect of accidental explosions)

Asher Gehl 9913569

270

Appendix C

Appendix C
Maximum Single Degree Of Freedom Response Charts
TABLE C1: Maximum response of elastic Single-Degree-of-Freedom system for triangular load

(US Department of the Army Technical Manual, TM5-1300. Design of structures to resist the effect of accidental explosions)

Asher Gehl 9913569

271

Appendix C

TABLE C2: Maximum deflection of Elasto-Plastic, Single-Degree-of-Freedom system for triangular load (US Department of the Army Technical Manual, TM5-1300. Design of structures to resist the effect of accidental explosions)

Asher Gehl 9913569

272

Appendix C

TABLE C3: Maximum response time of Elasto-Plastic, Single-Degree-of-Freedom system for triangular load

(US Department of the Army Technical Manual, TM5-1300. Design of structures to resist the effect of accidental explosions)

Asher Gehl 9913569

273

Appendix D

APPENDIX D
Drag Coefficient Tables
TABLE D1: Drag Coefficient, CD, for various shapes

(US Department of the Army Technical Manual, TM5-1300. Design of structures to resist the effect of accidental explosions)

Asher Gehl 9913569

274

Appendix D

TABLE D2: Drag Coefficient, CD, for various shapes

(Kinney, Graham, 1962 Graham Explosive Shocks in Air )

Asher Gehl 9913569

275

Appendix E

APPENDIX E
Results from Conwep
TABLE E1: Output Data from CONWEP for a 100kg weight of ANFO, Hemispherical Surface Burst at a Standoff Distance of 5m

Asher Gehl 9913569

276

Appendix E

TABLE E2: Conwep Screen Display

Asher Gehl 9913569

277

Appendix E
E3 CONWEP Assumptions/References:

Hyde (1991) (Presented verbatim)

The airblast parameters (peak incident & reflected pressure and impulse, positive phase duration, and time of arrival) are calculated using the equations found in ARBRL-TR02555, "Airblast Parameters from TNT Spherical Air Burst and Hemispherical Surface Burst", by Charles N. Kingery and Gerald Bulmash, April 1984. This report contains a compilation of data from explosive tests using charge weights from less than 1 kg to over 400,000 kg. The authors used curve-fitting techniques to represent the data with polynomial equations, which are used by this program. These equations are presented in graphical form only in Figure 3-4 and 3-7 of TM 5-855-1, or Figures IV-2 and IV-5 of AFM 88-58. The program assumes an exponential decay of pressure with time of the form: Ps(t) = Pso*[1 - (t-ta)/to]* exp[-(t-ta)/] where Ps(t) = pressure at time t Pso = peak incident pressure to ta = positive phase duration = arrival time = decay coefficient

Asher Gehl 9913569

278

Appendix E The above equation is usually referred to as the modified Friedlander equation. For more information on this and other approximations, see "Explosions in Air", by Wilfred Baker, University of Texas Press, 1973. Using the peak pressure, impulse, and duration from the equations explained above, the program iterates to find the decay coefficient , which has dimensions of time. The program then uses Friedlander's equation to find pressure values at various time steps. Peak dynamic pressure, peak particle velocity, shock density, and specific heat ratio are calculated using procedures found in Appendix IV-2 of AFM 88-58, with the following ambient conditions assumed: Po = 14.696 psi co = 1116. ft/sec 0 = 0.0765 lb/ft
B B

(ambient pressure) (speed of sound) (air density) (specific heat ratio)

o = 1.4

Asher Gehl 9913569

279

Appendix F

APPENDIX F
DESIGN FLOW CHART
Figure F1: Design Process Flow Chart (G.C Mays and P.D Smith 1995, Blast Effects on Buildings)

Asher Gehl 9913569

280

Appendix G

APPENDIX G
Results For Design Of Columns Of Varying Cross Sectional Shape Subject To Blast Load
TABLE G1: Output Data from MathCAD Worksheet 1

REINFORCED CONCRETE PROTECTION CATEGORY 1

WORKSHEET 1 RC CATEGORY 1

D page 1 0.5

REINFORCEMENT RATIO 0.5%

CD page 6 2.05 1.55 1.2 1 0.5 0.1


B B

ru/P

td/T

page 12 50 35 25 18 7.5 3.5

page 13 7.4 5.204 3.7 2.7 1.1 0.5

tm/td
page 14 2.5 2 1.8 1.7 1.1 0.9

OK?

0.225 0.276 0.327 0.366 0.52 0.786 0.333 0.408 0.483 0.541 0.77 1.16 0.462 0.565 0.67 0.75 1.067 1.6

0.8 0.8 0.8 0.8 0.8 0.8 0.979 0.979 0.979 0.979 0.979 0.979 1.158 1.158 1.158 1.158 1.158 1.158

no no no no yes yes no no no yes yes yes no yes yes yes yes yes

0.6

2.05 1.55 1.2 1 0.5 0.1 2.05 1.55 1.2 1 0.5 0.1

30 20 13 9 3 1.4 15 9 5 4 1.9 1.1

3.7 2.45 1.47 1.1 0.37 0.17 1.56 0.937 0.52 0.42 0.2 0.115

1.8 1.6 1.3 1 0.9 0.6 1.2 1 0.9 0.7 0.6 0.35

0.7

Asher Gehl 9913569

281

Appendix G

TABLE G2: Output Data from MathCAD Worksheet 2

REINFORCED CONCRETE PROTECTION CATEGORY 2


WORKSHEET 2 1% steel CD 2.05 1.55 1.2 1 0.5 0.1 CD 2.05 1.55 1.2 1 0.5 0.1 dc 501 439 393 364 290 221 dc 401 351 314 291 232 178 tm/td 0.93 0.99 1.05 1.08 1.21 1.37 tm/td 0.73 0.776 0.82 0.85 0.94 1.057 OK? no no no no no no no no no no no no

2% steel

TABLE G3: Output Data from MathCAD Worksheet 3

STRUCTURAL STEEL PROTECTION CATEGORY 1

WORKSHEET 3 STEEL SECTION

CD 2.05 1.55 1.2 1 0.5 0.1

Zmin (mm3) 4200000 3405000 2871000 2570000 1804000 1194000

Asher Gehl 9913569

282

Appendix G

CHART G1: Results from MathCAD Worksheet 1 REINFORCED CONCRETE PROTECTION CATEGORY 1
CONCRETE DESIGN RESULTS CATEGORY 1 : DUCTILITY VS DRAG COEFFICIENT
60

50

40 D=0.5 30 D=0.6 D=0.7 Minimum duct ilit y f or design 20

10

0 2.05 1.55 1.2 1 0.5 0.1

D R A G C OEFF I C I EN T

CONCRETE DESIGN CATEGORY 1: ROTATION VS DRAG COEFFICIENT


8

5 Rotation (deg) D=0.5 4 D=0.6 D=0.7 theta(min) 3

0 2.05 1.55 1.2 Drag Coefficient 1 0.5 0.1

Asher Gehl 9913569

283

Appendix G
CHART G2: Results from MathCAD Worksheet 2 REINFORCED CONCRETE PROTECTION CATEGORY 2
CONCRETE DESIGN CATEGORY 2: DRAG COEFFICIENT VS DISTANCE BETWEEN COMPRESSION AND TENSION REINFORCEMENT CENTROIDS NOTE: PRESENTED FOR COMPARISON ONLY - SECTIONS DO NOT SATISFY DESIGN CRITERIA
600

500

400

Depth (mm)

300

1% steel 2% steel

200

100

0 0 0.5 1 DRAG COEFFICIENT 1.5 2 2.5

CONCRETE DESIGN CATEGORY 2: DRAG COEFFICIENT VS tm/td


3.5

2.5

1% steel 2% steel minimum ratio

tm/td
1.5 1 0.5 0 2.05 1.55 1.2 1 0.5 0.1 DRAG COEFFICIENT

Asher Gehl 9913569

284

Appendix G
CHART G3: Results from MathCAD Worksheet 3 STRUCTURAL STEEL PROTECTION CATEGORY 1
STEEL COLUMN DESIGN: DRAG COEFFICIENT VS MINIMUM PLASTIC SECTION MODULUS
4500000 4000000 3500000 3000000 2500000

2000000 1500000 1000000 500000 0 2.05 1.55 1.2 1 0.5 0.1 D R A G C OEF F I C I EN T

Asher Gehl 9913569

285

Das könnte Ihnen auch gefallen