Upload
phungxuyen
View
262
Download
0
Embed Size (px)
Citation preview
STATIC AND DYNAMIC PERFORMANCE
OF LIGHTWEIGHT HYBRID COMPOSITE
FLOOR PLATE SYSTEM
A.M. Chanaka Madushan Abeysinghe
B.Sc. (Civil Engineering, Honours)
A thesis submitted in partial fulfilment of the requirements for the degree of
Doctor of Philosophy
School of Civil Engineering & Built Environment
Science and Engineering Faculty
Queensland University of Technology
September - 2012
Static and dynamic performance of lightweight hybrid composite floor plate system i
Keywords
Hybrid composite floor plate system, Glass-fiber Reinforced Cement, Polyurethane,
Steel, Experimental testing, Finite Element modelling, Static performance, Dynamic
performance, Human-induced loads, Vibration, Design guidelines
ii Static and dynamic performance of lightweight hybrid composite floor plate system
Abstract
In the modern built environment, building construction and demolition consume a
large amount of energy and emits greenhouse gasses due to widely used conventional
construction materials such as reinforced and composite concrete. These materials
consume high amount of natural resources and possess high embodied energy. More
energy is required to recycle or reuse such materials at the cessation of use.
Therefore, it is very important to use recyclable or reusable new materials in building
construction in order to conserve natural resources and reduce the energy and
emissions associated with conventional materials. Advancements in materials
technology have resulted in the introduction of new composite and hybrid materials
in infrastructure construction as alternatives to the conventional materials. This
research project has developed a lightweight and prefabricatable Hybrid Composite
Floor Plate System (HCFPS) as an alternative to conventional floor system, with
desirable properties, easy to construct, economical, demountable, recyclable and
reusable. Component materials of HCFPS include a central Polyurethane (PU) core,
outer layers of Glass-fiber Reinforced Cement (GRC) and steel laminates at tensile
regions. This research work explored the structural adequacy and performance
characteristics of hybridised GRC, PU and steel laminate for the development of
HCFPS.
Performance characteristics of HCFPS were investigated using Finite Element (FE)
method simulations supported by experimental testing. Parametric studies were
conducted to develop the HCFPS to satisfy static performance using sectional
configurations, spans, loading and material properties as the parameters. Dynamic
response of HCFPS floors was investigated by conducting parametric studies using
material properties, walking frequency and damping as the parameters. Research
findings show that HCFPS can be used in office and residential buildings to provide
acceptable static and dynamic performance. Design guidelines were developed for
this new floor system. HCFPS is easy to construct and economical compared to
conventional floor systems as it is lightweight and prefabricatable floor system. This
floor system can also be demounted and reused or recycled at the cessation of use
due to its component materials.
Static and dynamic performance of lightweight hybrid composite floor plate system iii
Publications
Abeysingh, C.M., D.P. Thambiratnam and N.J. Perera, Development of an
Innovative Hybrid Composite Floor System, In Thirteenth International Conference
on Civil, Structural and Environmental Engineering Computing, B.H.V. Topping and
Y. Tsompanakis, Editors. 2011, Civil-Comp Press, Stirlingshire, Scotland: Chania,
Crete, Greece.
Abeysingh, C.M., D.P. Thambiratnam and N.J. Perera, Investigation of hybridized
polyurethane, glass fibre reinforced cement and steel laminate in structural floor
plate systems, In The First International Postgraduate Conference on Engineering,
Designing and Developing the Built Environment for Sustainable Wellbeing. 2011,
Queensland University of Technology. p. 249-253.
Abeysingh, C.M., D.P. Thambiratnam and N.J. Perera, Innovative Hybrid Composite
Floor Plate System In AES-ATEMA’2012 Tenth International Conference, Y.M.
Haddad, Editor. 2012: McGill University, Montreal, Canada.
Abeysingh, C.M., D.P. Thambiratnam and N.J. Perera, Flexural performance of an
innovative Hybrid Composite Floor Plate System comprising Glass–fibre Reinforced
Cement, Polyurethane and steel laminate. Composite Structures 2013, 95: 179-190.
Thambiratnam D.P. , Perera N.J. , Abeysinghe C.M. , Huang M-H ,De Silva S.S.,
Human Activity-Induced Vibration in Slender Structural Systems. Structural
Engineering International 2012, 22(2): 238-245.
Abeysingh, C.M., D.P. Thambiratnam and N.J. Perera, Dynamic performance
characteristics of an innovative Hybrid Composite Floor Plate System under human-
induced loads. Composite Structures (2012),
http://dx.doi.org/10.1016/j.compstruct.2012.09.015.
iv Static and dynamic performance of lightweight hybrid composite floor plate system
Table of Contents
Keywords ................................................................................................................................................. i
Abstract ................................................................................................................................................... ii
Publications ........................................................................................................................................... iii
Table of Contents ................................................................................................................................... iv
List of Figures ..................................................................................................................................... viii
List of Tables ........................................................................................................................................ xii
List of Abbreviations ........................................................................................................................... xiv
List of Symbols .................................................................................................................................... xiv
Statement of Original Authorship ....................................................................................................... xvii
Acknowledgements ........................................................................................................................... xviii
CHAPTER 1: INTRODUCTION ....................................................................................................... 1
1.1 INTRODUCTION ....................................................................................................................... 1
1.2 HCFPS ......................................................................................................................................... 3
1.3 RESEARCH AIMS AND OBJECTIVES.................................................................................... 5
1.4 SIGNIFICANCE AND SCOPE ................................................................................................... 5 1.4.1 Significance ...................................................................................................................... 5 1.4.2 Scope ................................................................................................................................ 6
1.5 THESIS OUTLINE...................................................................................................................... 7
CHAPTER 2: LITERATURE REVIEW ........................................................................................... 9
2.1 CURRENT COMPOSITE CONSTRUCTION APPLICATIONS AND THEIR LIMITATIONS9 2.1.1 Fiber Reinforced Composites (FRC) ................................................................................ 9 2.1.2 Sandwich Composites..................................................................................................... 10 2.1.3 Hybrid composites .......................................................................................................... 14
2.2 IMPORTANCE OF DEVELOPING A HYBRID COMPOSITE FLOOR PLATE SYSTEM .. 15
2.3 PROPERTIES AND ADVANTAGES OF SELECTED MATERIALS FOR HCFPS .............. 18 2.3.1 Steel 18 2.3.2 Polyurethane (PU) .......................................................................................................... 20 2.3.3 Glass fiber Reinforced Cement/Concrete (GRC) ........................................................... 22
2.4 SUMMARY AND IMPLICATIONS ........................................................................................ 24
CHAPTER 3: METHODOLOGY .................................................................................................... 27
3.1 HCFPS DEVELOPMENT ......................................................................................................... 27
3.2 HCFPS CONFIGURATION AND SYMMETRY .................................................................... 27
3.3 EXPERIMENTAL TESTING ................................................................................................... 29 3.3.1 Test Panel ....................................................................................................................... 29 3.3.2 Heel impact test .............................................................................................................. 30 3.3.3 Static load test................................................................................................................. 31 3.3.4 Cyclic loading test .......................................................................................................... 32 3.3.5 Material testing ............................................................................................................... 34
3.4 FINITE ELEMENT (FE) METHOD OF ANALYSIS .............................................................. 35 3.4.1 Pre-processing ................................................................................................................ 35 3.4.2 Analysis and solutions .................................................................................................... 35 3.4.3 Post-processing ............................................................................................................... 36
Static and dynamic performance of lightweight hybrid composite floor plate system v
3.5 FE MODEL VALIDATION ...................................................................................................... 36
3.6 STATIC PERFORMANCE AND PARAMETRIC STUDIES .................................................. 37
3.7 DYNAMIC ANALYSIS OF HCFPS......................................................................................... 38 3.7.1 Human induced walking loads ........................................................................................ 38 3.7.2 Dynamic analysis and floor classification ...................................................................... 40 3.7.3 Acceptance criteria ......................................................................................................... 41 3.7.4 Damping ......................................................................................................................... 44
3.8 DYNAMIC PERFORMANCE AND PARAMETRIC STUDIES OF hcfps ............................. 45
3.9 DESIGN GUIDELINES ............................................................................................................ 45
3.10 SUMMARY ............................................................................................................................... 46
CHAPTER 4: EXPERIMENTAL TESTING .................................................................................. 47
4.1 MATERIAL PROPERTY INVESTIGATION .......................................................................... 47 4.1.1 Formulation of GRC ....................................................................................................... 47 4.1.2 Tensile testing for GRC .................................................................................................. 48 4.1.3 Four-point bending tests for GRC .................................................................................. 48 4.1.4 Cylinder compression test for GRC ................................................................................ 49 4.1.5 Composition of PU ......................................................................................................... 50 4.1.6 Compression test for PU ................................................................................................. 50 4.1.7 Tensile test for PU .......................................................................................................... 51 4.1.8 Three- point bending tests for PU core ........................................................................... 52 4.1.9 Tensile test for steel laminate ......................................................................................... 53
4.2 FABRICATION OF HCFPS TEST SPECIMEN ...................................................................... 53
4.3 SUPPORTING ARRANGEMENT FOR THE TEST PANEL .................................................. 55
4.4 DATA ACQUISITION SYSTEM ............................................................................................. 56
4.5 DYNAMIC TESTING ............................................................................................................... 57 4.5.1 Heel impact test .............................................................................................................. 57 4.5.2 Walking test .................................................................................................................... 61
4.6 STATIC LOAD TESTING ........................................................................................................ 62 4.6.1 Test set up, instrumentation and static load test.............................................................. 62 4.6.2 Static load test results and discussion ............................................................................. 63 4.6.3 Deflection ductility ......................................................................................................... 65
4.7 CYCLIC LOAD TESTING ....................................................................................................... 65 4.7.1 Test set up, instrumentation and cyclic load test ............................................................ 66 4.7.2 Test results and discussion for cyclic loading test .......................................................... 67 4.7.3 Comparison of cyclic loading test results with static load test results ............................ 67
4.8 TESTING OF GRC-PU-GRC COMPOSITE PANEL .............................................................. 68 4.8.1 Test sample size and fabrication ..................................................................................... 68 4.8.2 Test set up and instrumentation ...................................................................................... 69 4.8.3 Static load test results and discussion ............................................................................. 70
4.9 SUMMARY ............................................................................................................................... 72
CHAPTER 5: DEVELOPMENT AND VALIDATION OF FE MODELS ................................... 75
5.1 FE MODEL DEVELOPMENT AND VALIDATION USING DYNAMIC TEST RESULTS
OF HCFPS TEST PANELS .................................................................................................................. 75 5.1.1 Material properties for dynamic analysis ........................................................................ 75 5.1.2 Model description ........................................................................................................... 76 5.1.3 Free vibration analysis and validation with first natural frequency ................................ 76 5.1.4 Linear transient dynamic analysis ................................................................................... 77 5.1.5 Heel impact load function ............................................................................................... 77 5.1.6 Application of the damping to FE models ...................................................................... 78 5.1.7 Dynamic analysis validation with acceleration response of heal impact test .................. 79 5.1.8 Dynamic analysis and validation with walking loads ..................................................... 80
vi Static and dynamic performance of lightweight hybrid composite floor plate system
5.2 FE MODEL DEVELOPMENT AND VALIDATION USING STATIC TEST RESULTS of
HCFPS TEST PANEL .......................................................................................................................... 82 5.2.1 Material properties for static analysis ............................................................................. 82 5.2.2 GRC Material models ..................................................................................................... 82 5.2.3 4 PU material model ....................................................................................................... 83 5.2.4 Steel material model ....................................................................................................... 83 5.2.5 Model description ........................................................................................................... 83 5.2.6 Static analysis, validation and discussion ....................................................................... 84
5.3 FE MODEL DEVELOPMENT AND VALIDATION FOR GRC-PU-GRC COMPOSITE
PANEL .................................................................................................................................................. 87 5.3.1 FE model ........................................................................................................................ 87 5.3.2 Material properties .......................................................................................................... 88 5.3.3 FE model validation ....................................................................................................... 89
5.4 SUMMARY ............................................................................................................................... 90
CHAPTER 6: STATIC PERFORMANCE OF HCFPS ................................................................. 91
6.1 DEVELOPMENT OF HCFPS ................................................................................................... 91 6.1.1 Section configuration for static performance studies ...................................................... 92
6.2 VARIABLES IN PARAMETRIC STUDIES ............................................................................ 93 6.2.1 Section configurations .................................................................................................... 93 6.2.2 FE model for HCFPS ...................................................................................................... 93 6.2.3 Material Properties ......................................................................................................... 94 6.2.4 Loading conditions ......................................................................................................... 95
6.3 RECTANGULAR BEAM AND TAPERED BEAM ................................................................ 96 6.3.1 FE modelling .................................................................................................................. 96 6.3.2 FE analysis results and discussion .................................................................................. 97
6.4 FLEXURAL PERFORMANCE ................................................................................................ 99 6.4.1 FE modelling .................................................................................................................. 99 6.4.2 Properties of GRC .......................................................................................................... 99 6.4.3 Properties of Steel ......................................................................................................... 100 6.4.4 Properties of PU ........................................................................................................... 101 6.4.5 FE abalysis, results and discussion ............................................................................... 101
6.5 COMPARISON OF HCFPS WITH STEEL-DECK COMPOSITE FLOOR SYSTEM USING
STIFFNESS AND SELF-WEIGHT.................................................................................................... 108
6.6 DETERMINATION OF HCFPS SECTION PROPERTIES USING ANALYTICAL
METHODS ......................................................................................................................................... 109 6.6.1 Linear elastic deflection of HCFPS .............................................................................. 111 6.6.2 Stresses in individual component materials .................................................................. 112 6.6.3 Properties of cracked section ........................................................................................ 113
6.7 SHEAR PERFORMANCE ...................................................................................................... 116
6.8 PERFORMANCE OF GRC-PU-GRC PANEL AND SLAB JOINT ...................................... 117 6.8.1 FE modelling of GRC-PU-GRC panel ......................................................................... 117 6.8.2 FE model for slab joint of the adjacent HCFPS panels ................................................ 119 6.8.3 Properties of materials .................................................................................................. 119 6.8.4 FE analysis results and discussion ................................................................................ 119
6.9 SUMMARY ............................................................................................................................. 123
CHAPTER 7: DYNAMIC PERFORMANCE OF HCFPS ........................................................... 125
7.1 STRUCTURAL CONFIGURATION...................................................................................... 125
7.2 DYNAMIC PERFORMANCE OF HCFPS (SINGLE PANEL APPROACH) ....................... 126 7.2.1 FE modelling ................................................................................................................ 127 7.2.2 Material properties ........................................................................................................ 127 7.2.3 Mass of the HCFPS ...................................................................................................... 128 7.2.4 Free vibration analysis .................................................................................................. 129 7.2.5 Parameters that influence the first mode natural frequency of the HCFPS .................. 131
Static and dynamic performance of lightweight hybrid composite floor plate system vii
7.2.6 Damping ....................................................................................................................... 132 7.2.7 Mathematical load model for human induced loads ..................................................... 132 7.2.8 FE transient dynamic analysis ...................................................................................... 133 7.2.9 Results from parametric study and discussion .............................................................. 133 7.2.10 Vibration assessment of HCFPS ................................................................................... 137
7.3 DYNAMIC PERFORMANCE OF HCFPS FLOORS WITH THE STRUCTURAL FRAME141 7.3.1 FE modelling ................................................................................................................ 142 7.3.2 Material properties ........................................................................................................ 142 7.3.3 Mass of the HCFPS floor model ................................................................................... 143 7.3.4 Free vibration analysis of HCFPS floor model ............................................................. 143 7.3.5 Damping ....................................................................................................................... 145 7.3.6 Mathematical load model for human induced loads ..................................................... 145 7.3.7 FE transient dynamic analysis ...................................................................................... 146 7.3.8 Results from parametric study and discussion .............................................................. 146
7.4 CONCLUSIONS...................................................................................................................... 148
CHAPTER 8: CONCLUSIONS AND RECOMMENDATIONS ................................................. 151
8.1 CONTRIBUTION FROM THIS RESEARCH ........................................................................ 151
8.2 DISCUSSION AND SUMMARY ........................................................................................... 152 8.2.1 Experimental Testing and FE model validation ............................................................ 152 8.2.2 Static design of HCFPS ................................................................................................ 154 8.2.3 Dynamic performance of HCFPS ................................................................................. 156 8.2.4 Design guidelines.......................................................................................................... 157 8.2.5 Supporting and connection methods ............................................................................. 160 8.2.6 Limitation of design guidelines .................................................................................... 161 8.2.7 Manufacturing and casting guide .................................................................................. 161 8.2.8 Implications .................................................................................................................. 162 8.2.9 Future work................................................................................................................... 163
BIBLIOGRAPHY ............................................................................................................................. 165
viii Static and dynamic performance of lightweight hybrid composite floor plate system
List of Figures
Figure 1-1: Development of HCFPS using component materials ........................................................... 3
Figure 1-2: Formulation of HCFPS ......................................................................................................... 4
Figure 2-1: Sandwich construction configurations [17] ........................................................................ 10
Figure 2-2: De-lamination failure of a sandwich panel [6] ................................................................... 16
Figure 2-3: Wrinkling failure of sandwich panels [9] ........................................................................... 16
Figure 2-1: Hybrid material configuration in HCFPS ........................................................................... 17
Figure 2-2: How to optimize the performance HCFPS using component material characteristics ....... 18
Figure 2-3: Stress strain behaviour of high tensile steel of three similar specimens [22] ..................... 19
Figure 3-1: Proposed HCFPS panel configuration and symmetry ........................................................ 28
Figure 3-2: Graphical representation of building floor using HCFPS panels ....................................... 28
Figure 3-3: 3200 mm span HCFPS test panel configuration ................................................................. 29
Figure 3-4: A typical acceleration response [35] ................................................................................. 30
Figure 3-5: Static load test arrangement ............................................................................................... 31
Figure 3-6 Load cycles and steps of loading for cyclic loading test [39] .............................................. 32
Figure 3-7: A typical load-deflection curve for six load cycles [39] ..................................................... 33
Figure 3-8: Parameters to calculate the Repeatability [39] .................................................................. 34
Figure 3-9: Dynamic vertical force due to walking step by a person [57]. ......................................... 38
Figure 3-10: Frequency weighted RMS acceleration base curve [55] .................................................. 42
Figure 3-11 Development stages of HCFPS ......................................................................................... 46
Figure 4-1: (a) Tensile testing for GRC specimen, (b) Stress-strain relationships for GRC in
tension .................................................................................................................................. 48
Figure 4-2: (a) Four-point bending test for GRC, (b) Load-deflection plots for GRC in four-
point bending tests ............................................................................................................... 49
Figure 4-3: (a) Compression test for GRC, (b) Stress-strain relationships for GRC in
compression ......................................................................................................................... 50
Figure 4-4: (a) Stress-strain relationships f or PU in compression, (b) Compression testing for
PU core ................................................................................................................................. 51
Figure 4-5: (a) Tensile test for PU core, (b) Stress-strain behaviours of PU in tension ........................ 52
Figure 4-6: (a) Three-point bending test for PU core, (b) Load-deflection plots from bending
test of PU core ...................................................................................................................... 52
Figure 4-7: (a) Dimensions of tensile test specimen, (b) Tensile testing for steel laminate, (c)
Stress-strain relationships for steel laminate in tension ....................................................... 53
Figure 4-8: Section dimensions of HCFPS test panel .......................................................................... 54
Figure 4-9: Casting steps of HCFPS test panel ..................................................................................... 55
Figure 4-10: Test panel supporting arrangement .................................................................................. 56
Figure 4-11: Data acquisition system .................................................................................................... 56
Figure 4-12: Test setup and data acquisition system for heel impact test ............................................. 57
Figure 4-13: (a) Heel impact at the mid-span, (b) 5g Accelerometer and 25 mm LVDT at the
bottom HCFPS panel............................................................................................................ 58
Static and dynamic performance of lightweight hybrid composite floor plate system ix
Figure 4-14: Typical heel impact displacement response at mid-span for panel 1 ................................ 58
Figure 4-15: Typical heel impact acceleration response at mid-span for panel 1 ................................. 59
Figure 4-16: Typical heel impact acceleration response at mid-span for panel 2 ................................. 59
Figure 4-17: Typical heel impact acceleration response at mid-span for panel 3 ................................. 59
Figure 4-18: Typical FFT analysis of an acceleration response ........................................................... 61
Figure 4-19: Typical acceleration response at the mid-span for walking test ....................................... 62
Figure 4-20: Loading arrangement for the static load test of HCFPS panel .......................................... 63
Figure 4-21: Load-deflection behaviour of HCFPS panels ................................................................... 64
Figure 4-22: Cracking and failure due to the flexure ............................................................................ 65
Figure 4-23: Cyclic loading test for HCFPS panel ................................................................................ 66
Figure 4-24: Cyclic behaviour of HCFPS panel for first 6 loading cycles ............................................ 67
Figure 4-25: Cyclic behaviour of HCFPS panel .................................................................................... 68
Figure 4-26: GRC-PU-GRC composite panel slab in HCFPS assembly.............................................. 68
Figure 4-27: Determined sectional configuration of composite panel .................................................. 69
Figure 4-28: Loading test setup ............................................................................................................. 70
Figure 4-29: Supporting conditions for loading test .............................................................................. 70
Figure 4-30: Load-deflection curves for A type panels ......................................................................... 71
Figure 4-31: Load-deflection curves for B type panels ......................................................................... 71
Figure 4-32: Failure of one test panel ................................................................................................... 71
Figure 5-1: FE model for dynamic analysis .......................................................................................... 76
Figure 5-2: Mode shape for the first natural frequency ......................................................................... 77
Figure 5-3: Heel impact load function .................................................................................................. 78
Figure 5-4: Computed and measured acceleration responses due to heal impact.................................. 80
Figure 5-5: Computed and measured acceleration responses due to walking ....................................... 81
Figure 5-6: GRC material model ........................................................................................................... 83
Figure 5-7: FE model of HCFPS panel for static loading test ............................................................... 84
Figure 5-8: FE mesh of HCFPS panel ................................................................................................... 84
Figure 5-9: FE model validation with experimental results .................................................................. 85
Figure 5-10: Flexural cracks in the beam of HCFPS at the failure ....................................................... 86
Figure 5-11: Stress and strain distribution at the mid-span of the HCFPS ............................................ 86
Figure 5-12: FE model for GRC-PU-GRC panel .................................................................................. 88
Figure 5-13: Validation for FE model for static test (Type A panel) .................................................... 89
Figure 5-14: Validation for FE model for static test (Type B panel) .................................................... 89
Figure 6-1: Graphical representation of building floor using HCFPS panels ........................................ 91
Figure 6-2: Proposed supporting methods for HCFPS floor to structural frame ................................... 91
Figure 6-3: Proposed HCFPS panel configuration and symmetry ........................................................ 92
Figure 6-4: HCFPS section for parametric study .................................................................................. 92
Figure 6-5: GRC fill replacing PU core in the vicinity of supports ....................................................... 92
Figure 6-6: Section parameters for parametric study ............................................................................ 93
Figure 6-7: FE model of HCFPS ........................................................................................................... 94
x Static and dynamic performance of lightweight hybrid composite floor plate system
Figure 6-8: Proposed HCFPS panel configuration and symmetry ........................................................ 97
Figure 6-10: Variation in stress in the GRC layer (with GRC 10) with change in "a" .......................... 98
Figure 6-11: Continuous glass fiber mesh ............................................................................................. 99
Figure 6-12: GRC material model ....................................................................................................... 100
Figure 6-13: Stress-strain relationship for High strength and mild steel ............................................. 100
Figure 6-14: Load -deflection behaviour of A, B and C type sections ................................................ 102
Figure 6-15: A typical flexural crack development in the beam of the HCFPS .................................. 105
Figure 6-16: Stress and strain distribution at the mid-span of the HCFPS along X-X ........................ 106
Figure 6-17: A typical flexural stress distribution (in GRC and PU) at the mid-span (Type A) ......... 107
Figure 6-18: Load-deflection plots of steel deck floor system and HCFPS ........................................ 109
Figure 6-19: Parameters used to define the properties of HCFPS section .......................................... 110
Figure 6-20: Load-deflection comparison between FE and analytical methods ................................. 111
Figure 6-21: Stresses in component materials at mid span section of HCFPS along X-X .................. 112
Figure 6-22: Tensile crack development in the beam of the HCFPS (Type B) ................................... 114
Figure 6-23: Parameters used to define the properties of cracked HCFPS section ............................. 114
Figure 6-24: Stress distribution of cracked HCFPS section along X-X .............................................. 115
Figure 6-25: Shear stress in GRC outer shell at the vicinity of support .............................................. 116
Figure 6-26: Shear zone of in the HCFPS cross-section ..................................................................... 116
Figure 6-27: GRC-PU-GRC slab of HCFPS ....................................................................................... 118
Figure 6-28: Layer thickness of GRC-PU-GRC panel ........................................................................ 118
Figure 6-29: FE models of GRC-PU-GRC panel ................................................................................ 118
Figure 6-30: Typical detail of the adjacent slab connection ................................................................ 119
Figure 6-31: FE models of contact joint.............................................................................................. 119
Figure 6-32: Performance GRC-PU-GRC panel with GRC 5 and PU 20 ........................................... 121
Figure 6-33: Performance of slab joint with GRC 5 and PU 20.......................................................... 122
Figure 7-1: HCFPS floor plate with steel frame.................................................................................. 125
Figure 7-2: Prefabricated HCFPS panel .............................................................................................. 126
Figure 7-3: Sectional configuration parameters .................................................................................. 126
Figure 7-4: FE model of HCFPS panel .............................................................................................. 127
Figure 7-5: Typical mode shape for the first mode natural frequency ............................................... 130
Figure 7-6: RMS acceleration for section type A, and 3 m span ........................................................ 135
Figure 7-7: RMS acceleration for section type B, and 5 m span ......................................................... 136
Figure 7-8: RMS acceleration for section type C, and 7.5 m span ...................................................... 137
Figure 7-9: Structural configuration of four bay floor model using HCFPS ....................................... 141
Figure 7-10: FE model of four bay HCFPS floor ................................................................................ 142
Figure 7-11: First four modes shape of HCFPS floor model using GRC 10 and PU 75 ..................... 144
Figure 7-12: RMS acceleration of HCFPS floor model with section type C (non-activity panel) ...... 146
Figure 7-13: RMS acceleration of HCFPS floor model with section type C (activity panel) ............. 147
Figure 8-1: Optimum configuration of HCFPS ................................................................................... 154
Figure 8-2: Parameters used to define the properties of HCFPS section ............................................ 158
Static and dynamic performance of lightweight hybrid composite floor plate system xi
Figure 8-3: Parameters used to define the properties of cracked HCFPS section ............................... 159
Figure 8-4: Proposed supporting methods for HCFPS floor to structural frame ................................. 161
Figure 8-5: Cross-section of the HCFPS panel .................................................................................. 161
xii Static and dynamic performance of lightweight hybrid composite floor plate system
List of Tables
Table 2-1: Typical values of steel properties [22] ................................................................................. 19
Table 2-2: Typical values of PU properties [14, 25] ............................................................................. 21
Table 2-3: Typical values of GRC properties [29, 32, 33] .................................................................... 23
Table 3-1: standard test methods for material property investigation ................................................... 35
Table 3-2: Loading cases [48] ............................................................................................................... 37
Table 3-3: Super imposed permanent dead loads for an office floor .................................................... 37
Table 3-4: Design parameters for walking and running loads [49, 55] ................................................. 39
Table 3-5: Recommended response factors [56] ................................................................................... 42
Table 3-6: VDV values (m/s1.75
) for vertical direction vibrations [60] ............................................... 43
Table 4-1: Formulation of GRC ............................................................................................................ 47
Table 4-2: Damping ratios for the HCFPS test panel ............................................................................ 60
Table 4-3: Experimental First natural frequency of HCFPS panels ...................................................... 61
Table 4-4: Summary of material properties obtained from the material testing .................................... 72
Table 5-1: Component material properties for the dynamic analysis .................................................... 75
Table 5-2: Validation of first natural frequency .................................................................................... 77
Table 5-3: Mass proportional stiffness proportional damping for FE model ....................................... 79
Table 5-4: Component material properties for the GRC-PU -GRC , FE model. ................................... 88
Table 6-1: HCFPS section and span parameters ................................................................................... 93
Table 6-2: Properties of Steel ................................................................................................................ 94
Table 6-3: Properties of GRC [12] ........................................................................................................ 95
Table 6-4: Properties of PU [25] ........................................................................................................... 95
Table 6-5: Self-weight of the HCFPS floors ......................................................................................... 96
Table 6-6: Loading cases [48] ............................................................................................................... 96
Table 6-7: HCFPS section and span parameters ................................................................................... 97
Table 6-8: Properties of Steel ................................................................................................................ 97
Figure 6-9: Shear stress in GRC outer shell for type B section ............................................................. 98
Table 6-9: Properties of of GRC ......................................................................................................... 100
Table 6-10: Serviceability deflection of the HCFPS floor with PU 20 ............................................... 102
Table 6-11: Factor of Safety (FOS) for flexural performance of HCFPS with PU 20 and MSteel ........ 103
Table 6-12: Factored deflection to account for creep and shrinkage deformation .............................. 108
Table 6-13: Comparison of stresses in component materials under 5 kPa load .................................. 113
Table 6-14: Properties of of GRC ....................................................................................................... 115
Table 6-15: Shear Capacity of HCFPS sections .................................................................................. 117
Table 6-16: Sectional configuration of GRC-PU-GRC panels ........................................................... 118
Table 6-17: Properties of GRC and PU ............................................................................................... 119
Table 6-18: Performance of 1000 mm one-way span ......................................................................... 121
Table 6-19: Performance of 500 mm cantilever span ......................................................................... 121
Static and dynamic performance of lightweight hybrid composite floor plate system xiii
Table 6-20: Performance of 1000 mm one-way span with slab joint at the centre ............................. 122
Table 7-1: Spans and section dimensions ............................................................................................ 126
Table 7-2: Properties of GRC [12] ...................................................................................................... 128
Table 7-3: Properties of PU [25] ......................................................................................................... 128
Table 7-4: Properties of Steel .............................................................................................................. 128
Table 7-5: Super imposed permanent dead loads for an office floor .................................................. 129
Table 7-6: Mass of the HCFPS floors ................................................................................................. 129
Table 7-7: Modal frequencies of HCFPS panels ................................................................................. 130
Table 7-8: Comparison of first mode natural frequency ..................................................................... 131
Table 7-9: Parameters for the load model[55] ..................................................................................... 133
Table 7-10: VDV assessment of HCFPS using na ............................................................................... 140
Table 7-11: Properties of GRC [12] .................................................................................................... 143
Table 7-12: Properties of PU [25] ....................................................................................................... 143
Table 7-13: Properties of Steel ............................................................................................................ 143
Table 7-14: Modal frequencies of HCFPS model ............................................................................... 145
Table 7-15: Mass proportional damping (α) and stiffness proportional damping (β) for ζ= 5% ........ 145
Table 7-16: Minimum number of activities required generates lower probability adverse
comment of Floor model ................................................................................................... 147
xiv Static and dynamic performance of lightweight hybrid composite floor plate system
List of Abbreviations
HCFPS Hybrid Composite Floor Plate System
GRC Glass-fiber Reinforced Cement
PU Polyurethane
FE Finite Element
FRC Fiber Reinforced Composites
GFRP Glass Fiber Reinforced Polymer
SPS Sandwich Plate System
HCB Hybrid Composite Beam
ASTM American Society for Testing and Materials
AS Australian Standards
ISO International Organization for Standardization
AISC American Institute of Steel Construction
EI Flexural rigidity
LVDT Linear-Variable-Displacement-Transducers
DAS Data Acquisition System
RMS Root-Mean-Square
FFT Fast Fourier Transformation
MSteel Mild Steel
TSteel High Strength Steel
GRC 5, 10, 15, 20 Types of GRC (5, 10, 15, 20 represents modulus of elasticity in GPa)
PU 20, 75, 150, 360 Types of PU (20, 75, 150, 360 represents modulus of elasticity in MPa)
List of Symbols
E, Es Elastic modulus
G, Gs Shear modulus
ρ, ρs Density
τ, τs Shear strength
Lus Ultimate load of HCFPS (load at the yielding point)
Gk Dead Load (kPa)
Qk Imposed Load (kPa)
Static and dynamic performance of lightweight hybrid composite floor plate system xv
F(t) Dynamic load due to human activities
Q Static weight of walking people (kN, kPa)
αn Fourier coefficient
n nth harmonic
f, fp Step frequency or load frequency or activity frequency (Hz)
υn Phase angle
I Second moment of area
m Effective mass
L Span of HCFPS
υn Phase angle
arms Root-Mean-Square acceleration
aw,rms Frequency weighted Root-Mean-Square acceleration
VDV Vibration Dose Value
Ta Exposure period (16 hour day or 8 hour night)
na Number of time a activity will take place in exposure period
apeak Peak acceleration
Ao, A1, … An Acceleration (g)
ζ Structural damping
σcy Compressive yield stress
εcy Compressive yield strain
εcu Ultimate compressive strain
σcr Structural damping
εcr First cracking tensile strain
σtu Tensile stress at the end of tensile model
εtu Ultimate tensile strain
ν Poisons ratio
bb Width of the beam
db Depth of the beam
h Depth HCFPS section
L Width of the slab
yi Distance to centroid (from the bottom) of HCFPS
bi Width of individual components
di Depth of individual components
A Net area of component material
Ai Net area of ith
component material
xvi Static and dynamic performance of lightweight hybrid composite floor plate system
A Area of the transformed section
tPU Thickness of PU in the slab
tGRC, top Thickness of GRC layer at the top
tGRC Thickness of GRC layer
tGRC ESS Thickness GRC layer Either Sides of Steel laminate
tSteel Thickness of Steel laminate
EPU Elastic modulus of PU
EGRC Elastic modulus of GRC
ESteel Elastic modulus of Steel
n1 ESteel / EPU
n2 EGRC / EPU
VCapacity Shear capacity
f1, f2 Modal frequency (Hz)
.
Static and dynamic performance of lightweight hybrid composite floor plate system xvii
Statement of Original Authorship
The work contained in this thesis has not been previously submitted to meet
requirements for an award at this or any other higher education institution. To the
best of my knowledge and belief, the thesis contains no material previously
published or written by another person except where due reference is made.
Signature:
Date: _________________________
QUT Verified Signature
xviii Static and dynamic performance of lightweight hybrid composite floor plate system
Acknowledgements
I was able to complete this research project due to the great support of many sources.
Any contribution to the engineering profession as result of this research project have
been a joint effort of myself, my supervisory team and others who support this
project in various ways. My greatest thanks and appreciation goes to following
people.
First, I would like to express my sincere gratitude to my principal supervisor, Prof.
David Thambiratnam for giving me this opportunity together with his motivation,
great support and excellent guidance to carry out my research work successfully. I
thank him for steering me towards the goals of this research project and for helping
me to overcome the difficulties encountered during the candidature. I would never
have completed without his support and encouragement. I would also like extend
sincere thank to my associate supervisor Adjunct Prof. Nimal Perera for his
professional guidance, rigorous discussions, valuable advices and useful suggestions
given during entire period of time. I wish to thank Dr. Ashar Nasir, for his advices
and useful suggestions for this research. I would also like to thank Prof. Andy Tan
for being panel member for both my confirmation of candidature and final seminar.
Second, I gratefully acknowledge the financial support given by BEE for providing
the Post Graduate Research scholarship to carry out the research work. I would like
to express my special thanks to DomeShells Australia Pte Ltd and Mr. Chris Brown,
who contributed graciously all necessary test materials to carry out experimental
testing. That was an invaluable support for this research to reinforce the findings.
I would also wish to acknowledge the members of the technical staff at QUT for their
generous support given me during my experimental testings. In particular, I wish to
thank to Mr. Greg Tevelen, Mr. Gregory Paterson, Mr. Noel Hartnett, Mr Matthew
Mackay, Mr. Jonathan James and Mr. Anthony Morris for their patience and endless
support in carrying out the experimental testing. In addition, I am also thankful to all
academic and non-academic staff of faculty for the support given in many ways.
Static and dynamic performance of lightweight hybrid composite floor plate system xix
I would also like to gratefully acknowledge Dr Bernadette Bellette for her time and
support given me improving my writing skills. My acknowledgement is further
extended to my fellow researchers and friends for their valuable support in my
research.
Finally, I would like to thank my family and my loving wife Rajima Dilmani for
their love and support during the difficult times, and greatly appreciate all the help
provided. The completion of this thesis would have not been possible without their
encouragement and patience.
xx Static and dynamic performance of lightweight hybrid composite floor plate system
Chapter 1: Introduction 1
Chapter 1: Introduction
1.1 INTRODUCTION
In the modern built environment, building construction and demolition consume a
large amount of energy and emits greenhouse gasses due to the widely used
conventional construction materials such as reinforced concrete, composite concrete
and steel. These materials possess higher embodied energy and more energy is
required to recycle or reuse them at the cessation of use [1]. Moreover, commercially
useful life of buildings in the modern era is becoming shorter due to the higher
demand of available building sites with ever changing land use and lack of suitability
of buildings for developing demands. On average, commercially useful life of
building is less than 50 years although actual useful design life may extend to 100
years or more [2]. However, many buildings are structurally modified or retrofitted
for change of within 20 years. Therefore, it is very important to use recyclable or
reusable new materials in building construction in order to reduce the energy and
emissions associated with conventional materials. In addition, growing limitation in
natural resources, which are used to manufacture conventional building materials,
can also be reduced by developing recyclable or reusable building materials. In this
context, this research project develops an innovative hybrid floor system with
desirable properties, such as lightweight, easy to construct, economical,
demountable, recyclable and reusable.
Advancements in material technology have created a trend for using hybrid
composite materials in infrastructure construction. Hybrid composite and sandwich
materials offer desirable properties, such as being lightweight, easy to construct,
economical, demountable, recyclable and reusable. They are efficient systems,
offering high stiffness-to-weight ratios and high strength-to-weight ratios. Hybrid
composite and sandwich materials have been used effectively instead of conventional
materials in many engineering applications. Typically, sandwich panels comprise of
mainly two materials, which are high strength thin skins and a middle soft core.
Hybrid composite systems comprise of several hybridised materials to form
composites. Hybrid composite components have been used effectively in automotive
2 Chapter 1: Introduction
and aerospace applications [3, 4], while sandwich panels have been used in
aerospace, marine and civil constructions, such as cladding wall panels [5], floor
panels[6-8] and roof panels [9], bridge decks [8], transportation applications [10]
and dome-type shell structures [11]. However, sandwich panels can only be used as
short span floor structures as they are slender and deflect excessively when used as
long span panels. Moreover, sandwich panels exhibit de-lamination and wrinkling
failure under flexural loading. As a consequence, use of sandwich panels in floor
plate construction is limited. Despite the potential to use hybrid composites in floor
construction to overcome the limitation of sandwich panels, hybrid composite
components in floor plate construction has not been adequately investigated and
developed.
Admittedly, Glass-fiber Reinforced Cement (GRC), Polyurethane (PU) and thin steel
laminate have the potential to be used as component materials in hybrid structural
floor plates because they have been successfully used in sandwich construction
applications. GRC is a fiber-reinforced composite material comprising of alkali-
resistant glass fiber, cement and sand as the major constituents [12]. It is lighter than
conventional concrete and can be used to produce thin outer layers of sandwich
panels [11, 12]. PU is a common lightweight foam core material, which has been
extensively utilized in sandwich construction. A wide range of properties can be
achieved with different densities of PU [13]. There is thus a potential for the
development of an innovative hybrid composite floor System using GRC, PU and
steel laminates.
Despite this, applicability of hybridized GRC, PU and steel in structural floor
systems has not yet been studied. For the first time, this research has developed a
new light weight structural floor plate system, called Hybrid Composite Floor Plate
System (HCFPS) as illustrated in Figure 1-1 to investigate the potential application
of GRC, PU and steel hybrids. Configuration of the HCFPS is designed to combine
the inherent strength characteristics of the constituent materials and to eliminate their
weaker characteristics. This research investigates the strength, serviceability and
performance characteristics of HCFPS that can be used in high performance light
weight floor plates for buildings.
Chapter 1: Introduction 3
Figure 1-1: Development of HCFPS using component materials
1.2 HCFPS
HCFPS exploits the inherent beneficial properties of component materials in such a
manner to optimise the overall performance of the floor plate system. Proposed
HCFPS is assembled using component materials as shown in the Figure 1-2. Width
of HCFPS is limited to 2 m to suit prefabrication and transportation requirements but
adjacent panels are connected using the slab joint shown Figure 1-2. Length of
HCFPS can be varied by changing the material properties and sectional
configuration. A cold-formed thin perforated steel laminate is placed at the bottom of
the beam to improve tensile strength. Perforated steel laminate is used to enhance the
bonding between GRC and steel laminate. The PU core is replaced with a GRC fill in
the vicinity of the supports as shown in Figure 1-2, to enhance the support bearing
capacity of the panel.
When HCFPS panel is subjected to bending, compressive stresses occur mainly in
the slab and tensile stresses occur in the bottom steel laminate. Shear stresses occur
across the web of the beam due to the loading. GRC and PU exhibit better
performance under compressive and shear stresses [12, 14]. Hence, GRC and PU are
profiled and located to attract compressive and shear stresses in the slab and beams
of HCFPS (as shown in the Figure 1-2). Stability of HCFPS section is provided by
the PU core, because it provides connection between and lateral support for thin
GRC layers to prevent the lateral buckling. The steel laminate and GRC attract
4 Chapter 1: Introduction
higher stresses sue to their elastic modulus, which is significantly higher than that of
PU. In this way, beneficial inherent properties of individual component materials are
combined to achieve optimum performance in the HCFPS.
Figure 1-2: Formulation of HCFPS
HCFPS has high potential for use as a structurally efficient and high perfromance
flooring system. In order to verify these attributes, the static performance of HCFPS,
such as flexural capacity, shear capacity, deflection limits and ductility need to be
investigated when subjected to dead and imposed loading. In addition, HCFPS is
lightweight structure and therefore, needs to be evaluated and designed to comply
with the dynamic performance for comfort and serviceability in relation to human
perception in compliance with acceptable standards.
This research investigates the performance characteristics of steel, PU and GRC in
HCFPS using Finite Element (FE) method simulations supported by experimental
testing. ABAQUS code is used to develop FE models in order to simulate HCFPS
floor configurations and spans. The FE modelling techniques are validated using
experimental testing. Parametric studies are carried out to develop new HCFPS with
design guidelines.
Chapter 1: Introduction 5
1.3 RESEARCH AIMS AND OBJECTIVES
The main aim of this research is to develop a HCFPS comprising GRC, PU and steel
laminates and determine the performance characteristics.
Additional objectives are to:
o Determine properties of constituent materials
o Experimentally test HCFPS panels to investigate static, dynamic,
cyclic loading and post yield behaviours, mode of failure and ductility.
o Develop FE models and validate by experimental testing
o Perform parametric studies on HCFPS panels in relation to material
properties of GRC, PU, steel and their hybrids to quantify static
performance requirement of HCFPS. FE techniques will be used to
study the influence of parameters such as sectional configuration,
spans, loading limits and support condition
o Determine the vibration characteristic of HCFPS under human-
induced loads
o Investigate and determine the required enhancement of material
properties of hybrid materials for an optimum design that will satisfy
static and dynamic design criteria
o Development of design guidelines
1.4 SIGNIFICANCE AND SCOPE
1.4.1 Significance
This project is significant as the research findings will lead to the development of an
innovative hybrid structural floor plate system with several desirable properties such
as being lightweight, easy to construct, economical, demountable, recyclable and
reusable. Component materials used have been well developed and tested in
standalone applications. However, hybridized use of these materials for the proposed
application is unprecedented. For the first time, GRC, PU and steel hybrids are used
to develop a lightweight and efficient floor plate system with good vibration
performance.
HCFPS can be developed as a prefabricated floor system that can be manufactured in
an offsite factory under controlled conditions to achieve a product of superior quality
6 Chapter 1: Introduction
with low embodied energy. This floor system is approximately 50% lighter in weight
compared to conventional concrete floors. Therefore, this product is easy to
transport, handle and erect. HCFPS has the potential to revolutionize the construction
of structural floor systems by replacing slow, labour intensive and low quality
construction materials with factory based manufacturing process. Manufactured floor
plates can be assembled with simple connections on site eliminating the heavy,
cumbersome and time consuming material handling, transporting and erecting
processes while minimising safety hazards. HCFPS can be demounted and reused in
other applications or can be recycled as a whole component at the end of
commercially useful life of the buildings.
Additionally, lightweight property of HCFPS floor plates results in reduced load on
the supporting beams and columns. Thereby, sizes of such load-bearing members can
be reduced, yielding economical advantages.
Moreover, lightweight HCFPS floor and reduced size of load bearing members
results in lower mass for building structures. As a consequence, such buildings offer
better performance during earthquakes. In an earthquake, ground shaking generates
internal forces within the buildings called inertial forces, causing damage to building
structures. Magnitude of inertial forces are proptional to the mass of a structure [15].
Hence, the lower the mass of the building lower the strength demand for seismic
loads.
Findings from this research, therefore, provide an efficient, lightweight, economical,
and sustainable structural flooring system that can be recycled and reused as a whole
system compared to conventional floor systems. HCFPS offers multifunctional
structural properties, making it a viable alternative for conventional reinforced
concrete and composite deck floors. It is therefore a product that addresses social and
environmental needs of the global community using material and manufacturing process
with low energy content.
1.4.2 Scope
In this research project, the suitability of HCFPS for structural flooring is
determined. FE simulation was used with limited experimental testing. Material
properties are obtained from limited mechanical testing and available resources and
Chapter 1: Introduction 7
published manufacturers’ data. Static and dynamic performance characteristics of
HCFPS are investigated. Non-linear behaviour of the HCFPS is investigated under
static performance requirements. Post yield behaviour, mode of failure and ductility
measurements of HCFPS is investigated for the tested specimens. Linear elastic
behaviour of HCFPS is considered for investigating dynamic performance. Fire
resistance and thermal insulation have not been investigated in this research but data
can be obtained from existing information based on studies carried out on such
materials.
1.5 THESIS OUTLINE
This thesis consists of eight chapters and their content is as follows:
Chapter 1: An introduction and background to the research topic, configuration of
HCFPS, aims and objectives, scope and significance of the research have been
described in this chapter.
Chapter 2: This chapter reviews current application of lightweight composites
(sandwich and hybrid) in civil engineering construction. The limitation and
weaknesses of those composites for use in structural floor plate construction are
reviewed. Reasons and importance of developing a HCFPS are also discussed.
Chapter 3: This chapter describes research methodology and the various stages
through which the HCFPS was developed. It also makes references to appropriate
codes or standards and literature used in development.
Chapter 4: This chapter presents the experimental investigations conducted on 3200
mm span HCFPS panels and GRC-PU-GRC composite panels. Material property
investigation, test panels fabrication method, test setup, instrumentation and
experimental results are explained.
Chapter 5: This chapter presents the development of FE models using ABAQUS
and subsequent validation of FE models using experimental test results generated in
Chapter 4.
Chapter 6: This chapter presents the development of HCFPS to satisfy static
performance requirements. Details of parametric studies, which were conducted to
8 Chapter 1: Introduction
investigate the static performance characteristics and strength capacity of HCFPS,
are explained.
Chapter 7: This chapter investigates dynamic response of HCFPS under human
induced vibration. Findings shows that lightweight HCFPS can be used in residential
and office buildings by evaluating its vibration performance using acceptable
perceptibility limits provided in current design guidelines and standards.
Chapter 8: This final chapter highlights the main contribution of this research, major
findings and recommendations for future research.
Chapter 2: Literature review 9
Chapter 2: Literature review
This chapter will review current application of lightweight composites (sandwich and
hybrid) in civil engineering construction. Research studies that have been carried out
for the development of such composites and their properties will be discussed. The
limitation and weaknesses of those composites for use in structural floor plate
construction will also be reviewed. Furthermore, the reasons and importance of
developing a Hybrid Composite Floor Plate System (HCFPS) will be discussed.
Material that can be used to develop HCFPS and their properties and advantages will
finally be reviewed.
2.1 CURRENT COMPOSITE CONSTRUCTION APPLICATIONS AND
THEIR LIMITATIONS
Advancements in material technology have created a trend for using composite
materials in infrastructure construction. Composite materials are comprised of two or
more materials of different properties, in order to obtain more efficient final
products. The main types of composites are fiber reinforced composites, sandwich
composites and hybrid composites. Composite materials offer desirable properties
such as being lightweight, easy to construct, economical, demountable, recyclable
and reusable. They are efficient systems, offering high stiffness-to-weight ratios and
high strength-to-weight ratios [10]. Composite materials have been used effectively
instead of conventional materials in infrastructure construction. Composite
constriction applications and studies that are related to civil structures will be
reviewed in this section.
2.1.1 Fiber Reinforced Composites (FRC)
FRC are formed by imbedding fibers into a binder material. Continuous or
discontinuous fibers, which are glass, metal and ceramics, have been used in FRC.
Composite fibers act as main load attracting members, whilst binder material holds
the fibers in the intended orientation [16]. FRC have been used in civil construction
for architectural decorating applications, repair and rehabilitation of existing
structures and sandwich construction applications [5]. In this context, FRC
applications related only to the sandwich and hybrid composite construction have
10 Chapter 2: Literature review
been considered. Common FRC materials used in such applications are Glass-Fiber
Reinforced Polymer (GFRP) and Glass-fiber Reinforced Cement (GRC). Their
applications and properties in relation to sandwich and hybrid composite
constructions application will be discussed in section 2.1.2.
2.1.2 Sandwich Composites
Sandwich panels are comprised of two thin, high strength, outer skin laminates and a
central core. The skins attract compressive and tensile stresses that result from
bending, whilst the core not only keeps the two skins apart but also transmits shear
forces to the supports [9]. Skins of sandwich panels are made of aluminium, steel and
FRC such as GRFP and GRC. Cores of the sandwich panels are made as either solid,
honeycomb, truss, Z-shaped, C-shaped, or I-shaped cores using metallic or polymeric
materials such as Polyurethane (PU) as shown in Figure 2-1. The configurations of
the skins are either flat or lightly profiled, but wide variations in the core structure
allow sandwich panels to be used in different applications. Flexural stiffness of
sandwich panels can be increased by increasing the distance between the face plates.
As a consequence, normal flexural stresses in component material can be reduced
[17].
Figure 2-1: Sandwich construction configurations [17]
There are a number of applications of sandwich construction, as a consequence of
combining different facing and core materials. Major sandwich construction
applications in civil constructions are cladding wall panels [5, 9], floor panels [6-8],
roof panels [9], bridge decks [17], transportation applications [10] and dome-type
shell structures [11].
Chapter 2: Literature review 11
In the last three decades, lightweight sandwich panels, using a thin outer skin with a
low density core, have been developed and used for wall claddings and roof panels.
According to Davies [9] thin steel sheets are the most common material used as skin
for lightweight wall and roof sandwich panels. Aluminium, plywood, chipboard and
gypsum board are also recommended for this. Core materials are basically
categorised into two types, those being bonded and foamed cores. Bonded cores
typically consist of either polystyrene or mineral fibers, which are bonded to the face
plate using a suitable adhesive [9, 18]. Formed cores are typically obtained by
mixing two liquids and no adhesive is necessary as the mixture adheres strongly to
surfaces of skins during chemical process of hardening [9].
According to Davies [9], the load-bearing capacity of lightweight sandwich panels
depends directly on the face and core materials, and secondly, on the adhesion
between the faces and the core. In addition to these, load bearing-capacity is
influenced by the structural dimensions and structural system, including the lengths
of spans, the support widths and the fastening system. Strength with respect to failure
modes has to be determined in order to evaluate the load bearing capacity of
sandwich panels. Failure modes includes tensile failure of the faces, wrinkling failure
of the faces (due to compressive stress), shear failure of the core or the adhesion
between the core and face, crushing failure of the face and core at a support, tensile
or shear failure of fasteners [9].
Davies [9] further explains that the compressive skins of lightweight sandwich panels
can exceed the buckling stresses and result in wrinkling failure. Wrinkling failure is a
significant issue in lightweight panels under bending, because most sandwich panel
consist of thin outer skins. This may determine the ultimate strength of a sandwich
panel. Sandwich panel design, therefore, is governed mostly by the compressive
stress capacity of sandwich skins. Sandwich panels can only be used as short span
floor structures because they become slender and deflect excessively when used as
long span panels. Moreover, sandwich panels exhibit de-lamination and wrinkling
failures under flexural loading. As a consequence there is limited use of sandwich
panels in floor plate construction [9].
Flexural performance of sandwich panels, which are comprised of GFRP skins and
ribs along with a PU core, have been investigated as building claddings and floor
12 Chapter 2: Literature review
panels by Fam et at. [6]. GFRP ribs were used in between the GFRP skins to enhance
strength and stiffness. Static loading test of full scale 2500 mm span sandwich panel.
31.6 kg/m3 density and 75 mm thick PU foam core along with 1.6 mm GRFP skins
were used for the sandwich panels. One panel without ribs and others with various
internal ribs between two skins were prepared. Panels were tested for one-way
bending with a span of 2300 mm. A uniformly distributed load was applied to the
panel as four equally spaced line loads along the span. This loading system
facilitated curvature of the panel during the deformation, whist maintaining the
uniform loads at each contact point. Fam et al. [6] demonstrated that, flexural
strength and stiffness of the panels can be increased substantially by adding GRFP
ribs. Simplified expressions have been proposed to calculate the deflection, including
terms for shear and flexural deflection. Strength of the 2.5m span panel exhibited
equivalent capacity to those of a similar sized reinforced concrete panel. De-
lamination and buckling failure of compressive faces were observed. This was
because of the 1.6 mm GRFP layer in the compressive faces, which was subjected to
buckling under compressive stress. Even though, this sandwich panel utilised ribs to
enhance the stiffness and strength, there could be excessive deflection problems with
longer spans over 3m. Increasing sandwich thickness or material properties may not
result in an economical structure [6].
Large scale sandwich panels (9145 × 2440 × 78 mm) consisting of GFRP skins that
are connected by GFRP ribs and PU core, were investigated for the building cladding
applications by Sharaf et.al [5]. 8220 mm span sandwich panels with central support
were tested to investigate the flexural performance, by applying uniformly
distributed loads using air pressure and equally spaced line loads. Clear span of 4110
mm achieved a 7.5 kPa, but the respective deflection was about 80 mm. Sharaf et.al
[5] concluded that these panels can be effectively used as cladding panels of
buildings. Further, these panels exhibited wrinkling failure of compression faces and
core shear failure at supports at ultimate loading. Although these panels were capable
of achieving approximately 7.5 kPa ultimate loading, such panels cannot be used in
floor construction because of the excessive deflection [5].
Sandwich panels made of glass fiber reinforced polymer skins and modified phenolic
core were investigated for the building floors by Islam et al. [7]. Core density was
850 kg/m3 and the thickness of fiber composite skin was 1.8 mm of the sandwich
Chapter 2: Literature review 13
panel. One- and two- way behaviour of 450 mm × 950 mm and 15 mm thick panel
was investigated under point and uniformly distributed loads. Point load was applied
into an area of 100 mm × 100 mm at the centre of panels using a hydraulic jack.
Uniformly distributed load was applied to the panel by using high pressure air bags.
Air bags were placed in between large metal pate and sandwich panel and pressure
was increased to apply the distributed load. These panels can be used instead of
timber decking in floor constructions with timber joist. However, long span floor
structures cannot be constructed with this kind of sandwich panels because overall
thickness these panels are 15 mm and can be vulnerable to excessive deflections
under as long spans.
The Sandwich Plate System (SPS) is a composite system, comprising of metal plates
and a high density PU core. Current and potential SPS applications include ship
repair, ship components, maritime overlays, new bridge construction, bridge deck
repair/rehabilitation, grandstand floors, stadium risers, and building floor systems [8,
17]. Braun et al. [8] has conducted static and dynamic tests on a SPS double L
shaped prototype (12.2m span) riser section for a grandstand. They observed good
static and dynamic performance when compared with stiffened steel or pre-stressed
concrete alternatives. Applicability of SPS in bridge construction has been
investigated by Cousins et al. [17]. They observed that SPS could reach its full
plastic capacity in flexure or compression without local buckling of either faceplate.
However, SPS uses thicker steel plates (about 4 to 5 mm) and shorter spans (less than
4 m). SPS have been proposed to be used with steel beams to obtain longer spans
[17]. If SPS is used for long spans without steel beams, local buckling effects may
result. However, a proper study has not been carried out to evaluate the performance
characteristics of SPS in floor structures, especially for long spans. SPS may not be
economical due to the use of two steel face plates in structural flooring.
Sandwich panels comprised of GRFP and a high density balsa wood core have been
investigated under concentrated loads for civil infrastructure and heavy freight
transportation applications by Dawood et al. [10]. This study investigated different
methods to enhance punching shear capacity of the sandwich panels. This study
demonstrated that indentation resistance of sandwich panels depends on sandwich
configuration and core characteristics. Detail study has not been conducted to
14 Chapter 2: Literature review
investigate the flexural performance of such sandwich panels and applicability of
floor structures.
Performance characteristics of a sandwich panel comprised of two GRC faces with a
lightweight PU foam core have been investigated for their application in Dome
Shells Structures [11, 19, 20]. Studies demonstrated that this structure combines the
structural qualities of the compound curved shell and the sandwich panel. A typical
Dome Shells Structure includes a diameter of 6-10 m, with 2.1 m vertical walls
rounding to maximum height of 3.2-4.0 m at its centre. Experimental investigations
of full scale (8 m diameter) dome, which is comprised of 6 mm thick inner/outer
layer of GRC and a 55 mm thick PU core sandwich panel, has been carried out. High
structural qualities of the sandwich section and compound curved shell exhibited
very good performance under static and cyclonic wind loading. Furthermore,
excellent thermal and insulation properties were also observed. DomeShell structures
use low density PU core with outer GRC faces. For dome type structures, entire
sandwich sections are subjected to compression stresses [11].
Flexural performance of GRC-PU-GRC, which consist of 52 mm central PU core
and 6 mm GRC outer faces, have been investigated by Gaston et al. [20]. 1200 mm
span 400 mm wide simply supported sandwich panel was tested one way bending by
applying a central line load. The panel did not demonstrate wrinkling or de
lamination failure in the sandwich panel but overall failure resulted due to the tensile
failure of the bottom GRC layer. Therefore, these sandwich panels may not be used
in long span floor construction because GRC cannot undertake higher tensile stress.
However, short span floor panels (1000 mm span) may be achieved using GRC-PU-
GRC panel.
2.1.3 Hybrid composites
Hybrid composite systems are comprised of several materials hybridised to form
composites. Hybrid composite components have been investigated and used
limitedly in infrastructure constructions that are related to civil engineering.
The possibility of a Hybrid-Composite Beam (HCB) ( 9 m span with 750 mm
depth) as a structural member for USA railway bridges was investigated in 2003
[21]. This beam consisted of a fiber-reinforced polymer outer shell, lightweight core,
Chapter 2: Literature review 15
arch shaped concrete compression reinforcement and carbon, glass or steel tension
reinforcement. Compression arch reinforcement consists of Portland cement
concrete, which was pumped into profiled arch shaped conduit placed along the
beam. Arch was connected to the bottom tensile reinforcement at the end of the
beams and crest of the arch was at the top surface of the beam at mid-span. Tension
reinforcements was laid at the very bottom of the beam and anchored to the
compression reinforcement at the ends. Beam shell and conduit for the compression
reinforcement were fabricated monolithically using vacuum assisted resin transfer
method [21].
An excel worksheet has been used for the development of the HCB, thereby ultimate
moment and shear capacity were determined. Plastic neutral axis depth was
determined, assuming that plane section remains plane and perpendicular to the
plastic neutral axis. It was also assumed that compression reinforcement reaches to
the ultimate capacity under bending whilst all other materials remain elastic range at
this state [21]. Component material's strains were calculated based on the plastic
neutral axis depth and concrete strain (0.003) and stresses were then calculated using
strains.
Fabrication and design of HCB is complex and was investigated as bridge decks.
This kind of system may not be suitable for building floors as slab components have
to be cast separately and only be used as a beam. Arch type compression
reinforcement may not be able to provide for shallow depth beam that may be
required in floor construction.
2.2 IMPORTANCE OF DEVELOPING A HYBRID COMPOSITE FLOOR
PLATE SYSTEM
As explained in the previous section, sandwich panels have been used in civil
engineering applications, such as cladding wall panels, short span floor panels and
roof panels, bridge decks transportation applications and dome-type shell structures.
A hybrid composite beam system has been investigated as structural member for
railway bridges. Sandwich panels can only be used as short span floor structures
because they become slender and deflect excessively when used as long span panels.
Moreover, sandwich panels exhibit de-lamination and wrinkling failures under
flexural loading as shown in Figures 2-2 and 2-3. Thin skin laminates exhibit
16 Chapter 2: Literature review
buckling failure as they are quite weak under compressive stress [9]. As a
consequence there is a limited use of sandwich panels in floor plate construction.
Despite the potential to use hybrid composites in floor construction to overcome the
limitations of sandwich panels, their applicability in floor plate construction has not
been adequately investigated and developed.
Figure 2-2: De-lamination failure of a sandwich panel [6]
Figure 2-3: Wrinkling failure of sandwich panels [9]
In most of the above sandwich panel applications, PU is used as a lightweight foam
core material because it bonds with the outer faces during the foaming process. [6, 8,
11]. A wide range of properties can be achieved with different densities of PU [13].
Thin steel laminates offers better performance under tensile stress but is subjected to
buckling under compressive stress [9]. In contrast, GRC offers better compressive
properties but is weaker in tension [20]. GRC is lighter in weight than conventional
concrete [12]. There is a higher potential to develop a hybrid system comprising
GRC, PU, and Steel laminates as a structural floor system. However, applicability of
hybridized GRC, PU and steel in structural floors system has not yet been investigated.
This research investigates the strength, serviceability and performance characteristics of
GRC, PU and steel hybrids that can be used in high performance lightweight floor plates
for buildings. Parametric studies will be carried out to develop a new floor plate system,
Hybrid Composite Floor Plate System (HCFPS), and design guidelines.
Chapter 2: Literature review 17
Proposed HFPS is assembled using component materials as shown in the Figure 2-1.
A low density PU core at the centre results in a lightweight structure. In the hybrid
assembly, individual component materials are combined to offset any weakness and
achieve the optimum performance. A cold formed thin perforated steel laminate is
placed at the bottom of the beam to improve tensile strength. GRC and PU are
configured to attract compressive and shear stress due to their superior performance
under such stresses [12, 14].
Figure 2-1: Hybrid material configuration in HCFPS
It is well understood that there are a number of weaknesses in available composite
applications with the use of GRC PU thin steel laminates. However, in proposed
HCFPS optimum performance is obtained by mobilising the strength characteristics
of the component materials to offset weakness in composite action and thereby
deliver superior performance as illustrated in Figure 2-2.
18 Chapter 2: Literature review
Figure 2-2: How to optimize the performance HCFPS using component material characteristics
2.3 PROPERTIES AND ADVANTAGES OF SELECTED MATERIALS
FOR HCFPS
GRC, PU and Steel are the component materials selected to develop HCFPS.
Properties of these materials will be discussed in this section.
2.3.1 Steel
Perforated steel laminate is used as tensile reinforcement to the HCFPS. Cold formed
steel facing have been used extensively as a outer facing of sandwich construction
applications [9, 18]. However, in such sandwich panels load bearing capacity is
usually determined by wrinkling in compression face and shear failure of the core
rather than yielding of the tensile face [9]. Thin steel sheets exhibit poor performance
under compressive stress but better performance under tensile stress in sandwich
constriction applications. Therefore, steel laminate is configured in the tensile zone
Weakness of component materials in available
applications
Tensile failure of GRC sheets
Wrinkling failure of thin steel laminates
Shear failure of PU core
Crushing failure of the core at the
supports
How to eliminate the weakness and combine the
positive inherant properties in HCFPS
Supplement with steel laminate as a reinforcement in tensile zone
Use GRC layes and PU core for
compression zone as GRC-PU-GRC sandwich panel
Combine GRC and PU to
improve shear strength in shear
zones
Supplement with GRC fill at
support regions to replace PU core
Chapter 2: Literature review 19
of the HCFPS and it is embedded in the GRC layer. Perforated steel laminate can be
used in order to enhance the bonding between GRC and steel laminate.
Mechanical properties
Available properties of mild steel and high strength steel plates are listed in Table
2.1. Typical stress-strain behaviour of high strength steel is shown in Figure 2-3.
Table 2-1: Typical values of steel properties [22]
Name Density
Kg/m3
Young's
Modulus
GPa
Shear
Modulus
GPa
Shear
Strength
MPa
Tensile
strength
MPa
Poisson'
s ratio
MPa
Mild Steel
High Strength Steel
7800
7800
210
210
80
80
105
230
210
550
0.3
0.3
Figure 2-3: Stress strain behaviour of high tensile steel of three similar specimens [22]
Environmental impact
The steel sheet can be protected from corrosion using a layer such as zinc or zinc-
aluminium. The metal layer is then protected with an organic (plastic) layer.
However, corrosion protection is not essential as the steel laminate is embedded in
GRC layer in HCFPS. Steel can be recycled and reused.
20 Chapter 2: Literature review
2.3.2 Polyurethane (PU)
PU foam has been selected as the core material of HCFPS. It has been used as a core
material in sandwich construction with steel plates to fulfil both insulating and
structural requirements [23]. PU is a substance categorized as a polymer based on
its chemical structure. It is manufactured by combining a polyol, isocyanate, a
blowing agent and an activator through a controlled chemical reaction as a liquid
form [9, 24]. Liquid foam starts to expand rapidly and harden after 3-6 minutes of
mixing. During the process, PU adheres to the face material [9].
Mechanical properties
PU can be formulated and processed into foams of varying densities. The foams are
identified either as flexible or rigid with the density ranging from as little as 16 kg/m3
to as much as 1240 kg/m3 [14, 24]. As a consequence mechanical properties vary
over this density range.
Sharma et al [14] have investigated the relationship between mechanical properties
and density of PU (density range 35 kg/m3 to 1200 kg/m
3). The relationship between
Elastic modulus (E), Shear modulus ), Shear strength (τ) and compressive (Xc)
strength with the density (ρ) is given in following equations [14, 25].
Equation 2-1
Equation 2-2
Equation 2-3
Equation 2-4
In above equations properties of solid PU: Density (ρs) is 1200 kg/m3, Elastic
modulus ( ) = 1.6 GPa, Shear modulus ( ) = 640 MPa, Shear strength = 38
MPa and Compressive strength ( ) = 127 MPa [14].
Chapter 2: Literature review 21
Compressive and tensile strength of PU foam increase with density [6, 23].
However, tensile strength and compressive strength are marginally similar for lower
densities. Table 2-1 summarise the properties of PU with deferent densities.
Table 2-2: Typical values of PU properties [14, 25]
Density
Young's
Modulus
MPa
Shear
Modulus
MPa
Compressive
Strength
MPa
Shear
Strength
MPa
Tensile
strength
MPa
Poisson's
ratio
MPa
44
64
100
200
300
500
1.2
10.9
23.4
76.1
151.6
361.2
1.5
2.5
5.7
21.2
45.9
121.3
0.125
0.412
0.743
2.415
4.800
11.50
0.172
0.426
0.556
1.807
3.600
8.580
0.150
0.340
1.000
-
-
-
0.3
0.3
0.3
0.3
0.3
0.3
Vibration and sound damping
PU can be used as a sound and vibration damping material because of the glass
transition property. Damping property is enhanced as PU can be deformed repeatedly
and still maintain its original shape [24]. Experimental investigation conducted on a
12 m long SPS grandstand double riser unit with a 32mm PU core and 3.8 mm face
plates, showed good dynamic performance with 2.1% damping ratio [8]. The
damping ratio would be 1% for steel or pre stressed concrete structures of similar
span[8]. This improvement was obtained using a very thin PU layer. Transmission
loss of PU is significant when it is used as core material and can be used in noise
reduction [26]. Yoon et al. [26] has observed marginal transmission loss for
sandwich panel, which is comprised of steel outer sheet and PU core. Hence PU can
be used as a core material in HCFPS not only to improve the damping properties but
also to reduce noise due to human walking.
Environmental impact
PU has low environmental impact because it is inert and non toxic and can be
recycled [27]. PU can be successfully recycled to usable formats. There are various
methods that are currently used, such as mechanical recycling, chemical processing,
thermo chemical processing, and energy recovery [28]. PU is derived from oil and
22 Chapter 2: Literature review
has a recoverable energy value that, in some cases, can be comparable to coal and
slightly less than fuel oil [28].
2.3.3 Glass fiber Reinforced Cement/Concrete (GRC)
GRC have been selected as the outer shell material of the HCFPS. GRC is a
composite material consisting of ordinary Portland cement, Silica sand and water that
is mixed with alkali-resistant glass fibers [29]. The percentage, arrangement and
method of application of Glass fiber can be engineered to suit a wide range of
applications. The advantage of glass fiber is for carrying the tensile force, thus
overcoming the main disadvantage of cement, which has low tensile strength. Spray
and premix are the two methods that have so far been developed for the fabrication
of GRC components.
The method of premix
Constituents (mortar and pre-cut glass fiber) are mixed together into a paste in this
processes and cast by pressing, concreting and vibration [12, 30]. Fiber content of
3.5-4% in total weight is used for this process to facilitate workability and
compaction [12, 31].
Method of spray
Mortar paste and chopped glass fiber are simultaneously deposited from dual spray-
heads into suitable moulds in this process. This method currently account for a large
percentage of GRC production. Three ways of spraying GRC can be seen; manual
spray, mechanized spray and spray-dewater process [30]. Pre-cut glass fiber is
between 25 mm and 40 mm in length and contain about 5% of the total weight of the
mix [12].
Mechanical and Material properties of GRC
Mechanical properties of the GRC varies according to the production method and the
percentage of fiber present in the mix [29, 30]. Table 2-2 presents mechanical
properties for GRC that are cast using two mixing methods. Optimum of glass fiber
content is 5% for spray up method and 3.5% for premix method in total weight [29,
32, 33]
Chapter 2: Literature review 23
It has been shown that the addition of a polymer to GRC improves the performance
and mechanical properties [34]. Improvement in bending strength and ultimate
tensile strength is marginal for polymer modified GRC compared to normal GRC.
Furthermore, polymer addition substantially increases the durability under natural
weather conditions [34].
Table 2-3: Typical values of GRC properties [29, 32, 33]
Mechanical Property GRC
spray up
GRC
premix
Dry density (kN/m3) 19-21 18-20
Compressive strength (MPa) 50-80 40-60
Young modulus (GPa) 10-20 13-18
Poisson ratio 0.24 0.24
Tension
Ultimate tensile strength (MPa)
Shear
8-11
4-7
Inter lamina shear strength (MPa)
In plane shear strength (MPa)
Bending
Ultimate strength (modulus of rupture) (MPa)
Elastic limit (limit proportionality) (MPa)
3-5
8-11
20-30
7-11
N.A
4-7
10-14
5-8
In the methods explained above, chopped glass fibers are added to the mix and
therefore, fiber arrangement in the matrix is random. Although GRC shows tensile
strength values between 4 to 11 MPa, it provides much better performance under
compression. Ferreira [12] showed that GRC provides greater ductility under
compression when compared with plain mortar.
Environmental impact
Environmental impact of GRC is less than that of reinforced concrete [32]. The main
constituents of GRC are based on naturally occurring earth oxides such as silica
sand, Portland cement and glass fibers. Course aggregates are not a constituent of
GRC because it is manufactured as thin layers (6 mm -15 mm). GRC has reduced
cement and aggregate usage than reinforced concrete [32]. As a consequence GRC is
lightweight and has lower embedded energy. GRC can be recycled using lower
energy at the cessation of use.
24 Chapter 2: Literature review
2.4 SUMMARY AND IMPLICATIONS
This chapter has discussed the currently available lightweight sandwich and hybrid
composite applications in civil engineering construction using current literature.
Their properties were reviewed to find limitations to their use in structural floor plate
construction. Reasons and the importance of developing a Hybrid Composite Floor
Plate System (HCFPS) were discussed considering those limitations of current
composite applications. Properties of materials that are selected to develop HCFPS
were also discussed. The followings conclusions and arguments are made following
the literature review.
Sandwich panels, which are comprised of thin outer shell and central core, have been
studied and used in wall panels, short span floor panels and roof panels, bridge decks
and dome-type shell structures. A hybrid composite beam has been investigated as
structural member for railway bridges. Sandwich panels can only be used as short
span floor structures, because they become slender and deflect excessively when
used as long span panels. They exhibit de-lamination and wrinkling failures under
flexural loading as thin skin laminates are weaker under compressive stress.
Although hybrid composites may be used in floor construction to overcome the
limitation of sandwich panels, it has not yet been investigated.
GRC, PU and steel have been extensively investigated in sandwich construction
applications. However, applicability of hybridized GRC, PU and steel in structural
floors system has not yet been investigated. This research investigates the strength,
serviceability and performance characteristics HCFPS using GRC, PU and steel, to
develop a lightweight floor plates for buildings.
GRC, PU and steel laminates exhibit a number of weaknesses in available composite
applications. However, in proposed HCFPS optimum performance is obtained by
mobilising the strength characteristics of the component materials to offset
weaknesses in proper hybrid configuration and thereby obtain optimum performance.
Properties of steel, PU and GRC have been well investigated separately and that data
can be used to investigate performance characteristics of hybridized GRC, PU and
steel laminate.
Chapter 2: Literature review 25
Steel, GRC and PU can be recycled to usable formats. Therefore, HCFPS can be
developed as an environmentally friendly structure, which can also be recycled at the
cessation of use.
26 Chapter 2: Literature review
Chapter 3: Methodology 27
Chapter 3: Methodology
This chapter describes research methodology and the various stages through which
the HCFPS was developed. It also makes references to appropriate codes or
standards and literature used in development.
3.1 HCFPS DEVELOPMENT
This research developed a Hybrid Composite Floor Plate System (HCFPS) using
steel, Polyurethane (PU) and Glass-fiber Reinforced Cement (GRC) according to the
following stages. In first stage, configuration of the HCFPS using component
materials was determined as explained in Sections 1.2 and 2.2. In second stage,
experimental testing of HCFPS panels was conducted to investigate static, dynamic,
cyclic loading and post yield behaviours, mode of failure and ductility. Component
material testing was also conducted to obtain their properties. FE models were
developed and validated using test results generated from experimental testing. In
third stage, parametric studies were performed and HCFPS were developed to satisfy
static performance requirements. FE techniques were used to study the influence of
parameters such as sectional configuration, spans, loading limits and support
condition. The vibration characteristics of HCFPS under human-induced loads were
investigated in the next stage and the required improvements for an optimum design
that satisfy both static and dynamic design criteria were determined. Finally,
guidelines were developed for the design of HCFPS panels.
3.2 HCFPS CONFIGURATION AND SYMMETRY
The proposed HCFPS was assembled using component materials as shown in Figure
3-1. The width of the HCFPS was limited to 2 m to suit prefabrication and
transportation requirements, however it can be varied if necessary. Length of the
HCFPS can be varied by changing the material properties and sectional
configuration. HCFPS, which uses a steel beam column frame system to support it,
can be graphically represented in a building floor as depicted in Figure 3-2. Joint
details between adjacent HCFPS panels and connection methods for supporting
beams will be described Chapter 6, Section 6.8.2.
28 Chapter 3: Methodology
Figure 3-1: Proposed HCFPS panel configuration and symmetry
Figure 3-2: Graphical representation of building floor using HCFPS panels
A half of the HCFPS section can be investigated for the research studies as shown in
Figure 3-1 as the panel is symmetrical and hence results for section with a single
beam are applicable to capture the behaviour of proposed HCFPS panel
configuration with double beams. Therefore, a HCFPS panel with a single beam was
used for the purpose of static and dynamic experimental investigations and computer
model validation. Moreover, experimental testing on HCFPS panel with single beam
was more economical and efficient. Validated FE model was used for the parametric
studies to investigate static and dynamic performance characteristics. Single beam
HCFPS section used for all static performance investigations is adequate as static
behaviour of HCFPS floor panel with multiple beams will be similar. However,
dynamic behaviour of single beam HCFPS could vary from the HCFPS floor panel
with multiple beams due to the change in model mass. Nevertheless, validated FE
model for the single beam HCFPS panel was extended to investigate overall dynamic
Symmetry axis
HCFPS Floor
Steel beams and
columns
Chapter 3: Methodology 29
behaviour of proposed HCFPS configuration with two beams (refer to Figure 3-1)
and building floors using HCFPS panels (refer to Figure 3-2).
3.3 EXPERIMENTAL TESTING
3.3.1 Test Panel
Experimental investigations were conducted on three 3200 mm span HCFPS panel
with single beam as shown in Figure 3-3a. Span and section configuration were
determined by a preliminary FE analysis. Manual casting procedure was used to cast
the test panel and the beam was tapered (Figure 3-3(a)) to facilitate the casting
process. The PU core was replaced with a 100 mm GRC in the vicinity of the
supports, as shown in Figure 3-3(b). Perforated steel laminate was used to enhance
the bonding between GRC and steel laminate.
(a) Section dimensions
(b) GRC fill replacing PU core near supports
Figure 3-3: 3200 mm span HCFPS test panel configuration
30 Chapter 3: Methodology
3.3.2 Heel impact test
The heel impact test is a widely used test to investigate the dynamic behaviour of
floors. A person of average weight stands at the centre span of the structure, and to
raise their feet onto their toes approximately 50 mm off the surface and makes a
sudden impact with their heels [22]. This creates a sudden impact force on the floor.
The floor acceleration response is recorded in order to calculate the natural
frequencies and the damping coefficient.
Heel impact tests were performed on all three test panels and an average person (70
kg) was asked to create the heel impact [22] in this study. A 5g accelerometer was
used at mid-span to acquire the vibration response (as time acceleration plots).
Equation 3-1 presented by Ellis [35] was used to calculate the damping coefficient
(ς) from the time- acceleration plots. In this equation, A0 and An are the amplitudes
of ―n‖ successive peaks of the acceleration-time response plot (refer to Figure 3-4).
Damping obtained from this equation is termed ―log decrement damping‖. Murray
[36] stated that model damping, or true damping, is one-half to two-thirds of the
value of the log decrement damping. The first natural frequency of the test floor
panels was calculated by performing a Fast Fourier Transformation (FFT) analysis
for acceleration plots [22, 37, 38].
n
oe
A
A
nlog
2
1
Equation 3-1
Figure 3-4: A typical acceleration response [35]
Chapter 3: Methodology 31
3.3.3 Static load test
Static loading tests were carried out in order to obtain the load deflection curve to
calibrate the finite element model. Failure modes and ductility of the specimen were
investigated under static loading. Static loading test of the HCFPS panel was
performed according to the loading arrangement shown in Figure 3-5 (a and b). Two
test panels were tested under static loads. Steel plates 10 mm thick and 100 mm
wide, were placed under the slab and beam at the supports, and the plates were
supported by solid circular steel bars, as shown in Figure 3-5(b). Adjustable jacks
were used to support the steel bar under the slab as seen in Figure 3-5(c).
(b) Longitudinal view of loading arrangement supporting method
(c) Side view of the supporting arrangement
Figure 3-5: Static load test arrangement
(a)Loading arrangement
32 Chapter 3: Methodology
Loads were applied as four line loads along the span using steel spreader beams
through an arrangement shown in Figure 3-5(a). This arrangement was adequate to
simulate a uniformly distributed load (within the means of our testing facilities).
Furthermore, it enabled curvature of the panel during loading, whilst maintaining
uniform loads at the contact locations, similar to the test set up in [15]. Linear-
variation-displacement-transducers (LVDTs), with 0.01 mm sensitivity, were placed
at centre of span to measure deflections. Details of test procedure and
instrumentation will be discussed in Chapter 4.
3.3.4 Cyclic loading test
Cyclic loading test for one HCFPS test panel was conducted according to the
method given in [39]. This method can be adopted to evaluate the performance of
slabs comprised of new materials. The same loading and supporting arrangement that
was used for the static loading test was used for the cyclic loading test. Tests were
conducted using a hydraulic loading system, consisting of a displacement controlled
Moog actuator to control the cyclic loading. Six loading cycles were conducted as
shown in Figure 3-6. Applied load at 50% of the ultimate load (Lus) for the first two
cycles, 75 % (Lus) for next two cycles and 100% (Lus) for the final two cycles was
applied. Ultimate load (Lus) of HCFPS test panel was determined by the static load
testing. Duration of loading steps was maintained throughout the cyclic loading test,
as shown in Figure 3-6. A minimum of 10% of Lus was maintained during the
unloading. Deviation from linearity and Repeatability [39] for the HCFPS test panels
were calculated as explained below and results will be presented in Chapter 4.
Figure 3-6 Load cycles and steps of loading for cyclic loading test [39]
Chapter 3: Methodology 33
Deviation from linearity
Typical load-deflection results for cyclic load test are shown in Figure 3-7. Deviation
from linearity is a measure of non-linear behaviour of the tested member. Acceptable
value should be less than 25%. and this is calculated using Figure 3-7 and Equation
3-2 [39].
Figure 3-7: A typical load-deflection curve for six load cycles [39]
Equation 3-2
Equation 3-3
Repeatability
Repeatability of deflection is the ratio of deference between maximum and residual
deflections for a set of two identical cycles. Repeatability of the member deflection
can be calculated using Equation 3-4 and Figure 3-8 [39]. Repeatability is calculated
for the three sets of similar cycles. If the repeatability is greater than 95% it is
considered acceptable [39].
34 Chapter 3: Methodology
Figure 3-8: Parameters to calculate the Repeatability [39]
Equation 3-4
where: is the maximum deflection in Cycle B under a load of Pmax,
is the
residual deflection after Cycle B under a load of Pmin,
is the maximum
deflection in Cycle A under a load of Pmax, and is the residual deflection after
Cycle A under a load of Pmin.
3.3.5 Material testing
Mechanical properties of the materials used in the test panels were determined
separately by conducting a comprehensive material testing program. Standard testing
methods given in ASTM, ISO and Australian Standards (AS) were used as shown in
the Table 3-1. Only a tensile test was conducted for the steel laminate, as steel
laminate in the test panel acts as reinforcement and therefore is subjected to tensile
stress. Test samples and testing instruments were selected according to standard test
methods. Detailed testing procedure and instrumentation will be discussed in Chapter
4.
Chapter 3: Methodology 35
Table 3-1: standard test methods for material property investigation
GRC PU Steel
Tensile Strength
Tensile modulus
Flexural Strength
Flexural Modulus
Compressive strength
Compressive modulus
ASTM D 3039 [40]
ASTM D 3039 [40]
ASTM C 947 [41]
ASTM C 947 [41]
AS 1012.9 [42]
AS 1012.9 [42]
ISO 1926 [43]
ISO 1926 [43]
ASTM D 790 [44]
ASTM D 790 [44]
ASTM C 365 [45]
ASTM C 365 [45]
AS 1391 [46]
AS 1391 [46]
-
-
-
-
3.4 FINITE ELEMENT (FE) METHOD OF ANALYSIS
FE analysis was used extensively in this research to simulate HCFPS behaviour and
thus investigate its performance characteristics. It is a cost- and time-efficient
method compared with physical experiments. FE simulations have been extensively
used for studies with the support of limited experimental testing.
The commercially available finite element program ABAQUS 6.9-1, (Dassault
Systèmes Simulia Corp. [47]) was used in this research project with ABAQUS CAE
as the pre- and post-processor for the FE simulations [47]. ABAQUS involves the
following three major phases.
3.4.1 Pre-processing
Using ABAQUS CAE pre-processing of the FE models was conducted. At this stage,
FE models were simulated by defining an appropriate finite element mesh, assigning
suitable material properties and applying boundary conditions (restraint or
constraints) and loads. An input file (.inp) was then prepared to submit for analysis.
3.4.2 Analysis and solutions
Analyses of the input files were conducted using ABAQUS 6.9-1, which is installed
in the High Performance Computing (HPC) facility in QUT. HPC consists of a 400
processor SGI Altix XE Cluster with 960 GB memory.
Post-processing
Analysis &
Solutions
Pre-processing
36 Chapter 3: Methodology
In this phase, the submitted input data is assembled into matrix format and the
numerical analysis is executed. The matrix assembly process of a multi-degree of
freedom system is governed by Equation 3-5.
tFxKxCxM Equation 3-5
In above equation, )(tF is the applied load vector, [M] is the mass matrix, [C] is the
structural damping matrix, [K] is the stiffness matrix, }{x is the displacement vector,
}{x is the velocity vector ))(( dttdxx and }{x
is the acceleration vector
))(( 22 dttxdx .
The element type and material properties are used to define the mass matrix [M],
structural damping matrix [C] and stiffness matrix [K]. Damping matrix was derived
using Reyleigh proportional damping method as a combination of the mass and
stiffness matrices [22]. Acceleration vector }{x , velocity vector }{x and
displacement vector }{x are developed based on the boundary conditions. Applied
external loads on the system are used to define the applied load vector )(tF .
In the analysis and solution phase, the above equation is analysed using displacement
method or stiffness method and solved for displacement and stresses of the system.
In ABAQUS analysis, the output was obtained as a result file (.odb).
3.4.3 Post-processing
Post-processing analysis of output files (.odb) was conducted using ABAQUS CAE
graphically and numerically to interpret the results.
3.5 FE MODEL VALIDATION
A Finite element model of the experimentally tested HCFPS was developed for the
experimentally tested panel. This model was validated with static and dynamic
experimental results. Load-deflection behaviour, first natural frequency and
acceleration response were used for validation and detailed procedure will be
presented in Chapter 5.
Chapter 3: Methodology 37
3.6 STATIC PERFORMANCE AND PARAMETRIC STUDIES
The validated model was further developed using FE techniques in order to
investigate static performance characteristics of HCFPS by conducting parametric
studies. Sectional configuration, material properties and span lengths were used as
variable parameters in this study. First, sectional configurations were determined for
different spans (3m to 7.5m) for two loading conditions as given in Table 3-2.
Superimposed dead load, which gives a total of 1.0 kPa, is summarised in Table 3-5
[48]. This value may be increased up to 2.0 kPa in residential buildings due to a
higher density of partition walls and heavier floor finishes. In the present study, a
uniform superimposed dead load of 1.0 kPa was for parametric studies, which is
likely to be present on this floor system according to Table 3-3. However, analysis
results will demonstrate that higher super imposed load can also be incorporated
easily for this floor system as required.
Table 3-2: Loading cases [48]
Floor Type Gk Dead Load (kPa)
(Fixed partition and finishes) Qk Live Load (kPa)
Residential Self-weight +1.0 1.5
Offices or work areas Self-weight +1.0 3
Table 3-3: Super imposed permanent dead loads for an office floor
Type of loading Load, kPa
Floor finishes acoucstic insulation + cladding
Suspended ceiling
Suspended services
Lightweight partition, furniture and equipments
Fire protection
0.25
0.10
0.15
0.35
0.15
Flexural and shear stresses and deflection limits were investigated as the
performance characteristics to obtain the optimum configuration. Ductility and creep
and shrinkage deformation of the HCFPS were also evaluated for HCFPS.
Calculation procedure for the design of HCFPS was determined. This process will
be presented in Chapter 6.
38 Chapter 3: Methodology
3.7 DYNAMIC ANALYSIS OF HCFPS
Developed HCFPS configurations, which satisfy static performance, were analysed
under human induced loads. Serviceability and comfort criteria given in current
design standards and research outcomes were used to evaluate the dynamic
performance of HCFPS. Lightweight HCFPS can be excited under human induced
loads and should be designed to comply with the comfort and serviceability
requirements for human perception. International codes, which define vibration
response limits and the method used to evaluate the dynamic performance are
discussed in this section.
3.7.1 Human induced walking loads
During any human activity, force is repeatedly applied to the floor causing a dynamic
excitation in the floor. A typical pattern of a human induced force - time history is
shown in Figure 3-9.
A number of research studies have been conducted by Bachmann et al.[49], Allen et
al. [50], Maguire et al. [51], Ginty et al. [52], Simith [53], and Ellis et al. [54] to
obtain numerical formulae to express human induced dynamic forces. ISO 10137
(2007) [55], Ellis et al.[54] and Smith et al. [56] provides the latest numerical
expressions (Refer to Equation 3-6) to find vertical dynamic loads resulting from
various human induced actions, such as walking , running and jumping. Idealised
load model given by Equation 3-6 closely matches with the dynamic load resulted
from the actual human loads, which demonstrates in Figure 3-9. Hence, this load
model was selected to use in this research for define walking loads on the HCFPS
panel in FE analyses.
Figure 3-9: Dynamic vertical force due to walking step by a person [57].
Chapter 3: Methodology 39
Equation 3-6
In above equuation, F(t) is the dynamic force, Q is the static weight of the
participating person, n is the Fourier coefficient corresponding to nth
harmonic, fp
is the pacing frequency, t is the time and υn is the phase angle of the nth
harmonic, n
is the integer designating harmonic of the fundamental and k is the number of
harmonics that characterise the forcing function in the frequency range of interest.
Q should be to be taken as 746 N according to [56]. Bachmann et al. [49] suggested
different values of fp for various pacing of walking. Numerical coefficients for those
paces can be found in ISO 10137 [55] as shown in Table 3-4. It is suggested in [55]
to use phase shift of 900 for the harmonics in order to obtain conservative results.
The above explained parameters are for evaluating dynamic load due to a single
person. However, dynamic response of a floor structure due to a walking crowd is
normally is similar to the individual effect as Ellis [58] observed similar acceleration
levels due to a single person walking and group of people walking. This is mainly
because of the un-coordinated nature of walking human activities [55, 58] and weight
of passive people increasing the effective model mass. ISO 10137 (2007) [55]
suggests a coordination factor to incorporate the dynamic crowd effect. In this
method, dynamic load due to the walking group can be represented by applying a
coordination factor C(N) to the forcing function (refer to equation 3-7).
Table 3-4: Design parameters for walking and running loads [49, 55]
Mode of walking fp (Hz) Numerical coefficient for 1st four harmonics
1 2 3 4
Slow Walk
Normal Walk
Fast Walk
Running
1.7
2.0
2.4
3.0
0.26
0.37
0.52
1.40
0.1
0.1
0.1
0.4
0.06
0.06
0.06
0.10
0.06
0.06
0.06
-
)().()( NCtFtF N Equation 3-7
For example, if the movements of a group of people are un-coordinated, the
coordination factor becomes:
40 Chapter 3: Methodology
NNNC /)( Equation 3-8
Where N is the number of participants
3.7.2 Dynamic analysis and floor classification
Modal frequency
Modal frequency of HCFPS can be determined by free vibration analysis using FE
techniques. Additionally, simplified analytical methods suggested in the literature
(refer to Equation 3-9) can also be used. However, such methods are applicable only
for beams. Therefore, an adjusted equivalent beam section for HCFPS panel can be
used, but this method will be approximate compared to FE techniques.
32 mL
EIKf n
n
Equation 3-9
In Equation 3-9, EI is dynamic flexural rigidity of the member, m is the effective
mass, L is the span of the member and nK is a constant representing support.
High frequency and low frequency floors
HCFPS is a floor system associated with a lower mass and as a consequence, the first
natural frequency of the system is higher and thus categorised as a high frequency
floor. High frequency floors are defined as floors having a first natural frequency of
10 Hz which is greater than the fourth harmonic of the walking frequency (maximum
of 2.5 Hz) [56]. Preliminary FE studies shows that HCFPS gave a first natural
frequency greater than 10 Hz. As a consequence, HCFPS does not exhibit resonance
vibration due to human activities. Load due to walking in high frequency floors, acts
as an impulsive force, which diminishes before the next step [56].
Steady state and transient state
Response of a floor structure can be categorised as either a transient or steady state
response. Steady state response occurs once the waveform settles down in the
structure. Transient response is when the response of the floor occurs before the
steady state. Resonance can occur with a higher vibration response in low frequency
floors. In contrast, in high frequency floors the transient response is more significant
than the steady state response as resonance cannot occur and applied forces behave
Chapter 3: Methodology 41
like a series of impulses [56]. Therefore, transient response of the HCFPS is used to
assess the required human perception limits.
Vibration response
Response of a floor structure is measured in terms of acceleration and [55, 56, 59,
60] provides acceleration limits to satisfy human perception. The two values to
represent present the acceleration of a system are peak acceleration and Root-Mean-
Square (RMS) acceleration. Peak acceleration is the highest value of acceleration
resulting from an excitation. Peak acceleration cannot give an indication as to the
duration of time that the floor is subjected to the peak acceleration value. In contrast,
RMS acceleration is an average measurement of the acceleration time -history as
shown in the Equation 3-10. Smith et al. [56] stated that sharp peaks of acceleration
are less significant with lower RMS acceleration. Acceleration time-history can be
obtained from the dynamic FE analysis of the HCFPS and is used to calculate RMS
acceleration.
Equation 3-10
In Equation 3-10: T is the period under consideration, a(t) is the acceleration function
and t is the time.
3.7.3 Acceptance criteria
RMS acceleration limits
ISO 10137 [55] and BS 6472 [60] provide base values of RMS accelerations in
relation to a base curve, which is frequency weighted, to assess human response to
vibration. Moreover, a series of multiplying factors is also presented to adjust the
base curve to suit various environments in which the floors are used. Frequency
weighting factor is used as human perception to vibration varies with frequency [56].
The frequency weighted RMS acceleration base curve for the vertical direction
vibration is presented in Figure 3-10. RMS acceleration values above the curve are
perceptible for humans and values below the curve are imperceptible and do not
cause vibration problems. The base value of RMS acceleration is 5 × 10-3
m/s2 and
because of the frequency weighting, acceptable base value increases for the lower
and higher frequencies as shown in the figure.
42 Chapter 3: Methodology
Figure 3-10: Frequency weighted RMS acceleration base curve [55]
Response Factor (R)
This base curve is not used directly in practical applications and rather the evaluation
method suggested is the Response factor (R) method. The R of a floor is the ratio
between calculated or measured frequency weighted RMS acceleration ( and
base acceleration value [56].
Equation 3-11
Acceptable R values are suggested for floors depending on their use in [56]. HCFPS
is expected to be used in residential and commercial floors. Hence, response factors
suggested in Table 3-5 were used for dynamic evaluation of HCFPS.
Table 3-5: Recommended response factors [56]
Place Response factor
Residential
Offices
Workshops
Shopping malls
Day
Night
Day
Night
Day
Night
Day
Night
2-4
1.4
4-8
4-8
8
8
4
4
Chapter 3: Methodology 43
Vibration Dose Value (VDV)
Primarily, R values are determined by considering the continuous vibration of a
floor. In this context, human induced excitation has to be applied as a continuous
load to the floor by using the load model suggested in Equation 3-6. This will give
the worst possible loading scenario and response. If the RMS acceleration calculated
from this method gives a R value less than the values in Table 3-5, the floor vibration
is acceptable. However, continuous loading and vibration are uncommon and human
induced walking loads are intermittent [56]. Therefore, it is suggested to consider the
intermittent nature of the vibrations if the calculated R value is higher than
acceptable values.
In the process of vibration assessment cumulative measure of vibration response for
intermittent activities is more reliable and needs to be used for determining
perceptible levels. ISO 10137 [55] and BS 6472 [60] provide perceptive tolerance
levels for the intermittent vibrations using Vibration Dose Value (VDV ) and can be
calculated using Equation 3-12. Where, is the frequency weighted acceleration
and T is the to total period of the day during which vibration may occur [60] .
Acceptable VDV values suggested in BS 6472 [60] are shown in Table 3-6 for the
probable adverse effect.
Equation 3-12
Table 3-6: VDV values (m/s1.75
) for vertical direction vibrations [60]
Place Low probability of
adverse comment Adverse comment possible
Buildings 16h day
Buildings 8h night
0.2 - 0.4
0.13
0.4-0.8
0.26
Ellis [61] suggests a method (refer to Equation 3-14) to calculate VDV values of a
walking activity by using the in the design stage of floors and this method
was used in this study. Where, na is the number of time the activity will take place in
the exposure period and Ta is the duration of an activity (time taken to walk along the
floor)
Equation 3-13
44 Chapter 3: Methodology
Suggested VDV values can be substituted in Equation 3-14 by rearranging Equation
3-13 and thus calculate the number of occurrences (na) for evaluation purposes. The
number of occurrences (na) for that activity can then be evaluated depending on the
floor use. If na is unlikely to occur in service, floor will not exceed to threshold VDV
value. In this way the dynamic performance of HCFPS floor was evaluated for use in
office and residential floors. Dynamic evaluation of HCFPS panels will be presented
Chapter 7.
Equation 3-14
3.7.4 Damping
All physical systems have some inherent damping. Damping can be either external or
internal. The material or contact areas within structures such as bearings and joints,
are classified as internal damping materials. External contacts, such as non-structural
elements, are classified as external damping materials. In a floor structure, not only
the components of the structural system but also the non-structural components such
as finishes, partitions and standing objects play a major role in providing damping.
The amount of damping in a structure is determined by a damping ratio. Damping of
floors can greatly decrease the response of a structure driven at a resonant
frequency[62]. However, HCFPS may not give resonant response at a high frequency
due to human induced loads, and impulsive response might be reduced with the
damping.
There are a number of damping levels reported in the literature for different
structural types. In general, damping for bare composite floors is reported to be
between 1.5% - 1.8% [63, 64]. However, use of partitions on the finished floor
system may provide higher damping for the floor. Murray [36] identified a mild
damping level of 3% for an office without permanent partitions. Higher damping
could also arise in a floor with permanent, drywall partitions and this may be as
much as 5% - 6% [36]. Elnimeiri et al.[65] also recommended a damping coefficient
of 4.5% - 6% for finished floors with partitions. Moreover, Brownjohn [66] showed
that damping could increase to 10% depending on the position of items such as
cabinets, bookcases and desks. On the other hand, Sachse [67] proved that the
presence of stationary humans will increase the damping of the structure up to 12%.
Chapter 3: Methodology 45
Different damping arrangements have been considered within the literature to
prevent large vibration responses. Ljunggren [68] used resilient ceiling connected to
a lightweight floor by a visco elastic material and found a more substantial effect on
damping than the floor itself. Further, Visco-Elastic (VE) dampers can be used
effectively to prevent higher serviceability displacement and acceleration in steel
deck composite floors [22].
Damping characteristics of a sandwich structure can be enhanced by the
polyurethane core. Experimental investigation conducted on a 12 m long SPS
grandstand double riser unit with of 32 mm PU core and 3.8 mm face plates, showed
good dynamic performance with 2.1% damping ratio as given in reference [8]. The
damping ratio would be 1% for steel or pre stressed concrete structures of similar
span[8]. This improvement was obtained using a thin PU core. Thicker PU core of
the HCFPS enhances damping properties.
Damping in the floor structure is difficult to determine and only approximate values
can be provided depending on the floor type and external damping. However, by
using partition and ceiling, external damping can be increased by about 2%
according to the above literature. The inherent damping properties of HFPS were
investigated using experimental studies and external damping due to partition was
considered as 2% for parametric studies.
3.8 DYNAMIC PERFORMANCE AND PARAMETRIC STUDIES OF
HCFPS
Dynamic performance characteristics were investigated for the HCFPS sections,
which had met static performance requirements, using FE techniques as explained in
Section 3.7. RMS acceleration, Response factor (R) and VDV were used as dynamic
performance assessment measurements. Parametric studies were conducted to
investigate the influence parameters on acceleration response and VDV values.
Dynamic evaluation of HCFPS panels will be presented Chapter 7.
3.9 DESIGN GUIDELINES
Guidelines were developed to design the HCFPS system. Thickness of hybrid
components, density of PU, sectional configurations and loading were considered as
the variables to obtained optimum section for particular span. Design calculation
46 Chapter 3: Methodology
procedure was developed as a design tool. Design guidelines and manufacturing and
installation guidance of prefabricated HCFPS panels will be presented Chapter 8.
3.10 SUMMARY
HCFPS was developed in stages as described in this section. The entire process is
illustrated in the flow chart in Figure 3-11.
Figure 3-11 Development stages of HCFPS
Dtermination of proper configuration for HCFPS using componenet materials
Experimental testing and FE model development and validation for static and dynamic performance
Development of HCFPS to satisfy static performance by condcting paramertric studies using validated FE techniques
Dynamic performance evaluation of HCFPS using validated FE techniques
Development of design gudelines
Chapter 4: Experimental Testing 47
Chapter 4: Experimental Testing
This chapter presents the experimental investigations conducted on 3200 mm span
HCFPS panels and GRC-PU-GRC composite panels, which is top slab of HCFPS.
Material property investigation, fabrication of test panels, test setup, instrumentation
and experimental results are presented.
4.1 MATERIAL PROPERTY INVESTIGATION
A comprehensive test program was carried out to determine the material properties
of the component materials (GRC, PU and steel) used to manufacture HCFPS test
specimens. Both PU and GRC were tested in tension, compression and bending,
whilst the steel sheet was tested in tension. This section presents the details of the
experimental test program and results.
4.1.1 Formulation of GRC
GRC is a cementitious matrix, comprising of cement, sand, water, admixtures and
short-length alkali-resistant glass fibers [31]. All GRC samples were cast using a pre-
mix production method and the formulation of the constituent materials is given in
Table 4-1.
Table 4-1: Formulation of GRC
Constituent materials Percentage of total weight
(%)
Cement (general purpose grey)
Sand (fine washed)
Metakaolin (Power Pozz)
Polymer (Vinnapas 512T)
Super plasticizer
Water
Pre- cut alkali resistant glass fiber
33.8
33.8
8.5
4.1
1.0
14.8
4.0
GRC consists of a random fibre arrangement because glass fibers are mixed with the
cement sand mortar. Therefore, GRC can be treated as an isotropic material and
techniques required by the analysis of GRC stresses, strains and deflections are
identical to those used with isotropic materials [69]. All the GRC samples were
48 Chapter 4: Experimental Testing
selected randomly irrespective to the casting direction. Random specimens exhibitted
similar results in each tensile, compression and bending supporting this argument.
4.1.2 Tensile testing for GRC
The maximum thickness of the GRC layers used for the HCFPS test specimens and
tensile test specimen was 10 mm. Sample size was selected as 250×25 mm
according to ASTM 3039 [40]. Five tensile test specimens were used to represent all
GRC batches. Five GRC batches were prepared using the formulation given in Table
4-1 to cast HCFPS test panels (refer to section 4.2) and five tensile coupons were
cast from the each batch. Uniaxial tension tests were carried out using an Instron
5569 series Mechanical Tester, at a loading rate of 0.5 mm/min as shown in Figure 4-
1a. Longitudinal strains were measured using the built-in extensometer of the test
machine and the stress-strain relationships are illustrated in Figure 4-1b. Average
cracking tensile strength and tensile modulus were obtained as 3.1 MPa and 5.0 GPa,
respectively. Figure 4-1a shows a typical failure mode of a test coupon.
(a) (b)
Figure 4-1: (a) Tensile testing for GRC specimen, (b) Stress-strain relationships for GRC in
tension
4.1.3 Four-point bending tests for GRC
Four-point bending tests were carried out to investigate the flexural behaviour of
GRC. Test specimens for the four-point bending tests were 10×25×200 mm with 152
mm clear span according to ASTM C947 [41]. Testing was conducted using the
same test machine, as per section 4.1.2, with the loading set up shown in Figure 4-2a,
at a 1 mm/min displacement rate. Five test samples representing all GRC batches
0.00
0.50
1.00
1.50
2.00
2.50
3.00
3.50
4.00
0 0.002 0.004 0.006 0.008 0.01 0.012
Str
ess (
MP
a)
Strain (mm/mm)
Chapter 4: Experimental Testing 49
were used and resultant load deflection plots are given in Figure 4-2b. All the test
samples exhibited a similar initial linear behaviour, but beyond the yielding point
two samples exhibited slightly different behaviour. This could be due to the random
fiber arrangements in the test specimens. Flexural modulus was calculated using the
initial linear behaviour of the plots according ASTM C947 [41] and gave an average
value as 4.8 GPa. The average force at which the force-deflection curve deviates
from linearity was 160N. This value was used to calculate the flexural strength of
GRC as 9.7 MPa, according to the method given in [41].
(a) (b)
Figure 4-2: (a) Four-point bending test for GRC, (b) Load-deflection plots for GRC in four-
point bending tests
4.1.4 Cylinder compression test for GRC
Compression tests were performed to obtain the stress-strain behaviour of GRC in
compression. Three cylindrical specimens of 100 mm diameter and 200 mm height,
representing all GRC batches, were used for this test. Specimens were tested using
Universal Tinius Olsen test machine as shown in Figure 4-3a. Cross-head
displacement was used to calculate the strains. The stress-strain relationships for
GRC in compression are illustrated in Figure 4-3b. Compressive modulus and
compressive strength were calculated as 5.9 GPa and 19.6 MPa respectively.
50 Chapter 4: Experimental Testing
(a) (b)
Figure 4-3: (a) Compression test for GRC, (b) Stress-strain relationships for GRC in
compression
4.1.5 Composition of PU
AUSTHANE AUE 757 rigid medium-density PU foam (density = 99.8 kg/m3) was
used as the central core of the HCFPS test panel. This foam is generally obtained by
mixing AUE 757 Polyol and ECOISO-GP Isocyanate liquids under controlled
conditions.
Although the PU production was conducted in layer wise pouring process, test
samples were selected randomly from the from the foam core irrespective to the
casting procedure. Random specimens exhibitted similar results in each tensile,
compression and bending demonstrating isotropic nature of PU and this was further
observed by other researchers in [6]
4.1.6 Compression test for PU
Compression tests were conducted to obtain the compressive behaviour of the PU
core. Five prism-shape PU foam coupons, which were 70×70 mm in cross-section
and 50 mm thick, representing all PU batches, were manufactured and tested
according to the ASTM C365-03 [45]. The tests were carried out using an Instron
5569 series Mechanical Tester, by attaching flat loading platens (refer to Figure 4-4a)
with a displacement rate of 0.5 mm/min. The built-in extensometer of the test
machine was used to measure the strains. Stress-strain relationships are presented in
Figure 4-4b, which shows an initial linear response, then a plastic response, followed
0
5
10
15
20
25
0 0.005 0.01 0.015 0.02 0.025 0.03 S
tress (
MP
a)
Strain (mm/mm)
Chapter 4: Experimental Testing 51
by a strain-hardening behaviour with the increase in strain. The measured plastic
compressive strength and compressive modulus were 0.5 MPa and 26.0 MPa
respectively..
(a) (b)
Figure 4-4: (a) Stress-strain relationships f or PU in compression, (b) Compression testing for
PU core
4.1.7 Tensile test for PU
To determine tensile properties of the PU core, tensile tests were conducted
according to ISO 1926 [43]. Five prism-shaped test specimens, with 10 × 20 mm
cross section and 150 mm length, were tested in the same test machine, as outlined in
section 4.16, and as shown in Figure 4-5a. The built-in extensometer was used to
measure the strain at a loading rate of 0.5 mm/min. It was possible to adopt this test
method for the 99.8 kg/m3 density PU, since it had sufficient stiffness to facilitate
gripping in the test machine. Tensile failure of all test samples occurred around the
centre of the test sample, as depicted in Figure 4-5a. Tensile stress- strain curves are
presented in Figure 4-5b, and average tensile modulus and tensile strength were
obtained as 19.5 MPa and 0.9 MPa respectively.
0
0.5
1
1.5
2
2.5
3
3.5
0 0.2 0.4 0.6 0.8 S
tress (
MP
a)
Strain (mm/mm)
52 Chapter 4: Experimental Testing
(a) (b)
Figure 4-5: (a) Tensile test for PU core, (b) Stress-strain behaviours of PU in tension
4.1.8 Three- point bending tests for PU core
To investigate the flexural behaviour of PU, three-point bending tests were
conducted as per ASTM D 790 [44], using five 20×10× 160 mm test specimens. An
Instron 5544A mechanical testing machine was used to measure the load-deflection
behaviour of test specimens by applying a central load at a rate of 2 mm/min as
shown in Figure 4-6a. Load-deflection plots from these tests are illustrated in Figure
4-6b. Maximum average load sustained by test specimens during the bending was
19.1 N. The Elastic modulus and flexural strength were calculated as 22.4 MPa and
2.0 MPa respectively, using the method provided in [44].
(a) (b)
Figure 4-6: (a) Three-point bending test for PU core, (b) Load-deflection plots from
bending test of PU core
0
5
10
15
20
25
0 5 10 15 20 25
Load
(N
)
Deflection (mm)
Chapter 4: Experimental Testing 53
4.1.9 Tensile test for steel laminate
Tensile tests were conducted on three specimens of the steel laminate. Each
specimen had a thickness of 1 mm with dimensions as shown Figure 4-7a. They were
prepared and tested according to AS 1391 [46], using an Instron 5569 series
Mechanical Tester as shown in Figure 4-7b. Stress-strain relationships obtained from
the tests are shown in Figure 4-7c, from which modulus of elasticity and yield
strength of steel laminate were determined as 209.9 GPa and 201.0 MPa respectively.
(a)
(b) (c)
Figure 4-7: (a) Dimensions of tensile test specimen, (b) Tensile testing for steel laminate, (c)
Stress-strain relationships for steel laminate in tension
4.2 FABRICATION OF HCFPS TEST SPECIMEN
A manual casting method was used to cast the HCFPS test specimens with a step by
step process. Details of section configuration for the 3200 mm span HCFPS test
panel are explained in section 3.3.1. The beam of the test panel was tapered (Figure
4-8) to facilitate the casting process. To enhance the support bearing capacity of the
test panel, the PU core was replaced with a 100 mm GRC in the vicinity of the
supports, as shown in section 3.3.1, Figure 3-2b. Perforated steel laminate was used
to enhance the bonding between GRC and steel laminate. Cover for the steel
laminate was approximately 6 mm in the test specimens, but this was only for the
0
50
100
150
200
250
300
350
400
450
0 0.05 0.1 0.15 0.2 0.25
Str
ess (
MP
a)
Strain (mm/mm)
54 Chapter 4: Experimental Testing
experimental testing. The cover can be increased in practical applications, if
necessary, depending on the environmental conditions.
Figure 4-8: Section dimensions of HCFPS test panel
Three test specimens were cast using a plywood mould (Figure: 4-9 a). At first, a
6mm thick GRC layer was applied at the bottom of the beam and a 3mm thick
perforated steel plate was placed on top (Figure 4-9b). A second, 6mm thick GRC
layer was applied on top of the steel plate, extending to 10mm thick GRC layers
along the sides of the beam and bottom of the slab (Figure 4-9c). After allowing 2
days of curing time, a central PU core was poured on top of the GRC layers (Figure
4-9d) and allowed to harden for 24 hours. Then the top of PU core was grinded and
levelled off (Figure 4-9e) to obtain the required thickness. Finally, a 10 mm thick
top GRC layer was placed on the hardened PU core (Figure 4-9f). Artificial bonding
agents were not used between GRC and PU, as these two materials achieved a good
bonding during PU hardening and the GRC curing processes (as also evidenced
during the tests). Specimen was left for four week before the testing. Curing time for
the GRC is same as the concrete [69] and PU does not need specific curing time as
cementitious materials. Therefore, curing time of the HCFPS panels will be similar
to the conventional concrete panels. However, this floor system is proposed as a
prefabricated floor system and hence curing time is not significant as in-situ
concrete.
Chapter 4: Experimental Testing 55
(a) Ply wood Mould (b) Perforated steel laminate
(c) Placing bottom GRC layer (d) Pouring PU core
(e) Grinding and leveling top of PU core (f) Placing top GRC layer
Figure 4-9: Casting steps of HCFPS test panel
4.3 SUPPORTING ARRANGEMENT FOR THE TEST PANEL
For all static and dynamic tests the same supporting arrangement was used. A
specially designed arrangement was used to support the "T" shaped ends of test
panels. The HCFPS panel was supported as shown in Figure 4-10. Steel plates,
which were 10 mm thick and 100 mm wide, were placed under the slab and beam at
the supports to distribute the loads uniformly. The steel plates were supported by
solid circular steel bars, as shown in this Figure. Adjustable jacks were used to
support the steel bar under the slab, also seen in Figure 4-10.
56 Chapter 4: Experimental Testing
Figure 4-10: Test panel supporting arrangement
4.4 DATA ACQUISITION SYSTEM
Static and dynamic test were conducted for the HCFPS test panels. Displacements,
acceleration and applied load data were acquired using the Data Acquisition System
(DAS) shown in Figure 4-11.
Figure 4-11: Data acquisition system
Chapter 4: Experimental Testing 57
The DAS consisted of a data recording computer running on Microsoft Windows XP
with a data logger and LabVIEW software was used for data acquisition. The
maximum capacity of the DAS was 50000 samples per second but the maximum
sampling rate required in the experimental tests was 2049. DAS setup provided the
faster and more accurate data for the experimental testing. Method of data acquisition
is discussed under dynamic and static experimental testing separately.
4.5 DYNAMIC TESTING
Two types of dynamic testing were conducted: heel impact and walking tests. Heel
impact tests were the primary type of human induced loads that were used to excite
the HCFPS test panel. Walking tests were also conducted to investigate the vibration
response.
4.5.1 Heel impact test
Heel impact tests can be used to obtain vibration characteristics of the floor panels
[22, 70]. Heel impact tests were performed on all three HCFPS test panels to obtain
the acceleration response. Test set up and data acquisition system are shown in
Figure 4-12 and supporting arrangement was used as explained section 4.3. To apply
the heel impact, an average person (70 kg) was asked to stand at the mid-span of test
panel and to raise their heels approximately 50 mm and produce a sudden impact
(Refer to Figure 4-13a ) [22]. A 5g accelerometer and 25 mm LVDT were used to
acquire acceleration and deflection data at mid-span as shown in Figure 4-13b. Data
was acquired for 3 second period at a rate of 2049 samples per second.
Figure 4-12: Test setup and data acquisition system for heel impact test
58 Chapter 4: Experimental Testing
(a) ( b)
Figure 4-13: (a) Heel impact at the mid-span, (b) 5g Accelerometer and 25 mm LVDT at the
bottom HCFPS panel
Three heel impacts were performed for each panel and similar acceleration and
displacement responses were obtained. Displacement time histories at the mid-span
were also similar for each panel and a typical displacement-time history plot is
shown in Figure 4-14. Typical acceleration responses of the 5g accelerometer at mid-
span for each HCFPS test panel are presented in Figure 4-15, 4-16 and 4-17.
Figure 4-14: Typical heel impact displacement response at mid-span for panel 1
Chapter 4: Experimental Testing 59
Figure 4-15: Typical heel impact acceleration response at mid-span for panel 1
Figure 4-16: Typical heel impact acceleration response at mid-span for panel 2
Figure 4-17: Typical heel impact acceleration response at mid-span for panel 3
60 Chapter 4: Experimental Testing
Equation 4-1 presented by Ellis [35] was used to calculate the damping coefficient
(ς) from the time- acceleration plots. Damping obtained from this equation is ―log
decrement damping‖. Murray [36] stated that modal damping or true damping is one-
half to two-thirds of the value of the log decrement damping. In this context the first
four successive peaks were used to estimate the damping. Table 4-2 shows the
average damping ratios for each panel according to test results. True damping ratio
of HCFPS test panel can be taken as 5%.
n
oe
A
A
nlog
2
1
Equation 4-1
In above equation, A0 and An are the amplitudes at n successive peaks apart where
A0 is the first peak and n=4 of the acceleration-time response plot.
Table 4-2: Damping ratios for the HCFPS test panel
Panel Log decrement
damping Average True damping
Panel 1
Panel 2
Panel 3
Test 1
Test 2
Test 3
Test 1
Test 2
Test 3
Test 1
Test 2
Test 3
10.48 %
10.10 %
10.05 %
10.04 %
10.10 %
9.80 %
10.25 %
10.35 %
10.60 %
10.20 %
9.98 %
10.40 %
5.10 %
4.99 %
5.20 %
The first natural frequency of the test floor panels was be calculated by performing a
Fast Fourier Transformation (FFT) analysis for acceleration plots [22, 37, 38]. Power
Spectrum was obtained from the FFT analysis. The first peak that can be found in the
Power Spectrum is the first natural frequency of the test panel. Typical FFT analysis
for an acceleration response (acceleration response shown in Figure 4-15) is shown
in Figure 4-18. FFT analysis was conducted for the other acceleration responses and
average first natural frequency was obtained as 22.8 Hz as shown in the Table 4-3.
Chapter 4: Experimental Testing 61
Figure 4-18: Typical FFT analysis of an acceleration response
Table 4-3: Experimental First natural frequency of HCFPS panels
Panel First natural frequency (Hz) Average
Panel 1
Panel 2
Panel 3
22.83
22.91
22.66
22.8 Hz
4.5.2 Walking test
Walking tests were conducted for the HCFPS test panels to obtain the acceleration
response at the mid-span. Test set up and supporting arrangement were as described
for the heel impact test (explained in section 4.5.1). An average weight person
walked along the test panel at an average speed. In general, average frequency of
walking is 2 Hz, and therefore, average walking speed is calculated as 1.5 m/s
(stride length is 750 mm) [56]. The time spent to walk along the 3200 mm span test
panel was approximately 2.2 seconds. Hence the walking speed on the test panel was
also approximately 1.5 m/s. A typical acceleration response at the mid-span is shown
in Figure 4-19 and exhibits behaviour of a high frequency floor. In high frequency
floors, applied forces due to walking steps behave like a series of impulses and decay
with time prior to the next step [56].
62 Chapter 4: Experimental Testing
Figure 4-19: Typical acceleration response at the mid-span for walking test
4.6 STATIC LOAD TESTING
Two test panels were used to carry out the static loading test. Tests were carried out
in order to obtain the load-deflection curve to calibrate the finite element model.
Failure modes and ductility of the specimen were investigated during the testing.
4.6.1 Test set up, instrumentation and static load test
Loads were applied as four line loads along the span using 1000 mm steel spreader
beams through an arrangement shown in Figure 4-20 (dimensions and spacing are
given in section 3.3.3). A supporting arrangement was used as explained in section
4.3. Two HCFPS panels were tested in bending with a clear span of 3100 mm. Load
was applied to the panel using a hydraulic pump, which was attached to the loading
frame, as shown in Figure 4-20. A 30 kN load cell was used to measure the load.
This arrangement was adequate to simulate a uniformly distributed load (within the
means of our testing facilities) It enabled the curvature of the panel during loading,
whilst maintaining the uniform loads at the contact locations, similar to the test set up
in [15]. It was hence possible to adequately capture the flexural behaviour of the
HCFPS using this loading arrangement. In order to distribute the loads uniformly, 10
mm rubber pads were placed between the steel spreader beams and the surface of the
Chapter 4: Experimental Testing 63
HCFPS panel. Three Linear-variable-displacement-transducers (LVDTs), with 0.01
mm sensitivity, were placed at centre of span to measure the deflections.
Figure 4-20: Loading arrangement for the static load test of HCFPS panel
4.6.2 Static load test results and discussion
Loading was continued till it was observed that the HCFPS panel was close to failure
because deflection started increase excessively with crack development. Load versus
mid-span deflection curves for the two test specimens obtained from the static
loading tests are presented in Figure 4-21. Cracking of the bottom GRC layer of the
beam started to occur at a load of 12.5 kN. This was considered as the yielding point
of the HCFPS panel.
Load-deflection results from the static load tests, presented in Figure 4-21, show a
smooth transition from elastic to plastic behaviour, but they do not show clearly a
yielding point. However, flexural cracks were observed during the testing, at 12.5 kN
applied loading in both panels. Vertical cracks then developed on either side of the
beam, as shown in Figure 4-22 (a). Loading was continued until mid-span deflection
reached approximately 50 mm, which was the maximum measurable limit of the
LVDT. At this deflection, cracks in the GRC layer, in the beam of HCFPS test panel,
started to widen (Figure 4-22 (b)) because of the plastic deformation of the steel
laminate. However, complete collapse did not occur, even at the 50 mm central
deflection, as the steel laminate continued to deform plastically with the loading.
Hydraulic Jack
30 kN Load Cell
64 Chapter 4: Experimental Testing
Furthermore, there was no failure in the slab and the failure occurred only in the
beam. If, the loading had continued beyond the 50 mm deflection at mid-span, there
could be a complete collapse of the HCFPS panel, due to the ultimate failure of the
steel laminate. However, acceptable ductility for the HCFPS panel can be determined
(as explained in Section 4.6.3), with the load-deflection results up to 50 mm mid-
span deflection. Hybrid integrity remained during the entire test and there was no de-
lamination between layers. There was no shear or support bearing failure in HCFPS
specimens even at 50 mm deflection.
Figure 4-21: Load-deflection behaviour of HCFPS panels
In order to establish the linear range of the HCFPS, its load deflection behaviour was
determined from the analysis of a simply supported beam, using an equivalent
flexural stiffness. This equivalent flexural stiffness (EI) of the HCFPS was calculated
as 5.76×10^11
Nmm2, using the values of the Elastic modulus of the component
materials. Four equal loads (with similar spacing as in the test setup shown in Figures
3-4a and 4-9) were used to obtain the load-deflection plot shown in Figure 4-21. This
behaviour can be considered as the linear load-deflection of HCFPS. Load-deflection
results of static loading test, presented in Figure 4-21, demonstrated that the
deflection increased approximately linearly up to a load of 12.5 kN. After this point
the deflection increased non-linearly. The force at the point on the load-deflection
curve where it noticeably deviated from linearity, was considered as the yielding load
as shown in Figure 4-21. This is further supported by the experimental observation,
where flexural cracks started to form at 12.5 kN applied loading.
0
5
10
15
20
25
0 10 20 30 40 50
Load,
kN
Deflection, mm
Test 2
Test 1
Deflection of a beam with equal EI to HCFPS
Chapter 4: Experimental Testing 65
(a) Vertical cracks in the beam of test panel
(b) Failure only in the beam of HCFPS test panel
Figure 4-22: Cracking and failure due to the flexure
4.6.3 Deflection ductility
Ductility of a structural member can be considered as a measure of its ability to
undergo deformation without a substantial reduction in the flexural capacity [71].
One method of quantifying the ductility is the displacement ductility index, which is
the ratio of ultimate deflection to the deflection at yielding [71] . Ultimate deflection
was considered as 45 mm and deflection at the yielding point 9.6 mm from Figure 4-
21. Deflection ductility index was hence calculated as 4.7 for the HCFPS test panels,
which is acceptable for a structural member.
4.7 CYCLIC LOAD TESTING
Cyclic loading test was conducted for the final test panel. Testing was conducted
according to the method given in Section 3.1.4. Deviation from linearity and
repeatability were calculated to evaluate the cyclic loading performance of the test
panel.
66 Chapter 4: Experimental Testing
4.7.1 Test set up, instrumentation and cyclic load test
Test was conducted using a hydraulic loading system, consisting of a Moog actuator
as shown in Figure 4-23. Supporting arrangement as explained in section 4.3 was
used. Displacement controlled moog actuator was used to control the cyclic loading
but load was applied using the arrangement explained in Section 4.6.1. LVDTs, with
0.01 mm sensitivity, were placed at centre of span to measure the deflections.
Cyclic loading test was conducted according to the method explained in Section
3.3.4. Ultimate load (Lu) for the cyclic loading was considered as the maximum load
carrying capacity of the HCFPS panel before any failure, which as determined by the
static load testing was 12.5 kN (refer to Section 4.6.2). First, six load cycles were
conducted according to load step method and rate of loading shown in Figure 3-5. 50
% of the ultimate load was applied for the first two load cycles. Then, 75% of
ultimate load was applied for the second two cycles and finally 100% of the ultimate
load was then applied for the last two load cycles. Minimum loading of 1.25 kN,
which is 10% of 12.5 kN, was maintained during the unloading cycles. Load-
deflection behaviour for those six cyclic loading steps is presented in Figure 4-24.
Figure 4-23: Cyclic loading test for HCFPS panel
Chapter 4: Experimental Testing 67
Figure 4-24: Cyclic behaviour of HCFPS panel for first 6 loading cycles
4.7.2 Test results and discussion for cyclic loading test
The load-deflection plots for the cyclic load test are shown in Figure 4-24. From
these plots the repeatability and deviation from linearity of the member deflection
(before yielding) were evaluated according to [39], which is explained in Section
3.3.4. Repeatability was 99 %, which was more than the recommended minimum
limit of 95%. Deviation from linearity was 10%, which was less than maximum
recommended limit of 25%. Therefore, HCFPS test panel exhibited acceptable cyclic
loading behaviour.
4.7.3 Comparison of cyclic loading test results with static load test results
Additional loading cycles were carried out with 1 kN increments to investigate cyclic
behaviour of the HCFPS panel beyond the yield point. Two loading and unloading
cycles were conducted for each increment up to 20.5 kN. As before, a minimum of
1.25 kN loading was maintained during the unloading. At the end of the cyclic
loading tests, the applied load was increased to obtain a span deflection of 45 mm in
order to compare the load-deflection behaviour (of the panel) under cyclic loading
with that under static loading as shown in Figure 4-25. The panel subjected to the
cyclic load test exhibited similar load-deflection behaviour as the (other) two panels
subjected to static loading (refer to Figure 4-25). Hence, the cyclic load test results
are considered as reliable.
0
2
4
6
8
10
12
14
0 2 4 6 8 10
Load,
(kN
)
Displacement, (mm)
68 Chapter 4: Experimental Testing
Figure 4-25: Cyclic behaviour of HCFPS panel
4.8 TESTING OF GRC-PU-GRC COMPOSITE PANEL
GRC-PU-GRC composite panel, as shown in Figure 4-25, acts as top slab of the
HCFPS hybrid assembly. Performance of the GRC-PU-GRC composite panel must
to be investigated to determine the structural capacity of the HCFPS slab. Bending
tests of the GRC-PU-GRC composite panel were conducted to investigate structural
behaviour and failure modes.
Figure 4-26: GRC-PU-GRC composite panel slab in HCFPS assembly
4.8.1 Test sample size and fabrication
800 mm × 400 mm panels were selected to test for one-way bending within the
means our testing facilities. Test results were used to validate the FE model and
investigate structural capacity of GRC-PU-GRC panel with 1000 mm one-way and
0
5
10
15
20
25
0 10 20 30 40 50
Load (
kN
)
Deflection (mm)
Cyclic loading
Static test 1
Static test 2
Chapter 4: Experimental Testing 69
500 mm cantilever span used in HCFPS configuration (refer to Figure 4-26). Two
composite panels with different layer thicknesses were used for the testing. The
sectional configurations of these two types of sections are shown in Figure 4-27 and
two numbers of panels were cast for each type. The GRC-PU-GRC composite panel
specimens were cast layer wise. First, the bottom layer of the panel was cast and
allowed one day for curing and the PU layer was poured on top. Then the top layer
was cast on the hardened PU core. Composition of the GRC layer was the same as
explained in section 4.1.1 and the density of the polyurethane was 62 kg/m3.
a) Type A b ) Type B
Figure 4-27: Determined sectional configuration of composite panel
4.8.2 Test set up and instrumentation
The loading tests of the panels were carried out using a Tenious Olsen Hydraulic
testing machine (300kN Capacity). The loading arrangement of the test consisted of a
spreader circular steel roller to achieve an evenly distributed central line load across
the whole panel (refer to Figure 4-28). Two supports, allowing translations along the
span and rotations about the span, were used. Solid, high strength steel circular rods
(20 mm) arrangement were used as supports (refer to Figure 4-29). Supports were
positioned to obtain a span of 700 mm.
To measure the mid-span vertical deflection of each test panel, a 20 mm linear-
LVDT was attached at the centre of the panel. The deflection reading was directly
obtained from a data acquisition computer.
70 Chapter 4: Experimental Testing
Figure 4-28: Loading test setup
Figure 4-29: Supporting conditions for loading test
4.8.3 Static load test results and discussion
Mid-span deflections at 50 N load increments were acquired using a data acquisition
system. This static test was done on each of the 4 panels and load-deflection data
were recorded. Four static tests were performed on each panel and the load-
deflection graphs are shown in Figure 4-30 and Figure 4-31. Failure modes of
composite panel were flexural failure at the mid-span as shown in Figure 4-32.
Linear load-deflection behaviour until the failure was observed for both types of
composite panels. Approximately linear load-deflection was observed for all the
panels. Although GRC-PU-GRC panel exhibited a linear behaviour until the failure,
this panel is only a component of HCFPS. Hybrid behaviour of HCFPS exhibited a
ductile behaviour as seen in the experimental testing and there was no failure in the
GRC-PU-GRC slab until the overall failure of the HCFPS panel (refer to Section
4.6.2). There was no delimitation during the loading test, until the flexural failure as
Chapter 4: Experimental Testing 71
shown in Figure 4-32. Load-deflection data was used to validate FE model of the
GRC-PU-GRC composite panel.
Figure 4-30: Load-deflection curves for A type panels
Figure 4-31: Load-deflection curves for B type panels
Figure 4-32: Failure of one test panel
0
500
1,000
1,500
2,000
2,500
3,000
3,500
4,000
4,500
5,000
0 2 4 6 8
Load,
N
Deflection, mm
0
500
1000
1500
2000
2500
3000
0.0 1.0 2.0 3.0 4.0 5.0
Deflection, mm
Load,
N
72 Chapter 4: Experimental Testing
4.9 SUMMARY
A comprehensive testing program was undertaken to investigate the behaviour of
HCFPS panel and GRC-PU-GRC composite panel. Dynamic tests, static loading
tests and cyclic loading test of 3200 mm span HCFPS panels were conducted.
Experimental studies also included comprehensive material tests for the constituent
materials.
Material properties for the three component materials (GRC, PU and steel) that were
obtained from the material testing program are summarised in Table 4-4.
Table 4-4: Summary of material properties obtained from the material testing
GRC
(MPa)
PU
(MPa)
Steel
(MPa)
Tensile Strength
Tensile modulus
Flexural Strength
Flexural Modulus
Compressive strength
Compressive modulus
3.1
5.0× 103
9.7
4.8 ×103
19.6
5.9 ×103
0.9
19.5
2.0
22.4
0.5
26.0
201.0
209.9×103
-
-
-
-
Heel impact test was conducted for the HCFPS panel and the damping ratio and first
natural frequency were dtermined as 5% and 22.8 Hz respectively. Walking tests
were also conducted on the HCFPS panel and acceleration responses were obtained.
Acceleration response exhibited a series of impulses and the response decays with
time before the next step, indicating the behaviour of a high frequency floor [56].
Therefore, in HCFPS floors, transient response is more significant than the steady
state.
Maximum load carrying capacity of the HCFPS panel before failure was determined
by the static load testing as 12.5kN. The HCFPS exhibited ductile behaviour and
flexural failure in the beam at mid-span. There was tensile failure of the outer GRC
layer and plastic yielding of the steel laminate. HCFPS displayed a deflection
ductility of 4, which is acceptable for floor plates. There were no support bearing or
shear failures during the testing of HCFPS panels. De-lamination between layers did
not occur until failure. Hence, HCFPS shows flexural failure under distributed loads.
Chapter 4: Experimental Testing 73
Cyclic loading tests yielded repeatability of 99% and deviation from the linearity of
10%. These values are within the limits given in [39] , showing acceptable cyclic
loading performance of this floor plate
GRC-PU-GRC composite panel, which is top slab of the HCFPS, exhibited linear
load-deflection behaviour and flexural failure due to the loading.
Results obtained from the experimental testing were used to validate FE models
explained in chapter 5.
74 Chapter 4: Experimental Testing
Chapter 5: Development and Validation of FE models 75
Chapter 5: Development and Validation of FE models
This chapter presents the development of FE models and their subsequent validation
using experimental test results generated in Chapter 4. The commercially available
finite element program ABAQUS 6.9-1 [47] was used with ABAQUS CAE as the
pre- and post-processor for the FE simulations, as described in section 3.4.
5.1 FE MODEL DEVELOPMENT AND VALIDATION USING DYNAMIC
TEST RESULTS OF HCFPS TEST PANELS
FE model for the HCFPS test was developed to conduct free vibration analysis and to
simulate heel impact test. First natural frequency was obtained using free vibration
analysis and validated the FE model with experimentally obtained value. "Modal
Dynamic" analysis procedure available in ABAQUS [47] was used to conduct
transient modal dynamic analysis for heel impact and walking excitation.
Acceleration response of the FE model was obtained from this analysis and
compared with experimental response.
5.1.1 Material properties for dynamic analysis
Dynamic analysis of HCFPS was conducted using linear elastic properties of the
component materials. Materials responds within elastic limits for free vibration and
human induced vibration analysis [56] and hence, properties listed in Table 5-1 were
used in the analysis. These properties were obtained from material test results
presented in Section 4.9.
Table 5-1: Component material properties for the dynamic analysis
Properties PU GRC Steel
Density (kg/m3) 99.8 1983 7800
Elastic Modulus (Mpa) 22.4 5000 210,000
Poisons ratio (ν) 0.3 0.24 0.3
Poisons ratio (ν) for the PU was taken from [72] for the density of 99.8 kg/m3 and ν
of GRC was taken as 0.24 from [73].
76 Chapter 5: Development and Validation of FE models
5.1.2 Model description
HCFPS test panel was modelled as shown in Figure 5-1 using ABAQUS [47]. 10
mm thick and 100 mm wide steel plates were modelled under the HCFPS panel
(refer to Figure 5-1(a)), in order to simulate the experimental conditions at the
supports. Translations in X, Y and Z directions and rotations in Y and Z directions
were restrained at one end whilst translations in X and Y direction and rotations in Z
and Y directions were restrained at the other end (refer to Figure 5-1 (b)) in order to
simulate the restraint from the steel bars. 3 mm thick perforated steel laminate was
used in the test panel with 30% openings. Effective thickness for steel laminate in the
FE model was taken as 2.1 mm in order to account for the plate openings. C3D8R
eight node liner brick elements were used in the FE model for all parts along with
reduced integration and hourglass control [74]. Convergence study was conducted to
select the optimum mesh size for the FE model. Maximum element size was 25 mm,
and the FE mesh is shown in Figure 5-2 (b).
(a) Support
(b) Mesh
Figure 5-1: FE model for dynamic analysis
5.1.3 Free vibration analysis and validation with first natural frequency
Free vibration analysis of the above described model was conducted and natural
frequencies were obtained. First and second mode natural frequencies were obtained
as 23.64 Hz and 27.66 Hz respectively. Mode shape for the first natural frequency is
Chapter 5: Development and Validation of FE models 77
shown in Figure 5-2 (which is the bending mode of HCFPS panel). Table 5-2 shows
that the difference in first mode natural frequency obtained from the FE analysis and
experimentally obtained value is only 0.84 Hz (3%). Therefore, FE model is
adequately validated by the experimental results.
Figure 5-2: Mode shape for the first natural frequency
Table 5-2: Validation of first natural frequency
First natural frequency (Hz)
Experimental - Avegrage 22.80
FE results 23.64
5.1.4 Linear transient dynamic analysis
Response of floor structure can be categorised as transient and steady state responses
as described in section 3.7.2. Transient response is the response of the floor before
the steady state. In high frequency floors (>10 Hz) such as HCFPS, transient
response is more significant than the steady state response because resonance cannot
occur and applied forces behave like a series of impulses [56]. Applied heel impact
load of the HCFPS test panel was simulated in the FE model using transient dynamic
analysis to obtain the acceleration response and this was compared with the
experimental response.
5.1.5 Heel impact load function
Applied heel impact load of the HCFPS test panel was simulated in the FE model.
Murray [75] suggested a triangular load function to simulate the load due to a heel
impact test by an average weight person. Load function is a 2670 N initial load that
linearly decreases to zero over a period of 0.05 seconds, which represents an impulse
of 67 Ns [75]. This method was used successfully by Jetann [70] to validate a FE
model for the acceleration response of heel drop test performed on a post-tensioned
concrete slab. Suggested load function was applied to the FE model of HCFPS test
78 Chapter 5: Development and Validation of FE models
specimen as shown in Figure 5-3. This represents the typical heel impact test of the
panel and corresponding acceleration is shown in Figure 4-16.
Figure 5-3: Heel impact load function
5.1.6 Application of the damping to FE models
Generally, the damping of a structural system is defined as the model damping ratio ζ
(as described in Section 3.7.4). In transient dynamic analysis problems, the damping
matrix cannot be defined as damping ratio, but rather an explicit damping matrix was
defined by Clough et al. [76]. Damping matrix is assumed to be proportional to the
combination of the mass and stiffness matrix and this method is also known as
Reyleigh proportional damping method (combination is shown Equation 5-1). This
method has also been used to incorporate the damping to floor models [22].
KMC Equation 5-1
In above equation, [C] is the system damping matrix, [M] is the mass matrix, [K] is
the stiffness matrix, is the mass proportional damping and is the stiffness
proportional damping.
According to the Reyleigh proportional damping method, relationship between
damping ratio ζ and frequency fn of nth
mode can be obtained in terms of and
as described in equation 5-2 [76].
22
n
n
n
f
f
Equation 5-2
0
500
1000
1500
2000
2500
3000
0 0.5 1 1.5 2 2.5 3
Fo
rce (
N)
Time (Seconds)
Chapter 5: Development and Validation of FE models 79
If the damping ratios and frequencies for two consecutive modes (mth
and nth
) are
known, and can be obtained by solving two simultaneous equation as shown
in equation 5-3.
nn
mm
n
m
ff
ff
1
1
2
1 Equation 5-3
It can be assumed that the variation of damping ratios for first two natural
frequencies is minor such that (ζ1= ζ2= ζ) [22]. Hence, Equation 5-3 can be
rearranged as Equation 5-4 and and can be calculated by substituting ζ and two
natural frequencies f1 and f2.
1
2 21
21
ff
ff
Equation 5-4
Average damping ratio obtained from the experimental results in Section 4.5.1 (5%)
and first two natural frequencies obtained in section 5.1.3 (f1 = 23.64 Hz and f2 =
27.66 Hz) were used to calculate the mass proportional damping ( ) and stiffness
proportional damping ( ) using the equation 5-4. The calculated and are
shown in Table 5-3.
Table 5-3: Mass proportional stiffness proportional damping for FE model
Damping ratio of 5%
1.245
0.002
5.1.7 Dynamic analysis validation with acceleration response of heal impact test
Heel impact load function described above was applied with Relaigh damping to
obtain acceleration response of the FE model. "Modal Dynamic" analysis procedure
available in ABAQUS [47] was used to conduct linear transient modal dynamic
analysis. Acceleration obtained from the FE analysis was compared with
experimental response shown in Figure 5-4. This demonstrated excellent validation,
as peak acceleration values and time duration of the response decay of the FE
acceleration time-history agreed with the experimental response generated in Section
80 Chapter 5: Development and Validation of FE models
4.5.1. Hence, FE models can be used to predict the dynamic response of the HCFPS
panels.
(a) FE acceleration time history
(b) Heel impact acceleration response at mid-span for panel 2
Figure 5-4: Computed and measured acceleration responses due to heal impact
5.1.8 Dynamic analysis and validation with walking loads
Walking of a single person along the HCFPS test panel was simulated using FE
techniques. Pan et al. [77] suggested a method to model the steps of a single person
walking by using point loads. Distance between successive strides was calculated
using the pacing frequency. Hence, footfall interval is 0.5 s for a pacing frequency of
2 Hz and thus gives a stride length of 750 mm [57, 78]. Average point load applied
on a floor due to a single step at an average walking speed (2 Hz) is 616 N [57, 78].
Chapter 5: Development and Validation of FE models 81
This load model was applied to the FE model of HCFPS test panel with damping
data as described in section 5.1.6. The acceleration response was obtained as shown
in Figure 5-5 (a), by conducting transient dynamic analysis. Acceleration response
obtained from FE analysis is similar to the experimental walking response shown in
Figure 5-5 (b) Therefore, FE models can be used simulate the walking activities and
determine vibration response.
(a) Computed acceleration response at mid-span for single person walking
(b) Experimentally measured acceleration response at mid-span for single person walking
Figure 5-5: Computed and measured acceleration responses due to walking
82 Chapter 5: Development and Validation of FE models
5.2 FE MODEL DEVELOPMENT AND VALIDATION USING STATIC
TEST RESULTS OF HCFPS TEST PANEL
FE model for the HCFPS test panel were developed and non-linear analysis were
used to simulate static load tests. Linear and non-linear properties of the materials
were used in the FE analysis using the "Static" analysis procedure available in
ABAQUS [47]. Load-deflection response of the FE model was obtained from the
analysis and compared with experimental results.
5.2.1 Material properties for static analysis
Material properties for the static analysis were based on the material tests that were
conducted on each component material. Following linear and non-linear properties
were used for GRC, PU and steel for the static analysis of FE models.
5.2.2 GRC Material models
GRC layer of the HCFPS test panel is subjected to compressive, flexural and tensile
stresses under bending. Soranakom et. al [79] suggested a material model to define
all such states of GRC (Figure 5-6) and this model was used in this study. Material
model behaviour of GRC was modelled in ABAQUS by modifying the concrete
damage plasticity model, which is defined for similar behaviour of the concrete.
Linear and non-linear tensile behaviour of GRC was modelled similar to the tensile
test results, as they (refer to 0) agree with the material model behaviour. Although
there is a gradual decrease of compressive stress after reaching a maximum stress
according to compression testing of GRC (section 4.1.4, Figure 4-3), it is assumed as
constant after the peak, as per the suggested material model. This did not affect the
overall results of the analysis as the GRC did not exhibit a compressive failure in the
experimental testing (up to mid-span deflection of 50 mm) or in the FE results
explained in section 5.2.6. Therefore, this approximation was considered reasonable
for FE modelling and analysis.
Values for the material model obtained from the GRC material tests are as follows:
compressive yield stress (σcy) = 19.6 MPa, compressive yield strain (εcy) = 0.0040,
ultimate compressive strain (εcu) = 0.03, cracking tensile strength (σcr) = 3.1 MPa,
first cracking tensile strain (εcr) =0.00062, tensile stress at the end of tensile model
(σtu) = 1 MPa, ultimate tensile strain (εtu) = 0.01 and modulus of elasticity (E=) 4.99
GPa). Poisons ratio of GRC is taken as 0.24 from [73].
Chapter 5: Development and Validation of FE models 83
Figure 5-6: GRC material model
5.2.3 4 PU material model
PU core was modelled using linear elastic properties of PU (E=22.4, ν =0.3). Poisons
ratio (ν) for the PU was taken from [72] for the density of 99.8 kg/m3. Non-linear
properties of PU were not necessary, since it attracts low stresses due to the lower
elastic modulus compared to the GRC and steel. This is explained further in Section
5.2.6.
5.2.4 Steel material model
Steel laminate in the HCFPS is subjected to tensile stress as it acts as reinforcement.
Elastic properties (E=209.9 GPa, ν =0.3) and plastic stress and strain values were
used in ABAQUS, as obtained from the tensile tests (refer to section 4.1.9). 3mm
thick perforated steel laminate was used in the test panel with 30% openings.
Effective thickness for steel laminate in the FE model was taken as 2.1 mm in order
to account for the plate openings.
5.2.5 Model description
Considering the symmetry of the test panel along the span, a half model of HCFPS
panel was developed with appropriate boundary conditions as illustrated in Figure 5-
7. At the centre of the beam, translations along the Z and X directions and rotations
about the X, Z and Y axes were restrained. 10 mm thick and 100 mm wide steel
plates were modelled under the HCFPS panel at the supports. Translations were
restrained in the Y direction at the supports as shown in Figure 5-7. This model
simulates the test setup, as steel plates were placed under the HCFPS panel as shown
in Figure 4-20 in section 4.6.1. Load spreader beams were also modelled and the load
was applied as illustrated in Figure 5-7.
84 Chapter 5: Development and Validation of FE models
Figure 5-7: FE model of HCFPS panel for static loading test
C3D8R eight node liner brick elements were used in the FE model for all parts along
with reduced integration and hourglass control [74]. The FE model was meshed as
shown in Figure 5-8. Fine mesh was used in the beam, which exhibited flexural
failure during the testing. Density of the mesh was determined by conducting a
convergence study.
Figure 5-8: FE mesh of HCFPS panel
5.2.6 Static analysis, validation and discussion
Static analysis of the FE model was conducted with the above material models.
Perfect bonding was assumed between each of the materials. This assumption was
supported by the experimental investigation in which no de-lamination was observed
until failure (refer to Sections 4.6.2 and 4.8.3).
Load-deflection behaviour was compared with experimental results to validate the
FE model. FE model exhibited a very good agreement in not only the linear
behaviour but also in the non-linear behaviour captured during the tests, as shown in
Chapter 5: Development and Validation of FE models 85
Figure 5-9. The FE model exhibits a linear behaviour up to an applied load of 14 kN.
From this point onwards, it exhibits a non-linear behaviour. This value of the load
matched reasonably well with the experimental yielding point (refer to Section
4.6.2), with the small difference due to non uniformities of the material layers in the
experimental panel.
Figure 5-9: FE model validation with experimental results
Furthermore, FE results also exhibited a flexural failure in the central span of the
HCFPS panel beam. Damage due to the tensile failure of the GRC in the FE model is
illustrated in Figure 5-10. The damage parameter of GRC has been defined as the
ratio of cracking strain to the total strain in ABAQUS. GRC and steel laminate
follow the non-linear tensile behaviour after yielding as observed in both the FE
analysis and experimental testing. Further, stresses in individual materials in other
parts of the HCFPS panels did not exceed their capacities. This was also observed in
experimental testing as there was no resulting shear or support bearing failure. The
analysis was conducted only until the mid-span deflection reached 45 mm for the
process of validation, as the computation time increased significantly beyond this
point. Since non-linear behaviour could be predicted up to a deflection ductility
index of 4, which is the ratio of 45 mm to 11 mm, FE prediction was considered as
adequate for further analysis. Thus, FE models can be used for predicting the
behaviour of HCFPS.
0
5
10
15
20
25
0 10 20 30 40 50
Load (
kN
)
Deflection (mm)
Test 2
Test 1
FE model behaviour
86 Chapter 5: Development and Validation of FE models
Figure 5-10: Flexural cracks in the beam of HCFPS at the failure
Flexural stress and strain distributions along the cross-section at mid-span of HCFPS
at the yielding point (at the applied load of 14 kN and 11 mm deflection, refer to
Figure 5-9) were obtained from FE model as shown in Figure 5-11. Steel laminate
starts to yield at this point as it reaches to tensile yielding stress of 200 MPa (refer to
Figure 4-7). The yielding point obtained in experimental testing of HCFPS panel (at
the applied load of 12.5 kN and 9.9 mm deflection, refer to Figure 4-21) therefore
matches the FE results reasonably well.
PU core attracts insignificant tensile and compressive stresses due to its lower elastic
modulus according Figure 5-11. As PU has a lower tensile capacity, this hybrid
configuration facilitates the avoidance of tensile stress in the PU core. Similarly,
compressive stress in the compression zone of the slab is mostly attracted to the GRC
layer, (though in Figure 5-11 this is not distinct due to the scale).
Figure 5-11: Stress and strain distribution at the mid-span of the HCFPS
Chapter 5: Development and Validation of FE models 87
At the applied load of 14 kN, compressive stress of the PU core in the slab of
HCFPS, is (Figure 5-11) 0.01 MPa, which is less than the plastic compressive
strength of 0.505 MPa presented in Figure 4-3 (Section 4.1.4). However, PU cannot
be neglected from the FE analysis as PU acts as a core and maintains the integrity of
the sectional configuration of the HCFPS.
Steel laminate acts as reinforcement for the HCFPS by attracting high tensile stress
of 200 MPa (Figure 5-11). The lower most GRC layer at mid-span cracks at the
applied load of 14 kN and hence tensile stress in that layer is zero. The top most
GRC layers in the slab of HCFPS panel attract compressive stress. The compressive
stress then distributes over the area of top slab resulting in lower stress
concentrations in the top most GRC layer of the HCFPS. This can be seen in Figure
5-10 where flexural cracks appear only in the beam of the HCFPS without any
compression failure in the slab. Compressive stress at the top GRC layer is 5.8 MPa
at 14 kN (Figure 5-11) load and the compressive strength of GRC obtained from the
experimental testing is 19.6 MPa. The FE results also showed that the compressive
stress in the top GRC layer did not reach the compressive strength even at 45 mm
mid-span deflection. This behaviour was further supported during the experimental
testing in which compression failure was not evident in the slab of the HCFPS panel.
5.3 FE MODEL DEVELOPMENT AND VALIDATION FOR GRC-PU-GRC
COMPOSITE PANEL
FE models for the GRC-PU-GRC composite test panels were developed and
conducted linear static analysis to simulate static load test explained in Section 4.8.
Linear properties of the materials were used to conduct the FE analysis in ABAQUS
[47]. Load-deflection response FE model was obtained from that analysis and
compared with experimental results
5.3.1 FE model
Layered three-dimensional FE models were developed for the GRC-PU-GRC
composite panel as shown Figure 5-12 (test panel dimensions are given in Section
4.8.1). Translations in X, Y and Z directions and rotations in Y and Z directions
were restrained at one end whilst translations in X and Y direction and rotations in Z
and Y directions were restrained at the other end (refer to Section 4.8.2) in order to
simulate the restrain from the steel bars.
88 Chapter 5: Development and Validation of FE models
C3D8R eight node liner brick elements were used in the FE model for all parts along
with reduced integration and hourglass control [74]. The FE model was meshed as
shown in Figure 5-12. Density of the mesh was determined by conducting a
convergence study. Mesh was of 10 x 10 x 20 mm size providing 3200 elements.
Perfect bond was assumed to occur between GRC and PU layers in FE model as
there was no de-lamination during the loading test until the flexural failure, as
explained in Section 4.8.3.
Figure 5-12: FE model for GRC-PU-GRC panel
5.3.2 Material properties
The material properties of GRC were obtained from the material testing and
properties of 62 kg/m3
density PU were determined using Equation 5-5 [14]. Material
properties are listed in Table 5-4.
Equation 5-5
(Where modulus of solid PU ( ) and density of solid PU are considered to be
1.6 GPa and 1200 kg/m3 respectively).
Table 5-4: Component material properties for the GRC-PU -GRC , FE model.
Properties PU GRC
Density (kg/m3) 62 1983
Elastic Modulus (Mpa) 10.5 5000
Poisons ratio (ν) 0.3 0.24
Poisons ratio (ν) for the PU was taken from [72] for the density of 62 kg/m3 and ν of
GRC was taken as 0.24 from [73].
Chapter 5: Development and Validation of FE models 89
5.3.3 FE model validation
Validation was conducted by comparing the load-deflection curve of the FE model
with experimental load-deflection curves. Mean experimental mid-span deflection
values at various loads were obtained from the panels separately for type A and Type
B. The load-deflection data was then plotted in the graph, as seen in Figure 5-13 and
Figure 5-14. FE models generated results that were in agreement with the mean
load–deflection curves. Thus, the validation produced a satisfactory agreement up to
the yielding point of the panels. The validated layer model was used for further
investigations. Parametric study will be conducted to determine structural capacity of
GRC-PU-GRC composite panel as one-way and cantilever spans, which are
component parts of HCFPS (Presented in Chapter 6).
Figure 5-13: Validation for FE model for static test (Type A panel)
Figure 5-14: Validation for FE model for static test (Type B panel)
0
500
1,000
1,500
2,000
2,500
3,000
3,500
4,000
4,500
5,000
0 1 2 3 4 5 6 7
Panel 2
Panel 1
FE Behaviour
Load,
N
Deflection, mm
0
500
1000
1500
2000
2500
3000
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
Panel 1
Panel 2
FE Behaviour
Deflection, mm
Load,
N
90 Chapter 5: Development and Validation of FE models
5.4 SUMMARY
FE models were developed for the experimentally tested HCFPS panel (3200 mm
span) and GRC-PU-PU composite panel, which is top slab of HCFPS. These models
were validated with previously generated static and dynamic experimental results.
FE models for the HCFPS test panels were developed and validated using first mode
natural frequency and acceleration response. First mode natural frequency obtained
from the FE analysis agreed with the experimentally obtained value with only 0.84
Hz (3%) difference. Acceleration response of FE model obtained from heel impact
excitation and walking load agreed well the experimentally obtained responses.
Hence, FE models can be used to simulate the walking activities and determine the
vibration response of HCFPS panels.
FE models of HCFPS exhibited a very good agreement with experimental load-
deflection in not only the linear behaviour but also in the non-linear behaviour. FE
results also exhibited a flexural failure in the central span of the HCFPS panel beam
as observed in experimental testing. Failure resulted due to the tensile failure of GRC
and yielding of steel laminate in the flexural zone. Stresses in individual materials in
other parts of the HCFPS panels excluding flexural zone, did not exceed their
capacities. This was also observed in experimental testing as there was no resulting
shear or support bearing failure. Nonlinear, FE analysis was conducted to obtain a
deflection ductility index of 4. Thus, FE models can be used to predict the linear and
non-linear behaviour of HCFPS.
Steel laminate acts as reinforcement for the hybrid by attracting high tensile stress
and yielding of HCFPS panel occurs by yielding of steel laminate. PU core attracted
negligible tensile or compressive stress due to lower elastic modulus but the integrity
of HCFPS section provides by acting as infill material.
FE model was developed for GRC-PU-GRC composite test panel and load-
deflection response of FE model was validated with experimental results. This can be
used to investigate the performance of the top slab of the HCFPS.
Validated FE models will be used for further parametric studies to develop HCFPS
floor system in Chapter 6 and 7.
Chapter 6: Static performance of HCFPS 91
Chapter 6: Static performance of HCFPS
This chapter presents the development of HCFPS to satisfy static performance
requirements by combining the beneficial inherent properties of individual
component materials to achieve optimum performance. Details of parametric studies,
which were conducted to investigate the static performance characteristics and
strength capacity of HCFPS, are explained.
6.1 DEVELOPMENT OF HCFPS
HCFPS can be graphically represented in a building floor as shown in Figure 6-1.
The structural supporting system can be either steel or Reinforced Concrete (R/C).
HCFPS can be fixed to the structural frame as shown in Figure 6-2. Static
performance characteristics of HCFPS were investigated considering these
supporting conditions.
Figure 6-1: Graphical representation of building floor using HCFPS panels
Figure 6-2: Proposed supporting methods for HCFPS floor to structural frame
92 Chapter 6: Static performance of HCFPS
6.1.1 Section configuration for static performance studies
The proposed HCFPS panel can be prefabricated as shown in Figure 6-3. Width of
the HCFPS was limited to 2 m to suit prefabrication and transportation requirements.
Span length of the HCFPS was varied by changing the material properties and
sectional configuration. A HCFPS panel with single beam (refer to Figure 6-4) was
used to investigate static performance as the proposed HCFPS panel is symmetrical
(refer to Figure 6-3). PU core was replaced with GRC in the vicinity of the supports
to enhance the support bearing capacity, as shown in Figure 6-3. Performance
characteristics of the GRC-PU-GRC panel (slab section of the HCFPS) and slab joint
were also investigated separately.
Figure 6-3: Proposed HCFPS panel configuration and symmetry
Figure 6-4: HCFPS section for parametric study
Figure 6-5: GRC fill replacing PU core in the vicinity of supports
Axis of symmetry
Chapter 6: Static performance of HCFPS 93
6.2 VARIABLES IN PARAMETRIC STUDIES
Stiffness and deflection limits, and flexural, shear and support bearing capacity were
investigated as static performance criteria to develop HCFPS. Parametric studies
were conducted on HCFPS using validated FE techniques. Sectional configuration,
properties of component materials, span lengths, loading limits and support
conditions were used as variable parameters in this study.
6.2.1 Section configurations
Three types of HCFPS section configurations were used for the parametric study as
shown in Figure 6-6 and Table 6-1.
Figure 6-6: Section parameters for parametric study
Table 6-1: HCFPS section and span parameters
Type of
Section
Span
(m)
b
(mm)
a
(mm)
t
(mm)
h
(mm)
A
B
C
3.0
5.0
7.5
100
150
200
100
150
200
60
80
80
200
350
450
6.2.2 FE model for HCFPS
Validated FE techniques were used to evaluate the static performance characteristics
of HCFPS. Half models of HCFPS panels were developed, considering the symmetry
of the HCFPS panels along the span (refer to Figure 6-4 and 6-5), with appropriate
boundary conditions as shown in Figure 6-7. Translations along the Z and X
directions and rotations about the X, Z and Y axes were restrained at the centre of the
beam. Translations were restrained in the X, Z and Y directions at the supports, as
shown in Figure 6-7, to obtain the supporting conditions demonstrated in Figure 6-2.
94 Chapter 6: Static performance of HCFPS
Figure 6-7: FE model of HCFPS
C3D8R eight node liner brick elements were used in the FE model for all
components of the HCFPS, along with reduced integration and hourglass control
[74]. The FE model was meshed as shown in Figure 6-7. Density of the mesh was
determined by conducting a convergence study.
6.2.3 Material Properties
Properties of steel, GRC and PU, listed in Tables 6-2, 6-3 and 6-4, were used for
parametric studies. Properties of high strength steel were obtained as explained in
Section 2.3.1. Properties of Mild steel were obtained from experimental testing in
Section 4.1.9. Material properties can be enhanced by changing the constituents of
GRC [29] and by increasing the density of PU [25]. Properties of GRC were obtained
as explained in Section 2.3.3 [12] and properties of PU were obtained as per Section
2.3.2 [25]. HCFPS can be developed as a economical floor system by using currently
available and widely used material properties of both GRC and PU. The accepted
Young's modulus of GRC is 10 GPa or higher [12]. Properties of GRC 10 can be
obtained economically either by method of spay or premix [12] (refer to Section
2.3.3). Low density PU core is economical and the lowest density for the HCFPS was
selected as 100 kg/m3 (PU 20). Therefore, higher attention was paid to use GRC 10
and PU 20 to develop HCFPS as economical floor system.
Table 6-2: Properties of Steel
Name Density
Kg/m3
Young's
Modulus
GPa
Shear
Modulus
GPa
Shear
Strength
MPa
Tensile
strength
MPa
Poisson's
ratio
M Steel (Mild Steel)
T Steel (High Strength Steel)
7800
7800
210
210
80
80
105
230
210
550
0.3
0.3
Chapter 6: Static performance of HCFPS 95
Table 6-3: Properties of GRC [12]
Name Density
Kg/m3
Young's
Modulus
GPa
Compressive
Strength
MPa
Shear
Strength
MPa
Tensile
strength
MPa
Poisson's
ratio
GRC 5
GRC 10
GRC 15
GRC 20
1900
1900
1900
1900
5
10
15
20
20
30
40
50
3
4
5
6
3
4
5
6
0.24
0.24
0.24
0.24
Table 6-4: Properties of PU [25]
Name Density
Kg/m3
Young's
Modulus
MPa
Shear
Modulus
MPa
Compressive
Strength
MPa
Shear
Strength
MPa
Tensile
strength
MPa
Poisson's
ratio
PU 20
PU 75
PU 150
PU 360
100
200
300
500
23.4
76.1
151.6
361.2
5.7
21.2
45.9
121.3
0.743
2.415
4.800
11.50
0.556
1.807
3.600
8.580
1.000
-
-
-
0.3
0.3
0.3
0.3
6.2.4 Loading conditions
HCFPS floors were developed for use in residential and office floors. Self-weight of
the HCFPS, using the material properties described in Section 6.2.3, were calculated
as shown in Table 6-5. Average self-weight of the HCFPS can be used as 1 kPa. Two
loading conditions as given in Table 6-6 were used according to AS 1170.1 [48].
Dead and imposed loads were used as combination of (1.25 × Dead load) and (1.5 ×
Imposed Load) according to AS 1170.0 [80]. Performances of HCFPS panels were
investigated under service and ultimate loads.
Performance characteristics of the GRC-PU-GRC composite, which is the top slab
section of HCFPS, will be separately investigated in Sections 6.8. under distributed
and concentrated loads. Concentrated loads were used for residential and office
floors as 1.8 kN and 3.5 kN respectively [48] to investigate the behaviour of GRC
layer and PU core at the location of applied load.
96 Chapter 6: Static performance of HCFPS
Table 6-5: Self-weight of the HCFPS floors
Type of
Section
Type of
GRC
Mass (kPa)
PU 20 PU 75 PU 150 PU 360
A
B
C
GRC 5
GRC 10
GRC 20
GRC 5
GRC 10
GRC 20
GRC 5
GRC10
GRC 20
0.59
0.69
0.78
0.63
0.78
0.89
0.67
0.87
1.01
0.76
1.04
1.25
Table 6-6: Loading cases [48]
Floor Type
Dead Load (kPa)
(self weight + fixed
partition and finishes)
Imposed
Load
(kPa)
Service load
(kPa)
Ultimate
load
(kPa)
Residential 1.0 + 1.0 1.5 3.5 4.75
Offices or work areas 1.0 + 1.0 3 5.0 7.00
6.3 RECTANGULAR BEAM AND TAPERED BEAM
6.3.1 FE modelling
The proposed HCFPS configuration consists of a rectangular beam. This section
discusses why a rectangular beam section was selected by comparing the
performance of HCFPS sections consisting of tapered and rectangular beams. Three
types of sectional configurations were selected to investigate the performance of
rectangular and tapered beam sections. Spans and section parameters (a, b, d and h)
were determined as shown in Figure 6-8 and Table 6-7. ―a‖ was changed in stages
from a rectangular section to tapered section in increments of 25 mm for each type of
section. Material properties were used as given in Table 6-8 (obtained from Tables 6-
2, 6-3 and 6-4). FE models were developed as described in section 6.2.2.
Chapter 6: Static performance of HCFPS 97
Figure 6-8: Proposed HCFPS panel configuration and symmetry
Table 6-7: HCFPS section and span parameters
Type of
Section
Span
(m)
b
(mm)
a
(mm)
t
(mm)
h
(mm)
A
B
C
3.0
5.0
7.5
100
150
200
100-200
150-250
200-300
60
80
80
200
350
450
Table 6-8: Properties of Steel
Properties PU GRC Steel
Name
Density (kg/m3)
E (Mpa)
Poisons ratio
PU 20
100
0.24
0.3
GRC 5
1900
5, 000
0.24
GRC 10
1900
10,000
0.24
M Steel
7800
210,000
0.3
6.3.2 FE analysis results and discussion
Linear static analyses were conducted to investigate the performance characteristics
of HCFPS. Uniformly distributed loads were applied incrementally (0.5 kPa) from 0
to 10.0 kPa in the analysis. Deflection and flexural, shear and support bearing
capacities of each type of section were studied for each section configuration.
Deflection and flexural capacity did not change significantly with dimension ―a‖ the
for the three types of sections.
A minor variation in shear stress distribution across the beam (refer to Figure 6-8)
was observed with changes in ―a‖. GRC and PU transfer shear stresses across the
beam section of HCFPS. PU core attracts lower shear stress due to the lower elastic
modulus and GRC outer shell transfers higher shear stress across the section. A
98 Chapter 6: Static performance of HCFPS
typical shear stress distribution (under applied uniformly distributed loads of 7.0
kPa) in the GRC outer layer (across the beam) is shown in Figure 6-8 (for type B
HCFPS section). Ultimate load under 2 kPa dead load and 3 kPa imposed load is 7.0
kPa (refer to Table 6-1). Shear stresses in the GRC outer shell were obtained for all
HCFPS panels by changing ―a‖ under 7.0 kPa load as shown in Figure 6-9. The
stress concentration in the web of the beam decreased with increase of dimension "a"
(refer to Figures 6-9 and 6-10). Only a component of shear force transfers through
the outer GRC layer in the tapered beam, as the GRC layer makes a small angle with
the vertical direction. In contrast, high shear for transfers to the outer GRC layer of
rectangular beam as the GRC layer is vertical. However, the variation of the stress
concentration due to this scenario is minor and all HCFPS panels with rectangular
beams did not exceed the shear capacity of GRC at the expected design ultimate load
of 7.0 kPa (for office floor). Shear capacity of "GRC 10" is 4 MPa, according to
Table 6-2. GRC 10, GRC, 15 or GRC 20 can be used to obtain shear stress capacity
greater than 4 MPa (refer to Table 6-3) in GRC layer of HCFPS. Therefore, with
those materials, rectangular beam section can be used for HCFPS.
Figure 6-9: Shear stress in GRC outer shell for type B section
Figure 6-10: Variation in stress in the GRC layer (with GRC 10) with change in "a"
Continuous glass fiber mesh (refer to Figure 6-11) can be embedded in the GRC
layers instead of chopped glass fibers to improve shear and flexural capacity.
Continuous glass fiber mesh has been used for Domeshells structure construction
[81] to improve the performance of GRC layers at high stress concentrated zones.
Stress carrying capacity of GRC can be improved up to 10-15 MPa [82] by
embedding a continuous glass fiber mesh. Hence, continuous glass fiber mesh can be
0
0.5
1
1.5
2
2.5
3
3.5
Shear
Str
ess (
MP
a)
100 125 150 175 200 225 250 275 300
3 5 7.5 Span (m)
a (mm)
Chapter 6: Static performance of HCFPS 99
used in high shear stress concentrated zones of the HCFPS beam (refer to Figure 6-9)
to improve the shear bearing capacity and to control possible crack.
Figure 6-11: Continuous glass fiber mesh
The use of rectangular beam sections for HCFPS is economically advantageous as
less material is required. HCFPS sections with tapered beams did not offer
significant improvement in flexural, deflection or shear performance compared to
rectangular beams. Hence a rectangular beam section ("T" shaped) has been selected
for the development of HCFPS. Depth ―h‖, slab thickness ―t‖, width of the beam "b"
and properties of materials will be used for further parametric studies to develop
HCFPS.
6.4 FLEXURAL PERFORMANCE
Flexural performance was the governing design criteria of HCFPS panels as a
flexural failure was observed during experimental testing (refer to sections 4.6.2 and
5.2.6). Flexural performance of A, B and C type HCFPS panels with rectangular
beams (refer to Table 6-1) were studied.
6.4.1 FE modelling
Material properties and applied load were used as variable parameters in the study.
Applied load was increased to obtain linear and non-linear deflection in all HCFPS
panels by using linear and non-linear properties of component materials. FE models
were developed as described in Section 6.2.2.
6.4.2 Properties of GRC
Linear and non-linear material properties of GRC were used in FE analysis using
appropriate material models (refer to Figure 6-12) explained in Section 5.2.2.
Properties for the material model were obtained, as listed in Table 6-9. Parameters
include: compressive yield stress (σcy), compressive yield strain (εcy), ultimate
compressive strain (εcu), cracking tensile strength (σcr), first cracking tensile strain
100 Chapter 6: Static performance of HCFPS
(εcr), tensile stress at the end of tensile model (σtu) and ultimate tensile strain (εtu) and
modulus of elasticity (E). Density and Poisson's ratio were obtained from Table 6-3.
Figure 6-12: GRC material model
Table 6-9: Properties of of GRC
Name E,Ec
GPa
σcy
MPa
σcr
MPa
σtu
MPa εcy εcu εcr εtu
GRC 5
GRC 10
GRC 15
GRC 20
5
10
15
20
20
30
40
50
3
4
5
6
1
1
1
1
0.0040
0.0030
0.0027
0.0025
0.03
0.04
0.04
0.04
0.00060
0.00040
0.00033
0.00030
0.01
0.01
0.01
0.01
6.4.3 Properties of Steel
Two types of steel were used as given in Table 6-2. Steel laminate in the HCFPS is
subjected to tensile stress. Stress-strain relationships for both types of steel are shown
in Figure 6-13. Elastic properties and plastic stress and strain values were used in
ABAQUS.
(a) High strength steel (T steel) [22] (b) Mild steel (M Steel)
Figure 6-13: Stress-strain relationship for High strength and mild steel
0
50
100
150
200
250
300
350
400
450
0 0.05 0.1 0.15 0.2 0.25
Str
ess (
MP
a)
Strain (‰)
Chapter 6: Static performance of HCFPS 101
6.4.4 Properties of PU
The PU core was modelled using linear elastic properties as given in Table 6-4. Non-
linear properties of PU were not modelled as it attracts insignificant stresses due to
the very low elastic modulus compared to GRC and steel. This will be explained
further in Section 6.4.5.
6.4.5 FE abalysis, results and discussion
Uniformly distributed loads were applied increment of 0.5 kPa from 0 to 10.0 kPa.
Non-linear static analyses were conducted to investigate the flexural response.
Load-deflection behaviour
Load-deflection behaviour of A, B and C type sections were obtained as shown in
Figure 6-14. Load-deflection plots with modulus of elasticity of 10 GPa and 15 GPa
lay between the two load-deflection plots (GRC 5 and GRC 20) for each section
shown in Figure 6-14.
Stiffness and deflection of the HCFPS can be controlled by adjusting material
properties and section configuration. Stiffness of the HCFPS sections improved with
an increase in elastic modulus of GRC from 5 GPa to 20 GPa for all types, as seen in
Figure 6-14. Yielding limits of the HCFPS sections improved with high strength steel
laminate in comparison to Mild steel laminate. The stiffness or yielding limit of the
HCFPS sections did not significantly change due to the increase in density and
modulus of elasticity of PU. Hence, density of the PU core was maintained as 100
kg/m3
for all sections.
The deflection performance was evaluated at service load using usual criteria of
Span/250 (under both instantaneous and long term total loads) and Span/360 (under
superimposed post construction loads) given in current design standards. The
deflection assessment was carried out at service load without the ultimate load
factors as shown in Table 6-10. This shows that deflections of HCFPS sections
sunder service loads are well below deflection control limit of span /360 and
Span/250. Hence HCFPS can be used to satisfy deflection performance in floor
plates. Deflections of HCFPS panels can be controlled by changing section
configuration and material properties.
102 Chapter 6: Static performance of HCFPS
Applied ultimate load on office floor (qult, Office= 7 kPa) and residential floor (qult, Resi= 4.75 kPa)
Figure 6-14: Load -deflection behaviour of A, B and C type sections
Table 6-10: Serviceability deflection of the HCFPS floor with PU 20
Type of
Section
Type of
GRC
Residential Office Deflection
control limits
Service
load
(kPa)
Deflection
mm
Service
load
(kPa)
Deflection
mm Span
360
Span
250
A
B
C
GRC 5
GRC 20
GRC 5
GRC 20
GRC 5
GRC 20
3.1
3.1
3.2
3.2
3.3
3.3
2.8
1.3
4.9
2.2
12.6
5.5
4.6
4.6
4.7
4.7
4.8
4.8
4.1
2.0
7.2
3.3
18.3
8.0
8.3
8.3
13.9
13.9
20.8
20.8
12.0
12.0
20.0
20.0
30.0
30.0
0
5
10
15
20
25
0 10 20 30 40
Load (
kP
a)
Deflection (mm)
GRC 5, PU 20, M Steel
GRC 20, PU 20, M steel
Type A Section
Yeilding point
qult,Resi
qult,Office
0
5
10
15
20
25
0 10 20 30 40 50
Load (
kP
a)
Deflection (mm)
GRC 5, PU 20, T Steel
GRC 20, PU 20, T steel
Type A Section
Yeilding point
qult,Office
qult,Resi
0
5
10
15
20
25
30
35
0 10 20 30 40
Load (
kP
a)
Deflection (mm)
GRC 5, PU 20, M Steel GRC 20, PU 20, M steel
Type B Section
Yeilding point
qult,Resi
qult,Office
0
5
10
15
20
25
30
35
0 10 20 30 40 50
Load (
kP
a)
Deflection (mm)
GRC 20, PU 20, T Steel GRC 5, PU 20, T steel
Type B Section
qult,Office
Yeilding point
qult,Resi
0
5
10
15
0 10 20 30 40 50
Load (
kP
a)
Deflection (mm)
GRC 5, PU 20, M Steel
GRC 20, PU 20, M steel
Type C Section
Yeilding point
qult,Resi
qult,Office
0
5
10
15
0 10 20 30 40 50 60
Load (
kP
a)
Deflection (mm)
GRC 5, PU 20, T Steel
GRC 20, PU 20, T steel
Type C Section Yeilding point
qult,Office
qult,Resi
Chapter 6: Static performance of HCFPS 103
Load carrying capacity of HCFPS
The ultimate load limits in an office floor (7.0 kPa) and residential floor (4.75 kPa)
(refer to Table 6-2) are shown in Figure 6-14. Results show that HCFPS floors
remain within linear elastic region up to the ultimate load. At the ultimate conditions,
individual materials did not exceed their capacities (including shear zone). Ultimate
load carrying capacity of HCFPS sections are shown in Figure 6-14 (yielding point).
Therefore, a Factor of Safety (FOS) for the flexural performance of HCFPS can be
defined as ratio of ultimate load carrying capacity/applied ultimate load. Factor of
safety of each section configuration is shown in Table 6-11. FOS was estimated only
for the material combination of MSteel, PU 20, GRC 20 and GRC 5. FOS higher than
these values when TSteel is used because ultimate load bearing capacity is high as
shown in the Figure 6-14. This factor can be used as per the design requirements by
adjusting section configuration and material properties. Therefore, HCFPS can be
used to satisfy flexural performance requirements in floor plates.
Furthermore, superimposed dead load included in the applied ultimate load is 1.0 kPa
but this can be increased, if required and such increases (up to a point) will also
provide a satisfactory behaviour as evident from Figure 6.14. The ultimate load
carrying capacity of the HCFPS can be improved using "T" steel, section
configuration and material properties according to Figure 6-14.
Table 6-11: Factor of Safety (FOS) for flexural performance of HCFPS with PU 20 and MSteel
Type of
Section
Type of
GRC
Residential (qult,Resi= 4.75 kPa) Office (qult,Office= 7 kPa)
Ultimate load
carrying capacity
(kPa)
FOS
Ultimate load
carrying capacity
(kPa)
FOS
A
B
C
GRC 5
GRC 20
GRC 5
GRC 20
GRC 5
GRC 20
8
10
8
10
7.5
7.8
1.7
2.1
1.7
2.1
1.6
1.6
8
10
8
10
7.5
7.8
1.2
1.4
1.2
1.4
1.1
1.1
Moreover, capacity reduction factor , which is the ratio between envisaged failure
load and theoretical failure load, or partial factor of safety ( m ), as defined in design
codes [83] can be incorporated into the designs of HCFPS . However, this is separate
104 Chapter 6: Static performance of HCFPS
area of research and has to be investigated considering the individual material
characteristics of GRC, PU and steel and hybrid behaviour of these in relation to the
environmental degradation.
Ductility
Ductility of a structural member can be considered as a measure of its ability to
undergo deformation without a substantial reduction in flexural capacity [71]. One
method of quantifying the ductility is the displacement ductility index, which is the
ratio of ultimate deflection (at the final failure ) to the deflection at yielding [71].
Another way of defining the ductility of a structural member is in terms of strain
energy as the ratio of energy at ultimate strain to the energy at the maximum elastic
displacement [84]. Deflection ductility index method was used in this research as it
was straightforward and directly applicable for the HCFPS panels.
Yielding point deflections of A, B and C types of HCFPS panels are shown in Figure
6-14. FE analyses were conducted to obtain a reasonable deflection beyond the
yielding point as shown in Figure 6-14. Analyses were terminated with deflections
shown in Figure 6-14, as computational time was increased significantly of FE
models. However, it can be expected that deflection may further extend with the steel
laminate yielding. Analysis termination point deflection can be considered as
ultimate deflection (at the final failure) for HCFPS panels. Using the data deflection
ductility index of 4 or higher value for all HCFPS panels can then be estimated
conservatively. Deflection ductility index of HCFPS is hence acceptable for a
structural floor plate.
Failure modes of HCFPS
FE analysis exhibited a flexural failure in the central span of the beam of the all types
of HCFPS (refer to Figure 6-15). GRC and steel laminate show non-linear tensile
behaviour after yielding. Stresses in individual materials in other parts of the HCFPS
panels excluding flexural zone, did not exceed their capacities. Similar behaviour
was observed in experimental testing (refer to Section 4.6.2) and FE validation of test
results (refer to Section 5.2.6). Non-linear behaviour was obtained for all panels up
to a deflection ductility index of 4 as can be seen in Figure 6-15. Stresses in
individual materials in other components (excluding flexural zone) of the HCFPS
panels did not exceed their capacities even after passing yielding point.
Chapter 6: Static performance of HCFPS 105
Figure 6-15: A typical flexural crack development in the beam of the HCFPS
Stress-strain distribution across mid-span section
Stress and stress-distribution across the HCFPS section at the mid span were used to
evaluate the performance of a hybrid assembly. Stress and strain distribution at mid-
span were obtained as that was the identified critical zone under the bending.
Flexural stress and strain distributions along the cross-section at the mid-span of
HCFPS, at the applied load of 5 kPa were obtained for all HCFPS sections from FE
model as shown in Figure 6-16. A typical stress distribution in FE model is shown in
Figure 6-17. Material types of (GRC 5, PU 20 and M steel) were used for all HCFPS
panels to obtain stress and strain diagrams.
According to flexural stress and strain distributions at the centre of the HCFPS panel,
PU core attracted insignificant tensile or compressive stress due to the comparatively
low elastic modulus as shown in Figure 6-17. PU has low tensile capacity and this
hybrid configuration facilitates the avoidance of tensile stress in the PU core.
Therefore, only linear elastic properties of PU can be used in the FE analysis as its
stresses do not exceed the yielding limits. However, PU acts as a core and maintains
the integrity of the sectional configuration of the HCFPS. Steel laminate acts as
reinforcement for the hybrid by attracting high tensile stress. GRC layers in the slab
of HCFPS panel attract compressive stress (refer to Figure 6-17). FE results
demonstrated that the compressive stress in the top GRC layer did not reach
compressive strength even at 40 mm mid-span deflection for all types of HCFPS
sections. This deflection can be used to obtain deflection ductility of 4. Similar
behaviour was observed during experimental testing in which compression failure
was not evident in the slab of the HCFPS panel (refer to Section 4.6.2).
106 Chapter 6: Static performance of HCFPS
(a) Stress and strain distribution at the mid-span of the A, B and C type HCFPS under 5
kPa
(b) Typical stress and strain distribution at the mid-span of the HCFPS
Figure 6-16: Stress and strain distribution at the mid-span of the HCFPS along X-X
0
50
100
150
200
-10 10 30 50 70 90 110
Section D
epth
(m
m)
Stress (MPa)
Type A Section
0
50
100
150
200
-0.00075 -0.00025 0.00025 0.00075
Section d
epth
(m
m)
Strain
N/A
Type A Section
0
50
100
150
200
250
300
-10 10 30 50 70 90 110
Section D
epth
(m
m)
Stress (MPa)
Type B Section
0
50
100
150
200
250
300
-0.00075 -0.00025 0.00025 0.00075
Section d
epth
(m
m)
Strain
N/A
Type B Section
0
50
100
150
200
250
300
350
400
450
-10 10 30 50 70 90 110 130
Section D
epth
(m
m)
Stress (MPa)
Type C Section
0
50
100
150
200
250
300
350
400
450
-0.001 -0.0005 0 0.0005 0.001
Section d
epth
(m
m)
Strain
N/A
Type C Section
Chapter 6: Static performance of HCFPS 107
Figure 6-17: A typical flexural stress distribution (in GRC and PU) at the mid-span (Type A)
When the HCFPS panel is subjected to bending, compressive stresses occur mainly
in the slab, while tensile stresses occur in the bottom steel laminate. Shear stresses
occur across the web of the beam. As GRC and PU exhibit better performance under
compressive and shear stresses [12, 14], they are profiled and located to attract
compressive and shear stresses in the slab and beam of the HCFPS as shown in
Figure 6-3. The continuous GRC layer along the edges provides an encasement to the
HCFPS. Overall, the integrity of the HCFPS section is maintained by the PU core
as it provides a connection between the GRC layers and buckling support for the thin
GRC layers. However, higher tensile, compressive and shear stresses are attracted to
the steel laminate and GRC as their elastic modulus are significantly higher than that
of the PU. Therefore, beneficial inherent properties of individual component
materials based on their performance capability were combined to achieve optimum
performance of the HCFPS.
Creep and shrinkage of HCFPS
Creep deformation of a structural member is defined as the low deformation
followed by the initial elastic deformation under sustained loads [29]. Experimental
test was not conducted to investigate creep and shrinkage behaviour of the HCFPS.
However, GRC is considered as a material which capable of sustaining loads over
prolong periods [29]. Creep deformation in PU core is minimal as it attracts low
stresses in the HCFPS due to the low elastic modulus. HCFPS section contains steel
laminate reinforcement hence GRC and PU are not directly subjected to creep
deformation. Therefore, creep deformation can be considered as conservatively low.
However, for the design purposes elastic deformation obtained from the analysis was
multiplied by a conservative factor of 1.5 to consider any possible creep deformation
with sustained loading as shown in Table 6-12. This shows that creep and shrinkage
108 Chapter 6: Static performance of HCFPS
deflection factor of 1.5 can be allowed for HCFPS floors and factored deflection can
be still maintained below the deflection control limits. In this context, suitable
material combination needs to be determined so that factored deflection does not
exceed the yielding point of HCFPS (refer to Figure 6-14). Acceptable creep and
shrinkage performance can be achieved for all types of HCFPS with GRC 10, which
is the commonly used type of GRC with modulus of elasticity of 10 GPa. Hence,
GRC 10 is recommended to use in HCFPS construction.
Table 6-12: Factored deflection to account for creep and shrinkage deformation
Type
of
Section
Type of
GRC
Residential Office Deflection
control limits
Service
load deflection
(mm)
Factored
deflection
mm
Service
load deflection
(mm)
Factored
deflection
mm
Span
360
Span
250
A
B
C
GRC 5
GRC 20
GRC 5
GRC 20
GRC 5
GRC 10
GRC 20
2.8
1.3
4.9
2.2
12.6
8.3
5.5
4.2
2.0
7.4
3.3
20
12.5
8.3
4.1
2.0
7.2
3.3
18.3
12.2
8.0
6.2
3.0
10.8
5.0
27.5
18.5
12.0
8.3
8.3
13.9
13.9
20.8
20.8
20.8
12.0
12.0
20.0
20.0
30.0
30.0
30.0
6.5 COMPARISON OF HCFPS WITH STEEL-DECK COMPOSITE
FLOOR SYSTEM USING STIFFNESS AND SELF-WEIGHT
Stiffness and self-weight of HCFPS were compared with an existing conventional
floor system, steel-deck composite floor system with a 3000 mm one way span [85].
Details of testing and material properties of the steel-deck composite floor system are
presented in [22, 85].
Type A (3 m span) HCFPS panel, described in Table 6-1, was used in this study.
Elastic properties of GRC, along with non-elastic properties, were obtained from
Table 6-3 and 6-9. Elastic properties of PU (as described in Section 5.2.3) were
obtained from Table 6-3. Elastic and plastic properties of steel were selected from
Section 6.4.3 and Table 6-2. FE modelling was conducted as explained in Section
6.2.2 and static analyses were conducted by changing material properties.
Chapter 6: Static performance of HCFPS 109
Steel-deck composite floor system panel was tested by applying central line load
with a clear span of 3000 mm between supports. HCFPS panel was also loaded with
a central line load to obtain the load-deflection response. Stiffness of HCFPS can be
increased to achieve a stiffness close to that of steel deck composite floor system by
improving material properties. Stiffness of HCFPS is similar to steel deck composite
floor system by using material types of GRC 20, PU 20 and M steel (refer to Tables
6-2, 6-3 and 6-4) as shown in Figure 6-18.
Figure 6-18: Load-deflection plots of steel deck floor system and HCFPS
Self-weight of HCFPS test panels were compared with that of a steel-deck composite
system. Self-weight of a 3200 mm span, 1000 mm wide HCFPS panel was
approximately 190 kg using material types of GRC 20, PU 20 and M steel (refer to
Tables 6-2, 6-3 and 6-4). For the same size, self-weight of a steel deck composite
slab with 100 mm thick concrete deck was estimated at 793 kg. Therefore, HCFPS
panels are approximately 70% lighter than conventional steel deck composite slabs.
6.6 DETERMINATION OF HCFPS SECTION PROPERTIES USING
ANALYTICAL METHODS
Section properties of HCFPS were determined by using analytical methods and these
properties can be used to predict the flexural behaviour of HCFPS. Analytical
methods were validated using FE results as explained in this section.
Parameters of the HCFPS section are shown in Figure 6-19 and were used to
determine the properties of the section. Transformed section, which has modulus of
elasticity of PU, was determined using modular ratios (ESteel / EPU) and (EGRC / EPU).
Neutral axis depth ( ) and Second moment of area ( ) of HCFPS section can be
0
5
10
15
20
25
0 10 20 30 40 50 60
Load (
kN
)
Deflection (mm)
HCFPS with GRC 20, PU 20 and M steel
Steel Deck Composite Floor
110 Chapter 6: Static performance of HCFPS
determined by using the Equation 6-1 and 6-3. Equivalent flexural stiffness of
HCFPS is defined by Equation 6-4.
Figure 6-19: Parameters used to define the properties of HCFPS section
bb – Width of the beam
db – Depth of the beam
h –Depth HCFPS section
L –Width of the slab
yi -Distance to the centroid (from the bottom) of individual components
bi -Width of individual components
di -Depth of individual components
A - Net area of component material
Ai - Net area of ith component material
-Area of the transformed section
tPU – Thickness of PU in the slab
tGRC, top –Thickness of GRC layer at the top
tGRC –Thickness of GRC layer
tGRC ESS –Thickness GRC layer Either Sides of Steel laminate
tSteel – Thickness of Steel laminate
EPU – Elastic modulus of PU
EGRC – Elastic modulus of GRC
ESteel – Elastic modulus of Steel
n1 – ESteel / EPU
n2 – EGRC / EPU
Note- All the areas (A and Ai) of these equations are calculated as net areas of the
component materials
Equation 6-1
Chapter 6: Static performance of HCFPS 111
Equation 6-2
Equation 6-3
Equation 6-4
6.6.1 Linear elastic deflection of HCFPS
Equivalent flexural stiffness of HCFPS section, , can be used to determine
deflection under loading. Equivalent beam section was assumed using values
of HCFPS to determine the deflection. For example, mid-span deflection due to a
uniformly distributed load can be determined using Equation 6-5. Calculated
deflections using this equation for A, B and C type of sections were compared with
the FE results as shown in Figure 6-20.
Equation 6-5
Figure 6-20: Load-deflection comparison between FE and analytical methods
0
5
10
15
20
25
0 10 20 30 40
Load (
kP
a)
Deflection (mm)
FE GRC 5, PU 20, M Steel
FE GRC 20, PU 20, M steel
Analytical GRC 5, PU 20, M Steel
(a) Type A Section
0
5
10
15
20
25
30
35
0 10 20 30 40
Load (
kP
a)
Deflection (mm)
GRC 5, PU 20, M Steel FE
GRC 20, PU 20, M steel FE
GRC 5, PU 20, M Steel Analytical
GRC 20, PU 20, M Steel Analytical
(b) Type B Section
0
5
10
15
20
25
0 10 20 30 40 50
Load (
kP
a)
Deflection (mm)
GRC 5, PU 20, M Steel FE
GRC 20, PU 20, M steel FE
GRC 5, PU 20, M steel Analytical
GRC 20, PU 20, M steel, Analytical
(c) Type C Section
112 Chapter 6: Static performance of HCFPS
Comparison demonstrates that this analytical method can be used to predict the linear
elastic deflection of HCFPS. Equation 6-5 calculates only the flexural deflection of
HCFPS and it can be used to calculate the deflection of HCFPS similar to FE results.
This implies that shear deflection of the HCFPS is very small due to uniformly
distributed loads and can be excluded from design requirements.
6.6.2 Stresses in individual component materials
Normal bending stress in each component material at the mid-span of the HCFPS
(shown in Figure 6-21) can be determined by using Equations 6-6, 6-7 and 6-8.
Stresses were derived, assuming that plane section at the mid-span of the HCFPS
remains plane. Stresses in component materials of A, B and C type of HCFPS
sections obtained from FE analysis under 5 kPa loading, were compared with stresses
obtained from analytical methods as shown in Table 6-13. Relevant stress
distributions obtained from FE analysis for A, B and C type of HCFPS sections are
shown in Figure 6-16. Comparison shows that analytical equations can used to
predict stresses in component materials with close agreement to FE results. The small
deference between the analytical and FE results may be due to simplifications in the
analytical methods. FE models were supported as face retrains at the supports (refer
to Figure 6-7) but analytical results were obtained by assuming those supports as a
line supports. However, analytical methods can be used to calculate the approximate
stresses in component materials for design purposes.
Figure 6-21: Stresses in component materials at mid span section of HCFPS along X-X
Equation 6-6
Equation 6-7
Chapter 6: Static performance of HCFPS 113
Equation 6-8
Table 6-13: Comparison of stresses in component materials under 5 kPa load
Type
of
Section
Type of materials used σSteel
MPa
σGRC,Bottom
MPa
σGRC,Top
MPa
A
B
C
GRC 5, PU 20 M Steel
GRC 20, PU 20 M Steel
GRC 5, PU 20 M Steel
GRC 20, PU 20 M Steel
GRC 5, PU 20 M Steel
GRC 20, PU 20 M Steel
FE
101.1
60.1
101.2
59.1
120
75
Analytical
109.3
64.9
107.6
66.3
127.7
79.9
FE
2.7
5.1
2.4
6.1
3
5.9
Analytical
2.8
6.5
2.7
6.4
3.2
6.5
FE
1.8
2
3.6
4.1
4.7
5.5
Analytical
1.95
2.5
4.0
4.6
4.8
5.7
6.6.3 Properties of cracked section
Moment capacity of HCFPS sections can be determined as the applied moment at the
yielding point. GRC outer shell in the beam can be cracked (by exceeding stress
tensile capacity of 6.5 MPa for GRC 20 and 4.0 MPa for GRC 5, refer to Table 6-3)
and) as the applied moment is increased. Figure 6-22 shows the damaged zone
(tensile cracks) of the GRC outer layer in the beam of the HCFPS (Type B) at the
centre span. The damage parameter of GRC has been defined as the ratio of cracking
strain to total strain. However, yielding of the HCFPS sections starts with the
yielding of steel laminate. This was observed in experimental testing (refer to Section
4.6.2) and FE validation of test results (refer to Section 5.2.6). Therefore, moment
capacity of HCFPS sections can be calculated by using a cracked section as shown in
Figure 6-23. Tensile capacity is provided only by the steel laminate and compression
capacity is provided by the slab section as shown in Figure 6-22. This can be
confirmed as outer GRC layer cracks with the increase of moment, as shown in
Figure 6-17. Tensile capacity of the PU is insignificant compared to steel and GRC
and can also be excluded from the analysis. Plastic neutral axis depth ( ) and
Second moment of area ( ) of HCFPS section can be determined by using the
Equation 6-9 and 6-11. Equivalent flexural stiffness of HCFPS is
defined from Equation 6-12.
114 Chapter 6: Static performance of HCFPS
Figure 6-22: Tensile crack development in the beam of the HCFPS (Type B)
Figure 6-23: Parameters used to define the properties of cracked HCFPS section
Equation 6-9
Equation 6-10
Equation 6-11
Equation 6-12
Yielding stress of the steel can be used to calculate the moment capacity of HCFPS
section as shown in Equation 6-13. Stress of GRC layer at the top surface (
(refer to Equation 6-15) can be determined from the strain obtained from Equation 6-
14. It was assumed that stress variation at point of yielding is linear across the
section.
Equation 6-13
Equation 6-14
Chapter 6: Static performance of HCFPS 115
Equation 6-15
Figure 6-24: Stress distribution of cracked HCFPS section along X-X
Moment capacity ( ) and were obtained from FE analysis and
compared with results obtained from analytical method as shown in Table 6-10.
Comparison shows that analytical results closely agree with the FE results. The small
difference in the analytical and FE results may be due to simplification of the
analytical methods. FE models were supported as face retrains at the supports (refer
to Figure 6-7) but analytical results obtained by assuming as line supports. GRC and
PU below the slab section of the HCFPS was excluded from the analytical
calculation and that may also have affected this difference. However, this
comparison shows that moment capacity of HCFPS sections can be estimated
approximately using analytical methods.
Table 6-14: Properties of of GRC
Type of
Section Type of materials
kNm
σGRC,Top
MPa
A
B
C
GRC 5, PU 20 M Steel
GRC 20, PU 20 M Steel
GRC 5, PU 20 M Steel
GRC 20, PU 20 M Steel
GRC 5, PU 20 M Steel
GRC 20, PU 20 M Steel
FE
10
10.5
28.5
30
49
49.5
Analytical
8.5
9.6
25
26
44.5
45.5
FE
3.11
5.1
6.4
7.9
6.5
7.8
Analytical
3.01
5.2
6.7
8.1
6.8
8.1
116 Chapter 6: Static performance of HCFPS
6.7 SHEAR PERFORMANCE
Shear capacity of HCFPS section can be estimated by observing the shear stress
distribution in its beam at shear zones. Typical shear stress distribution in the outer
GRC layer is shown in Figure 6-24. Shear resistance across the beam is provided by
GRC layer and PU core. Hence, shear capacity of HCFPS sections can be estimated
as the shear capacity of the PU core and GRC layer in shear flow zone as shown in
Figure 6-25. Although there is a high shear stress concentred zone at the vicinity of
support in the outer GRC layer (refer to Figure 6-25), continuous glass fiber mesh
can be embedded instead of chopped glass fibers to improve shear capacity of GRC
layer as explained in Section 0.
Figure 6-25: Shear stress in GRC outer shell at the vicinity of support
Figure 6-26: Shear zone of in the HCFPS cross-section
Shear capacity of the HCFPS can be estimated using Equation 6-17. In that equation,
τGRC and τPU are shear capacity of GRC and PU and can be obtained from Tables 6-3
and 6-4. AGRC3 and APU2 are the areas of GRC and PU in the shear zone (refer to
Figure 6-26). Table 6-11 shows the shear capacity of each section and shows that
applied shear load is lower than the shear capacity of HCFPS sections.
Equation 6-17
Chapter 6: Static performance of HCFPS 117
Table 6-15: Shear Capacity of HCFPS sections
Type of
Section Type of materials
kN
Applied shear force
with 5.75 kPa load
(refer to table 6-1)
A
B
C
GRC 10, PU 20 M Steel
GRC 20, PU 20 M Steel
GRC 10, PU 20 M Steel
GRC 20, PU 20 M Steel
GRC 10, PU 20 M Steel
GRC 20, PU 20 M Steel
PU
4.75
4.75
15.32
15.32
29.32
29.32
GRC
8.4
14
16.2
27
22.2
37
Total
13.15
18.75
31.52
42.32
51.52
66.32
8.625
8.625
14.375
14.375
21.5
21.5
HCFPS has higher shear capacity and its inherent shear capacity is adequate for
office and residential floor loading. Although, there is high shear concentration in the
GRC layer at the vicinity of support due to its higher elastic and shear modulus
compare to the PU core, shear bearing capacity of GRC can be improved by
embedding continuous glass fiber mesh in high shear stress concentrated zones of
GRC in the HCFPS (refer to Figure 6-25). High shear stress concentrated zone can
be taken as 0.15 × Span length from the support. Hence it is recommended to supply
a continuous glass fiber mesh in high shear stress concentrated zones of the HCFPS
(refer to Figure 6-9) to improve the shear bearing capacity of GRC. Therefore,
HCFPS provides high shear capacity and high safety factor against a shear failure.
This is further supported by experimental observations as there was no shear failure
in HCFPS test specimens (refer to Section 4.6.2)
6.8 PERFORMANCE OF GRC-PU-GRC PANEL AND SLAB JOINT
6.8.1 FE modelling of GRC-PU-GRC panel
1000 mm one-way span and 500 mm cantilever span of GRC-PU-GRC composite
panel is configured in HCFPS hybrid assembly as shown in Figure 6-21.
Performance characteristics of GRC-PU-GRC panel were investigated using
validated FE techniques. Two types of panels were used in the study, as shown in
Figure 6-28 and Table 6-16. These panels were used in types A, B and C HCFPS
panels as discussed in Section 6.2.1. FE models were developed for 1000 mm one-
118 Chapter 6: Static performance of HCFPS
way and 500 mm cantilever spans as shown in Figure 6-29. Fully fixed boundary
conditions were used to represent the continuity of GRC-PU-GRC panel in the
HCFPS configuration. FE modelling was conducted as explained in Section 5.3.1.
Figure 6-27: GRC-PU-GRC slab of HCFPS
Figure 6-28: Layer thickness of GRC-PU-GRC panel
Table 6-16: Sectional configuration of GRC-PU-GRC panels
Panel Name t1 GRC (mm) t2 GRC (mm) t PU (mm) t Overall (mm)
60 mm Panel 15 10 35 60
80 mm Panel 15 10 55 80
Figure 6-29: FE models of GRC-PU-GRC panel
Chapter 6: Static performance of HCFPS 119
6.8.2 FE model for slab joint of the adjacent HCFPS panels
Adjacent HCFPS panels can be connected in a building floor as shown in Figure 6-
30. FE model was developed to investigate the performance of the contact joint as
shown in Figure 6-31. Contact between two panels and fully fixed boundary
conditions at the ends were defined in ABAQUS (refer to Figure 6-31). FE
modelling was conducted as explained in Section 5.3.1.
Figure 6-30: Typical detail of the adjacent slab connection
Figure 6-31: FE models of contact joint
6.8.3 Properties of materials
Linear elastic properties of component materials were used as listed in Table 6-17.
Table 6-17: Properties of GRC and PU
Properties PU GRC
Name
Density (kg/m3)
E (Mpa)
Poisons ratio
PU 20
100
0.24
0.3
GRC 20
1900
20,000
0.24
GRC 5
1900
5,000
0.24
6.8.4 FE analysis results and discussion
Performance characteristics of each model were evaluated under uniformly
distributed and point loads. 0 to 10 kPa distributed loads were applied in 0.5 kPa
120 Chapter 6: Static performance of HCFPS
increments. 0 to 10 kN point loads were applied in 0.5 kN increments at the centre of
the one-way span and on the contact location with the adjacent slab (refer to Figure
6-30).
Concentrated loads were also used for residential and office floors as 1.8 kN and 3.5
kN respectively [48]. Performance characteristics of GRC-PU-GRC composite slab
section was separately investigated. Concentrated loads were used to investigate the
behaviour of GRC layer and PU core at the location of applied load.
GRC-PU-GRC panel
Maximum distributed design ultimate load on office floors and residential floors are
7.0 kPa and 4.75 kPa respectively (refer to Section 6.2.4). Linear- load deflection
behaviour and failure due to flexural cracking of the GRC was observed in
experimental testing for the GRC-PU-GRC panel. Performance of the composite
panel was obtained for loads up to 10 kPa, which is greater than ultimate design load.
Material capacity and serviceability deflection limit of span/360 (refer to Figure 32)
was not exceeded up to a distributed load of 10 kPa applied to 60 mm and 80 mm
panels spanning one way over the 1000 mm fixed support and 500 mm cantilever.
Maximum principle stresses in the GRC and PU core remain in their individual
capacities under uniformly distributed load or point load (refer to Figure 32).
Maximum stresses were obtained to an applied load of 8.5 kPa at 0.85 (7.0/8) factor
of safety for design purposes. Point load of 3.5 kN was applied at the centre of
spanning panel to obtain deflections and stresses. Stresses in individual material and
deflection remained within the capacity under the concentrated load as shown in
Figure 6-32 (b).
Table 6-18 and 6-19 summarise the results of the parametric studies. Parametric
studies demonstrate that the GRC-PU-GRC composite panel can be used as the slab
of the HCFPS with a high factor of safety. All GRC types (GRC 5, GRC 10, GRC,
15 and GRC 20) can be used with PU 20 for GRC-PU-GRC panel construction based
on parametric study results.
Chapter 6: Static performance of HCFPS 121
(a) Deflection and stresses under 10 kPa load for 60 mm panel spanning one-way over 1000 mm
(b) Deflection and stresses with 3.5 kN point load for 60 mm panel spanning one-way over 1000 mm
(c) Deflection and flexural stress with 10 kPa distributed load for for 60 mm panel spanning one-way
over 1000 mm
Figure 6-32: Performance GRC-PU-GRC panel with GRC 5 and PU 20
Table 6-18: Performance of 1000 mm one-way span
Name
Service load deflection
mm
Maximum stress in GRC
a MPa
at 5 kPa at 3.5 kN Span/360 at 8.5 kPa at 3.5 kN
60 mm
GRC 5 and PU 20
GRC 20 and PU 20
80 mm
GRC 5 and PU 20
GRC 20 and PU 20
1.0
0.6
0.4
0.3
3.5
2.6
3.3
2.4
2.7
2.7
2.7
2.7
0.50
0.40
0.45
0.38
1.6
1.3
1.7
1.2
Table 6-19: Performance of 500 mm cantilever span
Name
Deflection
mm
Maximum stress in GRC
MPa
at 5 kPa Span/250 at 8 kPa
60 mm
GRC 5 and PU 20
GRC 20 and PU 20
80 mm
GRC 5 and PU 20
GRC 20 and PU 20
1.5
1.4
1.2
1.1
2.0
2.0
2.0
2.0
1.4
1.1
1.3
1.0
122 Chapter 6: Static performance of HCFPS
Slab joint of the adjacent HCFPS panels
The slab joint exhibited acceptable performance under distributed point loading.
Performance of the composite panel was obtained up to a 10 kPa applied load, which
is greater than ultimate design load. Individual materials remain within their capacity
under 10 kPa load (refer to Figure 27). ). Maximum stresses were obtained to applied
load of 8.5 kPa at 0.85 (7.0/8) factor of safety for design purposes. Table 6-15
summarises the results of the parametric study for slab joint. All GRC types (GRC 5,
GRC 10, GRC, 15 and GRC 20) can be used with PU 20 for GRC-PU-GRC panel
and slab joint of HCFPS panel based on parametric study results with a high safety
factor. Point load of 3.5 kN was applied on the slab joint to obtain deflections and
stresses. Stresses in individual material and deflection remained within the capacity
under the concentrated load as shown in Figure 6-32 (b).
(a) Deflection and stress under 10 kPa load for slab joint (60 mm) panel
(b) Deflection and stress under 2.0 kN point load on slab joint (60 mm) panel
Figure 6-33: Performance of slab joint with GRC 5 and PU 20
Table 6-20: Performance of 1000 mm one-way span with slab joint at the centre
Name
Deflection
mm
Maximum stress in GRC
MPa
at 5 kPa at 3.5 kN Span/360 at 8.5 kPa at 3.5 kN
60 mm
GRC 5 and PU 20
GRC 20 and PU 20
80 mm
GRC 5 and PU 20
GRC 20 and PU 20
1.4
1.3
1.1
1.0
4.5
2.8
2.7
2.5
2.7
2.7
2.7
2.7
1.50
1.30
1.20
1.00
3.01
2.85
2.75
2.51
Chapter 6: Static performance of HCFPS 123
6.9 SUMMARY
Parametric studies were conducted to develop the HCFPS. Sectional configurations,
spans, loading and material properties were used as the parameters in the study.
A rectangular beam was selected for the development of HCFPS. Rectangular beam
sections offered acceptable shear performance. Use of rectangular beam is also
economically advantageous, as less material is required. HCFPS has high shear
capacity and its inherent shear capacity is adequate for office and residential floor
loading.
The static performance of HCFPS is governed by overall stiffness of the HCFPS,
flexural behaviour of GRC and yielding of steel laminate. Stiffness and flexural
performance of the HCFPS can be improved by enhancing the material properties.
Load-deflection behaviour HCFPS floors remain within linear elastic region up to the
ultimate applied load in residential and office floors. At the ultimate applied
conditions, individual materials did not exceed their capacities (including shear
zone). FOS for the flexural performance of HCFPS can be defined as ratio of
ultimate load carrying capacity (at the yielding point)/applied ultimate load. A FOS
of 1.5 or greater value can be obtained for the HCFPS by using appropriate material
combinations. Therefore, HCFPS can be used to satisfy flexural performance
requirements in floor plates.
HCFPS exhibited a flexural failure in the central span of the beam of the HCFPS as
observed in experimental testing. GRC and steel laminate shows non-linear tensile
behaviour after yielding. Non-linear behaviour was obtained for all panels up to a
deflection ductility index of 4 (approximately). Hence ductility index of HCFPS is
acceptable for a structural floor plate.
Elastic deformation obtained from the analysis was multiplied by a conservative
factor of 1.5 to consider any possible creep deformation with sustained loading for
design purposes. Creep and shrinkage deflection factor of 1.5 was allowed for
HCFPS floors and factored deflection can be still maintained below the deflection
control limits.
124 Chapter 6: Static performance of HCFPS
PU core attracted negligible tensile and compressive stress due to lower elastic
modulus as seen from stress distribution results. Steel laminate acts as reinforcement
for the hybrid by attracting high tensile stress. GRC layers in the slab of HCFPS
panel attract compressive stress. The inherent properties of individual component
materials in this floor system have been combined to achieve optimum performance
based on their beneficial strength characteristics.
GRC-PU-GRC composite panel slab joint in the slab of the HCFPS exhibited
excellent structural capacity and stiffness under uniformly distributed and point
loads.
Flexural and shear capacity, linear elastic deflection and component material stresses
can be derived using suggested analytical methods in this chapter.
Accepted properties of GRC (GRC 10, GRC 15 and GRC 20) are recommended to
construct HCFPS along with PU 20 based on the static performance investigation.
Applicability GRC 5 is limited for HCFPS, as it has low elastic modulus and tensile
and shear capacity. Properties of GRC 10 can be obtained economically either by
method of spay or premix [12]. Low density PU 20 (100 kg/m3) core is economical
to use in HCFPS. Therefore, GRC 10 and PU 20 can be used to develop an
economical HCFPS to satisfy static performance.
HCFPS is lightweight floor system approximately 50-70% lighter than the equivalent
conventional composite slabs. This floor system can be used as a viable alternative to
conventional floor system since it meets structural performance requirements and has
many desirable properties. Longer spans can be obtained, if necessary, by changing
the material properties of component materials and the sectional configuration.
Chapter 7: Dynamic performance of HCFPS 125
Chapter 7: Dynamic performance of HCFPS
Chapter 6 demonstrated that HCFPS panels can be used in floor construction to
satisfy the static performance requirements, such as deflection limits, and shear and
flexural strength capacities. This chapter investigates dynamic response of HCFPS
and identifies characteristics that influence acceleration response under human
induced vibration by conducting parametric studies using experimentally validated
FE models. This chapter also shows that lightweight HCFPS can be used in
residential and office buildings by evaluating its vibration performance using
acceptable perceptibility limits provided in current design guidelines and standards.
7.1 STRUCTURAL CONFIGURATION
HCFPS floor can be represented schematically as shown in Figure 7-1. Dynamic
performance of HCFPS floors was evaluated using two approaches. First, dynamic
performance of typical prefabricated HCFPS (refer to Figure 7-2) was evaluated
using the appropriate boundary conditions, which result as a consequence of the
adjacent HCFPS panels and supporting structural frame. Second, HCFPS floor plate
with supporting structural frame, as shown in Figure 7-1, was used for the dynamic
performance assessment. FE modelling and dynamic analysis was conducted for both
approaches using validated FE techniques described in Chapter 5.
Figure 7-1: HCFPS floor plate with steel frame
126 Chapter 7: Dynamic performance of HCFPS
Figure 7-2: Prefabricated HCFPS panel
7.2 DYNAMIC PERFORMANCE OF HCFPS (SINGLE PANEL
APPROACH)
Three section configurations (Type A, B and C sections), which previously satisfied
static performance requirements (in Chapter 6), were selected to investigate the
dynamic performance. Spans and section configuration parameters are shown in
Figure 7-3 and Table 7-1. Dynamic analysis was conducted for 2 m wide
prefabricated HCFPS panels with their total spans.
Figure 7-3: Sectional configuration parameters
Table 7-1: Spans and section dimensions
Section Type Span
(m)
b
(mm)
t
(mm)
h
(mm)
A
B
C
3.0
5.0
7.5
100
150
200
60
80
80
200
350
450
Chapter 7: Dynamic performance of HCFPS 127
7.2.1 FE modelling
A typical FE model of single HCFPS panel used in the dynamic analysis is shown in
Figure 7-4. Boundary conditions of the FE model were selected to simulate
connections to the steel beams and adjacent slabs (refer to Figure 7-1). Translations
along the Z and X directions and rotations about the Z axis were restrained at slab
joint connections to simulate the restrain from the adjacent slabs. Translations along
the X, Y and Z were restrained at end supports to simulate connection with the steel
beams.
Figure 7-4: FE model of HCFPS panel
C3D8R eight node liner brick elements were used in the FE models for all
components of the HCFPS, along with reduced integration and hourglass control
[74]. The FE models were meshed as shown in Figure 7-4. Density of the mesh was
determined by conducting a convergence study.
7.2.2 Material properties
Properties of GRC, PU and steel used in Chapter 6 for the static performance
investigations (for the A, B and C type of HCFPS sections) were used in the dynamic
performance investigations. Materials respond within elastic limits for free vibration
and human induced vibration analyses according to [56]. Hence, dynamic analysis of
FE models of the HCFPS was conducted using linear elastic properties of the
component materials as listed in Tables 7-2, 7-3 and 7-4.
Properties of PU had minimal effect on the structural capacity of HCFPS section
(refer to Section 6.4.5). However, elastic modulus of PU changes with density [14]
and properties of PU, as listed in Table 7-3, were used to investigate their influence
on the dynamic performance of HCFPS.
128 Chapter 7: Dynamic performance of HCFPS
Table 7-2: Properties of GRC [12]
Name Density
Kg/m3
Young's Modulus
GPa
Poisson's ratio
GRC 5
GRC 10
GRC 20
1900
1900
1900
5
10
20
0.24
0.24
0.24
Table 7-3: Properties of PU [25]
Name Density
Kg/m3
Young's Modulus
MPa
Poisson's ratio
PU 20
PU 75
PU 150
PU 360
100
200
300
500
23.4
76.1
151.6
361.2
0.3
0.3
0.3
0.3
Table 7-4: Properties of Steel
Name of material Density
Kg/m3
Young's Modulus
GPa
Poisson's ratio
Steel 7800 210 0.3
7.2.3 Mass of the HCFPS
The dynamic performance of office and residential floors should be investigated
using a floor mass that represents actual service loads [56]. The floor mass should be
equivalent to the summation of the self-weight of the floor plate and superimposed
dead loads due to finishes, ceiling, services and partitions. 10% of the nominal
imposed loads (1.5 kPa for residential and 3.0 kPa for office floors) can also be
added as permanent loads [56]. Superimposed dead loads, which gives a total of 1.0
kPa, are summarised in Table 7-5 [48]. This value may be increased up to 2.0 kPa in
residential buildings due to a higher density of partition walls and heavier floor
finishes. In the present study, a uniform superimposed dead load of 1.0 kPa was used
in the dynamic analysis of HCFPS floors. Table 7-6 shows the self weight of HCFPS
floors and mass of the floor including 1 kPa uniform superimposed dead load.
Chapter 7: Dynamic performance of HCFPS 129
If the superimposed load is higher than 1.0 kPa, the natural frequencies will be
lowere and the damping properties of the HCFPS would be better due to the high
equivalent floor mass (refer to Section 3.7.4). Therefore, dynamic characteristics
were investigated with 1.0 kPa superimposed as this is the average minimum load
expected on HCFPS floors and as dynamic performance will be better with high floor
mass.
Table 7-5: Super imposed permanent dead loads for an office floor
Type of loading Load, kPa
Floor finishes acoucstic insulation + cladding
Suspended ceiling
Suspended services
Lightweight partition, furniture and equipments
Fire protection
0.25
0.10
0.15
0.35
0.15
Table 7-6: Mass of the HCFPS floors
Type of
Section
Type of
GRC
Mass (kg/m2)
PU 20 PU 75 PU 150 PU 360
A
B
C
GRC 5
GRC 10
GRC 20
GRC 5
GRC 10
GRC 20
GRC 10
GRC 20
w1
60.0
71.0
79.5
w2
162.0
172.9
181.4
w1
64.5
79.8
91.5
w2
166.5
181.8
193.4
w1
69.1
88.7
103.5
w2
171.0
190.7
205.5
w1
78.2
106.5
127.5
w2
180.1
208.5
229.4
w1- Self weight w2- Mass with 1kPa load
7.2.4 Free vibration analysis
Parameters as described above were varied in the FE model of A, B and C type
HCFPS sections. Free vibration analysis was conducted to obtain the modal
130 Chapter 7: Dynamic performance of HCFPS
frequencies as shown in Table 7-7. Typical mode shape for the first mode natural
frequency is shown in Figure 7-5, which is the bending mode of HCFPS.
First natural frequency (f1) of HCFPS floors increases with the increase of elastic
modulus GRC (refer to Table 7-7). However, f1 decreases with the increase of PU
density, despite the increase of modulus of elasticity of PU with the density. f1 of
types A, B and C HCFPS floors are greater than 10 Hz (refer to Table 7-7). The
maximum possible fourth harmonic of the walking frequency (2.4 Hz) is lower than
the first mode natural frequency [56] of HCFPS floors. Hence resonant vibration is
unlikely to occur and HCFPS can be categorised as high frequency floors.
Figure 7-5: Typical mode shape for the first mode natural frequency
Table 7-7: Modal frequencies of HCFPS panels
Type of
Section
Type of
GRC
Type of PU
PU 20 PU 75 PU 150 PU 360
A
B
C
GRC 5
GRC 10
GRC 20
GRC 5
GRC 10
GRC 20
GRC 10
GRC 20
f1 (Hz)
30.1
34.6
36.8
18.5
22.4
27.5
13.5
16.9
f2 (Hz)
35.5
40.8
43.4
21.8
26.4
32.4
19.3
24.2
f1 (Hz)
25.0
29.7
36.1
18.0
21.7
26.6
13.0
16.1
f2 (Hz)
32.7
38.3
47.3
23.6
28.4
34.8
21.6
26.6
f1 (Hz)
24.6
29.2
35.5
17.5
21.1
25.8
12.5
15.5
f2 (Hz)
32.0
38.0
46.3
27.7
27.4
33.5
21.8
26.8
f1 (Hz)
24.0
28.3
39.3
16.9
20.1
24.5
11.7
14.3
f2 (Hz)
32.6
38.5
46.6
23.0
27.3
33.3
21.5
26.2
High frequency floors are defined as floors having a first natural frequency of 10 Hz
or more, which is greater than the fourth harmonic of the walking frequency
(maximum of 2.4 Hz) [56]. Dynamic loads due to walking on high frequency floors,
Chapter 7: Dynamic performance of HCFPS 131
act as impulsive forces, which diminish before the next step [56]. This behaviour of
HCFPS was demonstrated in the experimental testing (refer to Section 4.5.2) and FE
modal validation (refer to Section 5.1.8).
7.2.5 Parameters that influence the first mode natural frequency of the HCFPS
The first mode natural frequency of a beam may be determined by a simplified
analytical method (refer to Equation 7-1) suggested in [56] and is used to identify the
parameters that influence the first mode natural frequency of the HCFPS. In this
equation, EI is flexural rigidity of the member, m is the effective mass (for a simply
supported beam 50% of the total mass) and L is the span. First mode natural
frequency of type A HCFPS (3 m span) and HCFPS test panel (3 m span) (refer to
Sections 4.2 and 4.6.2) was calculated using Equation 7-1. Equivalent beam sections
of HCFPS were determined using the flexural rigidity (EI), for the test panel: EI
=6.76×10^11 Nmm2 and for type A HCFPS: EI =1.67×10^12 Nmm
2 (with PU 20
and GRC 10). HCFPS test panels were weighed and average weight was recorded as
202 kg and self-weight of the type A HCFPS section was calculated as 180 kg. These
parameters were substituted in to Equation 7-1 to obtain the first mode natural
frequency as shown in Table 7-8.
312 mL
EIf
Equation 7-1
Table 7-8: Comparison of first mode natural frequency
First natural frequency (Hz)
HCFPS test Panel Type A HCFPS
Only self weight With 1 kPa load
Experimental - Avegrage
FE method
Analytical method
22.80
23.64
24.74
-
40.10
41.15
-
30.1
30.5
The parameters that influence the first mode natural frequency of the HCFPS can be
identified as the flexural rigidity and self-weight. Flexural rigidity of the HCFPS
panel is similar to that of conventional floor systems (refer to Section 0), but the self-
weight of HCFPS panel is lower. The self-weight of a 3.2m span, 1m wide
conventional steel-deck composite floor system is 793kg and first mode natural
132 Chapter 7: Dynamic performance of HCFPS
frequency 13 Hz [22, 85]. In contrast, the self-weight of the same sized HCFPS panel
is approximately 200 kg (approximately 70 % decrement compared to convetional
systems). Thus, the lightweight HCFPS floors with high first natural frequency
categorised as high frequency floors.
7.2.6 Damping
Damping ratio of concrete and steel-deck composite bare floor systems are in the
range of 1.5 to 2 % [22, 63, 64, 85]. Damping ratio of conventional floors with
finishes and partitions is in the range of 4.5% - 6% [36, 65]. There is an approximate
increase of 2 to 4% in damping ratio due to the finishes and partition in floor
structures. True damping level of 5% of HCFPS is considered conservative based on
experimental tests on the bare test panels. HCFPS floors will have superimposed
dead loads higher than 1.0 kPa in practical applications due to partitions, suspended
services and floor finishes. Hence, a damping ratio of 5 % is feasible for HCFPS
floors in service. Vibration responses of HCFPS floors were comparatively studied
using damping ratios of 3% and 5% in the parametric studies.
True damping ratios of (ζ) 5% and 3% were incorporated into the FE models as
explicit damping matrix for the dynamic analysis. Clough et al. [76] defined the
damping in transient dynamic analysis problems using an explicit damping matrix,
by incorporating the damping ratio. Damping matrix is assumed to be a combination
of mass proportional damping (α) and the stiffness proportional damping (β), known
as Reyleigh proportional damping method [76]. This method is explained in detail in
Section 5.1.6 and Equation 7-2 was used to calculate α and β using the first two
natural frequencies f1 and f2 (in Table 7-7) The calculated α and β values for damping
ratio of 3% and 5% for each type of HCFPS floor were incorporated in FE models.
1
2 21
21
ff
ff
Equation 7-2
7.2.7 Mathematical load model for human induced loads
Continuous walking is the worst possible loading scenario that can be used in design
studies [56]. A mathematical load model to simulate continuous human induced load
F(t) [55] is given in Equation 7-3 (refer to Section 3.7.1) and used to excite the FE
models. Two loading cases were considered; one person walking (Q=745 N [56]) and
Chapter 7: Dynamic performance of HCFPS 133
a group of people walking (Q= 0.75 KPa). Three types of walking were studied as
listed in Table 7-9.
kn npn tnfQtF 1 )2sin(1)( Equation 7-3
In the above Eq., F(t) is the dynamic force, Q is the static weight of the participating
person, n is the Fourier coefficient corresponding to nth
harmonic, fp is the pacing
frequency, t is the time and n is the phase angle of the nth
harmonic, n is the integer
designating harmonic of the fundamental and k is the number of harmonics that
characterise the forcing function in the frequency range of interest. Phase shift of 900
for each harmonic was used [55]. The numerical values of the first four Fourier
coefficients used to model the human walking load are listed in Table 7-9.
Table 7-9: Parameters for the load model[55]
Mode of
walking fp (Hz)
Numerical coefficient for 1st four
harmonics
1 2 3 4
Slow Walk
Normal Walk
Fast Walk
1.7
2.0
2.4
0.26
0.37
0.52
0.1
0.1
0.1
0.06
0.06
0.06
0.06
0.06
0.06
7.2.8 FE transient dynamic analysis
Parameters as described above were varied in the FE model of A, B and C type
HCFPS sections. Transient dynamic analysis was performed using continuous
walking loads to obtain the acceleration response. Transient response is more
appropriate than the steady state response in high frequency floors because resonance
cannot occur and applied forces behave like a series of impulses [56].
7.2.9 Results from parametric study and discussion
Acceleration responses due to a single person walking and due to a dynamic
distributed load, which represents a group of people, were observed to be similar.
This kind of behaviour is supported by published findings from previous studies [58].
Walking path was changed to different positions on the slab, however no significant
134 Chapter 7: Dynamic performance of HCFPS
change in the acceleration response was evident. Although the walking path was
changed perpendicular to the span direction of the (beam of the) HCFPS,
acceleration response did not exhibit a significant variation. The highest acceleration
response for continuous walking (induced at different positions) was observed at the
mid span of HCFPS panels, and this was used for subsequent vibration assessments.
RMS acceleration
As described in Section 3.7.3, response of a floor structure is evaluated in terms of
peak acceleration and Root-Mean Square (RMS) acceleration (arms). Peak
acceleration is the highest value of acceleration resulting from an excitation.
However, it does not provide a measure of the duration of the response. In contrast,
RMS acceleration is an average measurement of the acceleration time-history, as
shown in Equation 7-4. Smith et al. [56] stated that sharp peaks of acceleration are
less significant with lower (arms). T is the period under consideration in Equation 7-4
and (T= 1 second) was used as suggested by Smith et al [56]. This covers at least one
complete cycle of acceleration due to walking activities [56]. Acceleration response
was obtained from the FE analysis and arms values were calculated using Equation 7-
4. All three HCFPS panels were high frequency floors and thus arms acceleration
must be frequency-weighted using a factor of 8/f1 (refer to Section 3.7.3) to obtain
the frequency-weighted RMS acceleration (aw,rms) [55, 56]. Variation of aw,rms values
(due to continuous walking) of A, B and C sections with change in parameters are
shown in Figures 7-6, 7-7 and 7-8.
T
rms dttaT
a0
2)(1
Equation 7-4
aw,rms values vary in parallel with change in forcing frequency in type A, B and C
floors. The lowest aw,rms can be seen for slower walking and the highest aw,rms is seen
for fast walking for all types of HCFPS panels (refer to Figures 7-6, 7-7 and 7-8).
aw,rms changed marginally with an increase in properties of PU core at each forcing
frequency and damping level. Decrement of 15 to 20 % in aw,rms resulted from
increasing PU density from 100 kg/m3 to 500 kg/m3. (refer to Figures 7-6, 7-7 and
7-8). aw,rms decreased by approximately 50 % when the damping ratio was changed
from 3% to 5%, as can be seen in Figures 7-6, 7-7 and 7-8. Similarly, aw,rms
Chapter 7: Dynamic performance of HCFPS 135
decreased by approximately 40 % with an increase in modulus of elasticity of GRC
from 10 GPa to 20 GPa.
Figure 7-6: RMS acceleration for section type A, and 3 m span
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
1.7 1.7 2 2 2.4 2.4
aw
,rm
s (
ms
-2)
fp (Hz)
500
300
200
100
5% 3% 5% 3% 5% 3% ζ
fp = Forcing frequency , ζ - Damping ratio
PU Density kg/m3
R = 8
R = 4
5% 3% 5% 3% 5% 3% ζ
fp = Forcing frequency , ζ - Damping ratio
PU Density kg/m3
R = 8
R = 4
(a) GRC 5, (E=5GPa)
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
1.7 1.7 2 2 2.4 2.4
aw
,rm
s (
ms
-2)
fp (Hz)
500
300
200
100
5% 3% 5% 3% 5% 3% ζ
fp = Forcing frequency , ζ - Damping ratio
PU Density kg/m3
R = 8
R = 4
(b) GRC 10, (E=10GPa)
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
1.7 1.7 2 2 2.4 2.4
aw
,rm
s (
ms
-2)
fp (Hz)
500
300
200
100
5% 3% 5% 3% 5% 3% ζ
fp = Forcing frequency , ζ - Damping ratio
PU Density kg/m3
R = 8
R = 4
(c) GRC 20, (E=20GPa)
136 Chapter 7: Dynamic performance of HCFPS
Figure 7-7: RMS acceleration for section type B, and 5 m span
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
1.7 1.7 2 2 2.4 2.4
aw
,rm
s (
ms
-2)
fp (Hz)
500
300
200
100
5% 3% 5% 3% 5% 3% ζ
fp = Forcing frequency , ζ - Damping ratio
PU Density kg/m3
R = 8
R = 4
(a) GRC 5, (E=5GPa)
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
1.7 1.7 2 2 2.4 2.4
aw
,rm
s (
ms
-2)
fp (Hz)
500
300
200
100
5% 3% 5% 3% 5% 3% ζ
fp = Forcing frequency , ζ - Damping ratio
PU Density kg/m3
R = 8
R = 4
(b) GRC 10, (E=10GPa)
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
1.7 1.7 2 2 2.4 2.4
aw
,rm
s (
ms
-2)
fp (Hz)
500
300
200
100
5% 3% 5% 3% 5% 3% ζ
fp = Forcing frequency , ζ - Damping ratio
PU Density kg/m3
R = 8
R = 4
(c) GRC 20, (E=20GPa)
Chapter 7: Dynamic performance of HCFPS 137
Figure 7-8: RMS acceleration for section type C, and 7.5 m span
7.2.10 Vibration assessment of HCFPS
Response factor method (R)
As explained in Section 3.7.3, ISO 10137 [55] and BS 6472 [60] provides base
values of RMS acceleration in relation to a base curve, which is frequency-weighted,
to assess human response to vibration at different frequencies. The base value of
RMS acceleration of 5 × 10-3
m/s2 is suggested for floors with their first natural
frequency between 4 to 8 Hz. Calculated RMS acceleration must be frequency
weighted using a factor of 8/f1 for high frequency floors in order to assess vibration
using the base acceleration level.
However, the base curve is not used directly in practical applications. Instead, the
Response factor (R) method is suggested for vibration assessments. R of a floor is the
0
0.02
0.04
0.06
0.08
0.1
0.12
1.7 1.7 2 2 2.4 2.4
aw
,rm
s (
ms
-2)
fp (Hz)
500
300
200
100
5% 3% 5% 3% 5% 3% ζ
fp = Forcing frequency , ζ - Damping ratio
PU Density kg/m3
R = 8
R = 4
(a) GRC 10, (E=10GPa)
0
0.02
0.04
0.06
0.08
0.1
0.12
1.7 1.7 2 2 2.4 2.4
aw
,rm
s (
ms
-2)
fp (Hz)
500
300
200
100
5% 3% 5% 3% 5% 3% ζ
fp = Forcing frequency , ζ - Damping ratio
PU Density kg/m3
R = 8
R = 4
(a) GRC 20, (E=20GPa)
138 Chapter 7: Dynamic performance of HCFPS
ratio between calculated or measured frequency weighted RMS acceleration (aw,rms)
and the base acceleration value [56]. Acceptable R values are given for residential
floors and office floors as 4 (acceptable aw,rms = 0.02 ms -2
) and 8 (acceptable aw,rms
= 0.04 ms-2
) respectively [56]. These levels of aw,rms are shown in Figures 7-6, 7-7
and 7-8.
Type A (3 m span) HCFPS floor exhibits an aw,rms value less than 0.04 ms-2
with 5%
damping level when comprised with GRC having elastic modulus of 10 GPa (GRC
10) and 20 GPa (GRC 20) at all walking frequencies. Change in PU density does not
offer significant advantage in reducing aw,rms and density of 100 kg/m3 (PU 20) can
be used to obtain an aw,rms value less than 0.04 ms-2
. Both type B (5 m span) and
Type C (7.5 m span) resulted in an aw,rms value less than 0.04 ms-2
with 5% damping
level and GRC with 20 GPa (GRC 20) elastic modulus at all walking frequencies.
Vibration Dose Value (VDV) method
aw,rms values were calculated using the continuous loading model described in
Section 5.3. However, continuous loading and vibration are uncommon and human
induced walking loads are intermittent in residential and office buildings [56]. A
cumulative measure of the vibration response for intermittent activities is more
reliable in residential and office floors and needs to be used for assessing the
vibration performance [56] as explained in Section 3.7.3. ISO 10137 [55] and BS
6472 [60] provide perceptible tolerance levels for intermittent vibrations using the
Vibration Dose Value (VDV) which can be calculated using Equation 7-5, in which
aw is the frequency weighted acceleration and T is the total duration of the vibration
[60].
41
0
4)(T
w dttaVDV Equation 7-5
Ellis [61] suggested an alternate procedure, as given in Equation 7-6 to calculate
VDV values of a walking activity by using the aw,rms ( refer to Section 3.7.3) during
design stage of floor plates and this was used in the present study. In Equation 7-6, na
is the number of times the activity will take place in the exposure period (number of
occurrences) and Ta is the duration of an activity (time taken to walk along the floor).
This method is a reverse calculation procedure to obtain number of concurrences (na)
of human activities required reach to perceptible threshold VDV value. aw,rms can be
Chapter 7: Dynamic performance of HCFPS 139
obtained from the dynamic analysis of the floor plate under human activity. Ta can be
estimated using the floor configuration and walking speed. VDV values are defined
for floor plates for two exposure periods as day and night (refer to Section 3.7.3).
Hence, estimated na can be used as assessing parameter by considering the floor plate
use during the exposure period.
Equation 7-6
Maximum aw,rms acceleration for 5 % damping, under forcing frequencies of 2.4 Hz,
2.0 Hz and 1.7 Hz are shown in Tables 7-10. Those maximum aw,rms values were
obtained for each type of HCFPS panel from Figures 7-6, 7-7 and 7-8 with properties
of materials (PU 20 and GRC 10), required for the static performance (refer to
Section 6.9). aw,rms values were estimated using sinusoidal continuous loading with
one second duration [56] and can be considered as a more accurate representation of
vibration induced by human action (of one intermittent event) to calculate VDV.
Ta values for each span were estimated using a walking speed of 1.1 m/s for fp= 1.7
Hz, 1.5 m/s for fp= 2 Hz, and 2.5 m/s for fp= 2.4 Hz) [56]. Acceptable thresholds of
VDV values are provided in BS 6472 [60] for the "low probability of adverse
comment" as 0.4 m/s1.75
for a 16 hour day and 0.13 m/s1.75
for an 8 hour night for
buildings in service. The number of activities (na) that need to be performed to reach
the acceptable thresholds of VDV were estimated as shown in Table 7-10. A single
walking activity every minute in an office or residential floor during the day is an
unlikely occurrence. Walking activities in residential floors are minimal during the
night. The calculated na values in Table 7-10 are therefore unlikely to occur in
residential and office floors and hence HCFPS floors will not exceed threshold VDV
values. VDV analysis considers the probability of a number of occurrences of
vibration induced activities and provides a reliable estimate of the response based on
the parameter na.
140 Chapter 7: Dynamic performance of HCFPS
Table 7-10: VDV assessment of HCFPS using na
Section type fp (Hz) Ta (s)
a w,rms (ms-2
)
(PU 20
GRC 10
ζ=5%)
VDV (ms-1.75
) na
A (Span 3.0m)
B (Span 5.0m)
C (Span 7.5m)
2.4
2.0
1.7
2.4
2.0
1.7
2.4
2.0
1.7
1.2
2.0
2.7
2.0
3.3
4.5
3.0
5.0
6.8
0.038
0.028
0.022
0.05
0.045
0.033
0.055
0.048
0.042
0.4 (16 h day)
0.13 (8 h night)
0.4 (16 h day)
0.13 (8 h night)
0.4 (16 h day)
0.13 (8 h night)
0.4 (16 h day)
0.13 (8 h night)
0.4 (16 h day)
0.13 (8 h night)
0.4 (16 h day)
0.13 (8 h night)
0.4 (16 h day)
0.13 (8 h night)
0.4 (16 h day)
0.13 (8 h night)
0.4 (16 h day)
0.13 (8 h night)
47,850 (50 per min)
533
97,396 (100 per min)
1086 (1 per min)
189,299(197 per min)
2111 (2 per min)
9,578 (10 per min)
107
8,795 (9 per min)
97
22,435 (23 per min)
250
7, 518 (8 per min)
83
4,510 (5 per min)
50
5,658 (6 per min)
63
Optimum sectional configurations for static performance requirements and material
properties; PU 20 and GRC 10 (refer to Tables 7-2, 7-3 and 7-4) with 5 % damping
were adequate to provide acceptable vibration performance of all types of HCFPS
panels. Hence density of PU core can be maintained at 100 kg/m3 (PU 20) for all
HCFPS panels to obtain acceptable vibration performance. Vibration response of
HCFPS under walking activities does not exceed the perceptible threshold of VDV
that was estimated in accordance with BS 6472 [60]. Satisfying this criterion
provides further confirmation that HCFPS can achieve acceptable vibration response
under human induced loads. Lightweight HCFPS floor systems can therefore be
effectively designed for use in office and residential buildings to provide acceptable
vibration performance.
Dynamic performance assessment of HCFPS panels was conducted using a single
panel with appropriate boundary conditions. Assessment shows that HCFPS
Chapter 7: Dynamic performance of HCFPS 141
designed to meet static performance requirements are also adequate to provide
acceptable dynamic performance. However, this is further supported by conducting a
dynamic assessment of full scale HCFPS floor with supporting structural frame.
7.3 DYNAMIC PERFORMANCE OF HCFPS FLOORS WITH THE
STRUCTURAL FRAME
Dynamic behaviour of HCFPS floor (Type C, 7.5 m span) with a structural frame
was studied and compared with the results obtained from the single HCFPS panel
behaviour in Section 7.2. Structural configuration of the floor model is shown in
Figure 7-9. There are four 8.0 m × 7.5 m bays and C type HCFPS panels were used
in the 7.5 m span direction, as shown in Figure 7-9 (a). 457 UB 52 steel beam was
used as primary beams (B1) and columns (C1).
(a) Isometric view of HCFPS floor model
(b) Plan view of HCFPS floor model
Figure 7-9: Structural configuration of four bay floor model using HCFPS
B1 - Main Beam 475 UB 52
B2 - Secondary Beam 305 UB 42
C1 - Column 475 UB 52
142 Chapter 7: Dynamic performance of HCFPS
7.3.1 FE modelling
FE model of four bay HCFPS floor used in the dynamic analysis is shown in Figure
7-10. Type C HCFPS panels were modelled as described in Section 7.2. HCFPS
panels were connected to the steel frame using pin connection at the joint. It is
recommended to use the height of the columns of the FE model up to their inflection
points [22, 56]. Typically inflection points of columns located at the mid-height of
the columns in multi-story construction [56]. Columns were modelled up to a height
of 1.5 m either side of the floor plate and fixed boundary conditions were used at
opposite ends. The connections were restrained at the beam and column joint node
to avoid rigid body movements of the FE model.
C3D8R eight node liner brick elements were used in the HCFPS panels for all
components of the HCFPS, along with reduced integration and hourglass control
[74]. Primary and secondary beams were modelled using beam element of 3 DOF,
3B2- linear beam elements. Density of the mesh was determined by conducting a
convergence study.
Figure 7-10: FE model of four bay HCFPS floor
7.3.2 Material properties
Properties of GRC, PU and steel used in the dynamic performance investigations
HCFPS floor model is shown in Tables 7-11, 7-12 and 7-13. Dynamic analysis of
FE models of the HCFPS was conducted using linear elastic properties of the
component materials as described in Section 7.2.2. Changes in properties of PU
resulted in minimal effect on the structural capacity of HCFPS section (refer to
Section 6.4.5). Furthermore, acceleration level did not significantly reduce due to the
Chapter 7: Dynamic performance of HCFPS 143
increase in PU density. Hence, PU 20 and PU 75 were used in the parametric studies
on the HCFPS floor model.
Table 7-11: Properties of GRC [12]
Name Density
Kg/m3
Young's Modulus
GPa
Poisson's ratio
GRC 10
GRC 20
1900
1900
10
20
0.24
0.24
Table 7-12: Properties of PU [25]
Name Density
Kg/m3
Young's Modulus
MPa
Poisson's ratio
PU 20
PU 75
100
200
23.4
76.1
0.3
0.3
Table 7-13: Properties of Steel
Name Density
Kg/m3
Young's Modulus
GPa
Poisson's ratio
T Steel 7800 210 0.3
7.3.3 Mass of the HCFPS floor model
Mass of the HCFPS floor model was used as the summation of self-weight and kPa
superimposed dead load (refer to Section 7.2.3) in the dynamic analysis.
7.3.4 Free vibration analysis of HCFPS floor model
Parameters as described above were varied in the HCFPS FE model and free
vibration analysis was conducted to obtain the modal frequencies as shown in Table
7-14. Typical mode shapes for the first four natural frequencies are shown in Figure
7-11.
144 Chapter 7: Dynamic performance of HCFPS
(a) First mode natural frequency 12.47 Hz
(b) Second mode natural frequency 13.17 Hz
(c) Third mode natural frequency 16.10Hz
(d) Fourth mode natural frequency 16.98 Hz
Figure 7-11: First four modes shape of HCFPS floor model using GRC 10 and PU 75
First natural frequency (f1) of HCFPS floor model (refer to Figure 7-11 a) was
greater than 10 Hz (refer to Table 7-14) for all material combinations. The maximum
possible fourth harmonic of the walking frequency (2.4 Hz) is lower than the first
Chapter 7: Dynamic performance of HCFPS 145
mode natural frequency [56] of HCFPS floor model. As consequences resonant
vibration is unlikely.
First natural frequency of Type C HCFPS section (7.5. m span) using GRC 10 and
PU 75 was obtained as 13.0 Hz (refer to Table 7-7). Hence, first natural frequency
obtained from single panel approach closely matches that of obtained with structural
frame (12.47 Hz, refer to Figure 7-11 (a)).
Table 7-14: Modal frequencies of HCFPS model
Type of
GRC
Type of PU
PU 20 PU 75
GRC 10
GRC 20
f1 (Hz)
12.97
15.94
f2 (Hz)
13.70
17.23
f1 (Hz)
12.47
15.33
f2 (Hz)
13.17
16.57
7.3.5 Damping
The true damping ratio (ζ) of 5% was incorporated into the FE models as explicit
damping matrix for the dynamic analysis as described in Section 7.2.6 using first two
natural frequencies in Equation 7-2. Mass proportional damping (α) and the stiffness
proportional damping (β) were calculated as shown in Table 7-15.
Table 7-15: Mass proportional damping (α) and stiffness proportional damping (β) for ζ= 5%
Type of
GRC
Type of PU
PU 20 PU 75
GRC 10
GRC 20
α
0.6662
0.8280
β
0.0037
0.0030
α
0.6405
0.7963
β
0.0039
0.0031
7.3.6 Mathematical load model for human induced loads
Continuous walking is the worst possible loading scenario that can be used in design
studies [56]. A mathematical load model to simulate continuous human induced load
F(t) [55] was used as explained in Equation 7-3 (refer to Section 7.2.7), and used to
excite the FE models.
146 Chapter 7: Dynamic performance of HCFPS
7.3.7 FE transient dynamic analysis
Parameters as described above were varied in the FE model of A, B and C type
HCFPS sections. Transient dynamic analysis was performed using continuous
walking loads to obtain the acceleration response. Human induced walking loads
were applied on one panel of the four bay HCFPS floor model (refer to Figure 7-9)
7.3.8 Results from parametric study and discussion
Acceleration responses due to a single person walking and due to a dynamic
distributed load, which represents a group of people, were observed to be similar as
discussed in Section 7.2.9. Walking path was changed to different positions on the
selected activity bay, however no significant change in the acceleration response was
evident. The highest acceleration responses due to walking loads resulted at the
centre of the activity panel. Non-activity panels show comparatively lower
acceleration level as shown in Figure 7-12. Maximum acceleration values shown in
Figure 7-13 recorded in the floor model were used for vibration assessment. aw,rms
acceleration predicted in Type C section (single panel approach in Section 7.2.9)
closely agrees with the HCFPS floor model results. Maximum aw,rms values (with
GRC 10 and PU 20) for 5 % damping, under forcing frequencies of 2.4 Hz, 2.0 Hz
and 1.7 Hz are presented in Tables 7-16. VDV assessment was conducted as
described in Section 7.2.9.
Figure 7-12: RMS acceleration of HCFPS floor model with section type C (non-activity panel)
0
0.02
0.04
0.06
0.08
0.1
0.12
1.7 2 2.4
aw
,rm
s (
ms
-2)
fp (Hz)
200
100
5% 3% 5% 3% 5% 3% ζ
fp = Forcing frequency , ζ - Damping ratio
PU Density kg/m3
R = 4
(a) GRC 10, (10GPa) non-activity panel
Chapter 7: Dynamic performance of HCFPS 147
Figure 7-13: RMS acceleration of HCFPS floor model with section type C (activity panel)
Table 7-16: Minimum number of activities required generates lower probability adverse
comment of Floor model
Section type fp (Hz) Ta (s)
a w,rms (ms-2
)
(PU 20
GRC 10
ζ=5%)
VDV (ms-1.75
) na
C (Span 7.5m)
2.4
2.0
1.7
3.0
5.0
6.8
0.047
0.043
0.035
0.4 (16 h day)
0.13 (8 h night)
0.4 (16 h day)
0.13 (8 h night)
0.4 (16 h day)
0.13 (8 h night)
8,178 (9 per min)
91
7,004 (7 per min)
78
11,733 (12 per min)
130
0
0.02
0.04
0.06
0.08
0.1
0.12
1.7 2 2.4
aw
,rm
s (
ms
-2)
fp (Hz)
200
100
5% m 5% 5% ζ
fp = Forcing frequency , ζ - Damping ratio
PU Density kg/m3
R = 8
R = 4
(a) GRC 10, (10GPa) In activity panel
0
0.02
0.04
0.06
0.08
0.1
0.12
1.7 2 2.4
aw
,rm
s (
ms
-2)
fp (Hz)
200
100
5% 5% 5% ζ
fp = Forcing frequency , ζ - Damping ratio
PU Density kg/m3
R = 8
R = 4
(a) GRC 20, (20GPa) In activity panel
148 Chapter 7: Dynamic performance of HCFPS
The calculated na values in Table 7-16 are, therefore, unlikely to occur in residential
and office floors and, hence, HCFPS floors will not exceed the threshold VDV
values. VDV analysis considers the probability of a number of occurrences of
vibration induced activities and provides a reliable estimate of the response based on
the parameter na. Moreover, predictions that were made using single panel HCFPS
panels are adequate to determine the dynamic performance of HCFPS.
Assessment shows that HCFPS that are designed to satisfy static performance
requirements are adequate to provide acceptable dynamic performance under human
induced walking loads.
7.4 CONCLUSIONS
Dynamic response of HCFPS floors was investigated by conducting parametric
studies using experimentally validated FE models. Material properties, walking
frequency and damping were used as parameters in the parametric study.
The first mode natural frequency of HCFPS floors is greater than 10 Hz and hence
HCFPS can be categorised as a high frequency floor system. The maximum possible
fourth harmonic of the walking frequency (2.4 Hz) is lower than the first mode
natural frequency [56]. As a consequence resonant vibration is a unlikely to occur.
Flexural rigidity of the HCFPS panel is similar to that of conventional floor system
but the self-weight of HCFPS panel is lower by approximately 50-70%. The first
mode natural frequency of HCFPS floors is therefore higher due to its lighter weight.
Acceleration responses due to a single person walking and due to a dynamic
distributed load, which represents a group of people, were observed to be similar.
aw,rms increased with the increase of walking frequency. aw,rms changed marginally
due to the increase in properties of PU core. aw,rms decreased by approximately 50 %
when the damping ratio was changed from 3% to 5%. aw,rms decreased by
approximately 40 % with an increase in modulus of elasticity of GRC from 10 GPa
to 20 GPa.
Its excellent damping ratio of 5% is considered conservative based on experimental
tests on bare HCFPS test panels. HCFPS floors will have superimposed dead loads
Chapter 7: Dynamic performance of HCFPS 149
higher than 1.0 kPa due to partitions, suspended services and floor finishes. Hence, a
damping ratio of 5 % is feasible for HCFPS floors in service.
Dynamic analysis of HCFPS panels was conducted using a single panel with
appropriate boundary conditions. This is further supported by the dynamic analysis
of full scale HCFPS floor plate with supporting structural frame exhibiting similar
results of first natural frequency and aw,rms to the single panel approach. Therefore,
dynamic analysis with single panel can be used as efficient method to evaluate
dynamic performance of HCFPS.
Optimum sectional configuration and material properties (GRC 10, GRC 20 and PU
20) to satisfy static performance requirements were adequate to provide acceptable
vibration performance. Vibration response of HCFPS under walking activities does
not exceed the perceptible threshold of VDV that was estimated in accordance with
BS 6472 [60]. Satisfying this criterion provides evidence that HCFPS can achieve
acceptable vibration response under human induced loads.
HCFPS panels design to satisfy static performance using GRC 10 and PU 20 will
offer acceptable vibration performance under human induced walking loads.
Properties of GRC 10 can be obtained economically either by method of spay or
premix [12]. Low density PU 20 (100 kg/m3) core is economical to use in HCFPS.
Therefore, GRC 10 and PU 20 can be used to develop an economical HCFPS to
satisfy static and dynamic performance. Design guidelines for the HCFPS are
presented presents in Chapter 8 based on static and dynamic performance
characteristics.
150 Chapter 7: Dynamic performance of HCFPS
Chapter 8: Conclusions and recommendations 151
Chapter 8: Conclusions and recommendations
8.1 CONTRIBUTION FROM THIS RESEARCH
This research has developed a novel lightweight Hybrid Composite Floor Plate
System (HCFPS) using GRC, PU and steel laminate as component materials. The
research determined the performance characteristics of HCFPS using Finite Element
(FE) simulations supported by experimental testing. Parametric studies were
conducted to develop the HCFPS in order to satisfy static performance requirements
using sectional configurations, spans, loading and material properties as the
parameters. Dynamic response of HCFPS was investigated by conducting parametric
studies using material properties, walking frequency and damping as the parameters.
HCFPS offers acceptable vibration performance under human induced walking loads
for use in office and residential buildings. Lightweight HCFPS floor systems can
therefore be effectively designed for use in office and residential buildings to provide
acceptable performance characteristics.
The main findings of this study are listed below:
Optimum hybrid sectional configuration using component materials was determined.
Design of the HCFPS is governed by overall stiffness, flexural behaviour of the
HCFPS and yielding of steel laminate.
Stiffness and flexural performance of the HCFPS can be improved by enhancing the
material properties and changing sectional configuration.
HCFPS displayed a deflection ductility of 4, which is acceptable for floor plates.
Flexural and shear capacity, linear elastic deflection and component material stresses
can be derived using suggested analytical calculation procedure. Hence, design of the
HCFPS can be conducted using these methods.
HCFPS is a lightweight floor system approximately 50% lighter than equivalent
conventional concrete composite floor slabs.
152 Chapter 8: Conclusions and recommendations
The first mode natural frequency of HCFPS floor system is greater than 10 Hz and
hence HCFPS can be categorised as a high frequency floor system. Therefore,
resonant vibration due to human induced loads is unlikely to occur.
HCFPS offers excellent damping ratio of 5% (bare floor) due to the central PU core.
As a consequence, acceleration levels due to human induced loads is reduced.
Vibration response of HCFPS under walking activities does not exceed the
perceptible threshold of VDV that was estimated in accordance with BS 6472 [60].
Dynamic assessment of HCFPS can be conducted either by modelling a prefabricated
HCFPS panel with appropriate boundary conditions or by modelling HCFPS floor
plate with the structural frames as both methods give similar VDV assessment
results.
Dynamic assessment shows that HCFPS designed to satisfy static performance
requirements with (GRC 10 and PU 20) are adequate to provide acceptable dynamic
performance. Therefore, design guidelines of the HCFPS were developed using the
static performance requirements.
8.2 DISCUSSION AND SUMMARY
This research developed HCFPS according to the following stages. The first stage
involved investigating static, dynamic, cyclic loading and post yield behaviours,
mode of failure and ductility of the HCFPS panels. Component material testing was
also conducted to obtain their properties. FE models were developed and validated
using results from experimental testing. In the second stage, parametric studies were
performed and HCFPS were developed to satisfy static performance requirements.
The vibration characteristic of HCFPS under human-induced loads was investigated
in the next stage. Finally, guidelines were developed for the design of HCFPS panels.
8.2.1 Experimental Testing and FE model validation
A comprehensive testing program was undertaken to investigate the behaviour of
HCFPS panel and GRC-PU-GRC composite panel. Dynamic tests, static loading
tests and cyclic loading tests of 3200 mm span HCFPS panels were conducted.
Experimental studies also included a comprehensive material testing program to
obtain the properties of constituent materials.
Chapter 8: Conclusions and recommendations 153
Heel impact test was conducted for the HCFPS panel and the damping ratio and first
natural frequency were estimated as 5% and 22.8 Hz respectively. Walking tests
were also conducted on the HCFPS panel and acceleration responses were obtained.
Acceleration response exhibited a series of impulses and the response decays with
time before the next step, indicating the behaviour of a high frequency floor [56].
HCFPS exhibited ductile behaviour and flexural failure in the beam at mid-span after
the initial yielding. There was tensile failure of the outer GRC layer and plastic
yielding of the steel laminate. HCFPS displayed a deflection ductility of 4, which is
acceptable for floor plates. There were no support bearing or shear failures during
testing of the HCFPS panels. Furthermore, de-lamination between layers did not
occur until failure. Experimental testing exhibited flexural failure due to the loading.
Cyclic loading test for one HCFPS test panel was conducted according to the
method given in ACI [39]. This method can be used to evaluate the cyclic loading
performance of slabs comprised of new materials. Cyclic loading tests yielded
repeatability of 99% and deviation from linearity by 10%. These values are within
the limits given in [39] , showing acceptable cyclic loading performance of this floor
plate.
GRC-PU-GRC composite panel acts as the top slab of the HCFPS hybrid assembly.
Performance of the GRC-PU-GRC composite panel was experimentally tested to
investigate the behaviour under applied loads. GRC-PU-GRC composite panel
exhibited linear load-deflection behaviour and flexural failure due to the loading.
FE models for the HCFPS test panels were developed and validated using first mode
natural frequency and acceleration response. First mode natural frequency obtained
from the FE analysis agreed with the previously obtained experimental value with
only 0.84 Hz (3%) difference. Acceleration response of FE model obtained from heel
impact excitation walking load agreed with the experimentally obtained response.
Hence, FE models can be used to simulate walking activities and thus determine the
vibration response.
FE model of HCFPS exhibited a very good agreement with experimental load-
deflection in not only the linear behaviour, but also in non-linear behaviour. FE
results also exhibited a flexural failure in the central span of the HCFPS panel beam
154 Chapter 8: Conclusions and recommendations
as observed in experimental testing. FE model was developed for GRC-PU-GRC test
panel and load-deflection response of FE model was validated with experimental
results. FE model can therefore be used to investigate static performance of HCFPS.
8.2.2 Static design of HCFPS
Parametric studies were conducted to develop the HCFPS using validated FE
techniques. Sectional configurations, spans, loading and material properties were
used as parameters in the study.
Optimum configuration was determined for the HCFPS as shown in Figure 8-1.
Beneficial inherent properties of individual component materials were combined to
achieve optimum performance in the HCFPS.
Figure 8-1: Optimum configuration of HCFPS
Compressive stresses occur mainly in the slab and tensile stresses occur in the
bottom steel laminate under bending according to flexural stress and strain
distributions at the centre of the HCFPS panel. High tensile stresses are attracted to
steel laminate and compressive stresses are attracted to the top GRC, because their
elastic moduli are significantly higher than the PU. Compressive and tensile stresses
of the PU core are negligible due to the low elastic modulus according to flexural
stress and strain distributions at the centre of the HCFPS panel. Hence, a PU core
provides minimal influence on the structural capacity of the HCFPS. However,
overall integrity of HCFPS section is obtained by the PU core, as it provides a
Chapter 8: Conclusions and recommendations 155
connection between GRC layers. Furthermore PU core offers lateral support for thin
GRC layers to avoid lateral buckling. Core material is therefore important for
practical applications but alternative materials may be used instead of the PU core.
Shear stresses occur across the web of the beam due to loading. PU core was
replaced with GRC fill in the vicinity of the supports as shown in Figure 8-1 to
enhance the support bearing capacity. GRC and PU exhibit better performance under
compressive and shear stresses [12, 14]. Hence, GRC and PU are profiled and
located to attract compressive and shear stresses in the slab and beams of HCFPS (as
shown in the Figure 1-2). "T" beam sections (with rectangular beams) offered
acceptable shear performance. Use of rectangular beam is also economically
advantageous as less material is required. Shear capacity of the HCFPS sections are
higher than the applied shear force and may not significantly influence on the overall
design.
GRC-PU-GRC composite slab of the HCFPS and slab joint, which is used to connect
adjacent HCFPS panels, exhibited excellent structural capacity and stiffness under
uniformly distributed and point loads. GRC-PU-GRC panels with overall thickness
of 60 mm and 80 mm are suggested to be used in HCFPS depending on the loading
requirements and overall design of the HCFPS.
Design of this new floor system is governed by overall stiffness, flexural behaviour
of the HCFPS and yielding of steel laminate. Stiffness and flexural performance of
the HCFPS can be improved by enhancing the material properties and, hence, similar
stiffness to conventional floor systems can also be obtained. Deflection ductility
index of 4 can be obtained for HCFPS panels, which is an acceptable level of
ductility as a floor slab. Creep and shrinkage deformation of HCFPS has not been
experimentally tested but it could be minimal due to presence of steel laminate and
composite action of HCFPS. However, it is suggested to multiply the linear elastic
deflection of HCFPS under service loads using a conservative factor of 1.5 to
account any possible creep deformation under the sustained loads or shrinkage
deformation. Deformation of the HCFPS under the service loads are well below the
span/360 and span/250 limits suggested in design codes, and hence the factored
service loads deflection dose not exceed such limits.
156 Chapter 8: Conclusions and recommendations
Accepted and currently avaiable properties of GRC (GRC 10, GRC 15 and GRC 20)
are recommended to construct HCFPS along with PU 20 based on the static
performance investigation.
Flexural and shear capacity, linear elastic deflection and component material stresses
can be derived using the analytical methods summarised in Section 8.2.4 and the
accuracy of these methods were determined using the parametric studies.
8.2.3 Dynamic performance of HCFPS
Dynamic response of HCFPS floors was investigated by conducting parametric
studies using experimentally validated FE models. Material properties, walking
frequency and damping were used as parameters in the parametric study.
Self-weight of HCFPS panel is approximately 50% of conventional
composite/concrete floor systems. The first mode natural frequency of HCFPS floors
is therefore higher due to its lighter weight. The first mode natural frequency of
HCFPS floors is greater than 10 Hz and thus HCFPS can be categorised as a high
frequency floor system. The maximum possible fourth harmonic of the walking
frequency (2.4 Hz) is lower than the first mode natural frequency [56]. As a
consequence resonant vibration is unlikely to occur.
Damping ratio of 5% can be considered conservative based on experimental tests on
bare HCFPS test panels. HCFPS floors will have superimposed dead loads higher
than 1.0 kPa due to partitions, suspended services and floor finishes. Hence, a
damping ratio of 5 % is more feasible for HCFPS floors in service.
Dynamic performance assessment of HCFPS panels was conducted using a single
panel with appropriate boundary conditions. Vibration assessment of HCFPS was
conducted for human induced walking loads that are possible in residential and office
floors. VDV [56, 60] method was used to account for the intermittent nature of such
activities and vibration response of HCFPS under walking activities was checked
with the threshold of VDV values given in BS 6472 [60]. Assessment shows that
HCFPS that are designed to satisfy static performance requirements are adequate to
provide acceptable dynamic performance under human induced walking loads.
Single panel approach is further supported by the dynamic assessment of full scale
Chapter 8: Conclusions and recommendations 157
HCFPS floor plate with supporting structural frame exhibiting similar behaviour to
the single panel approach.
HCFPS can be designed to satisfy static performance requirements for applications in
residential and office floors. Static design of HCFPS (with material types of GRC 10
and PU 20) will offer acceptable vibration performance under human induced
walking loads. Lightweight HCFPS floor systems can therefore be effectively
designed for use in office and residential buildings to provide acceptable
performance characteristics. Design guidelines for the HCFPS will be presented
based on static performance requirements. Acceptable dynamic performance can be
obtained for HCFPS, which design to satisfy static performance with material types
of GRC 10 and PU 20. However, if required, dynamic performance assessment of
HCFPS floor can be conducted using the method discussed in this thesis.
8.2.4 Design guidelines
Design guidelines of HCFPS were developed using the static performance
requirements, such that the design provides acceptable dynamic performance in the
use of residential and office floors. It is recommended to be use elastic modulus of
GRC and PU equal or greater than 10 GPa (GRC 10) and 22.4 MPa (PU 20)
respectively to obtain the acceptable dynamic performance.
Design guidelines and calculation procedure were developed for the HCFPS system
using analytical equations suggested in Chapter 6 to determine static performance.
Thickness of hybrid components, density of PU, sectional configurations and loading
were considered as the variables to obtain the optimum section for a particular span.
Parameters of the HCFPS section are shown in Figure 8-2 and used to determine the
properties of the section. Neutral axis depth ( ) and Second moment of area ( ) of
HCFPS section can be determined by using Equations 8-1 and 8-3. Equivalent
flexural stiffness of HCFPS is defined by Equation 6-4. Elastic deflection of
the HCFPS can be determined using section properties of HCFPS. For example,
deflection at the centre due to a uniformly distributed load can be found using
Equation 8-5.
158 Chapter 8: Conclusions and recommendations
Furthermore, this suggested simplified design method can be used to develop a excel
sheet and that can be used as a tool to determine the sizes of the HCFPS panels
depending on the span length and applied loads.
Figure 8-2: Parameters used to define the properties of HCFPS section
yi -Distance to the centroid (from the bottom) of individual components
bi -Width of individual components
di -Depth of individual components
-Area of the transformed section
n1 – ESteel / EPU
n2 – EGRC / EPU
Note that all the areas of this equation are net areas of the component materials.
Equation 8-1
Equation 8-2
Equation 8-3
Equation 8-4
Equation 8-5
Linear elastic stresses in each component material at the mid-span of the HCFPS can
be determined by using Equations 8-6, 8-7 and 8-8.
Chapter 8: Conclusions and recommendations 159
Equation 8-6
Equation 8-7
Equation 8-8
Moment capacity of HCFPS sections can be calculated by using a cracked section as
shown in Figure 8-3. Tensile capacity is provided only by the steel laminate and
compression capacity is provided by the slab, as shown in Figure 8-3. Plastic neutral
axis depth ( ) and Second moment of area ( ) of HCFPS section can be
determined by using the Equations 8-9 and 8-11. Equivalent flexural stiffness
of HCFPS is defined from Equation 8-12.
Figure 8-3: Parameters used to define the properties of cracked HCFPS section
Equation 8-9
Equation 8-10
Equation 8-11
Equation 8-12
Yielding stress of the steel can be used to calculate the moment capacity of HCFPS
section, as shown in Equation 8-13. Stress of in GRC layer at the top surface
160 Chapter 8: Conclusions and recommendations
( (refer to Equation 8-15) can be determined using strain. Strain of the GRC
layer can be obtained by using Equation 8-14. It was assumed that stress variation at
point of yielding is linear across the section.
Equation 8-13
Equation 8-14
Equation 8-15
Shear capacity of the HCFPS can be estimated using Equation 8-16.
Equation 8-16
HCFPS has higher shear capacity and its inherent shear capacity is adequate for
office and residential floor loading. Although, there is high shear concentration in the
GRC layer at the vicinity of support due to its higher elastic and shear modulus
compare to the PU core (refer to Section 6.3.2), shear bearing capacity of GRC can
be improved by embedding continuous glass fiber mesh in high shear stress
concentrated zones of GRC in the HCFPS (refer to Figure 6-25). High shear stress
concentrated zone can be taken as (0.15 × Span length) from the support in the beam
of HCFPS. Hence it is recommended to supply a continuous glass fiber mesh in high
shear stress concentrated zones of the HCFPS to improve the shear bearing capacity
of GRC.
8.2.5 Supporting and connection methods
The structural supporting system can be either steel or Reinforced Concrete (R/C).
HCFPS can be fixed to the structural frame as shown in Figure 8-4 as simply support
connections. HCFPS panels should be connected to the supporting structure using the
beam (refer to Figure 8-5) because any hogging moment that could arise due to the
connection would transfer through the continuous glass fiber mesh, which is
provided to enhance the shear capacity of the HCFPS at the support.
Chapter 8: Conclusions and recommendations 161
Figure 8-4: Proposed supporting methods for HCFPS floor to structural frame
Figure 8-5: Cross-section of the HCFPS panel
8.2.6 Limitation of design guidelines
These analytical methods have only been validated for the spans less than 7.5 m.
Material properties of GRC and PU must to be maintained as follows to obtain the
acceptable dynamic performance. Modulus of elasticity of GRC should be greater
than 10 GPa (GRC 10). Density of PU should be greater than 100 kg/m3 and elastic
modulus greater than 22.4 Mpa (PU 20). Simply supported conditions should be
assumed for the design of HCFPS and the analytical methods are applicable only
under such conditions.
8.2.7 Manufacturing and casting guide
HCFPS test panels were cast using a manual casting procedure. However, a
production process may be developed for automated manufacturing to enhance
production efficiency. In this context, GRC layers can be cast in controlled
conditions and placed at the intended positions. Mechanized spray techniques can be
used to apply the GRC[30]. In this process mortar paste and chopped glass fiber are
simultaneously deposited from dual spray-heads into suitable moulds. This method
162 Chapter 8: Conclusions and recommendations
currently accounts for a large percentage of GRC production. Modulus of elasticity
up to 20 GPa can be achieved using pre-cut glass fiber between 25 mm and 40 mm in
length and containing 5% of the total weight of the mix [12]. Steel laminate can be
embedded in the bottom layer of GRC (refer to Figure 8-1) during the spray process.
Subsequently, PU core can be injected in between the GRC outer shell. PU is
manufactured by combining a polyol and isocyanate, a blowing agent and an
activator, through a controlled chemical reaction in liquid form [9, 24]. Liquid foam
starts to expand rapidly and becomes hard after 3-6 minutes of mixing. During the
process, PU adheres to the GRC outer shell [9].
Alternative materials can be used for the PU core as it provides the minimal impact
on the structural capacity. However, the bonding between GRC and the core material
must be achieved to avoid the de-lamination. Damping of the HCFPS floors must
also be assessed with alternative materials.
8.2.8 Implications
HCFPS can be used as a viable alternative to conventional floor system since it
meets structural performance requirements and has many desirable properties such as
lightweight, easy to construct, economical, demountable, recyclable and reusable.
HCFPS offers such properties as a consequence of component materials (GRC, PU
and steel laminate)
HCFPS can be developed as an economical floor system by using widely used
properties of GRC and PU in current composite industry. Properties of GRC 10 can
be obtained economically either by method of spay or premix [12]. Low density PU
20 (100 kg/m3) core is used in current sandwich construction applications [11] and
HCFPS can be developed using PU 20 as an economical structure. HCFPS panels
design to satisfy static performance using GRC 10 and PU 20 will offer acceptable
vibration performance under human induced walking loads. Therefore, GRC 10 and
PU 20 can be used to develop an economical HCFPS.
HCFPS can be developed as a prefabricated floor system that can be manufactured in
an offsite factory under controlled conditions to achieve a product of superior quality
with low embodied energy. This floor system is approximately 50% lighter in weight
compared to conventional concrete floors. Therefore, this product is easy to
Chapter 8: Conclusions and recommendations 163
transport, handle and erect. HCFPS has the potential to revolutionize the construction
of structural floor systems by replacing slow, labour intensive and low quality
construction materials, with factory based manufacturing process. Manufactured
floor plates can be assembled with simple connections on site eliminating the heavy,
cumbersome and time consuming material handling, transporting and erecting
processes, while minimising safety hazard. HCFPS can be demounted and reused in
other applications or can be recycled as a whole component at the end of its
commercially useful life.
Additionally, lightweight property of HCFPS floor plates result in reduced load on
the supporting beams and columns. Thereby, sizes of such load-bearing members can
be reduced, yielding economical advantages.
Moreover, lightweight HCFPS floor and reduced size of load bearing members
results in lower mass for building structures. As a consequence, such buildings offer
better performance during earthquakes. In an earthquake, ground shaking generates
internal forces within the buildings called inertial forces, which in turn, cause the
most damage to building structures. Magnitude of inertial forces are proptional to
the mass of a structure [15]. Hence, the lower the mass of the building lower the
demand of seismic loads.
In conclusion this research, provided an efficient, lightweight, economical, and
sustainable structural flooring system that can be recycled and reused as a whole
system, compared to conventional floor systems. HCFPS offers multifunctional
structural properties, making it a viable alternative for conventional reinforced
concrete and composite deck floors. It is therefore a product that addresses social and
environmental needs of the global community using material and manufacturing process
with low energy content.
8.2.9 Future work
This research shows that HCFPS can be used in residential and office floor
construction using FE modelling with the support of limited experimental testing.
Following future research items are suggested:
164 Chapter 8: Conclusions and recommendations
Full scale HCFPS floor plate can be experimentally tested to further
investigate performance characteristics and reinforce the findings from this
research.
HCFPS has been designed for simply supported conditions, but further
improvements can be investigated for restrained and continuous supports.
Shear behaviour was identified as insignificant design criteria of the design of
HCFPS according to the current research findings (as a result of HCFPS
configuration). However, shear behaviour of the HCFPS can be
experimentally investigated to further support for this.
Dynamic performance of HCFPS can be further evaluated under human
induced rhythmic activities.
Fire performance of HCFPS can be evaluated experimentally by conducting
fire tests.
Further research investigation can be conducted to determine some sizing
rules for HCFPS using the suggested simplified design guidelines.
Design of HCFPS can be improved by incorporating the capacity reduction
factors in relation partial safety factors of materials.
Static and dynamic behaviour of the slab joint (refer to Figure 8-1) can be
further investigated experimentally using actual HCFPS floor plate.
Bibliography 165
Bibliography
1. Joseph, P. and S. Tretsiakova-McNally, Sustainable Non-Metallic Building
Materials. Sustainability 2010;2(2): 400-427.
2. O’Connor, J., Survey on actual service lives for North American buildings, In
Woodframe Housing Durability and Disaster Issues conference. 2004: Las
Vegas.
3. Harras, B., R. Benamar and R.G. White, Experimental and theoretical
investigation of the linear and non-linear dynamic behaviour of a glare 3
hybrid composite panel Journal of Sound and Vibration 2002;252(2): 281-
315.
4. Mamalis, A.G., D.E. Manolakos, M.B. Ioannidis, D.P. Papapostolou, P.K.
Kostazos, and D.G. Konstantinidis, On the compression of hybrid sandwich
composite panels reinforced with internal tube inserts: experimental.
Composite Structures 2002;56(2): 191-199.
5. Sharaf, T. and A. Fam, Experimental Investigation of Large-Scale Cladding
Sandwich Panels under Out-of-Plane Transverse Loading for Building
Applications. Journal of Composites for Construction 2011;15(3): 422-430.
6. Fam, A. and T. Sharaf, Flexural performance of sandwich panels comprising
polyurethane core and GFRP skins and ribs of various configurations.
Composite Structures 2010;92(12): 2927-2935.
7. Islam, M.M. and T. Aravinthan, Behaviour of structural fibre composite
sandwich panels under point load and uniformly distributed load. Composite
Structures 2010;93(1): 206-215.
8. Braun, D., S.J. Kennedy, D.J.L. Kennedy and D.E. Allen, Sandwich Plate
System Risers for Stadia, SSRC 2002 Annual Stability Conference, 2002,
Seattle, Washington, USA.
9. Davies, J.M., Lightweight Sandwich Construction: John Wiley & Sons Ltd,
2008.
10. Dawood, M., W. Ballew and J. Seiter, Enhancing the resistance of composite
sandwich panels to localized forces for civil infrastructure and transportation
applications. Composite Structures 2011;93(11): 2983-2991.
11. Nasir, A.M., D.P. Thambiratnam and C. Button, Performance Characteristics
of Compound Curved Sandwich Shell Structures. Journal of the International
Association for Shell and Spatial Structures 2012;53(1): 19-30.
12. Ferreira, J.G. and F.A. Branco, GRC mechanical properties for structural
applications, in Instute Superior Techno. 2004: Portugal. p. 1049-001.
13. Thirumal, M., D. Khastgir, N.K. Singha, B.S. Manjunath and Y.P. Naik,
Effect of Foam Density on the Properties of Water Blown Rigid Polyurethane
Foam. Journal of Applied Polymer Science 2008;108: 1810-1817.
14. Sharma, R.S. and V.P. Raghupathy, Shear Properties of Polyurethane Foams
with Different Densities. International Journal of Materials Sciences
2008;3(3): PP 289-297.
15. Cooper, P., P. Hoby and N.K. Prinja, How to - do seismic analysis using
finite elements: Glasgow : NAFEMS, 2007.
16. Harris, B., Engineering Composite MaterialsLondon: The Institute of
Materials, 1999.
166 Bibliography
17. Cousins, T.E., T.M. Murray and Harris D. K., Use of Sandwich Plate System
In a Virginia Bridge. 2009, Virginia Transport and Research Council:
Charlottesville, Virginia.
18. Chong, K.P. and J.A. Hartsock, Structural analysis and design of sandwich
panels with cold-formed steel facings. Thin-Walled Structures 1993;16(1-4):
199-218.
19. Gaston, G., D.P. Thambiratnam and C. Button, Computer Analysis of
Compound Curved Sandwich Shell Structures, QUT Research Week 2005.
Queensland University of Technology Research Week International
Conference, 4-5 July 2005, Brisbane, Australia, 2005.
20. Gaston, G., D.P. Thambiratnam, C. Button and A. Nasir. Performance
Characteristics of Compound Curved Sandwich Shell Structures. Structural
Engineering, Mechanics and Computation, 2004.
21. Hillman, J.R., Investigation of a Hybrid-composite Beam System., in High
Speed Rail IDEA Program. 2003, Transportation Reserch Board of National
Acadamies: Chicago.
22. Sandun De Silva, S., Vibration Characteristics of Steel-Deck Composite
Floors under Human Exitation, in School of Urban Development, Faculty of
Build Environment and Engineering. 2007, QUT Brisbane.
23. Witkiewicz, W. and A. Zielinski, Properties of The Polyurathane (PU) Light
Foams. 2006, Ship Design and Research Centre, Department of Materials
Science, Corrosion and Environment Protection: Gdansk, Porland.
24. What is Polyurethane. [Internet]. Available from:
http://www.wisegeek.com/what-is-polyurethane.htm.
25. Sharma, R.S. and V.P. Raghupathy, A Holistic Approach to Static Design of
Sandwich Beams with Foam Cores. Journal of Sandwich Structures and
Materials 2008;10(5): 429-441.
26. Yoon, K.H., S.T. Yoon and O.O. Park, Damping Properties and Transmission
Loss of Polyurethane. ( I. Effect of Soft and Hard Segment Compositions).
Applied polymer science 1999;75: pp 604-611.
27. Ashby, M. and K. Johnson, Polyurethane (PU), In Material and Design. 2002,
Butterworth-Heinemann. p. 201-202.
28. Zia, K.M., H.N. Bhatti and I.A. Bhatti, Methods for polyurethane and
polyurethane composites, recycling and recovery: A review. Reactive &
Functional Polymers 2007;67: pp 675-692.
29. Recommended Specification forManufacture, Curing and Testing of
Glassfibre Reinforced Concrete (GRC) Products, 2002. GRC Industry Group
of National Precast Concrete Association Australia.
30. Brookes, A.J. and M. Maarten, Glass-fibre reinforced cement, In Cladding of
Buildings. 2008, Taylor & Francis. p. 71-89.
31. Majmdar, A.J. and R.W. Nurse, Glass Fibre Reinforced Cement. Materials
Science and Engineering 1974;15: 107-127.
32. GRC In Action. International Glassfibre Reinforced Concrete Association
[Internet], 2003. Available from:
http://www.bcmgrc.com/general_grcinaction.pdf.
33. Recommended practice for glass fiber reinforced concrete panelsChicago, Ill:
Prestressed Concrete Institute, 1987.
34. Bijen, J., Improved Mechanical Properties of Glass Fibre Reinforced Cement
by Polymer Modification. Cement & Concrete Composites 1990;12(5): 95-
101.
Bibliography 167
35. Ellis, B.R., Dynamic monitoring, in Monitoring and Assessment of
Structures. 2001a: New York. p. 8-31.
36. Murray, T.M., Floor Vibrations: Tips for designers of office buildings
Structure 2000;26-30.
37. Paskalov, A. and S. Reese, Deterministic and probabilistic floor response
spectra. Soil Dynamics and Earthquake Engineering 2003;23(7): 605-618.
38. Rainer, J.H., Dynamic tests on a steel-joist concrete-slab floor. Canadian
Journal of Civil Engineering 1980;7(2): 213-224.
39. Strength evaluation of existing concrete buildings, In ACI manual of concrete
practice. 2005, American Concrete Institute: Farmington Hills, Michigan. p.
437R-25-28.
40. ASTM D3039/D3039-93. Standard test method for tensile properties of
polymer matrix composite materials, 2000. ASTM International: West
Conshohocken, United States.
41. ASTM C947Standard Test Method for Flexural Properties of Thin-Section
Glass-Fiber-Reinforced Concrete (Using Simple Beam With Third-Point
Loading), 2003. ASTM International: West Conshohocken, United States.
42. AS 1012.9: Methods of testing concrete - Determination of the compressive
strength of concrete specimens, 1999. Standards Australia: Sydney, NSW,
Australia.
43. ISO 1926:2005, Rigid cellular plastics — Determination of tensile properties,
2005. International Organization for Standardization: Switzerland.
44. ASTM D790 - 10 Standard Test Methods for Flexural Properties of
Unreinforced and Reinforced Plastics and Electrical Insulating Materials,
2003. ASTM International: West Conshohocken, United States.
45. ASTM C365-03. Standard test method for flatwise compressive strength of
sandwich cores, 2004. ASTM International: West Conshohocken, United
States.
46. AS 1391, Metallic materials—Tensile testing at ambient temperature, 2007.
Standards Australia: Sydney, NSW, Australia.
47. Abaqus Analysis User's Manual, Version 6.9-1: ABAQUS Inc.
48. AS/NZS 1170.1, Structural design actions, Part 1: Permanent, imposed and
other actions, 2002. Standards Australia: Sydney, NSW, Australia.
49. Bachmann, H. and W. Ammann, Vibrations in Structures induced by Man
and MechanicsSwitzerland: IABSE - AIPC - IVBH, 1987.
50. Allen, D.E. and G. Pernica. Control of Floor Vibration. Construction
Technology Update No. 22 [Internet], 1998. Available from: http://dsp-
psd.pwgsc.gc.ca/collection_2008/nrc-cnrc/NR25-2-22E.pdf.
51. Maguire, J.R. and T.A. Wyatt, Dynamics - an introduction for civil and
structural engineersLandon: Thomas Telford Publishing, 1999.
52. Ginty, D., J.M. Derwent and T. Ji, The frequency ranges of dance-type loads.
The Structural Engineer 2001;79/no 6: 27-31.
53. Smith, J.W., Dynamic Loading and Design of Structures, In Human-induced
vibration. 2002, Spon Press: New York. p. 285-306.
54. Ellis, B.R. and T. Ji, The response of structures to dynamic crowd loads, in
Digest 426. 2004: UK.
55. ISO 10137, Bases for Design of Structures- Serviceability of Buildings
Against Vibrations, 2007. International Standardisation Organisation: Geneva
Switzerland.
168 Bibliography
56. Smith, A.L., S.J. Hicks and P.J. Devine, Design of Floors for Vibration: A
New Approach.Ascot, Berkshire, UK: The Steel Construction Institute, 2007.
57. Guidelines on the Design for Floor Vibration due to Human Actions Part II:
Floor Vibration due to Walking LoadsHong Kong: Structural engineering
branch, Architectural services department (ArchSD), 2009.
58. Ellis, B.R., On the response of long span floors to walking loads generated by
individuals and crowds. The Structural Engineer 2000;78: 17-25.
59. Murray, T.M., D.E. Allen and E.E. Ungar, AISC Steel Design Guide Series
11: Floor Vibrations Due to Human ActivityChicago, Illinois: American
Institute of Steel Construction, 2003.
60. BS 6472-1:2008, Guide to evaluation of human exposure to vibration in
buildings. Vibration sources other than blasting 2008. British Standards: BSI,
UK.
61. Ellis, B.R., Serviceability evaluation of floor vibration induced by walking
loads. Structural Engineer 2001;79(21): 30-36.
62. Mahrenholtz, O. and H.A.C. Bachmann, Appendix C: Damping, In Vibration
Problems in Structures : Practical Guidelines. 1995, Birkhauser Verlag:
Germany,. p. 157-168.
63. Osborne, K.P. and B.R. Ellis, Vibration design and testing of a long-span
lightweight floor. The Structural Engineer 1990;68(15): 181-186.
64. Wyatt, T.A., Design gude on the vibration of floorAscot, Berkshire, UK: The
Steel Construction Institute UK, 1989.
65. Elnimeiri, M. and H. Iyenga, Composite floor vibrations: predicted and
measured in Steel Structures. 1989, American Society of Civil Engineers: San
Francisco, CA, USA.
66. Brownjohn, J.M.W., Energy Dissipation from Vibrating Floor Slabs due to
Human-Structure Interaction. Shock and Vibration 2001;8: 315-323.
67. Sachse, R., Modelling effects of human occupants on modal properties of
slender structures. The Structural Engineer 2002;80(5): 2.
68. Ljunggren, F. and A. Agren, Development of a new damper to reduce
resonant vibrations in lightweight steel joist floors. Applied Acoustics
2002;63(11): 1267-1280.
69. A Recommended Practice, Design, Manufacture and Installation of Glass
Reinforced Concrete (GRC), 2006. GRC Industry Group of National Precast
Concrete Association Australia.
70. Jetann, C., Development of the Response Coefficient-Root Function Method
for the Transient Vibration Serviceability Design of Flat Conctere Floors, in
School of Urban Development, Faculty of Build Environment and
Engineering. 2010, QUT Brisbane.
71. Park, R. and D. Ruitong, Ductility of Doubly Reinforced Concrete Beam
Sections. ACI Structural Journal 1988;85(2): 217-225.
72. Rinde, J.A., Poisson's ratio for rigid plastic foams. Journal of applied polymer
science 1970;14(8): 1913-1926.
73. West, J.M., A.J. Majumdar and R.C. de Vekey, Lightweight glass-reinforced
cement. Composites 1980;11(1): 19-24.
74. Sun, E.Q., Shear locking and hourglassing in msc nastran, abaqus, and ansys.
2006, MSC Software.
75. Murray, T.M., Design to Prevent Floor Vibrations. AISC Engineering Journal
1975;12(3): 82-87.
Bibliography 169
76. Clough, R.W. and J. Penzien, Dynamic of structuresNew York: McGraw-
Hill, Inc, 1993.
77. Pan, T., X. You and C. Lim, Evaluation of Floor Vibration in a
Biotechnology Laboratory Caused by Human Walking. Journal of
Performance of Constructed Facilities 2008;22(3): 122-130.
78. Živanović, S., A. Pavic and P. Reynolds, Vibration serviceability of
footbridges under human-induced excitation: a literature review. Journal of
Sound and Vibration 2005;279(1–2): 1-74.
79. Soranakom, C. and B. Mobasher, Correlation of tensile and flexural
responses of strain softening and strain hardening cement composites.
Cement and Concrete Composites 2008;30(6): 465-477.
80. AS/NZS 1170.0, Structural design actions, Part 0: General principles, 2002.
Standards Australia: Sydney, NSW, Australia.
81. DomeShells Technology, Superior Building System for the Changing World.
by Domeshells Australia Pty. Ltd. [Internet]. Available from:
http://www.domeshells.com.au/.
82. Peled, A. and B. Mobasher, Tensile Behavior of Fabric Cement-Based
Composites: Pultruded and Cast. Journal of Materials in Civil Engineering
2007;19(4): 340-348.
83. BS 8110-1, Part 1: Code of practice for design and construction, 1997.
British Standards: BSI, UK
84. Iskhakov, I. and Y. Ribakov, Two-layer pre-stressed beams consisting of
normal and steel fibered high strength concrete. Materials & Design
2008;29(8): 1616-1622.
85. Sandun De Silva, S. and D.P. Thambiratnam, Dynamic characteristics of
steel-deck composite floors under human-induced loads. Computers &
Structures 2009;87: 1067-1076.