A FRACTURE MECHANICS - BASED APPROACH FOR MODELING DELAMINATION OF SPRAY - APPLIED FIRE - RESISTIVE MATERIALS FROM STEEL STRUCTURES By Amir Arablouei A DISSERTATION Submitted to Michigan State University in partial fulfilment of the requirements for the degree of Civil Engineering - Doctor of Philosophy 201 5 ABSTRACT A FRACTURE MECHANICS - BASED APPROACH FOR MODELING DELAMINATION OF SPRAY - APPLIED FIRE - RESISTIVE MATERIALS FROM STEEL STRUCTURES By Amir Arablouei Steel structures exhibit lower fire - resistance due to high thermal conductivity of steel and rapid deterioratio n of strength and stiffness properties of steel with temperature. Therefore, steel structures are to be provided with fire insulation to achieve required fire resistance. This is often achieved through spray applied fire resistive materials (SFRM) that are externally applied on steel surface. The main function of SFRM is to delay temperature rise in steel, and thus slow down the degradation of stiffness and strength properties of steel when exposed to fire. Delamination of fire insulation can occur during service life of the structure due to exposure to harsh environmental conditions or due to poor bond properties at the interface of steel and SFRM. Further, high deformation levels in structural members due to extreme loading conditions such as earthquake, impact or explosion can lead to delamination of fire insulation from steel structures. Fire that can develop as a secondary event following an earthquake, explosion or impact (primary events) can cause significant damage and destruction to the steel struct ure if SFRM applied on the steel members experience fire insulation loss during primary events. For instance, combined effects of impact or blast and ensuing fire could lead to the progressive collapse of structure as in the case of the terrorist attacks o n the World Trade Center buildings (NIST, 2005) and collapse of Piper Alpha platform in North Sea (1988). In this research, an experimental - numerical approach is adopted to investigate delamination of fire insulation from steel structures subjected to stat ic loading and also extreme loading conditions such as seismic, impact and blast loading. The cohesive zone behavior at the interface of SFRM and steel is determined through static fracture tests conducted for three types of SFRM namely, mineral fiber - base d, gypsum - based and Portland cement - based SFRM. Subsequently, dynamic impact tests are carried out on beams insulated with above three types of SFRM to assess performance of SFRM under dynamic loading and also to assess the effect of strain rate on cohesiv e zone properties. A fracture mechanics - based numerical model, that can simulate crack initiation and propagation at the interface of steel and fire insulation, is developed in ANSYS and LS - DYNA for low and high strain rate loading conditions, respectively . The numerical approach is validated against both material and structural level tests. The validated numerical model is subsequently applied to quantify the effect of critical factors governing delamination phenomenon namely, fracture energy, elastic modu lus and thickness of SFRM. Results from parametric studies under static loading were utilized to identify the critical factors governing delamination of fire insulation from steel structures. Further, these results formed the basis for defining a delamination characteristic parameter that incorporates material - related governing factors in a single parameter and maintains interdependency between them. Results obtained from parametric study under impact loading is also utilized to estimate the dynamic increase factor (DIF) on fracture energy at the interface of steel and SFRM. Eventually, the delamination characteristic parameter is modified to capture differences in the nature of seismic and blast loading conditions, i.e. the way the stres ses are transferred to the interface of steel and SFRM. iv This research is dedicated to my beloved wife , Khadijeh and beautiful daughter , Hannah . Without their emotional support I would not have been able to accomplish this research . v ACKNOWLEDGEMENT S I would like to especially appreciate my advisor, Professor Venkatesh Kodur for his supports during the course of my study at Michigan State University. His continuous supports and understanding helped me to overcome many obstacles I enco untered with, over the past years. Undoubtedly, t he training I received at Michigan State University under his supervision will remain a life - time treasure for me. I would also like to appreciate my PhD committee members, Prof. Parviz Soroushian, Prof. Alejandro Diaz and Prof. Nizar Lajnef for the ir time and providing valuable guidance on improving this research. For her boundless support, love, and encouragement, I am thankful to my wife Khadijeh Rostami. She thoughtfully accompanied me along this journey. My ultimate regards goes to Mr. Siavosh Ravanbakhsh for his unlimited and astonishing supports during the experimental program in this research. I would like to extend my thanks to Ms. Margaret C onner, Ms. Laura Post, Ms. Mary Mroz and Ms. Laura Taylor for all the help they provided. I am very much thankful to Prof. Emin Kutay for providing the high speed camera . I would like to t hank my friends Ata Babazadeh, Anuj Shakya, Sudhir Varma, Ankit Agrawal and Esam Aziz for helping me during the impact experiments. Th e experiments presented in this thesis are partially supported through AISC Faculty Fellowship from American Institute of Steel Construction to Prof. Ko dur and Michigan State University. Any opinions, findings, and conclusions or recommendations expressed in this paper are those of the authors and do not necessarily reflect the views of the sponsors. vi TABLE OF CONTENTS LIST OF TABLES ................................ ................................ ................................ ........................... x LIST OF FIGURES ................................ ................................ ................................ ........................ xi KEY TO ABBREVIATIONS ................................ ................................ ................................ ..... xxiv CHAPTE R 1 ................................ ................................ ................................ ................................ ... 1 1 INTRODUCTION ................................ ................................ ................................ ................... 1 1.1 General ................................ ................................ ................................ ............................. 1 1.2 Role of SFRM i n Fire Performance of Steel Structures ................................ ................... 2 1.3 SFRM Categories and its Performance under Applied Loading ................................ ...... 5 1.4 Potential Loading Scenarios Leading to Delamination of SFRM from Steel Structures . 9 1.5 Mechanisms of Fracture and Delamination of SFRM ................................ ................... 12 1.6 Consequences of Fire Insulation Delamination ................................ ............................. 1 5 1.7 Research Objectives ................................ ................................ ................................ ....... 16 1.8 Anticipated Research Impact ................................ ................................ ......................... 18 1.9 Scope and Outline ................................ ................................ ................................ .......... 19 CHAPTER 2 ................................ ................................ ................................ ................................ . 20 2 STATE - OF - THE - ART REVIEW ................................ ................................ ......................... 20 2.1 General ................................ ................................ ................................ ........................... 20 2.2 Experimental Studies ................................ ................................ ................................ ...... 21 2.2. 1 Material Level Tests ................................ ................................ ............................... 21 2.2.2 Structural Level Tests ................................ ................................ ............................. 27 2.3 Numerical Studies ................................ ................................ ................................ .......... 30 2.4 Codes of Practice ................................ ................................ ................................ ............ 42 2.5 Knowledge Gaps ................................ ................................ ................................ ............ 47 CHAPTER 3 ................................ ................................ ................................ ................................ . 50 3 EXPERIMENTAL STUDY ................................ ................................ ................................ .. 50 3.1 Gene ral ................................ ................................ ................................ ........................... 50 3.2 Determination of Fracture Process Zone Properties for SFRM ................................ ..... 51 3.2.1 Test Procedures to Evaluate Cohesive Laws over FPZ ................................ .......... 51 3.2.2 Materials and Specimen Geometry ................................ ................................ ......... 53 3.2.3 Experimental Setup for Fracture Mode - I ................................ ................................ 54 3.2.4 Experimental Setup for Fracture Mode - II ................................ ............................... 56 3.2.5 Elastic Modulus Tests ................................ ................................ ............................. 58 3.3 Results from Fracture Tests ................................ ................................ ........................... 58 3.4 Drop Mass Impact Test ................................ ................................ ................................ .. 69 3.4.1 Selection of Experimental Approach ................................ ................................ ...... 69 vii 3.4.2 Impact Test Set - up ................................ ................................ ................................ .. 70 3. 4.3 Test Specimens ................................ ................................ ................................ ....... 72 3.5 Results from Impact Tests ................................ ................................ .............................. 74 3.5.1 Impact Force ................................ ................................ ................................ ........... 75 3.5.2 Displacement ................................ ................................ ................................ ........... 78 3.5.3 Extent of Delamination of Fire Insulation ................................ .............................. 80 3.6 Summary ................................ ................................ ................................ ........................ 84 CHAPTER 4 ................................ ................................ ................................ ................................ . 85 4 NUMERICAL MODELI NG ................................ ................................ ................................ . 85 4.1 General ................................ ................................ ................................ ........................... 85 4.2 Fracture Mechanics - Based Numerical Model ................................ ................................ 86 4.2.1 Characterizing SFRM Delamination using Fracture Mechanics ............................ 86 4.2.2 Implementation of Fracture Mechanics in Finite Element Model .......................... 87 4.2.3 Modeling Delamination of SFRM during Seismic Loading ................................ ... 89 4.2.3.1 Finite Element Discretization ................................ ................................ .............. 90 4.2.3.2 Contact Interaction ................................ ................................ .............................. 95 4.2.3.3 Material Constitutive Model ................................ ................................ ............... 96 4.2.3.4 Material Properties ................................ ................................ ............................ 100 4.2.3.5 Mesh Size ................................ ................................ ................................ .......... 100 4.2.3.6 Nonlinear Solution Predicaments ................................ ................................ ...... 101 4.2.4 Modeling Delamination of SFRM during Impact and Blast Loading .................. 102 4.2.4.1 Finite Element Discretization ................................ ................................ ............ 102 4.2.4.2 Contact Interaction ................................ ................................ ............................ 103 4.2.4.3 Material Constitutive Model ................................ ................................ ............. 104 4.2.4.4 Material Properties ................................ ................................ ............................ 106 4.2.4.5 Mesh Size ................................ ................................ ................................ .......... 107 4.2.4.6 Pitfalls in Explicit Solution ................................ ................................ ............... 107 4.3 Thermal - Structural Numerical Model ................................ ................................ .......... 108 4.3.1 Structural Analysis ................................ ................................ ................................ 109 4.3.2 Thermal Analysis ................................ ................................ ................................ .. 110 4.3.3 Material Constitutive Model ................................ ................................ ................. 113 4.3.4 Temperature - Dependent Material Properties ................................ ........................ 114 4.3.5 Fire Scenario ................................ ................................ ................................ ......... 119 4.4 Validation of Numerical Model ................................ ................................ ................... 119 4.4.1 Validation of Fracture Mechanics - Based Numerical Model ................................ 120 4.4.1.1 Direct Fracture Tests ................................ ................................ ......................... 120 4.4.1.2 Indirect Fracture Tests ................................ ................................ ....................... 127 4.4.1.3 Steel plate - SFRM Assembly under Tension ................................ ..................... 130 4.4.1.4 Cantilever Column Subjected Quasi - Static Loading ................................ ........ 133 4.4.1.5 Beam - Column Assembly Subjected to Seismic Loading ................................ . 138 4.4.1.6 Beam Subjected to Impact Loading ................................ ................................ .. 146 4.4.1.6.1 Insulated Beam Behavior ................................ ................................ ............. 146 4.4.1.6.2 Non - insulated Beam Behavior ................................ ................................ ..... 155 viii 4.4.1.6.3 Effect of Boundary Conditions ................................ ................................ .... 157 4.4.1.7 Beam - Column Subjected to Blast Loading ................................ ....................... 160 4.4.2 Validation of Thermal - Structural Numerical Model ................................ ............ 163 4.5 Summary ................................ ................................ ................................ ...................... 168 CHAPTER 5 ................................ ................................ ................................ ............................... 170 5 PARAMETRIC STUDY ................................ ................................ ................................ ..... 170 5.1 General ................................ ................................ ................................ ......................... 170 5.2 Factors Gove rning Delamination of SFRM from Steel Structures .............................. 171 5.3 Delamination of Fire Insulation from Slender Steel Truss ................................ .......... 172 5.3.1 Analysis Details ................................ ................................ ................................ .... 173 5.3.2 Effect of Type of SFRM on Delamination ................................ ........................... 174 5.3.3 Effect of Variation in Cohesive Zone Parameters on Delamination ..................... 180 5.3.4 Effect of Variation in SFRM Elastic Modulus ................................ ..................... 189 5.3.5 Effect of Variation in SFRM Thickness ................................ ............................... 192 5.3.6 Parameter for Characterizing Delamination of SFRM ................................ ......... 194 5.4 Delamination of Fire Insulation from Steel Beam - Column Assembly under Seismic Loading ................................ ................................ ................................ ................................ .... 197 5.4.1 Analysis Details ................................ ................................ ................................ .... 199 5.4.2 Effect of Type of SFRM ................................ ................................ ....................... 200 5.4.3 Effect of Cohesive Strength and Fracture Energy ................................ ................ 201 5.4.4 Effect of Elastic Modulus of Insulation ................................ ................................ 211 5.4.5 Effect of Insulation Thickness ................................ ................................ .............. 213 5.4.6 Additional Sensitivity Analysis ................................ ................................ ............ 216 5.4.7 Delamination Characteristic Parameter for Seismic Loading ............................... 218 5.5 Delamination of Fire Insulation from a Beam under Impact Loading ......................... 220 5.5.1 Analysis Details ................................ ................................ ................................ .... 221 5.5.2 Approach for Predicting Dynamic Increase Factor on Fracture Energy of SFRM 222 5.5.3 Numerical Predicaments ................................ ................................ ....................... 222 5.5.4 Quantified Dynamic Increase Factors ................................ ................................ ... 224 5.6 Delamination of Fire Insulation from Steel Beam - Column Assembly under Blast Loading ................................ ................................ ................................ ................................ .... 229 5.6.1 Analysis Details ................................ ................................ ................................ .... 230 5.6.2 Dynamic Response of the Beam - Column to Blast Load ................................ ...... 232 5.6.3 Effect of Blast Overpressure ................................ ................................ ................. 23 4 5.6.4 Effect of Elastic Modulus ................................ ................................ ..................... 237 5.6.5 Effect of Fracture Energy ................................ ................................ ...................... 240 5.6.6 Parameter for Characterizing Delamination of Fire Insulation under Blast Loading 242 5.7 Summary ................................ ................................ ................................ ...................... 244 CHAPTER 6 ................................ ................................ ................................ ............................... 247 6 CONSEQUENCES OF FIRE INSULATION DELAMINATION ................................ ..... 247 ix 6.1 General ................................ ................................ ................................ ......................... 247 6.2 Post - earthqu ake Fire Response of a Moment - Resisting Frame ................................ ... 248 6.2.1 Analysis Procedure to Obtain Load - Displacement Relationship .......................... 248 6.2.2 Analysis Procedure to Quantify Time to Failure ................................ .................. 260 6.3 Post - Blast Fire Response of a Beam - Column ................................ .............................. 266 6.4 Strategies to Overcome Consequences of Delam ination ................................ ............. 280 6.5 Summary ................................ ................................ ................................ ...................... 280 CHAPTER 7 ................................ ................................ ................................ ............................... 282 7 CONCLUSIONS AND RECOMMENDATIONS ................................ .............................. 282 7.1 General ................................ ................................ ................................ ......................... 282 7.2 Key Findings ................................ ................................ ................................ ................ 283 7.3 Resear ch Impact and Practical Implications ................................ ................................ 287 7.4 Recommendations for Future Research ................................ ................................ ....... 288 REFERENCES ................................ ................................ ................................ ........................... 291 x LIST OF TABLES Table 1.1 Material ingredients of CAFCO300 and CAFCO400 ................................ .................. 13 Table 3.1 Three type of SFRM utilized in experiments ................................ ................................ 54 Table 3.2 Cohesive zone model parameters obtained in experiment for three types of SFRM ... 58 Table 3.3 Fire insulated steel specimens and test variables ................................ ......................... 74 Table 4.1 Material properties used for verification of thermal - structural model ....................... 164 Table 5.1 Cohesive zone model parameters obtained in experiments for three types of SFRM 177 xi LIST OF FIGURES Figure 1.1 SFRM applied on steel structural elements ................................ ................................ ... 6 Figure 1.2 Delamination of fire insulation from steel structures (observed in World Trade Center) ................................ ................................ ................................ ................................ ......................... 8 Figure 1.3 Illustration of stress build up in moment resisting frame subjected to cyclic loading 10 Figure 1.4 Blast load on long steel truss and beam - column members ................................ .......... 12 Figure 1.5 Progression of cracks leading to delamination of SFRM from steel surface (Development of fracture process zone) ................................ ................................ ....................... 14 Figure 2.1 Normal and shear bond experiments carried out by Chen et al. (2010) ...................... 22 Figure 2.2 Normal bond experiments performed by Braxtan and Pessiki (2011a) ...................... 24 Figure 2.3 Experimental setup designed by NIST to measure fracture energy at steel - SFRM interface (Tan et al. (2011)) ................................ ................................ ................................ .......... 25 Figure 2.4 Load - displacement curves for different initial crack size measured by Tan et al. (2011) in their fracture tests ................................ ................................ ................................ .......... 26 Figure 2.5 Test setup of an exterior beam - column assembly to measure delamination of SFRM (Braxtan and Pessiki (2011b) ................................ ................................ ................................ ........ 28 Figure 2.6 Delamination of SFRM type Blaze Shield - II from bottom flange of the beam - column assembly tested by Braxtan and Pessiki (2011b) ................................ ................................ .......... 29 Figure 2.7 Test set - up for fire insulated column test (Wang et al. (2013)) ................................ ... 29 Figure 2.8 Debonding and fracture of SFRM from steel column at high levels of quasi - static load (Wang et al. (2013)) ................................ ................................ ................................ ...................... 30 Figure 2.9 Effect of partial loss of fire insulation on fire resistance of steel columns (Tomecek and Milke (1993)) ................................ ................................ ................................ ......................... 32 Figure 2.10 Effect of partial loss of fire insulation on fire resistance of steel columns (Ryder et al. (2002)) ................................ ................................ ................................ ................................ ...... 33 xii Figure 2.11 Temperature distribution over the cross section of steel column as a consequence of missing fire insulation from flange (Kwon et al. (2006)) ................................ ............................. 36 Figure 2.12 Capacity reduction of the column versus fire duration (Kwon et al. (2006)) ............ 37 Figure 2.13 Effect of fire insulation damage on fire resistance of a moment resisting frame (Gu and Kodur, (2011)) ................................ ................................ ................................ ........................ 38 Figure 2.14 Effect of fire insulation damage on fire resistance of a steel column (Dwaikat and Kodur, (2012)) ................................ ................................ ................................ .............................. 39 Figure 2.15 Effect of fire insulation damage on moment - rotation response of beam - column connection (Dwaikat and Kodur, (2012a)) ................................ ................................ ................... 40 Figure 2.16 Numerical modeling of fire insulation delamination from steel surface (Dwaikat and Kodur, (2011)) ................................ ................................ ................................ .............................. 41 Figure 2.17 Normal bonding test between steel and fire insulation based on ASTM E736 (2011) ................................ ................................ ................................ ................................ ....................... 44 Figure 2.18 Drop mass test for measuring durability of fire insulation under accidental impact loading based on ASTM E760 (2011) ................................ ................................ .......................... 45 Figure 2.19 Point load test for measuring durability of fire insulation under service loading conditions based on ASTM E759 (2011) ................................ ................................ ...................... 45 Figure 3.1 Test plate geometry for measuring fracture parameters ................................ .............. 55 Figure 3.2 Schematic of test assembly for determination of CZM parameters ............................ 56 Figure 3.3 Test set up designed for measuring CZM parameters at steel - SFRM ......................... 57 Figure 3.4 Normal force - displacement relationship (fracture mode - I) ................................ ......... 60 Figure 3.5 Shear force - displacement relationship (fracture mode - II) ................................ .......... 60 Figure 3.6 Fracture at steel - SFRM interface observed in the experiments ................................ ... 62 Figure 3.7 Interfacial cohesive strength for three types of SFRM ................................ ................ 66 Figure 3.8 Interfacial critical fracture energy for three types of SFRM ................................ ....... 66 xiii Figure 3.9 Determination of bilinear cohesive law based on experimental results ...................... 67 Figure 3.10 Bilinear cohesive law for fracture mode - I determined from experiments ................ 68 Figure 3.11 Bilinear cohesive law for fracture mode - II determined from experiments ............... 68 Figure 3.12 Schematic view of drop weight impact test set - up ................................ .................... 71 Figure 3.13 Schematic view of specimen ................................ ................................ ..................... 73 Figure 3.14 Force - time history recorded during impact tests on the beams insulated with different types of SFRM ................................ ................................ ................................ .............................. 76 Figure 3.15 Vertical displacement time history recorded during impact tests at the end of the beams insulated with different types of SFRM ................................ ................................ ............. 79 Figure 3.16 Extent of delamination of different types of SFRM from bottom flange of the I - beams subjected to impact loading ................................ ................................ ............................... 82 Figure 3.17 Illustration of high speed camera snapshots during impact on steel beams insulated with different types of SFRM (v=8.05 m/sec) ................................ ................................ .............. 82 Figure 4.1 Cohesive zone constitutive model for SFRM ................................ .............................. 99 Figure 4.2 True stress - true strain relationship for A992 - Gr. 50 steel ................................ ......... 107 Figure 4.3 Time - dependent engineering stress - strain relationship for S350 steel (EC3, 2005) . 115 Figure 4.4 Time - dependent modulus of elasticity for structural steel (EC3, 2005) ................... 116 Figure 4.5 Time - dependent thermal strain for structural steel (carbon steel) (EC3, 2005) ........ 116 Figure 4.6 Variation of thermal conductivity with temperature for structural steel (carbon steel) (EC3, 2005) ................................ ................................ ................................ ................................ . 117 Figure 4.7 Variation of Specific heat with temperature for structural steel (carbon steel) (EC3, 2005) ................................ ................................ ................................ ................................ ........... 117 Figure 4.8 Variation of thermal conductivity with temperature for gypsum - based SFRM (Kodur and Shakya, 2013 ................................ ................................ ................................ ........................ 118 xiv Figure 4.9 Variation of specific heat with temperature for gypsum - based SFRM (Kodur and Shakya , 2013) ................................ ................................ ................................ .............................. 118 Figure 4.10 Finite element model of SFRM - steel assembly in mode - I fracture experiments .... 121 Figure 4.11 Finite element model of SFRM - steel assembly in mode - II fracture experiments ... 122 Figure 4.12 Comparison of force - displacement relationship predicted from numerical model with measured values from experiments for gypsum - based SFRM ................................ ................... 124 Figure 4.13 Comparison of force - displacement relationship predicted from numerical model with measured values from experiments for Portland cement - based SFRM ................................ ...... 125 Figure 4.14 Comparison of force - displacement relationship predicted from numerical model with measured values from experimen ts for mineral fiber - based SFRM ................................ ........... 126 Figure 4.15 Experimental setup for fracture tests carried out by NIST to measure fracture energy of SFRM ................................ ................................ ................................ ................................ ...... 128 Figure 4.16 Finite element model for fracture test conducted by NIST ................................ ..... 128 Figure 4.17 Measured and predicted force - displacement response at the steel - insulation interface for material level tests conducted at NIST ................................ ................................ .................. 129 Figure 4.18 Experimental setup for plate covered with mineral fiber - based SFRM (Braxtan and Pessiki, 2011a) ................................ ................................ ................................ ............................ 130 Figure 4.19 Finite element model for plate covered with mineral fiber - based SFRM ............... 131 Figure 4.20 Delamination length predicted and measured on a plate covered with mineral fiber - based SFRM ................................ ................................ ................................ ................................ 132 Figure 4.21 Experimental set - up for quasi - static loading of a cantilever column insulated with SFRM (Wang et al., 2013) ................................ ................................ ................................ .......... 134 Figure 4.22 Finite element model for quasi - static loading of a cantilever column insulated with SFRM ................................ ................................ ................................ ................................ .......... 135 Figure 4.23 Load - displacement response of the insulated steel column tested by Wang et al. (2013) ................................ ................................ ................................ ................................ .......... 136 xv Figure 4.24 Delamination and fracture of SFRM from steel column at high levels of static load ................................ ................................ ................................ ................................ ..................... 137 Figure 4.25 Test setup of an exterior moment frame assembly to measure delamination of SFRM (Braxtan and Pessiki (2011b) ................................ ................................ ................................ ...... 140 Figure 4.26 Finite element model of beam - column assembly tested by Braxtan and Pessiki (2011b) ................................ ................................ ................................ ................................ ........ 141 Figure 4.27 Cyclic displacements applied at beam tip for validation of SFRM delamination ... 142 Figure 4.28 Comparison of predicted and measured load versus percent drift in insulated beam - column assembly ................................ ................................ ................................ ......................... 143 Figure 4.29 Comparison of predicted and measured extent of delamination in beam - column assembly insulated with gypsum - based SFRM ................................ ................................ .......... 144 Figure 4.30 Comparison of predicted and measured extent of delamination in beam - column assembly insulated with mineral fiber - based SFRM ................................ ................................ .. 145 Figure 4.31 Numerical model of experimental setup for fire insulated beams in drop mass impact test ................................ ................................ ................................ ................................ ............... 147 Figure 4.32 Comparison of experimental and predicted impact force for beams insulated with mineral fiber - based SFRM subjected to two different impact veloci ties ................................ ... 149 Figure 4.33 Comparison of experimental and predicted impact force for beams insulated with gypsum - based SFRM subject to two different impact velocities ................................ ............... 150 Figure 4.34 Comparison of experimental and predicted impact force for beams insulated with Portland cement - based SFRM subject to two different impact velocities ................................ .. 151 Figure 4.35 Comparison of experimental and predicted vertical deflection at one end of beams insulated with mineral fiber - based SFRM and subjected to two different impact velocities ..... 152 Figure 4.36 Comparison of experimental and predicted vertical deflection at one end of beams insulated with gypsum - based SFRM and subjected to two different impact velocities ............. 153 Figure 4.37 Comparison of experimental and predicted vertical deflection at one end of beams insulated with Portland cement - based SFRM and subjected to two different impact velocities 154 xvi Figure 4.38 Comparison of experimental and predicted responses for a steel beam without insulation ................................ ................................ ................................ ................................ ..... 156 Figure 4.39 Numerical models created for evaluating the effect of supports on beam behavior in drop ................................ ................................ ................................ ................................ ............. 158 Figure 4.40 Effect of including support details in numerical predictions ................................ ... 159 Figure 4.41 Schematic view of blast tests on steel beam - column conducted by Nassr et al. (2011) ................................ ................................ ................................ ................................ ..................... 160 Figure 4.42 Finite element model of beam - column tested by Nassr et al. (2011) ...................... 161 Figure 4.43 Comparison between numerical prediction and experimentally measured deflection at mid - span of the beam - column under blast loading ................................ ................................ . 162 Figure 4.44 Location of thermocouples in cross section of the concrete filled column SQ24 ... 164 Figure 4.45 Finite element Model for concrete filled steel column ................................ ............ 165 Figure 4.46 Temperature prediction ................................ ................................ ........................... 166 Figure 4.47 Temperature ( ° K) distribution at cross section ................................ ........................ 167 Figure 4.48 Axial deformation of the column ................................ ................................ ............ 168 Figure 5.1 Progression of crack at the interface of steel and SFRM (based on cohesive zone model concept) ................................ ................................ ................................ ............................ 174 Figure 5.2 Finite element model of bottom chord of truss encapsulated with SFRM ................ 175 Figure 5.3 Percentage of delamination progression on outer sides of truss chord with respect to average axial strain developed in truss member ................................ ................................ ......... 176 Figure 5.4 Crack propagation pattern at the interface of steel and gypsum - based SFRM at different strain levels ................................ ................................ ................................ ................... 178 Figure 5.5 Fracture energies released during delamination for gypsum ................................ ..... 179 Figure 5.6 Schematic view of various analyses cases for studying the sensitivity of delamination progression to CZM parameters ................................ ................................ ................................ .. 181 xvii Figure 5.7 Effect of fracture energy of gypsum - based SFRM on initiation and progression of delamination in truss member under tension ................................ ................................ .............. 183 Figure 5.8 Effect of fracture energy of Portland cement - based SFRM on initiation and progression of delamination in truss me mber under tension ................................ ...................... 184 Figure 5.9 Effect of fracture energy of Mineral fiber - based SFRM on initiation and progression of delamination in truss member under tension ................................ ................................ .......... 185 Figure 5.10 Effect of cohesive strength of gypsum - based SFRM on initiation and p rogression of delamination in truss member under tension ................................ ................................ .............. 186 Figure 5.11 Effect of cohesive strength of Portland cement - based SFRM on initiation and progression of delamination in truss member under tension ................................ ...................... 187 Figure 5.12 Effect of cohesive strength of Mineral fiber - based SFRM on initiation and progression of delamination in truss member under tension ................................ ...................... 188 Figure 5.13 Effect of SFRM elastic modulus on initiation and progression of delamination in truss member under tension (Gypsum - based SFRM) ................................ ................................ . 190 Figure 5.14 Effect of SFRM elastic modulus on initiation and progression of delamination in truss member under tension (Portland cement - based SFRM) ................................ .................... 190 Figure 5.15 Effect of SFRM elastic modulus on initiation and progression of delamination in truss member under tension (Mineral fiber - based SFRM) ................................ ......................... 191 Figure 5.16 Effect of SFRM thickness on initiation and progression of delamination in truss member under tension (Gypsum - based SFRM) ................................ ................................ ......... 192 Figure 5.17 Effect of SFRM thickness on initiation and progression of delamination in truss member under tension (Portland cement - based SFRM) ................................ ............................. 193 Figure 5.18 Effect of SFRM thickness on initiation and progression of delamination in truss member under tension (Mineral fiber - based SFRM) ................................ ................................ .. 193 Figure 5.19 Strain ductility demand of steel at delamination initiation versus delamination parameter ................................ ................................ ................................ ................................ ..... 196 Figure 5.20 Strain ductility demand of steel at complete delamination versus delamination parameter ................................ ................................ ................................ ................................ ..... 196 xviii Figure 5.21 Finite element model of beam - column assembly used for parametric study under seismic ................................ ................................ ................................ ................................ ......... 198 Figure 5.22 Cyclic displacements applied at beam tip for parametric study .............................. 199 Figure 5.23 Percentage of delamination of three types of SFRM applied on the beam in a beam column assembly subjected to seismic loading ................................ ................................ .......... 201 Figure 5.24 Effect of fracture energy of gypsum - based SFRM on delamination progression on the beam of beam - column assembly subjected to seismic loading ................................ ............. 203 Figure 5.25 Effect of fracture energy of Portland cement - based SFRM on delamination progression on the beam of beam - column assembly subjected to seismic loading .................... 204 Figure 5.26 Effect of fracture energy of mineral fiber - based SFRM on delamination prog ression on the beam of beam - column assembly subjected to seismic loading ................................ ........ 205 Figure 5.27 Effect of cohesive strength of gypsum - based SFRM on delamination progression on the beam of beam - column assembly subjected to seismic loading ................................ ............. 207 Figure 5.28 Effect of cohesive strength of Portland cement - based SFRM on delamination progression on the beam of beam - column assembly subjected to seismic loading .................... 208 Figure 5.29 Effect of cohesive strength of mineral fiber - based SFRM on delamination progression on the beam of beam - column assembly subjected to seismic loading .................... 209 Figure 5.30 Effect of elastic modulus of gypsum - based SFRM on delamination progression on the beam of beam - column assembly subjected to seismic loading ................................ ............. 212 Figure 5.31 Effect of elastic modulus of Portland cement - based SFRM on delamination progression on the beam of beam - column assembly subjected to seismic loading .................... 212 Figure 5.32 Effect of elastic modulus of mineral fiber - based SFRM on delamination progression on the beam of beam - co lumn assembly subjected to seismic loading ................................ ........ 213 Figure 5.33 Effect of thickness of gypsum - based SFRM on delamination progression on the beam of beam - column assembly subjected to seismic loading ................................ .................. 215 Figure 5.34 Effect of thickness of Portland cement - based S FRM on delamination progression on the beam of beam - column assembly subjected to seismic loading ................................ ............. 215 xix Figure 5.35 Effect of thickness of mineral fiber - based SFRM on delamination progression on the beam of beam - column assembly subjected to seismic loading ................................ .................. 216 Figure 5.36 Effect of change in material properties of gypsum - based SFRM on delamination progression on the beam of beam - column assembly subjected to seismic loading .................... 217 Figure 5.37 Effect of change in material properties of Portland cement - based SFRM on delamination progression on the beam of beam - column assembly subjected to seismic loading ................................ ................................ ................................ ................................ ..................... 218 Figure 5.38 Extent of delamination as a function of parameter E.t/G c ................................ ....... 220 Figure 5.39 Extent of delamination ratio versus dynamic increase factor for CZM properties of mineral fiber - based SFRM ................................ ................................ ................................ .......... 225 Figure 5.40 Extent of delamination ratio versus dynamic increase factor for CZM properties of gypsum - based SFRM ................................ ................................ ................................ .................. 226 Figure 5.41 Extent of delamination ratio versus dynamic increase factor for CZM properties of Portland cement - based SFRM ................................ ................................ ................................ .... 226 Figure 5.42 Numerical and experimental illustration of extent of delamination in beam insulated with Mineral fiber - based SFRM (DIF=1.00) ................................ ................................ .............. 227 Figure 5.43 Numerical and experimental illustration of extent of delamination in beam insulated with Gypsum - based SFRM (DIF=1.41) ................................ ................................ ..................... 228 Figure 5.44 Numerical and experimental illustration of extent of delamination in beam insulated with Portland cement - based SFRM (DIF=2.32) ................................ ................................ ......... 229 Figure 5.45 Finite element model of beam - column insulated with SFRM utilized in parametric study under blast loading ................................ ................................ ................................ ............ 231 Figure 5.46 Mid - span deflection of the beam - column insulated with SFRM under different blast overpressure level ................................ ................................ ................................ ....................... 233 Figure 5.47 Illustration of delamination of mineral fiber - based SFRM from steel beam - column over the first 10 ms of the blast scenario ................................ ................................ .................... 234 Figure 5.48 Extent of delamination on steel column as a function of blast overpressure for mineral fiber - based SFRM ................................ ................................ ................................ .......... 235 xx Figure 5.49 Extent of delamination on steel column as a function of blast overpressure for gypsum - based SFRM ................................ ................................ ................................ .................. 236 Figure 5.50 Extent of delamination on steel column as a function of blast overpressure for Portland cement - based SFRM ................................ ................................ ................................ .... 236 Figure 5.51 Extent of delamination on steel column as a function of elastic modulus for mineral fiber - based SFRM ................................ ................................ ................................ ....................... 238 Figure 5.52 Extent of delamination on steel column as a function of elastic modulus for gypsum - based SFRM ................................ ................................ ................................ .......................... 239 Figure 5.53 Extent of delamination on steel column as a function of elastic modulus for Portland cement - based SFRM ................................ ................................ ................................ ................... 239 Figure 5.54 Extent of delamination on steel column as a function of fracture energy for mineral fiber - based SFRM ................................ ................................ ................................ ....................... 241 Figure 5.55 Extent of delamination on steel column as a function of fracture energy for gypsum - based SFRM ................................ ................................ ................................ ................................ 241 Figure 5.56 Extent of delamination on steel column as a function of fracture energy for Portland cement - based SFRM ................................ ................................ ................................ ................... 242 Figure 5.57 Extent of delamination on steel beam - column under blast loading as a function of delamination characteristic parameter ................................ ................................ ........................ 244 Figure 6.1 Finite element model of thermal analysis carried out to compute temperature time history of beam - column ................................ ................................ ................................ .............. 249 Figure 6.2 Temperature (K) distribution in the beam insulated with gypsum - based SFRM exposed to standard fire (ASTM E119) ................................ ................................ ...................... 251 Figure 6.3 Temperature evolution in the beam and column subjected to standard fire (ASTM E119) ................................ ................................ ................................ ................................ ........... 252 Figure 6.4 Finite element model of beam - column assembly used for studying effect of temperature rise on capacity reduction of moment connection due to loss of fire insulation during seismic loading ................................ ................................ ................................ ............................ 253 Figure 6.5 Effect of temperature on load - displacement relationship of beam - c olumn assembly endured 25% delamination over the plastic hinge region on the beam ................................ ...... 254 xxi Figure 6.6 Effect of temperature on load - displacement relationship of beam - column assembly endured 50% delamination over the plastic hinge region on the beam ................................ ...... 254 Fig ure 6.7 Effect of temperature on load - displacement relationship of beam - column assembly endured 75% delamination over the plastic hinge region on the beam ................................ ...... 255 Figure 6.8 Effect of temperature on load - displacement relationship of beam - column assembly endured 100% delamination over the plastic hinge region on the beam ................................ .... 255 Figure 6.9 Effect of delamination percentage over the plastic hinge region of the beam on load - displacement relationship of beam - column assembly at temperature of 200 °C ........................ 256 Figure 6.10 Effect of percentage of delamination over the plastic hinge region of the beam on load - displacement relationship of beam - column assembly at temperature of 400 °C ............... 256 Figure 6.11 Effect of percentage of delamination over the plastic hinge region of the beam on load - displacement relationship of beam - column assembly at temperature of 600 °C ............... 257 Figure 6.12 Effect of percentage of delamination over the plastic hinge region of the beam on load - displacement relationship of beam - column assembly at temperature of 800 °C ............... 257 Figure 6.13 Effect of percentage of delamination over the plastic hinge region of the beam on load - displacement relationship of beam - column assembly at temperature of 1000 °C ............. 258 Figure 6.14 Plastic strain distribution in beam - column assembly at T=200 °C with 25% delamination over the plastic hinge region ................................ ................................ ................. 258 Figure 6.15 Plastic strain distribution in beam - column assembly at T=400 °C with 50% delamination over the plastic hinge region ................................ ................................ ................. 2 59 Figure 6.16 Plastic strain distribution in beam - column assembly at T=600 °C with 75% delamination over the plastic hinge region ................................ ................................ ................. 259 Figure 6.17 Plastic strain distribution in beam - column assembly at T=800 °C with 100% delamination over the plastic hinge re gion ................................ ................................ ................. 260 Figure 6.18 Finite element model of beam - column assembly used for quantifying failure time during fire following earthquake ................................ ................................ ................................ 261 Figure 6.19 Displacement of the beam during fire following an earthquake that has undergone insulation damage du ring seismic loading ................................ ................................ .................. 262 xx ii Figure 6.20 Displacement of the beam during fire following an earthquake that has undergone insulation damage during seismic loading ................................ ................................ .................. 262 Figure 6.21 Plastic strain distribution in beam - column assembly at the time of failure (0% d elamination of SFRM over the plastic hinge region) ................................ ................................ 264 Figure 6.22 Plastic strain distribution in beam - column assembly at the time of failure (25% delamination of SFRM over the plastic hinge region) ................................ ................................ 264 Figure 6.23 Plastic strain distribution in beam - co lumn assembly at the time of failure (50% delamination of SFRM over the plastic hinge region) ................................ ................................ 265 Figure 6.24 Plastic strain distribution in beam - column assembly at the time of failure (75% delamination of SFRM over the plastic hinge region) ................................ ................................ 265 Figure 6.25 Plastic strain distribution in beam - column assembly at the time of failure (100% delamination of SFRM over the plastic hinge region) ................................ ................................ 266 Figure 6.26 Finite element model of cross section of column for thermal analysis ................... 267 Figure 6.2 7 Temperature distribution ( ° K) in the cross section of column with full SFRM after 2 hours of exposure to standard fire (ASTM E119) ................................ ................................ ...... 269 Figure 6.28 Temperature distribution ( ° K) in the cross section of column with 25% delamination of SFRM after 2 hours of exposure to standard fire (ASTM E119 ) ................................ ........... 269 Figure 6.29 Temperature distribution (°K) in the cross section of column with 50% delamination of SFRM after 2 hours of exposure to standard fire (ASTM E119) ................................ ........... 270 Figure 6.30 Temperature distribution (°K) in the cross section of column with 75% dela mination of SFRM after 2 hours of exposure to standard fire (ASTM E119) ................................ ........... 270 Figure 6.31 Temperature distribution (°K) in the cross section of column with 100% delamination of SFRM after 2 hours of exposure to standard fire (ASTM E119) ..................... 271 Figure 6.32 Fi nite element of the column to simulate effect of temperature rise after blast loading on structural response of the column ................................ ................................ .......................... 271 F igure 6.33 Vertical displacement of the column during fire following blast ............................ 273 Figure 6.34 Time to failure of the column exposed to fire following blast as a function of percentage delamination occurred during blast loading ................................ ............................. 273 xxiii Figure 6.35 Axial fo rce developed in the column during fire following blast ........................... 275 Figure 6.36 Effect of axial restraint on vertical expansion of column exposed to fire following blast (100% delamination) ................................ ................................ ................................ .......... 275 Figure 6.37 Displacement vector (mm) of column endured 25% delamination during fire following blast loading ................................ ................................ ................................ ............... 276 Figure 6.38 Displacement vector (mm) of column endured 50% delamination during fire following blast loading ................................ ................................ ................................ ............... 276 Figure 6.39 Displacement vector (mm) of column endured 75% delamination during fire foll owing blast loading ................................ ................................ ................................ ............... 277 Figure 6.40 Displacement vector (mm) of column endured 100% delamination during fire following blast loading ................................ ................................ ................................ ............... 277 Figure 6.41 Plastic strain distribution over the column undergone 25% delamination of fire insulation and subjected to fire following blast loading ................................ ............................. 278 Figure 6.42 Plastic strain distribution over the column undergone 50% delamination of fire insulation and subjected to fire following blast loading ................................ ............................. 278 Figure 6.43 Plastic strain distribution over the column undergone 75% delamination of fire insulation and subjected to fire following blast loading ................................ ............................. 279 Figure 6.44 Plastic strain distribution over the column undergone 100% delamination of fire insulation and subjected to fire following blast loading ................................ ............................. 279 xxiv KEY TO A BB REVIATIONS FPZ : Fracture Process Zone CZM: Cohesive Zone Model VCCT : Virtual C rack C losure T echnique LEFM : Linear Elastic Fracture Mechanics CTOD : Crack Tip Opening Displacement SFRM : Sprayed Applied Fire Resistive Material LVDT: Li near V ariable D ifferential T ransformer DIF : Dynamic Increase Factor I 1 : F irst i nvariant of s tress tensor J 2 : S econd invariant of stress deviator tensor : Internal friction angle G nc : Critical fracture energy at normal mode G tc : Critical fracture energy at tangential mode c : Normal cohesive streng th c : Tangential cohesive strength n,c : Normal failure displacement t ,c Tangential failure displacement : Total mixed - mode relative displacement n : Separation in normal direction t : Separation in tangential direction n,o : Normal separation at normal cohesive strength xxv t,o : Tangential separation at tangential cohesive strength q : heat flux Q : H eat source k: Thermal conductivity h f : F ilm coefficient T B : bulk temperature of adjacent fluid (air) T f : Fire temperature : Stefan - Boltzman constant 1 CHAPTER 1 1 INTRODUCTION 1.1 General Steel is one of the primary materials used in structural framing of buildings due to numerous advantages steel offers such as high strength - to - weight ratio, high level of ductility and ease in fabrication and construction process. However, steel structures do not exhibit good fire - resistance due to high thermal conductivity of steel and rapid deterioration of strength and stiffness properties of steel with temperature. Hence, to maintain stability and integrity of steel structures during fire, steel structu res are to be provided with fire insulation to achieve required fire resistance. This is often achieved through spray applied fire resistive materials (SFRM) that are externally applied on steel surface. SFRM is widely used as fire insulation material due to number of advantages it offers over other insulation materials, including low thermal conductivity, light weight, cost - effectiveness and ease of application (Kodur and Shakya, 2013). The main function of SFRM is to delay the temperature rise in steel, and thus slow down the degradation of stiffness and strength properties of steel when exposed to fire. 2 1.2 Role of SFRM in F ire P erformance of S teel S tructures Fire performance of a steel st ructure during normal loading conditions strongly relies on the quality of SFRM, equipment, workmanship and the application process. Delamination of fire insulation can occur during service life of the structure due to exposure to harsh environmental condi tions , deterioration in material properties over the time or due to poor initial bond properties at the interface of steel and SFRM. Further, any change in functionality of structure can conse quently increase the load level on f ire insulated structural mem ber and hence can be a potential factor for inducing cracking and delamination of SFRM during service life . Therefore, to achieve good performance from SFRM during a fire event , the delamination of SFRM under service loading conditions should be minimized. This entails utilizing SFRM with high fracture resistance, controlling the quality of application, and monitoring the condition of SFRM on structural members on a regular basis. Fire can not only occur during normal loading condition, but also can develo p following an extreme loading condition that strikes the structure. Fire following an earthquake is one of the possible scenarios to be accounted for in the design of structural systems (Mousavi et al. 2008). Post - earthquake fires caused numerous fataliti es and high fire losses in many previous earthquakes. As an example, in the aftermath of Hyougoken - Nambu earthquake (Kobe, Japan, 1995) 7000 buildings were destroyed by post - earthquake fires alone (Faggiano, 2007). Further, e xplosion and impact are the oth er possible loading scenarios to be considered in the design of critical steel structures such as tall buildings, petrochemical facilities and offshore platforms. Fire that can develop as a secondary event following an explosion or impact (primary events) can cause significant damage and destruction to the structure. The combined effects of impact or blast and ensuing fire could lead to the progressive failure of structure as in the case of 3 the terrorist attack on the World Trade Center buildings (NIST, 200 5) and collapse of Piper Alpha platform in North Sea (1988). These impactful events have shown that, although explosion, impact , earthquake and subsequent fire s are rare events in structures, their ramifications can be disastrous which include, but not limited to personnel casualties, environmental damage and considerable property losses. Consequently , p ost - earthquake , post - impact and post - blast fire consideratio n has been drawing attention over the past few past years as part of an emerging trend towards enhancing structural resiliency under multi - hazard scenarios . Substantial inelastic actions in structures during an earthquake , impact and blast can impose large deformation in structural and non - structural elements. During such extreme loading events, there is therefore a high possibility that active fire protection systems get compromised by ruptured water supply piping system and delayed response for firefighting (Mousavi et al. 2008). In such scenarios, adequate fire resistance of structure is the only line of defense for overcoming the damage or collapse of structural systems. In other words , the fire performance of steel structures relies entirely on the effectiveness of fire insulation applied on structural members. Given the fact that the fire performance of steel structures relies entirely on the effectiveness of fire insulation applied on structural members, a crucial question that can be raised is whether the fire insulation will remain in - place after massive energy transfer to structure during seismic, blast or impact loading. The answer to this question is negative since the role of SFRM, as a protective layer during fire following above extreme loading conditions, can be compromised if the energy transferred to the structure by seismic, impact and blast loading , can cause fracture and delamination of fire insulation from steel surface. Both experiments and field observations have 4 shown that SFRM can delaminate unde r static, cyclic and blast loading (Braxtan and Pessiki, 2011b; Wang et al., 2013; NIST 2005). Und er seismic, impact and blast loading, dynamic interfacial stresses developed at the SFRM - steel interface in the highly stressed zones of structural elements can open the cracks that are inevitably left over from SFRM application process. Once initiated, th eses cracks can rapidly propagate along the interface of steel and SFRM leading to delamination of SFRM from steel surface. Therefore, efficiency of SFRM during fire following earthquake, blast and impact, entails assuring stable dynamic fracture resistance at steel - SFRM interface such that SFRM would not delaminate during these impulsive loading events or at least the extent of delamination would be minim al . An additional key question is that whether the SFRM types, currently utilized in steel co nstruction, possess enough fracture toughness to resist against fracture and delamination under the action of seismic, impact and blast loading conditions. Further, if the current SFRM types are vulnerable and hence can be dislodged from steel surface, wha t types of material properties would be required to avoid the delamination of fire insulation from steel structures. Owing to the lack of answers to above questions, current fire safety provisions do not address the effect of multiple hazards su ch as fire following earthquake or impact, or blast on fire resistance of structures. For evaluating post - earthquake, post - blast and post - impact fire performance of steel structures, it is of crucial importance to have comprehensive knowledge regarding the extent of SFRM damage during primary event of earthquake , impact and blast loading. In current practice, it is assumed that the SFRM will not debond or disintegrate and will continue to maintain its integrity 5 throughou t the fire following earthquake, impact or blast. In fire resistance analysis, thermal response of steel structures is evaluated by assuming SFRM to be perfectly intact during the fire exposure. This serious shortcoming in current provisions necessitates developing a robust approach to predict the delamination of SFRM from steel surface under the action of extreme loading events on a structure. Developing such knowledge is one of the imperative steps towards rational design and assessment of post - earthquake, post - impact and post - blast fire performa nce of steel structures. 1.3 SFRM C ategories and its P erformance under A pplied L oading Spray - applied fire - resistive material (SFRM) is commercially available in cementitious and mineral fiber - based forms. Cementitious - based SFRM is further grouped under two categories; gypsum - based SFRM that comprises gypsum and vermiculite, and Portland - cement based SFRM that is composed of Portland cement and vermiculite. Mineral fiber - based fire insulation comprises of Portland cement and mineral wool fiber mixture. Cement itious and mineral - fiber - based SFRM are delivered to the construction site as wet - mix and dry - mix, respectively. Figure 1 . 1 shows SFRM applied on steel structural elements. There are number of factors that can influence the decision of building owners on choosing what type of SFRM should be used in a specific buil ding. Fire engineers in close collaboration with structural engineers determine the thickness and type of SFRM to be applied on the steel structure to provide the desired protection against fire. However, this decision is mainly made based on the fire - rati ng s prescribed in current codes and standards. Fire - rating for beams and columns are affected by only thermal properties (i.e. thermal conductivity and specific heat) of SFRM, hence the mechanical properties (i.e. elastic modulus and fracture energy ) of SF RM are 6 not given any consideration when designing the fire protection. Therefore, the final decision may not necessarily lead to selection of SFRM with high bonding properties. Figure 1 . 1 SFRM applied on steel structural elements b) Mineral fiber - based SFRM applied on trusses in a floor assembly a) Gypsum - based SFRM applied on beams and columns in a moment - resisting frame 7 Mechanical p erformance of SFRM is highly dependent upon its integrity, constitutive ingredients, and the manner in which insulation is prepared and applied to the steel surface. During application of SFRM on steel structural members, microscopic cracks can develop within bulk SFRM itself, and also at the i nterface between steel and SFRM, mainly due to high shrinkage and low tensile strength of SFRM. Poorly bonded SFRM can be dislodged under combination of permanent dead loads and frequent ly applied loading - unloading cycles of live loads. For instance, Figure 1 . 2 illustrates delamination of fire insulation from steel beam and truss elements under service loading conditions , which were observed in World Trade Center towers during inspection by Port Authority of New York in 1993 . E ven in case of a good bond conditions, SFRM may experience some level of delamination from steel surface due to the fact that steel structures undergo high level of deformations under extreme loading conditions . As a result of such large strains developed in steel, strain compatibility can no longer be held at steel - SFRM interface. Consequently, existing microcracks within SFRM can widen and prop agate to the steel - SFRM interface leading to partial or full delamination of fire insulation. 8 Figure 1 . 2 Delamination of fire insulation from steel structures (observed in World Trade Center) a) Delamination of insulation from steel beam (source: Fire Protection Engineering Magazine) b) Delamination of insulation from truss system in floor assembly Missing Insulation 9 1.4 Potential L oading S cenarios L eading to D elamination of SFRM from S teel S tructures Special steel moment - resisting frames, which have gained vas t attention in earthquake prone regions, are assigned the highest response modification factor (R) (NEHERP, 2009) and thus are expected to experience very large deformations. In steel moment frames subjected to earthquake loading, as shown in Figure 1 . 3 , plastic hinges are formed in beams at the vicini ty of columns, as well as at bottom of the columns, thereby the nonlinear actions in the structure is accommodated. Owing to large cyclic strains that develop in steel resulting from its high ductility demands during earthquake loading cycles, strain compa tibility can no longer be maintained at the interface of steel and SFRM , as mentioned before . Since significant amount of strain energy is dissipated in the plastic hinge region and the beam cross section is highly distorted due to likely flange and web lo cal buckling, considerable amount of energy also gets released at the interface of steel and applied SFRM. Further, e ffect of local buckling in flange and web on the extent of delamination are therefore expected to be quite significant. However, in practi ce, regardless of the ductility demands anticipated to develop in the plastic hinge zones of the structure, SFRM with same properties and thickness is applied on the entire structural frame. This is applied by neglecting the fact that the level of strains developed at the steel - SFRM interface can vary over a broad range along the beam span . In other words, SFRM applied to plastic hinge region of a beam (in the vicinity of supports) will demand quite high lev el of fracture propert ies in order to remain in - place during cyclic loading. Under the action of seismic loading, crack tips are frequently subjected to tensile and shear stresses arising from steel deformations, either due to plastification or sudden defo rmation because of local buckling. Consequently, the damage accumulation during each loading cycle can substantially affect the extent of delamination in plastic hinge zones. 10 Figure 1 . 3 Illustration of stress build up in moment resisting frame subjected to cyclic loading Moment - resisting frames encapsulated with SFRM can also be subjected to debris impact as a result of an internal or external explosion or blast over pressure applied on the surface of the structural elements . Direct debris impact can locally damage the insulation applied on the members. Further, the impact kinetic energy, depending on the mass and velocity of debris, can cause delamination on the unexposed areas of the member due to stress wave propagation throughout the impacted member. The blast pressure generated during an explosion can affect both insulated beams and columns by imposing direct pressure on the SFRM applied on the steel surface. This type of loading can lead to sign ificant local damage to SFRM as well as plastic M p SFRM s >> y Strain compatibility becomes very hard to maintain Strain distribution Moment Curvature g a) Steel moment frame subjected to seismic excitation b) Strain distribution in insulated steel section c) Moment curvature behavior during cyclic loading Steel Plastic hinge region 11 d eformation at both ends and mid - span of the member causing further indirect delamination from the member s . In modern high - rise buildings, not only the beams and beam - columns are prone to experience fire insulation damage, but also long steel truss system s are susceptible to encounter insulation damage du e to direct debris impact or blast overpressure . Long span steel trusses are commonly utilized in the floor assemblies of high - rise buildings thereby acco mmodating large open spaces without any interruption from columns. These floors, while supporting their self - weight along with dead and live loads, provide lateral stability to the exterior walls and columns and distribute wind load among the exterior wall s. The steel truss members are usually encapsulated with SFRM to achieve required fire resistance in floor systems. The truss members, due to high slenderness and small cross sectional sizes , are flimsy and thus are more vulnerable to insulation damage as compared to heavy steel sections utilized for columns and beams. Further, a truss system covers a larger area than columns and beams, hence the probability of debris impact on truss members during explosion or impact scenarios are higher than that in co lumns and beams. In the event of explosion , as illustrated in Error! Reference source not found. , generated blast overpressure can substantially increase the internal strains in truss members endangering the integrity of SFRM applied on truss members. In addition, under the action of impact loading, numerous trusses can completely fail le aving adjoining trusses heavily overloaded. The above explained loading scenarios will be considered in studying delamination of SFRM from steel structures in this research. The fracture and delamination of SFRM will be investigated on steel moment - resisting frames subjected to seismic loading on flimsy truss member s subj ected to extreme deformation, on beam s subjected to impact loading and on beam - column s subjected to blast overpressure. 12 Figure 1 . 4 Blast load on long steel truss and beam - column members 1.5 Mechanisms of F racture and D elamination of SFRM Spray applied fire resistive materials, as c ementitious materials , can be considered as two - phase composites comprising of a homogeneous phase and a particle phase (Modeer, 1979). Hence, in cementitious SFRM, the matrix (homogeneous phase) is composed of hydrated cement gels or gypsum paste , and the vermiculite particles (particle phase) form the reinforcement. This way the fracture properties of SFRM can be taken to be the average of individual properties of the two phases and the interfacial bond between the phases (Cotterell and Mai, 1996). Cl ose examination of material constituents of SFRM reveals that nearly 70 percent of SFRM is composed of Concrete floor (thickness of 101.6 mm) 18.29 m 1016 mm 736 mm Explosion Blast over pressure P(t) t Strain t Axial strain at bottom chord P max t Mid - span deflection a a Section a - a SFRM Truss cross section 13 gypsum or cement , both of which are cementitious materials . Therefore , static and dynamic fracture mechanics of SFRM is expected to be analogous to the o ne developed for a cementitious material. For example, ingredient s of a frequently utilized gypsum - based SFRM, which is known as C AFCO 300, are provided in Table 1 . 1 . Also, composition of a frequently utilized Portland cement - based SFRM, known as C AFCO 400, is listed in Table 1 . 1 . Table 1 . 1 Material ingredients of CAFCO300 and CAFCO400 Chemical name Weight % CAFCO300 CAFCO400 Portland cement - 40 - 70 Calcium Sulfate, Hemihydrate 50 - 75 - Vermiculite 15 - 35 15 - 40 Cellulose 1 - 10 - Calcium Carbonate 1 - 10 10 - 30 Quartz 0 - 5 <1 The l oading scenarios il lustrated in Figure 1 . 3 and Figure 1 . 4 can lead to crack initiation and propagation at the interface of SFRM and steel. This crack initiation and propagation phenomenon can be explained using fracture mechanics principles developed for cementitious materials (Cotterell and Mai, 1996). Figure 1 . 5 depicts a typical vicinity of crack at steel - SFRM interface and associated fracture process zone (FPZ) developed at the crack tip. Within the FPZ, microcracking and debonding between the homo geneous phase and the particle phase occurs causing strain - softening behavior in this zone. Delamination is initiated when the cohesive stress at the SFRM - c ) and subsequently progresses until the cohesive stress reaches zero value, the point at which delamination is completed. It should be noted that, stress - displacement relationship (cohesive laws) over the FPZ is one of the essential input to fracture mechanics - based numerical models. Hence, determination of theses cohesive 14 laws (i.e. the stress - displacement relationships over FPZ) is one of the primary objectives in this research. Figure 1 . 5 Progression of cracks leading to delamination of SFRM from steel surface (Development of fracture process zone) During dynamic impulsive loading conditions (blast or impact), two additional factors play a crucial role on the crack formation and its development within the bulk SFRM, as well as at the interface of SFRM and steel (i.e. over t he FPZ) . These factors are so - called, strain rate dependency of material behavior and structural inertia forces. The effect of former factor is related to rate dependency of initiation and propagation of micro - cracks in SFRM and can be explained using frac ture kinetics theory (Krausz and Krausz, 1988), whereas, influence of latter factor is associated with significant variations in state of strains and stresses in the material. According to fracture kinetics theory, micro - crack growth is dominated by activa tion energy. For an inherent micro - crack at SFRM - steel interface, even when the structure is at rest (no load Cohesive zone tip Potential crack propagation path Steel SFRM f Cohesive surfaces c Traction free crack surfaces T T Tensile stresses Fracture process zone (FPZ) SFRM 15 applied), bond - breaking and bond - healing processes occur at atomic level resulting in forward and backward movement of crack - tip line, respectivel y. However, at macroscopic level, no net change in the crack size is observed since crack - tip progression and shrinkage occur at the same frequency. When the structure is subjected to loading, the number of bond - breaking steps surpasses the number of bond - healing steps due to external energy supply, leading to crack progression in macro scale. If the load is applied within a very short time (as in the cases of impact or blast), since the number of bond - breaking phases is assumed to be constant in time for a given material, the total number of excessive bond - breaking phases will be smaller than the case when the applied load is quasi - static. As a result, the apparent cohesive strength as well as cohesive critical fracture energy of material (SFRM) will be h ig her (Krausz and Krausz, 1988).That means, during high strain rate loading conditions the fracture properties of steel - SFRM interface is expected to enhance. However, the enhancements observed in the material fracture properties owing to the effects of rate - dependency of material should always be distinguished from the In this research, it is attempted to characterize the cohesive stress - displacement relationship over the FPZ developed at steel - SFRM interface, and also estimate the effect of high loading rate on these fracture properties. Developing this knowledge will make it possible to explore the delamination of fire insulation from steel structures subjected to various loading scenarios. 1.6 Consequences of F ire I nsulation D elamination The consequences of SFRM delamination from steel structural elements can be significantly sever e . In moment - resisting frames subjected to seismic loading , damage to fire insulation over 16 the plastic hinge zone in beams can lead to high heat transfer to beam and significantly diminish the beam capacity, which can result in excessive deformation of beam. More importantly, delamination of fire insulation on beams opens a path for heat to be transferred to adjacent columns , which otherwise would remain less prone to heat penetration . This can significantly affect the column capacity during post - earthquake fire (Braxtan and Pessiki, 2011 b ). Further, b eams and beam - columns in steel frames, when subjected to blast overpressure, can undergo extreme deformations. Consequently, lack of fire insulation on these structural elements during fire following the explosion, can accelerate the adverse effect of fire and thereby extremely jeopardize the structural stability of building. Beams can suffer very high deformation leading to centenary action and horizontal pull - in in columns. Global buckling in beam - columns is thus accelerated as a result of horizontal force applied from beams and also due to dir ect effect of temperature rise. In high - rise buildings , the floor asse mbly c an experience large deformations due to softening in steel truss over a very short time. This can jeopardize the stability of adjacent columns through centenary action in floor assembly and eventually the entire structural stability of building can b e compromised. For instance, the progressive collapse of WTC twin towers was partially attributed to loss of fire insulation resulting from high impact and blast loads (FEMA 2002, NIST 2005). This incident has led to a major debate with respect to the role of fire insulation on structural integrity and resiliency of high - rise buildings under extreme loading events (NIST, 2005). 1.7 Research Objectives Based on above discussion, it is clear that there is lack of understanding on the initiation and propagation of damage and delamination in fire insulation applied on steel structures during static, cyclic and impulsive loading as encountered during service condition s , earthquake , impact 17 and explosion, respec tively. The main aim of this research is to develop fundamental understanding on the fracture mechanics and delamination of fire insulation from steel structures. The knowledge gap described in the previous section will be filled by pursuing following obje ctives: - Carry out detailed state - out - the - art review on the delamination of fire insulation from steel structures and the consequences this phenomenon can impose on structures. - Conduct material level experiments to determine fracture properties at the inter face of steel and fire insulation. In particular, develop cohesive stress - displacement relationship over the fracture process zone at steel - SFRM interface. The results of these experiments will provide essential material property input for numerical models . - Perform drop weight impact test s on fire insulated beams to characterize the delamination of different types of fire insulation from steel members subjected to impulsive loads causing high stain rate. Using an experimental - numerical approach estimate th e effect of high loading rate on the fracture properties over the fracture process zone at steel - SFRM interface. - Develop a fracture mechanics - based numerical approach for modeling crack initiation and propagation at the interface of fire insulation and s teel structures. Subsequently, validate the developed numerical model by comparing the model predictions against test data at both material and structural levels. - Car ry out a set of parametric studies to identify critical factors governing delamination of fire insulation from steel structures subjected to seismic and blast loading. In doing so, quantify the extent of delamination over the structural members as a function of the governing factors. 18 - Define a new delamination characteristic parameter for fire insulation, which can account for all critical factors governing delamination through one parameter. Thereafter, relate delamination initiation limits and delamination extent on the structural mem bers to this parameter. - Develop a thermal - structural numerical model that can simulate effect of SFRM delamination on fire performance of steel structures during fire following earthquake, impact and blast loading scenarios. 1.8 Anticipated R esearch Impact Current approach is unable to rationally assess post - earthquake , post - impact and post - blast fire performance of steel structures, partly due to limited knowledge on delamination of fire insulation from steel structures subjected to such loading scenarios. Further, the performance of fire insulation in terms of its adhesion to steel surface is only evaluated based upon normal bonding stress. The proposed Ph.D. research aims to produce two main results. First, the proposed experimental - numerical approach init iates the application of fracture mechanics in evaluating delamination of fire insulation from steel structures in a practical scale. Second, since proposed study aims to identify the critical factors governing delamination phenomenon at steel - SFRM interfa ce, the outcomes of research can also be useful for those researchers who are attempting to rectify the drawbacks associated with current fire insulation by developing new fire insulation materials. Further, by relating the delamination initiation limits and extent of delamination at the critical location s o f structural elements to the new parameter, a more rational approach of differentiating among different fire insulation products and t heir application in different situations will be possible for practicing engineers. 19 1.9 Scope and Outline The current research is carried out to achieve the above objectives, results of which are presented in seven chapters in this dissertation. Chapter 1 pr ovides the basic background with respect to issue of delamination of fire insulation from steel structures and its consequences. Chapter 2 details the state - of - the - art research on the fracture and delamination of fire insulation from steel structures where both experimental and numerical research results, as well as current code provisions are compiled and the knowledge gaps are underlined. Experimental program, encompassing static fracture tests and drop mass impact tests, are detailed in C hapter 3 and the outcomes are discussed. The fracture mechanics - based numerical model and its validation are outlined in Chapter 4 where implementation of fracture mechanics into the finite element model is outlined. In validation section of C hapter 4, predictions from th e numerical model are compared against data from experiments conducted in this research, as well as other studies. Chapter 5 deals with performance of three types of SFRM, widely utilized in steel construction, under static and dynamic loading. The paramet ric study, in terms of critical factors governing delamination, is presented in C hapter 5. Chapter 6 addresses the consequences of fire insulation damage and delamination from steel structures. Results obtained from thermal - structural analysis, when the st ructure is exposed to fire following extreme loading events, are detailed in this chapter. Eventually, Chapter 7 summarizes the major outcomes form this research and outlines the future potential research areas. 20 CHAPTER 2 2 STATE - OF - THE - ART REVIEW 2.1 General In current provisions, there is no methodology to account for the effect of delamination of fire insulation delamination on performance of steel structures during fire following earthquake, impact and blast. This is mainly because of limited studie s carried out, both at material and structural levels, on delamination and fracture mechanisms of fire insulation and its role on fire performance of steel structures. After collapse of world trade center in 2001 there were some initial studies on fracture properties of SFRM and ever since limited research results has been published. At material level, experiments have been focused on measuring normal bond strength, which is usually reported in material specifications. At structural level, the results have been limited to measuring extent of delamination over the structural elements (i.e. beam and column) subjected to quasi - static cyclic loading. There has been no numerical model developed for modeling delamination of SFRM from steel surface on a practical s cale. Further, there is no research, either experimental or numerical, on delamination of fire insulation from steel 21 structures during impact and blast loading. This section provides a state - of - the - art review on experimental and numerical studies with resp ect to delamination of fire insulation from steel structures. The current provisions in codes and standard are also reviewed. 2.2 Experimental S tudies The limited experimental studies carried out on fracture performance of fire insulation materials can be divi ded into two groups; tests carried out at material level and experiments conducted at structural level. 2.2.1 Material L evel T ests At material level, Chen et al. (2010) carried out tests to evaluate mechanical and interfacial properties of one type of SFRM, na mely YC3, including compressive strength, tensile strength, normal bond strength and shear bond strength. However, they did not measure load - displacement response at SFRM - steel interface and reported only maximum strength attained at fracture. The authors also carried out static tests on small scale specimens insulated with SFRM, namely tensile, compression and bending tests. Their results showed delamination of SFRM from steel surface under the applied loading. Figure 2 . 1 shows the experimental setup adopted by Chen et al. (2010) to measure the normal and shear bond strength at the interface of SFRM type YC3 and steel substrate. In test set - up for normal bonding strength, the SFRM was applied on a short T - shaped steel specimen, while another T - shape steel profile with the same size was glued on the top face of the SFRM. The specimen was positioned in the material testing machine w hile being clamped by upper and lower jaws of the machine and load was applied through the bottom jaw as is depicted in Error! Reference source not found. a. The normal bond strength was defined as the maximum 22 load attained during failure divided by the SFRM - steel interface area subjected to tensile stresses. In shear bond tests, a steel plate is sandwiched by two sets of SFRM - steel plate assembly which were glued to the central plate as shown in Error! Reference source not found. b. The load was applied on the central plate and the shear bond strength was defined as maximum load carried by the system divided by the SFRM - steel interface area subjected to shear stresses. The measured normal and shear bonding strength for this type of SFRM are 40 kPa and 70 kPa, respectively. The measured mechanical properties including density, elastic modulus, compressive strength and tensile strength were 550 kg/m 3 , 32.43 MPa, 590 kPa and 50 kPa, respectively. Figure 2 . 1 Normal and shear bond experiments carried out by Chen et al. (2010) a) Normal bonding test b) Shear bonding test 23 Braxtan and Pessiki (2011a) evaluated normal bond strength of SFRM types Cafco300 (wet - mix) and Blaze Shield II (dry - mix) through tests on small scale steel coupons. The steel plates, insulated with SFRM, were subjected to tensile yielding at various strai n ductility demands. Once a certain strain level was attained, the plates were unloaded. Subsequently, normal bond tests were performed on the SFRM, and thereby degradation of the bond strength at SFRM - steel interface as a function of tensile yielding in s teel was evaluated. They also studied the effect of surface mill finish of steel on normal bond performance. However, they did not provide any load - displacement response at SFRM - steel interface. The experimental set up used by Braxtan and Pessiki (2011a) a nd the relationship between average normal bond strength versus average strain on plate is illustrated in Figure 2 . 2 . Based on these tests, Braxtan and Pessiki (2011a) reported that when SFRM is applied on steel that has mill scale, the adhesive strength of the SFRM degrades rapidly once the steel yields. They attributed the rapid degradation of the adhesive strength to the debonding of the mill scale from the steel as the steel yields. They concluded that the normal bond strength is three times higher for Cafco300 than the normal bond strength for Blaze Shield II. Also, in tensile tests carried out on steel plates covered with fire insulation, they fo und that SFRM can detach from the steel plate after loading beyond yield. Further, delamination of SFRM was more prevalent in the plates sp r ayed with Blaze Shield II than in the plates sprayed with Cafco300. 24 Figure 2 . 2 Normal bond experiments performed by Braxtan and Pessiki (2011a) a) Test setup and specimens b) Bond strength vas strain level in steel substrate WM - SB: Wet mix on sand blasted plates WM - M: Wet mix on normal plates DM - SB: Dry mix on sand blasted plates DM - M:Dry mix on normal plates 25 The above discussed tests reported in the literature are strength - based and thus they do not address the effect of interfacial cracks on bond performance. Tan et al. (2011) proposed a new test me thod for measuring adhesion of SFRM on steel to overcome some of the current limitations in ASTM E736 (2006) for characterizing the SFRM - steel bond performance. This test method is based on linear elastic fracture mechanics (LEFM) approach and assumes pre - existing flaws at SFRM - steel interface. Figure 2 . 3 illustrates the schematics of single - arm cantilever test specimen utilized by Tan et al. (2011) in their fracture experiments. In the tests, field conditions were simulated for the application of SFRM on steel. While holding the SFRM in - place, the end of s teel coupon was peeled - off with a constant displacement rate of 0.1 mm/s and the corresponding applied load was measured. Series of loading and unloading cycles were simulated to study the relation among fracture energy and initial crack length. The record ed load - displacement curves for different initial crack sizes are shown in Figure 2 . 4 . Figure 2 . 3 Experimental setup designed by NIST to measure fracture energy at steel - SFRM interface (Tan et al. (2011)) SFRM Thickness of SFRM Steel substrate 5 mm 25 mm Non - bonded starter crack Steel substrate 25.4 mm 210 mm P (a) SFRM - steel substrate assembly Hole for attaching to test machine (b) Steel substrate dimensions 26 Tan et al. (2011) adopted two different approaches to deduce the fracture energy; an analytical solution based on theory of beam on elastic foundation and an experimental compliance calibration method. Results from these two approaches were in good agreement. The measured critical fracture energy for gypsum - based SFRM was in the range of 2 J/m 2 to 6 J/m 2 , while this was between 6 J/m 2 to 12 J/m 2 for compositely reinforced fibrous SFRM. It should not be overlooked that this test method only takes into consideration the critical fracture energy in normal fracture mode for evaluation of delamination at interface and does not take into account the frictional mode in evaluating delamination. Further, as will be outlined later, the application of linear elastic fracture mechanics for cementitious materials, which develop a large fracture process zone at cra ck tip, is not accurate. Figure 2 . 4 Load - displacement curves for different initial crack size measured by Tan et al. (2011) in their fracture tests Zhang and Li (2014) introduced a fire - resistive engineered cementitious composite (FR - ECC) to address the current issue of lack of durability (adhesion and cohesion) of SFRM on steel structures. In particular, they studied the effectiveness of employing ac rylic polymer latex as 27 admixtures and interfacial adhesive to enhance interfacial fracture properties of FR - ECC at its interface with steel substrate. The interfacial fracture resistance was evaluated by utilizing a fracture test proposed by Tan et al. (20 11). Based on the measured critical fracture energy between latex modified FR - ECC matrix and steel, they reported that using latex as admixture and interfacial adhesive can efficiently improve the interfacial critical fracture energy at the inte rface of FR - ECC and steel by 54% and 147 %, respectively. Further, they attributed the enhanced adhesive properties to the change in composition and microstructure of interfacial transition zone (ITZ) between latex modified FR - ECC matrix and steel. 2.2.2 Structural L evel T ests At structural level, Braxtan and Pessiki (2011b) studied damage pattern in SFRM applied on a beam - column assembly subjected to quasi - static cyclic loading through large - scale experiments, where the cyclic loading represented a strong seismic event. Su bstantial damage of SFRM in bottom and top flanges and partial damage of SFRM in web of beam were observed. Figure 2 . 5 illustrates the overall geometr y and member sizes for the beam - column assembly connection. In moment resistant steel frames subjected to lateral forces, inflection points form at the mid - height of the columns and at the mid - span of the beams. An exterior beam - column assembly was tested by Braxtan and Pessiki (2011b) and inflection points were simulated through attaching the column to a reaction wall by pin supports. A vertical load was applied at the beam tip. Lateral torsional buckling in beam was prevented by providing enough lateral s upports. This beam - column assembly was subjected to the cyclic displacement - controlled loading protocol as per ATC procedure ( FEMA 461 , 2007 ). Based on their cyclic monotonic tests, Braxtan and Pessiki (2011b) reported that at story drift of 3% and 4 %, SFRM damage is localized in the beam flanges where large inelastic deformation 28 and local instabilities occur. According to their observation for beam - column assembly insulated with SFRM type Blaze Shield - II, the SFRM on the beam web remained intact thro ughout the duration of the test. However, it was found during the post - testing inspection that the SFRM was delaminated over most of the beam web. In case of SFRM type Cafco300, the extent of delamination over the flanges is less as compared to SFRM type B laze Shield - II. Figure 2 . 6 shows the delamination of SFRM type Blaze Shield - II from bottom flange of the beam. Figure 2 . 5 Test setup of an exterior beam - column assembly to measure delamination of SFRM (Braxtan and Pessiki (2011b) Wang et al. (2013) conducted experiments to investigate failure pattern of SFRM type YC3 applied on steel cantilever columns under large quasi - static cyclic moments, induced at the bottom of the steel column. Figure 2 . 7 depicts the experimental set - up designed and used by Wang et al. (2013). They concluded that adhesion of SFRM to steel remains weak so that noticeable delamination occurs under large moments. In addition, they also inferred that cyclic L beam =4.42 m Inter - story Drift Angle H =3.05 m Beam: W 24X55 , Column: W 12X20 Shear Tab plate: PL 508x127x12.7 mm Continuity plate: PL 275x76x10 mm Web doubler plate: PL 933x241x6 mm a) Exterior beam - column assembly b) Steel moment frame 29 loading intensifies the extent of damage owing to damage accumulation effects. However, the effect of damage accumulation has not yet been quantified. The observed damage and delamina tion of fire insulation in their tests is shown in Figure 2 . 8 . Figure 2 . 6 Delamination of SFRM type Blaze Shield - II from bottom flange of the beam - column assembly tested by Braxtan and Pessiki (2011b) Figure 2 . 7 Test set - up for fire insulated column test (Wang et al. (2013)) Reaction frame Column and its foundation 30 Figure 2 . 8 Debonding and fracture of SFRM from steel column at high levels of quasi - static load (Wang et al. (2013)) 2.3 Numerical S tudies Most of the previous numerical studies focused on studying the effect of partial loss of fire insulation on the fire resistance of steel structural members. In these studies, damage mechanism in SFRM and causes of interfacial delamination of fire insulatio n from steel surface were not taken into consideration. However, the consequences of arbitrary insulation loss were quantified. Tomecek and Milke (1993) studied the effect of partial loss of fire insulation from flange and web of steel columns on the fire resistance of the columns using computer program FIRES - T3. The authors carried out 2D thermal analysis to compute the temperature evolution over the cross section of the column. No structural analysis was performed; instead, the average steel temperature over the cross section was used to determine the time to failure of the columns based on the criteria outlined in ASTM E119 (2014) . Based on their analysis, Tomecek and Milke (1993) found that fire resistance of steel column can appreciably decrease in cas e of insulation loss and the level of reduction depends on the extent of insulation loss, the size of column and Peel off Vertical crack Vertical crack 31 the position of protection loss. The fire resistance degradation of columns depends on the initial fire - rating of the columns such that columns with higher fire - rating undergo higher fire resistance reduction. For instance, 2% insulation loss on a one - hour - rating and three - hour - rated column (W10X49) leads to 10% and 28% decrease in fire resistance of the column, respectively. Further, columns wit h heavy sections experience less reduction in fire resistance compared to columns with small sections. Figure 2 . 9 depicts the finite element model al ong with the fire resistance reduction curves for W10X49 column. Ryder et al. (2002) also investigated the reduction in the fire resistance of steel columns due to the loss of SFRM directly on the column using FIRES - T3 computer program. They performed 3D t hermal analysis to predict temperature distribution within the column over the time. The computed thermal field is used in conjunction with thermal endpoint criteria specified in ASTM E119 to estimate the fire resistance of the column. Their results showed that insulation loss, though to a very small extent, can significantly influence the fire resistance of a steel column. Further, they concluded that the reduction in fire resistance is mainly affected by the extent of insulation loss rather than the size of the column. The schematics of missing SFRM from flange and web of the column, along with the temperature time - history at exposed flange and web, is shown in Figure 2 . 10 . Kwon et al. (2006) investigated the effect of SFRM removal from both web and flange of a steel column. They utilized A baqus software to conduct thermal and structural analysis. Based on their numerical results, they concluded that the loss of even small amount of SFRM caused a reduction in strength of the column and the consequences of SFRM removal from the flange was found to be more severe than the removing the SFRM from the web. 32 Figure 2 . 9 Effect of partial loss of fire insulation on fire resistance of steel columns (Tomecek and Milke (1993)) a) Fire insulation loss from flange c ) Fire resistance of W 10X49 versus percentage loss of fire insulation from flange b ) Fire insulation loss from web d ) Fire resistance of W 10X49 versus percentage loss of fire insulation from flange and web 33 Figure 2 . 10 Effect of partial loss of fire insulation on fire resistance of steel columns (Ryder et al. (2002)) a) Missing SFRM in flange b ) Missing SFRM in web c ) Temperature at exposed flange surface of W 6X16 column d ) Temperature at exposed web surface of a W 6X16 column 34 depicted in Figure 2 . 11 along with the temperature evolution over the time in different locations of the cross section. The reduction in structural capacity of the column as a functio n of fire duration, for different fire insulation missing scenarios is shown in Figure 2 . 12 Gu and Kodur (2011) carried out parametric studies on s ix - story steel - framed building to illustrate the effect of insulation damage on fire response of a steel structure. In their analysis, realistic fire scenarios, loading, and failure criteria were taken into consideration. Figure 2 . 13 shows the steel frame considered in the analysis along with the insulation damage pattern, deformation of frame for 10% insulation damage a nd the fire resistance reduction of the frame as a function of insulation damage percentage. Based on their analysis results, they concluded that the fire resistance of a steel - framed structure is significantly influenced by the extent of insulation loss, type of fire scenario, and level of lateral load. Gu and Kodur (2011) also highlighted that the insulation damage can result in faster deterioration in the structural response of framed buildings under the combined effect of fire and lateral loading. Dwaik at and Kodur (2012) developed a simplified approach for predicting temperature rise in steel sections with locally damaged fire insulation and validated their approach against numerical simulations of ANSYS finite element software. Based on the fire resist ance analysis on a W14x145 steel column, they showed dramatic reduction in plastic capacity of column due to 5% loss of 25 mm applied SFRM insulation, as shown in Figure 2 . 14 . The fire resistance of this column decreased from 180 minutes to 90 minutes due to 5% loss in fire insulation. Keller and Pessiki (2012) conducted an analytical case study to evaluate the effect of SFRM delamination patterns obser ved in experiments carried out by Braxtan and Pessiki (2011b) on thermo - mechanical response of steel moment beam - column assembly during post - earthquake 35 compartment fire exposure. Figure 2.15 illustrated their finite element model developed in A baqus software and the moment - rotation response of the beam - column connection after being exposed to fire scenarios with different duration. As is clear in Figure 2 . 15 , significant temperature - induced softening occurs in moment - rotation response of beam - column connection, and as a consequence, flexibility of the structural system for sideway motion is increased resulting in intensified drift demands under th e action of residual post - earthquake destabilizing forces. Dwaikat and Kodur (2011) performed 2D finite element analysis adopting a cohesive zone approach to model spontaneous initiation and propagation of delamination at SFRM - steel interface under static and impact loads. They studied delamination under three loading cases, including pure tension, pure bending moment and drop mass at the tip of a cantilever beam. Figure 2 . 16 depicts schematics of the models analyzed in ANSYS software and the delamination percentage under tensile loading condition. They concluded that interfacial tensile stresses at SFRM - steel interface are lower in case of thin layers of insulation and also thickness of SFRM can be optimized with respect to impact energy. 36 Figure 2 . 11 Temperature distribution over the cross section of steel column as a consequence of missing fire insulation from flange (Kwon et al. (2006)) b) Temperature time - history for l p =b f /16 (b f =flange width) c) Temperature time - history for l p =b f (b f =flange width) a) Finite element model of the column (W14X109) and the missing insulation form flange (l p ) 37 Figure 2 . 12 Capacity reduction of the column versus fire duration (Kwon et al. (2006)) a) Capacity reduction for different fire insulation missing scenarios b) Column deformation for l p =b f and fire duration of 90 min 38 Figure 2 . 13 Effect of fire insulation damage on fire resistance of a moment resisting frame (Gu and Kodur, (2011)) a) Steel frame analyzed b ) Assumed insulation damage c ) Ground floor of the frame exposed to fire d ) Deformation of steel frame under 10% insulation damage e ) Fire resistance reduction 39 Figure 2 . 14 Effect of fire insulation damage on fire resistance of a steel column (Dwaikat and Kodur, (2012)) a) Fire insulated steel column and missing insulation b ) Temperature rise in steel c ) Reduction in plastic capacity of steel column 40 Figure 2 . 15 Effect of fire insulation damage on moment - rotation response of beam - column connection (Dwaikat and Kodur, (2012a)) a) Finite element model of beam - column connection b ) Moment - rotation response of beam - column connection after being exposed to different fire scenarios 41 Figure 2 . 16 Numerical modeling of fire insulation delamination from steel surface (Dwaikat and Kodur, (2011)) a) Different loading cases b ) Delamination percentage as a function of loading t: fire insulation thickness t p : steel plate thickness 42 The above literature review shows that previous experimental and numerical studies, though provided valuable understanding on fire insulation delamination, have two major disadvantages. First, most of the previous researchers performed strength - based studies and hence did not adopt a fracture m echanics approach towards describing the cracking and delamination of fire insulation. Second, most of the experiments, both at material and structural levels, have been carried out under static and cyclic monotonic loadings. There has been no research on establishing dynamic and rate - dependent fracture properties of SFRM. Further, there have been no experimental and numerical studies on dynamic delamination of SFRM from steel structures subjected to high strain rate loading. 2.4 Codes of P ractice The durabilit y requirements for insulation materials are specified in codes and standards for buildings. Also, there are some recent reports, which highlight the role of critical properties of insulation materials in achieving satisfactory fire performance of steel str uctures. A number of ASTM tests are currently used to gauge the durability and integrity of SFRM under normal life of structure; during construction process; and under extreme conditions (such as earthquake and severe fires). A major drawback of most of th ese tests is that they are not fundamentally linked to materials science (Bentz et al., 2009) and they do not measure many of critical engineering parameters that are necessary for understanding the mechanics of SFRM under severe loading conditions, such a s fracture energy and the debonding stresses. For instance, the current method for testing the cohesive/adhesive properties of SFRM, ASTM E736 (2011), consists of a disk with a hook for hanging a weight, and that disk is attached to the SFRM through a qui ck setting adhesive. The SFRM material must withstand a minimum weight 43 before it is dislodged. The weakness of this method is that it provides only one value of failure load without any distinction whether the failure is due to poor adhesion, or poor cohes ion. Figure 2 . 17 illustrated the test method prescribed in ASTM E736 for measuring bond strength between fire insulation and steel substrate. ASTM E760 (2011) is another standard, which specifies a test method for evaluating the SFRM performance under impact loads. This standard requires that no visible cracks or spalling of the SFRM should be observed when it is subjected to the following prescribed impact test. The impact test is performed using an impactor of leather bag with mass of 27.7 kg dropped from a height of 1.2 m on the middle of a 3.6 m free span insulated cellular steel deck with concrete topping, as shown in Figure 2 . 18 . As obvious, the impact simulated in this test is comparable to opping heavy objects on floors, and thus, the prescribed test does not represent severe impacts that would result from blast or earthquake loading. The performance of SFRM under service deflection is assessed by the ASTM E759 (2011) standard. A steel deck - concrete slab assembly, similar to that used in ASTM E760 standard mentioned above, is also used in this standard. A point load is applied at the center of the slab assembly with the insulation applied at the bottom surface (tension side) of the steel deck . The SFRM is deemed to satisfy the test if cracks or dislodging due to the induced deflection is not observed until a deflection limit of L /120 is reached, as illustrated in Figure 2 . 19 . 44 Figure 2 . 17 Normal bonding test between steel and fire insulation based on ASTM E736 (2011) 45 Figure 2 . 18 Drop mass test for measuring durability of fire insulation under accidental impact loading based on ASTM E760 (2011) Figure 2 . 19 Point load test for measuring durability of fire insulation under service loading conditions based on ASTM E759 (2011) 46 Eurocode 3 (2010) does not give any specific requirements for the durability characteristics of insulation and coating materials, including their maintenance ar e not given in Eurocode 3, of a structural member is subjected to impact damage from mov ing vehicles, the handling of merchandise or other activity, the fire protective covering shall be protected by corner guards or by a substantial jacket of metal or other noncombustible material to a height adequate to provide tes that fire insulation in structures that are susceptible to extreme loading events, should be protected to avoid any damage. Recent reports by NIST (2005) and Federal Emergency Management Agency (FEMA 2002) on the collapse of WTC buildings highlighted the need for satisfactory fire insulation performance insulation occurred not only in locations where direct debris impact happened, but also in perimeter columns NIST recommends the development of appropriate criteria, test methods and standards: i) for the in - service performance of SFRM used to insulate steel structural components; and ii) to ensure that these materials, as installed, confirm to conditions in tests used to establish the fire resistance rating of components, assemblies, and systems. In addition, FEMA report on WTC building performance study concludes that the performance of spray - applied fire protection material played a crucial role in the collapse of twin towers (WTC). It also concludes that, adhesion and cohesion characteristics of SFRM are not well understood, and that there is an urgent need for developing performance ba sed requirements for SFRM. 47 Based on recent recommendations of NIST (2005), U.S. General Services Administration (GSA , 2010 ) has introduced updated provisions for the use of robust fireproofing materials in steel framed buildings. The proposed provisions re quire fireproofing materials to have bond strength of 20.6 kPa for buildings below a height of 128 m, and 47.9 kPa for buildings above a height of 128 m. Also, based on these recommendations by NIST, amendments were made in the IBC (2012) code to increase the bond strength for fireproofing by nearly three times greater than currently required for buildings 75 - 420 feet in height and seven times greater for buildings more than 420 feet in height. 2.5 Knowledge Gaps Based on the above review, it is clear that t here is limited data on mechanical properties of SFRM, especially fracture properties. Also, there is lack of understanding on the initiation and propagation of damage and delamination in fire insulation applied on steel structures during cyclic and impuls ive loading as encountered during earthquake, impact and explosion, respectively. Further, there is lack of numerical models to predict the delamination phenomenon at SFRM - steel interface in steel structures subjected to seismic, impact and blast loading. Hence, further research is needed in following key areas: - Fracture properties need to be determined in mode - I fracture and mode - II fracture for different types of SFRM commonly applied on steel structures. These experiments can not only deliver bond streng th but also provides fracture toughness and fracture ductility over the fracture process zone. This data will provide core input to the numerical models dealing with delamination of SFRM from steel structures in practical scale. - A numerical approach for mo deling delamination of fire insulation from steel structures subjected to seismic, impact or blast loading conditions on practical scale is not reported 48 in literature. Such a numerical model, once validated against experiments at material and structural le vel, can be used to carry out extensive parametric studies to identify critical factors governing delamination of fire insulation from steel structures subjected to seismic, impact or blast loading. - Delamination of fire insulation from steel structures su bjected to blast loading is not studied literature. In order to investigate this issue, experimental study needs to be carried out either by directly exposing the insulated steel elements to blast overpressure or conducting impact tests (drop mass tests) t o generate a high stain rate field similar to the one expected during explosion events. - In current practice, normal bonding strength, along with density, are the only mechanical properties which are reported in material specifications and there are standa rd test methods to measure these properties. However, there are other factors, namely tangential bond strength, normal critical fracture energy, tangential critical fracture energy, elastic modulus and SFRM thickness that can influence initiation and propa gation of cracks at the interface of fire insulation and steel surface. A more rigorous parameter is needed to account for all critical factors governing delamination phenomenon by maintaining interdependency among different factors. - In previous numeric al studies, the effect of fire insulation damage has been accounted for by arbitrarily choosing the location and amount of missing insulation. However, a more realistic evaluation of fire performance of steel structures during fire following earthquake, or impact or explosion is to be carried out in which the delamination extent is adopted from the results of fracture mechanics - based numerical model. Hence, the fracture mechanics - based numerical model should be combined with a thermal - structural 49 model to si mulate the effect of extreme loading and subsequent fire, sequentially. This type of analysis has not been performed thus far. - The above stated knowledge gaps are to be overcome to enhance the understanding on fire insulation delamination from steel struct ures and also to develop a fracture mechanics - based approach to study the effect of critical parameters on initiation and progression of delamination of fire insulation from steel structures subjected to static and dynamic loading conditions. This disserta tion is designed to undertake required studies for overcoming the above knowledge gaps. 50 CHAPTER 3 3 EXPERIMENTAL STUDY 3.1 General The previous experimental and numerical studies, though provided valuable understanding on delamination of fire insulation, have two major disadvantages. First, most of the previous researchers adopted strength - based test approaches and hence did not adopt a fracture mechanics approach for evaluating cracking and delamination of fire insulation. Second, most of the experiments, both at material and structural levels, have been carried out under static or cyclic monotonic loading. There has been no research on establishing dynamic and rate - dependent fracture properties of SFRM. Further, there have been no experim ental and numerical studies on dynamic delamination of SFRM from steel structures subjected to high strain rate loading. The experimental program undertaken in this study is divided into two parts. In the first part, the constitutive relations of SFRM over the fracture process zone, namely cohesive laws is determined using a series of static fracture tests. The fracture tests are conducted on three types of SFRM commonly utilized in current buildings. In the second part, drop mass impact tests are 51 carried o ut to investigate the dynamic delamination of SFRM from steel beams, insulated with the very three types of SFRM, under impulsive loading conditions. In this chapter, first the test procedures adopted to determine fracture process zone properties along wit h the obtained results are presented. Subsequently, test procedures, designed and applied for performing a drop mass impact test on SFRM - insulated beams, with corresponding results, are described. 3.2 Determination of F racture P rocess Z one P roperties for SFRM As outlined in C hapter 1, to predict the crack propagation at SFRM and steel interface it is indispensable to establish the cohesive stress - displacement relationship over the fracture process zone (FPZ) at steel - SFRM interface, namely cohesive laws. In thi s section, test procedures adopted to determine these cohesive laws are presented. 3.2.1 Test P rocedures to E valuate C ohesive L aws over FPZ There are, in general, two approaches for obtaining stress - displacement relationship in FPZ of cementitious materials; dir ect approach and indirect approach. In direct approach, stress - displacement response is measured by means of a tension test for Mode I fracture (Peterson, 1985; Reinhardt, 1987; Gu o and Zhang, 1987). In this method, although pre - existing flaws start to gro w at discrete locations during initial stages of loading, localization of deformation occurs in the FPZ once the maximum load has been attained (Cotterell and Mai, 1996). Specimen dimensions must be large enough to accommodate full development of the FPZ a cross the area undergoing the tensile loading. Stress - displacement relationship obtained from a tension test can directly generate all three parameters of cohesive law namely, cohesive stiffness, cohesive strength and fracture energy. No further numerical work is therefore required for extracting cohesive laws over FPZ. However, the fracture evolution across the tensile area must be uniform; 52 - displacement curve (Hordijk et al., 1987). There are various indirect methods proposed in the literature for deriving stress - displacement measurement of J - integral and crack tip opening displacement (CTOD) to obtain the stress - displacement curve. In this method, two specimens have to be used which makes interpretation of results difficult due to inhomogeneous behavior of cementitious materials. Indirect approach has also extensively been used for composite mate rials and interface of two materials (Sorensen and Jacobsen, 2003; Gordnian et al., 2008; Lee et al. 2010; Valoroso et al. 2013). In recent years, Double Cantilever Beam (DCB) specimens (ASTM D5528, 2013) and End Notched Flexure (ENF) (ASTM WK22949, 2009) specimens are widely utilized to extract the fracture energy in pure modes I and II, respectively. However, fracture energy is the only outcome from these tests. Two other parameters of cohesive laws, namely cohesive stiffness and cohesive strength, are th erefore to be determined through numerical modeling. To extract theses parameters, an ideal stress - displacement curve is assumed and numerical simulation is carried out. The predicted overall load - displacement relationship is compared to the experimental b ehavior and this iterative process is repeated until the best agreement is obtained between experimental and simulation results. However, due to mesh sensitivity of cohesive solutions, the above explained computational effort can be quite significant and t he predicted cohesive zone properties may not be accurate. In fact, sensitivity analyses with respect to cohesive parameters may not be successful for some sample geometries (Alfano et al., 2011). With respect to fire insulation, DCB and ENF tests cannot b e used because SFRM does not contribute to structural capacity (strength) of SFRM - steel assembly. That is, delamination at 53 SFRM - steel interface will not cause any softening in the overall load - displacement relationship. Single Cantilever Beam (SLB) specime ns proposed by Tan et al. (2011) to measure fracture energy of SFRM in pure Mode I entails using a very thin steel substrate (0.35 mm) which may therefore does not account for the strain - softening in FPZ. Further, no testing procedure has so far been proposed for measurement of fracture energy in pure Mode - II at SFRM - steel interface. Therefore, direct approach is adopted in this study to establish cohesive laws for mode - I and mode - II delamination at steel - SFRM interface. 3.2.2 Materials and S pecimen G eometry For evaluating fracture - based cohesive properties, three types of commercially available SFRM that are commonly used in building applications, have been selected. The generic type of these three SFRMs is summarized in Table 3 . 1 . Figure 3 . 1 illustrates the overall plate geometry and laboratory. After 6 - weeks of curing, the specimens were carefully shipped to Michigan State cut to the desired dimensions. The clear space left between specimens is large enough to fit the clamps in between for constraining the specimen plate into t he testing machine. Tensile test specimens measured 76.2 x 76.2 x 25.4 mm and shear test specimens measured 101.6 x 25.4 x 25.4 mm. With respect to size of specimens, it was attempted to adopt as large specimens as possible to reduce the size effects and h ence generate as realistic data as possible which can be applicable in practice. The issue of size effect has been studied by Bazant (1984), Bazant and Kazemi (1990) and Bazant and Kazemi (1991) for concrete and rock. 54 Table 3 . 1 Three type of SFRM utilized in experiments Name Type of SFRM A Medium density gypsum - based B Medium density Portland cement - based C Mineral - fiber - based 3.2.3 Experimental S etup for F racture M ode - I A special test setup was designed for undertaking fracture tests on SFRM insulated steel plates to measure normal cohesive stress - displacement response. Details of the test specimens and the testing procedure designed for measuring fracture mode - I properties at steel - SFRM interface is depicted in Figure 3 . 2 a and Figure 3 . 3 a. A plywood block with thickness of 15 mm is carefully drilled at the center to which an eyebolt is screwed in. The wooden block is glued on top of tensile specimen using wood glue. After gluing wooden block to SFRM , wood surface is leveled and clamped to the steel plate to make a perfectly flat surface. After 24 - hours, the specimens are unclamped and prepared for testing. The test is carried out on an electromechanical material testing system (MTS) shown in Figure 3 . 3 a. The steel plate is clamped to an I - beam, which is connected to bottom actuator, to prevent deformation of plate during the test. The eyebolt is c onnected to the upper rigid block using a shackle - eye nuts - threaded rod assembly. Special care is taken to ensure that no eccentricity exist between MTS loading direction and specimen center. Displacement - controlled load is applied on the specimens and loa d - displacement relationship is recorded while the loading rate is kept constant at 1µm/sec. Test is terminated once the full fracture of SFRM occurs and specimen can no longer withstand any further load. 55 Figure 3 . 1 Test plate geometry for measuring fracture parameters a) 3D view of SFRM samples b) Test plate and SFRM sample dimensions Tensile test sample Shear test sample 56 Figure 3 . 2 Schematic of test assembly for determination of CZM parameters 3.2.4 Experimental S etup for F racture M ode - II Direct shear test is conducted to measure the stress - displacement response in fracture mode - II at the interface of SFRM and steel plate. The specimen details and testing method is illustrated in Figure 3 . 2 b and Figure 3 . 3 b. The SFRM block is pushed against the steel plate thereby inducing direct shear stresses at the interface of steel and SFRM. Fixed upper jaw Wooden block Wood glue Eye bolt s crewed into wood SFRM Steel substrate C - clamp I - beam Lower jaw Downward displacement control loading Upward displacement control loading Fixed upper jaw Steel plate SFRM Steel substrate Lower jaw a) Test schematic for model - I delamination b) Test schematic for model - II delamination Steel bar 57 Figure 3 . 3 Test set up designed for measuring CZM parameters at steel - SFRM a) Test set up for model - I delamination b) Test set up for model - II delamination 58 The width of the test specimen along loading direction was chosen to be small enough to preclude cohesive failure within the SFRM. A small gap was introduced be tween the loading plate and test specimen plate so that friction between two plates is eliminated. The test is carried out through a displacement control loading technique with a constant displacement rate of 1µm/sec. The stress - displacement recording is c ontinued until the SFRM block is fully delaminated from steel surface and the total applied load returns to zero value. 3.2.5 Elastic M odulus T ests The elastic modulus for three types of SFRM was determined by conducting compression tests on SFRM blocks of 50.8 mm x50.8 mm x50.8 mm size. Displacement controlled loading was applied with a constant displacement rate of 1µm/sec. The measured elastic modulus on three types of SFRM is listed in Table 3 . 2 . Table 3 . 2 Cohesive zone model parameters obtained in experiment for three types of SFRM SFRM type c (kPa) c (kPa) G cn (J/m 2 ) G ct (J/m 2 ) G ct /G cn K n (kPa/mm) K t (kPa/mm) n t E (MPa) Manufacture c range (kPa) A 22.9 49.6 2.2 7.9 32.8 4.2 57.3 107.9 1.73 2.98 11.5 7.2 - 20.5 B 52.8 107.3 2.0 33.7 74.4 2.2 57.4 162.6 1.40 2.11 38.4 20.8 - 409.6 C 13 24.6 1.9 4.3 22.5 5.2 39.3 61.4 2.03 4.63 2.6 7.2 - 17.9 3.3 Results from F racture T ests The force - displacement relationships recorded from tensile tests are shown in Figure 3 . 4 for three types of SFRM insulated specimens. It is apparent that the response of Portland cement - based SFRM is relatively brittle as compared to the gypsum - based and mineral fiber - based SFRM. For gypsum - based and mineral fiber - based SFRM types, interfacial force rises almost linearly to critical cohesive strength, and subsequently, decreases with increasing normal displacement. The softening behavior observed in the force - displacement curves, confirms that 59 the size of fully developed FPZ is noticeable for gypsum - based and mineral fiber - based SFRM. Even in the case of Portland cement - based SFRM, there is no rapid load drop as would be the case for elastic - brittle materials. For this SFRM type, f orce - displacement curve is nonlinear up to the peak load which is followed by a sharp drop to 20 percent of peak load as can be seen from force - displacement response in Figure 3 . 4 . Then the load response slowly diminishes to zero with further interfacial deformation. Ultimate fracture displacement in this case is higher than those obtained for gypsum - based and mineral fiber - based SFRM. In all specimen s, prior to reaching cohesive strength (peak load) there was no sign of crack development throughout the specimen. However, once damage is localized and FPZ formation starts, cracks started appearing in decay phase of force - displacement curve and get fully developed and visible upon reaching to the failure displacement. Delayed development of FPZ can be attributed to the fact that in plain tensile specimens no initial crack or notch is introduced. Consequently, this is no focus point for the formation of FP Z and thus dispersion of initial microcracking occurs. Since SFRM is substantially softer than steel (E SFRM =30 MPa<