1994 Punching shear strength of reinforced and post ...
Transcript of 1994 Punching shear strength of reinforced and post ...
University of WollongongResearch Online
University of Wollongong Thesis Collection University of Wollongong Thesis Collections
1994
Punching shear strength of reinforced and post-tensioned concrete flat plates with spandrel beamsChoong-Luin ChiangUniversity of Wollongong
Research Online is the open access institutional repository for theUniversity of Wollongong. For further information contact ManagerRepository Services: [email protected].
Recommended CitationChiang, Choong-Luin, Punching shear strength of reinforced and post-tensioned concrete flat plates with spandrel beams, Doctor ofPhilosophy thesis, Department of Civil and Mining Engineering, University of Wollongong, 1994. http://ro.uow.edu.au/theses/1259
PUNCHING SHEAR STRENGTH OF REINFORCED AND POST-TENSIONED CONCRETE FLAT PLATES WITH
SPANDREL BEAMS
A thesis submitted in fulfilment of the requirements
for the award of the degree of
Doctor of Philosophy
from
JT* 0%. F*
THE UNIVERSITY OF WOLLONGONG
by
C H O O N G - L U I N C H I A N G , PEng., B.E. (Honours Class I),
Grad IEAust
DEPARTMENT OF CIVIL AND MINING ENGINEERING
1994
"The world and all that is in
it belong to the Lord; the earth and all who live on
it are his. He built it on deep waters
beneath the earth and laid its foundations in the
ocean depths."
- Psalms 24.1-2
11
DECLARATION
I declare that this work has not been submitted for a degree to any university or
such institution except where specifically indicated.
Choong-Luin Chiang
August 1994
iii
ACKNOWLEDGMENTS
The author deeply appreciates the opportunity provided by his thesis supervisor,
Associate Professor Yew-Chaye Loo, in the pursuit of this doctoral degree. The close
supervision, frank and fruitful discussions, and fatherly guidance given by Professor Loo
have been instrumental in honing the author's research skills over the course of this
study.
Sincere gratitudes are extended to the University of Wollongong for the scholarship.
The thesis research forms an integral part of an award of the Australian Research
Council funded project of Professor Loo. The author is grateful to the Council for
supporting the monumental task of this research project, over the past three-and-a-half
years.
The author also acknowledges the generous contributions in material and services
provided by the following firms in Wollongong and Sydney:
- Acrow Pty Ltd;
- Alliance Equipment Pty Ltd, Wollongong;
- Austress PSC, Sydney;
- Cleary Bros. Concrete Pumping;
- Nippy Crete Concrete; and
- VSL Prestressing, Sydney, in particular Mr. Jeff Lind (former Senior Engineer).
The technical staff of the Department of Civil and Mining Engineering, University of
Wollongong, in particular, Messrs. N. Gal, R. Webb, F. Hornung (who has since left),
iv
G. Caines, C. Allport and A. Grant are remembered for their invaluable assistance in the
laboratory work.
Acknowledgments, gestures of friendship and comradeship are also extended to special •
groups of current and ex-fellow students, sharing a common goal in model construction:
- J.R.A. Wagg and F. Stillone (Undergraduate class of '91);
- T.V. Chau, S.A. Clarkson, T. Mustaffa and D. Chaisumdet (Undergraduate class of
'92);
- H.T. Pengelly, K.F. Lei, H.M. Hii and J. O'Connors (Undergraduate class of '93); and
- A. LaManna, A. Wachtel, S. Merange and G. Canlis (Undergraduate class of '94).
Special acknowledgment is due to my wife, Sheila, for her constant support and
encouragements given, in sickness and in health, be it near or far, throughout the period
of this study. Her typing and proof-reading skills have been a God-send.
Finally, the author is indebted to all his other family members in Malaysia, USA and
Hong Kong for their understanding and encouragements, right from the very beginning.
v
ABSTRACT
In the design of reinforced or prestressed concrete flat plates with spandrel beams, one
of the most critical aspects that needs to be considered is the punching shear strength, Vu
of the slab-column-spandrel connections. Research on this behaviour has been carried
out in recent years by various engineering academics and professionals. One of the more
encompassing analytical approach in determining Vu , is the procedure developed by Loo
and Falamaki* for reinforced concrete slab-column-spandrel connections at the edge-
and corner-column positions. In the case of prestressed concrete flat plates, there is as
yet no reliable procedure for the determination of Vu , particularly for the above stated
slab-column-spandrel connections. Thus, the main objective of the present study is to
develop an analytical procedure for the prediction of Vu , for these types of structural
connections. This procedure is an extension of the analytical approach developed by
Loo and Falamaki* whereby the effect of prestress is taken into account and
incorporated into the appropriate equilibrium equations.
The development albeit an extension of a reliable analytical method requires accurate
test results from large-scale models with proper boundary conditions. Tests up to failure
were carried out on four prestressed post-tensioned concrete flat plate models, two with
spandrel beams of different widths and depths, while the other two have a torsion strip
and a free edge respectively. These half-scale models, each weighing approximately 5
tonnes, represent two adjacent panels at the corner of a typical flat plate structure and
* Loo, Y.C. and Falamaki, M. (1992), "Punching Shear Strength Analysis of Reinforced Concrete
Flat Plates With Spandrel Beams", ACI Structural Journal, V.89, No.4, July-August, 1992, pp.375-
383.
vi
Abstract
were tested under simulated uniformly distributed loads. For ease of construction, each
of the flat plate models was supported on six concrete column sections, which were cast
and rigidly connected onto prefabricated steel sections at the bottom instead of having
concrete columns. Specially designed load cells were used to measure the three reaction
components at the hinged base of each of the column supports. Strain gauges were also
attached to the reinforcing bars and prestressed tendons of the slab. The strains and
other electrical signals were logged using a Hewlett Packard 3054A data acquisition
control system via a Hewlett Packard 9826 computer.
Besides the experimental work, three concurrent studies were also carried out. The first
was a comparative study of the various methods of punching shear strength analysis of
reinforced concrete flat plates with spandrel beams. It was found that the Wollongong
Approach developed by Loo and Falamaki [1992] was superior to alternative methods
including that recommended by the Australian Standard for Concrete Structures (AS
3600-1988). The second was a parametric study of the effect of spandrel beam on the
punching shear strength analysis, using the Wollongong Approach. From this study, a
range of parametric limitations were identified within which the Wollongong Approach
may be applied.
Finally, based on the experimental results, a theoretical study was undertaken which led
to the development of a prediction procedure for the punching shear strength, Vu of
post-tensioned concrete flat plates with spandrel beams, the details of which are
presented herein. As with the Wollongong Approach, the proposed procedure (referred
to in this thesis as the "Modified Wollongong Approach") is applicable to the analysis of
failures at the corner- and edge-column positions. In addition to the consideration of
parameters such as the overall geometry of the connection, the concrete strength and
vii
Abstract
the slab restraint on the spandrel, the procedure also takes into account the following
parameters:
(1) the size and location of the prestressed tendons,
(2) the magnitude of the applied prestress, and
(3) the profiles of the draped tendons.
In comparing the accuracy and reliability of the proposed analytical procedure with the
Australian Standard approach, the latter was found to be less reliable in that it
consistently overestimates the value of Vu , especially at the corner-column positions.
This is also consistent with the compared results of the comparative study on reinforced
concrete flat plates with spandrel beams.
vin
Table of Contents
TABLE OF CONTENTS
Title Page i Declaration iii Acknowledgments iv
Abstract vi
Table of Contents ix
List of Figures xv List of Tables xix Notation xxii
1. Introduction ...1 1.1 General Remarks.. 1
1.2 The Prediction of Vu 2
1.3 Aims of Research 3
1.4 Scope of Research and Experimental Work 4
1.5 Theoretical Background 8
1.5.1 Punching shear .8
1.5.2 Theoretical aspect 10
1.6 Overall Layout of Thesis 13
2. Literature Review 15 2.1 Background of Previous Work 15
2.2 Codes of Practice 18
2.3 Methods of Punching Shear Strength Analysis 19
2.3.1 Australian Standard A S 3600-1988 19
2.3.2 American Concrete Institute ACI318-89 19
2.3.3 British Standard BS 8110:1985 20
2.3.4 The Wollongong Approach 20
2.3.5 Others 21
2.4 Previous Experimental Work on Post-Tensioned Concrete Flat Plates 22
2.4.1 J.F.Brotchie and F.D.Beresford [ 1967] 22
2.4.2 N.H.Burns and R.Hemakom [1977] 22
2.4.3 Regan [1985] 23
2.4.4 D.A.Foutch, W.L.Gamble and H.Sunidja [1990] ....23
2.4.5 A.E.Long and D.J.Cleland [1993] 24
2.5 Comments on Previous Experimental Work 24
ix
Table of Contents
3. Performance of Analytical Methods for Punching Shear
Strength of Reinforced Concrete Flat Plates 25
3.1 General Remarks 25
3.2 Objectives and Scope 25
3.3 Methods of Analysis - A Review 27
3.3.1 Australian Standard Procedure [SAA, 1988] 27
3.3.2 The ACI Method [ACI, 1989] 29
3.3.3 The British Standard Method [BSI, 1985] 31
3.3.4 The Wollongong Approach [Loo and Falamaki, 1992] 33
3.4 Presentation and Comparison of Results 34
3.4.1 Model tests 34
3.4.2 Predicted results. 37
3.4.3 Comparison: Flat plates with torsion strips 39
3.4.4 Comparison of ACI and British Standard methods incorporating
reduction factors: Flat plates with torsion strips 41
3.4.5 Comparison: Flat plates with spandrel beams 44
3.5 Conclusions 46
4. Parametric Study of the Effect of Spandrel Beams on the
Punching Shear of Strength of Reinforced Concrete Flat
Plates 47
4.1 Introduction 47
4.2 Formulas for Spandrel Strength Parameter 47
4.3 Generating Models of Varying Features of Spandrel Beam 49
4.3.1 Varying the width of spandrel beam 49
4.3.2 Varying the depth of spandrel beam 54
4.3.3 Varying the area of longitudinal steel reinforcement in spandrel
beam 58
4.4 Summary of Parametric Study Results 62
5. Theoretical Approach to Incorporate Effect of Prestress in the
Analysis of Punching Shear 64
5.1 Effect of Prestress on Punching Shear 64
5.1.1 General 64
5.1.2 Current concrete codes of practice 67
5.1.3 Effect of prestress in the Wollongong Approach 70
5.1.4 Equilibrium equations 73
5.1.5 Spandrel beam parameters 75
X
Table of Contents
5.1.6 Interaction equation 78
5.1.7 Modifications to interaction equation 80
5.2 Evaluation of Ultimate Moment and Shear 81
5.3 Analysis of Punching Shear 83
5.3.1 Corner connections with spandrel beam 83
5.3.2 Edge connections with spandrel beam 84
5.3.3 Connections with torsion strip 85
5.4 Ultimate Flexural Strengths Mju and Mm for Post-tensioned Concrete
Sections 86
5.5 Losses of Prestress 88
5.5.1 Immediate losses 88
5.5.2 Time-dependent losses 91
5.5.3 Summary of prestress losses considered in the
calculation of Vu 92
6. Test Program - An Introduction to the Test Models 93
6.1 General 93
6.2 Details of Models 93
6.2.1 Dimensions 94
6.2.2 Support system 94
6.3 Design of Models 96
6.3.1 Design of prestressing and reinforcement 96
6.3.2 Prestressing and reinforcement details 102
6.3.3 Cable ducting and anchorage system 107
6.3.4 Concrete specifications 109
6.3.5 Design of spandrel beam 109
7. Test Program - Construction, Instrumentation and Testing of
Mod e l s 113
7.1 General Remarks 113
7.2 Construction of Models 114
7.2.1 Setting up of formwork 114
7.2.2 Cable ducts 118
7.2.3 Placement of ducts and reinforcement 122
7.2.4 Casting and curing of concrete and removal of formwork 127
7.2.5 Post-tensioning of prestressing tendons 129
7.2.6 Grouting 131
7.3 Instrumentation of Models 133
XI
Table of Contents
7.3.1 Strain gauges 134
7.3.2 Load cells 137
7.3.3 Whiffle-tree loading system 137
7.3.4 Deflection measurement 139
7.4 Testing of Models 141
7.4.1 During load application 141
7.4.2 After load application 143
7.4.3 Precautions prior to and during testing 144
7.4.4 Problems encountered during model testing 145
8. Presentation and Discussion of Test Results 146
8.1 General 146
8.2 Crack Patterns and Failure Mechanism 147
8.2.1 Model PI 147
8.2.2 Model P2 150
8.2.3 Model P3 153
8.2.4 Model P4 .'.157
8.3 Deflections 160
8.3.1 Model PI 160
8.3.2 Model P2 162
8.3.3 Model P3 163
8.3.4 Model P4 165
8.4 Punching Shear Strength and Unbalanced Moments 167
8.4.1 General 167
8.4.2 . Experimental results for shear, moment and torsion 168
8.4.3 Failure modes at the corner- and edge-column connections 172
8.4.4 Comparison of measured vertical column reactions and
total applied load 173
8.5 Observations 176
9. Determination of Flexural Strength of Critical Slab Strips 177
9.1 General Remarks 177
9.2 Flexural Behaviour of Uncracked and Cracked Sections 177
9.2.1 Uncracked section analysis 179
9.2.2 Cracked section analysis 180
9.3 Flexural Strength Theory 181
9.3.1 Formulas for the prestressed tendon strain components 181
9.3.2 Determination of Mu for a prestressed concrete section 182
xii
Table of Contents
9.4 Comparison of Measured and Calculated Values for Mj and Mm 188
10. Calibration of the Prediction Formulas for Spandrel
Parameter 191
10.1 General Remarks 191
10.2 Research Scheme for Determination of Semi-empirical Formulas
for \|/ 192
10.3 Measurement of the Slab Restraining Factor \\f 193
10.3.1 Derivation of equations to measure \|/ 193
10.3.2 Measured values of \j/ 197
10.4 Calibration of Semi-empirical Formulas for \[/ 199
10.5 Comparison Between Measured and Predicted Values of \|/ 204
11. Analysis of Punching Shear Strength 206
11.1 General Remarks 206
11.2 The Prediction Formulas for Vu 207
11.2.1 Corner-connections with spandrel beam 207
11.2.2 Edge-connections with spandrel beam 208
11.2.3 Connections with torsion strip or free edge 209
11.3 Computer Program for Vu 209
11.4 Comparison of Results 211
11.4.1 Predicted punching shear results by the proposed approach
(or the Modified Wollongong Approach) 212
11.4.2 Predicted punching shear results by codes of practice
and comparison 212
11.5 Comments on the Accuracy and Reliability of the Proposed Approach
(or the Modified Wollongong Approach) 216
11.5.1 Flat plates without spandrel beams (Models PI and P4) 216
11.5.2 Flat plates with spandrel beams (Models P2 and P3) 218
11.6 Summary 219
12. Conclusion 221
12.1 General Remarks 221
12.2 Performance and Limitations of the Wollongong Approach 222
12.3 Experimental and Theoretical Studies of Post-tensioned
Concrete Flat Plates 223
12.3.1 Deformation and Failure Mechanisms 223
12.3.2 Prediction of Vu by the proposed approach
xiii
Table of Contents
(Modified Wollongong Approach) 224
12.3.3 Accuracy of the proposed approach
(Modified Wollongong Approach) 225
12.4 Recommendations for Further Study 225
References 227
Appendix A Calibration of Prestressing Jack 234
Appendix B Deflection Measurements 237
Appendix C Plots of Experimental Moment, Shear and Torsion Results for
Post-tensioned Concrete Flat Plate Models 249
Appendix D Tabulations of Experimental Moment, Shear and Torsion Results for
Post-tensioned Concrete Flat Plate Models 254
Appendix E Flexural Strength Theory 259
Appendix F Strain Measurement Results 266
Appendix G Sample Calculations of Mj Using Measured Tendon Strain 272
Appendix H Sample Calculations for Punching Shear Strength Using the
Proposed Approach (Modified Wollongong Approach) 275
Appendix I Details of Spandrel Beams or Torsion Strips and
Critical Slab Strips 284
Appendix J Listings of Source Code for BASIC Computer Programs
to analyse Vu 286
Papers Published Based on This Thesis 307
xiv
List of Figures
LIST OF FIGURES
Fig. 1.1 Typical half-scale flat plate floor Fig. 1.2 Test models Fig. 1.3 Types of shear failures in a slab Fig. 1.4 Free-body diagrams for edge and corner connections Fig. 3.1 Relative strength of concrete cubes and cylinders Fig. 3.2 Flat plate model Wl with a typical spandrel beam section details
Fig. 3.3 Measured Vu versus predicted Vu - models with torsion strips Fig. 3.4 Measured Vu versus predicted Vu - models with spandrel beams Fig. 3.5 Measured Vu versus predicted Vu by A C I and British methods
incorporating reduction factors - models with torsion strips Fig. 4.1 Generated models with varying width of spandrel beam
Fig. 4.2 Graph of punching shear Vu versus spandrel strength parameter 5
of generated models with varying width bsp and shallow spandrel
beams
Fig. 4.3 Graph of punching shear Vu versus spandrel strength parameter 8
of generated models with varying width bsp and very deep
spandrel beams Fig. 4.4 Generated models with varying depth of spandrel beam
Fig. 4.5 Graph of punching shear Vu versus spandrel strength parameter 8
of generated corner-connection models with varying Dj
Fig. 4.6 Graph of punching shear Vu versus spandrel strength parameter 8
of generated edge-connection models with varying!)/
Fig. 4.7 Generated models with varying amount of longitudinal steel reinforcement in spandrel beam
Fig. 4.8 Graph of punching shear Vu versus spandrel strength parameter 8
of generated corner-connection models with varying A[s Fig. 4.9 Graph of punching shear Vu versus spandrel strength parameter 8
of generated edge-connection models with varying A/iV Fig. 5.1 Effect of prestress on principal stresses in the centroid of a slab
strip Fig. 5.2 Forces acting in the corner column region of a post-tensioned
concrete flat plate Fig. 5.3 Forces acting in the edge column region of a post-tensioned
concrete flat plate Fig. 5.4 Forces, strains and stresses in a prestressed concrete section
Fig. 6.1 Typical column support details Fig. 6.2 Design layout of cable ducts Fig. 6.3 Design layout of steel reinforcement Fig. 6.4 Profile of cable ducts along short span of flat plate model
xv
List of Figures
Fig. 6.5 Profile of cable ducts along long span of flat plate model Fig. 6.6 Stress versus strain plot for 5.05 m m diameter prestressed tendon Fig. 6.7 Stress versus strain plot for 10-rnm diameter steel reinforcing bar Fig. 6.8 Tendon and reinforcement details at column strips of post-
tensioned concrete flat plate models Fig. 6.9 Shape and section view of cable duct Fig. 6.10 Anchorage system Fig. 6.11 Stress versus strain plot for 6-mm diameter steel bar Fig. 6.12 Details of a typical spandrel beam (for Model P2) Fig. 7.1 General set-up of a flat plate model Fig. 7.2 Details of slab formwork and the supporting frame Fig. 7.3 Typical details of formwork and support system for spandrel
beam at the slab edge Fig. 7.4 Bending process of duct to final shape Fig. 7.5 Typical fabricated cable duct Fig. 7.6 Typical 12-mm bearing plate Fig. 7.7 P V C grout vent for cable ducts Fig. 7.8 Tendons protruding from the edge formwork Fig. 7.9 Top steel reinforcement over corner column A Fig. 7.10 Top steel reinforcement over edge column B Fig. 7.11 Spandrel beam reinforcement for Model P2 Fig. 7.12 Casting of concrete slab for flat plate model Fig. 7.13 Concrete test cylinders Fig. 7.14 Barrel and wedge used for anchorages
Fig. 7.15 Post-tensioning of tendons Fig. 7.16 Grouting of prestressed tendons Fig. 7.17 Hewlett-Packard 9826 computerised data acquisition system Fig. 7.18 Locations of strain gauges on steel reinforcement Fig. 7.19 Locations of strain gauges on tendons in cable ducts Fig. 7.20 Close-up view of load cells at the base of steel column Fig. 7.21 General view of whiffle-tree loading system Fig. 7.22 Dial gauges set up on top of slab to measure deflections Fig. 7.23 Set up of the Linear Variable Differential Transformer (LVDT) Fig. 7.24 Set up of synchronization control system for the two double
acting hydraulic rams Fig. 7.25 Post-mortem examination of test Model P3 after failure
Fig. 8.1 Crack pattern on top of Model PI Fig. 8.2 Crack pattern at soffit oi Model PI Fig. 8.3 Shear failure (Model PI - column A) Fig. 8.4 Crack pattern on top and sides of Model P2 Fig. 8.5 Crack pattern at soffit and sides of Model P2
Fig. 8.6 Flexural failure (Model P2) Fig. 8.7 Shear failure (Model P3 - column B) Fig. 8.8 Crack pattern on top and sides of Model P3
xvi
List of Figures
Fig. 8.9 Crack pattern at soffit and sides of Model P3 Fig. 8.10 Shear failure (Model P4 - column B) Fig. 8.11 Crack pattern on top and sides of Model P4
Fig. 8.12 Crack pattern at soffit and sides of Model P4 Fig. 8.13 Deflection measurement stations
Fig. 8.14a Load-deflection plots - Model PI (Locations 1, 5 and 6) Fig. 8.14b Load-deflection plots - Model PI (Locations 1, 3 and 4) Fig. 8.14c Load-deflection plots - Model PI (Locations 7, 8 and 9) Fig. 8.15a Load-deflection plots - Model P2 (Locations 1, 5 and 6) Fig. 8.15b Load-deflection plots - Model P2 (Locations 2, 3 and 4) Fig. 8.15c Load-deflection plots - Model P2 (Locations 7, 8 and 9) Fig. 8.16a Load-deflection plots - Model P3 (Locations 1, 5 and 6) Fig. 8.16b Load-deflection plots - Model P3 (Locations 2, 3 and 4) Fig. 8.16c Load-deflection plots - Model P3 (Locations 7, 8 and 9) Fig. 8.17a Load-deflection plots - Model P4 (Locations 1, 5 and 6)
Fig. 8.17b Load-deflection plots - Model P4 (Locations 2, 3 and 4) Fig. 8.17c Load-deflection plots - Model P4 (Locations 7, 8 and 9) Fig. 8.18 Experimental shear, moment and torsion recorded at corner-
connections A and C and at edge-connection B for Model P2 Fig. 9.1 Flow chart to determine the theoretical ultimate flexural strength
of prestressed concrete slab strips Fig. 9.2 Flow chart to determine the flexural strength of prestressed
concrete slab strips using strain measurements Fig. 9.3 Graph of measured versus calculated Mj and Mm results Fig. 10.1 Research scheme to establish the semi-empirical formulas to
determine \|/
Fig. 10.2 Graphs of y versus log10(78) for connections A and C , and \\f
versus logio(A-S) for connections B
Fig. 11.1 Measured versus predicted Vu for post-tensioned concrete flat plate models without spandrel beams (PI and P4)
Fig. 11.2 Measured versus predicted Vu for post-tensioned concrete flat plate models with spandrel beams (P2 and P3)
Fig. Cl.l Plot of moment transfer to columns A, B and C versus applied load up to failure (Model PI)
Fig. CI.2 Plot of shear strength around columns A, B and C versus applied
load (Model PI) Fig. CI.3 Plot of torsional moment at' columns A, B and C versus applied
load (Model PI) Fig. C2.1 Plot of moment transfer to columns A, B and C versus applied
load (Model P3) Fig. C2.2 Plot of shear strength at columns A, B and C versus applied load
(Model P3) Fig. C2.3 Plot of torsional moment at columns A, B and C versus applied
load (Model P3)
xvu
List of Figures
Fig. C3.1 Plot of moment transfer to columns A, B and C versus applied load (Model P4)
Fig. C3.2 Plot of shear strength at columns A, B and C versus applied load (Model P4)
Fig. C3.3 Plot of torsional moment at columns A, B and C versus applied load (Model P4)
Fig. E. 1 Typical strain and stress distribution of a prestressed concrete cross-section
Fig. E.2 Idealised stress block at ultimate moment conditions Fig. E.3 Strain distribution of prestressed tendon at three stages of
loading
xviii
List of Tables
LIST OF TABLES
Table 3.1 Table 3.2 Table 3.3 Table 3.4 Table 3.5 Table 3.6
Table 3.7
Table 3.8
Table 3.9
Table 3.10 Table 3.11 Table 3.12 Table 4.1
Table 4.2
Table 4.3
Table 4.4
Table 5.1 Table 6.1
Table 6.2
Table 7.1 Table 8.1 Table 8.2
Table 8.3
Table 8.4 Table 8.5 Table 8.6 Table 8.7
Comparison of punching shear results - models with torsion strips Comparison of punching shear results - models with spandrel beams Pearson correlation coefficients for models with torsion strips Deviations of predicted Vu for models with torsion strips Frequency distribution of predicted Vu for models with torsion strips Comparison of factored punching shear results by A C I and British
Standard methods - models with torsion strips Pearson correlation coefficients for models with torsion strips (after incorporating reduction factors) Deviations of predicted Vu for models with torsion strips (after incorporating reduction factors) Frequency distribution of predicted Vu for models with torsion strips (after incorporating reduction factors) Pearson correlation coefficients for models with spandrel beams Deviations of predicted Vu for models with spandrel beams Frequency distribution of predicted Vu for models with spandrel beams Typical generated data and results from parametric study of varying
width of spandrel beam in Model M2-A Typical generated data and results from parametric study of varying
depth of spandrel beam in Model M2-A Typical generated data and results from parametric study of varying longitudinal reinforcement in spandrel beam of Model M2-A
Summary of parametric study results of limits or ranges of 8 for reliable
punching shear strength analysis Summary of formulas for prestress losses used in the analysis of Vu Summary of tendon and reinforcement details in the critical column strips
and spandrel beam Yield strengths of steel reinforcement used for longitudinal bars and
closed ties in spandrel beam Laying sequence of cable ducts Concrete strengths of half-scale models Experimental shear, moment and torsion recorded at corner-connections
A and C and at edge-connection B for Model P2 Measured punching shear, unbalanced moments and applied loads at
failure for post-tensioned concrete flat plate models Comparison of vertical column reactions and applied loads for Model PI
Comparison of vertical column reactions and applied loads for Model P2 Comparison of vertical column reactions and applied loads for Model P3 Comparison of vertical column reactions and applied loads for Model P4
xix
List of Tables
Table 9.1 Strain measurements at ultimate for all models, at front face of columns and midspans of critical slab strips
Table 9.2 Measured and calculated values of Mj and Mm for all models
Table 10.1 Variables used to determine the slab restraining factor, \j/
Table 10.2 Results for the calculated variables and the slab restraining factor, \j/
Table 10.3 Calculated values of 8 , 7 , X , log^^S) and log^o(^8) for all post-
tensioned concrete flat plate models PI to P4
Table 10.4 Coefficients of determination ( R2) and correlation coefficients (R) of the
plots of \j/ versus log 10(78) and \j/ versus log10(X8)
Table 10.5 The measured and predicted values of \|/
Table 11.1 Punching shear strength predictions for post-tensioned concrete flat plate models PI to P4 using the proposed approach (Modified Wollongong
Approach) Table 11.2 Predicted punching shear strength results using the methods
recommended by A S 3600-1988, ACI318-89 andBS 8110:1985
Table 11.3 Deviations of predicted Vu for models without spandrel beams Table 11.4 Deviations of predicted Vu for models with spandrel beams Table B. 1 a Deflection measurements of Model PI (using rulers) Table B. 1 b Deflection measurements of Model PI (using LVDTs) Table B .2a Deflection measurements of Model P2 (using rulers) Table B.2b Deflection measurements of Model P2 (using LVDTs) Table B.3a Deflection measurements during first day of testing of Model P3 (using
rulers) Table B.3b Deflection measurements during second day of testing of Model P3
(using rulers) Table B.3c Deflection measurements during first day of testing of Model P3 (using
dial gauges) Table B.3d Deflection measurements during second day of testing of Model P3
(using dial gauges) Table B.3e Deflection measurements during first day of testing of Model P3 (using
L V D T s ) Table B.3f Deflection measurements during second day of testing of Model P3
(using L V D T s ) Table B.4a Deflection measurements during first day of testing of Model P4 (using
rulers) Table B.4b Deflection measurements during second day of testing of Model P4
(using rulers) Table B.4c Deflection measurements during first day of testing of Model P4 (using
dial gauges) Table B.4d Deflection measurements during second day of testing of Model P4
(using dial gauges) Table B.4e Deflection measurements during first day of testing of Model P4 (using
LVDTs)
xx
List of Tables
Table B.4f Deflection measurements during second day of testing of Model P4
(using LVDTs) Table D. 1 Shear, moment and torsion at columns A, B and C for Model PI Table D.2 Shear, moment and torsion at columns A, B and C for Model P2 Table D.3a Shear, moment and torsion at columns A, B and C for Model P3 - Day
One Table D.3b Shear, moment and torsion at columns A, B and C for Model P3 - Day
T w o Table D.4a Shear, moment and torsion at columns A, B and C for Model P4 - Day
One Table D.4b Shear, moment and torsion at columns A, B and C for Model P4 - Day
T w o Table F. 1 Strain measurements of reinforcing steel for Model PI Table F.2a Strain measurements of reinforcing steel for Model P2 Table F.2b Strain measurements of prestressed tendons for Model P2 Table F.3a Strain measurements of prestressed tendons for Model P3 - Day One Table F.3b Strain measurements of prestressed tendons for Model P3 - Day Two Table F.4a Strain measurements of prestressed tendons for Model P4 - Day One Table F.4b Strain measurements of prestressed tendons for Model P4 - Day Two Table F.4c Strain measurements of reinforcing steel for Model P4 - Day Two Table 1.1 Details of the spandrel beams or torsion strips Table 1.2 Column dimensions of post-tensioned concrete flat plate models
Table 1.3 Details of the critical slab strips Table 1.4 Details of prestressing in critical slab strips
xxi
Notation
NOTATION
a = the segment length of the critical shear perimeter parallel to the vector
direction of M
A = the area of the critical section
A' = the area of concrete under idealized rectangular compressive stress block AFL = anchorage friction losses
Ag = the gross area of the prestressed concrete section AL], AL2 = anchoring losses A/v = the total area of the longitudinal steel reinforcement in the spandrel beam
Ms.gen - the generated longitudinal steel area Ap = the area of prestressed tendons Api , AP2 = the area of the prestressed tendons in the main and transverse bending
directions Asc = the area of the compressive steel in the slab Ast = the area of tensile steel reinforcement in a cross-section Asti = the area of negative steel reinforcement in critical slab strip Astm = the area of positive steel reinforcement in critical slab strip At = the area of the rectangle defined by the centres of the four longitudinal
corner bars of the spandrel beam A't = area of the transformed cross-section Aws = the area of closed ties in the spandrel beam
b = the width of the critical slab strip
bSp = the width of spandrel beam bsp,gen = the generated width of the spandrel beam bw = the width of the spandrel beam
C = the resultant compressive force of magnitude
Cc = compressive force in the concrete above the neutral axis
Cs = a steel component of the resultant compressive force
d = the effective depth of the reinforced concrete slab dc = the effective depth of the compressive reinforcement
de = the effective depth of prestressed concrete slab de j = the depth of negative steel reinforcement in critical slab strip
dn = the depth to the neutral axis of a section dp = the depth of the prestressed tendons with respect to the extreme
compressive fibre dpi = the depth of the prestressed tendons in the main bending direction dp2 - the depth of the prestressed tendons in the transverse bending direction
xxii
Notation
ds = the depth of the tensile steel with respect to the extreme compressive fibre
dsc = the depth of the compressive steel with respect to the extreme compressive fibre
dsp = the effective depth of spandrel beam
dspi = the depth of tensile steel reinforcement in spandrel beam D, = the depth of spandrel beam
Di,gen = the generated depth of the spandrel beam Ds = the overall depth of the slab
e = the eccentricity of prestressing force from the centroid of the uncracked section
ej = the eccentricities of the prestressed tendons in the main bending direction e2 = the eccentricities of the prestressed tendons in the transverse bending
direction Ep = the modulus of elasticity of prestressed tendon Es = the modulus of elasticity of steel
fcf
h
J
kj ,k2 ,
k4,k5 , and ky
I
fc?>
**
lp, ls and lsc
Lpa
M,
M2
= the tensile stress of the concrete section
= the characteristic cube strength of the concrete Jcu
fey = the limiting concrete shear stress fly = the yield strength of the longitudinal steel reinforcement in the spandrel
beam /^y = yield strength of closed ties in the spandrel beam
= the moment of inertia of the gross prestressed concrete section
= the analogous polar moment of inertia
= the Jk-factors used for determining Vu (Modified Wollongong Approach)
= the lever arm between the internal compressive and tensile resultants = the lever arms between the concrete compressive force and the respective
forces at the prestressed and non-prestressed steel levels
= distance along the tendon from the jacking end
= the negative yield moment of the slab over the front segment of the
critical perimeter = the negative yield moment of the slab over the side segment of the critical
perimeter
xxiii
Notation
MCj = the total unbalanced moments transferred to the column centre in the main moment directions
M(22 = the total unbalanced moments transferred to the column centre in the transverse moment directions
Mcr - the cracking moment to be resisted by the section when the first crack is formed
Mm - the bending strength of the critical slab strip at the corresponding mid-
span M0 = decompression moment, that is the moment necessary to produce zero
stress in the concrete at the extreme tension fibre
MSp = pure bending for the spandrel beam Mu = ultimate moment M M V = pure bending in spandrel beam at ultimate
M = the design bending moment at the column
n = the modular ratio of steel reinforcement
n} = the number of prestressed tendons at the front face of the column in the
main bending direction n2 = the number of prestressed tendons at the side face of the column in the
transverse bending direction np = the modular ratio of prestressed tendons
pa = the force in the tendon at any point Lpa (in metres) from the jacking end
pe = the final or effective prestressing force Pel* Pe2 = tne prestressing forces being applied at the front faces of the column-slab
connections
Pe]x,Peiy = prestress components Pe2x ' P&y p. = the force in the tendon immediately after transfer p = the prestressing force imposed at the jack
P - the vertical component of the prestress usually acts in the opposite
direction to the load induced shear
s = the spacing of closed ties in spandrel beam
T = the resultant tensile force j 2 = the negative yield moment of the torsional moment over the side segment
of the critical perimeter
Tp = tensile forces in the prestressing tendon Ts = tensile forces in the reinforcing steel Tsp = pure torsion for the spandrel beam
T^ = P u r e torsion in spandrel beam at ultimate
xxiv
the critical shear perimeter
the perimeter of the rectangle defined by the centres of the four longitudinal comer bars of the spandrel beam
design shear force acting at the centre of the column the shear force over the front segment of the critical perimeter the shear force over the side segment of the critical perimeter the ultimate shear resistance of a concrete section which is uncracked the ultimate shear resistance when the section is cracked in flexure the vertical component of the effective prestress force at the section which is generally deemed to be negligible and may be ignored
pure shear for the spandrel beam the punching shear strength of reinforced and prestressed concrete flat
plates the measured punching shear strength of prestressed concrete flat plates the punching shear strength where there is no moment transfer to the
column pure shear in spandrel beam at ultimate
the horizontal dimensions for the rectangular perimeter of the rectangle defined by the centers of the four longitudinal comer bars of the spandrel
beam the vertical dimensions for the rectangular perimeter of the rectangle defined by the centers of the four longitudinal comer bars of the spandrel
beam the distance from the centroid of the critical section to the point where
VJTOZX acts
the depth of the centroidal axis of the uncracked transformed section the distance of the centroidal axis to the extreme tensile fibre
the product of the longitudinal steel ratio
the sum in radians of the absolute values of all successive angular
deviations of the tendon over the length Lpa the transverse steel ratio
the ratio of the long side to the short side of the loaded area
an angular deviation or wobble term which depends on the duct diameter
the spandrel strength parameter
the elastic concrete strain
the extreme fibre compressive strain at ultimate moment
strain of the non-prestressed tensile steel
strain measurement of the prestressed tendons
elastic strain in the prestressing tendon
xxv
Notation
^sc
<5cp
Gpu
'sc
'scu
'su
concrete strain at tendon level
the strain of tendon at ultimate load condition
g = the strain in the compressive steel reinforcement
g = the strain of non-prestressed tensile steel at ultimate
O y . 0 2 = residual proportion of prestress after deducting friction losses
0 L = the diameter of longitudinal steel bar
0 T = the diameter of closed tie
7 = the column width factor
yv = the portion of the unbalanced m o m e n t transferred to the column
7" = the ratio of the depth of the idealised rectangular compressive stress block to the depth of the neutral axis at the ultimate strength in bending
xjjf = the slab restraining factor
X = the slab thickness factor
jl = a friction curvature coefficient which depends on the type of duct
v c = a permissible shear stress carried by the concrete
* = the m a x i m u m shear stress at the critical section V 9p = an angle of inclination of the prestressing tendon
01 = an angle of inclination of the prestressing tendon in the main bending
direction
02 = an angle of inclination of the prestressing tendon in the transverse bending
direction
p = the percentage of tension steel of the slab in the direction perpendicular to
the edge cr = the normal compressive stress
Gj = principal tensile stress
<j, = the tensile stress of the section due to the applied prestress
the concrete stress at the tendon level
the ultimate stresses in the bonded tendons
cr = the stress in the compressive reinforcement
cr = the ultimate stresses in the compressive steel
cr = the ultimate stresses in the non-prestressed tensile
cr = ratio of the effective tendon prestress to the ultimate strength of tendons pe
1 = shear stress
coo = the additional transverse strength of spandrel beam due to restraining
effects of the slab tan 00 = friction loss per unit length in tendon
£,,£.- = elastic prestress losses
xxvi
Chapter I : Introduction
CHAPTER 1
INTRODUCTION
1.1 General Remarks
Flat plates are slab structures without drop panels and column capitals at the column and
slab connections. From an aesthetic and economic point of view, the flat plate structure
has the edge over other forms of slab systems due to the significant savings in
construction costs and an aesthetically pleasing sight from the slab soffit. In addition to
these, the elimination of beams and girders reduces the clear spacing of multi-storey
buildings thus creating additional floor space for a given building height. For these
reasons, flat plates are widely used for multi-storey structures such as office buildings
and car-parks. The structural capacity of the slab-column connection can be enhanced
by the introduction of spandrel beams spanning the free edges of the slab.
Besides being normally reinforced, flat plates may also be prestressed. Post-tensioned
concrete floors form a majority of all prestressed concrete construction and are
economically competitive with reinforced concrete slab structures in the popular medium
to long-span construction range.
Prestressing overcomes many of the disadvantages associated with reinforced concrete
slabs. Deflection, which is usually a major design consideration, is better controlled in
post-tensioned slabs. B y a careful control of cambers by prestressing, "deflection" may
be reduced substantially or eliminated altogether. Besides controlling the deflection,
prestress also inhibits cracking and may be used to produce near crack-free floor
systems. Prestress has also been regarded as one of the options to improve the punching
shear capacity of concrete slabs.
1
Chapter I : Introduction
Punching shear failure is referred to the local shear failure which would occur around the
loaded area or the column support. It is one of the most critical considerations when
determining the thickness of flat plates at the column-slab intersection.
This study is focused on the reliability of the various methods of predicting the punching
shear strength, Vu , of reinforced concrete flat plates. In addition, it emphasises on the
development of a reliable prediction method for Vu of post-tensioned concrete flat plates
with or without spandrel beams.
1.2 The Prediction of Vu
The analysis of the punching shear strength, Vu , of the slab-column-spandrel
connections of reinforced and post-tensioned concrete flat plates, at the edge- and
comer-column positions, has been closely studied by engineering academics and
professionals in recent years. In so far as the development of a reliable method of
predicting Vu for a slab-column-spandrel connection is concerned, several analytical
methods have been developed and proposed for reinforced concrete flat plates.
Unfortunately, most of these methods do not encompass the effects of prestress in the
determination of Vu for post-tensioned concrete flat plates. Those which do seem not to
work satisfactorily. Therefore a detailed study is required to be undertaken so as to
develop a reliable analytical method for post-tensioned concrete flat plates. A general
analytical method for Vu , is said to be reliable if it can predict both the punching shear
strength of the connections as well as the accompanying failure mechanisms.
The problem associated with predicting the punching shear strength Vu , may be
summarised by posing the following questions:
• How would the strength of the spandrel beams affect the magnitude of Vul
2
Chapter I : Introduction
• H o w would the characteristic compressive strength of concrete affect the
failure mechanism and the punching shear strength at the edge- and comer-
column positions?
• What are the effects of the spandrel strength on the mechanisms of failure?
• Would the most suitable critical shear perimeter be the same as for reinforced
concrete flat plates?
• How to incorporate the effect of prestress into the governing equilibrium
equations?
• Would the interaction of torsion, bending and shear in the spandrel beam in the
vicinity of the connection be similar for both reinforced and post-tensioned
concrete flat plates?
1.3 Aims of Research
The purpose of this research is to study the structural behaviour of post-tensioned
concrete flat plates with or without spandrel beam, with regard to their punching shear
strength capabilities. This is done by testing half-scale post-tensioned concrete flat plate
models to failure under uniformly distributed loads. From the results of these laboratory
experiments, essential data may be obtained for establishing an accurate prediction
procedure for the punching shear strength of post-tensioned flat plates.
An analytical procedure is to be developed for the evaluation of the punching shear
strength of corner and edge column-slab connections of post-tensioned concrete flat
plates with spandrel beams or torsion strips. The proposed procedure is an extension to
3
Chapter I : Introduction
an earlier analytical approach developed by Loo and Falamaki [1992] of which the scope
of research was focused on reinforced concrete flat plates with spandrel beams or
torsion strips. This method is also referred to as the "Wollongong Approach".
In addition, a separate comparative study is also made on the accuracy and reliability of
the Wollongong Approach in predicting the punching shear strength of reinforced
concrete flat plates. In this study, the Wollongong Approach is compared to three
national concrete codes of practice, namely AS 3600-1988 [SAA, 1988], ACI318-89
[ACI, 1989] andBS 8110:1985 [BSI, 1985].
Based on the Wollongong Approach, a parametric study is made on how the spandrel
beam in terms of width, depth and amount of longitudinal steel reinforcement, affects the
spandrel strength parameter. This is used in turn to study the effects of the spandrel
beam on the Wollongong Approach in determining the punching shear strength of
reinforced concrete flat plates.
From the collected experimental data on post-tensioned concrete flat plates, a study is
also carried out to incorporate the effect of prestressing into the above analytical
procedure to evaluate the punching shear strength of post-tensioned concrete flat plates
with or without spandrel beams.
1.4 Scope of Research and Experimental Work
The research is based mainly on the analytical procedure for punching shear strength of
reinforced concrete flat plates detailed in the Wollongong Approach [Loo and Falamaki,
1992]. This is aided by experimental work conducted on post-tensioned concrete flat
plate models with spandrel beams.
4
Chapter I : Introduction
The flat plate models are constructed to a half-scale representation of a typical flat plate
floor structure. The continuity of each slab model is closely simulated by having two
panels supported by six columns. The focus of this study is on the comer and edge
connections of the models which were of post-tensioned concrete construction with
bonded tendons.
The models are closely studied in their failure mechanisms in punching shear at the
column-slab and column-slab-beam connections. The deflections at every loading stage
up to failure and the crack patterns of the models are also studied and discussed.
The plan view of the half-scale flat plate floor system is presented in Fig. 1.1 whereas the
experimental flat plate models are illustrated in Fig. 1.2 . In Fig. 1.2 are shown the
loading points which simulate the uniformly distributed loads and the locations where
vertical deflections are measured.
5
Chapter 1: Introduction
rt -e-• • • •
• • • •
• • • •
• • • i "
• F B G - H • • r
.A _B ,
2700 -««« m>-
2700 -* m~-
,c
m
rI
1
2700 - ^ tm-
.D
2701 -mt
1
•
•
Jj
F,
3 % * - •
r O in so CN
O in VO CN
O ' in VO CN
o l
m vo CN
in vo CN
1'
o3 I
Cuj
O • i—i ut—>
o CD
00
Fig. 1.1 - Typical half-scale flat plate floor
6
Chapter I : Introduction
Section a-a
.80
80 J t160
.80 U22JT-20
Models Pi and P4
Model P2
r
OO <N
U 2
Column
400 x550
7
_ 400 x400
-H250H* T Model P3
6
NQ
2
Column H H
200 x 400 -
5
3
•400x 300
o m
1 o uO
B|-400x 300
H H |
• EL
Spandrel beam $ or
torsion strip
o
725. 1325 1325
3525
200 x -300 •
o o r-
im
Whiffle-tree loading arrangement at the bottom of the slab
^ II ' II
t
Legend
4 Locations for deflection • measurement
gj Loading points for testing
CS Column strips
Note: All measurements are in milimetres
Downward direction of applied loadings
Fig. 1.2-Test models
7
Chapter 1: Introduction
1.5 Theoretical Background
1.5.1 Punching shear
In the design of slabs and other similar stmctures such as footings, the shear strength
often determines the required thickness of the member, especially in the vicinity of a
concentrated load or a column support [Gilbert and Mickleborough, 1990].
Consider a column-slab connection shown in Fig. 1.3 . Shear failure may occur on one
of two critical sections. The slab may act essentially as a wide beam and shear failure
may occur across the entire width of the member, as illustrated in Fig. 1.3a . This is
beam-type shear (or one-way shear) failure and the shear strength calculations are
similar to that of a beam. The critical section for this type of shear failure is usually
assumed to be located at a distance equal to the effective depth of the slab, from the face
of the column or concentrated load. Beam-type shear is often critical for footings but
rarely causes concern in the design of floor slabs.
An alternative type of shear failure may occur in the vicinity of a concentrated load or
column. This is illustrated in Fig. 1.3b . The slab fails in a local area around the column,
which forms a truncated cone or pyramid caused by the critical diagonal tension crack
around the column [Park and Gamble, 1982]. This is known as punching shear failure
(or two-way shear failure). The mechanism of punching shear failure may be described
as follows.
The first crack to form in the slab when subjected to applied load, is a roughly circular
tangential crack around the perimeter of the column due to negative bending moments
in the radial direction. This is followed by radial cracks (due to negative bending
moments in the tangential direction) which extend from the perimeter of the column (see
8
Chapter 1 : Introduction
Fig. 1.3b). These diagonal tension cracks tend to originate near the mid-depth of the
slab, and are therefore more similar to web-shear cracks than flexure-shear cracks. The
stiffness of the slab surrounding the cracked region tends to preserve the shear transfer
to the column by aggregate interlock at higher loads. Punching shear failure may occur
eventually, when the applied load is high enough to overcome the slab stiffness in the
vicinity of the cracked region.
u.uuiu.un
Assumed critical section t
z
(a) Beam-type shear failure
Assumed critical section
PWWi t
Section
Actual failure mode
Radial cracks
Tangential cracks
(b) Punching shear failure
Fig. 1.3 - Types of shear failures in a slab
Chapter 1: Introduction
The critical section assumed in some concrete codes [SAA, 1988 and ACI, 1989], and
the cracks that occur in the concrete in the actual failure mode, are shown in Fig. 1.3b .
The critical section for punching shear is usually taken to be geometrically similar to the
loaded area and located at a distance half the effective depth away from the face of the
loaded area [SAA, 1988 and ACI, 1989], although the distance may be specified as 1.5
times the effective depth [BSI, 1985]. The critical section (or failure interface) is
assumed to be perpendicular to the plane of the slab.
1.5.2 Theoretical aspect
The theoretical aspect of this research is aimed at developing a prediction procedure for
the punching shear in post-tensioned concrete flat plate. Following the work by Loo and
Falamaki [1992], there are three groups of undetermined forces at the column-slab-
spandrel connection in a flat plate. These forces are shown acting on free-body
diagrams in Fig. 1.4 and are briefly described below.
(a) Mj and V2
These are internal actions to be measured experimentally and for which semi-empirical
formulas have to be developed. They are the negative yield moment of the slab and the
shear force respectively, over the front segment of the critical perimeter.
(b) MCi and MC2
These are actions to be computed using known structural analysis procedures. They are
the total unbalanced moments transferred to the column centre in the main and
transverse moment directions respectively.
10
Chapter I : Introduction
Spandrel beam
Column
Critical perimeter
Spandrel beam
Column
Critical perimeter
(a) Corner connection (b) Edge connection
(c) Typical section
Fig. 1.4 - Free body diagrams for edge and corner connections
3 0009 03132123 0
11
Chapter 1 : Introduction
(c) M2,V2,T2, and Vu
These are actions to be determined in conjunction with (a) and (b) using the three
equilibrium equations plus an interaction equation for torsion, shear and moment acting
on the spandrel or torsion strip [Loo and Falamaki, 1992].
Note that M2 , V2 and T2 are the negative yield moment of the slab, shear force and
torsional moment respectively over the side segment of the critical perimeter. Finally, Vu
is the punching shear strength of the slab.
In order to establish the interaction equation in (c), the following parameters, ©0 and \\i
have been determined previously [Loo and Falamaki, 1992]. While co0 is the additional
transverse strength of spandrel beam (if any) due to restraining effects of the slab, \\i is
the slab restraining factor.
In order to fully develop the prediction procedure for the punching shear strength Vu ,
semi-empirical formulas have to be established for M;, V}, and \\i .
The main objective of the ultimate load tests of a series of nine half-scale reinforced
concrete flat plate models undertaken by Falamaki and Loo [1992] was to provide
experimental data for setting up these semi-empirical formulas. This was similarly
carried out for post-tensioned concrete flat plate models, in which the semi-empirical
formulas for \i7 was re-calibrated to take the effect of prestress into consideration in the
determination of Vu .
Further discussion is undertaken in Chapter 5 on how to incorporate the effects of
prestress into the punching shear prediction approach introduced by Loo and Falamaki
[1992],
12
Chapter 1 : Introduction
1.6 Overall Layout of Thesis
Following the introduction presented in this chapter, Chapter 2 gives a review of current
and past literature relevant to the research on punching shear, Vu , of reinforced as well
as post-tensioned concrete flat plates with or without spandrel beams. Based on
published results, a comparative study is made in Chapter 3 on the performance and
reliability of the current methods of analysing Vu . The methods considered are the
Australian Standard AS 3600-1988 [SAA,1988], the American Concrete Institute
ACI318-89 [ACI,1989], the British Standard BS 8110:1985 [BSI, 1985] and the
Wollongong Approach [Loo and Falamaki, 1992].
A parametric study is carried out in Chapter 4 on the effects of varying some variables of
the spandrel beam, on the spandrel strength parameter, 8 . This is undertaken as a step
towards determining the effects on the punching shear, Vu (in the Wollongong
Approach) for reinforced concrete flat plates. From this study, the limits of the
parameter, 8 , are defined for the applicability of the prediction procedure for analysing
Vu-
In Chapter 5, an overview is made as to how the effects of prestress are included in the
analysis of Vu of post-tensioned concrete flat plates by the major codes of practice
mentioned above. A theoretical approach that incorporates the effects of prestress into
the Wollongong Approach is also introduced. The effects are taken into consideration in
the three basic equilibrium equations and the interaction equation for the forces acting at
the column-slab-spandrel connections. In this chapter, the semi-empirical equations
from the Wollongong Approach are also described and discussed with emphasis on
applicability to post-tensioned concrete flat plate models.
13
Chapter I : Introduction
The overall experimental approach and the construction procedure for the post-
tensioned concrete flat plate models are described fully in Chapter 6 and Chapter 7
respectively. The test results obtained for all the post-tensioned concrete flat plate
models are presented in Chapter 8. These include crack propagation, the deflections and
the measured punching shear strengths.
The determination of the ultimate flexural strengths (based on the measured strains of
steel and tendons) of the critical slab strips [Falamaki and Loo, 1992] of post-tensioned
concrete flat plates are shown in Chapter 9. In this chapter, a comparison is also made
between the measured and calculated moments at ultimate for the critical slab strips.
From this comparison, the reliability of using the measured moments in analysing Vu , is
determined.
The semi-empirical formulas to determine the slab restraining factor, \j/ for the prediction
of Vu , in post-tensioned concrete flat plates are calibrated in Chapter 10, using the
method employed by Falamaki [1990]. Using these calibrated formulas and the available
test data, the punching shear strength, Vu for both comer- and edge-column
connections are determined in Chapter 11. All the predicted Vu results using the
proposed approach (or the Modified Wollongong Approach) for post-tensioned
concrete flat plates are then compared with the experimental results to determine the
accuracy of the proposed prediction procedure. In this chapter, comparisons are also
made on the reliability of the analytical methods recommended by the Australian,
American and British concrete codes for the punching shear strength of post-tensioned
concrete flat plates.
Finally, the conclusions on the outcome of the theoretical and experimental aspects of
the study of the punching shear strength of flat plates are presented in Chapter 12
together with recommendations for further study.
14
Chapter 2 : Literature Review
CHAPTER 2
LITERATURE REVIEW
2.1 Background of Previous Work
The earliest code recommendation on flat slabs was an empirical design method included
in the 1920 ACI Code. Some designers viewed this empirical approach with scepticism
as it may not have been based on valid experimental results, although some revisions of
this method have been included in subsequent codes in many countries including the
previous British Code CP110:1972.
In order to overcome the limitations on the applicability of such empirical methods, an
alternative approach to slab design was developed and introduced in the Californian
Code of 1933. With some minor modifications, it became the now widely known
equivalent frame method which was incorporated into the ACI codes from 1941 to
1963.
A major reconsideration of slab design was undertaken in America in the late fifties and
early sixties. A series of small-scale slabs, each comprising nine 1.5 m x 1.5 m square
panels, was tested at the University of Illinois [Sozen and Siess, 1963]. It included a flat
slab without column capitals or drop panels i.e. a flat plate, but with beams at all free
edges and two flat slabs with capitals and drop panels. In addition, a larger scale flat
slab without capitals or drop panels was also tested by the Portland Cement Association
[Guralnick and LaFraugh, 1963].
These experimental studies involved careful and thorough instrumentation, which
together with complementary analytical studies, gave rise to the improved equivalent
15
Chapter 2 : Literature Review
frame method of design used in the following and current A C I code [ACI, 1989]. This
approach unifies the design of all two-way slab systems and in respect of flat slabs,
introduces an allowance for the relative flexibility of the slab-column connections.
This research was a significant step towards an improved slab design approach although
since then further tests on multi-panel slabs have been undertaken with no significant
advances in the design methods.
Throughout the period of the development of flat slabs, there has been a secondary type
of experimental work concerned with local conditions at the slab-column junctions and
aimed largely at establishing expressions for resistance to punching shear. Thirty-nine
reinforced concrete slab models of 1.8m x 1.8m size were tested at the University of
Illinois by Elstner and Hognestad [1956], from which 34 models failed in punching
shear. The test specimens have generally been intended to represent regions of slabs
around individual columns bounded by lines of radial contraflexure. The reduction of
the area of slab to be modelled has enabled the scale of these specimens to be at times
much larger than that of most multi-panel work.
Originally, attention was focused on internal columns loaded symmetrically, but more
recent studies have included internal columns (subjected to unsymmetrical loading), edge
columns and comer columns [Zaghlool et al, 1970]. The scope of work has also
extended to prestressed slabs [Brotchie and Beresford, 1967, Bums and Hemakom,
1977, Regan, 1985, and Foutch, Gamble and Sunidja, 1990] and lightweight slabs
[Beresford and Blakey, 1963]. Tests of these types are the basis for the code of practice
equations for punching shear resistance. These equations have been amended and
extended as more data became available. As an example of this process, AS 3600-1988
[SAA, 1988] is the first Australian code to include provisions for spandrel beams in flat
plates.
16
Chapter 2 : Literature Review
Until very recently, there have been two separate strands of development: tests of multi-
panel structures and tests of isolated slab-column connections.
The former have the disadvantage of requiring a very large amount of work in order to
obtain essentially a single test result in terms of ultimate load behaviour. Testing
techniques are unavoidably complicated if distributions of moments in the slabs and
reactions and moments in the columns are to be determined with any degree of accuracy.
The specimens are either relatively large or built to such a small scale that there must be
doubts as to the quantitative significance of the results obtained.
The latter type of specimen has the drawback of having arbitrary boundary conditions
and may not represent the behaviour of a similar region in a real structure, where lines of
contraflexure can shift as the loading varies. In all cases where unbalanced moments are
transmitted from the slab to a column, there is a possibility that the specimen may fail in
flexure when the moment resistance is fully utilised. On the other hand, in a real flat-
plate, the moment might remain constant as the shear increased with further loading.
Due to these limitations inherent in past testings, since then there have been several
research projects using fairly large scale specimens (of size ranging from 3m x 1.8m to
3m x 3m), which may be termed "partial structures". Slabs supported by four corner
columns have been tested by Zaghlool et al [1970] and Ingvarsson [1974]. Kinnunen
[1971] and Long et al [1978] have used specimens supported by two opposite edge
columns or by one edge and one internal column. Models of internal regions extending
from a column to lines of maximum span moments have also been tested by Vanderbilt
[1972] and Long and Masterson [1974].
A series of laboratory tests were conducted by Rangan and Hall [1983 a] and Rangan
[1987] on half-scale models of edge panels of a reinforced concrete flat plate floor.
17
Chapter 2 : Literature Review
These were the basis for the punching shear design provisions contained in the
Australian Standard for Concrete Stmctures, AS 3600-1988 [SAA, 1988] for reinforced
as well as prestressed concrete flat plates.
2.2 Codes of Practice
There are existing codes which cover the methods of analysis of the punching shear in
reinforced as well as post-tensioned concrete flat plate structures. Among them are the
Australian Standard AS 3600-1988 Concrete Structures [SAA, 1988] and the American
Concrete Institute ACI 318-89 [ACI, 1989].
The punching shear strength analysis and design of reinforced and prestressed concrete
flat plates with spandrel beams are not included in the American Concrete Institute code
of practice [ACI, 1989]. The current Australian Standard AS 3600-1988 [SAA, 1988]
provides a method for computing the punching shear strength of both reinforced and
prestressed concrete flat plates with torsion strips and spandrel beams. The method for
prestressed concrete flat plate is a simplistic extension of the procedure developed for
reinforced concrete structures. This was done by including the nominal prestressed
concrete stress in the relevant equations. The accuracy and reliability of these equations
have not been verified although the basic procedure for reinforced concrete flat plates
was shown to be at times grossly inadequate [Loo and Falamaki, 1992].
Other national code of practice for concrete stmctures also reviewed herein is the British
Standard BS 8110:1985 [BSI, 1985]. The recommendations for the method of
determining the shear strength of prestressed beams are also applicable to prestressed
slabs in the British Standard method, whereby consideration is given to uncracked or
cracked sections in flexure. An analytical method recently developed by Loo and
18
Chapter 2 : Literature Review
Falamaki [1992] for predicting the punching shear of reinforced concrete flat plates is
also discussed along with other relevant literature on this topic.
Experimental work on post-tensioned concrete flat plates undertaken by researchers
which have been published in Australia, United Kingdom and North America such as
ACI Stmctural Journal are reviewed in Section 2.4 . The models that were tested and
the reported findings of their stmctural behaviour at failure loads are also discussed in
Section 2.5 .
2.3 Methods of Punching Shear Strength Analysis
2.3.1 Australian Standard AS 3600-1988 [SAA, 1988]
The provisions for punching shear in the Australian Code have been developed from the
results of a series of laboratory tests conducted by Rangan and Hall [1983a] on half-
scale reinforced concrete edge column-slab specimens.
The extension of Rangan and Hall's proposals to cover prestressed concrete slabs has
been described as logical and simple [Gilbert and Mickleborough, 1990]. The
procedures which are based on a simple model of the slab-column connection will be
discussed in more details in Chapter 3 (Section 3.3.1) for reinforced concrete slabs, and
in Chapter 5 (Section 5.1.2a) for prestressed concrete flat plates.
2.3.2 American Concrete Institute ACI318-89 [ACI, 1989]
For prestressed concrete slabs under combined bending and shear, the ACI code
recommends an analytical method which is a modified form of the procedures for
19
Chapter 2 : Literature Review
reinforced concrete slabs. These procedures are based on the proposals of Di Stasio
and Van Buren [I960].
Gilbert and Mickleborough [1990] stated that although the ACI approach provides
reasonable accuracy with test results [Hawkins, 1974], it is essentially a linear working
stress design method and has fallen from favour over the past 20 years. Such
approaches are described as irrational, too cumbersome and difficult to use, particularly
in the case of edge and comer columns both with or without spandrel beams.
However, it is noted that the ACI method does not specify any provisions for the
presence of spandrel beams in reinforced and prestressed concrete slabs.
2.3.3 British Standard BS8110 1985 [BSI, 1985]
In the current British code, the effect of prestressing on punching shear is not clearly
specified for flat plates although it mentioned that the approach is similar to method of
determining shear resistance for prestressed concrete beams.
2.3.4 The Wollongong Approach [Loo and Falamaki, 1992]
The scope of the Wollongong Approach did not include the effect of prestress in the
determination of the punching shear strength of flat plates. One of the aims of this study
is to incorporate the prestressing effect into this approach.
20
Chapter 2 : Literature Review
2.3.5 Others
The European CEB-FIP Model Code [1978] has provision for determining the punching
shear at a distance of half the effective depth from the face of the loading area. This is
consistent with the Australian and the ACI Codes.
However, the formula recommended can only be used for circular columns with
diameter not exceeding 3.5 times the effective depth. It may not be applicable for
rectangular columns with perimeter exceeding 11 times the effective depth and with a
ratio of length to breadth which exceeds 2.0. It considers the prestress as a permanent
compressive force whereby the structure is then treated as a reinforced section subject to
an applied axial force to enhance its shear resistance.
Some other researchers have developed formulas for calculating the punching shear of
flat plates. Most of them are based on the existing codes of practice. For example,
Regan [1981] has developed an expression which is similar to the British Standard but
with the critical perimeter located 1.25 times the effective depth away from the column
faces. In his approach, the influence of prestress on the punching shear strength is
accounted for by adding the decompression load to the punching resistance of a
geometrically similar slab without prestress. The decompression load is the load required
to annul the effect of the prestress, i.e. to give a zero stress at the extreme fibres at the
top of the slab [Regan, 1985].
Chana [1991] suggested that the simplest method of introducing the effect of prestress
on the punching shear resistance is through the use of an effective reinforcement ratio.
The actual ratio of prestressed reinforcement is converted to an equivalent ratio of
ordinary steel reinforcement in the calculation for punching shear strength. This is
similar to the original proposal by the Concrete Society [1979], whose approach has
21
Chapter 2 : Literature Review
been described as "optimistic" as it considers that the applied prestress can still achieve
its ultimate state prior to the punching shear failure of the slab.
2.4 Previous Experimental Work on Post-Tensioned Concrete Flat Plates
2.4.1 J.F. Brotchie and F.D. Beresford [1967]
Brotchie and Beresford [1967] tested a large scale ( 8 m x 14 m) flat plate model with
15 columns (5x3 arrangement) at 3 m spacing. The slab was prestressed by draped
unbonded tendons without having any non-stressed steel reinforcement.
The model was tested to failure under uniformly distributed load and sustained a
flexural failure. The shear capacity of the model was sufficient to prevent a punching
shear failure.
2.4.2 N.H. Burns and R. Hemakom [1977]
Bums and Hemakom tested a one-third scale flat plate model having 16 columns (4x4
arrangement). It was also prestressed by unbonded tendons. The physical behaviour of
the model was observed up to the point of collapse.
Punching shear failure occurred only at the interior columns whereas the exterior
columns failed by developing positive and negative yield lines near the supports and
midspan respectively. Extensive cracks were recorded around the interior columns but
not around the exterior columns.
22
Chapter 2 : Literature Review
The authors concluded that the shear strength of the slab was affected by the concrete
compressive stress (P/A), the amount of bonded reinforcement and the tendon
arrangement.
2.4.3 P.E. Regan [1985]
Regan [1985] carried out and reported fifteen tests of bonded post-tensioned concrete
slabs of 180 mm to 225 mm thick, and of 3 m x 1.5 m size. These models were
prestressed only in one direction and normally reinforced in the transverse direction.
Concentrated upward loads were applied through steel plates located at the centre of the
models, while reactions were provided by four tie bars close to the short edges. All but
three of the tests were reported to have failed in punching shear, and that the applied
prestress have contributed to the punching shear strength of the models.
2.4.4 D.A. Foutch, W.L. Gamble and H. Sunidja [1990]
Foutch et al [1990] conducted tests on four two-third scale single-column flat plate
models which were prestressed with unbonded tendons.
Two of their models failed in flexural mode while the other two failed in shear. They
concluded that these tests support the use of prediction formulas for shear stress which
includes a contribution from the effect of prestress for an edge column-slab connection.
The current ACI code only allows the effect of prestress to be taken into account for
interior columns.
23
Chapter 2 : Literature Review
2.4.5 A.E. Long and D.J. Cleland [1993]
Long and Cleland [1993] tested five 14-scale model slabs that closely simulated an
exterior panel for the purpose of studying the failure characteristics of edge panels in
post-tensioned unbonded slabs. All the models are of identical size and supported by
two edge columns but with varying levels and distribution of prestress.
All the models were reported to have failed in a punching shear mode and it was found
that higher prestress levels increase the ultimate punching shear capacity.
2.5 Comments on Previous Experimental Work
The models tested by all the researchers mentioned in Section 2.4 have one common
feature: they were post-tensioned by unbonded tendons, as allowed by the ACI318-89
[1989]. The Australian Standard AS3600-i988 does not recommend unbonded tendons
in post-tensioning. All unbonded tendons have to be subsequently grouted after full
prestressing of the tendons.
Another common oversight by the previous experiments on post-tensioned concrete flat
plates is the effect of having spandrel beams along the edge of the slab on the punching
shear strength of the column-slab-spandrel connections. Further, punching shear failure
behaviour at comer- and edge-column positions was not the main subject of these earlier
studies.
The present thesis research covers both these topics which were neglected by previous
researchers.
24
Chapter 3 : Performance of Analytical Methods
CHAPTER 3
PERFORMANCE OF ANALYTICAL METHODS FOR PUNCHING
SHEAR STRENGTH OF REINFORCED CONCRETE
FLAT PLATES
3.1 General Remarks
The punching shear strength df concrete flat plates is one of the topics of intensive
research in recent years by various concrete stmctures researchers. This chapter
reviews four current methods of analysing the punching shear strength at the corner-
and edge-column positions of reinforced concrete flat plates. They include those
recommended in the Australian Standard AS3600-1988 [SAA, 1988], the American
Concrete Instimte ACI318-89 [ACI, 1989] and the British Standard on Concrete
Practices (BS8110) [BSI, 1985] as well as the approach developed at the University of
Wollongong, Australia [Loo and Falamaki, 1992]. Based on half-scale model test
results, a comparative study of these four analysis methods is made with regard to their
Umitation, accuracy and reliability. It is found that the Wollongong Approach in general
gives the best performance in predicting the punching shear strength of flat plates with
torsion strips and those with spandrel beams. The Australian Standard procedure
performs just as satisfactorily for flat plates with torsion strips but tends to be unsafe for
those with spandrel beams. Both the A C I and the British methods are applicable only to
flat plates with torsion strips; they also tend to give unsafe predictions for the punching
shear strength.
3.2 Objectives and Scope
Various major national codes of practice for concrete stmctures have provisions for the
25
Chapter 3 : Performance of Analytical Methods
analysis of punching shear strength of concrete flat plates. However, there appears to be
no published material available on the relative accuracy and reliability of these methods
of analysis. The relative merits or otherwise of the various methods would be useful for
the design engineers, especially those who practise internationally.
A brief review is given herein of the analytical method recommended in
the Australian Standard AS3600-1988: Concrete Stmctures [Standard
Association of Australian, S A A 1988];
• the American Concrete Institute, Building Code Requirements for Reinforced
Concrete Stmctures A C I 318-89 and Commentary - A C I 318R-89, [ACI,
1989];
the British Standard B S 8110: Part 1, "Code of Practice for Design and
Construction" , [British Standards Institute, BSI 1985].
In addition, an analytical method recently developed by Loo and Falamaki [1992] (the
Wollongong Approach) is also discussed.
Based on- the half-scale model test results published by Rangan and Hall [1983a],
Rangan [1990b] and Falamaki and Loo [1992], the relative accuracy and reliability of all
four methods of analysis are investigated. The comparison is carried out by plotting the
experimental punching shear strength values against the predicted values due to each of
the four methods, where applicable. Pearson correlation coefficients are also obtained to
illustrate the strength of the relationship between each prediction method and the
experimental values.
Other standard statistical indicators such as maximum, average and standard deviations
as well as frequency distribution are also employed to determine the relative
26
Chapter 3 : Performance of Analytical Methods
performance of the four analytical methods in question. The comparative study provides
considerable insight into the relative merits or otherwise of the four methods of analysis.
3.3 Methods of Analysis - A Review
Of the four methods included in this study, the Australian Standard procedure and the
Wollongong Approach cover reinforced concrete flat plates with spandrel beams or
torsion strips whereas the A C I and the British methods are limited to cases with free
edges only (i.e. torsion strips without closed ties). These methods are briefly described
in the following sections.
3.3.1 Australian Standard Procedure [SAA, 1988]
The Australian Standard procedure is mainly the work of Rangan [1990a]. It is based
heavily on the experimental study conducted by Rangan and Hall [1983a] on half-scale
models of the edge panels of flat plate floors. According to Rangan [1987], the
equations developed are deemed to consider the following cases :
(a) a slab in the vicinity of an interior column, without spandrel beams and closed
ties;
(b) a slab in the vicinity of an edge column, with a spandrel beam on the side
provided with minimum closed ties;
(c) a heavily loaded slab in the vicinity of an edge column, with a spandrel beam as
in part (b) above but having more than the minimum closed ties;
(d) a slab in the vicinity of an edge column, without spandrel beam and closed ties,
i.e. a free edge.
The analysis procedure for an edge-column connection is assumed valid for a corner-
27
Chapter 3 : Performance of Analytical Methods
column connection.
When the moment to be transferred is zero, the equation for punching shear only
incorporates four parameters, namely the critical perimeter, the effective depth of slab,
the column dimensions and the compressive strength of concrete. The critical shear
perimeter is defined as the (failure) section half the effective depth away from the faces
of the column.
The punching shear strength where there is no moment transfer is given as:
Vuo = udfcv ...(3-1)
where u is the critical shear perimeter, d is the effective depth of the slab and fcv is the
Umiting concrete shear stress for which relevant formulas are given in the Australian
Standard.
When there is moment transfer to the column, in addition to the above-mentioned
parameters, the equation also considers the total unbalanced moment, and the shear
force to be transferred to the column support.
In the presence of spandrel beam and closed ties, other parameters taken into account
include the dimensions and yield strength of closed ties, as well as the area and perimeter
defined by the longitudinal bars at the four comers of the closed ties.
For a typical slab with moment transfer to the column, and having a spandrel beam with
the minimum requirement of closed ties, the punching shear strength is given as:
28
Chapter 3 : Performance of Analytical Methods
1-2 [ ] Vm Vu = —* -C
3"2) M u
[ 7 + — ^ ] 2 y*abw
where (DiJDs) is the ratio of the depths of the beam and the slab; M and V are
respectively the design bending moment and shear force at the column; a is the segment
length of the critical shear perimeter parallel to the vector direction of My and bw is the
width of the spandrel beam.
This approach has been derived based on two assumptions [Rangan, 1987] :
(a) the distribution of the shear stress is taken to be uniform along the critical
perimeter;
(b) for both the edge-column and comer-column positions, a constant slab
restraining factor of 4 is adopted for the slab-column connections with spandrel
beams or torsion strips with closed ties.
Some shortcomings of these assumptions have been identified and discussed by Loo and
Falamaki [1992]. For further details of this method, the reader is referred to the
Australian Standard [SAA, 1988].
3.3.2 The ACI Method [ACI, 1989]
The method recommended in the ACI code for punching shear design under combined
shear and moment is mainly based on the work of Di Stasio and Van Buren [I960]. In
this method, the moment is assumed to be transferred to the column by flexure of the
slab and by torsion of the slab edge strip or torsion strip [Cope and Clark, 1984]. Also,
29
Chapter 3 : Performance of Analytical Methods
the moment is transferred along a critical section around the column, similar to the
Australian Standard procedure, over which the punching shear failure is assumed to
occur.
In the presence of an unbalanced moment, consideration is given to the effect of torsion
by applying an analogous polar moment of inertia of the critical section, and a calculated
fraction of the net moment to be transferred to the column by this torsion. The
maximum shear stress at the critical section is expressed as:
* = r JvMly (3_3) A J
where A is the area of the critical section, / is the analogous polar moment of inertia,
Y v is the portion of the unbalanced moment transferred to the column and y is the
distance from the centroid of the critical section to the point where Vmax acts.
In addition, the ACI method stipulates that the maximum shear stress vL* is not to
exceed a permissible shear stress, vc carried by the concrete which is:
vc = ( 0.166 + ^-)ifc (< 0.33 VZ) -O-4)
where Pc is the ratio of the long side to the short side of the loaded area.
Note that in this approach, the effects of the torsion strips with or without closed ties
are taken into consideration but those of the spandrel beams cannot be included.
Furthermore, according to ACI-ASCE Committee 426 [1977], the predicted results are
safe where the ratio of the long and short sides of a rectangular column or loaded area
30
Chapter 3 : Performance of Analytical Methods
is higher than 2.0 .
In this simplistic approach, the main parameters are the compressive strength of concrete
and the dimensions of the critical perimeter. It does not take other factors into account
such as the slab restraining effect and the transverse strength of the torsion strip or the
spandrel beam.
3.3.3 The British Standard Method [BSI, 1985]
The general approach in determining the punching shear strength recommended in
BS8110:1985 is similar to the previous British Concrete Code CP 110:1972 [BSI,
1972]. The values for the design nominal shear stresses in CP110 are based on the
recommendations of the Institution of Stmctural Engineers Shear Study Group [1969].
As in the Australian Standard, it also assumes that the distribution of shear around the
column is uniform, although this is not the case in general [Loo and Falamaki, 1992].
According to this method, the punching shear stress is given as
^c=C^Tu^f- .-(3-5)
where p is the percentage of tension steel of the slab in the direction perpendicular to the
edge and/CM is the characteristic cube strength of concrete.
The relationship between the compressive cube strength and compressive cylinder
strength is presented in Fig. 3.1. The cylinder strengths are between 76% to 96% of the
corresponding cube strengths [Neville, 1978].
31
Chapter 3 : Performance of Analytical Methods
It is important to note that the compressive cylinder strength results obtained for the
reinforced concrete models tested in Wollongong and the University of New South
Wales are to be converted to their respective compressive cube strengths, before
applying the British Standard method in determining their punching shear strengths.
In contrast to the Australian and the ACI codes, BS8110 takes the shear failure
perimeter to be a rectangle with its sides at 1.5 times the average effective depth from
the column faces. The parameters included in calculating the punching shear are the
steel ratio of slab, effective depth and cube root of the compressive strength of concrete
(where the concrete strength must exceed 25 MPa). Further allowance is made in the
form of a reduction factor of 1.25 to be applied to the calculation in determining the
punching shear strength.
r % OS
% * « — " ^ OO <3 l>
m V
J - >
5» O
> ca m 4J rl PH
a o u
60
SO
40
30
20
10
-
-
-
/
V
c
y /
i
/
f
a/
f
1 1
/
1
0 10 20 30 40 50 60
Compressive cylinder strength (MPa)
Fig. 3.1 - Relative strength of concrete cubes and cylinders
As in the ACI method, the effects of spandrel beam on the punching shear strength
cannot be accounted for in the BS8110 procedure. There are other limitations :
32
Chapter 3 : Performance of Analytical Methods
(a) compressive strength of concrete not to exceed 40 M P a ;
(b) effective depth of slab not to exceed 400 m m ;
(c) main steel ratio of the slab not to exceed 3%.
Similar to the ACI approach, the British method is also considered simplistic in that
factors such as the restraining factor of the slab and the transverse strength of the torsion
strip which are not taken into consideration.
3.3.4 The Wollongong Approach [Loo and Falamaki, 1992]
The Wollongong Approach is the result of a comprehensive research into the punching
shear strength of reinforced concrete flat plates. Applicable to flat plates with spandrel
beams or torsion strips, the approach comprises explicit formulas through which the
punching shear strength at the comer- and edge-column connections can be determined.
A detailed presentation of the method has been made elsewhere [Loo and Falamaki,
1992] and it is not repeated herein. However, it may be noted that the Wollongong
Approach takes full account of the following important parameters and effects that were
neglected or left out in the other three methods :
(a) The slab restraining factor is not a constant but a function of the strength of the
spandrel beam.
(b) The distribution of shear force around the critical perimeter is not uniform but a
function of the slab reinforcement ratio, clear span of the slab, and the size of the
spandrel beam in relation to the slab.
(c) The interaction of torsion, shear and moment in the spandrel beam.
(d) Flat plates with deep spandrels have a bending mode of failure very much
different from those with shallow beams and they are carefully defined.
33
Chapter 3 : Performance of Analytical Methods
3.4 Presentation and Comparison of Results
3.4.1 Model tests
As part of a comprehensive study of the punching shear strength of concrete flat plates,
a total of nine half-scale reinforced concrete models were constmcted and tested to
failure over a period of three years (from 1987 to 1990). Test program of such a scale is
not believed to have been attempted elsewhere previously. Full details of this
experimental work and its results have been reported by Falamaki and Loo [1992]. For
ease of reference, the plan dimensions of the models and a typical cross section of the
slab-column-beam connection are presented in Fig. 3.2a and 3.2b respectively.
-O
Column
Column
F G _
X
A B
H
C
i
\
i
y
725
2650
.
3
l
•««£
<
2700
540C
X
Imm
2700 mm -• >
"••>
(a) Plan
3375
Fig. 3.2 - Flat plate model Wl with a typical spandrel beam section details
(continued next page)
34
Chapter 3 : Performance of Analytical Methods
Longitudinal reinforcement Slab reinforcement
Slab
Closed ties
(nominal reinforcement)
Spandrel beam
(b) Section x-x
Fig. 3.2 - Flat plate model Wl with a typical spandrel beam section details
(continued from previous page)
Table 3.1 - Comparison of punching shear results - models with
torsion strips
Colum
n
W5-A
M5-A
W5-B
M5-B
W5-C
M5-C
R3-A
R4-A
R90-A
R90-B
R90-C
_ * Type
C/NS
C/NS
E/NS
E/NS
C/NS
C/NS
E/NS
E/NS
E/NS
E/NS
E/NS
Mtuasured
\
(kN)
32.59
34.42
71.24
87.81
37.30
38.36
88.76
75.42
120.20
108.20
21.90
A S 3600-1988
Predicted
(kN)
37.01
40.70
53.04
49.79
43.66
42.82
87.23
85.03
114.95
148.35
24.12
Mean:
MeasmssJ Predicted
0.88
0.85
1.34
1.76
0.85
0.89
1.02
0.89
1.05
0.73
0.91
1.015
ACI318-89
Predicted v«
(kN)
57.50
69.95
91.18
144.23
80.22
90.23
130.17
103.84
108.02
136.38
58.58
Measured Predicted
0.57
0.49
0.78
0.61
0.46
0.43
0.68
0.73
1.11
0.79
0.37
0.638
BS8110:1985
Predicted
(kN)
41.20
64.32
80.05
173.66
53.27
77.49
166.60
127.85
130.88
138.43
59.22
Measured Predicted
0.79
0.54
0.89
0.51
0.70
0.50
0.53
0.59
0.92
0.78
0.37
0.647
Wollongong
Predicted V u
(kN)
35.21
29.37
74.62
94.21
35.07
53.91
94.90
83.73
76.60
144.24
23.00
Measured Predicted
0.93
1.17
0.95
0.93
1.06
0.71
0.94
0.90
1.57
0.75
0.95
0.987
C-corner column; E-edge column; NS-without spandrel beam.
35
Chapter 3 : Performance of Analytical Methods
Table 3.2 - Comparison of punching shear results - models with
spandrel beams
Column
R90-D
Wl-A
W2-A
W3-A
W4-A
Wl-B
W2-B
W3-B
W2-C
W3-C
W4-C
M2-A
M3-A
M4-A
M2-B
M3-B
M4-B
M3-C
M4-C
Type**
E/S C/S
C/S
C/S
C/S
E/S
E/S E/S C/S C/S
C/S
C/S C/S
C/S
E/S
E/S
E/S C/S
C/S
Measured
(kN)
36.20 0
50.15
48.08
43.38
47.07
117.63
120.36
93.57
45.17
44.33
46.32
53.90
25.70
58.97
123.22
76.50
130.24
24.30
60.09
AS 3600-1988
Predicted
Vu (kN)
51.71
119.07
120.94
70.42
96.12
146.92
150.05
94.99
113.54
73.38
82.93
82.90
127.31
114.77
116.24
214.32
137.82
131.89
102.75
Mean:
Measured
Predicted
0.70
0.42
0.40
0.62
0.49
0.80
0.80
0.98
0.40
0.60
0.56
0.65
0.20
0.51
1.06
0.36
0.95
0.18
0.58
0.593
Wollongong
Predicted
Vu (kN)
45.89
55.05
58.23
42.91
55.85
148.53
105.87
115.76
44.44
38.36
49.55
35.36
10.59
46.64
98.15
78.07
97.18
32.90
58.58
Measured
Predicted
0.79
0.91
0.83
1.01
0.84
0.79
1.14
0.81
1.02
1.16
0.93
1.52
2.42
1.26
1.26
0.98
1.34
0.74
1.03
1.027
The test to failure of these nine models, each weighing about five tonnes, provides a
total of 24 punching shear strength results, including 16 for comer-column and 8 for
edge-column positions. Also included in the present study are the punching shear
strength results from two similar models tested by Rangan and Hall [1983a] and from
four single edge-column specimens tested by Rangan [1990b]. These make a total of 16
results for comer-column connections and 14 for edge-column connections.
These test results are tabulated in column 3 of Tables 3.1 and 3.2 . Note that the
** S-with spandrel beam. 0 This specimen did not fail in punching; the reported result is the maximum value.
36
Chapter 3 : Performance of Analytical Methods
Wollongong results are those prefixed with the letter W or M and those of Rangan and
associates, with an R. Note that the four results R90-A, B, C and D are for single
column specimens. Out of these 30 results, 11 are for flat plates with torsion strips
while the rest are for those with spandrel beams of various sizes.
3.4.2 Predicted results
The four methods of analysis reviewed herein, namely the Australian Standard
procedure, the ACI method, the British Standard recommendations and the Wollongong
Approach, are used to compute the punching shear strength of the slab-column and slab-
column-beam connections of the models. Remember that the British and the ACI
methods are applicable to flat plates with torsion strips only.
Table 3.1 presents the predicted Vu values by all the four methods for the models with
torsion strips; Table 3.2 contains the predicted results by the Australian Standard
procedure and the Wollongong Approach for the models with spandrel beams. Also
included in the tables are the ratios of the measured and predicted results.
In Fig. 3.3, the measured Vu are plotted against the predicted values for the models with
torsion strips. The corresponding correlation plots for the models with spandrel beams
are given in Fig. 3.4 . Superimposed on these figures are the 30-percentile lines. In
these plots, a correlation point falling below the 45° equality line represents an
overestimate or unsafe prediction. Further, a point falling below the -30% line means
that the analysis method has overestimated the strength by more than 30%. The -30%
line may be taken as a safety criterion to reflect the Australian Standard stipulation of a
shear capacity reduction factor of 0.7 .
37
Chapter 3 : Performance of Analytical Methods
200
+30% Equality (underestimate) ,
-30% a / (overestimate)
Legend A S 3600-1988 ACI318-89 B S 8110:1985 Wollongong
0 50 100 150 200
Predicted V u (kN)
Fig. 3.3 - Measured Vu versus predicted Vu - models with torsion strips
250
200
55
> T3
P ca cd
150
lOO -
50
O
+30% (underestimate)
Equality
-30% (overestimate)
- A,
Legend A S 3600-1988
A Wollongong
1 ' 1 ' I n
O 50 100 150 200 250
Predicted Vu (kN)
Fig. 3.4 - Measured VM versus predicted Vu - models with spandrel beams
38
Chapter 3 : Performance of Analytical Methods
The Pearson correlation coefficients [Nie, Hull and Bent, 1975] for the correlation plots
in Fig. 3.3 and 3.4 are summarised in Tables 3.3 and 3.4 respectively. Note that the
closer the correlation coefficient is to unity, the stronger is the relationship between the
measured and the predicted results.
3.4.3 Comparison: Flat plates with torsion strips
In Table 3.1 for models with torsion strips, it can be seen that with a mean ratio of the
measured and predicted Vu very close to unity, either of the two Australian methods is
superior to the ACI and British methods. The ACI method, on average, overestimates
the punching shear strength by 56.7% * while the British approach overestimates by
54.6% *.
It is apparent in Fig. 3.3 that the predictions by the Wollongong Approach and the
Australian Standard procedure has one correlation point each above the +30% line.
While the former has two predictions below the -30% line, the latter has only one below
the same -30% line. Contrary to these, "the ACI procedure has 8 out of 11 correlation
points falling below the -30% line and the British method has 8 out of 11. The
frequency distributions of these points are presented in Table 3.5 .
A survey of the Pearson correlation coefficients in Table 3.3 leads to the conclusion that
the strength of the relationship between each of the four prediction methods is almost as
high as the other three. However, a reading of the standard deviation values in Table
3.4 shows that both the Australian methods deviate far less than the ACI or the British
l Percentage overestimate for ACI method = ( - 1 ) x 100% = 56.7%
0.638 1 Percentage overestimate for British approach = ( - 1 ) x 100% = 54.6%
0.647
39
Chapter 3 : Performance of Analytical Methods
method. Further, the m a x i m u m and average deviations of both Australian methods are
much smaller than the other two methods.
With a comparatively smaller deviation in addition to a safe and consistent distribution in
their Vu predictions, both the Australian methods are by far superior to those
recommended by the ACI and the British Standard for flat plates with torsion strips.
Table 3.3 - Pearson correlation coefficients for models with torsion strips
Methods of analysis
A S 3600 -1988
ACI318-89
BS 8110:1985
Wollongong
Correlation coefficient
0.868
0.840
0.836
0.860
+ Table 3.4 - Deviations of predicted Vu for models with torsion strips
Methods of analysis
AS 3600 -1988
ACI318-89
BS 8110:1985
Wollongong
Maximum
deviation
- 0.43
+ 1.67
+ 1.70
+ 0.41
Average
deviation
+ 0.04
+ 0.72
+ 0.67
+ 0.05
Standard
deviation
18.18
36.69
44.14
18.19
+ Maximum deviation = the maximum value of (Predicted Vu - Measured V„)/Measured Vu Average deviation = the average value of (Predicted Vu - Measured VJ/Measured Vu
Standard deviation = lit (Predicted Vu - Measured VUY
40
Chapter 3 : Performance of Analytical Methods
Table 3.5 - Frequency distribution of predicted Vu for models
with torsion strips
Methods of
analysis
AS 3600 -1988
ACI318-89
BS 8110:1985
Wollongong
Over-estimates
30 - 60 %
1
8
8 2
0 - 30 %
6 2
3
6
Under-estimates
0-30% 3
1
-
2
30 - 60 %
1
-
-
1
3.4.4 Comparison of A C I and British Standard methods incorporating reduction
factors: Flat plates with torsion strips
The mean ratios of measured and predicted values of Vu for the ACI and British
Standard methods may be applied as reduction factors to bring their respective
predictions of Vu closer to the measured Vu values. For example, the predicted Vu
values by the A C I procedure are to be multiplied by a factor 0.638, whereas the
corresponding predictions by the British Standard method should be multiplied by a
factor of 0.647 . The factored punching shear results for both the A CI and British
Standard methods are shown in Table 3.6 .
With a mean ratio of the measured and predicted Vu very close to unity, both the ACI
and British methods have improved considerably in their prediction of Vu when the
appropriate reducing factors are introduced.
41
Chapter 3 : Performance of Analytical Methods
Table 3.6 - Comparison of factored punching shear results by ACI and
British Standard methods - models with torsion strips
Column
W5-A
M5-A
W5-B M5-B
W5-C
M5-C R3-A
R4-A
R90-A R90-B
R90-C
Type*
C/NS
C/NS
E/NS
E/NS
C/NS
C/NS E/NS
E/NS E/NS
E/NS E/NS
Measured
V„ (kN)
32.59 34.42 71.24
87.81
37.30
38.36 88.76 75.42
120.20
108.20 21.90
ACI318-89
Predicted
V„ (kN)
36.69 44.63
58.17 92.02
51.18
57.57 83.05
66.25 68.92
87.01 37.37
Mean:
Measured
Pr edicted
0.89 0.77
1.22
0.95 0.73
0.67 1.07
1.14 1.74
1.24
0.59 1.001
BS 8110:1985
Predicted
V„ (kN)
27.17 41.62
51.79 112.36 34.47 50.14
107.79 82.72
84.68 89.56 38.32
Measured
Pr edicted
1.20 0.83
1.38
0.78 1.08
0.77 0.82 0.92 1.42
1.21 0.57 0.997
This is clearly illustrated in Fig. 3.5 wherein the resulting predictions by the A C I
procedure has 7 correlation points within the - 3 0 % to + 3 0 % envelope whereas the
British Standard method has 10 correlation points within the same envelope. While the
former has three predictions below the - 3 0 % line, the latter has only one below the same
- 3 0 % line.
A reading of Table 3.7 shows that the strength of the relationship between the A C I and
British methods with the introduction of the reduction factors remains unchanged with
correlation coefficients of 0.840 and 0.836 respectively. The smaller deviations and
narrower frequency distributions of these points (Table 3.8 and 3.9 respectively) indicate
that these modified results obtained are safer and more reliable.
42
Chapter 3 : Performance of Analytical Methods
+30% (underestimate)
150
125 -
55
to
CO CO
5S
100
/ */k. O^ (overestimate)
50 -
Equality
3 0 %
Legend A ACI318-89 n BS 8110:1985
I ' I ' I ' I ' I '
O 25 50 75 100 125 150
Predicted V u (kN)
Fig. 3.5 - Measured Vu versus predicted Vu by A C I and British methods
incorporating reduction factors - models with torsion strips
By incorporating the reduction factors of 0.638 and 0.647 respectively to the ACI
procedure and the British Standard method, their predictions of Vu are more
satisfactory. This means that both the A C I and the British Standard methods may be
confidently used to predict Vu provided the appropriate reduction factors are applied.
Table 3.7 - Pearson correlation coefficients for models with torsion strips
(after incorporating reduction factors)
Metihods of analysis
ACI318-89
B S 8110:1985
Correlation coefficient
0.840
0.836
43
Chapter 3 : Performance of Analytical Methods
Table 3.8 - Deviations of predicted Vu for models with torsion strips
(after incorporating reduction factors)
Methods of analysis
ACI318-89
BS 8110:1985
Maximum
deviation
+ 0.71
+ 0.75
Average
deviation
+ 0.10
+ 0.08
Standard deviation
19.79
17.85
Table 3.9 - Frequency distribution of predicted Vu for models with torsion
strips (after incorporating reduction factors)
Methods of
analysis
ACI318-89
BS 8110:1985
Over-estimates
30 - 60 %
3
1
0 - 30 %
3
5
Under-estimates
0 - 30 %
4
5
30 - 60 %
1
-
3.4.5 Comparison: Flat plates with spandrel beams
For the Wollongong Approach, the mean ratio of the measured and predicted Vu given
in Table 3.2 is 1.027 . This is far more satisfactory than the Australian Standard
procedure which has a mean ratio of 0.593 (indicating that, on average, the procedure
overestimates the punching shear strength of flat plates with spandrel beams by 68.6%).
The superiority of the Wollongong Approach is confirmed in Fig. 3.4 where all the
correlation points except two, lie within the 30 percent lines; the exceptions are
conservative predictions. For the Australian Standard procedure, out of 19 correlation
points, 14 are below the - 3 0 % line. This underscores the unsafe tendency of its
predictions.
44
Chapter 3 : Performance of Analytical Methods
Table 3.10 - Pearson correlation coefficients for models with
spandrel beams
Methods of analysis
AS 3600 - 1988
Wollongong
Correlation coefficient
0.434
0.891
It is obvious in Table 3.10 that the Wollongong Approach correlates well with the test
results. The Pearson correlation coefficient of 0.891 is much higher than that for the
Australian Standard procedure (0.434). From the additional statistical comparisons
shown in Tables 3.11 and 3.12, it can be further concluded that the Wollongong
Approach, with a much smaller deviation and safer distribution in its predictions, is more
accurate and reliable than the Australian Standard procedure in analysing Vu for flat
plates with spandrel beams.
++ Table 3.11 - Deviations of predicted Vu for models with spandrel beams
Methods of analysis
AS 3600 -1988
Wollongong
Maximum
deviation
+ 4.43
-0.59
Average
deviation
+ 1.10
- 0.01
Standard
deviation
60.18
15.43
Table 3.12 - Frequency distribution of predicted Vu for models
with spandrel beams
Methods of
analysis
AS 3600-1988
Wollongong
Over-estimates
30 - 60 %
14
-
0 - 30 %
4
9
Under-estimates
0 - 30 %
1
8
30 - 60 %
-
2
45
Chapter 3 : Performance of Analytical Methods
Falamaki and Loo [1991] have observed that the Australian Standard procedure was
developed without adequate support of model test results. In particular, none of the test
results quoted by Rangan [1987] was for models with spandrel beams. This may be the
main reason for the poor performance by the Australian Standard procedure.
3.5 Conclusions
For the punching shear strength analysis of flat plates at the corner- and edge-column
positions, the Wollongong Approach is found to be accurate and reliable. The
performance is consistent for flat plates with spandrel beams and with torsion strips.
The Australian Standard procedure on the other hand gives accurate and reliable
predictions for flat plates with torsion strips but at times gravely over-estimates the
strength of flat plates with spandrel beams.
Both the ACI and the British Standard methods are limited to flat plates with torsion
strips without closed ties. Further, they tend to give unsafe predictions. On average, the
ACI prediction is 1.57 "times the measured strength and the corresponding value for the
British Standard method is 1.55 times. On the other hand, by applying the reduction
factors of 0.638 and 0.647 respectively to the ACI and the British Standard methods, the
resulting predictions of Vu are more accurate and satisfactory. This means that both the
ACI and the British methods may be confidently used to predict Vu for flat plates with
torsion strips, provided that the appropriate reduction factors are applied.
46
Chapter 4 : Parametric Study of Spandrel Beams
CHAPTER 4
PARAMETRIC STUDY OF THE EFFECT OF SPANDREL BEAMS
ON THE PUNCHING SHEAR STRENGTH OF REINFORCED
CONCRETE FLAT PLATES
4.1 Introduction
This chapter describes a parametric study of the Wollongong Approach [Loo and
Falamaki, 1992] in analysing the punching shear, Vu , of reinforced concrete flat plates
with spandrel beams. This study is focused on how the spandrel strength parameter may
affect the analysis of Vu , using the Wollongong Approach. The investigation is
undertaken by varying the spandrel beam dimensions and the total area of steel
reinforcement in the spandrel beam, keeping other variables constant, such as overall
depth of slab, column size, and'compressive strength of concrete.
It is shown that, in general, the punching shear increases with deeper and wider spandrel
beam but only up to a certain point, beyond which the formulations are not applicable.
The same also applies to the amount of longitudinal steel reinforcement in the spandrel
beam whereby limitations exist, within which the Wollongong Approach formulas may
be applied.
4.2 Formulas for Spandrel Strength Parameter [Falamaki and Loo, 1992]
The spandrel strength parameter as represented by 5 is the product of the longitudinal
steel ratio, a , the transverse steel ratio (3 , and the ratio of dspld where dsp and d are
respectively the effective depths of the spandrel beam and the slab of the reinforced
concrete flat plates. Thus,
47
Chapter 4 : Parametric Study of Spandrel Beams
d o = a p - f ...(4-1)
a
where,
a = ^!lA_ .(4.2) 200000 v J
= C/f[co + co0]
50000
The variables /4/^ and fly are the total area and yield strength of the longitudinal steel
reinforcement in the spandrel beam. Ut refers to the perimeter of the rectangle defined
'by the centers of the four longitudinal corner bars of the spandrel beam (i.e. Ut = 2
[x+y]), where x and y are the horizontal and vertical dimensions respectively for the
rectangular perimeter.
While the transverse spandrel strength parameter is defined as:
co = ^LsL ...(4-4)
its additional transverse strength co0 was provided by the slab restraint for the spandrel
rotation and has been established empirically by Loo and Falamaki [1992] to be a
constant value, or
to0= 446 N / m m •••(4-5)
Note that Aws and fwy are the area and yield strength of closed ties in the spandrel
beam. The derivation of the Eqs. (4-2) and (4-3) for a and P respectively, as well as the
48
Chapter 4 : Parametric Study of Spandrel Beams
definitions for © and co0 are fully explained elsewhere [Falamaki, 1990] and are
elaborated further in Section 5.1.5.
4.3 Generating Models of Varying Features of Spandrel Beam
The generation of models with different characteristics of spandrel beam is undertaken
with the aid of a simple BASIC computer program, specifically written to run, save and
retrieve the various data input. It is also able to provide data output in ASCII format so
as to facilitate data import to other software uses such as LOTUS 123, Cricket Graph
and Microsoft Word for Windows.
Three types of model generation were undertaken, namely, by varying the spandrel beam
width and its corresponding depth as well as varying the amount of longitudinal steel
reinforcement in the spandrel beam. All these variations have the effect of influencing
the spandrel strength parameter 6 , which is then used to study the effect on the
determination of punching shear Vu by the Wollongong Approach.
4.3.1 Varying the width of spandrel beam
The width of spandrel beam (denoted as bsp) may be related to the parameter 5 in Eq.(4-
1) through the variable x , the horizontal dimension of Ut in Eq.(4-3), which in turn is
directly related to the parameter P . Since 5 and P are directly proportional to each
other, the parameter 8 may be used to relate the varying width bsp to the punching
shear, Vu .
The width of the spandrel beam for ten generated models are set by generating a range
of data for bsp as shown
49
Chapter 4 : Parametric Study of Spandrel Beams
bsp.gen = bsp - 10° + i20 ...(4-6)
where bsp,ge„ is the generated width of the spandrel beam, bsp is the spandrel width based
on a given model and / refers to the sequence from 1 to 10 for the ten generated
models. Fig. 4.1 shows ten generated models of varying width of spandrel beam.
All the data for the generated models remains unchanged except for the variable bSp
which in turn will generate a range of the parameter 5 to be used to determine the
punching shear, Vu of the generated models. For example, typical generated data and
results for models based on ModelM2-A are as shown in Table 4.1 .
Longitudinal reinforcement
Spandrel beam Closed ties
Varying width
Column
Fig. 4.1 - Generated models with varying width of spandrel beam
50
Chapter 4 : Parametric Study of Spandrel Beams
Table 4.1 - Typical generated data and results from parametric study of
varying width of spandrel beam in Model M2-A
Generated model (/)
1
2
3
4
5
6
7
8
9
bSp (mm)
140.0
160.0
180.0
200.0
220.0
240.0
260.0
280.0
300.0
a
1.0790
1.0790
1.0790
1.0790
1.0790
1.0790
1.0790
1.0790
1.0790
P 3.0097
3.4589
3.9081
4.3574
4.8066
5.2558.
5.7050
6.1542
6.6034
5
3.4846
4.0047
4.5248
5.0448
5.5649
6.0850
6.6051
7.1252
7.6453
r«(N) -70966
17581
31731
35356
35711
34741
33169
31313
29330
From Table 4.1, it can be seen that for the first generated model with small spandrel
width (i.e. bSp = 140 m m ) , the Wollongong Approach produces a negative Vu . From
the second to fifth model, Vu increases with bsp , after which Vu steadily decreases.
This is clearly illustrated in Fig. 4.2a with a plot of Vu versus 5 .
Three categories of the parametric study of the varying width have been identified which
are dependent on the depth of the spandrel beam or the ratio dsplde . Xi dspld < 1.2 , the
generated models have shallow spandrel beams; if 1.2 < dSpld < 2.5, they are deemed to
be with intermediate depth of spandrel beams; where dspld > 2.5, they are considered to
be with very deep spandrel beams.
Fig. 4.2a shows a plot for punching shear Vu versus spandrel strength parameter 5 for
generated corner-connection models with varying width bSp and shallow spandrel beams
(dspld < 1.2) using selected models M2-A , W3-A , and W3-C. The corresponding plot
for edge-connection models with shallow beams is presented in Fig, 4.2b using models
W3-B and M2-B .
51
Chapter 4 : Parametric Study of Spandrel Beams
V u (N)
100000
50000 -
-500O0
Ocn.M2-A Oen.W3-A Ocn.W3-C
-100000 r—•—| —i j r
2.0 4.0 6.0 8.0
Spandrel strength parameter 5
(a) Corner-connection models
Vu (N) 200000
100000
-100000 -
-200000
-300000
-400000
Oon.M2-B -JII— Gcn.W3-B
1 I ' I ' I ' I ' I ' 3.0 4.0 5.0 6.0 7.0 8.0 9.0
Spandrel strength parameter 5
(b) Edge-connection models
Fig 4.2 - Graph of punching shear Vu versus spandrel strength parameter 5
of generated models with varying width bsp and shallow spandrel beams
52
Chapter 4 : Parametric Study of Spandrel Beams
It may be seen from Fig. 4.2 that the lower limit range of 8 for realistic analysis of Vu in
corner-connection models with shallow spandrel beams is between 3.6 and 3.9.
Similarly, the lower limit range is nearly identical (at between 3.4 and 4.0) for edge-
connection models with shallow spandrel beams.
Fig. 4.3a shows a plot for punching shear Vu versus spandrel strength parameter 5 for
generated corner-connection models with very deep spandrel beams (dspld > 2.5) using
selected models M3-A , W2-A , and W2-C. The corresponding plot for edge-connection
models with very deep beams is presented in Fig. 4.3b using models W2-B and M3-B .
Vu (N) lOOOOO -o— Oen.M3-A
-+— Oen.W2-A -a— Oen.W2-C
50000
-50000 -
-100000
10.0 20.0 30.0 40.0
Spandrel strength parameter 5
(a) Corner-connection models
Fig 4.3 - Graph of punching shear Vu versus spandrel strength parameter 8
of generated models with varying width bsp and very deep spandrel beams
(continued next page)
53
Chapter 4 : Parametric Study of Spandrel Beams
V u (N)
200000
150000
100000
50000 -
0 -
-50000
-lOOOOO
-150000 -
-200000
-o— Oen.M3-B -*— Oen.W2-B
18.0
i — ] i | i
lO.O 20.0 30.0 40.0
Spandrel strength parameter 5
(b) Edge-connection models
Fig 4.3 - Graph of punching shear Vu versus spandrel strength parameter 8
of generated models with varying width bsp and very deep spandrel beams
(continued from next page)
Fig. 4.3 shows that the range of 5 for realistic analysis of Vu in corner-connection models
with very deep spandrel beams is between 20.0 and 27.0 . On the other hand, only a
lower limit is identified (at 5 = 18.0) for edge-connection models with very deep
spandrel beams.
For generated corner-connection models with intermediate spandrel depth (not
graphically shown here), the range of 5 for realistic analysis of Vu is between 10.0 and
15.0 while the lower limit of 5 for the corresponding edge-connection models is 9.5 .
4.3.2 Varying the depth of spandrel beam
It is obvious in Eq.(4-1) that the depth of spandrel beam (denoted as D,) may be directly
related to the parameter 5 by the effective depth of the spandrel beam, dsp. The depth
54
Chapter 4 : Parametric Study of Spandrel Beams
D, may also be related to 5 through the variable^ , the vertical dimension of Ut, which
is a variable for determining the parameter P . From these two direct relationships, the
parameter 8 may also be used to relate the varying depth Dy to the punching shear Vu .
The depth of the spandrel beam for ten generated models are set by generating a range
of data for D, as shown
Dl,eu= Ds + (/- l)Dj •••(4-7)
where DJgen is the generated depth of the spandrel beam, Ds is the overall depth of the
slab, D; is the original spandrel depth based on a given model and / refers to the
sequence from 1 to 10 for the ten generated models. The generated models with varying
depth of spandrel beam is presented graphically in Fig. 4.4 .
Varying depth
D,6 Di,5 D,i4 D|3 D,,7
Longitudinal reinforcement
Closed ties
D u
D 1.9 D, 1,10'
Column
Fig. 4.4 - Generated models with varying depth of spandrel beam
55
Chapter 4 : Parametric Study of Spandrel Beams
All the data for the generated models remains unchanged except for the variable D,
which in turn will generate a range of the parameter 5 to be used to determine the
punching shear, Vu , of the generated models. For example, typical generated data and
results for models based on ModelM2-A are as shown in Table 4.2 .
From observation of Table 4.2, it can be seen that for the first two generated models,
one with a torsion strip (i.e. D, = 100 m m ) and the other with a shallow spandrel depth
(i.e. D1 = 103.6 m m ) , the Wollongong Approach produces a negative Vu . From the
third model onwards, Vu increases numerically with D} . This is clearly illustrated in Fig.
4.5 with a plot of Vu versus 5 .
The graphical relationship between the spandrel strength parameter, 8, and the calculated
punching shear results of the generated models shown in Fig. 4.5 is based on selected
corner-connection models M2-A , M3-A , Wl-A , W2-A , M3-C and W2-C .
Table 4.2 - Typical generated data and results from parametric study of
varying depth of spandrel beam in Model M2-A
Generated
model (/)
1
2
3
4
5
6
7
8
9
10
Dj (mm)
100.0
103.6
107.2
110.8
114,4
118,0
121.6
125.2
128.8
132.4
a
1.0790
1.0790
1.0790
1.0790
1.0790
1.0790
1.0790
1.0790
1.0790
1.0790
P 3.9531
4.0339
4.1148
4.1956
4.2765
4.3574
4.4382
4.5191
4.5999
4.6808
8
3.7004
3.9549
4.2166
4.4855
4.7616
5.0448
5.3353
5.6328
5.9376
6.2495
VU(N)
-16470
-3986
7279
17502
26823
35356
43194
50412
57075
67306
56
Chapter 4 : Parametric Study of Spandrel Beams
Vu (N) 150000
10OOOO -
50000
-5OOO0
-1OOOO0
-150000 -
-200000
-*— Oen.M2-A — — Qon.M3-A -O-Oon.Wl-A -»— Oen.W2-A -x—- Oen.M3-C
Ocn.W2-C
35.0
O.O 10.O 20.0 30.0 40.0
Spandrel strength parameter 5
Fig 4.5 - Graph of punching shear Vu versus spandrel strength parameter 8
of generated corner-connection models with varying Dj
It can been seen that the Wollongong Approach has limitations, beyond which the
analysis of Vu would not be feasible. From Fig. 4.5, it can be observed that the lower
limit for realistic analysis of Vu ranges from 8 = 3.0 to 8 = 9.0 whereas the upper
limit ranges from 8 = 25.0 to 8 = 35.0 .
A similar graphical representation for generated edge-connection models based on
selected edge-connection models M2-B , M3-B , Wl-B and W2-B is shown in Fig. 4.6 .
In this case it can be observed that the lower limit for realistic analysis of Vu ranges
from 8 = 4.0 to 8 = 11.0 whereas the upper limit is at 8 = 29.0 .
57
Chapter 4 : Parametric Study of Spandrel Beams
-*— <3en.M2-B -*— Oen.M3-B -O- Gen.Wl-B -A— Gen.W2-B
V u (N)
300000
200O00 -
100000
O
-100000
-20OO00
-300000
-400000
O.O lO.O 20.0 30.0 40.0
Spandrel strength parameter 5
Fig 4.6 - Graph of punching shear Vu versus spandrel strength parameter 8
of generated edge-connection models with varying Dj
4.3.3 Varying the area oflongitudinal steel reinforcement in spandrel beam
As the variable for longitudinal steel reinforcement (denoted as A!s) is also directly
related to the spandrel strength parameter, 8 in Eq.(4-1), the parametric study of the
effect of A/s on the punching shear, Vu , may be undertaken by relating the parameter,
8 to Vu for the generated models.
The area of the longitudinal steel in the spandrel beam for say, eight generated models
are set by generating a range of data for Als as shown:
, Als Ah, gen = 2Ah - ' -j
...(4-8)
58
Chapter 4 : Parametric Study of Spandrel Beams
where Aisgen is the generated longitudinal steel area, A/s is the original longitudinal steel
area based on a given model and /' refers to the sequence from 1 to 8 for the eight
generated models with varying A[s (see Fig. 4.7).
Longitudinal reinforcement Ajs
Spandrel beam
• Original longitudinal reinforcement
O Additional longitudinal reinforcement
Fig. 4.7 - Generated models with varying amount of longitudinal steel
reinforcement in spandrel beam
This means that, all the data for the generated models remains unchanged except for the
variable Als which in turn will generate a range of the parameter 8 to be used to
determine the punching shear, Vu , of the generated models. For example, typical
generated data and results for models based on Model M2-A are as shown in Table 4.3 .
59
Chapter 4 : Parametric Study of Spandrel Beams
Table 4.3 - Typical generated data and results from parametric study of
varying longitudinal reinforcement in spandrel beam of Model M2-A
Generated model (/)
1
2
3
4
5
6
7
Als (mm2)
791.68
678.59
565.49
452.39
339.29
226.20
113.10
a
1.8882
1.6184
1.3487
1.0790
0.8092
0.5395
0.2697
P 4.3574
4.3574
4.3574
4.3574
4.3574
4.3574
4.3574
8
8.8285
7.5673
6.3061
5.0448
3.7836
2.5224
1.2612
P«(N)
79142
69427
55371
35356
6276
-39076
-125269
From observation of Table 4.3, it can be seen that for the first five generated models, the
calculated Vu reduces numerically with a reducing A/s. From the sixth model onwards,
the Wollongong Approach produces a negative Vu . This is clearly illustrated in Fig. 4.8
with a plot of Vu versus 8 .
The graphical representation shown in Fig. 4.8 is the relationship between the spandrel
strength parameter, 8 , and the calculated punching shear results of the generated models
based on selected corner-connection models M2-A , M3-A , Wl-A , W2-A , M3-C and
W2-C . It can be seen that the Wollongong Approach has limitations beyond which the
analysis of Vu would not be feasible. From Fig. 4.8, it can be observed that the lower
limit for realistic analysis of Vu ranges from 8 = 4.0 to 8 = 6.0 whereas the upper
limit ranges from 8 = 25.0 to 8 = 34.0 .
60
Chapter 4 : Parametric Study of Spandrel Beams
Vu (N) 100000
-100000
-200000
Oon.M2-A Qen.M3-A Gen.Wl-A Oen.W2-A Gen.M3-C Oon.W2-C
O.O 10.0 20.0 30.0 40.0
Spandrel strength parameter 5
Fig 4.8 - Graph of punching shear Vu versus spandrel strength parameter 8
of generated comer-connection models with varying A[s
Vu (N) 250000
200000
150000 -
100000 -
50000
-50000 -
•100000
*— Gon.M2-B •— Oen.M3-B
Oen.Wl-B W2-B
0.0 lO.O 20.0 30.0 40.0
Spandrel strength parameter 5
Fig 4.9 - Graph of punching shear Vu versus spandrel strength parameter 8
of generated edge-connection models with varying A\s
61
Chapter 4 : Parametric Study of Spandrel Beams
A similar graphical representation for generated edge-connection models based on
selected edge-connection models M2-B , M3-B , Wl-B and W2-B is shown in Fig. 4.9 .
In this case it can be observed that the lower limit for realistic analysis of Vu ranges
from 8 = 3.0 to 8 = 8.0 whereas the upper limit is at 8 = 28.0 .
4.4 Summary of Parametric Study Results
From the limited parametric study of the Wollongong Approach in analysing Vu of
reinforced concrete flat plates with spandrel beams, it has been shown that, in general,
the punching shear increases with deeper and wider spandrel beam but only within some
boundary limits beyond which the analyses are not feasible. It should be noted that the
parametric study was undertaken, based not only on one actual model tested, but on
several models by varying the features of their respective spandrel beams. Therefore,
one would not expect the upper or lower limits of 8 to be perfectly a certain value for
the group of models considered. A range of upper and lower limits of 8 were found for
the several models considered, so as to be used as a guide within which the analysis of
Vu may be reliable. The same also applies to the amount of longitudinal steel
reinforcement in the spandrel beam.
All these variations have the effect of influencing the spandrel strength parameter 8
directly, which is conveniently used to study its relationship with the punching shear
strength Vu of the generated models. The results obtained from the parametric study are
summarised in Table 4.4 . It is noted that any generated models beyond the limits of 8
will not produce any realistic or reliable punching shear results.
62
Chapter 4 : Parametric Study of Spandrel Beams
Table 4.4 - Summary of parametric study results of limits or ranges of 8 for
reliable punching shear strength analysis
T y p e of parametric study of the
spandrel b e a m
(1) Varying the width bsp of
spandrel beam
(a) Shallow beams
dSpld <\.2
(b) Intermediate
depth of beams
1.2 < dspld < 2.5
(c) Very deep beams
dspld > 2.5
(2) Varying the depth D} of spandrel beam
(3) Varying the area of longitudinal
reinforcement Ais of spandrel beam
Corner-connections of generated models
Lower limit or
range of 5
3.6 to 3.9
10.0 to 15.0
20.0
3.0 to 9.0
4.0 to 6.0
Upper limit or
range of 8
-
-
27.0
25.0 to 35.0
25.0 to 34.0
Edge-connections of generated models
Lower
limit or
range of 8
3.4 to 4.0
9.5
18.0
4.0 to 11.0
3.0 to 8.0
Upper limit or
range of 8
-
-
-
29.0
28.0
63
Chapter 5 : Theoretical Approach
CHAPTER 5
THEORETICAL APPROACH TO INCORPORATE EFFECT OF
PRESTRESS IN THE ANALYSIS OF PUNCHING SHEAR
5.1 Effect of Prestress on Punching Shear
5.1.1 General
One of the fundamental reasons of introducing prestress into structural concrete
elements is to reduce the amount of cracks associated with a structure under severe
loading conditions. W h e n punching shear is likely to occur, the first sign is the
appearance of a roughly circular tangential crack around the perimeter of the column
[Park and Gamble, 1982]. This is followed by radial cracks which extend from the
perimeter of the column. These grow and are accompanied by more circumferential
cracks in the form of a quarter- or half-truncated cone around the loaded area or the
column support [Falamaki and Loo, 1992].
The effect of prestress on the formation and direction of inclined shear cracks can be
seen by examining the stresses acting on a small element located at the centroidal axis of
an uncracked beam or slab strip as shown in Fig. 5.1 [Gilbert and Mickleborough,
1990].
By using a simple Mohr's circle construction, the principal stresses and their directions
are readily determined. W h e n the principal tensile stress at reaches the tensile strength
of concrete, cracking occurs and the cracks form in the direction perpendicular to the
direction of o t ,
64
Chapter 5 : Theoretical Approach
W h e n the prestress is zero, a{ is equal to the shear stress T and acts at 45° to the strip
axis, as shown in Fig. 5.1a . If diagonal cracking occurs, it will be perpendicular to the
principal tensile stress, i.e. at 45° to the strip axis. When the prestress is not zero, the
normal compressive stress cr (=P/A) reduces the principal tension o"j , as illustrated in
Fig. 5.1b . The angle between the principal stress direction and the strip axis increases,
resulting in a less inclined crack if cracking occurs.
Therefore, prestress may improve the effectiveness of any transverse reinforcement that
may be used to enhanced the shear strength of the structural element. As the prestress
caused the inclination of the crack to be smaller, a larger number of vertical stirrups or
ties are crossed by the crack and consequently a larger tensile force can be carried across
the crack.
If the prestressing tendon is inclined at an angle dp , the vertical component of the
prestress Pv (-P sin Qp = P Qp) usually acts in the opposite direction to the load which
induces shear. The force Pv may therefore be included as a significant part of the shear
strength of the cross-section.
65
Chapter 5: Theoretical Approach
~3Z 7 7
j ^
At A: o =
J W
^5= SSSJ
"PT
- — and A
x = lb
(a) IfP = 0: 11 Shear
- 1 A , (90-ev) = 45°
(b) IfP>0
Shear
Normal
v/(90^ev)>45°
o.
Fig. 5.1 - Effect of prestress on principal stresses in the centroid
of a slab strip
66
Chapter 5 : Theoretical Approach
5.1.2 Current concrete codes of practice
(a) Australian Standard AS 3600 -1988 [SAA, 1988]
The effect of prestress is taken into consideration in the Australian Standard in the
following manner. In the case of no bending moment transfer to the column support in
the direction considered, the equation for calculating the ultimate shear strength of the
slab is a modification of Eq.(3-1).
Vuo= ude[fcv+ 0.3 Gcp] ...(5-1)
where the additional term, acp is the average intensity of effective prestress in the
concrete in each direction. The effect of prestress on the shear strength of the slab is
taken 0.3 times the effective prestress, on the basis of recommendation by the ACI318-
89 [ACI, 1989] as shown in Eq.(5-2). As identified in Chapter 3, u is the length of the
critical perimeter whereas de is the effective depth of the prestressed slab.
When there is moment transfer to the column, the procedures are the same as those for
reinforced concrete flat plates. The only difference is the additional prestress parameter
Gc in Eq.(5.1) . This is applied to the appropriate equations for various cases such as
Eq.(3.2) for a typical slab having a spandrel beam with the minimum requirement of
closed ties. The associated equations may be found elsewhere [SAA, 1988].
(b) American Concrete Institute ACI318-89 [ACI, 1989]
According to the ACI code, for slabs prestressed in two directions, in which no portion
of the column cross-section is closer to a discontinuous edge than four times the slab
67
Chapter 5 : Theoretical Approach
thickness, the shear strength is calculated based on the principal tensile stress approach,
that is:
Vuc = t 0.29 J7c + 0.3 acp ] ude + Vp ...(5-2)
where Gcp is as described in Eq.(5-1) and Vp is the vertical component of the effective
prestress force at the section, which is generally deemed to be negligible and may be
ignored.
Based on research carried out by ACI-ASCE Committee 423 [1983] and Burns and
Hemakom [1977], the ACI method considers 0.3 times the effective prestress in the
analysis of Vu . This recommendation was adopted by the Australian Standard AS 3600-
1988 [SAA, 1988], as shown in Eq.(5.1).
Other limitations include that f'c should not be greater than 35 MPa and acp in each
direction should not be less than 1 MPa or greater than 3.5 MPa.
Where some portions of the column cross-section is close to a discontinuous edge by
within four times the slab thickness and in the presence of an unbalanced moment, the
calculation of Vu is the same as for a normally reinforced concrete slab section (by
applying Eq.3.3 as described in Chapter 3). Apparently, the ACI method considers that
the effect of prestress is negligible when such conditions exist.
(c) British Standard BS 8110 : 1985 [BSI, 1985]
The recommendations for the method of determining the shear strength of beams are
also applicable to slabs in the British Standard method, whereby consideration is given
to uncracked or cracked sections in flexure, as indicated by Kong and Evans [1987].
68
Chapter 5 : Theoretical Approach
The ultimate shear resistance of a concrete section which is uncracked in flexure is given
by:
Vco = 0.67 bDs -Jo? +0.8 acpat ...(5-3)
where b is the width and Ds is the overall depth of the slab section, acp is the concrete
compressive stress due to prestress and at is the principal tensile stress determined
from
a, = 0.24 JZ -(5-4)
Note that f' refers to the characteristic cube strength of the concrete in this case. On
the other hand, the ultimate shear resistance when the section is cracked in flexure may
be calculated as:
Vcr = (1 - 0.55 f-) vc bde + Mo-Tf > 0J V/« H -(5-5)
J p v
where,
^SL - ratio of the effective tendon prestress to the ultimate strength of
tendons;
v = design shear stress as from Eq.(3-5) in which the total tension steel
area includes the prestressing tendons;
dp = effective depth to the centroid of the tendons;
= decompression moment, i.e. the moment necessary to produce zero
stress in the concrete at the extreme tension fibre;
e
M,
69
Chapter 5 : Theoretical Approach
V and Mv = design shear force and design bending moment at the section under
consideration.
For a cracked section analysis, in which case M* is larger than M0 , the ultimate shear
resistance is to be taken as the lesser of Eq.(5-3) and Eq.(5-5).
5.1.3 Effect of prestress in the Wollongong Approach [Loo and Falamaki, 1992]
In order to fully understand the forces acting on the column-slab connection in a post-
tensioned flat plate, two free-body diagrams based on the Wollongong Approach are
shown in Figs. 5.2 and 5.3 for forces acting at a corner- and edge-column positions
respectively.
Some parameters shown have already been briefly described in Chapter 1. They include
Mj, Vj, M2 \ V2 , T2, MC1, MC2 and Vu.
Mj and Vj are the negative yield moment of the slab and the shear force respectively
over the front segment of the critical perimeter, whereas M2 , V2 and T2 are the
negative yield moment of the slab, shear force and torsional moment respectively over
the side segment of the critical perimeter.
It should be noted that the critical perimeter as defined by Loo and Falamaki [1992] for
the direct transfer of the slab bending moment and shear force to the column is also
defined in Figs. 5.2 and 5.3 . The front segment of the perimeter is located 0.5de from
the front face of the column, where de is the effective slab depth. The side segments of
the critical perimeter coincide with the column side face(s).
70
Chapter 5 : Theoretical Approach
MQI and MQ2 ar& the total unbalanced moments transferred to the column centre in the
main and transverse moment directions respectively and finally, Vu is the punching shear
strength of the slab.
Additional parameters on the effect of prestress shown in Figs. 5.2 and 5.3 are the
prestressing force applied, the overall area of prestressed tendons and their
eccentricities.
Pe] and Pe2 are the effective prestressing forces acting at the front and side faces of the
column-slab connections respectively. The subscripts x and y denote the horizontal and
vertical components of these applied prestressing forces. Apj and AP2 are the area of
the prestressed tendons while ej and e2 represent the eccentricities of the prestressed
tendons in the main and transverse directions respectively. The term nj refers to the
number of prestressed tendons at the front face of the column in the main bending
direction. Similarly, n2 is the number of prestressed tendons at the side face of the
column in the transverse bending direction.
Other variables shown include the dimensions of the column (C/ and C2), the width and
depth of the spandrel beam (bsp and Df), the overall and effective depths of the slab (Ds
and de), the height of the column Lj and the horizontal reaction component at the base
of the column.
The procedure undertaken to incorporate the effect of prestress into the prediction
formulas developed by Loo and Falamaki [1992] is elaborated in the following sections.
71
Chapter 5: Theoretical Approach
prestressing tendons area Ar
Prestressing force Pe2 (per n 2 wires)
y
A
Prestressing force Pel (per nj wires)
prestressing tendons area A p 2
Prestressing tendon
L -I Dc
Slab
Spandrel beam
Column View 'x'
Ci " bsp/2
Fig. 5.2 - Forces acting in the corner column region of a
post-tensioned concrete flat plate
72
Chapter 5 : Theoretical Approach
de/2
v Prestressing u force P
(per n wires)
Prestressing force P e 2
(per n 2 wires)
Prestressing force Pel
(per nj wires)
prestressing tendons area A p i
prestressing tendons area A, P2
Fig. 5.3 - Forces acting in the edge column region of a
post-tensioned concrete flat plate
5.1.4 Equilibrium equations
By referring to Fig. 5.2, for a corner connection, the three equilibrium equations may be
derived as shown.
By considering the equilibrium of vertical forces at the ultimate state, i.e. I*Fy = 0,
Vu = Vj + V2 - niPely ...(5-6)
B y enforcing the equilibrium of moments, w e have for IM0 = 0 ,
73
Chapter 5 : Theoretical Approach
MC1 = T2 + M1+ [VJ - ni Pely] [Cj - (tsp ' de)] - nj Pelx e, + MT ...(5-7)
and for, ET 0 = 0 ,
MC2 M2
h [Ll - (^ ' D'>]
rhPeix -(5-8)
" MC2= U iDi-Ds)-
n2Pe2*Ll -(5"9)
2L;
In Eq.(5-7),
= ^ [ / f r - P , ] + E Q - * g ] (5_10) r 2L; 2
for position O not coinciding with 0;, and
= M C / [ D 7 - Z ) 5 ] ...(5-11) r 2Lj
for position (9 coinciding with Oh i.e. C/ = bsp
Similarly, for the edge connection as shown in Fig. 5.3,
Vu= V1 + 2V2 -niPely -(5-12)
74
Chapter 5 : Theoretical Approach
MC1 = 2T2 + Mj + [V7 - n, Pely] [Cj - (b*p-de)] . nj Pe}xej + MT ...(5-13)
and
"«" i^}Si 2L7
Before discussing the interaction equation involving torsion, shear and bending in the
spandrel beam, the spandrel beam parameters are explained first.
5.1.5 Spandrel beam parameters
The formulation for the slab restraining factor \|/ has been discussed in detail in previous
work by Loo and Falamaki [1992] for reinforced concrete flat plates. It is dependent on
the strength parameter of the spandrel, the longitudinal reinforcement factor and the
transverse reinforcement factor which are represented as 8, a and p respectively.
The equation for 8 has been identified in Chapter 4 and its derivation is briefly discussed
here. The factor 8 is expressed as:
8 = o d A -(5-15)
in which, the effective depths of spandrel beam and slab, dsp and de respectively, are
determined using Eq.(5-34), except that the term Ap2 is replaced by Api for de . The
terms for a and (3 are derived from the ratio of the longitudinal and transverse spandrel
75
Chapter 5 : Theoretical Approach
strength to their respective minimum strength requirement as governed by the Australian
Standard AS 3600-1988 [SAA, 1988] for concrete structures. They are:
a = r AlJl\ ...(5-16)
and,
P = ut [ffl + <°ol -(5-17)
S min
The terms for Ais and fiy are the total area and yield strength of the longitudinal steel
reinforcement, and Aws and fwy are the area and yield strength of closed ties in the
spandrel beam. The term Ut is defined in Section 4.2 .
Loo and Falamaki [ 1992 ] simplified the equations for a and |3 to:
a = ^hy ...(5-18) 200000
and
= Ut[a> + coQ] ...(5-19)
50000
The transverse spandrel strength parameter is defined as:
0) = Ksfwy ...(5-20)
76
Chapter 5 : Theoretical Approach
while its additional transverse strength co0 is due to slab restraint for the spandrel
rotation and is related to the concrete cracking strength. It has been established
empirically by Loo and Falamaki [1992] to be a constant value, or
ta0= 446 N/mm ...(5-21)
Eq.(5-21) is regarded to be generally applicable, since any slack would be taken up by
the formula for evaluating the slab restraining factor, \\t [Loo and Falamaki, 1992].
From these expressions, the semi-empirical formulas to determine v|/ for both edge and
corner column positions were established for reinforced concrete flat plate models. The
application of the same semi-empirical formulas for post-tensioned concrete flat plate
models was not feasible, as shown by Lei [1993]. The Wollongong Approach gave an
over-estimated punching shear result for the corner column position whereas a negative
value was given for the edge column position. The reason for the inapplicability of the
Wollongong Approach for the post-tensioned concrete flat plate models was due to the
semi-empirical formulas for \\J, which do not take the effect of prestress into account, as
mentioned earlier in Section 1.5.2 (c). Therefore, this necessitates the testing of post-
tensioned concrete flat plate models to obtain data for establishing new sets of semi-
empirical formulas to determine the slab restraining factor, \|/. These will be elaborated
further in Chapters 7 and 10 respectively.
For information, the semi-empirical formulas for the slab restraining factor v|/ applied to
reinforced concrete flat plates are re-produced below. For edge columns,
75 3
yedge = 2.65 - 1.71 log10 [(—) 8] -.(5-22) .5
For corner columns where C2 > Cj,
11
Chapter 5 : Theoretical Approach
^corner = 10.70 - 4.45 log1Q [T8] ...(5-23)
and for C 2 < Ch
ycomer = 5.04 - 3.12 log10 [yb] ...(5-24)
In the above expressions, the column width factor
T-[f]J|
5.1.6 Interaction equation
The interaction equation which takes into account the effects of torsion, shear and
bending in the spandrel beam is based on the theory originated by Elgren, Karlsson and
Losberg [1974]. This interaction formula, which described the failure of an isolated
beam with the compression zone on top, may be derived from:
*&_ + [-3_]2 _ £ _ -JL. + [--*-]' —- -?«-=/ ...(5-26)
Astflydsp 2At Awf^ Astfiy 2dsp A-wsfwy At fly
where Ast is the area of tensile steel reinforcement and At is the area of the rectangle
defined by the centres of the four longitudinal corner bars of the spandrel beam.
A f After substituting Ab = 2Ast and co = ^ into the Eq.(5-26), the following
s
expression is obtained.
78
Chapter 5 : Theoretical Approach
2M2 + tJL/ I J ^ + [A.J J. J ^ = . _ Asflydsp "2A^ ® ^//y ~ 2 < V W %/fy
In cases of pure bending, pure torsion and pure shear, the ultimate beam moment,
torsion and shear are respectively,
Mus = | Asfiydsp -(5-28)
Tus = 2A( | M £ = V 2 4 C 0 S -(5-29) 1 2Ut \Ut®
v--*J^-w-$i
This leads to an interaction equation which may be expressed as:
M2_ + i^L/ + [ii/ = 1 ...(5-31) M^ Tus Vus
In order to be more conservative, the above interaction equation may be linearised as
[Elgrenetal, 1974]:
Ml_ + 1l_ + Yl_ = j ...(5-32) M T V iVJ-us xus rus
79
Chapter 5 : Theoretical Approach
5.1.7 Modifications to the interaction equation
In order to take into account of the established spandrel parameters in Section 5.1.5,
modifications have to be made to the interaction equation obtained from Section 5.1.6.
Furthermore, the simplistic expression for pure bending will need to be modified to
include other relevant factors. Thus,
Mm = [A„fly + Ap2f„) [1 - 0.6 ^ ± M j s l ] d s p ...(5-33) ®sp®spJc
where the effective depth of the spandrel beam,
dsp = A-stflyd-s + A.p2Jpy
dp (5-34)
Astfly + \lfpy
By substituting (co + C0o) for ® and ^Msfly for AiJiy, we can derive the expressions
for pure bending, torsion and shear respectively for the spandrel beam as:
Msp = VM„, -(5-35)
T„ = 200000^- , p ^ -(5-36) sp Ut V 2
Vsp = 200000 J ^ - V o W -(5-37)
Incorporating these relationship, the interaction equation becomes:
80
Chapter 5 : Theoretical Approach
M2_ + Il_ + Yl_ = , ...(5.38) M T V irisp *sp Ysp
or in its complete form
Ml + ^ = ^ = + $- = 1 -(5-39)
VM™ 200000±J<®L 2 0 0 0 0 0 ^ ^
i.e. — S — + - 7 ^ = = [¥ - - ^ ] 200000 i^. ...(5-40)
V2C/, J 217,
5.2 Evaluation of Ultimate M o m e n t and Shear
The procedure for evaluating the ultimate moment and shear acting at the front face of
the column and over the front segment of the critical perimeter, i.e. Mj and Vi
respectively is through the use of semi-empirical formulas developed by Loo and
Falamaki [1992]. These semi-empirical formulas are used to analyse the post-tensioned
flat plate models. The analysed Vu results can be compared with the experimental results
to test the suitability of utilising the same semi-empirical formulas or whether some
modifications may be necessary to ensure a closer prediction of the shear strength of the
models. This is carried out in Section 11.3 .
For corner connections,
Mi = Mr ...(5-41) lYl1, corner lvllu v
81
Chapter 5 : Theoretical Approach
and
Mr + M Vitamer = '3-66 + 5.04 ^ ^ L » ...(5-42) u+
for dsp/de < 1.2 and for ^i/tie > 1.2 ,
Mr + M ^corner = '0-50 + 4.70 Mlfiorner " ...(5-43)
Similarly for edge connections,
Mledge = M]u -(5-44)
The corresponding ultimate shear is:
Vledge = 9.35 + 3.49 ¥lf& -(5-45) 4
for Jjp/cie < 1.2 and for d^de > 1.2 ,
* W = 2.84 + 5.62 Ml'edgeT + Mm -(5-46) 4
In the above equations, M\u and M m refer to the bending strengths of the critical slab
strip at the front segment of the critical perimeter and at the corresponding mid-span
respectively. The analysis of Af/„ and Mm is briefly discussed in Section 5.4 .
82
Chapter 5 : Theoretical Approach
5.3 Analysis of Punching Shear
In order to analyse the punching shear strength of the post-tensioned concrete flat plate,
we need to consider the three equilibrium equations developed in Section 5.1.4 and the
interaction equation (5-40) derived in Section 5.1.7 . With these four equations, we can
proceed to solve simultaneously for the four unknown quantities, namely, Vu , T2 , V2
and M2 •
5.3.1 Corner connections with spandrel beam
Firstly, we consider the analysis formulas for corner connections with spandrel beam.
By condensing the equilibrium equations (5-6), (5-7) and (5-9) and the interaction
equation (5-40), we have for the punching shear strength
y = kfe^ ' ^4 + ^7 (5-47) k3 - k4k5
where,
k, = 200000 [ i w - W s - ^ * , ) ] U L ...(5-48) Mus V Vcorner
k2 = k6Mcl - MliCorner - kylCj - (bsp'de)] + nj Pelx ej ...(5-49)
42Ut k3 = -^r ...(5-50)
83
Chapter 5 : Theoretical Approach
K = J ^ -(5-5D
k5 = 9LL^!L ...(5-52)
k6 = 1 - [Dl ' D*] ...(5-53) 6 2L}
and,
ky = V1>corner - n2 Pely -(5-54)
Note that in Eqs.(5-48) and (5-49), MC1 and MC2 are the unbalanced moments in the
main'and transverse moment directions to be transferred to the centre of column,
respectively.
5.3.2 Edge connections with spandrel beam
In the case of edge connections, a similar procedure may be followed to obtain the
explicit equation for the punching shear strength. Condensing Eqs.(5-12), (5-13), (5-14)
and (5-40) gives:
v 2krk3k4 - k2k4 + kjky ...(5-55) u ' k3 - k4k5
where,
84
Chapter 5: Theoretical Approach
k, = 200000 [yedge - L \ J-2&- ...(5.56) Lus Wedge
iK'de)-'.edge
k2 = k6Mcl - Mledge - kyiCj - * e ] + nj Pelx ej . ...(5-57)
and,
*7 = VI>edge - nj Pely ...(5-58)
The other quantities are as those defined for the corner connections.
5.3.3 Connections with torsion strip
As generally defined, where dsp is not greater than de, the spandrel is referred to as a
torsion strip (with or without closed ties).
Equations for Vu derived in Sections 5.3.1 and 5.3.2 are applicable to flat plates with
spandrel beams, but they may be similarly applied to connections with torsion strips by
replacing Eq.(5-41) for a corner connection [Loo and Falamaki, 1992] with,
_ r 400 ] " M ...(5-59) Ml toner ~ lbsp + C2
lu
Similarly, the equations for Vu are also applicable to an edge connection with torsion
strips [Loo and Falamaki, 1992] except that Eq.(5-44) is replaced by
85
Chapter 5 : Theoretical Approach
Mltedge = i2b + c 1 M]u ...(5-60)
5.4 Ultimate Flexural Strengths Mju and Mm for Post-Tensioned Concrete
Sections
At the overload stage, the behaviour of a prestressed concrete flexural member is similar
to that of a reinforced concrete member such as a beam. The procedure used to
determine the moment capacity of a reinforced concrete section may be applied to a
prestressed concrete section, with the same initial assumptions associated with the
concrete compressive stress block.
Beyond the cracking load, the internal forces in a typical cracked section are as shown in
Fig. 5.4 . They consist of tensile forces Ts and Tp in the reinforcing steel and
prestressing tendon respectively, and a compressive force Cc in the concrete above the
neutral axis. If compressive reinforcement is also present, the compressive resultant Csc
acts in conjunction with Cc . Note that the section is at equilibrium at all load stages up
to the ultimate state when Mu prevails.
In adopting the concept of the critical strip widths [Loo and Falamaki, 1992], the
analysis of a flat plate takes the critical strip width to be equal to the width of the
corresponding column support. By definition, the bending moments of the critical slab
strip at the front segment of the critical perimeter and at the corresponding mid-span, are
Miu and Mm respectively.
The determination of these bending moments is described and elaborated in detail in
Chapter 9.
86
Chapter 5 : Theoretical Approach
o— — |
4°"
H H
t t ^ U
\
<
z
^ Q
-df
-O T?
-C? T?
-O T3CIO
1)
1) Uu
o C o U
cfl
c l-u U-J
Cfl
O
Cut
Cfl
U O o P-4
Fig. 5.4 - Forces, strains and stresses in a prestressed concrete section
87
Chapter 5 : Theoretical Approach
5.5 Losses of Prestress
The determination of prestress losses is based on the provisions given in the Australian
Standard AS 3600-1988 [SAA, 1988] and outlined in the Concrete Design Handbook
(in accordance to AS 3600-1988) published by Cement and Concrete Association of
Australia [1991]. This has also been discussed by Gilbert and Mickleborough [1992].
Losses may be classified as the immediate or short-term prestress losses and the time-
dependent or long-term losses of prestress.
The difference between the prestressing force imposed at the jack, Pjt and the force in
the tendon immediately after transfer, Pit is termed as the immediate loss. The gradual
loss of prestress that takes place with time is called the time-dependent loss. The final or
effective prestressing force, Pe remains after all losses.
5.5.1 Immediate losses
The main immediate losses are caused by:
• elastic deformation of the concrete as the prestress is transferred;
• friction along the draped tendon in a post-tensioned member;
• friction in the prestressing jack;
• friction in the anchorage; and
• anchorage slip.
(a) Loss due to elastic deformation
For a post-tensioned slab with more than one tendon and where the tendons are stressed
sequentially, the elastic deformation losses vary from one tendon to the other. They are
at a maximum in the tendon stressed first and a minimum (zero) in the tendon stressed
88
Chapter 5 : Theoretical Approach
last. These losses are usually small and, for practical purposes, the average elastic
shortening loss may be obtained as:
1 E P p = 2 ltGci -( 61)
where Ep and Ec are the moduli of elasticity of prestressed tendon and concrete
respectively while <5cl is the concrete compressive stress at the tendon level immediately
after transfer. GCI may be calculated as [Warner and Faulkes, 1988]:
C- = "f + f T 1 ] 6 -(5~62) Ag *g
The terms Ag and lg are the gross area and moment of inertia of the prestressed section,
and e is the eccentricity of prestressing force from the centroid of the uncracked
section.
(b) Loss due to friction along the tendon
For post-tensioned members, friction losses occur along the tendon during the stressing
operation. Friction between the tendon and the duct results in a gradual reduction in
prestress with distance Lpa along the tendon from the jacking end.
This loss depends on the total angular change of the tendon, the distance from the
jacking end and the size and type of the sheathing containing the tendons.
89
Chapter 5 : Theoretical Approach
The resultant prestress after loss due to friction along the tendon may be obtained by an
equation recommended by major concrete codes [ACI, 1989, BSI, 1985 and SAA,
1988].
Pa = p.^-a-tot + tpLpJ ...(5-63)
where,
Pa is the force in the tendon at any point Lpa (in metres) from the jacking end
P, is the force in the tendon at the jacking end
p. is a friction curvature coefficient which depends on the type of duct. For
galvanized steel duct, \i ~ 0.2.
atot is the sum in radians of the absolute values of all successive angular
deviations of the tendon over the length Lpa.
Pp is an angular deviation or wobble term which depends on the duct diameter.
For flat metal ducts: 0.016 < p> < 0.024.
(c) Loss due to friction in the jack
This is dependent on the type of jack being used. In this study, we may assume that the
loss is negligible.
(d) Loss due to friction in the anchorage
This loss (which is minimal) may be estimated as two percent of the jacking force
[Cement and Concrete Association of Australia, 1991].
AP = 0.02xPj ...(5-64)
90
Chapter 5 : Theoretical Approach
(e) Loss during anchoring
In post-tensioned members, some slip or draw-in occurs on transfer of prestressing force
from the jack to the anchorage. The amount of slip depends on the type of anchorage
used. For wedge-type anchorages, the maximum slip A / may be taken as 6 mm
[Cement and Concrete Association of Australia, 1991]. The loss in prestress due to
anchoring may be obtained from:
AP = 2 Ztctnti) ...(5-65)
where,
Z = V^7^ tan co
tan co = friction loss per unit length
A/ = anchorage slip + draw-in
= 6 mm for current commercially available systems.
5.5.2 Time-dependent losses
Since the interval between casting and testing of the post-tensioned flat plate models
was relatively short, i.e. not more than 40 days, it is assumed that the time-dependent
losses in prestress was quite minor. Thus, this gradual losses are not taken into
consideration in the calculation for the final effective prestress Pe in the tendons prior to
the ultimate load tests of the post-tensioned flat plate models.
In general, the time-dependent losses in prestress are caused by the gradual shortening
of the concrete at the tendon level due to creep and shrinkage, and by relaxation of the
tendon itself.
91
Chapter 5 : Theoretical Approach
5.5.3 Summary of prestress losses considered in the calculation of Vu
The formulas for prestress losses, to be used in the analysis of punching shear strength,
Vu , of post-tensioned concrete flat plate models in Chapter 11, are summarised and
presented in Table 5.1 .
Table 5.1 - Summary of formulas for prestress losses used in the
analysis of Vu
Immediate losses due to :
(a) Elastic deformation
(b) Friction along the tendon
(c) Friction in the anchorage
(d) Anchorage slip
Formulas used:
1 EP
P 2 Ec ct
p - p -MV-tot + Vp^pa]
AP = 0.02 x Pj
A P = 2 Z tan a*
92
Chapter 6 : Test Program - An Introduction
CHAPTER 6
TEST PROGRAM - AN INTRODUCTION TO
THE TEST MODELS
6.1 General
As part of a comprehensive study of the behaviour of concrete flat-plate structures, four
half-scale post-tensioned concrete models have been tested to failure under simulated
uniformly distributed load. These supplement similar tests on nine reinforced concrete
models [Falamaki and Loo, 1992]. Representing two continuous panels of a building
floor, each model contained six columns, including three corner and edge column-slab
connections with or without spandrel beam.
This chapter describes the overall layout of the flat plate models in terms of size and the
support system used. It also details the design of the models, with emphasis on the
layout of prestressing tendons and steel reinforcement, type of anchorage system used,
properties of materials used for tendons, ductings, steel and concrete, and finally the
design of the spandrel beam.
6.2 Details of Models
The details of the models in terms of dimensions, fabrication and instrumentation are
similar to the ultimate load tests conducted by Falamaki and Loo [1992] on reinforced
concrete flat plate models. The general layout of the test models for post-tensioned
concrete flat plates were illustrated in Figs. 1.1 and 1.2 (Chapter 1).
93
Chapter 6 : Test Program -An Introduction
6.2.1 Dimensions
In Fig. 1.1, the shaded area of the half-scale flat plate represents two adjacent panels at
the corner of a typical building floor. The test models shown in Fig. 1.2 were
constructed to represent these corner panels (denoted as ABGF and BCHG). Four
models were constructed and tested to failure. They are identified as models PI , P2 ,
P3 and P4 . Models PI and P4 were built without spandrel beam; Model PI a free edge
while Model P4 featured a torsion strip along the edge. On the other hand, Model P2
was constructed with a 300 mm by 160 mm spandrel beam, and Model P3 incorporated
a 250 mm by 120 mm spandrel beam.
Apart from the spandrel beam and torsion strip, all the models are identical in all other
aspects including having the same slab thickness of 80 mm. The overall dimensions of
the models including the cross sections of the slab models are given in Fig. 1.2 . These
dimensions are similar' to those adopted for the reinforced concrete models [Falamaki
and Loo, 1992].
6.2.2 Support system
In all the models, the restraining effects of the adjacent panels of the building (which
were curtailed) were simulated by enlarging the cross sections of columns F, G , H and
C in order to provide more stiffness in the redistributing of moments following the
yielding of the slab reinforcement. The sizes of the various column supports, which are
common in all the models are shown in Fig. 1.2
The models were tested in a self-contained steel space frame resting on the floor of the
laboratory. The columns of the model were extended downward to a theoretical
contraflexure point (at approximately half the column height below the floor). Since the
94
Chapter 6 : Test Program - An Introduction
columns above the slab were omitted, it was necessary to double the stiffness of each
column by halving the column height so as to resist twice the moments. For ease of
construction each of the flat plate model was cast onto six prefabricated steel column
sections (with equivalent stiffnesses to the cast-in-place concrete columns).
Top steel plate (bolted onto flat plate)
Fig. 6.1 - Typical column support details
The height of the columns in each model varies, depending on the depth of the spandrel
beam at the edge. For example, in Model PI, the height of the column from the centre
of the slab to the centre of support is 410 mm whereas it is 430 mm for Model P2.
Details of a typical support system is shown in Fig. 6.1 . The adjustable support system
95
Chapter 6 : Test Program - An Introduction
allowed for proper alignment of the columns both in the horizontal and vertical
directions. At all the support positions, the slab with the truncated concrete columns
were held monolithically to the reusable steel column sections by twelve 16 mm mild
steel bolts in column H and eight 16 mm bolts in each of the remaining columns. The
arrangement shown in Fig. 6.1 ensured a monolithic connection between the spandrel
beam (if existed), the column and the slab. This reusable column support system has
worked successfully in the tests of nine half-scale reinforced concrete models [Falamaki
and Loo, 1993].
6.3 Design of Models
The design of the models are divided into four main components, namely the design
layout of prestressing tendons and steel reinforcement in the slab, the design
specifications of prestress and reinforcement, other material requirements for cable ducts
and concrete strengths and finally, the design of a reinforced concrete spandrel beam for
Model P2 and Model P3 .
6.3.1 Design of prestressing and reinforcement
The typical model slab was designed with the collaboration of a Senior Engineer of VSL
Prestressing, Sydney in accordance with the Australian Concrete Standard AS 3600-
1988 [SAA, 1988]. The design layout of cable ducts and steel reinforcement, which are
identical in all the models are shown in Figs. 6.2 and 6.3 respectively. The design
superimposed service load for all the models was 8 kPa.
96
Chapter 6 : Test Program - An Introduction
10-mm perimeter bars (lapped 300 m m )
1-tendon carrying
duct
Legend
(4 ) cable duct number
yv location of anchorage / 3 2 \ as measured from slab
soffit, in m m PvTi soffit to the bottom '—' of ducting, in m m © ®
150 150©© © © ©
© © Note: All measurements are in milimetres
Fig. 6.2 - Design layout of cable ducts
97
Chapter 6 : Test Program - An Introduction
In Fig. 6.2, all the cable ducts spanning across both directions contained two
prestressing tendons except for the ducts running along the edges of the slab which
accommodates only one prestressing tendon. The size of the prestressing tendons used
was 5.05 mm diameter. The numerical values along the length of the cable ducts in both
directions indicate the profile of the ducts along the respective direction, that is the
required height of the duct measured from the soffit of the slab to the underside of the
duct.
The corresponding values at both ends of the cable ducts along the edges of the slab are
the heights of the duct measured from the soffit of the slab to the centreline of the duct.
Fig. 6.2 also shows that the cable ducts are spaced closer within the column strips at 250
mm and are spaced wider in between the column strips from 350 mm to 450 mm.
Besides the prestressing tendon layout, Fig. 6.2 also shows the layout of 10 mm
diameter reinforcing bars along the edges of the slab as perimeter bars, one on top of the
other. They are essentially required to strengthen the concrete edges during the post-
tensioning process.
As shown in Fig. 6.3, all the positive reinforcement bars are located over all the column
supports in both the longitudinal as well as the transverse directions. More
reinforcement are required for larger size columns such as column G and H . Negative
reinforcement are provided at the mid-span (bottom) of column strips along the main
bending direction.
All these additional steel bars are provided to increase the flexural strength at the column
supports as well as the critical column strips. Figs. 6.4 and 6.5 show the profiles of the
cable ducts along both directions in a cross-sectional view.
98
Chapter 6 : Test Program - An Introduction
4 No. 10-mm bars x 900 m m long (top)
2 No. 10-mm bars @ 250 (bottom) x 2000 m m long
4 No. 10-mm bars x 900 m m long (top)
W/,'
m F - m %
7\
2 No. 10-mm bars (top)
3 No. 10-mm bars @ 250 (bottom) x 2000 m m long
450
4 No. 10-mm bars x 900 m m long (top)
2 No. 10-mm bars x 900 m m long (top)
2 No. 10-mm bars (top)
*
il/
2 No. 10-mm bars @ 250 (bottom) x 2000 m m long
s .='
2 No. 10-mm bars (top)
/-
2 No. 10-mm bars (top)
o
2 No. 10-mm bars x 900 m m long (top) 450
Fig. 6.3 - Design layout of steel reinforcement
99
Chapter 6 : Test Program - An Introduction
H 00 < PQ
H 00 W
£
c o T3 C 3 B v G
os. 5.05
profile.
G
0 3 —
es 0 •0
£3 6
UO uC . O
oo L_* 0 cx Z 60
c CN-5 U- O O 3 P-. -a
4^ O •a 1 <=
1 D
*-• 1 g O CN g .
•ut m <D
ON
0 •uf
1
1 I un r _
V
\ \ cn \CN
cnl
1 1
0
on CN
ll 1 on / 0
CN/JCN
// // // //
•f j ^ " 0
- ^ "uf
uO -H J O C
1 No.5.
ting pro
O 3
e 6
a, CO
1 1
CN ^ " V ^ rn ^ ^
1 UJ
p
^ c 0 T3 C E
UO F-H
O uC P «o S O Z M CO -i.S,
u O O 3 U H T3
ID CN LO
ro
0
^r
0 0 00 CN
O
ro CN
0 OD CO
0 CN
r
0 uO
cn
0 0 LO
0
4-J
0 3 •a ul-H
O
I/] I-H
T3
3
O u—1
CO
U4-U
0 4—1
ufH
O 00
E 0 I-H UrH
I-H
OO
CU
0 u-Ou a 0 •a
e OJ H
4—1
O
2
00
0 bo Cuj 1-u
O -C O
a Cuj cd aj _C
13 l-u 4-J
a O
rC - 4 — >
O 4->
T3
§ OO CD
b0 CuJ t-i
O uC
e 0
4-J
aj -JO
CJ 00
0 4-J
a Cuj 1-4
T3 u t — >
0 cs 00
3 W)
> 0 J3 CuJ
J3 E-H
Fig. 6.4 - Profile of cable ducts along short span of flat plate models
100
Chapter 6 : Test Program - An Introduction
H
o oo
o
c o TJ c <D a B ^ UO W
uo o kH
o &< Z «* rZ .S
*4-l
0 3 — P-. T3
C O •0
ut—»
g ^ g uo «j ° C uo O
. l-u
O O. Z &0 c
l-H O
r 0 "=
H-, -a
Ui
00 10 U u ^ ULt ' m t
w .\\ 1 v\
- 4s-I 1
0 CN
ut CN
1 1 /CO
/
jr • * /
. d 1—
X on l\ c*"1
\CN
1
- r 1
• - \4
O CN
CN CN
1 '/ /
.^/y ^s. x*^ 00 0 "V
u * 1 ULt
UJ
E E 0 00
O •a c 4J 4-1
6 £ uo
uo
tfl O
z cN
— 0
s 0 U Du M C •a 0 3
tu T3
~ — -H
c CuJ
— CL,
on A
c 0 •a a co
g E UO
q „ UO *4
00'
0 z S.
6 -4
3 -1
W)
2
O O
LO
O
0 LO
O 00 N" N"
O CN C>> ro
0 uO
ro 10
0 0 co CN
O N" CN CN
O CO uO
O CN
O uO LO
O
CJ
-1 T3 UM U
00 Uu OJ
•
(1) hn 03 Uu
O X3 O rl Cu
T3 _ C tul 3 CD
s.s 4_i 13 ^ ,b 00 c ufH «
o ° 4-1 CD
S5 o O 00 4-J
o « Ufa ^ "4H O0
S CuJ 100 o
^ -C
ufi ° U. CuJ
4-J
o
2
"cd
o
C
cd u T3 UJ
O
c
3 bfl
>
o CuJ
<0 -C
H
CiH-a
Fig. 6.5 - Profile of cable ducts along long span of flat plate models
101
Chapter 6 : Test Program - An Introduction
6.3.2 Prestressing and reinforcement details
The 5.05 mm diameter post-tensioning tendon has a mean tensile strength of 1500 MPa.
A typical stress-strain plot is shown in Fig. 6.6 . Each of the tendons in the model was
stressed to 75% of the tensile strength. This required a jacking force of 23.03 kN, 25%
of which was applied 24 hours after casting the models.
Stress (MPa) 2000 -j—
1600 -
1200 - /
BOO - /
400 - /
O -f , -| 1 1 '
O 10 20 30
Strain (x O.OOl)
Fig. 6.6 - Stress versus strain plot for 5.05 mm diameter prestressed tendon
In the design, the following allowances were made for the post-tensioning tendons:
• friction coefficient, \L = 0.20;
wobble coefficient, (3P = 0.015 radians/m.
A wedge draw-in of 3 mm was also allowed.
102
Chapter 6 : Test Program - An Introduction
The value of 0.20 used for the friction coefficient, ji, in the design was as recommended
by the concrete codes [ACI, 1989, BSI, 1985 and SAA, 1988] for galvanized steel duct.
This was described in Part (b) of Section 5.5.1 .
On the other hand, the specified wobble coefficient, Pp of 0.015 was slightly less than
the recommended minimum of 0.016 . This indicated that the flat metal ducting used
would caused less frictional loss because each ducting carried only a maximum of two
prestressing tendons in the design layout for all the models (see Section 6.3.1).
A wedge draw-in or slip of 3 mm was specified, which was half of the maximum slip
allowed of 6 mm as described in Part (e) of Section 5.5.1 . This was due to the half-
scale size used for all the models tested.
The additional non-stressed steel reinforcement provided for the models were all of 10
mm diameter high tensile steel with an ultimate tensile strength of at least 450 MPa. A
typical stress-strain plot for the reinforcement is depicted in Fig. 6.7 . The minimum
concrete cover to reinforcement was specified at 15 mm.
103
Chapter 6 : Test Program - An Introduction
Stress (MPa)
600
0 20 40 60 80 Strain (x 0.001)
Fig. 6.7 - Stress versus strain plot for 1 0 - m m diameter steel
reinforcing bar
The details of the tendons and steel reinforcement (in. the critical column strips in the
direction perpendicular to the slab edge, and spandrel reinforcement) are summarised in
Table 6.1 . It should be read in conjunction with Fig. 1.2 which outlines the various
widths of the column and middle strips of the models [SAA, 1988]. Fig. 6.8 reflects the
details as summarised in Table 6.1 .
104
Chapter 6 : Test Program - An Introduction
Table 6.1 - Summary of tendon and reinforcement details in the critical
column strips and spandrel beam
Model .
PI
P2
P3
P4
Column
strip
CS1 and
CS3
CS2
CS1 and
CS3
CS2
CS1 and
CS3
CS2
CS1 and
CS3
CS2
Slab steel
Top
10 mm @ 60 mm + 2
L 10 mm + 2
@ 50 mm
Bottom
10 mm @ 250 mn + 2 - straight bars
10 mm @ 250 mr + 3 - straight bars
Prestressing tendon
3 No. ducts @
250 m m - 5.05 mi tendons + 5
3 No. ducts @ 250 m m - 5.05 mi
tendons + 10
As in Model PI
As in Model PI
As in Model PI
As in Model PI
As in Model PI
As in Model PI
Spandrel steel
None
2-10 m m at top and bottom with
6 m m closed ties
@ 300 mm spacings for a
300 m m x 160 m m deep spandrel beam
As in Model P2 but for a 250 m m x
120 m m deep
spandrel beam
2-10 m m at top and bottom
with 6 m m closed ties @ 250 m m
spacings for a 300
m m wide torsion strip
105
Chapter 6 : Test Program - An Introduction
Corner Front face of column columns A & C \ / T°Psteel
Bottom steel
Duct with 1 tendon
Ducts with 2 tendons each
Fig. 6.8 - Tendon and reinforcement details at column strips of
post-tensioned concrete flat plate models
106
Chapter 6 : Test Program - An Introduction
6.3.3 Cable ducting and anchorage system
The cable ducts used were made of 0.4 mm thick galvanized steel. Each duct consisted
of a 45 mm width strip placed on top of a 25 mm x 15 mm square corrugated strip with
two 10 mm wide edges. The two components of the ducts were held together with
water-proofing adhesive tapes. This duct size was used in order to accommodate at
least two 5.05 mm diameter prestressing tendons. The details of the cable ducts are
depicted in Fig. 6.9 .
10 mm 25 mm 10 mm \m%\ •j-^ • m*fm •{
/
sealed with waterproof P V C tape
Fig. 6.9 - Shape and section view of cable duct
15 mm
107
Chapter 6 : Test Program - An Introduction
Twelve-mm thick steel bearing plates for the anchorage points were fabricated with two
6-mm diameter holes (at 35 mm apart) which were required to provide access for the
two prestressing tendons. Only one 6-mm diameter hole is needed for anchorage points
that span parallel and nearest to the edge of the slab since the ducts at these locations
were to accommodate only one tendon each. Fig. 6.10 shows details of the anchorage
system.
Grout vent r- 2 - 10 m m <j) perimeter bars (lap 400 m m )
0.5mm relief all round
Si 5 m m (j) prestressing tendon
Spacer 12 m m thick-bearing plate
(a) Section along cable
50 mm 100 mm
Barrels and wedges
- «)
6 m m fl) hole
1-tendon bearing plate
-M
6 m m (J) hole
2-tendon bearing plate
(b) Section a-a
Fig. 6.10 - Anchorage system
108
diopter 6 : Test Program - An Introduction
6.3.4 Concrete specifications
The ready-mixed concrete used to cast the model was designed to provide a
compressive strength of 22 MPa at transfer of applied post-tensioning and a
characteristic strength of 32 MPa at 28 days, although the 28-day strength for Model P4
was lower than expected at 27.0 MPa. Details of concrete strength for the models are
given in Table 8.3 in Chapter 8. The size of aggregates used in the concrete mixes for
all the models was specified as 10 mm.
Each of the concrete strengths given is the average value taken from the compression
tests of four 100 x 200 mm cylinders for Model PI but 150 x 300 mm cylinders were
used for Model P2, P3 and P4 . For Model PI, two batches of these cylinders were
taken near the edge of the model, i.e. strips AF and CH while the other two near the
centre of each panel. In Model P2 and subsequent models P3 and P4, only three
batches were taken, one from each of the column strips AF, BG and CH. A minimum of
three sets of test cylinders were set aside for testing for each model.
Typical measured slumps ranged from 80 to 100 mm for Model PI and 100 to 120 mm
for models JP2 , P3 and P4 .
6.3.5 Design of spandrel beam
A reinforced concrete spandrel beam was designed and constructed for Model P2 and
ModelP3 The design was in accordance with the Australian Standard AS 3600-1988
[SAA, 1988], The concrete specifications were the same as those specified for the slab
design.
109
Chapter 6 : Test Program - An Introduction
For a spandrel beam of 300 m m wide and 160 m m deep or twice the slab thickness (in
Model P2), the design required, the use of eight 6-mm diameter ties at 300 mm spacings,
with four 10-mm longitudinal bars at the four corners of the spandrel beam. The design
for a spandrel beam of 250 mm wide and 120 mm deep (for Model P3) was identical to
that for Model P2 , as the closed tie requirement for both models was already at a
maximum for the given design load of 8 kPa. Therefore, the design of a spandrel beam
for Model P3 was essentially the same as in Model P2 except for a shallower spandrel
beam for Model P3 .
The 10-mm longitudinal bars are similar to those used for the slab steel reinforcement, of
which the yield strengths are shown in Table 6.2 . It also includes the yield strengths of
the 6-mm closed ties used in both spandrel beams (in Model P2 and P3) and in the
torsion strip (in Model P4). Each of the yield strength values shown was the average of
tensile strength test results for at least three steel bars of a given type and size.
A typical stress-strain plot for the 6-mm diameter steel used as closed ties is shown in
Fig.6.11 while details of the spandrel beam are depicted in Fig. 6.12 .
110
Chapter 6 : Test Program - An Introduction
Table 6.2 - Yield strengths of steel reinforcement used for
longitudinal bars and closed ties in spandrel beam
Model
PI
P2
P3
P4
Yield strength of 10-mm
steel bars (MPa)
512
538
578
555
Yield strength of
6-mm steel wire (MPa)
-
450
489
541
Stress (MPa)
600
500
400 -
300 -
200 -1
100 -
O •' i | i | i | i |
0 10 20 30 40 50 60
Strain (x 0.001)
Fig. 6.11 - Stress versus strain plot for 6-mm diameter steel bar
111
Chapter 6 : Test Program - An Introduction
10-mm bars
160 m m spandrel beam depth
15 mm cover all round
80 m m slab thickness
6-mm ties at 300 m m spacings
Fig. 6.12 - Details of a typical spandrel beam (for Model P2)
111
Chapter 7 : Test Program - Construction, Instrumentation and Testing
CHAPTER 7
TEST PROGRAM - CONSTRUCTION, INSTRUMENTATION
AND TESTING OF MODELS
7.1 General Remarks
This chapter details the construction, instrumentation, and testing of the post-tensioned
concrete flat plate models. Each of the models was cast-in-place and the test up to
failure was carried out in a self-contained space reaction frame. The test set-up as
shown in Fig. 7.1 is identical to that for the reinforced concrete models tested previously
[Falamaki and Loo, 1992].
Fig. 7.1 - General set-up of a flat plate model
113
Chapter 7: Test Program - Construction, Instrumentation and Testing
Instrumentation in the form of load cells, strain gauges, and linear variable differential
transformers (LVDTs) were used to measure the forces and displacement data of the
models up to the failure stage. A whiffle-tree loading system was used to simulate a
uniformly distributed load being applied on all the test models.
7.2 Construction of Models
The construction of the models commenced with the setting up of the simulated column
support system and the formwork for the slab and spandrel beam (if any). Fabrication of
the cable ducts and the cutting and bending of the steel reinforcement were undertaken
at about the same time. These were followed immediately by the placement of cable
ducts (with the prestressing tendons) and steel reinforcement onto the formwork of the
slab.
After casting and curing of the slab, post-tensioning was applied onto the prestressing
tendons in two stages, the first to be applied 24 hours after casting and the second after
the concrete strength has achieved the required strength for full transfer of prestress.
Grouting was necessary to ensure that the prestresed tendons in the ducts were
sufficiently bonded to the concrete slab.
7.2.1 Setting up of formwork
Before the formwork was set up, the column sections were supported into position on
top of the supporting steel pedestal by five dummy load cells. This adjustable support
system allowed for the proper alignment of the columns both in the horizontal and
vertical directions. A typical column support system is as shown in Fig. 6.1 .
114
Chapter 7 : Test Program - Construction, Instrumentation and Testing
Standard marine formply was used to fabricate the slab and edge formwork. The
advantage of this material was that it could be re-used for the subsequent models. The
formwork was supported on extendable steel beams which in turn were supported by
two channel sections spanning in the north-south direction.
Fig. 7.2 shows details of the support system for the slab formwork for each of the
models while Fig. 7.3 presents the formwork and support system used for the spandrel
beam construction in models P2 and P3 . The base formwork for the spandrel beam is
propped up by adjustable steel props.
Before bolting the steel channels to the test frame, they were held in position with the
aid of adjustable steel props and levelled to an accuracy of 0.5 mm. The sheets of
formply used were cut into various sizes, the largest being about a quarter of the area of
a single panel of the model. Once they were put together into their final positions, any
gaps in between were sealed with silicon sealant.
The edge formwork to the slab was constructed around the slab perimeter, and fixed in
position by steel angles and bolts so as to resist the expected outward pressure of the
fresh concrete during casting.
Twelve 16-mm mild steel threaded rods were slotted through the bolt holes of the steel
column H and eight 16-mm bolts at the remaining steel columns. They extended above
the top of the slab and were securely fastened with nuts onto a steel plate at the top and
the prefabricated steel section at the base (see Fig. 6.1).
115
Chapter 7 : Test Program - Construction, Instrumentation and Testing
Edge form Plastic tubing form
Steel plate form welded onto steel column
100 m m deep extendable steel beam
Steel bracing
Support frame (bolted to pedestal)
Steel column
Dummy load cells
Pedestal
Fig. 7.2 - Details of slab formwork and the supporting frame
116
Chapter 7 : Test Program - Construction, Instrumentation and Testing
Top of slab 300 m m
Spandrel beam
Channel screwed onto spandrel formwork and welded onto T-Plate
10 m m steel T-pIate welded on to pedestal to support spandrel formwork
Pedestal
Fig. 7.3 - Typical details of formwork and support
system for spandrel beam at the slab edge
117
Chapter 7: Test Program - Construction, Instrumentation and Testing
P V C tubings of 27-mm internal diameter were slotted over these rods in order to
provide spaces allowing for concrete movements during prestressing. The tubing was
evenly spaced around the rods by having nuts at the top and bottom ends of the rods.
These tubings were removed once the concrete has sufficiently hardened. Steel plates
were welded around the sides of the columns as the required form for the truncated
concrete column at these locations.
7.2.2 Cable ducts
Galvanized steel was used to fabricate the cable ducts because it can be easily cut and
bent to the required shape and it was flexible enough to be compressed transversely in its
final form to fit into the thin slab. The galvanized steel sheet used was of 1.2 m by 2.4 m
and of 0.4 mm thick. Two different strip sizes, i.e. 45-mm and 75-mm wide were cut
from each sheet so as to facilitate the bending process to fabricate the cross-sectional
shape as shown in Fig. 6.9 . The strip for the top was joined to the bent strip section by
using waterproof PVC adhesive tape. A typical bending operation of the duct to its final
square corrugated shape is shown in Fig. 7.4 .
Once every length of the ducting (2.4 m) was fabricated, the ductings were riveted end-
to-end (where required) so as to make up the necessary spans of the slab in both
directions. All the joints were sealed with masking or PVC adhesive tape. This also
applied to any gaps along the length of the ducts. A typical fabricated cable duct is
shown in Fig. 7.5 .
118
Chapter 7 : Test Program - Construction, Instrumentation and Testing
© 75 m m
75 m m wide strip of 0.4 m m galvanized steel sheet
©
hvvu>
7\ /
/
© \s777i
10 mm
t
©
iW/
, ^
\7777f -+—f 10 mm . c square 15 mm ^
1
©
VZ&
u ^ /
/
XSSSA
10 mm
J©
©
0 E3 -r\—+- 10 mm , c square 15 mm ^
10 mm 10 mm
~L_T I 15 mm 25 mm,
Final shape of bent ducting
Fig. 7.4 - Bending process of duct to final shape
119
Chapter 7 : Test Program - Construction, Instrumentation and Testing
Fig. 7.5 - Typical fabricated cable duct
To facilitate the positioning of the wires through the two-hole anchorages, the end of the
ducts were cut and splayed. The ducts for the one-hole anchorage did not require such
treatment. The 12-mm bearing plates for the anchorages (see Fig. 7.6) were fabricated
with the required 6-mm diameter holes and a welded knob on the inner face. The knob
enabled the duct to slot easily onto the plate to the required position.
Following this, holes were cut near the end of the ducts to insert the plastic grout vents
(see Fig. 7.7), which were secured to the ducting with adhesive tape. These were
required for grouting the tendons.
120
Chapter 7: Test Program - Construction, Instrumentation and Testing
Fig. 7.6 - Typical 12-mm bearing plate
Fig. 7.7 - P V C grout vent for cable ducts
121
Chapter 7: Test Program - Construction, Instrumentation and Testing
Finally, the ducts were bent to their required multiple curve profiles (see Figs. 6.4 and
6.5). The ducts were bent to the required profiles after having being placed onto chairs
of selected heights on the formwork and then tied securely down by wires through the
slab formwork.
7.2.3 Placement of ducts and reinforcement
The layout of the cable ducts was marked out on the top surface of the formwork with
masking tapes. Then holes were drilled through the edge formwork at the appropriate
points in order to fix the bearing plates as well as to allow the prestressing tendons to
extend through the anchorages and edge formwork. The 12-mm bearing plates at the
anchorages were then screwed onto the inside of the edge formwork. It was important
to ensure that the bearing plates were flushed with the edge formwork and perpendicular
to the base formwork. This was to ensure that the prestressing was properly distributed
when applied.
Spacer chairs of various sizes were glued onto the formwork for the ducts to sit on so as
to provide for the required profiles of the cable ducts. Since the ducts were relatively
stiffened by its shape, at various points thin wires were necessary to tie the ducts down
to satisfy the concrete cover requirement as well as to prevent them from being
displaced during casting.
The laying sequence of the ducting was worked out in a way as described in Table 7.1 .
This was necessary in view of the stiffness of the ducts which would prevent vertical
clearance of ducts at overlapping points. The laying sequence was found to be effective
in overcoming this difficulty.
122
Chapter 7 : Test Program - Construction, Instrumentation and Testing
Table 7.1 - Laying sequence of cable ducts
Laying sequence
1
2-
3
•4
5
6
7
8
9
10
11
12
13
14
15
Duct number*
22
23
21
10
11
9
30
12
14
13
29
15
24
20
25
Laying sequence
16
17
18
19
20
21
22'
23
24
25
26
27
28
29
30
Duct number*
19
1
2
8
28
16
18
26
3
7
17
27
6
5
4
Once each duct was set into place, the tendons were pushed into the duct by slotting
them through one edge formwork and out from the other end. The tendons extended
about 300 mm out from both edges of the formwork to allow for the ease of post-
tensioning at a later stage (Fig. 7.8).
The bearing plates were completely wrapped to the ducting with the use of masking tape
and silicon sealant. Then rubber tubes of 16-mm diameter were fitted over the vent of
the grout plates for the grouting of tendons as well as to prevent ingress of concrete
during casting.
* The numbering of cable ducts is in accordance with Fig. 6.2
123
Chapter 7 : Test Program - Construction, Instrumentation and Testing
Fig. 7.8 - Tendons protruding from the edge formwork
The 10-mm diameter steel reinforcing bars were cut and bent where necessary to the
required shape. These steel reinforcing bars are then placed into their final positions
near the columns, at mid-spans and at the perimeter of the slab. At the column regions,
the steel reinforcing bars were secured by tying them to the cable ducts.
The positive reinforcement bars near the mid-spans of column strips were set onto chairs
to maintain the proper vertical clearances and were secured by tying to adjacent ducts.
The perimeter bars were overlapped accordingly and set as close to the edge as possible.
Typical details are shown in Figs. 7.9 and 7.10 .
124
Chapter 7: Test Program - Construction, Instrumentation and Testing
Fig. 7.9 - Top steel reinforcement over corner column A
Fig. 7.10 - Top steel reinforcement over edge column B
125
Chapter 7 : Test Program - Construction, Instrumentation and Testing
The typical layout for the 10-mm longitudinal bars and 6-mm closed ties in the spandrel
beam is shown in Fig. 11 .
Fig. 7.11- Spandrel beam reinforcement for Model P2
126
Chapter 7 : Test Program - Construction, Instrumentation and Testing
1.2A Casting and curing of concrete and removal of formwork
lust before casting, the top of the prefabricated steel section and inner sides of the steel
formwork were lubricated with grease. The PVC tubings for the column bolts were also
lubricated with grease. The former was to allow for some movement in the concrete
under prestress whereas the latter was to ensure easier removal of tubings after concrete
has hardened. Rubber tubes over the grout plates were sealed up with masking tapes to
prevent entry of concrete during casting.
Fig. 7.12 - Casting of concrete slab for flat plate model
The total volume of ready-mixed concrete required for the test model and the twelve test
cylinders was about 2.0 cubic metres. The ready-mixed concrete was pumped to the
slab through a 100-mm diameter hose. An internal (poker) vibrator was used to
127
Chapter 7 : Test Program - Construction, Instrumentation and Testing
consolidate the concrete. Extra precautions were taken to prevent damages to the strain
gauges attached on the reinforcement during the casting and vibrating processes (Fig.
7.12).
The surface of the slab was finished immediately with a wooden float. Two to three
hours after casting, the top surface of the slab was made smoother with a steel trowel.
There was slight bleeding of the concrete as the mix was deliberately made wet in order
to be more workable around the cable ducts.
About 24 hours after casting, the concrete was cured by covering the top surface with
moist hessian and plastic sheets. At the same time the edge formwork of the model was
stripped to facilitate the prestressing process.
Fig. 7.13 - Concrete test cylinders
128
Chapter 7 : Test Program - Construction, Instrumentation and Testing
The test cylinders were cast at the same time and cured in the same manner as the slab
(Fig. 7.13). They were vibrated internally and were cured in the same way as the flat
plate model. The cylinders were stripped after approximately eight days. Following
this, the plastic tubing forms were removed from around the column rods. Finally, the
strain gauges were checked to identify those which were not in working order.
7.2.5 Post-tensioning of prestressing tendons
According to the design specifications, 25% of the total jacking force was applied
approximately 24 hours after the slab has been cast. Note that this requirement could
not be carried out for Model PI due to a developed fault in the jacking equipment.
Nevertheless, the omission of this procedure would not cause severe effect due to the
half-scale size of the- model (in consultation with a Senior Engineer of VSL
Prestressing).
In the calibration of the hydraulic jack to be used in applying the required prestressing in
the model, it was determined that for a total jacking force of 23.03 kN per tendon, the
equivalent pressure from the jack was 2800 psi (see Appendix A). Therefore, 25% of
this was 700 psi. The tensioning was carried out only at the live end for each
prestressing tendon.
Before the prestressing can proceed, the barrels and wedges for the anchorages were
slotted in tightly against the exposed 12 mm plate along the sides of the slab. The steel
barrels and wedges used for anchoring the ends of the tendons are shown in Fig. 7.14 .
Unfortunately in the case of Model PI, during the process of applying the 25%
prestressing, the jacking equipment developed a fault. It was then decided that the 25%
129
Chapter 7 : Test Program - Construction, Instrumentation and Testing
prestressing has to be omitted for Model PI. A full prestressing was then applied 8 days
after casting when the fault in the prestressing jack has been rectified. The problem did
not occur in the other models, in which the recommended 25% prestressing was applied
as scheduled in each case.
Fig. 7.14 - Barrel and wedge used for anchorages
The prestressing was carried out in the required sequence so as to ensure an evenly
distributed tensioning in the wires during this process. The prestressing order was in
accordance with the cable duct numbers (see Fig. 6.2) as follows :
Initial order 1, 3, 5, 7, 9, 11, 13;
then 14 to 30;
and followed by 2, 4, 6, 8, 10, 12.
130
Chapter 7 : Test Program - Construction, Instrumentation and Testing
Fig. 7.15 - Post-tensioning of tendons
Fig. 7.15 shows the post-tensioning operation being carried out. It was important to
note that after applying the full prestressing, a barrier made of timber plank and steel
plate was erected around the perimeter of the slab as a precautionary or safety measure
against any prestressing failure before and during testing.
7.2.6 Grouting
The tendons of the model were grouted the following week after full prestressing.
Grouting was necessary to ensure that the tendons in the ducts are sufficiently bonded to
the concrete part of the slab.
131
Chapter 7: Test Program - Construction, Instrumentation and Testing
air was The grout was pumped in through the grout vent at one end of the duct and
allowed to seep out from the vent at the other end (see Fig. 7.16). A force of about 3
k N was used to pump the grout into the ducts.
Fig. 7.16 - Grouting of prestressed tendons
Wherever a duct was blocked by concrete seepage during casting, the grout was
pumped in from the other end as well to ensure complete grouting of the tendons. O n
completion of grouting the tendons in each duct, the rubber tubes were sealed off by
tying them with wires. At the same time, the gaps around the bolts in the columns were
grouted to provide monolithic joint to each of the column-slab connections.
132
Chapter 7 : Test Program - Construction, Instrumentation and Testing
7.3 Instrumentation of Models
Electrical resistance strain gauges were attached to selected reinforcing bars and
prestressing tendons at strategic locations of the models so as to record the strains
during the testing stage.
Fig. 7.17 - Hewlett-Packard 9826 computerised data acquisition system
133
Chapter 7: Test Program - Construction, Instrumentation and Testi ng
Further instrumentation was undertaken by the placement of three load cells at the base
of each column support. These were used to measure the three reaction components at
each of the pin supports.
The uniformly distributed loading to be applied during testing was simulated by the
installation of the whiffle-tree loading system at the bottom of the slab. This system
allowed for hydraulic rams at the base to apply the required (tension) loading which
were also measured by load cells.
All the strains and load cell readings were logged using a Hewlett-Packard 3054A data
acquisition control system via a Hewlett-Packard 9826 computer (Fig. 7.17). The
vertical deflections were also measured at critical locations of the slab (see Section
7.3.4).
7.3.1 Strain gauges
The steel strains were measured at numerous locations in the test structure with
electrical resistance strain gauges (TML-PL-10-11) having a gauge factor between 2.06
to 2.11 and a length of 10 m m . A total of 18 such gauges were mounted on the 10-mm
steel reinforcing bars in Model PI. These strain gauges which were placed at the critical
perimeter (i.e. at half the effective depth d/2 from the front face of the columns) and at
the mid-spans of the critical column strips for Model PL The strain values recorded
were used to determine the ultimate moments at these locations (see Section 9.3.3).
For subsequent models, in addition to sixteen strain gauges on steel reinforcing bars
(Fig. 7.18), more strain gauges were used so as to compensate for any failed or faulty
ones at later load stages. A further 27 gauges of type Nl l-FA-5-120-11 (with gauge
134
Chapter 7: Test Program - Construction, Instrumentation and Testing
factor of approximately 2.10 and gauge length of 5 m m ) , were m o u n t e d on the tendons
on specific locations as shown in Fig. 7.19 .
9ZL
N 9
0S9<3
S<3£1
•u*|tN
777 ^ z. zrzrz
mm -
=3
LU oo 3 a
a T^uN
S Z%2 ZX zrz o s
mm
C
s 3 O
u
=3 —1
0s! o 00 3 <N a r-l
O Oi
s. -IK •u^l<N
L?zei
HH
Oi
o
ON Pi
oo
O
oo
Pi
t-Ou
00 3 a
a
S£££
& ; )3 oi •% un Oi o oo 3
a
Oi
o 00 3
a oi
u
^1 w -H t3icN
CN Oi
<%
5 LU 00 3 a
o
ffl o o o uO
^ s ^ l
<!
s E 3 o U i
a cd u—< C/J
r
CD
M 3 Cut
a
-HK
Note : Strain gauges near columns are placed on 10-mm d,a. top reinforcement at
bottom reinforcement.
Fig. 7.18 - Locations of strain gauges on steel reinforcement
135
Chapter 7: Test Program - Construction, Instrumentation and Testing
X
•;';/
'•• r •yS.
>/."'.•
•.'••/•"••
it.
M
sX CN
CN
o .•• c
•f-C
_; O
4
4
CN CN
CN
S CN
*. ON
CN <£ r
vo
n 1—1
2° CO
CM
ET
ON
oo
O
_rv \o
o uO
o •ut
o CO
u CN
u
'."•_."
'.••_'..'
;.;/
;•;/
.;•;•>;•
.••'•.'•••"
'•'.'••'.•
•'.'•'••'.•'
o
u
u
: PQ
Cut
a c C/u
o
s o U
Fig. 7.19 - Locations of strain gauges on tendons in cable ducts
136
Chapter 7: Test Program - Construction, Instrumentation and Testing
Before the strain gauges were glued onto the steel bars or tendon, the selected locations
have to be polished smooth and shiny by sandpaper. Once the gauges have been glued
onto the surface of the steel, waterproof sealant was applied over them to protect from
grease and moisture to ensure the effectiveness of the gauges.
The end-connectors of all the gauges were soldered onto connecting electrical wires,
which in turn were to be soldered onto cables connecting to the portable strain meters or
the computerised data acquisition system. It was very important to check the wirings
and connections for these strain gauges to make sure that they were all in good working
condition prior to the casting of the slab.
7.3.2 Load cells
Three specially designed electrical load cells were provided at the pin support
(contraflexure points) at the base of each of the columns to allow for the recording of
the three reaction components of the columns. The steel sections simulating the in-situ
concrete columns were each supported by a steel pedestal via these electrical load cells.
These load cells which are shown in Fig. 7.20 were placed before the slab formwork was
stripped.
7.3.3 Whiffle-tree loading system
After the slab formwork was set up, PVC dowels with a diameter of 16 mm were
screwed to the formwork (from underneath). These were to provide for the 32 points in
the slab which facilitate the installation of the whiffle-tree loading system at the base of
the slab. This arrangement is shown in Fig. 1.2 . A total of 32 100-mm square pads
were set on top of the slab to secure the supporting steel rods for the loading system.
137
Chapter 7: Test Program - Construction, Instrumentation and Testing
Fig. 7.20 - Close-up view of load cells at the base of steel column
Fig. 7.21 - General view of whiffle-tree loading system
138
Chapter 7: Test Program - Construction, Instrumentation and Testing
The whiffle-tree loading system was essentially a series of criss-crossing steel beams in
an inverted pyramid set-up of ascending beam sizes from top to bottom. Fig. 7.21
shows the general set-up of the whiffle-tree loading system.
Two 20-tonne double-acting hydraulic rams which were attached to the whiffle-tree
system at the base of each slab panel, were used to apply the downward pulling loading.
To measure the load density, each ram was attached to an electrical load cell.
Synchronization of the rams was achieved by pumping through one single hydraulic
control system.
7.3.4 Deflection measurement
The slab deflections under applied load were measured at nine locations as shown in Fig.
1.2 . Rulers were hung on the soffit of the slab at the nine specified locations for
deflection measurement for Model PL In subsequent models, dial gauges were used in
addition to the hung rulers at similar locations (Fig. 7.22). These dial gauges were set
up on top of the slab.
Linear Variable Differential Transformers (LVDTs) were also installed on top of the slab
at centre of each panel for finer deflection measurement (Fig. 7.23). These LVDTs were
initially calibrated (before testing of models) in such a way that the voltage readings may
be directly converted to provide the equivalent deflection measurement in millimetre.
139
Chapter 7: Test Program - Construction, Instrumentation and Testing
Fig. 7.22 - Dial gauges set up on top of slab to measure deflections
Fig. 7.23 - Set up of the Linear Variable Differential Transformer (LVDT)
140
Chapter 7 : Test Program - Construction, Instrumentation and Testing
The L V D T s were calibrated by connecting them to 10 volts D C power supply. Then, a
voltmeter was connected to the DC power supply to get the individual DC output for
each LVDT. During the calibration, the displacement point of each LVDT was
displaced over a micrometre rule to obtain 10 displacement values ranging from zero to
50 mm. These measured values were calibrated and recorded against the corresponding
voltage readings obtained from the voltmeter. These set of readings were then plotted in
a linear relationship of voltage against displacement in millimetre.
The voltage readings for the LVDTs were recorded at each loading stage during testing
of the models. The obtained readings, which were in voltages, were then converted to
millimetre displacements by using the voltage-millimetre plots.
7.4 Testing of Models
7.4.1 During load application
The loading system was connected up onto a 20-tonne double acting hydraulic ram at
the base of the slab which has a synchronized control system for the rams (Fig. 7.24).
The Hewlett-Packard computer was run to record the initial set of data as the zero strain
readings prior to commencement of testing.
141
Chapter 7 : Test Program - Construction, Instrumentation and Testing
Fig. 7.24 - Set up of synchronization control system for the two
double acting hydraulic rams
initial deflection readings were taken with a level for the hung rulers (accuracy to 0.5
mm) and dial gauges (accuracy to 0.05 mm). The initial deflection readings for the
LVDTs were also recorded. This was carried out by obtaining the calibrated voltage
reading corresponding to zero displacement reading in each LVDT. By adjusting the
LVDTs' vertical displacement points up or down, the calibrated voltage readings were
obtained and recorded as the initial zero deflection readings before the testing
commenced.
For the subsequent sets of load data, the test loads were applied at an increment of
approximately 5% of the expected failure load:
142
Chapter 7: Test Program - Construction, Instrumentation and Testing
Design load = 8 kPa.
i.e. load increment = 0.05 x 8 = 0.4 kPa
which is equivalent to: D-
0.4 kPa x ( 5.7 m x 3.53 m slab ) = 8 kN, say 10 kN.
This was done by raising the hydraulic pressure in the loading rams to the required level.
At every load stage, the applied load was held constant so that data for column
reactions, load densities, strain measurements and deflection readings were recorded.
The Hewlett-Packard computer was run to check the level of load density applied on the
model. Once the load density in the printout matched the required load level at that
stage, the computer was run to record the required set of column reactions and strain
data. At the same time, the deflection readings were taken at all the selected points.
Observations and markings were also made of any cracks formed as a result of the
applied load increment.
Generally, the first cracks occurred at around load stage 6 ( i.e. 60 kN ) for each of the
model. As the applied load neared the 100 kN mark, the subsequent load increment was
reduced from every 10 kN to every 5 kN until the model finally failed.
7.4.2 After load application
Load increment was continued up to slab failure whereby all recorded data were printed
out as hardcopy. The hydraulic system was then disconnected at the end of the testing.
A post-mortem examination of the test model was conducted whereby the pattern, size,
143
Chapter 7 : Test Program - Construction, Instrumentation and Testing
location and type of cracks before and after failure were identified and recorded (Fig.
7.25).
Fig. 7.25 - Post-mortem examination of test Model P3 after failure
7.4.3 Precautions prior to and during testing
After the full prestressing of the tendons, a barrier of timber planks and steel plate was
erected around the perimeter of the slab as a safety measure against any prestressing
failure before and during testing. This has to remain in place throughout the test of each
model.
For safety reasons, only authorised persons (with hard hats) were allowed to step on top
of and below the slab in between load stages, to mark the crack patterns. The hydraulic
144
Chapter 7: Test Program - Construction, Instrumentation ami Testing
pressure for the loading rams in between load stages was maintained steady (by an
experienced laboratory technician) while marking of cracks and deflection readings were
taken.
7.4.4 Problems encountered during model testing
There were insufficient number of nuts being screwed and tightened on the bottom
threaded ends of steel rods (for the whiffle-tree loading system). Consequently, this
resulted in the collapse of the loading system during the testing of Model P3 near the
ultimate load stage, requiring a resumption of the test the following day.
The loading system also failed during the testing of Model P4 , but this time due to the
wom-off end of a threaded steel rod (for the whiffle-tree loading system). This also
resulted in the test being carried out over two days.
These setbacks did not affect the test results recorded over a two-day period, as can be
seen in the deflection, stress and strain measurements in Appendices B, D and F. The
models exhibited sufficient repeatability to enable a repeated load to be applied from the
beginning until they reached the ultimate (beyond the last applied load on the first day of
testing).
145
Chapter 8: Presentation and Discussion of Test Results
CHAPTER 8
PRESENTATION AND DISCUSSION
OF TEST RESULTS
8.1 General
This chapter presents the test results obtained for slab deflections, unbalanced slab
moments to be transferred to the columns, as well as punching shear strength and the
crack propagation patterns of the models. In light of the test results, the ultimate
punching shear behaviour of post-tensioned concrete flat plates with or without spandrel
beams is also discussed.
The laboratory study has produced experimental data not hitherto available on the
behaviour of post-tensioned concrete flat plates with bonded tendons. These include the
load deflection plots at nine locations in each of the two-panel six-column slabs, the slab
crack patterns, the failure loads, the punching shear results and the ultimate unbalanced
slab moments.
This information provides an insight into the complicated punching shear failure
behaviour of the slab-column connections at the corner- and edge-column positions.
The data will be useful in an attempt to extend the punching shear analysis of concrete
flat plates with spandrel beams [Loo and Falamaki, 1992] to include the effects of
prestressing.
146
Chapter 8: Presentation and Discussion of Test Results
8.2 Crack Patterns and Failure Mechanism
8.2.1 Model PI
The crack patterns recorded at the top surface and the soffit of Model PI are shown in
Figs. 8.1 and 8.2 respectively. It may be seen that the crack patterns around the corner
columns A and C and the edge column B are characteristics of punching shear crack
failure. This is particularly obvious from the pattern on the top surface (Fig. 8.1).
The failure at each of the three columns is identified as the shear failure mode [Falamaki
and Loo, 1992] for which the failure zone takes the form of a truncated cone: for the
edge column it was a half cone; for the corner columns it was a quarter cone. The top
view of the failure cone for the corner column A may be seen in Fig. 8.3 . Note also that
most of the positive moment cracks were developed at the failure load which together
with the punching shear cracks formed the collapse mechanism.
The numbers marked along the cracks indicate the load stages during which the Viarious
cracks first occurred; and the letter F signifies the cracks recorded after failure.
147
Chapter 8 : Presentation and Discussion of Test Results
CM
ffi
r- cu
stage
ho ' '
flS *
u
C7M
r^
u.
U.
LU
IU
""UT
•ut
Crul
UTILMUT OM \
Fig. 8.1 - Crack pattern on top and sides of Model PI
148
Chapter 8 : Presentation and Discussion of Test Results
t-M
r-cu
Cut
u
•J
a) ii =3 u|-C
CU
3 3 r- U-i
|,JL. OS.1
Fig. 8.2 - Crack pattern at soffit and sides of Model PI
149
Chapter 8 : Presentation and Discussion of Test Results
Fig. 8.3 - Shear failure (Model PI - column A )
8.2.2 Model P2
Of Model P2, the crack patterns are recorded in Figs. 8.4 and 8.5 respectively for the
top and bottom surfaces of the model. It is clear that positive yield lines were developed
at the soffit along the midspans in both the longitudinal and transverse directions.
These were accompanied by a distinct negative yield line developed along the joint
between the slab and the spandrel beam (Fig. 8.6). Negative moment cracks also
appeared around columns F, G and H as well as along the line joining the two central
columns (B and G). These enabled the formation of the distinctly flexural collapse
mechanism in each of the two slab panels.
150
Chapter 8: Presentation and Discussion of Test Results
LU
3 3 o c o a CO JLU.
w u
~i? 12s ho \s
Crack (= 4 mm) Spandrel beams
N O
£0 C — Q. CO
s /. r
? s
• l i
H .11.
o ca u. U
\,
OJ
.1 tn •o Cut
1-3 bo '
.3 •a u H lu ID
^ 1
3 tu
a / . -"3
Fig. 8.4 - Crack pattern on top and sides of Model P2
151
Chapter 8: Presentation and Discussion of Test Results
r-- cu
•3
R !ul 1>
J3
S Ou 00
m / "*S
r*. / -•
i>
^ ^
r V *""* CM
<o —i^oo m
-OM « " ^
.-H
*n '—' _' —
«n
n. Nr,
Vi * — t ^ "
y 00
CO .—1
u£>
L? .2^2 0C>V
—H m-4
» V 00
vo - — i-H
*<• ->
^"2 ^
12 8 V ^ ^ 1515 A T 15tf ( 8
F \5 14^15^/4
vWUio?- > 15 7 10 8
Wo 7W0
Fig. 8.5 - Crack pattern at soffit and sides oi Model P2
152
Chapter 8: Presentation and Discussion of Test Results
Fig. 8.6 - Flexural failure (Model P2)
8.2.3 Model P3
In Model P3, the first positive yield line occurred at the soffit of each slab panel midway
between the columns and parallel to the spandrel beam. Further increases in load led to
the meeting of these yield lines. The formation of cracks at the midspan of the slab
soffit, extended gradually to the edges in the transverse directions. The increase in load
also cause crack formations at the vertical sides of the spandrel beam and slab. These
153
Chapter 8: Presentation and Discussion of Test Results
cracks were clearly the result of torsion in the spandrel. In addition, cracks were also
observed at the corner of the columns and then across the front face of columns,
particularly the corner-columns A, C and the edge-column B .
Model P3 finally failed by punching shear at the corner- and edge-column positions,
after the formation of a typical large yield line crack across the positive flexural
reinforcement at the midspan of the slab. Fig. 8.7 shows a typical punching shear failure
at the edge-column B of the model. The shear failure is clearly identified by the crack
pattern on the top of the slab as shown in Fig. 8.8 . The crack pattern of the slab soffit
is illustrated in Fig. 8.9 .
Fig. 8.7 - Shear failure {Model P3 - column B)
154
Chapter 8: Presentation and Discussion of Test Results
16 P^7 L
10 -15 mm crack width
Fig. 8.8 - Crack pattern on top and sides of Model P3
155
Chapter 8 : Presentation and Discussion of Test Results
Fig. 8.9 - Crack pattern at soffit and sides of Model P3
156
Chapter 8 : Presentation and Discussion of Test Results
8.2.4 Model P4
The crack pattern of Model P4 is very similar to Model PI, as both were constructed
without spandrel beams although Model P4 had a torsion strip at the edge spanning from
column A to column C. The punching shear failure occurred first at the edge-connection
B (see Fig. 8.10), followed closely by similar failures at columns A and C as illustrated in
the crack patterns at the top of the slab in Fig. 8.11.
As in the failure mechanism of Model P3, the first positive yield line in Model P4
occurred at the soffit of each slab panel midway between the columns and parallel to the
torsion strip. As the applied load increases these yield lines joined together, and cracks
at the midspan of the slab soffit also extend gradually to the edges in the transverse
directions (see Fig. 8.12).
Fig. 8.10 - Shear failure {Model P4 - column B)
157
Chapter 8: Presentation and Discussion of Test Results
2 V ^ - F ,18 J,
*\
19
s
[ Punching shear A Fi crack
Spalling of I Punching concrete surface v crack
shear
Fhinching shear
crack 22! ^ 2
Fig. 8.11 - Crack pattern on top and sides of Model P4
158
Chapter 8: Presentation and Discussion of Test Results
[u~ 0)
xn <*-*
1 3 3 F
23
t Liu
241
R 24
18
XL
"^T 19 ^W 20
Fig. 8.12 - Crack pattern at soffit and sides of Model P4
159
Chapter 8: Presentation and Discussion of Test Results
The first sign of punching shear failure occurred in the vicinity of the edge column B, in
which the crack width at the front face and corners of the column increased significandy
at failure. This was immediately followed by similar failures at the comer columns A and
C.
8.3 Deflections
The slab deflections under applied load were measured at nine locations as defined in
Fig. 8.13 . The deflection results for models PI, P2, P3 and P4 are detailed in Tables
B.l, B.2, B.3 and B.4 (Appendix B); the load-deflection plots up to failure load for
Model PI are presented in Figs. 8.14 . The corresponding load-deflection plots for
models P2, P3 and P4 are given in Figs. 8.15, 8.16 and 8.17 .
8.3.1 Model PI
For Model PI, the maximum deflection was measured at location 3 which was the centre
of panel ABGF with a deflection reading of 46.7 mm at the ultimate load. This was
followed by deflection readings of 45.3 mm and 42.0 mm at location 8 and location 5
respectively.
F| O
1 3 e o
_ _4 A ] - O
0. 5 e
B
7 ©
8 o
9 e
H
6 e
c
4 O Locations for deflection measurement
Fig. 8.13 - Deflection measurement stations
160
Chapter 8: Presentation and Discussion of Test Results
0 10 20 PO 40 50
Deflection (mm)
Fig. 8.14a - Load-deflection plots-
Model PI (Locations 1, 5 and 6)
O 10 20 30 40 50
Deflection (mm)
Fig. 8.14b - Load-deflection plots-
Model PI (Locations 2, 3 and 4)
0 10 20 30 40 50
Deflection (mm)
Fig. 8.14c - Load-deflection plots - Model PI (Locations 7, 8 and 9)
161
Chapter 8: Presentation and Discussion of Test Results
A survey of Fig. 8.14 indicates that the deflection is linear up to an applied load of 10.0
kPa for all locations. This is just above the design superimposed (service) load of 8 kPa.
8.3.2 Model P2
Similarly, the applied load versus deflection curves for Model P2 are shown in Fig.
8.15. In this case, the deflection is elastic until the applied load approached 10.5 kPa.
In the model, the maximum deflections occurred at location 8 (66.7 mm) and location 3
(66.3 m m ) , the centre of panels BCHG and ABGF respectively. Due to the stiffening
effect of the spandrel beam, the deflections at the edges of the two slab panels were
small all the way up to the failure load stage. This is evident in the plots for locations 4
and 9 in Figs. 8.15b and 8.15c .
A\i
PH
lied
ON
H *—r— 1 1
*^y
-o-l
-*-5
i
o
i
_
0 10 20 30 40 50
Deflection (mm)
Fig. 8.15a - Load-deflection plots
Model P2 (Locations 1, 5 and 6)
~ i — | — i — | — I 1 1 1 r
0 10 20 30 40 50 60 70
Deflection (mm)
Fig. 8.15b - Load-deflection plots
Model P2 (Locations 2, 3 and 4)
162
Chapter 8: Presentation and Discussion of Test Results
0 10 20 30 40 50 60 70
Deflection (mm)
Fig. 8.15c - Load-deflection plots - Model P2 (Locations 7, 8 and 9)
8.3.3 Model P3
For Model P3, the applied load versus deflection curves shown in Fig. 8.16 exhibit
similar pattern to that of Model P2. The six superimposed curves shown in Fig. 8.16 are
the load-deflection plots of deflection readings taken during the first and second day of
testing of the model. This was due to the failure of the loading system on the first day of
testing. As with the first two models, the deflection is elastic until the applied load
approached 10.0 kPa mark. As expected, the maximum deflections (combined first and
second day readings) occurred at location 3 (93.1 m m ) and location 8 (72.5 m m ) , the
centre of panels ABGF and BCHG respectively.
163
Chapter 8 : Presentation and Discussion of Test Results
Again, the stiffening effect of the spandrel beam caused the relatively small deflections at
the edges of the two slab panels up to the failure load stage, as indicated in the plots for
locations 4 and 9 in Figs. 8.16b and 8.16c.
cd
PH
M T3
a o u
PH
<
LIU
ed 2U PH M T3 O
V
PH 1 Q
O H <
o -c • i
JrlWil /I2J7
1
-•
-B-2 (Day 1) -*-•* (Day 1) —0-4 (Day 1) —*-2 (Day 2) -n-3 (Day 2) -*-4 (Day 2)
i i
10 20 30 40
Deflection (mm)
10 20 30 40 50
Deflection (mm)
Fig. 8.16a - Load-deflection plots
Model P3 (Locations 1, 5 and 6)
Fig. 8.16b - Load-deflection plots
Model P3 (Locations 2, 3 and 4)
164
Chapter 8: Presentation and Discussion of Test Results
J U
la 2 0
PH
« 15 _o
Applied
I-L
i i
0 -c
frm
lip
« I
/
v/
1
•
-a-7 (Day 1) -•-8 (Day 1) -H5-9 (Day 1) —•—7 (Day 2) - 0 8 (Day 2) —*—9 (Day 2)
i i
0 10 20 30 40 50
Deflection ( m m )
Fig. 8.16c - Load-deflection plots - Model P3 (Locations 7, 8 and 9)
8.3.4 Model P4
The applied load versus deflection curves for Model P4 are shown in Fig. 8.17. Again,
due to the failure of the loading system on the first day, the testing was resumed the next
day. This resulted in the superimposition of six curves for each load-deflection plot in
Fig. 8.17 . It can be seen that the deflection is elastic until the applied load has just
passed an applied load of 11.0 kPa. The maximum deflections (combined first and
second day readings) are at location 3 (132.0 m m ) and location 8 (85.0 m m ) , the centre
of panels ABGF and BCHG respectively.
165
Chapter 8: Presentation and Discussion of Test Results
cd
PH
•*3 ca o
20
15
10
•a 4H "P< PH
<
•
- ITT
-I
T ' 1
-B-
—o—
•
I (Day 1) 5 (Day 1) 5 (Day 1)
1 (Day 2) 5 (Day 2) 6 (Day 2)
1 r
20 40 60 80
Deflection (mm)
40 80 120 160 Deflection (mm)
Fig. 8.17a - Load-deflection plots
Model P4 (Locations 1, 5 and 6)
Fig. 8.17b - Load-deflection plots
Model P4 (Locations 2, 3 and 4)
20
30 60 Deflection (mm)
90
Fig. 8.17c - Load-deflection plots - Model P4 (Locations 1, 8 and 9)
166
Chapter 8: Presentation and Discussion of Test Results
It should be noted that the deflection measurements were recorded over two days during
the testing of the slabs Model P3 and Model P4 due to the sudden failure of the whiffle
tree loading system at an applied load of 21.0 kPa and 12.7 kPa, respectively. The
models were re-loaded the next day and all measurements were re-taken from the
beginning up to failure. The deflection results shown for Model P3 and Model P4 are
the measurements taken during the two days of testing of each model up to failure.
8.4 Punching Shear Strength and Unbalanced Moments
8.4.1 General
The construction and test dates for the four models are given in Table 8.1 which also
shows the average concrete cylinder strength test results.
. The punching shear strength, Vu , is defined as the ultimate column reaction at the
contraflexure point. The vertical load cell installed at the bottom of each test column
(see Fig. 6.1) provided the records for this reaction. Note that the three exterior
connections A, B and C (see Fig. 1.2) of each model were designed to fail at the same
ultimate load. For all practical purposes, this objective was achieved for all the models.
Table 8.1 - Concrete strengths of half-scale models
Model
PI
P2
P3
P4
Date of casting
01-10-91
08-07-92
28-09-93
05-05-94
Date of testing
06-11-91
09-09-92
26/27-10-93
09/10-6-94
Average cylinder strength, M P a
35.7
33.9
47.8
27.0
167
Chapter 8 : Presentation and Discussion of Test Results
8.4.2 Experimental results for shear, moment and torsion
The measured values of the unbalanced moments in the main and transverse bending
directions to be transferred to the column centre are termed as Mci and MQ2
respectively. The values of MQI and Mc2 are important parameters in the punching
shear strength analysis of concrete flat plates [Loo and Falamaki, 1992]. These moment
values were obtained as the products of the relevant horizontal column reactions and the
vertical distance between the column contraflexure point and the mid-depth plane of the
slab (Lj in Fig. 5.2). The horizontal reactions were measured using the corresponding
load cells located horizontally, as shown in Fig. 6.1 .
The measured stresses (for shear, moment and torsion) recorded at the three exterior
connections of Model P2 , under uniformly distributed load up to failure, are presented
in Table 8.2 . The recorded moment and torsion results refer to the measured
unbalanced moment transfer to the columns, MCj and MC2 respectively whereas the
recorded shear was the vertical reaction measured at the column until the ultimate value
of Vu , was reached.
Some plots of the experimental stresses (namely, moment, shear and torsion) against the
applied load up to failure, as recorded at the three exterior connections for Model P2 are
presented in Fig. 8.18 .
168
Chapter 8 : Presentation and Discussion of Test Results
Table 8.2 - Experimental shear, m o m e n t and torsion recorded at corner-
connections A and C and at edge-connection B for Model P2
Applied
load
(kPa)
0
1.37
3.28
4.37
5.69
6.96
9.86
11.27
12.48
13.21
14.69
15.86
16.07
17.23
18.44
19.61
Shear (kN) around column
A
0
3.01
4.23
6.11
7.86
9.91
14.41
17.73
19.74
21.18
24.07
26.08
26.64
28.52
30.23
32.02
B
0
10.20
13.79
18.34
23.82
28.89
38.53
43.58
47.96
50.92.
56.57
60.73
60.95
65.89
71.01
76.33
C
0
3.36
4.64
6.10
8.32
10.00
14.91
17.65
19.86
20.66
23.27
25.00
25.48
27.39
29.16
32.97
Moment (kN-m) at column
A
0
0.52
0.75
1.20
1.64
2.29
4.49
6.56
7.70
8.68
10.17
11.17
11.57
11.88
11.95
12.01
B
0
2.44
3.41
4.73
6.32
7.90
10.35
12.47
14.62
19.54
23.84
24.75
25.61
26.55
27.51
28.50
C
0
1.52
2.04
2.66
3.50
4.24
5.30
6.04
6.77
7.25
7.94
8.67
8.80
9.69
10.47
11.17
Torsion (kN-m) at column
A
0
1.15
1.68
2.34
3.04
3.79
4.85
5.45
5.68
5.87
6.15
6.31
6.59
7.06
7.12
7.35
B
0
0.04
0.05
0.04
0.05
0.07
0.14
0.14
0.10
0.15
0.09
0.14
0.17
0.27
0.32
0.28
c 0
0.05
0.30
0.62
1.14
1.57
3.34
4.59
5.46
5.92
7.49
8.11
8.26
8.71
9.19
9.51
It may be observed from Table 8.2 and Fig. 8.18 that the recorded moments and shears
at the edge-connection B were comparatively higher than both corner-connections A and
C due to the higher proportion of the applied load being taken up by connection B.
Conversely, the recorded torsion was the least at edge-connection B with the net
transverse unbalanced moments at or near zero due to the near symmetry of the two-
panel flat plate models. These trends in the stresses are generally observed in the other
models as well. Similar plots of stresses versus applied load for models PI, P3 and P4
169
Chapter 8 : Presentation and Discussion of Test Results
are presented in Appendix C. The corresponding tabulated results for these plots are
shown in Appendix D.
30 - Column A • Column B • Column C
T 0 5 10 15 20
Applied load (kPa)
(a) Plot of moment transfer to columns
A, B and C versus applied load up
to failure
O 5 IO 15 20
Applied load (kPa)
(b) Plot of shear strength at columns A,
B and C versus applied load up fo
failure
Fig. 8.18 - Experimental shear, moment and torsion recorded at corner-
connections A and C and at edge-connection B for Model P2
(continued next page)
170
Chapter 8 : Presentation and Discussion of Test Results
10 -°— Column A -*— Column B -*— Column C
O 5 IO 15 20
Applied load (kPa)
(c) Plot of torsional moment at columns A, B and C versus
applied load up to failure
Fig. 8.18 - Experimental shear, moment and torsion recorded at corner-
connections A and C and at edge-connection B for Model P2
(continued from previous page)
As a result of the ultimate load tests of four post-tensioned concrete flat plate models,
twelve punching shear strength results were available from the four models tested.
These are presented in Table 8.3 together with the measured values of Mci and Mc2
and the measured failure loads.
171
Chapter 8: Presentation and Discussion of Test Results
Table 8.3 - Measured punching shear, unbalanced moments and applied
loads at failure for post-tensioned concrete flat plate models
Connection
PI-A
Pl-B
Pl-C
P2-A
P2-B
P2-C
P3-A
P3-B
P3-C
P4-A
P4-B
P4-C
Measured
vu,m
26.93
68.75
27.72
32.02
76.33
32.97
30.97
73.35
31.78
27.64
66.43
28.22
Measured
MC1,
kN-m
7.39
15.25
8.70
12.01
28.50
11.17
12.25
33.30
13.54
7.34
16.42
9.55
Measured
MC2,
kN-m
7.26
0.24
6.14
7.35
0.28
9.51
7.29
0.28
8.23
6.31
0.13
5.10
Measured
failure load,
kPa
16.18
19.61
22.42
16.64
8.4.3 Failure modes at the corner- and edge-column connections
Models PI and P4 yielded smaller punching shear strength values for .all corner- and
edge-column connections, compared to that recorded for models P2 and P3 . This is
mainly due to the additional stiffening effect provided by the presence of spandrel beams
along the edge in models P2 and P3 .
* All numbers that end with A or C are corner connections; those that end with B are edge connections.
Unlike models P2 and P3, models PI and P4 did not have spandrel beams at the edge.
172
Chapter 8: Presentation and Discussion of Test Results
Model P2 underwent significant tendency towards a flexural failure mechanism (as
shown in Section 8.2) prior to the failure stage. At this stage, punching shear cracks
have developed well in advance particularly around the regions of the interior columns
as well as the coiner- and edge-column connections A , B and C . Therefore, when the
model finally failed, the corner- and edge-columns also failed almost simultaneously in
punching shear as a response to a sudden shift or re-distribution of the applied loading to
the exterior column supports. For all intended purposes, the comer- and edge-column
connections in Model P2 have failed in punching shear, albeit indirectly.
The failure load in Model P3 is slightly higher than that for Model P2, although Model
P3 has a smaller spandrel beam (of 250 mm wide by 120 mm deep) compared to that for
Model P2 (of 300 mm x 160 mm). The reason is that the compressive strength of
concrete for Model P3 is much higher at 47.8 MPa (see Table 8.1). Nevertheless,
Model P3 also failed in punching shear at the comer- and edge-column connections,
with punching shear results which are slightly less than that of Model P2 (see Table 8.3).
This is due to the shallower and narrower spandrel beam in Model P3 which has a lesser
stiffening effect in enhancing the punching shear resistance at these exterior column
connections.
8.4.4 Comparison of measured vertical column reactions and total applied load
As a check for the accuracy of the loading system used, the total applied load are
compared with the measured vertical column reactions in Tables 8.4, 8.5, 8.6 and 8.7 for
all the test models. It may be seen that the error is less than two percent in all cases.
This confirms the good accuracy of the loading and recording system.
173
Chapter 8: Presentation and Discussion of Test Results
Table 8.4 - Comparison of vertical column reactions and applied loads for
Model PI
Connection
Pl-A
Pl-B
Pl-C
Pl-F
Pl-G
Pl-H
ZVM =
V M,kN
26.93
68.75
27.72
33.55
71.76
30.48
259.19
Panel
North
South
ZP =
Applied load at failure, P, kN
130.35
131.09
261.44
Table 8.5 - Comparison of vertical column reactions and applied loads for.
Model P2
Connection
P2-A
P2-B
P2-C
P2-F
P2-G
P2-H
zvM =
V u,kN
32.02
76.33
32.97
32.44
72.28
34.25
280.29
Panel
North
South
•LP =
Applied load at failure, P, kN
139.58
141.51
281.09
Chapter 8: Presentation and Discussion of Test Results
Table 8.6 - Comparison of vertical column reactions and applied loads for
Model P3
Connection
P3-A
P3-B
P3-C
P3-F
P3-G
P3-H
ZV H =
V M , k N
30.97
73.35
31.78
38.71
84.46
36.45
295.72
Panel
North
South
ZP =
Applied load at
failure, P, k N
148.32
149.46
297.78
Table 8.7 - Comparison of vertical column reactions and applied loads for
Model P4
Connection
P4-A
P4-B
P4-C
P4-F
P4-G
P4-H
ZV„ =
Vu,kN
27.64
66.43
28.22
29.59
66.70
25.39
243.97
Panel
North
South
EP =
Applied load at
failure, P, k N
119.38
120.40
239.78
175
Chapter 8: Presentation and Discussion of Test Results
8.5 Observations
The failure mechanisms of the flat plate models are found to be greatly influenced by the
presence of the spandrel beam which enhanced the punching shear strength at the edge-
and comer-column positions. Spandrel beams of smaller width and depth possess less
stiffening effect to improve the punching shear resistance at the exterior column
positions.
The stiffening effect of the spandrel beam also contributes to smaller deflections at the
edges of the two slab panels up to the failure load stage.
176
Chapter 9: Determination of Ultimate Flexural Strength
CHAPTER 9
DETERMINATION OF ULTIMATE FLEXURAL STRENGTH
OF CRITICAL SLAB STRIPS
9.1 General Remarks
The critical slab strips (as briefly described in Section 5.4) are referred to slab strips of
width equal to the width of the corresponding column support, and span in the main
bending direction perpendicular to the spandrel beam.
As input to the analysis of punching shear Vu , the ultimate flexural strengths of the slab
strips are required to be calculated for the post-tensioned concrete flat plate models.
Based on a rational and well-established flexural strength theory, the strain-moment
relationship of a prestressed concrete section as described in Section 5.4 is used to
determine the ultimate moments Mj and Mm , at the front face of the critical perimeter
and the mid-span of the critical strip respectively. Procedures for determining the
moments are outlined for both uncracked and cracked sections of the slab strips.
9.2 Flexural Behaviour of Uncracked and Cracked Sections
In order to determine whether the section is uncracked or cracked, the applied and
cracking moments, M0 and Mcr , respectively are calculated. M0 is calculated by using
the direct design method of analysis [SAA, 1988]. The moment to be resisted by the
section, when the first crack is formed, is defined as the cracking moment, Mcr . It can
be determined from the expression:
177
Chapter 9 : Determination of Ultimate Flexural Strength
Kr = U'cf + °bp] (9.1}
where abp is the maximum compressive fibre stress of the section due to the applied
prestress and is given by:
Pe ?ee
°bp A; + 1Ty' -(9-2)
in which Pe and e are the effective prestressing force and the eccentricity of the
prestressed tendon respectively;^ is the distance of the centroidal axis to the extreme
tensile fibre; fcf is the tensile stress of the concrete section. According to AS 3600-
1988 [SAA, 1988],
fcf = 0.6 yjfc (9_3)
Further, It and A t are the moment of inertia and area of the transformed cross-section
respectively where,
2
It=~f- + bDs[yc- ^ - } 2 + (np-l)Ap[yc-dp}2
2 , /„ rs A r r -, 2 + in-I)An\yc-dsY + {n-l)Asc[yc-dsc}2 ...(9-4)
and
4 = bD, + (np-l)Ap + {n-l)Ast + {n-l)Asc ...(9-5)
The various terms used in Eqs.(9-4) and (9-5) are described as follows:
178
Chapter 9 : Determination of Ultimate Flexural Strength
n = the modular ratio of steel reinforcement (i.e. E/Ec);
np = the modular ratio of prestressed tendons (i.e. Ep/Ec);
Es , Ep and Ec = the moduli of elasticity of reinforcing bars and the tendon, and
the concrete modulus, respectively;
Ast, Ap and Asc = the area of tensile steel reinforcement, prestressed tendons, and
the compressive steel reinforcement (if any), respectively;
b and Ds = the width and depth of the section, respectively;
ds, dp and dsc = the depth of the tensile steel, prestressed tendons and the
compressive steel, with respect to the extreme compressive
fibre, respectively; and
yc = the depth of the centroidal axis of the uncracked transformed
section, which is,
r n 2
(n-1) [Ast ds + Asc dsc] + (n -1) [A d ] + —±-
(n-1) [Ast + Asc] + (np-l)Ap + bDs
...(9-6)
If M0 exceeds Mcr, then the section is cracked, and the analytical method as outlined in
Section 9.2.2 is used. Otherwise, an uncracked section analysis is utilised as described
below.
9.2.1 Uncracked section analysis
The analysis of an uncracked section involves the calculation of the moment resisted by
the transformed section. That is,
M = abp -^ -(9-7) yt
179
Chapter 9 : Determination of Ultimate Flexural Strength
in which the various terms are as defined earlier.
The moment of the uncracked section may also be determined experimentally through
the use of strain measurements of the prestressed tendons, Zp. The strain is then applied
to Eq.(9-2) to determine abp , [Gilbert and Mickleborough, 1990] or
„ - £ A (*, - D Ap S A {np -l)Ape
This experimentally determined value can then be used in Eq.(9-7) to obtain the
required moment M.
9.2.2 Cracked section analysis
When the tensile stress produced by the external moment M0 overcomes the
compression caused by the prestress, and the extreme fibre stress reaches the tensile
strength concrete, cracking occurs, i.e. M0 has reached or exceeded Mcr . The
determination of the flexural strength of the critical slab strip section up to its ultimate
state is based on the method used for the ultimate flexural strength of a typical beam
section as described by Gilbert and Mickleborough [1990].
This flexural strength theory which is to be applied to a prestressed concrete section is
discussed in Section 9.3; whereas the application of the method is elaborated in
Appendix G. In applying this theory, the non-yielding or yielding of tensile
reinforcement and the prestressed tendon are taken into account. This is done by
considering the stresses exerted by them at various load stages up to failure.
180
Chapter 9 : Determination of Ultimate Flexural Strength
9.3 Flexural Strength Theory
The flexural strength theory is a reasonable and well-established theory used to
determine the ultimate strength of a cross-section in bending, Mu [Gilbert and
Mickleborough, 1990]. It considers the strength of both the concrete and the steel in
the compressive and tensile parts of the cross-section. This theory assumes a linear
variation of strain over the cross-section, the absence of tensile stress and an idealised
stress-strain relationship of both concrete and steel. The detailed explanations of these
assumptions and the adoption of an idealised compressive stress block for concrete, as
well as the prestressed tendon strain components are described in Appendix E.
9.3.1 Formulas for the prestressed tendon strain components
The ultimate flexural strength of a section may be calculated accurately, by obtaining an
accurate estimate of the final strain in the prestressed tendons and the steel
reinforcement. For a bonded tendon, the ultimate tensile strain in the prestressed
tendon, Epu, is usually considered to be the sum of several sub-components.
With reference to Fig. E.3 (Appendix E), at loading stage la, the elastic concrete strain,
Ece may be determined from:
PI p2
z" = t [ Y + i ] •(9-9)
The corresponding elastic strain in the prestressing tendon, Epe (denoted as stage lb in
Fig. E.3) is:
181
Chapter 9 : Determination of Ultimate Flexural Strength
eDe = -&- ...(9-10) p A F
ApZp
At load stage 3, the concrete strain at tendon level may be expressed in terms of the
extreme compressive fibre strain, Ecu ( = 0.003) and the depth to the neutral axis at
failure, dn,as:
d„ - d„
e„r = eCM [ ~JLTJL 1 -(9-H)
dn
This concrete strain at tendon level, Ept may be measured experimentally by using the
recorded strain readings at failure from strain gauges attached on the prestressed
tendons during testing.
Therefore, the strain of tendon at ultimate load condition may be obtained as:
Zpu = Zpe + e « + Zpt ...(9-12)
9.3.2 Determination of Mu for a prestressed concrete section
In order to determine theoretically, the moment at ultimate Mu , using Eq.(E-l) in
Appendix E, the depth to the neutral axis dn and the corresponding final stress in tendon
0"pM must first be ascertained.
An iterative trial and error procedure is used to determine dn . About three iterations are
normally needed to obtain a good estimate of dn and hence Mu . The steps required to
carry out these calculations are as follows:
182
Chapter 9 : Determination of Ultimate Flexural Strength
(X) By referring to a doubly reinforced section as shown in Fig. E.2 (Appendix E),
the resultant compressive force consists of a steel component Cs (= <JSCASC) and
a concrete component Cc (= 0.85fcby"dn). The strain in the compressive
reinforcement is thus,
(d„ - d.r) e * = eCM , -(9-13)
dn
ff Esc < Esy (= fsylEs), then asc = ESCES , otherwise osc = fsy , where asc is the
strain in the compressive reinforcement.
(2) The resultant tensile force comprises of a prestressed component Tp (= GpuAp)
and a non-prestressed component Ts (= GstAst)- The strain of non-prestressed
tensile steel at ultimate is
e* = eCM (ds ' dn) -(9-14)
dn
If Est < Esy, then Gst = EstEs, otherwise ast =fsy.
(3) To calculate the depth to the neutral axis at ultimate dn , a trial and error
procedure is utilised. Successive values of dn are selected and tried until the
value which satisfies the horizontal equilibrium equation is determined.
TP + TS=CC+CS ...(9-15)
Assuming that all non-prestressed steel are at yield, then from Eq.(9-15),
Cc = Tp + Ts- Cs
i.e. 0.S5 f'cby"dn = apu Ap +fsy (Ast - Asc)
183
Chapter 9 : Determination of Ultimate Flexural Strength
resulting in,
0.85f'ci'bdn -fsv(Ast - Asc) <Jpu = — " sy st — ...(9-16)
AP
(A) The selected value for dn is also used to determine Epu from Eq.(9-12). When
the values of o>« and Epu satisfy the stress-strain relationship of the prestressing
tendon, the correct value of dn is found.
(5) Check that Esc > Esy and Est > ESy to ensure that the non-prestressed steel has
indeed yielded. Otherwise, GSC = ESCES and / or Gst = EstEs before applying
Eq.(9-16). More iteration for dn may be needed if this is the case.
(6) The ultimate moment may be determined by using Eq.(E-6) in Appendix E or
alternatively by taking moments about the extreme compressive fibre, i.e. at Ecu
level.
Mu = Cc(±y"dn) 4- Csdc - Tpdp - Tsds ...(9-17)
where dc, dp and ds are the depth of the compressive steel, prestressing tendon
and the non-prestressed tensile steel respectively.
The overall procedure for determining the theoretical ultimate moment may be
summarised as shown in the flow chart in Fig. 9.1 .
184
Chapter 9 : Determination of Ultimate Flexural Strength
<<-
C sta" 3
Compute: Mcr Eq.(9-1) Mg (From structural analysis)
Input data: b, D, ds,dp,fc, Pe ,e etc.
No Uncracked Section /Analysis
Compute: E« Eq.(9-9) Epe Eq.(9-10)
Yes
Cracked Section j\nalysis
Compute:
M Eq.(9-7)
C St°P 3
Loop: -1
i = D to 2
Select:
d„ = i
Compute:
Eq.(9-ll)
Epo and apU
are valid
Compute:
EpU Eq.(9-12)
EpU and o-pu are not valid
Check Yes ' A b s ( 0pu. Opuplot) N ^ No
- Minimum
Compute:
°"pu.piot (From stress-straiii curve plot for tendon)
Compute:
cTpU Eq.(9-16)
Yes Check
No V f s and f
\
sc *• ' s y /
> /
1
Compute:
M u Eq.(9-17)
(^ End ^)
Fig. 9.1 - Flow chart to determine the theoretical ultimate flexural strength
of prestressed concrete slab strips
185
Chapter 9 : Determination of Ultimate Flexural Strength
The resisting moment up to the ultimate or failure stage may be determined
experimentally by applying the strain measurements of the steel reinforcement and the
prestressed tendons to the appropriate equations described earlier. The method of
determining the resisting moment of the critical slab strips is similar to that shown in Fig.
9.1, except for the input of measured strain values for steel and tendons.
(Similar to Fig. 9.1^
prior to the loorj/
± y Loop: -1
i = D to 2
Yes
lg — ^sy ks
Yes
°"p ~ e py E p
Select:
d„ = i
No
f, = E S E S
No
°"p = eP^P
Compute:
M Eq.(9-17)
d„ is valid
C E"d 3
Yes
d„ is not valid
No
Compute:
"p.plot (From stress-strain curve plot for tendon)
ft
Compute:
Eq.(9-12)
Fig. 9.2 - Flow chart to determine the flexural strength of prestressed
concrete slab strips using strain measurements
186
Chapter 9 : Determination of Ultimate Flexural Strength
These strain values are then used in Eqs.(9-12) and (9-16) to determine the depth of the
neutral axis dn which is finally included in Eq.(9-17) to calculate the required resisting
moment. The overall approach is shown in Fig. 9.2 . The strain measurements of the
prestressed tendons to be applied in this approach are presented in Table 9.1.
Table 9.1 - Strain measurements at ultimate for all models, at front face of
columns and midspans of critical slab strips
Connection
Pl-A*
Pl-B*
Pl-C*
P2-A
P2-B
P2-C
P3-A
P3-B
P3-C
P4-A
P4-B
P4-C
Strain measurements at ultimate (used to calculate for Mj and Mm)
Gauge at front face of column
(forM;)
0.002050
0.001009
0.002023
" 0.002404
0.002197
0.002049
0.002681
0.002157
0.003065
0.000512
0.000815
0.000209
Gauge at midspan of critical
slab strip ( for Mm)
0.000933
0.001204
0.001031
0.002558
0.002018
0.002035 #
0.002422
0.003119
0.002751
0.000519
0.004772
0.000678
Note: All strain measurements are of prestressed tendons, unless otherwise stated.
The measured strains are read off gauges on steel reinforcement, as strain gauges were not placed on tendons in Model PI .
# This is measured strain of steel reinforcement, as the strain gauges on tendons were faulty at this location for Model P2-C .
187
Chapter 9 : Determination of Ultimate Flexural Strength
The strain measurements recorded up to failure of the steel reinforcing bars and
prestressed tendons for all the models are presented in Appendix F.
9.4 Comparison of Measured and Calculated Values for M; and Mm
The measured values for the ultimate moments at the front face of column, M; , and at
the midspan of the critical slab strip, Mm , are those determined using the experimentally
recorded strains of the prestressed tendons. On the other hand, the calculated values for
Mj and Mm , are computed by assuming that the prestressed tendons are yielding when
the model fails. A typical sample calculation of Mj using the measured tendon strain
for Model P2-B at the front face of the column support is presented in Appendix G.
The measured and calculated values of Mj and Mm , for all the models are presented
in Table 9.2 . It can be observed that 5 out of the 12 measured values for Mj are
determined with strain measurements of the prestressed tendons at yield. As a result,
these five measured values of Mj coincide with the calculated Mj values. On the
other hand, 6 out of the 12 measured values for Mm coincide with the calculated Mm
values.
The measured values for M1 and Mm correlate reasonably well with the corresponding
calculated values, as shown in Fig. 9.3 . The only exception is the correlation point for
Model P3-B, in which the measured Mj is less than the corresponding calculated value
by 15.25 % .
This is due to the exceedingly high concrete strength (of 47.8 MPa) being used in Model
P3 , which resulted in the tendon strain to be well below its yielding point when the
critical strip section (at the front face of column B) failed in punching shear (which
188
Chapter 9 : Determination of Ultimate Flexural Strength
happened well before it had reached its ultimate flexural capacity). The failure
mechanism of Model P3-B has been documented earlier in Chapter 8.
From the comparison of the measured and calculated flexural strength values of the
critical slab strips at failure, it can be deduced that the correlation of the measured to the
calculated results is reasonably good. Therefore, the measured results for M1 and Mm
may be confidently used for the prediction of Vu by employing the Modified Wollongong
Approach in Chapter 11.
Table 9.2 - Measured and calculated values of Mj and Mm for all models
Connection
Pl-A*
Pl-B*
Pl-C*
P2-A
P2-B
P2-C
P3-A
P3-B
P3-C
P4-A
P4-B
P4-C
Measured
Mi, kN-m
7.047
10.370
7.529
7.292
10.790
7.932
7.131
9.005
7.615
6.974
10.713
7.773
Mm, kN-m
9.079
- 11.098
9.749
9.709
11.731
10.509 #
9.325
11.243
9.840
9.636
11.625
10.029
Calculated
Mj, kN-m
7.114
10.540
7.625
7.581
10.790
8.152
7.131
10.626
7.615
6.974
10.713
8.002
Mm, kN-m
9.226
11.098
9.749
10.007
11.731
10.590
9.346
11.243
9.840
9.729
11.625
10.422
The measured results were based on strains measurement of steel reinforcement, as strain gauges were not placed on tendons in Model PI .
# This measured result was based on strain of steel reinforcement, as the strain gauges on tendons were faulty at this location for Model P2-C.
189
Chapter 9 : Determination of Ultimate Flexural Strength
12
B I
55
T3 U
ca u
s
-a
IO
8
8 IO
10.626
12 Mil and. M m , calculated
(kN-m)
Fig. 9.3 - Graph of measured versus calculated Mj and Mm results
190
Chapter 10 : Calibration of the Prediction Formulas
CHAPTER 10
CALIBRATION OF THE PREDICTION FORMULAS FOR
SPANDREL PARAMETER
10.1 General Remarks
In this chapter, the formulas for predicting the slab restraining factor on the spandrel
beam, \y , are calibrated. These can be used in Chapter 11 to compute the punching
shear strength values, Vu , of post-tensioned concrete flat plates using the Modified
Wollongong Approach developed in this thesis (in Chapter 5). It should be reiterated
that the Wollongong Approach in its original form [Loo and Falamaki, 1992] is
applicable only to reinforced concrete flat plates.
It has been stated in Section 5.1.5 that the semi-empirical equations established for
reinforced concrete flat plates could not be applied to post-tensioned concrete flat plates
[Lei, 1993]. Therefore, new semi-empirical formulas to determine the slab restraining
factor \}/ , for post-tensioned concrete flat plates have to be developed. Such
development can be done with the aid of the ultimate load test data on post-tensioned
concrete flat plate models.
The method of formulating the semi-empirical equations (5-22), (5-23) and (5-24) to
determine \|/ for reinforced concrete systems [Falamaki, 1990] is employed herein to
calibrate a new set of semi-empirical equations for the post-tensioned concrete flat
plates. The accuracy of the resulting semi-empirical equations is determined by
comparing the predicted values of \|/ with the corresponding measured values.
191
Chapter 10: Calibration of the Prediction Formulas
10.2 Research Scheme for Determination of Semi-empirical Formulas for \\f
The research scheme to establish the new semi-empirical formulas for \|/ is shown in Fig.
10.1 . The relevant equations in the Modified Wollongong Approach are set up in
Section 10.3.1 to provide an analytical guide for the determination of \|/ in Section
10.3.2 with the aid of test data obtained from the ultimate load tests of the half-scale
post-tensioned models (Chapters 8 and 9).
Section 10.3.1
Set up general equations as
guides for the measurement
of \|/
Section 10.4
Use measured \|/
to calibrate semi-empirical formulas
for \|/
(r
-)
• *
Obtain test data
Ji-kL
Section 10.3.2
Obtain measured
values of tj/
Section 10.5
Use test data to test
accuracy of semi-
empirical formulas for V|/
Fig. 10.1 - Research scheme to establish the semi-empirical formulas to
determine \|/
192
Chapter 10 : Calibration of the Prediction Formulas
The measured values of \|/ are used as a basis for the calibration of the semi-empirical
formulas for \\t. This is done via a linear regression analysis of the measured \|/ in
Section 10.4. The linear regression analysis is used for the curve fitting of these
measured values of \]/ in the scatter graphs of \|/ versus 8 or the spandrel strength
parameter (see Section 4.2 for definition). This is carried out for each of the following
sets of column positions, which are governed by the column dimensions [Falamaki,
1990]:
• the series "A" corner columns ( Q > C2);
• the series "C" corner columns (C; < C2); and
• the series "B" edge columns (C; < C2).
From the linear regression analysis, the estimates of the regression coefficients are
obtained for the resulting semi-empirical formulas. In Section 10.5, the semi-empirical
formulas are used to predict the values of \|/. These predicted \\f values are compared
with the measured ones to determine the accuracy of the resulting semi-empirical
formulas.
10.3 Measurement of the Slab Restraining Factor \|/
10.3.1 Derivation of equations to measure \J/
The measurement of the slab restraining factor, \|/ , involves the use of the three
equilibrium equations (5-6), (5-7) and (5-9) for corner connections or equations (5-12),
(5-13) and (5-14) for edge connections plus the interaction equation (5-40).
193
Chapter 10 : Calibration of the Prediction Formulas
In order to clarify the measuring procedure, the interaction equation (5-40) from Section
5.1.7 is re-produced below:
T2 V2 _ _ rw M2 n 0/wwl (ap + - p i — = [\|/ *-] 200000 l-^- ...(10-1) r Af i K p _ ' Mus ~ V V
Eq.(lO-l) is simplified by letting:
2 , ^2 A = AT + T / -(10-2)
and
A2 = 200000Va|3 ...(10-3)
Substituting Eqs.(10.2) and (10.3) into Eq.(lO.l) yields
A = [¥ - - ^ ] *2 - -(10-4)
^™ vV
In order to obtain an expression for \|/, Eq.(10-4) is re-arranged to take the form:
V A 2 - ^ A - ^ A 2 = 0 -(10-5) M„ us
This is a general quadratic equation of Jy, the solution of which gives:
194
Chapter 10: Calibration of the Prediction Formulas
a AK < M2 - -A) ± JAZ - 4A2( - -f- A2)
vv = T: -(10-6) 2 A 2
By squaring both sides of Eq.(10-6), the factor \|/ is determined as:
v2 A,K s2 M2 A + A4 + 4(A2y
v = [ L ^" ] 2 -dO-7)
zA2
By imposing \|/ = ymeasured > tne measured factor, and letting
A, = JA2 + 4(A2)
2 j ± -(10-8) us
Eq.(10-7) is simplified further, or
T measured L _ . J
It can be observed from Eq.(10-9) in conjunction with Eqs.(10-2), (10-3) and (10-8)
that all the variables, except M2 , V2 and T2 , can be determined using the following
equations as given in Chapters 5:
Eq.(5-18)for a;
Eq.(5-19) for (3;
• Eq.(5-33) for Mm\ and
195
Chapter 10 : Calibration of the Prediction Formulas
Eq.(5-34) for dsp;
Note that in Eq.(10.2), At is the area of the rectangle defined by the centres of the four
longitudinal corner bars of the spandrel beam (i.e. At = xy), and Ut is the perimeter of
the rectangle defined by the centres of the four longitudinal comer bars of the spandrel
beam (i.e. Ut = 2x + 2y).
The equations for M2 , V2 and T2 as given in Eqs.(lO-lO), (10-11) and (10-12) are
derived from the three equilibrium equations (5-9), (5-6) and (5-7) respectively, for
comer-column connections. Similarly, Eqs.(10-13), (10-14) and (10-15) for M2 , V2 and
T2 are based on the equilibrium equations (Eqs.5-14, 5-12 and 5-13) for edge-column
connections.
Thus, for comer-column connections,
M2= lMc2+n2Pe2xL1]U-^i^1^ -dO-10)
V-, = V - V, + n,P, ...(10-11) v2 yu.test yl T "-llely
and,
T2 = MCI - M, - [V, - n, PeIy) [C, - y ^ I + «, Ku », • MT -(10-12)
In Eq.(10-12), MT may be calculated using Eq.(5-10).
196
Chapter 10 : Calibration of the Prediction Formulas
For edge-column connections,
M2 = MC2 [1 - iD'2LD')] ...(10-13)
V2 = \ [ VUJest - V, + niPely ] ...(10-14)
and,
T2 = ^-[MC1-M1- [V, - n} Pely] [C, - ( ^ ~ ^ ] + n, Pe!x e, - MT } ...(10-15)
2 -u-*
Note that the expressions for M2 , V2 and T2 as shown from Eq.(lO-lO) to Eq.(10-15)
require the input of recorded test data for Ma , MC2 and Vutest (see Table 8.3 in
Section 8.4.2) as well as the measured values of M1 (see Table 9.2 in Section 9.4).
10.3.2 Measured values of \j/
Based on the test data recorded in Table 8.3 for MC1 , MC2 , Vujest and the calculated
M1 in Table 9.2, M2 , V2 and T2 may be determined via Eqs.(lO-lO), (10-11), (10-12),
(10-13), (10-14) and (10-15).
The values determine for M2 , V2 , T2 , a , P , Mus, dsp , At and Ut in Section 10.3.1
are then applied to Eq.(10-9) in order to obtain the measured values for the slab
restraining factor, \|/measured • The values obtained for At, Ut , dsp , a , p and Mm are
197
Chapter 10 : Calibration of the Prediction Formulas
presented in Table 10.1 whereas those of M2 , V2 , T2 , A , A} , A2 and the measured
results of \|/ are given in Table 10.2 .
Table 10.1 - Variables used to determine the slab restraining factor \j/
Model
Pl-A
P2-A
P3-A
P4-A
Pl-B
P2-B P3-B
P4-B
Pl-C P2-C
P3-C
P4-C
dsp (mm)
56.2
112.5
83.5
57.7
58.3
113.8 84.7
59.0
56.2 132.1
92.2
57.7
At (mm2)
9920 26784
17976 9920
9920 26784
17976 9920
9920 26784
17976
9920
Ut (mm)
576 712
596
576
576. 712
596 576
576 712
596 576
a
0.40212
0.84509 0.90792
0.87179
0.40212
0.84509 0.90792 0.87179
0.40212 0.84509
0.90792 0.87179
P 5.13792
6.95489 5.86560 5.84267
5.13792
6.95489
5.86560 5.84267
5.13792
6.95489 5.86560 5.84267
Mus (kNm)
3.6433 12.061
9.3109 5.7255
3.7906 12.208 9.4582
5.8728
3.6433 14.285 10.357 5.7255
Table 10.2 - Results for the calculated variables and
the slab restraining factor \\f
Model
Pl-A
P2-A
P3-A
P4-A
Pl-B
P2-B
P3-B
P4-B
Pl-C
P2-C
P3-C
P4-C
M2
(kNm)
15.535
14.555
15.659
15.386
0.2400
0.2540
0.2676
0.1300
14.410 16.474
16.531
14.170 1
y2
(kN)
-3.0885
-1.3986 -1.3441
-3.7455
25.296
9.9884
11.252
23.928
-4.7596
-4.6990
-3.8199
-5.8798
T2
(kNm)
-4.6013
-2.3340
-2.7488
-4.8295
2.7604
2.5474
5.2154
3.1989
-3.6111 -4.4933 -2.6884
-3.2659
A
(xlO3)
-391.82
-92.720
-133.97 -413.31
339.11
131.09
286.76
368.42
-318.08
-184.35
-139.79 -294.44
(x 103)
1250.3
1069.1 1204.6
1536.5
368.68
191.70
326.09 392.14
1186.9
1057.6
1174.6 1450.4
A2
(xlO3)
287.48 484.87
461.54
451.38
287.48 484.87
461.54
451.38
287.48
484.87 461.54
451.38
Ttneasured
2.22913
1.01380 1.34517 1.54804
1.51546
0.11080
0.44078 0.70977
2.28324
0.81085
1.25660
1.63961
198
Chapter 10 : Calibration of the Prediction Formulas
10.4 Calibration of Semi-empirical Formulas for \J/
The procedure used to calibrate a new set of semi-empirical equations for the slab
restraining factors, \|/ for post-tensioned concrete flat plates is similar to the method
employed by Loo and Falamaki [1992] for reinforced concrete systems. The semi-
empirical equations given in Eqs.(5-22), (5-23) and (5-24) for reinforced concrete flat
plates can be presented in the following general forms.
For edge columns,
Vt«, = A -BlogI0 [&S] -(10-16)
and for comer columns,
3
V — = A-B,w£)J£8] -C"»7> Cj U£>,
s
where A and B are coefficients that were ascertained from regression analysis of the
plots of the measured values of \|/ versus the corresponding logarithmic expression in
terms of the spandrel strength parameter, 5 (Eq.5-15).
It was established by Falamaki [1990] that the measured values of \|/ (i.e. ^measured)
should be inversely proportional to the corresponding spandrel strength parameters, 5 .
199
Chapter 10 : Calibration of the Prediction Formulas
Therefore, the general forms for \|/ as shown in Eqs.(10-16) and (10-17) are also
adopted herein to determine \}/ for post-tensioned concrete flat plates.
Besides the slab restraining factor, x\f, the calibration of the semi-empirical formulas also
involves the spandrel strength parameter, 8 (Eq.5-15) and the column dimension factor,
C 3 175 7, for comer-column connections where according to Eq.(5-25), y = [—] — .
ci VDs
For the edge-column connections, the slab thickness factor,
X = [£] .-(10-18)
With the introduction of 7 and X, we have from Eq.(10.16),
NW = A -Blogl0[Xb] -(10-19)
for edge columns, and from Eq.(10.17),
Corner = *-B logw [tf] -d0"20)
for comer columns.
The calculated values for 5 , 7 , A, and the corresponding logarithmic expressions of (78)
for comer connections, and (X5) for edge connections are presented in Table 10.3 .
200
Chapter 10 : Calibration of the Prediction Formulas
With the results in Table 10.2 and Table 10.3, the measured \\f values (from Table 10.2)
are plotted against log10Cy8) and log10(X5) from Table 10.3, for the comer and edge
connections respectively. These plots are presented in Fig. 10.2 . The linear regression
analysis is used for the curve fitting of these measured values of \j/ in the scatter graph of
y for each of the three column positions A, B and C (Section 10.2).
Table 10.3 - Calculated values of 8 , 7 , X , log10(75) and log10(A,5) for all
post-tensioned concrete flat plate models PI to P4
Model
Pl-A
P2-A
P3-A P4-A
Pl-B
P2-B
P3-B
P4-B
Pl-C
P2-C
P3-C P4-C
5
2.0633
11.725 7.8504
5.2036
2.1941
12.146 8.1492
5.4432
2.0633
13.758 8.6684
5.2036
7
0.286888 0.286888
0.286888
0.286888
-
2.295102
2.295102 2.295102
2.295102
X
-
0.823975 0.823975 0.823975
0.823975
-
Iog10 (7§)
-0.22773 0.52682
0.35260 0.17402
-
0.67536 1.49936 1.29874
1.07711
log10 OS)
-
0.25717 1.00033 0.82703
0.65177
-
A linear regression analysis is carried out for each of the plots given in Fig. 10.2 . The
"curve-fitting" process has been carried out with the aid of a standard computer
package* to create graphs and to determine the coefficients of determination (R2) . The
values of correlation coefficients, R were obtained (i.e. VR2) as an indication of the
adequacy of the curve-fittings.
"CA-Cricket Graph v. 1.3.1" developed for IBM and IBM-compatibles by Computer Associates and
Microsoft Corp.
201
Chapter 10 : Calibration of the Prediction Formulas
\p - 1.86-1.58 log (y5)
3.0
2.0
W
1.0 -
0.0
R"2 - 0.9957
-<—\—'—i—'—r-^ -0.25 O.OO 0.25 0.50 0.75
log (y6)
(a) Corner connections A
•K - 1.98-1.88 log (X5)
V
1.5 J
l.O -
0.5 -
o.o
R"2
Pl-B
P4-B "
1 1 '
= 0.9975
\ P3-B
\ P2-B
1 ' 1 ' 1 0.25 0.50 0.75 1.00 1.25
log (A.5)
(b) Edge connections B
v»V = 3.50-1.76log (y5)
3.0
2.0 -
V
1.0
0.0 -|—i—|—i—|—i—| i I r
O.SO 0.75 l.OO 1.25 1.50 1.75
log(yo)
(c) Corner connections C
Fig. 10.2 - Graphs of y versus log 10(78) for connections A and C and
\j/ versus logio(M>) for connections B
202
Chapter 10 : Calibration of the Prediction Formulas
Note that the closer the correlation coefficient is to unity, the better is the adequacy of
the fitted curve for the plots [Walpole and Myers, 1985]. The curve-fittings are shown
in Fig. 10.2a, 10.2b and 10.2c for connections A, B and C respectively, and the R values
are tabulated in Table 10.4 .
Table 10.4 - Coefficients of determination (R2) and correlation coefficients
(R) of the plots of \|/ versus logio (76) and \|/ versus logio (kd)
Connection
A B C
Position
Comer Edge Comer
R2
(from Fig. 10.2)
0.9957 0.9975 0.9946
R
0.9978 0.9987 0.9973
As shown in Table 10.4, with Rvalues of 0.9978, 0.9787 and 0.9973 for connections A,
B and C respectively, the linear curve-fitting for each of the plots is considered
adequate. The mathematical expressions for the fitted lines are taken to be the semi-
empirical formulas for \|/.
The semi-empirical formula to be applied for edge-column (connections B) is:
Vedge = 1-98 - 1-88 log10 [Xb] .(10-21)
whereas for comer-column (connections Q where C2>Ci,
Vcorner = 3.50 - 1.76 logw [7§] .(10-22)
203
Chapter 10 : Calibration of the Prediction Formulas
and for C 2 ^ C j , in connections A ,
Vcorner = lM ' l"58 l°SW M ...(10-23)
10.5 Comparison Between Measured and Predicted Values of y
The predicted values of \j/ are determined using the semi-empirical equations (10-21),
(10-22) and (10-23). These are presented and compared with the measured \j/ values in
Table 10.5 together with the ratio of measured to predicted values of \)/.
The punching shear strengths, Vu , at all the connections are analysed using the semi-
empirical formulas developed in Section 10.4 . The results of these analyses are
presented in Chapter 11. The relationship between the analysed Vu and the ratio " J f '
¥ predicted
is studied.
It was found that, when the ratio measured is closed to unity, the semi-empirical T predicted
formulas give an accurate prediction of Vu .
Also, when the ratio ^measaKi > 1.0 , the semi-empirical formulas produce a T predicted
conservative prediction of Vu .
204
Chapter 10 : Calibration of the Prediction Formulas
As shown in Table 10.5, with a mean value of 1.008 and a standard deviation of 0.047
for the ratio of measured to predicted values of V|/ , the resulting semi-empirical
equations give reasonably accurate predictions of \|/.
Table 10.5 - T h e measured and predicted values of \|/
Model
Pl-A
P2-A
P3-A P4-A
Pl-B P2-B P3-B
P4-B
Pl-C
P2-C P3-C
P4-C
T measured
2.22913
1.01380
1.34517 1.54804
1.51546
0.11080
0.44078 0.70977
2.28324
0.81085
1.25660
1.63961
T predicted
2.21981 1.02762
1.30289 1.58506
1.49651
0.09938 0.42519 0.75468
2.31137 0.86113 1.21422
1.60429
Mean
Standard Deviation
i measured
T predicted
1.004
0.987 1.032 0.977
1.013 1.115 1.037
0.940
0.988 0.942
1.035 1.022
1.008
. 0.047
205
Chapter 11 : Analysis of Punching Shear Strength
CHAPTER 11
ANALYSIS OF PUNCHING SHEAR STRENGTH
11.1 General Remarks
The punching shear strength, Vu , of the post-tensioned concrete flat plate models are
analysed herein using the Modified Wollongong Approach developed in Chapter 5 and
calibrated in Chapter 10. The Wollongong Approach [Loo and Falamaki, 1992] for
reinforced concrete flat plates was used as a basis for the development of the Modified
Wollongong Approach. This was achieved by incorporating the effects of prestress into
the Wollongong Approach.
The equations required for the analysis of Vu by the Modified Wollongong Approach are
summarised for ease of reference. The development and use of a computer program to
analyse the punching shear strength is also discussed. The computer program
encompasses the analytical methods recommended by A S 3600-1988 [SAA, 1988],
ACI318-89 [ACI, 1989], B S 8110:1985 [BSI, 1985] and the Modified Wollongong
Approach. T w o sets of sample calculations for Vu at the comer- and edge-column
connections using the Modified Wollongong Approach are presented in detail.
The predicted punching shear strength results for the eight comer- and four edge-
column connections of the post-tensioned concrete flat plate models, using the four
analytical methods are presented and compared with the experimental results. From
these comparisons, it is found that the Modified Wollongong Approach gives a more
accurate and reliable prediction of Vu than the other three methods.
206
Chapter 11 : Analysis of Punching Shear Strength
11.2 The Prediction Formulas for Vu
11.2.1 Corner connections with spandrel beam
For a corner column-slab-spandrel connection,
kfak4 - k2k4 + k3ky
k3-k4k5
where,
kAMn + noP^xLr)^ [ a|3 n 1 n. k} = 200000 [ycorner - 6K C2 2 " ] J—*- -dl-2)
lv*us V T corner
k2 = k6Mcl - M 7 , c o r w - [ Q - - ^ ] + n, Pei;c ej ...(11-3)
jk, = - A - -(H-4) V2<7,
A,2u7.
5 2
••(11-5)
ci - bsP ...(11-6)
207
Chapter 11 : Analysis of Punching Shear Strength
Wl - Ds\ T-i
h-l- M ^ ...(U-7)
and,
k7 = V1>corner- njPely ...(11-8)
11.2.2 Edge connections with spandrel beam
For an edge column-slab-spandrel connection,
v = 2k}k3k4 - k2k4 + k3hj ...(11-9)
" k3 - k4k5
where,
*, = 200000 lVedge - *£*] J-p_ ...d.-.O) mus V T edge
k2 = k6Ma - M1Mge - hflC, - L-^-l + „, PeIX e, ...(H-l D
and,
u T/ P ...(11-12) K7 ~ Vledge ' nlUly
208
Chapter 11 : Analysis of Punching Shear Strength
The other quantities are as those defined for the comer connections.
11.2.3 Connections with torsion strip or free edge
Equations for Vu presented in Section 11.2.1 are applicable to connections with torsion
strips except that Eq.(5-41) for a comer connection [Loo and Falamaki, 1992], is
replaced by
400 °25
M ^ - l - j ^ Miu -dl-13)
Similarly, for edge connections with torsion strips [Loo and Falamaki, 1992], the
equations given in Section 11.2.2 may be utilised to analyse Vu by replacing Eq.(5-44)
with
"* • i»?U" *>- -(1M4)
11.3 Computer Program for Vu
The computation for Vu is aided by the use of a BASIC computer program to run all the
required simple and explicit equations contained in the proposed prediction procedure
(referred to in this thesis as the Modified Wollongong Approach). The program is quite
versatile in usage, featuring user-friendly data input system as well as facilities for easy
storage and retrieval of data.
209
Chapter 11 : Analysis of Punching Shear Strength
Besides developing a computer program for the Modified Wollongong Approach, other
analytical methods such as the Australian Standard procedure, the ACI method and the
British Standard method are also programmed in a similar fashion. This enable a faster
and easier analysis of all the tested post-tensioned concrete models as well as to mn
computation on reinforced concrete models tested previously by Loo and Falamaki
[1992]. The listings of the source code for the four computer programs (for the
Modified Wollongong Approach, the Australian Standard procedure, the ACI method
and the British Standard method) are shown in Appendix J.
Two sets of typical sample calculations using the Modified Wollongong Approach, are
carried out with the aid of the program for models Pl-A and P3-B . These are
presented in Appendix H. The data input for the analysis of Vu for other post-
tensioned concrete models are presented in Appendix I.
Other features of the computer program for the Modified Wollongong Approach are:
• computation of the measured slab restraining factor, ti/measured (shown in Section
10.4); and
• analysis of the ultimate moments, M; and Mm .
Note also that the parametric study undertaken in Chapter 4 was made easier by writing
a simple interface routine to handle numerous punching shear analysis runs and data
management work.
210
Chapter 11 : Analysis of Punching Shear Strength
11.4 Comparison of Results
The predicted results by the Modified Wollongong Approach are presented, in
conjunction with those by the methods recommended in AS 3600-1988 [SAA, 1988],
ACI318-89 [ACI, 1989] andBS 8110:1985 [BSI, 1985].
Table 11.1 presents the predicted punching shear strength results by the Modified
Wollongong Approach. Also included are the ratios of the predicted and measured Vu ,
for the eight comer- and four edge-column connections of the post-tensioned concrete
flat plate models. The parallel results by the other three methods are shown in Table
11.2.
The relative accuracy and reliability of all four prediction methods are determined by
comparison with the measured values of Vu . The comparison is carried out by plotting
the experimental punching shear strength values against the predicted values from each
of the four prediction methods, where applicable. These are referred to as the
correlation plots. Standard statistical results (i.e. maximum, average and standard
deviations) are used to determine the relative performance of the four prediction
methods.
As in Section 3.4.2 (Chapter 3), superimposed on these figures are the 30-percentile
lines and the 45° equality line. Any point falling below the -30% line means that the
analysis method has overestimated the strength by more than 30%. Also, the -30% line
may be taken as a safety criterion to reflect the Australian Standard requirement of a
shear capacity reduction factor of 0.7 .
211
Chapter 11 : Analysis of Punching Shear Strength
11.4.1 Predicted punching shear results by the proposed approach (or the
Modified Wollongong Approach)
Table 11.1 presents the punching shear results for all the four post-tensioned concrete
flat plate models predicted using the Modified Wollongong Approach, herein referred to
as the proposed approach.
It can be observed in Table 11.1 that, with an overall mean ratio of predicted to
measured Vu of 0.987, the proposed approach gives reasonably accurate predictions of
the punching shear of all the models. The predictions are slightly more conservative for
models with spandrel beams (i.e. models P2 and P3) with a mean ratio of predicted to
measured Vu of 0.953 . The corresponding value is 1.021 for models without spandrel
beams (i.e. models PI and P4).
11.4.2 Predicted punching shear results by codes of practice and comparison
The punching shear strength results for the four flat plate models are compared with the
predicted values due to the Australian Standard procedure [AS 3600-1988] and the
methods recommended by the American Concrete Institute [ACI318-89] and the British
Standard Institute [BS 8110:1985]. These methods have been described earlier in
Section 5.1.2.
212
Chapter 11 : Analysis of Punching Shear Strength
Table 11.1 - Punching shear strength predictions for post-tensioned concrete
flat plate models PI to P4 using the proposed approach (Modified
Wollongong Approach)
Connection
Pl-A
Pl-B
Pl-C
P4-A
P4-B
P4-C
Measured Vu
(kN)
26.93
68.75
27.72
27.64
66.43
28.22
P2-A
P2-B
P2-C
P3-A
P3-B
P3-C
32.02
76.33
32.97
30.97
73.35
31.78
Predicted Vu
(kN)
26.35
67.71
29.32
31.70
71.95
24.68
Mean ratio =
34.05
70.54
42.63
23.58
68.83
23.46
Mean ratio =
Overall mean =
Ratio
Predicted / Measured
0.978
0.985
1.058
1.147
1.083
0.875
1.021
1.063
0.924
1.293
0.761
0.938
0.738
0.953
0.987
All numbers that end with A or C are corner connections; those that end with B are edge
connections. Unlike models P2 and P3 , models PI and P4 do not have spandrel beams.
213
Chapter 11 : Analysis of Punching Shear Strength
The outcomes of these codes of practice predictions are presented in Table 11.2 . It can
be seen that for the six slab-column connections of models PI and P4 (which are
without spandrel beams), both the Australian and the American code methods give a
mean ratio of the predicted and measured results of 1.619 and 1.892 respectively. This
means that the Australian Standard procedure gives a better prediction than the ACI
method, which overestimate Vu by as much as 2.46 times that of the corresponding
measured result (for comer connection Pl-C). On the other hand, the Australian
Standard procedure overestimates the strength by just below 2.15 times at the most (for
comer connection P4-C). Note that the Australian Standard results for edge
connections are generally good for models PI and P4 , at a predicted to measured Vu
ratio of 0.954 and 1.289 respectively.
A survey of Table 11.2 indicates that the British method with a mean ratio of the
predicted and measured results of 0.726, gives a conservative prediction overall for
models PI and P4 . The maximum underestimate is for edge connection B where the
prediction is about 60% of the measured value.
In the case of models P2 and P3 which are flat plates with spandrel beams for which the
ACI method and the British method are not applicable, only the Australian Standard
procedure is used for the comparison. This is shown in the lower portion of Table 11.2 .
The mean ratio of the predicted and measured results is 2.364 for models P2 and P3 .
These are more unsafe compared to the predictions for models PI and P4 (which are
without spandrel beams).
214
Chapter 11 : Analysis of Punching Shear Strength
Table 11.2 - Predicted punching shear strength results using the methods
recommended by A S 3600 -1988, ACI318-89 and B S 8110:1985
Connection
*
Pl-A
Pl-B
Pl-C
P4-A
P4-B
P4-C
P2-A
P2-B
P2-C
P3-A
P3-B
P3-C
Measured
Vu,kN
26.93
68.75
27.72
27.64
66.43
28.22
32.02
76.33.
32.97
30.97
73.35
31.78
Overal
AS 3600-1988
Predicted
V„,kN
42.75
65.56
45.63
57.81
85.62
60.56
Mean =
96.09
127.20
116.85
73.03
84.37
78.36
Mean =
[ mean =
Predicted
Measured
1.587
0.954
1.646
2.092
1.289
2.146
1.619
3.001
1.666
3.544
2.358
1.150
2.466
2.364
1.992
ACI318-89*
Predicted
41.84
101.35
68.12
54.90
99.79
67.07
Mean =
Predicted
Measured
1.554
1.474
2.457
1.986
1.502
2.377
1.892
For Model P2, the
mean of n Predicted
Measured
= 2.737
For Model P3, the
mean of n Predicted
Measured
= 1.991
BS 8110:1985*
Predicted Vu,kN
18.88
42.82
26.31
17.47
39.38
24.22
Mean =
Predicted
Measured
0.701
0.623
0.949
0.632
0.593
0.858
0.726
-
Note that the predictions by the Australian Standard procedure are less safe for Model
P2 with a deeper spandrel beam (of 160 mm) than those for Model P3 with a shallower
spandrel beam (of 120 mm). This is shown in the fifth column at the lower portion of
* All numbers that end with A or C are corner connections; those that end with B are edge
connections. Unlike models P2 and P3 , models PI and P4 do not have spandrel beams.
* ACI318-89 and BS 8110:1985 do not cover the punching shear analysis of flat plates with spandrel
beams.
215
Chapter 11 : Analysis of Punching Shear Strength
Table 11.2 . The mean ratio of the predicted and measured results for Model P2 is
2.737, which is higher than that for Model P3 at 1.991. Also, the higher overestimates
are for the comer connections P2-A and P2-C , with the predicted to measured Vu ratio
of 3.00 and 3.54 respectively.
11.5 Comments on the Accuracy and Reliability of the Proposed Approach (or
the Modified Wollongong Approach)
11.5.1 Flat plates without spandrel beams (Models PI and P4)
It is obvious in Fig. 11.1 that the predictions by the proposed approach has all its
correlation points well within the +30%- and -30%-percentile limits. In contrast, the
Australian Standard procedure has 2 and the British method has 3 (out of 6 points), in
those limits.
The four remaining correlation points for the Australian Standard procedure are
overestimates by more than 30% (i.e. below the -30% line); the three remaining points
for the British method are above the +30% line.
Contrary to these, the ACI procedure has all its correlation points falling below the
-30% line. This means that all its results are overestimates and by more than 30%.
A survey of the maximum, average and standard deviations in Table 11.3 leads to the
conclusion that the strengths of the prediction for all four methods (from the highest to
lowest) are in the following order : the proposed approach, British method, Australian
Standard procedure and ACI method.
216
Chapter 11 : Analysis of Punching Shear Strength
* AS 3600-1988 ° ACI318-89 H BS 8110:1985
Proposed approach
i ' \~—*—r O 30 60 90 120
Predicted Vu (kN)
Fig. 11.1 - Measured versus predicted Vu for post-tensioned concrete flat
plate models without spandrel beams (PI and P4)
+ Table 11.3 - Deviations of predicted Vu for models without spandrel beams
Prediction methods
A S 3600-1988
ACI318-89
BS 8110:1985
Proposed approach
Maximum deviation
+ 1.15
+ 1.46
-0.41
+ 0.15
Average deviation
+ 0.62
+ 0.89
-0.27
+ 0.02
Standard deviation
22.01
32.36
16.28
3.25
+ Maximum deviation = the maximum value of (Predicted Vu - Measured VJ/Measured Vu Average deviation = the average value of (Predicted Vu - Measured V„)/Measured Vu
Standard deviation = [Z (Predicted V M - Measured vuY
N
217
Chapter 11 : Analysis of Punching Shear Strength
11.5.2 Flat plates with spandrel beams (Models P2 and P3)
It may be observed in Fig. 11.2 that all 6 correlation points of the proposed approach fall
within the 30% percentile limits, whereas only one correlation point by the Australian
Standard procedure is bounded by the 30% line limits. The other correlation points are
all overestimates by more than 30% (i.e. below the -30% line).
It is obvious in Table 11.4 that the predictions by the proposed approach deviate far less
from the measured values than that by the Australian Standard procedure. This means
that the proposed approach produces a more accurate and reliable prediction of Vu for
flat plate models with spandrel beams than the Australian Standard procedure.
- AS 3600-1988 » Proposed approach
Equality
30% /(overestimate)
- " — i — • — i — • r O 30 60 90 120 150
Predicted V u (kN)
Fig. 11.2 - Measured versus predicted Vu for post-tensioned concrete flat
plate models with spandrel beams (P2 and P3)
218
Chapter 11 : Analysis of Punching Shear Strength
Table 11.4 - Deviations of predicted Vu for models
with spandrel beams
Prediction methods
A S 3600-1988
Proposed approach
Maximum deviation
+ 2.54
+ 0.14
Average
deviation
+ 1.36
+ 0.03
Standard deviation
54.45
2.94
11.6 S u m m a r y
For both types of post-tensioned concrete flat plate models (i.e. with and without
spandrel beams), the overall performance of the proposed approach is compared to the
Australian Standard procedure [SAA, 1988]. It is found that their relative performance
is similar to the findings for reinforced concrete flat plate models (see Chapter 3), in
which the Wollongong Approach [Loo and Falamaki, 1992] is compared to the
Australian Standard procedure.
It has been suggested in Section 3.4.5 that the Australian Standard procedure was
developed without adequate support of model test results particularly for those with
spandrel beams. This was believed to be the reason for its poor performance [Falamaki
and Loo, 1991 and Loo and Chiang, 1993]. It appears that the lack of reliability of the
Australian Standard procedure also extends to post-tensioned concrete flat plates.
The performance of the ACI method [ACI, 1989] for models without spandrel beams is
also poor (with overestimates as high as 2.46 times the measured values). Most likely,
this is due to the omission of the effect of prestress on the punching shear strength of
219
Chapter 11 : Analysis of Punching Shear Strength
exterior column connections (see Section 5.1.2b). O n the other hand, the British
method [BSI, 1985] may not be as accurate as the proposed approach, but its
predictions are conservative (with underestimated results as low as 59.3% of the
measured value).
220
Chapter 12 : Conclusion
CHAPTER 12
CONCLUSION
12.1 General Remarks
In Chapter 3, a comparative study is described of the relative performance of the
Wollongong Approach [Loo and Falamaki, 1992] in predicting the punching shear
strength of reinforced concrete flat plates. In this study, the Wollongong Approach is
compared with three national concrete codes of practice, namely A S 3600-1988 [SAA,
1988], ACI318-89 [ACI, 1989] and B S 8110:1985 [BSI, 1985]. Besides the relative
accuracy and reliability of the Wollongong Approach, some limitations of this approach
are also identified regarding the effects of varying spandrel beam features on the
punching shear strength of reinforced concrete flat plates. The conclusions drawn are
given in Section 12.2 .
As part of a long-term research effort on the structural behaviour and failure mechanisms
of post-tensioned concrete flat plates with spandrel beams, a series of four half-scale
models were constructed and tested to failure. With the aid of the experimental results,
the prediction procedure for reinforced concrete flat plates by Loo and Falamaki [1992] is
modified to analyse the punching shear strength at the comer- and edge-column positions
of post-tensioned concrete flat plates. This is referred to as the proposed approach (or the
Modified Wollongong Approach). Based on these related studies, conclusions can be
drawn in the following four areas:
(a) deformation and failure mechanisms of post-tensioned concrete flat plate models;
221
Chapter 12 : Conclusion
(b) development of the proposed approach, for the prediction of Vu of post-tensioned
concrete flat plates with and without spandrel beams; and
(c) the accuracy of the proposed approach.
These are elaborated in Sections 12.3.1, 12.3.2 and 12.3.3 respectively. Some
recommendations for further study are given in Section 12.4 .
12.2 Performance and Limitations of the Wollongong Approach [Loo and
Falamaki, 1992]
The Wollongong Approach in general gives the best performance in predicting the
punching shear strength of flat plates with torsion strips and those with spandrel beams.
The Australian Standard procedure performs just as satisfactorily for flat plates with
torsion strips but tends to be unsafe for those with spandrel beams (as it at times grossly
over-estimates the punching shear strength). The main reason for the inferior
performance by the Australian Standard provision is the inadequate support of model
test results, particularly for flat plates with spandrel beams.
Both the ACI and the British methods are applicable only to flat plates with torsion
strips, in which they also tend to give unsafe predictions of the punching shear strength.
However, by applying appropriate reduction factors of 0.638 and 0.647 respectively to
the ACI and British methods respectively, the accuracy of the predictions of Vu
improves dramatically (Section 3.4.4).
The parametric study of the Wollongong Approach in analysing Vu , of reinforced
concrete flat plates with spandrel beams shows that, the punching shear strength
222
Chapter 12: Conclusion
increases with deeper and wider spandrel beam. This is true only within some limits of
the spandrel strength parameter, 8 , beyond which the formulations are not feasible. The
same also applies to the amount of longitudinal steel reinforcement in the spandrel beam.
The lower and upper limits of 8 , in which the Wollongong Approach may be applied to
determine Vu , of reinforced concrete flat plates with spandrel beams are summarised
and presented in Table 4.4 (Chapter 4).
12.3 Experimental and Theoretical Studies of Post-tensioned Concrete Flat
Plates
12.3.1 Deformation and Failure Mechanisms
The patterns of deformation (crack propagation and deflection) and failure mechanisms
at mid-panels and at exterior slab-column connections of post-tensioned concrete flat
plate models with spandrel beams are similar to the reinforced concrete systems.
Depending on the stiffening effect offered by the spandrel beam or torsion strip along the
slab edge, the post-tensioned concrete flat plate models would fail in either shear or
flexural failure modes.
Both models without spandrel beam and the one with shallow spandrel (i.e. models PI,
P4 and P3 respectively) followed shear failure mode. O n the other hand, the model with
deep spandrel (i.e. Model P2) failed in a flexural mode. Note that when Model P2
finally failed, the corner- and edge-columns also failed almost simultaneously in
punching shear as a response to a sudden shift or re-distribution of the applied loading to
the exterior column supports, as described in (Section 8.4). It has also been shown that
223
Chapter 12 : Conclusion
the spandrel beam offered considerable stiffening effect in reducing the magnitude of
deflections, particularly along the edges of the slab.
12.3.2 Prediction of Vu by the proposed approach (Modified Wollongong
Approach)
j\n analytical procedure is developed in this thesis for the prediction of the punching
shear strengths of post-tensioned concrete flat plates at the edge- and comer-column
connections. This procedure, referred to as the proposed approach, involves the
following work.
(a) Adoption of the basic theoretical aspects of the Wollongong Approach as
developed by Loo and Falamaki [1992] (Section 5.1.3).
(b) Incorporation of the effect of prestress into the three basic equilibrium equations
plus an interaction equation to account for shear, moment and torsion in the
spandrel (Sections 5.1.4 and 5.1.7 respectively).
(c) " Re-calibration of the semi-empirical formulas for the slab restraining factor, \J/ ,
to be used as an integral part of the prediction formulas for Vu of post-tensioned
concrete flat plates (Section 10.4).
This process is aided by developing a BASIC computer program to calculate all the
required simple and explicit equations contained in the proposed prediction procedure.
The program is versatile in usage, featuring user-friendly routines for data input, storage
and retrieval.
224
Chapter 12 : Conclusion
12.3.3 Accuracy of the proposed approach (Modified Wollongong Approach)
It was shown in Tables 11.1 and 11.2 (Chapter 11) that the proposed approach
developed in this thesis gives a more accurate prediction than those of A S 3600-1988
[SAA, 1988].
For models without spandrel beams, the proposed approach produced the best
prediction results compared with those of A S 3600-1988 [SAA, 1988], ACI318-89
[ACI, 1989] and B S 8110:1985 [BSI, 1985], as shown in Tables 11.1 and 11.2 . The
proposed approach also performed better in predicting Vu for models with spandrel
beams than the Australian Standard procedure (Tables 11.1 and 11.2).
The Australian Standard procedure gave a slightly better prediction results for models
without spandrel beam than for models with spandrel beams. Nevertheless, it
overestimates the punching shear strength considerably. This is particularly the case at
comer-column connections (Table 11.2). It has been suggested that the Australian
Standard procedure was developed without adequate support of model test results
particularly for those with spandrel beams. This is believed to be the reason for its poor
performance.
12.4 Recommendations for Further Study
Further studies on the structural behaviour of post-tensioned concrete flat plates may be
focused on two aspects: analytical studies of proposed prediction procedure;
experimental studies of other types of model construction.
225
Chapter 12 : Conclusion
Based on the proposed approach, additional parametric studies may be carried out on
factors that may influence the analysis of Vu . These include the distribution of prestress
and other features of the slab and spandrel beam.
To supplement the model tests being carried out to date, other studies may cover:
(a) band beam slabs; and
(b) effects of repeated loads on post-tensioned concrete flat plates.
226
References
REFERENCES
American Concrete Institute (ACI 1989), Building Code Requirements for Reinforced
Concrete (ACI 318-89) and Commentary - ACI318R-89, Detroit, 353 p.
ACI-ASCE Committee 423 (1983), "Recommendations for Concrete Members
Prestressed With Unbonded Tendons (ACI423.3R-83)", Concrete International :
Design and Construction, V.5, No.7, July 1983, pp.61-76.
ASCE-ACI Committee 426 (1977), "Suggested Revisions to Shear Provisions for
Building Codes"', ACI Structural Journal, V.74, September 1977, pp.458-469.
Beresford, F.D. and Blakey, F.A. (1963), "Experimental Lightweight Flat Plate
Structure, Part VII - A Test to Destruction", Constructional Review, V.36, No.6,
June 1963, pp. 18-26.
British Standards Institution (BSI 1985), BS8110:1985 Structural Use of Concrete,
Part 1, Code of Practice for Design and Construction, London.
Brotchie, J.F. and Beresford, F.D. (1967), "Experimental Study of a Prestressed
Concrete Flat Plate Structure", Civ. Engg. Trans. IEAust, V.CE9, No.2, October
1967, pp.276-282.
Burns, H. and Hemakom, R. (1977), " Test of Scale Model Post-Tensioned Flat Plate",
Proc. ASCE, V.103, ST6, June 1977, pp.1237-1255.
CEB-FIP Model Code (1978), International System of Unified Standard Codes of
Practice for Structures, Comite Euro-International du Be ton, Paris, 348 pp.
227
References
Cement and Concrete Association of Australia (1991), Concrete Design Handbook (in
accordance withAS3600), 2nd Ed., Sydney, 1991, 237 p.
Chana, P.S. (1991), "Punching Shear in Concrete Slabs", The Structural Engineer,
Vol.69, No.15, 6 August 1991, pp.282-285.
Chiang, CL. and Loo, Y.C. (1993), "Punching Shear Tests of Half-Scale Post-
Tensioned Concrete Flat Plate Models with Spandrel Beams," Proc. 13th
Australasian Conference on Mechanics of Structures and Materials (ACMSM
13), Wollongong, July 1993, V.l, pp. 185-192.
Clarkson, S.A. (1992), Deflection of a Post-tensioned Concrete Flat Plate With
Spandrel Beams, B.Eng. thesis, University of Wollongong, New South Wales,
1992, 100 p.
Concrete Society (1979), Flat Slabs in Post-tensioned Concrete with Particular
Regard to the Use of Unbonded Tendons: Design Recommendations, London,
Technical Report 17, July 1979.
Cope, RJ. and Clark, L.A. (1984), Concrete Slabs Analysis and Design, Elsevier
Applied Science, London, 1984, 502 p.
Di Stasio, J. and Van Buren, M.P. (1960), "Transfer of Bending Moment Between Hat
Plate Floor and Column", ACI Structural Journal, V.57, September 1960,
pp.299-314.
228
References
Elgren, L., Karlsson, I., Losberg A. (1974), "Torsion-Bending-Shear Interaction for
Concrete Beams", Journal of the Structural Division, ASCE, Vol.100, N0.ST8,
August 1974, pp. 1657-1676.
Elstner, R.C. and Hognestad, E. (1956), "Shearing Strength of Reinforced Concrete
Slabs", ACI Structural Journal, V.28, No.l, July 1956, pp.29-58.
Falamaki, M. (1990), Punching Shear Strength of Reinforced Concrete Flat Plates
With Spandrel Beams, PhD thesis, University of Wollongong, New South Wales,
1990, 285 p.
Falamaki, M. and Loo, Y.C. (1991), Discussion 87-S15 , "Punching Shear Design in the
New Australian Standard for Concrete Structures", by Rangan, B.V. (1990) , ACI
Structural Journal, V.88, No.l, January-February 1991, pp.115-116.
Falamaki, M. and Loo, Y.C. (1992), "Punching Shear Tests of Half-Scale Reinforced
Concrete Flat Plate Models With Spandrel Beams", ACI Structural Journal,
V.89, No.3, May-June 1992, pp.263-271.
Falamaki, M. and Loo, Y.C. (1993), Authors' Closure (Discussion 89-S25) , "Punching
Shear Tests of Half-Scale Reinforced Concrete Hat Plate Models With Spandrel
Beams", by Falamaki, M. and Loo, Y.C. (1992) , ACI Structural Journal, V.90,
No.2, March-April 1993, p.219.
Foutch, D.A., Gamble, W.L. and Sunidja, H. (1990), "Tests of Post-Tensioned Concrete
Slab-Edge Column Connections", ACI Structural Journal, V.87, No.2, March-
April 1990, pp. 167-179.
229
References
Gilbert, RJ. and Mickleborough, N.C. (1990), Design of Prestressed Concrete, Unwin
Hyman Ltd., London, 1990, 504 p.
Guralnick, S.A. and LaFraugh, R.W. (1963), "Laboratory Study of a 45-foot Square
Hat Plate Structure", ACI Structural Journal, V.60, No.9, September, 1963,
pp. 1107-1184.
Hawkins, N.M. (1974), "Shear Strength of Slabs with Moment Transferred to
Columns", Shear in Reinforced Concrete, SP-42, Vol.2, American Concrete
Institute, Detroit.
Hii, H.M. (1993), Serviceability of a Prestressed Concrete Flat Plate, B.Eng. thesis,
University of Wollongong, New South Wales, 1993, 169 p.
Ingvarsson, H., (1974), Experimentellt studium av betongplattor understodda av horn
pelare (Experimental Study of Concrete Slabs Supported By Corner Columns),
Meddelande 111, Institutionen for Byggnadsstatik, Kungliga Tekniska Hogskolan,
Stockholm, 1974.
Institution of Structural Engineers (1969), Shear Strength of Reinforced Concrete
Beams, Report by the Shear Study Group, London, January 1969.
Kinnunen, S., (1971), Forsok med betongplattor understodda av pelare vid fh kant
(Tests on Concrete Slabs Supported by Columns at Free Edges), Rapport R2,
Statens Institut for Byggnadsstatik, Stockholm, 1971.
Kong, F.K. and Evans, R.H. (1987), Reinforced and Prestressed Concrete, 3rd Ed.,
Van Nostrand Reinhold (International), London, 1987, 508 p.
230
References
Lei, K.F. (1993), Punching Shear Strength Behaviour of Post-tensioned Concrete Flat
Plates - Model Studies, B.Eng. thesis, University of Wollongong, New South
Wales, 1993, 144 p.
Lin, T.Y. and Burns, N.H. (1981), Design of Prestressed Concrete Structures, 3rd Ed.,
John Wiley & Sons, New York, 1981, 646 p.
Long, A.E. and Masterson, D.M. (1974), "Improved Experimental Procedure for
Determining the Punching Shear Strength of Reinforced Concrete Hat Slab
Structures", ACI Special Publication SP42, Shear in Reinforced Concrete, V.2,
1974, pp.921-935.
Long, A.E., Kirk, D.W. and Cleland, DJ. (1978), "Moment Transfer and the Ultimate
Capacity of Slab-Column Structures", The Structural Engineer, Vol.56 A, No.4,
April 1978, pp.95-102.
Long, A.E. and Cleland, D.J. (1993), "Post-Tensioned Concrete Hat Slabs at Edge
Columns", AC/Materials Journal, Vol.90, No.3, May-June 1993, pp.207-213.
Loo, Y.C. and Chiang CL. (1993), "Methods of Punching Shear Strength Analysis of
Reinforced Concrete Hat Plates - A Comparative Study", International Journal of
Structural Engineering and Mechanics, Techno-Press, V.l, No.l, October 1993,
pp.75-86.
Loo, Y.C. and Falamaki, M. (1992), "Punching Shear Strength Analysis of Reinforced
Concrete Hat Plates with Spandrel Beams", ACI Structural Journal, V.89, No.4,
July-August, 1992, pp.375-383.
231
References
Neville, A.M. (1978), Properties of Concrete, 2nd Ed., Pitman, London, 1978, 687 p.
Nie, N.H., Hull, CH. and Bent, D. (1975), SPSS Statistical Package for the Social
Sciences, McGraw-Hill Book Co., New York, 1975, 675 p.
Park, R and Gamble, W.L. (1982), Reinforced Concrete Slab , John Wiley & Sons, New
York, 1982, 618 p.
Rangan, B.V. (1987), "Punching Shear Strength of Reinforced Concrete Slabs", Civil
Engg. Trans., IEAust, Vol.CE29, No.2, 1987, pp.71-78.
Rangan, B.V. (1990a), "Punching Shear Design in the New Australian Standard for
Concrete Structures", ACI Structural Journal, V.87, No.2, March-April 1990,
pp. 140-144.
Rangan, B.V. (1990b), "Tests on Slabs in the Vicinity of Edge Columns", ACI
Structural Journal, V.87, No.6, November-December 1990, pp.623-629.
Rangan, B.V. and Hall, A.S. (1983a), Forces in the Vicinity of Edge Columns in Flat
Plate Floors, Uni-Civ Report No.R-203, University of NSW, Kensington, Vol.1,
240 p. and Vol.2, 1983,149 p.
Rangan, B.V. and Hall, A.S. (1983b), "Moment and Shear Transfer Between Slab and
Column", ACI Structural Journal, V.80, No.3, May-June, 1983, pp.183-191.
Regan, P.E. (1974), "Design for Punching Shear", The Structural Engineer, Vol.52.,
No.6., June 1974, pp. 197-207.
232
References
Regan, P.E. (1981), Behaviour of Reinforced Concrete Flat. Slabs, CIRIA Report
No.89, Construction Industry Research and Information Association, London,
1981, 89 p.
Regan, P.E. (1985), "The Punching Resistance of Prestressed Concrete Slabs", Proc.
Instn Civ.Engrs, Part 2, 79, London, 1985, pp.657-680.
Sozen, M.A. and Siess, CP. (1963), "Investigation of Multiple-panel Reinforced
Concrete Floor Slabs: Design Methods - Their Evolution and Comparison", ACI
Structural Journal, V.60, No..8, August, 1963, pp.999-1027.
Standards Association of Australia (SAA 1988), Australian Standard 3600-1988
Concrete Structures , Sydney, 105 p.
Vanderbilt, M.D. (1970), "Shear Strength of Continuous Plates", Proc. ASCE, V.98,
ST5, May 1972, pp.961-973.
Walpole, R.E. and Myers, R.H. (1985), Probability and Statistics for Engineers and
Scientists, 3rd Ed., Collier Macmillan, New York, 1985, 639 p.
Warner, R.F. and Faulkes, K.A. (1988), Prestressed Concrete, 2nd Ed., Longman
Cheshire, Melbourne, 1988,433 p.
Warner, R.F., Rangan, B.V. and Hall, A.S., Reinforced Concrete, 3rd Edition,
Longman Cheshire, Melbourne, Australia, 1989, 553 p.
Zaghlool, R.F., de Pavia, H.A.R. and Glockner, P.G. (1970), "Tests of Reinforced
Concrete Hat Plate Floors", Proc. ASCE, V.96, ST3, May 1970, pp.487-507.
233
Appendix A
APPENDIX A
CALIBRATION OF PRESTRESSING JACK
Conversion factor : 1 mm = 0.039 in.
I N = 0.22431b.
1 kPa = 145 psi.
(IMPa = 0.145 psi)
A.l Actual Stress on Prestressing tendon
5.05 m m diameter (or 0.197") of prestressing tendon
giving a cross-sectional area, Ap
n [ 0.197 ]2 _ Mne. . A = — — = 0.0305 sq.in.
P 4
The required prestressing force to be applied on each tendon was 23.03 kN. This was
equivalent to,
23.03 x 103 N x — = 5164.7 lbs. 9.81
Therefore, the applied prestress on each tendon was,
234
Appendix A
5164.7
0.0305 = 169335.4 psi
A.2 Stress Exerted by Jack
Since we require a prestress force of 23.03 kN to be applied on each tendon, the stress to
be exerted by the jack in order to achieve the required force may be determined by the
following manner.
Jack cylinder
32 m m (1.26")
Cross-sectional area, A
. n [ 1.972 - 1.262 ] 1fic . A = = 1.8 sq.in
For a stress application of 2000 psi, the required force needed to be exerted on the jack
cylinder was,
2000x1.8 = 3600lbs. (or 16.4kN)
and for an applied stress of 3000 psi, the required force was,
235
Appendix A
3000x1.8 = 5400lbs. (or24.5 kN)
In order to get an applied force of 5164.7 lbs. or 23.03 kN on each tendon, the applied
stress by the jack was,
2000 + [ 23-03 ' 16A x 1000 psi ] = 2819 psi 24.5 - 16.4 F F
which is the calibrated jacking force to be applied during the post-tensioning process.
236
Appendix B
APPENDIX B
DEFLECTION MEASUREMENTS
B.l Deflection Measurements of Model PI
Table B.la -Deflection measurements of Model PI (using rulers)
Applied load
(kPa)
0
4.46
5.93
7.33
8.69
10.18
11.59
13.18
14.89
16.18
Deflection measurement (mm) using rulers for both panels at location
1 0
0.5
1.0
1.5
2.0
3.5
5.0
7.0
10.5
37.0
^2 0
0.1
0.3
0.4
0.5
0.9
1.5
3.3
5.0
8.0
. 4 0
0.2
0.5
0.7
1.0
2.0
3.0
5.0
8.0
34.0
5
0 0.7
1.5
2.0
2.5
4.2
6.0
9.0
13.0
42.0
6 0
0.2
0.5
0.8
1.0
2.0
3.0
5.5
8.0
13.0
7 0
0.1
0.2
0.3
0.5
0.9
1.5
2.5
4.0
7.0
9 0
0.2
0.4
0.4
0.5
1.5
2.5
3.2
4.0
34.5
Table B.lb -Deflection measurements of Model PI (using LVDTs)
Applied load (kPa) 0
4.46
5.93
7.33
8.69
10.18
11.59
13.18
14.89
16.18
Deflection measurement (mm) using L V D T
Centre of panel ABGF (at location 3)
0
0.55
1.10
1.64
2.88
5.20
7.67
11.10
24.66
46.68
Centre of panel BCHG (at location 8)
0
0.41
0.96
1.51
2.47
3.97
5.89
8.90
15.07
45.25
237
Appendix B
B.2 Deflection Measurements of Model P2
Table B.2a -Deflection measurements of Model P2 (using rulers)
Applied load
(kPa)
0
1.37
3.28
4.37
5.69
6.96
9.86
11.27
12.48
13.21
14.69
15.86
16.17
17.23
18.44
19.61
Deflection measurement (mm) using rulers for both panels at location
1
0
0.6
2.4
2.8
3.4
4.1
5.2
8.0
12.5
15.4
19.2
22.8
25.8
31.6
37.9
46.1
2
0
0.1
0.5
0.7
1.0
1.2
1.7
2.2
3.3
4.5
5.3
6.9
7.4
8.2
10.4
13.4
4
0
0.1
0.3
0.5
0.6
0.8
1.2
2.7
3.4
3.5
3.6
4.3
5.3
5.7
6.4
7.7
5
0
0.1
0.5
1.1
1.8
2.8
3.3
9.0
13.3 '
16.5
19.6
24.5
29.8
33.7
36.5
41.9
6
0
0.3
1.1
1.4
2.0
2.7
3.9
7.3
12.0
14.2
17.6
20.1
28.1
36.0
42.1
49.3
7
0
0.2
0.4
0.5
0.8
0.9
1.2
2.4
3.2
3.7
4.2
5.0
5.6
5.8
9.7
13.4
9
0
0.1
0.3
0.3
0.4
0.5
1.3
3.0
3.8
4.3
4.7
5.1
6.0
7.1
8.2
11.9
238
Table B.2b -Deflection measurements of Model P2 (using LVDTs)
Applied load
(kPa)
0
1.37
3.28
4.37
5.69
6.96
9.86
11.27
12.48
13.21
14.69
15.86
16.17
17.23
18.44
19.61
Deflection measurement (mm) using L V D T
Centre of panel ABGF (at location 3)
0
0.67
1.80
2.53
3.33
4.50
6.21
11.25
17.05
20.81
25.25
33.57
42.44
44.41
51.84
66.33
Centre of panel BCHG (at location 8)
0
1.14
1.74
3.43
3.62
4.62
6.28
11.72
17.75
22.05
25.98
34.61
43.18
45.28
52.85
66.68
Appendix B
B.3 Deflection Measurements of Model P3
Table B.3a - Deflection measurements during first day of testing
of Model P3 (using rulers)
Applied load
(kPa)
0
1.40
2.80
4.20
5.60
7.00
8.40
9.80
11.20
12.60
14.00
14.68
15.40
17.61
18.17
19.57
20.27
Deflection measurement (mm) using rulers for both panels at location
1
0
0
1.0
1.0
1.5
3.0
3.5
4.5
5.5
7.0
8.0
9.5
11.0
14.0
15.0
18.5
22.0
2
0
0
1.0
1.0
1.5
2.0
2.5
3.0
4.0
6.0
6.5
8.0
9.0
11.0
11.5
13.5
15.0
3
0
0
1.0
1.0
2.5
4.0
5.5
7.0
10.0
14.0
16.0
19.5
23.0
30.0
32.0
40.0
47.0
4
0
0
0
1.0
1.0
1.0
1.5
2.0
3.0
6.0
6.0
7.5
9.0
11.5
12.5
15.0
17.0
5
0
0
1.0
2.0
3.0
4.0
5.0
6.5
8.0
9.0
10.0
12.0
13.5
16.0
18.5
23.0
28.0
6
0
0
0
0
1.0
2.0
3.0
3.5
4.0
5.0
6.5
7.5
8.5
10.5
11.5
13.5
15.5
7
0
0
0
0.5
1.0
1.0
1.0
1.0
2.0
2.5
3.0
3.0
3.5
4.0
5.0
5.5
6.5
8
0
2.0
2.0
3.0
3.0
4.0
5.5
6.5
8.0
10.0
11.0
13.0
15.0
19.0
21.0
26.0
32.0
9
0
0
1.0
1.0
1.5
2.0
2.0
7.0
7.0
8.5
9.5
10.0
11.0
12.0
13.0
14.5
16.0
240
Appendix B
Table B.3b - Deflection measurements during second day of testing of Model P3 (using rulers)
Applied load
(kPa)
0
7.97
14.0
16.77
18.73
20.27
21.52
22.36
Deflection measurement (mm) using mlers for both panels at location
1
0
2.00
8.00
13.00
15.00
17.50
20.00
26.50
2
0
0.50
3.50
4.00
5.50
6.50
7.50
9.50
3
0
4.00
15.00
22.00
25.00
30.00
35.00
45.00
4
0
1.50
5.50
10.00
11.00
12.50
17.00
25.00
5
0
3.00
10.00
14.50
17.00
20.00
23.50
34.00
6
0
1.50
5.50
7.00
8.00
9.50
11.00
24.50
7
0
0.50
3.00
3.50
4.00
4.50
5.50
8.00
8
0
3.50
10.50
15.50
18.00
21.00
25.00
48.00
9
0
1.00
3.00
9.00
10.50
13.00
14.50
17.00
Table B.3c - Deflection measurements during first day of testing
of Model P3 (using dial gauges)
Applied load
(kPa)
0
1.40
2.80
4.20
5.60
7.00
8.40
9.80
11.20
12.60
14.00
14.68
15.40
17.61
18.17
19.57
20.27
Deflection measurement (mm) using dial gauges for both panels at location
1
0
0.25
0.64
1.13
2.44
2.53
3.53
4.34
5.56
6.92
7.89
9.12
10.80
13.09
14.34
17.94
22.09
2
0
0.13
0.22
0.45
0.76
1.15
1.76
2.17
3.15
4.76
5.73
7.11
8.41
10.09
10.75
12.67
14.06
4
0
0.16
0.39
0.70
0.97
1.36
1.91
2.49
3.76
4.72
6.97
8.34
9.78
12.43
13.50
16.58
18.98
5
0
0.32
0.93
1.72
2.60
3.52
4.84
5.79
7.46
8.59
9.80
11.39
13.10
16.60
17.86
20.82
22.41
6
0
0.24
0.63
1.11
1.66
2.16
3.15
3.43
4.35
5.46
6.27
8.49
9.51
11.59
12.38
14.05
16.75
7
0
0.80
1.14
1.33
1.50
1.68
1.94
2.15
2.51
3.09
3.41
3.99
4.48
5.05
5.50
6.47
7.40
9
0
0
0.12
0.27
0.41
0.62
0.93
1.24
2.01
2.58
3.99
4.82
5.68
7.14
7.78
9.51
11.31
Appendix B
Table B.3d - Deflection measurements during second day of testing
of Model P3 (using dial gauges)
Applied load
(kPa)
0
7.97
14.0
16.77
18.73
20.27
21.52
22.36
Deflection measurement (mm) using dial gauges for both panels at location
1
0
2.90
8.18
11.38
12.17
15.53
18.53
2
0
1.36
4.25
6.01
6.67
7.92
8.84
-
4
0
2.09
6.17
8.67
9.82
11.39
12.66
28.50
5
0
5.00
12.02
16.15
18.12
21.13
25.14
-
6
0
3.52
7.47
9.84
10.89
12.52
14.52
-
7
0
1.77
3.50
4.47
4.89
5.52
6.21
-
9
0
1.05
3.70
4.86
5.79
6.72
8.45
19.01
Table B.3e - Deflection measurements during first day of testing
of Model P3 (using LVDTs)
Applied load
(kPa)
0
1.40
2.80
4.20
5.60
7.00
8.40
9.80
11.20
12.60
14.00
14.68
15.40
17.61
18.17
19.57
20.27
Deflection measurement (mm) using L V D T
Centre of panel ABFG (at location 3)
0
0.30
0.70
1.90
3.10
4.50
5.90
7.40
10.10
12.90
16.70
18.90
22.90
27.90
32.40
40.10
46.30
Centre of panel BCGH (at location 8)
0
0.30
0.70
1.80
2.70
3.30
4.70
5.80
6.90
8.50
10.30
11.90
14.40
16.80
20.30
25.10
29.70
242
Appendix B
Table B.3f - Deflection measurements during second day of testing
of Model P3 (using L V D T s )
Applied load
(kPa)
0
7.97
14.0
16.77
18.73
20.27
21.52
22.36
Deflection measurement (mm) using L V D T
Centre of panel ABFG (at location 3)
0
6.20
17.50
24.50
27.70
32.90
51.40
93.10
Centre of panel BCGH
(at location 8) 0
7.10
15.40
20.00
22.50
25.70
33.70
72.50
B.4 Deflection Measurements of Model P4
Table B.4a - Deflection measurements during first day of testing
of Model P4 (using rulers)
Applied load
(kPa) 0
1.37
2.75
4.43
5.71
6.89
8.75
9.72
11.47
12.77
12.70
Deflection measurement (mm) using rulers for both panels at location
1 0
0
0.9
1.1
2.0
2.9
4.0
5.0
6.9
10.0
4.0
2 0
0
0.9
1.0
1.0
2.0
2.5
2.5
3.1
6.0
3.0
3 0
0.5
1.0
2.0
3.0
4.5
6.0
7.0
10.0
16.0
7.0
4 0
0
1.0
1.0
1.5
2.0
2.5
3.0
4.0
7.0
5.0
5 0
0.2
0.5
1.0
1.4
2.0
2.5
3.0
4.1
5.5
2.2
6 0
0
0
1.0
1.5
3.0
3.0
3.0
4.0
5.0
5.0
7 0
0
0
0
0.5
1.0
1.5
1.5
2.0
3.5
7.0
8 0
0
0.5
1.5
2.0
4.0
5.0
6.0
8.5
12.0
4.5
9 0
0
0
0
0.5
1.0
1.5
1.5
2.0
3.5
1.0
243
Appendix B
Table B.4b - Deflection measurements during second day of testing
of Model P4 (using rulers)
Apphed
load
(kPa)
0
0.67
1.49
2.11
2.92
3.38
4.4
5.76
7.06
8.39
9.37
11.03
12.92
14.33
15.64
16.04
16.64
Deflection measurement (mm) using rulers for both panels at location
1
0
0
0.5
0.5
1.3
1.4
2.0
2.7
3.5
4.5
5.5
6.5
7.8
10.0
12.5
15.5
66.5
2
0
0
0.2
0.2
0.2
0.2
1.2
1.2
1.7
2.2
3.2
3.2
4.2
5.2
7.2
8.2
14.7
3
0
0.1
1.0
1.0
2.0
2.0
3.5
5.0
6.0
7.5
9.5
11.0
13.0
17.0
22.0
26.0
131.0
4
0
0
0.4
0.9
0.9
0.9
1.4
1.9
2.7
2.9
3.9
4.7
5.9
6.9
8.9
9.9
36.9
5
0
0
0.1
0.4
0.6
0.6
1.0
1.5
2.0
2.5
3.0
3.5
4.0
5.5
7.0
8.0
45.0
6
0
0
1.0
1.0
1.5
2.0
2.0
3.0
3.1
4.0-
5.0
5.0
6.0
8.0
11.0
13.0
69.5
7
0
0
0.5
0.9
1.0
1.0
1.2
2.0
2.1
3.0
3.0
4.0
4.5
6.0
7.0
8.5
23.0
8
0
0.2
1.0
1.0
2.0
2.0
3.0
4.0
5.0
6.0
7.5
9.0
11.0
14.5
19.5
23.0
87.0
9
0
0
0.5
0.5
0.8
0.9
1.0
1.5
2.0
2.0
3.0
3.0
4.0
5.0
6.0
7.0
36.0
244
Appendix B
Table B.4c - Deflection measurements during first day of testing
of Model P4 (using dial gauges)
Applied load
(kPa)
0
1.37
2.75
4.43
5.71
6.89
8.75
9.72
11.47
12.77
12.70
Deflection measurement (mm) using dial gauges for both panels at location
1
0
0.45
0.97
0.97
1.38
1.68
2.88
3.79
4.90
5.80
1.76
2
0
0
0.71
0.89
0.91
1.74
2.20
2.29
2.87
5.37
2.96
4
0
0
0.56
0.89
0.89
1.42
2.03
2.31
3.67
6.63
1.00
5*
-
-
-
-
-
-
-
-
-
-
-
6
0
0.25
0.64
1.33
1.98
3.81
4.03
4.97
6.63
7.11
1.87
7
0
0
0.35
0.45
0.45
0.60
1.83
2.80
3.36
4.80
1.78
9
0
0
0.82
1.40
1.60
1.80
2.22
2.70
2.78
3.94
1.41
Dial gauge 5 was faulty during testing of Model P4
245
Appendix B
Table B.4d - Deflection measurements during second day of testing
of Model P4 (using dial gauges)
Applied load
(kPa)
0
0.67
1.49
2.11
2.92
3.38
4.4
5.76
7.06
8.39
9.37
11.03
12.92
14.33
15.64
16.04
16.64
Deflection measurement (mm) using dial gauges for both panels at location
1
0
0
0.60
0.90
1.20
1.62
1.94
2.50
3.00
4.65
5.25
6.84
7.20
8.91
10.82
10.90
-
2
0
0
0.11
0.17
0.29
0.39
0.98
1.05
1.54
2.01
2.98
3.31
4.15
4.98
7.10
8.10
-
'4
0
0.25
0.41
0.46
0.69
0.76
1.27
2.00
2.39
2.92
3.66
4.32
5.62
6.73
9.02
9.86
-
5*
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
6
0
0.10
0.36
0.40
0.72
0.81
1.33
1.98
2.49
3.22
4.30
4.38
7.10
8.43
9.49
12.96
-
7
0
0
0.90
1.70
1.70
1.70
2.00
2.00
2.60
3.39
3.51
3.63
4.02
5.00
6.88
7.68
-
9
0
0.10
0.28
0.41
0.60
0.68
1.07
1.65
2.31
2.91
3.48
3.61
4.50
4.62
6.01
6.86
-
246
Table B.4e - Deflection measurements during first day of testing
oi Model P4 (using LVDTs)
Applied load
(kPa)
0
1.37
2.75
4.43
5.71
6.89
8.75
9.72
11.47
12.77
12.70
Deflection measurement (mm) using L V D T
Centre of panel ABFG
(at location 3)
0
3.00
4.00
4.00
5.10
8.00
9.00
11.50
12.00
16.50
6.00
Centre of panel BCGH
(at location 8)
0
5.00
5.00
6.00
8.00
8.00
10.00
11.00
13.00
15.50
7.00
247
Appendix B
Table B.4f - Deflection measurements during second day of testing
of Model P4 (using LVDTs)
Applied load
(kPa)
0
0.67
1.49
2.11
2.92
3.38
4.4
5.76
7.06
8.39
9.37
11.03
12.92
14.33
15.64
16.04
16.64
Deflection measurement (mm) using L V D T
Centre of panel ABFG
(at location 3)
0
2.00
2.50
2.50
3.50
3.50
11.00
13.00
13.50
14.50
16.00
17.50
19.00
21.50
25.00
28.00
132.00
Centre of panel BCGH
(at location 8)
0
1.00
2.50
2.50
3.00
12.00
12.50
13.50
14.00
16.00
16.50
17.00
18.00
21.50
25.00
27.50
85.00
Appendix C
APPENDIX C
PLOTS OF EXPERIMENTAL MOMENT, SHEAR AND TORSION
RESULTS FOR POST-TENSIONED CONCRETE FLAT PLATE
MODELS
C.l Plots of Experimental Moment, Shear and Torsion Versus Applied Loads at
Corner-connections A and C and at Edge-connection B for Model PI
20
15
9 10
a o
s a
Column A Column B Column C
70
^ ' r 0 5 10 15 20 Applied load (kPa)
• Column A Column B •Column C
0 5 10 15 20 Applied load (kPa)
Fig. CI. 1 - Plot of moment transfer to columns A, B and C versus applied load (Model PI)
Fig. CI.2 - Plot of shear strength around columns A, B and C versus applied load (Model PI)
249
Appendix C
• Column A • Column B - Column C
T i '—r O 5 IO 15 20
Applied load (kPa)
Fig. C1.3 - Plot of torsional moment at columns A, B and C versus
applied load (Model PI)
C.2 Plots of Experimental M o m e n t , Shear and Torsion Versus Applied Loads at
Corner-connections A and C and at Edge-connection B for Model P3 *
25 -
"a 20 -55
m. 1 5 '
a 0
8
5 -
0 -i
— a — t,oiumn A —*— Column B
* — > 1 *-.
jl^P
f*=m | , i"
. .
25 -
f 20" 55
"~ 15 -<=l
B 0 10 j
5 -
0 1?
— ° — Column A —*— Column B — 0 — Column C
3 1 | 1 1
0 IO 20 30 Applied load (kPa)
0 IO 20 30 Applied load (kPa)
(a) Day O n e (b) Day T w o
Fig. C2.1 - Plot of moment transfer to columns A, B and C versus applied load (Model P3)
The stress measurements were recorded over two days of testing of the slab Model P3 due to the sudden failure of some bolted nuts in the whiffle-tree loading system at the near ultimate loading of 21.0 kPa. The model was re-loaded the next day and all measurements were re-taken from the beginning up to failure.
250
Appendix C
Column A Column B Column C
0 IO 20 30
Applied load (kPa)
(a) Day One
so -
70 -
60 -
5^, 50 -
g 40 "
8 0 3 0 -
20 -
IO -
O i
A
7 • \ '
— 0 — Column —*— Column —"— Column
1
A B C
O 10 20 30
Applied load (kPa)
(b) Day Two
Fig. C2.2 - Plot of shear strength around columns A, Band C versus applied load (Model P3)
- Column A - Column B - Column C
O IO 20 30
Applied load (kPa)
(a) Day One
• Column A - Column B • Column C
O IO 20 30
Applied load (kPa)
(b) Day Two
Fig. C2.3 - Plot of torsional moment at columns A, B and C versus applied load (Model P3)
251
Appendix C
C.3 Plots of Experimental Moment, Shear and Torsion Versus Applied Loads at Corner-connections A and C and at Edge-connection B for Model P4 **
15 - Column A - Column B Column C
20
' Column A ' Column B • Column C
0 5 10 15 Applied load (kPa)
(a) Day One
n ' i r~ O 5 IO 15 20 Applied load (kPa)
(b) Day Two
Fig. C3.1 - Plot of moment transfer to columns A, B and C versus applied load (Model P4)
DO
50 -
40 -55
^ 3 0 -«H
es u
rtA M 20 -
10 -
0 -t
— a — Column A
— a — Column C
r , , | u. 1
70
0 5 IO 15 Applied load (kPa)
(a) Day One
Column A Column B Column C
! 1 ' 1 ' ' 1
0 5 10 15 20 Applied load (kPa)
(b) Day Two
Fig. C3.2 - Plot of shear strength around columns A, B and C versus applied load
(Model P4)
The stress measurements were recorded over two days of testing of the slab Model P4 due to the sudden failure of a threaded steel rod in the whiffle-tree loading system at a loading of 12.7 kPa. The model was re-loaded the next day and all measurements were re-taken from the beginning up to failure.
252
-a— Column A -*— Column B -A— Column C
9
0 5 10 15
Applied load (kPa)
(a) Day One
B 55 ui.1 a o
-u—1
OT »H
o E-H
/
6
5
4
3
7.
1
O
-
-
A 7
— o — Column —*— Column — o — Column
_j_ . _i_iUL - i : 1
A B C
0 5 10 15 20
Applied load (kPa)
(b) Day Two
Fig. C3.3 - Plot of torsional moment at columns A, B and C versus applied load (Model P4)
253
Appendix D
APPENDIX D
TABULATIONS OF EXPERIMENTAL MOMENT, SHEAR AND
TORSION RESULTS FOR POST-TENSIONED CONCRETE
FLAT PLATE MODELS
Table D.l - Shear, moment and torsion at columns A, B and C for Model PI
Applied load
(kPa)
0
4.46
5.93
7.33
8.69
10.18
11.59
13.18
14.89
16.18
Shear (kN) at column
A
0
8.39
10.66
12.93
15.07
17.43
19.66
21.87
25.14
26.93
B
0
24.25
30.20
35.71
40.97
46.86
52.32
58.36
63.35
68.75
C
0
8.23
10.49
12.74
14.96
17.43
19.69
21.88
25.51
27.72
Moment (kN-m) at column
A
0
0.46
1.07
1.72
2.28
2.89
3.52
4.43
6.83
7.39
B
0
1.69
2.92
3.94
5.36
7.38
9.38
11.80
13.62
15.25
C
0
0.52
1.14
1.95
2.91
4.05
5.16
6.32
7.80
8.70
Torsion (kN-m) at column
A
0
0.85
1.45
2.04
2.88
-3.75
4.56
5.80
6.84
7.26
B
0
1.24
1.37
1.40
1.45
1.66
1.81
2.12
2.20
2.35
C
0
0.67
1.04
1.60
2.17
2.70
3.17
4.41
5.61
6.14
254
Appendix D
Table D.2 - Shear, moment and torsion at columns A, B and C for Model P2
Applied , load
(kPa)
0
1.37
3.28
4.37
5.69
6.96
9.86
11.27
12.48
13.21
14.69
15.86
16.07
17.23
18.44
19.61
Shear (kN) at column
A
0
3.01
4.23
6.11
7.86
9.91
14.41
17.73
19.74
21.18
24.07
26.08
26.64
28.52
30.23
32.02
B
0
10.20
13.79
18.34
23.82
28.89
38.53
43.58
47.96
50.92
56.57
60.73
60.95
65.89
71.01
76.33
C
0
3.36
4.64
6.10
8.32
10.00
14.91
17.65
19.86
20.66
23.27
25.00
25.48
27.39
29.16
32.97
Moment (kN-m) at column
A
0
0.52
0.75
1.20
1.64
2.29
4.49
6.56
7.70
8.68
10.17
11.17
11.57
11.88
11.95
12.01
B
0
2.44
3.41
4.73
6.32
7.90
10.35
12.47
14.62
19.54
23.84
24.75
25.61
26.55
27.51
28.50
C
0
1.52
2.04
2.66
3.50
4.24
5.30
6.04
6.77
7.25
7.94
8.67
8.80
9.69
10.47
11.17
Torsion (kN-m) at column
A
0
1.15
1.68
2.34
3.04
3.79
4.85
5.45
5.68
5.87
6.15
6.31
6.59
7.06
7.12
7.35
B
0
0.04
0.05
0.04
0.05
0.07
0.14
0.14
0.10
0.15
0.09
0.14
0.17
0.27
0.32
0.28
C
0
0.05
0.30
0.62
1.14
1.57
3.34
4.59
5.46
5.92
7.49
8.11
8.26
8.71
9.19
9.51
255
Appendix D
Table D.3a - Shear, moment and torsion at columns A, B and C for Model P3
- Day One
Applied load
(kPa)
0
1.40
2.84
4.35
5.69
7.14
8.81
10.04
11.79
13.00
14.03
15.22
17.52
18.16
19.51
20.49
Shear (kN) at column
A
0
1.75
3.89
6.12
7.95
10.01
12.32
14.16
16.86
19.40
21.01
22.77
25.17
26.44
28.06
29.63
B
0
4.99
10.83
17.11
22.70
28.81
35.63
40.65
47.39
50.73
54.41
59,20
64.14
66.14
70.19
72.62
C
0
1.86
4.02
6.28
8.27
10.48
13.01
14.91
17.74
20.35
22.08
24.29
28.18
29.42
31.99
33.58
Moment (kN-m) at column
A
0
0.35
0.76
1.11
1.31
1.54
1.99
2.65
3.88
5.36
6.13
7.10
8.19
9.02
10.57
11.27
B
0
0.76
1.59
2.71
3.91
5.61
7.49
8.90
11.01
13.18
15.28
17.65
20.31
28.26
24.69
27.46
C
0
0.51
1.16
2.04
2.94
4.26
5.53
6.48
7.05
7.68
8.37
9.18
9.94
10.94
11.70
12.29
Torsion (kN-m) at column
A
0
0.30
0.58
0.88
1.12
1.43
1.75
2.05
2.20
2.63
3.18
3.74
4.14
4.85
5.19
6.11
B
0
0.05
0.06
0.08
0.09
0.10
0.11
0.12
0.13
0.15
0.12
0.15
0.22
0.18
0.25
0.20
C
0
0.35
0.60
0.96
1.21
1.48
1.78
2.01
2.25
2.73
3.27
3.82
4.48
4.97
5.40
6.25
256
Appendix D
Table D.3b - Shear, moment and torsion at columns A, B and C for Model P3 - Day Two
Applied load
(kPa)
0
7.98
14.06
17.36
20.36
21.45
22.42
Shear (kN) at column
A
0
10.33
18.25
23.40"
26.60
28.43
30.97
B
0
22.12
44.95
56.26
66.33
70.01
73.35
c 0
10.65
18.61
24.01
27.49
29.70
31.78
Moment (kN-m) at column
A
0
5.14
7.84
9.81
10.53
11.94
12.25
B
0
8.28
15.60
19.60
25.11
30.96
33.01
C
0
5.25
8.46
10.51
12.23
13.11
13.54
Torsion (kN-m) at column
A
0
2.38
3.42
5.03
6.54
6.95
7.29
B
0
0.03
0.29
0.44
0.36
0.34
0.28
c 0
3.03
4.18
5.22
6.15
7.88
8.23
Table D.4a - Shear, moment and torsion at columns A, B and C for Model P4
- Day One
Applied load
(kPa)
0
1.37
2.75
4.43
5.71
6.89
8.75
9.72
11.47
12.77
12.70
Shear (kN) at column
A
0
2.49
4.94
7.82
9.87
11.66
14.55
16.29
18.70
21.23
21.10
B
0
5.29
10.85
17.51
22.80
27.82
35.74
40.18
47.44
52.21
51.94
C
0
2.12
4.25
6.99
9.16
11.06
14.29
16.11
18.80
21.33
21.24
Moment (kN-m) at column
A
0
0.53
1.14
1.94
2.58
3.36
4.45
5.11
5.83
5.25
5.21
B
0
0.25
0.61
1.24
2.22
3.92
6.44
7.58
10.25
12.66
12.60
C
0
0.37
0.74
1.41
2.10
2.98
4.63
5.18
6.73
7.80
7.77
Torsion (kN-m) at column
A
0
0.42^
0.96
1.55
1.97
2.24
2.85
3.17
3.85
5.76
5.71
B
0
0.05
0.03
0.03
0.07
0.15
0.15
0.12
0.12
0.15
0.13
c 0
0.32
0.73
1.35
1.82
2.22
2.89
3.07
3.74
4.15
4.14
257
Appendix D
Table D.4b - Shear, moment and torsion at columns A, B and C for Model P4
- Day Two
Applied load
(kPa)
0
0.67
1.49
2.11
2.92
3.38
4.4
5.76
7.06
8.39
9.37
11.03
12.92
14.33
15.64
16.04
16.64
Shear (kN) at column
A
0
0.92
2.28
3.32
4.63
5.38
7.04
9.35
11.49
13.68
15.03
18.00
21.06
23.72
25.86
26.65
27.64
B
0
2.22
5.46
8.01
11.3
13.24
17.38
22.86
28.23
33.71
37.77
44.57
52.16
57.53
62.71
64.05
66.43
C
0
0.84
2.08
3.14
4.47
5.44
7.08
9.34
11.55
13.85
15.84
18.36
21.55
24.03
26.24
27.21
28.22
Moment (kN-m) at column
A
0
0.20
0.55
0.83
1.21
1.46
1.97
2.61
3.28
3.94
4.37
5.10
5.78
6.40
6.90
7.08
7.34
B
0
0.14
0.5
0.89
1.41
1.71
2.64
4.12
5.61
7.15
8.46
10.27
12.47
14.26
15.86
15.83
16.42
C
0
0.11
0.35
0.67
1.11
1.44
2.05
2.84
3.65
4.49
5.23
6.14
7.26
8.07
8.79
9.21
9.55
Torsion (kN-m) at column
A
0
0.13
0.44
0.63
0.93
1.09
1.48
2.02
2.54
3.09
3.50
4.23
4.99
5.59
6.07
6.08
6.31
B
0
0.04
0.07
0.09
0.09
0.09
0.09
0.09
0.10
0.10
0.10
0.10
0.09
0.12
0.14
0.12
0.13
C
0
0.13
0.40
0.58
0.81
0.96
1.25
1.66
2.08
2.52
2.88
3.38
3.98
4.45
4.78
4.92
5.10
258
Appendix E
APPENDIX E
FLEXURAL STRENGTH THEORY
E.l Assumptions
The assumptions made in the flexural strength theory are:
• variation of strain on the cross-section is linear, i.e. strains in the concrete and
bonded steel are calculated on the assumptions that plane sections remain plane;
• concrete carries no tensile stress, i.e. tensile strength of the concrete is ignored;
and
• stress in the compressive concrete and steel (prestressed and non-prestressed)
are determined from actual or idealised stress-strain relationships for the
respective materials.
The load at which failure or collapse of a flexural member occurs is called the ultimate
load, and the corresponding moment is termed as the ultimate moment. As the applied
moment increases into the overload range, the neutral axis of the section rises and
eventually the material behaviour becomes non-linear.
259
...
Appendix E
Neutral axis
Centroidal axis
® Decompression moment M 0
@ Cracking moment Mc r
<3> Post-cracking moment Mpc
© Ultimate moment Mu
Cross section Strains at various load stages
'SC
Stage® Stage© Staged) Stage®
Stresses at different load stages
C
Resultant forces
Fig. E.l - Typical strain and stress distribution of a prestressed concrete
cross-section
The depth to the neutral axis of a section is denoted as dn in Fig. E.l . Non-prestressed
steel may yield if its strain est exceeds the yield strain £Sy (= fsylEs). The prestressed
tendon may enter the non-linear part of its stress-strain curve as its strain, Ep , increases.
260
Appendix E
At the ultimate moment state (see Fig. E.l), the stress distribution shows that the
resultant compressive force of magnitude C equals the resultant tensile force T and the
ultimate moment capacity is calculated from the internal couple,
Mu = Cl=Tl ...(E-l)
where / is the lever arm between the internal compressive and tensile resultants.
E.2 Idealised compressive stress block
The idealised stress block at ultimate state, shown in Fig. E.2, is recommended by
ACI318-89 and A S 3600-1988. Both of these Concrete Codes also take the extreme
fibre compressive strain Ecu at ultimate moment to be 8 C M = 0.003, although it has
been mentioned that the actual extreme fibre strain at ultimate may not be close to this
value [Gilbert and Mickleborough 1990]. The use of the idealised compressive stress
block as specified by these two codes is adopted in Section 9.3.2 in the analysis of
flexural strength of slab strips.
The parameter 7" is dependent on the compressive strength of the concrete and is
described in the Australian Standard A S 3600-1988 [SAA, 1988] as shown below:
For f'c < 28 MPa : y" = 0.85 ...(E-2)
For 7% > 28 M P a : y" = 0.85 - 0.007 ( f'c - 28 ) ^ 0 . 6 5 ...(E-3)
261
Appendix E
For the rectangular section shown in Fig. E.2, the hatched area A' ( = y"bdn ) is
therefore assumed to be subjected to a uniform stress of 0.85/c. The resultant
compressive force Cc is the volume of the rectangular stress block.
Cc= 0.85fcy bdn .(E-4)
Neutral axis
~~ Centroidal axis
Cross section Strains at ultimate
a
Jr.
cu
OL
'pu
Jr. Y"d„
3: mt
O,
Q.5Ydn
"*
pu
'sou Cc ~f~f
TP ^ 1
~f
Actual stresses Idealised stress block Resultant forces
Fig. E.2 - Idealised stress block at ultimate m o m e n t conditions
262
Appendix E
Based on Eq.(E-l), the ultimate moment is thus,
Mu = Tplp + Ts h - CSc lsc ...(E-5)
where lp , ls and lsc are the lever arms between the concrete compressive force and the
respective forces at the prestressed and non-prestressed steel levels. Mu may then be
expanded into the following form:
Mu~ vpu\[dp- 1j^]+*suAst[ds- ^1-a.AtY ~dc]
...(E-6)
where opu, osu and ascu are the ultimate stresses in the bonded tendons and non-
prestressed tensile and compressive steel respectively, which are determined from the
consideration of strain compatibility and equilibrium. The various depths shown, ie. dp,
ds and dc are the effective depths of the prestressing steel, the tensile reinforcement and
the compressive reinforcement (if any).
E.3 Prestressed tendon strain components
Unlike non-prestressed steel in reinforced concrete section, the strain in the tendons in
prestressed concrete section is not equal to the concrete at tendon level. The strain in
the bonded prestressing tendon at any stage of loading is equal to the strain caused by
the initial prestress plus the change in strain in the concrete at the tendon level.
263
Appendix E
Centroidal axis of
ia) uncracked section
<fij) o
Fig. E.3 - Strain distribution of prestressed tendon
at three stages of loading
E.3.1 Stage 1
The elastic concrete compressive strain, Ece (denoted by stage la in Fig. E.3) is caused
by effective prestress when there is no externally applied moment. The corresponding
elastic strain in the prestressing tendon, £pe is denoted as stage lb in Fig. E.3 .
E.3.2 Stage 2
The applied moment is sufficient to decompress the concrete at tendon level. The strain
in the tendon is equal to that in stage 1 plus a tensile increment of strain equal in
magnitude to Ece.
264
Appendix E
E.3.3 Stage 3
This stage corresponds to the ultimate load condition. The concrete strain at tendon
level, Ept , may be determined in terms of Ecu ( = 0.003) and dn by applying the
requirements of strain compatibility. This concrete strain at tendon level may also be
measured experimentally via attachment of strain gauges on the prestressed tendons
prior to casting and testing.
265
Appendix F
APPENDIX F
STRAIN MEASUREMENT RESULTS
F.l Model PI
Table F.l - Strain measurements of reinforcing steel for Model PI
Applied load
(kPa)
0
4.46
5.93
7.33
8.69
10.18
11.59
13.18
14.89
16.18
3.62
Strain readings ( x IO"6) at Gauge No.
Rl
1
15
34
56
183
518
828
1126
1742
2088
2050
R2
1
20
48
81
248
606
948
1269
2009
2538
X
R3
0
3
8
13
38
187
387
567
877
937
X
R4
-1
2
8
16
71
180
295
392
770
1009
X
R5
X
X
X
X
X
X
X
X
X
X
X
R6
0
14
30
46
90
310
716
1147
1712
2023
2517
R7
1
9
18
25
41
74
159
405
745
976
933
R8
X
X
X
X
X
X
X
X
X
X
X
R9
1
8
16
26
52
94
289
444
827
1167
1204
R10
0
7
15
23
50
104
194
323
606
888
732
Rll
0
3
6
8
11
22
67
376
769
950
768
R12
0
8
15
22
31
58
184
500
1080
1363
1031
"x" refers to disconnected or faulty strain gauges
266
Appendix F
F.2 Model P2
Table F.2a - Strain measurements of reinforcing steel for Model P2
Applied
load
(kPa)
0
1.37
3.28
4.37
5.69
6.96
9.86
11.27
12.48
13.21
14.69
15.86
16.07
17.23
18.44
19.61
Average strain readings ( x 10-6) at Gauge No.
Rl &R2
0
20
36
62 .
91
147
1221
2107
2542
2845
3370
3995
4201
4967
5834
1952
R3&R4
0
67
84
122
173
264
1107
1477
1693
1879
2210
2541
2693
2981
3360
3579
R5&R6
0
17
10
20
35
48
909
1478
1865
2112
2568
2965
3106
3465
3899
2035
R7&R8
0
27
42
61
97
321
1137
1509
1171
1334
1461
1537
1789
1411
1647
-267
R9 & RIO
0
-73
432
393
470
2051
9196
12033
2554
70
522
X
X
X
X
X
Rll &R12
0
62
95
145
198
241
604
820
910
952
912
885
873
860
927
303
"x" refers to disconnected or faulty strain gauges
267
Appendix F
Table F.2b - Strain measurements of prestressed tendons for Model P2
Applied load
(kPa)
0
1.37
3.28
4.37
5.69
6.96
9.86
11.27
12.48
13.21
14.69
15.86
16.07
17.23
18.44
19.61
Strain readings ( x IO"6) at Gauge No.
1
0
6
42
71
127
242
1201
2087
2522
2835
3350
3975
4181
4947
5814
1932
2
0
18
50
78
131
224
916
1448
1781
2029
2482
2900
3052
3449
4318
2404
4
0
17
55
106
197
392
1040
1410
1626
1812
2143
2474
2626
2914
3293
3512
5
0
17
53
106
190
350
746
1165
1496
1746
2131 '
2410
2525
2720
2991
2197
8
0
-3
7
20
39
67
1212
1827
2073
2153
2351
2550
2346
2413
1581
1599
9
0
-7
3
18
31
99
982
1461
1848
2095
2551
2948
3089
3448
3882
2049
10
0
14
36
62
109
190
511
851
1020
1180
1411
1586
1384
945
931
871
11
0
15
34
70
294
472
1110
1482
1144
1307
1434
1510
1762
1384
1620
2558
13
0
33
83
136
179
278
542
758
848
890
950
1123
1411
1798
1865
2018
17
0
505
466
543
2124
2254
9269
12106
2627
-3
595
X
X
X
X
X
"x" refers to disconnected or faulty strain gauges
268
Appendix F
F.3 Model P3
Table F.3a - Strain measurements of prestressed tendons for Model P3 - Day One
Applied load
(kPa)
0 1.40
2.80
4.20
5.60
7.00
8.40
9.80
11.20
12.60
14.00
14.68
15.40
17.61
18.17
19.57
20.27
Strain readings ( x 10-6) at Gauge No.
1 0
94 101 115 134 149 203 265 295 323 389 332 340 258 178 -182
-1008
4 0
51 71 80 64 30 -92 -173
-677
-2598
-3226
-3987
-4550
-5409
-5777
-4904
-3951
6 0
22 27 47 93 191 184 39 -108
-183
-238
-281
-245
-181
-176
6 -37
8 0
85 94 121 130 147 168 254 286 311 357 324 322 194 65 -211
-1178
9 0
27 54 96 116 128 147 199 261 347 159 96 -102
-217
-349
-654
-952
10 0
11 32 74 112 193 612 996 1320
1380
1419
2456
3247
3951
4597
6271
7520
16 0
68 122 175 264 417 623 851 966 1022
1175
1210
2162
3384
4620
6558
8512
18 0
32 81 96 120 238 421 764 848 990 1150
1981
2441
3748
4822
6018
7362
Table F.3b - Strain measurements of prestressed tendons for Model P3 - Day Two
Applied load
(kPa)
0 7.97
14.0
16.77
18.73
20.27
21.52
22.36
Strain readings ( x IO"6) at Gauge No.
1 0
-74 -152
-348
-687
-1157
-2147
-2681
4 0
-120
-254
-514
-649
-1230
-1692
-2157
6 0
-114
-227
-434
-666
-1160
-1584
-2021
8 0
-96 -255
-321
-741
-1548
-2145
-3065
9 0
-117
-254
-406
-612
-1641
-2570
-2963
10 0
121 239 315 653 1488
1912
2422
16 0
98 232 368 814 1678
2597
3119
18 0
132 352 451 962 1547
1963
2751
269
Appendix F
F.4 Model P4
Table F.4a - Strain measurements of prestressed tendons for Model P4 - Day One
Applied load
(kPa)
0 1.37
2.75
4.43
5.71
6.89
8.75
9.72
11.47
12.77
Strain readings ( x IO"6) at Gauge No.
1 0
-67
-29
-37
-155
-41
-59
-61
-149
-139
2 0
179
635
160
162
202
230
334
542
678
4 0
-388
-830
-998
-460
256
-2387
-2630
-2913
-3232
9 0
8
6
12
13
17
51
65
92
157
10 0
-48
8
10
-78
-84
-119
-139
-175
-32
11 0
0
-4
0
2
9
18
32
55
102
14 0
-101
197
297
-2
2
-44
-211
-967
-1160
18 0
-13
695
400
-106
-80
-72
-28
80
32
Table F.4b - Strain measurements of prestressed tendons for Model P4 - Day Two
Applied load
(kPa)
0 0.67
1.49
2.11
2.92
3.38
4.4 5.76
7.06
8.39
9.37
11.03
12.92
14.33
15.64
16.04
Strain readings ( x IO6) at Gauge No.
1 0
13
5
-31
-36
3
-36
-2
19
55
-23
7
-28
22
80
23
2 0
-225
-271
-274
-237
-228
-211
-109
-35
57
256
531
730
499
700
512
4 0
-37
-60
-59
-68
-15
142
237
199
1233
136
210
-380
1185
75
815
9 0
4
7
10
18
19
30
44
57
68
77
91
122
156
194
209
10 0
-44
-73
-74
-70
-80
-84
-98
-106
-255
-145
-118
-73
-393
-588
-519
11 0
4
8
12
18
8
27
41
52
63
72
85
111
153
211
266
14 0
47
37
60
79
114
157
9
28
-404
-146
-167
-42
-1301
-4032
-4772
18 0
287
310
100
27
386
345
266
317
-92
242
139
246
78
-544
-678
270
Appendix F
Table F.4c - Strain measurements of reinforcing steel for Model P4
- Day T w o
Applied load
(kPa)
0
0.67
1.49
2.11
2.92
3.38
4.4
5.76
7.06
8.39
9.37
11.03
12.92
14.33
15.64
16.04
16.64
Strain readings ( x IO"6) at Gauge No.
Rl
0
102
176
266
342
445
554
660
766
896
1031
1114
1251
1366
1438
1437
15747
R2
0
36
46
56
64
83
107
132
160
195
235
265
331
389
434
424
369
R3
0
67
92
122
155
194
253
265
317
375
433
527
651
881
1075
1447
X
R4
0
17
43
74
119
149
242
402
589
780
953
1143
1375
1584
1779
2153
X
R5
0
13
33
59
104
139
200
275
358
455
560
662
781
880
990
1182
X
R6
0
46
59
75
93
118
156
203
253
304
358
403
465
491
519
545
X
R7
0
11
23
39
72
102
159
228
290
325
306
329
376
174
124
124
27
R9
0
35
51
65
81
99
130
194
296
397
505
574
705
946
1117
1328
X
RlOa
0
7
17
23
32
41
92
166
246
325
415
473
578
710
804
918
997
RlOb
0
3
8
14
21
28
43
76
136
196
258
293
372
491
578
699
1129
"x" refers to disconnected strain gauges at failure of model
271
Appendix G
APPENDIX G
SAMPLE CALCULATIONS OF Mj USING MEASURED
TENDON STRAIN
G.l Data for Model P2-B
Model P2-B : at distance d/2 from the front face of edge-column position
p
/;
Es Ep
fsy
fpy b
D
ds dp
dc
=
—
—
=
=
=
—
=
=
=
=
2285 kg/m3
35.7 MPa
200000 MPa
166667 MPa
538 MPa
1500 MPa
400 mm
80 mm
60 mm
44 mm
0
AP
A-st
ASc e
Pi atot
V
PP Zpt.test
=
=
=
=
=
-
=
=
=
39.28 m m 2
314.16 m m 2
0
4mm 46.06 kN (2 tendons)
0
0.2
0.015
0.002197
G.2 Calculations for Mt in Model P2-B
Model P2-B : at edge-column position
(Refer to the flow chart in Fig. 9.2).
272
Appendix G
Effective prestress at the section after allowing for losses (see calculations in Chapter
H):
Pe = 41.89 k N
Eq.(9-1): Mcr = 2.265 k N m
From structural analysis of the model at failure using the simple direct design method,
the applied moment:
M0 = 2.981 k N m
It is a cracked section analysis, since M0 > Mcr.
Eq.(E-3): 7" = 0.80849
Eq.(9-9): 8ce = 0.000049
Eq.(9-10): £pe = 0.006399
From the iterative trial and error procedure, the depth to the neutral axis at ultimate is :
dn = 25 m m
where,
the yield strain for steel is: 8Sy = 0.00269
(Checking Eq. (9-14), indicates that the steel has yielded).
From data: £Pt = 0.002197
Eq. (9-12): £p = 0.008645
The yield strain for tendon is: £py = 0.008999 (i.e. tendons are not at yield).
The compressive forces acting in the section:
273
Appendix G
Eq. (E-4): Cc = 233.17 kN
The corresponding tensile forces:
Tp = 58.92 kN
Ts = 169.02 kN
The error estimate is: EC - ET = 5.23 kN (which is 2.2%).
Therefore,
Eq.(9-17): Mj = 10.790 kNm
274
Appendix H
APPENDIX H
SAMPLE CALCULATIONS FOR PUNCHING SHEAR STRENGTH
USING THE PROPOSED APPROACH (MODIFIED
WOLLONGONG APPROACH)
H.l Sample Calculations for Punching Shear at Corner-column Connection
This section details the typical data obtained (for corner column position Pl-A) from
experimental results in the testing oi Model PI up to failure. It should be noted that this
model was constructed without any spandrel beam nor torsion strip, i.e. it is with a free
edge. This is followed by a sample calculation for punching shear, Vu , at the corner-
column position Pl-A using the Modified Wollongong Approach as introduced in
Chapter 5.
H.l.l Data for Model Pl-A
Model Pl-A : at comer-column position without spandrel beam or torsion strip (i.e.
a free edge).
MC1
MC2 Mj
Mm
bSp
dsp
»1
Ds
=
-
-
-
—
zz
=
=
7.390 kNm
7.260 kNm
7.047 kNm
9.079 kNm
300 mm
60 mm
80 mm
80 mm
^ U ' . V
fly s
f'c Lc
k X
V
= 0
= 512 MPa
= 0
= 35.7 MPa
= 2320 mm
= 410 mm
= 248 mm
= 40 mm
275
Appendix H
de
Ci
c2
PJ
fpy Api
Ap2
nj
n2
" « • / rlnh
=
=
=
=
=
=
=
=
=
60 mm
300 mm
200 mm
23.03 kN
1500 MPa
19.64 mm2
19.64 mm2
1 No.tendon
1 No.tendon
235.62 mm2
Als Jwy
v r u.iest
dp j
dp2
8;
e2 atot
^
PP
=
=
=
zz
=
=
=
=
=
=
157.08 mm2
0
26.93 kN
41 mm
51 mm
0.0253
0.0365
0
0.2
0.015
H.1.2 Calculations for V„ in Model Pl-A
Model Pl-A : at corner-column position
The initial prestress (Pj = 23.03 kN ) was applied at eccentricities of :
ej - I m m (in the direction of M Q J )
ej = 1 1 m m (in the direction of M Q 2 )
Moduli of elasticity of concrete, Ec =
steel, Ep -
30208 M P a
166667 M P a
Prestress losses:
Eq.(5-61): Elastic losses, ^
^2
83.83 N
94.88 N
Eq.(5-62): Residual proportion of prestress after deducting friction losses,
«J>y = 0.99902
0 2 = 0.99940
276
Eq.(5-63): Anchorage Inction losses,
AFL = 460.6 N
Eq.(5-64): Anchoring losses, ALj - 1781.6 N
AL2 = 2263.9 N
Effective pres tresses,
Pel = (Pj.$})-E>}-AF-AL} = 20.68 kN
pe2 = iPj-^2)-^2-AF-AL2 = 20.20 kN
Eq.(5-34):
Effective depth of spandrel, dsp = 56.2 m m
Similarly, effective depth of slab, de= 56.3 m m
PeI.cosQj = 20.67 kN
PeJ.sinQj = 0.523 kN
Pe2.cosB2 = 20.18 kN
Pe2.sinQ2 = ()-737 kN
Spandrel dimensions and strength parameters:
At = xy = 9920 m m 2
Ut = 2 [x + y] = 576 m m
Avt = Aull = 78.54 m m 2
Eq.(5-33):
Eq.(5-20):
Eq.(5-21):
Eq.(5-18):
Eq.(5-19):
Eq.(5-15):
Eq.(5-25):
Mus
CO
co0
a
P 5
Y
3.643 kNr
(.)
446 N/mm
0.40212
5.13792
2.06328
0.28689
Prestress components: Pejx
Pely
Pe2x
Pe2v
277
Appendix H
From Eq.(10.15), the slab restraining factor is calculated as:
y = 2.21981
which is reasonably close to the measured y value shown in Table 10.2 (i.e. Vmeasurej
= 2.22913 ).
The ultimate moment and shear at a distance djl away from the front face of the
column:
Eq.(5-59): For dsp<X2de,
Ml,corner = 6.665 k N m
Eq.(5-42): For dspl de< 1.2,
Vl,corner = 30.54 kN
The -factors used for detenriining Vu are computed as shown below:
Eq.(5-53):
Eq.(5-54):
Eq.(5-48):
Eq.(5-49):
Eq.(5-50):
Eq.(5-51):
Eq.(5-52):
h ~-
h ~-
ki ~-
k2 =
*j =
k4 =
ks =
1.0
30018
-414017
- 394443
12.178
: 0.22086
0
Finally, the punching shear, VH , for Model PI-A is analysed by using the expression
from Eq.(5-47):
278
Appendix H
y - h jjj k4 - k2 k4 + kj k7 kj - k4 k$
which gives:
Vu = 26.35 kN
By comparison with the experimental punching shear, VHJest ( = 26930 N ), it gives a
ratio of the predicted to experimental punching shear:
Vu'Vu.test = 0.978
H.2 Sample Calculations for Punching Shear at Edge-column Connection
Typical data obtained (for edge column position P3-B) from experimental results in the
testing of Model P3 up to failure is illustrated in this section. This model was
constructed with a spandrel beam of 250 m m wide and 120 m m deep. This is followed
by a sample calculation for punching shear, Vu , at this model P3-B using the Modified
Wollongong Approach as shown above in Section 10.3 .
H.2.1 Data for Model P3-B
Model P3-B : at edge-column position with a spandrel beam of 250 mm wide and
120 m m deep.
MC1 = 33.000 k N m
MC2 = 0.280 k N m
Mj = 9.005 k N m
279
Am
fly s
= 28.27 m m 2
= 578 MPa
= 300 m m
Mm
bSp
dsp
Dj
Ds
de
Ci
c2
Pj
fpy
Api
Ap2
nj
n2
"•st sinh
-=
—
-
-
—
-
-
-
-
—
—
=
=
=
—
11.243 kNm
250 mm
94 mm
120 mm
80 mm
60 mm
300 mm
400 mm
23.03 kN
1500 MPa
39.28 mm2
19.64 mm2
2 No.tendons
1 No.tendon
314.16 mm2
/,
Lc
k X
V
*ls
Iwy y ' u.test
dp,
dP2
QJ
e2
°-tot
H
PP
=
=
=
=
=
=
=
—
ur:
=
=
=
=
=
=
47.8 MPa
2320 mm
450 mm
214 mm
84 mm
314.16 mm2
489 MPa
73350 N
41 mm
56 mm
0.0253
0
0
0.2
0.015
H.2.2 Calculations for Vu in Model P3-B
Model P3-B : at edge-column position
The initial prestress (Pj = 23030 N ) was applied at eccentricities of,
e, = 1 m m (in the direction of M^j)
e2 - 16 m m (in the direction of M Q 2 )
Moduli of elasticity of concrete, Ec
steel, Ep
34954 M P a
166667 M P a
Prestress losses:
Eq.(5-61): Elastic losses, t,j
^2
144.90 N
68.89 N
280
Eq.(5-62): Residual proportion ot prestress alter deducting friction losses,
Oy = 0.99902
<t>2 = 0.99208
Eq.(5-63): Anchorage friction losses,
AFL 460.6 N
Eq.(5-64): Anchoring losses, AL,
AL2
1781.6 N
2263.9 N
Effective prestresses:
Pel = iPj.<S>l)-£,!-AF-ALI
Pe2 = (Pj.^2)-t)2-AF-AL2
20.62 kN
20.05 kN
Eq.(5-34):
Effective depth of spandrel,
Similarly, effective depth of slab,
Prestress components: Pejx
Pely
Pe2x
Pely
dsp
4
=
=
=
=
= 87.2
,= 55.3
Pel.cosQj
Pel.&inQj
Pe2.cosQ2
Pe2.s\nQ2
mm
mm
20.61 kN
0.522 kN
20.054 kN
0
Spandrel dimensions and strength parameters:
At =
ut =
Ast ~
Eq.(5-33):
Eq.(5-20):
Eq.(5-21):
Eq.(5-18):
xy
2[x + y]
Alsll
Mus
CO
co0
a
17976 m m 2
596 m m
235.62 m m 2
9.4582 kNm
46.080
446 N/mm
0.90792
281
Appendix. H
Eq.(5-19): {3 = 5.8656
Eq.(5-15): 8 = 8.1492
Eq.(10-18): X = 0.82397
From Eq.(10.13), the slab restraining factor is calculated as:
V = 0.42518
which compares well with the measured \\i value shown in Table 10.2 ( i.e. Vmeasure(j =
0.44078 ).
The ultimate moment and shear at a distance djl away from the front face of the
column:
Eq.(5-44): For dsp> X.2de,
Ml.edge = 9-005 kNm
Eq.(5-46): For dsp/de> 1.2,
Vledge = 51-89 kN
The it-factors used for determining Vlt are computed as shown below:
Eq.(5-53): k6 = 0.95556
Eq.(5-54): k7 = 50846
Eq.(5-56): k{ = 280931
Eq.(5-57): k2 = 12264520
Eq.(5-50): k3 = 21.327
282
Appendix H
Eq.(5-51): . k4 = 0.26655
Eq.(5-52): k5 = 25.0
Finally, the punching shear, Vn , for Model P3-B is analysed by using the expression
from Eq.(5-55):
y _ 2kj kj k4 - k2 k4 + kj k7
kj - k4 k$
which gives:
Vu = 68.83 kN
This gives a ratio of the predicted to experimental punching shear as:
Vu/Vu.test = 0-938
283
Appendix I
APPENDIX I
DETAILS OF SPANDREL BEAMS OR TORSION STRIPS AND
CRITICAL SLAB STRIPS
Table 1.1 - Details of the spandrel beams or torsion strips
Connection
No.
Pl-A P2-A P3-A P4-A
Pl-B P2-B P3-B P4-B
Pl-C P2-C P3-C P4-C
(mm)
300
300
250
300
300
300
250
300
300
300
250
300
(mm)
80
160
120
80
80
160
120
80
80
160
120
80
dspl>
(mm)
60
134
94
60
60
134
94
60
60
134
94
60
Als>
(mm2)
157.08
314.16
314.16
314.16
157.08
314.16
314.16
314.16
157.08
314.16
314.16
314.16
Aws>
(mm2)
0
28.27
28.27
28.27
0
28.27
28.27
28.27
0
28.27
28.27
28.27
(mm)
0
300
300
250
0
300
300
250
0
300
300
250
Ay (MPa)
512
538
578
555
512
538
578
555
512
538
578
555
Jwy )
(MPa)
0
450
489
541
0
450
489
541
0
450
489
541
(mm)
248
248
214
248
248
248
214
248
248
- 248
214
248
y,
(mm)
40
108
84
40
40
108
84
40
40
108
84
40
(mm)
10
10
10
10
10
10
10
10
10
10
10
10
Table 1.2 - Column dimensions of post-tensioned
concrete flat plate models
Column
identification
(for all models)
A B C F G H
Cj,
(mm)
300
300
300
400
400
550
c2,
(mm)
200
400
400
200
400
400
284
Appendix I
-*— ^ —
Table 1.3 - Details of the critical slab strips
Connection
No.
Pl-A P2-A P3-A P4-A
Pl-B P2-B P3-B P4-B
Pl-C P2-C P3-C P4-C
b,
(mm)
200
200
200
200
400
400
450
400
400
400
450
400
(mm)
80
80
80
80
80
80
80
80
80
80
80
80
del'
(mm)
60
60
60
60
60
60
60
60
60
60
60
60
Astl>
(mm2)
235.62
235.62
235.62
235.62
314.16
314.16
314.16
314.16
235.62
235.62
235.62
235.62
Astm <
(mm2)
235.62
235.62
235.62
235.62
235.62
235.62
235.62
235.62
235.62
235.62
235.62
235.62
(mm)
2320
2320
2320
2300
2320
2320
2320
2300
2240
2240
2240
2210
Lj,
(mm)
410
430
450
450
410
430
450
450
410
430
450
450
MCI'
(kNm)
7.39
12.01
12.25
7.34
15.25
28.50
33.00
16.42
8.70
11.17
13.54
9.55
MC2,
(kNm)
7.26
7.35
7.29
6.31
0.24
0.28
0.28
0.13
6.14
9.51
8.226
5.10
Mj,
(kNm)
7.047
7.292
7.131
6.974
10.37
10.79
9.005
10.713
7.529
7.932
7.615
7.773
(kNm)
9.079
9.709
9.325
9.636
11.10
11.731
11.243
11.625
9.749
10.509
9.84
10.029
V u.test '
(kN)
26.93
32.02
30.97
27.64
68.75
76.33
73.35
66.43
27.72
32.97
31.78
28.22
Table 1.4 - Details of prestressing in critical slab strips
Connection
No.
Pl-A P2-A P3-A P4-A
Pl-B P2-B P3-B P4-B
Pl-C P2-C P3-C P4-C
(kN)
23.03
23.03
23.03
23.03
23.03
23.03
23.03
23.03
23.03
23.03
23.03
23.03
fpy '
(MPa)
1500
1500
1500
1500
1500
1500
1500
1500
1500
1500
1500
1500
Apl>
(mm2)
19.64
19.64
19.64
19.64
39.28
39.28
39.28
39.28
19.64
19.64
19.64
19.64
Ap2>
(mm2)
19.64
19.64
19.64
19.64
19.64
19.64
19.64
19.64
19.64
19.64
19.64
19.64
"1 !
(No)
1
1
1
1
2
2
2
2
1
1
1
1
n2,
(No) (mm)
41
41
41
41
41
41
41
41
41
41
41
41
dp2'
(mm)
51
51
51
51
56
56
56
56
51
51
51
51
(rad)
0.0253
0.0253
0.0253
0.0253
0.0253
0.0253
0.0253
0.0253
0.0253
0.0253
0.0253
0.0253
e 2 '
(rad)
0.0365
0.0365
0.0365
0.0365
0
0
0
0
0.0365
0.0365
0.0365
0.0365
atot<
(rad)
0
0
0
0
0
0
0
0
0
0
0
0
M-
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
(rad/m)
0.015
0.015
0.015
0.015
0.015
0.015
0.015
0.015
0.015
0.015
0.015
0.015
285
Appendix J
APPENDIX J
LISTING OF SOURCE CODE FOR BASIC COMPUTER
PROGRAMS TO ANALYSE Vu
J.l BASIC Program for the Wollongong Approach [Loo and Falamaki, 1992]
and the Modified Wollongong Approach
10 REM SAVE"TOO.BAS",A 20 Z=2.30258509299# 100 CLS: 103 LOCATE 1,5:PRINT "DATA" 104 GOSUB 9200 105 LOCATE 3,5:PRINT "MODEL ";MODEL$;"AT "COL$;"COLUMN" 106 GOSUB 9200 110 LOCATE 5,5:PRTNT "Mel = ";MCl:LOCATE 5,30:PRINT "Aws = ";AWS 120 LOCATE 6,5:PRINT "Mc2 = ";MC2:LOCATE 6,30:PPJNT "Fly = ";FLY 130 LOCATE 7,5:PRTNT "Ml = ";M1 :LOCATE 7,30:PRTNT "s = ";S 140 LOCATE 8,5:PRINT "Mm = ";MM :LOCATE 8,30:PRTNT "Fc = ";FC1 150 LOCATE 9,5:PRTNT "Bsp = ";BSP:LOCATE 9,30:PRTNT "Lc = ";LC 160 LOCATE 10,5:PRTNT "Dsp = ";DSP:LOCATE 10,30:PRINT "Ll = ";L1 170 LOCATE 11,5:PRINT "DI = ";D1 :LOCATE 11,30:PRINT "x = ";XS 180 LOCATE 12,5:PRINT "Ds = ";DS :LOCATE 12,30:PRTNT "y = ";YS 190 LOCATE 13,5:PRINT "De = ";DE :LOCATE 13,30:PRINT "Als = ";ALS 200 LOCATE 14,5:PRTNT "CI = ";C1 :LOCATE 14,30:PRINT "Fwy = ";FWY 210 LOCATE 15,5:PRINT "C2 = ";C2 :LOCATE 15,26:PRINT "Vu.test = ";VUT 220 PRINT : GOSUB 9200 230 LOCATE 20,5:PRINT "PRESS [ESC] TO PROCEED " 240 KY$=TNPUT$(1): IF KY$=CHR$(27) THEN 245 ELSE GOTO 10 245 CLS:LOCATE 1,5:PRINT "DATA (CONTINUED)..." 246 GOSUB 9200 250 LOCATE 3,5:PRINT "Pj = ";PJ :LOCATE 3,30:PRINT "dpi = ";DP1 260 LOCATE 4,5:PRINT "fpy = ";FPY :LOCATE 4,30:PRINT "dp2 = ";DP2 270 LOCATE 5,5:PRINT "Apl = ";AP1 :LOCATE 5,30:PRINT "atot = ";ALFAT 280 LOCATE 6,5:PRINT "Ap2 = ";AP2 LOCATE 6,30:PRINT "p = ";MIU 290 LOCATE 7,5:PRTNT "nl = ";N1 : LOCATE 7,30:PRINT "6p = ";BETAP 300 LOCATE 8,5:PRINT "n2 = ";N2 :LOCATE 8,30:PRTNT "91 = ";THETA1 310 LOCATE 9,5:PRINT "Ast,slab = ";ASLAB :LOCATE 9,30:PRINT "92 =
";THETA2 350 PRINT : GOSUB 9200 360 LOCATE 15,5:PRINT "PRESS [ESC] TO PROCEED " 370 KY$=TNPUT$(1): IF KY$=CHR$(27) THEN 380 ELSE GOTO 10
286
Appendix J
380 REM CHECK FOR RC OR PC ANALYSIS 390 IF PJ=0 THEN 400 ELSE GOTO 3000 400 REM REINFORCED CONCRETE SLAB ANALYSIS 405 REM COMPUTE At, Ut 410 AT=XS*YS : UT=2*(XS+YS) 420 REM 430 REM INITIAL EQUATIONS *************************************** 440 AST=ALS/2 450MY=(1-(.6*AST*FLY)/(BSP*DSP*FC1))*DSP*AST*FLY 460 REM TRANSVERSE STRENGTH OF SPANDREL
470 IF S=0 THEN W=0 : GOTO 490 480 W=AWS*FWY/S 490 REM ADDTIONAL TRANSVERSE STRENGTH OF SPANDREL
495 W0=446 500 REM 510 IF COL$="CORNER " GOTO 800 520 REM ANALYSIS AT EDGE COLUMN 530 GAMMA=(75/DS)A3 540 ALFA=ALS*FLY/200000! 550 BETA=UT*(W+W0)/50000! 560 DELTA=ALFA*BETA*DSP/DE : X1=GAMMA*DELTA 570 REM 575 REM LET DR=RATIO OF THE EFFECTIVE DEPTHS OF THE SPANDREL
AND SLAB 576 DR=DSP/DE 580 REM EVALUATE VI 590 IF DSP<=DE THEN M1=(600/(2*BSP+C2))A.9*M1 600 IF DR<=1.2 THEN V1=9.350001+3.49*M1*.001/LC ELSE
V1=2.84+5.62*(M1+MM)*.001/LC 610 REM FORMULA FOR SI 620 SIP=2.65-1.71*L0G(X1)/Z 630 REM COMPUTE CONSTANTS 640K6=1-(D1-DS)/(2*L1) 650 IF ALFA<0 OR BETA<0 OR SIP<=0 THEN 6000 660K1=200000!*(ALFA*BETA/SIP)A.5*(SIP-K6*MC2/MY) 670 K2=K6*MC 1 -(DE+BSP)/2* V1 * 1000! -M1 680 K3=AT/UT 690 K4=(DSP/(2*UT))A.5 700K5=(Cl-BSP)/2 710 REM ANALYSIS OF Vu 720K7=2*K3*K4*K1-K4*K2+K3*V1*1000! : K8=K3-K4*K5 730 VU=K7/K8 740 GOTO 1500 750 REM IfiC) R F M *****************************************
800 REM ANALYSIS AT CORNER COLUMN
287
Appendix J
810 ALFA=ALS *FLY/200000! 820 BETA=UT*(W+W0)/50000! 830 DELTA=ALFA*BETA*DSP/DE 840GAMMA=(C2/C1)A3*(75/DS)A.5 : X2=GAMMA*DELTA 850 REM LET DR=RATIO OF THE EFFECTIVE DEPTHS OF THE SPANDREL
AND SLAB 860 DR=DSP/DE 870 REM EVALUATE VI 880 IF DSP<=DE THEN M1=(400/(BSP+C2))A.25*M1 890 IFDR<=1.2 THEN Vl=-3.66+5.04*(Ml+MM)*.001/LC ELSE Vl=-
.5+4.7*(Ml+MM)*.001/LC 900 REM FORMULA FOR SI 910 IF C2>C1 THEN STP=10.7-4.45 *LOG(X2)/Z ELSE SIP=5.04-3.12*LOG(X2)/Z 920 REM 930 REM COMPUTE CONSTANTS 940 K6=1-(D1-DS)/(2*L1) 950 IF ALFA<0 OR BETA<0 OR SIP<=0 THEN 6000 960K1=200000!*(ALFA*BETA/SIP)A.5*(SIP-K6*MC2/MY) 970K2=K6*MC1-(DE+BSP)/2*V1*1000!-M1 980 K3=AT/UT 990 K4=(DSP/(2*UT))A.5 1000 K5=(Cl-BSP)/2 1010 REM ANALYSIS OF Vu 1020 K7=K3*K4*K1-K4*K2+K3*V1* 1000! : K8=K3-K4*K5 1030 VU=K7/K8 1040 REM 1500 REM OPTION FOR RESULTS 1510 CLS:LOCATE 5,10:PRTNT "RESULTS OPTIONS :" 1520 GOSUB 9100 1530 LOCATE 7,10:PRINT "1) SUMMARISED RESULTS " 1540 LOCATE 9,10:PRINT "2) FULL RESULTS " 1550 GOSUB 9100 1560 LOCATE 12,10:INPUT "SELECT 1-2 ";SEL% 1570 IF SEL%=1 THEN 2000 ELSE IF SEL%=2 THEN 1600 ELSE GOTO 1500 1600 IF PJ=0 THEN CHAIN "LOO.BAS"„ALL ELSE CHATN "JOO.BAS"„ALL 2000 REM SUMMARISED RESULTS 2005 CLS:LOCATE 1,10:PRTNT "RESULTS PRINTOUT" 2010 GOSUB 9100 2015 LOCATE 3,10:PRTNT "MODEL ";MODEL$;" - ";COL$:LOCATE
3,38:PRINT "COLUMN" 2020 LOCATE 4,10:PRTNT "LOO & FALAMAKI'S METHOD TO DETERMINE
Vu" 2025 GOSUB 9100 2030 LOCATE 6,10:PRTNT USING "a = ##.######";ALFA:LOCATE
6,40:PRINT USING "6 =##.######" ;BETA 2040 LOCATE 7,10:PRTNT USING "5 = ###.######";DELTA:LOCATE
7,40:PRINT USING > - ##.######";SIP
288
Appendix J
2050 LOCATE 8,10:PRINT USING "dl/d = ##.######";DR:LOCATE 8,40:PRINT USING "7 = ##.######" ;GAMMA
2060 LOCATE 9,10:PRINT USING "VI = ##.####AAAA";V1*1000:LOCATE 9,40:PRINT USING "Ml = ##.####AAAA";M1
2070 LOCATE 10,10:PRTNT USING "Mu.sp = ##.####AAAA";MY:LOCATE 10,40:PRTNT USING "log[78] = ##.######";LOG(Xl)/Z
2080 IF PJoO THEN LOCATE 11,10:PRINT USING "Pel = ######.##";PE1 -.LOCATE 11,40:PRTNT USING "Pe2 = ######.##" ;PE2
2090 IF PJoO THEN LOCATE 12,10:PRINT USING "Lossesl = ###.##";LOSSESl:LOCATE 12,40:PRINT USING "Losses2 = ###.##" ;LOSSES2
2110 GOSUB 9100 2120 LOCATE 14,10:PRINT USING "Kl = ##.####AAAA";Kl:LOCATE
14,40:PRINT USTNG "K4 = ##.######" ;K4 2130 LOCATE 15,10:PRTNT USING "K2 = ##.####AAAA";K2:LOCATE
15,40:PRINT USTNG "K5 = ##.######";K5 2140 LOCATE 16,10:PRINT USING "K3 = ###.######";K3:LOCATE
16,40:PRINT USING "K6 =##.######";K6 2145 IF PJoO THEN LOCATE 17,40:PRINT USING "K7 = ######.##";K7 2150 GOSUB 9100 2160 LOCATE 19,10:PRINT USING "Vu.test = ######.##";VUT 2170 LOCATE 20,10:PRINT USING "Vu,pred = ######.##";VU 2180 LOCATE 21,10:PRINT USING "Ratio Vu,test/Vu,pred = ##.###" ;VUT/VU 2190 GOSUB 9100 2200 REM 2210 REM 2220 LOCATE 23,10:PRINT "[PRESS ESC TO END]" 2230 KY$=TNPUT$(1): IF KY$=CHR$(27) THEN CLOSE #2:GOTO 6500 ELSE
GOTO 2000 2240 REM 2250 REM 3000 REM PRESTRESSED CONCRETE SLAB ANALYSIS 3010 REM INITIAL EOUATIONS ***************************************
3020 AST=ALS/2 3030DSPEFF=(AST*FLY*DSP+AP2*FPY*DP2)/(AST*FLY+AP2*FPY) 3040 DEEFF=(ASLAB*FLY*DE+AP1*FPY*DP1)/(ASLAB*FLY+AP1 *FPY)
3100 REM ECCENTRICITIES 3110 El=DPl-DS/2 : E2=DP2-DS/2 3200 REM ELASTIC LOSSES 3210EPCONC=2400A(1.5)*.043*SQR(FC1) : EPTENDON= 166667! 3220 SIGCPl=(Nl*PJ/(C2*DS))+(Nl*PJ*El*DS/2)/(C2*DSA3/12) :
SIGCP2=(N2*PJ/(Cl*DS))+(N2*PJ*E2*Dl/2)/(Cl*DlA3/12) 3230 ELASTICl=.5*(EPTENDON/EPCONC)*SIGCPl :
ELASTIC2=.5*(EPTENDON/EPCONC)*SIGCP2 3240 ELOSSl=ELASTICl*(APl): ELOSS2=ELASTIC2*(AP2) 3250 REM FRICTION LOSSES 3260 LPAl=(Cl+DEEFF/2)/1000 3270 IF COL$="CORNER " THEN LPA2=C2/1000 ELSE LPA2=2.65
289
Appendix J
3280 FLOSS I =EXP(-MIU*(ALFAT+BETAP*LPA1)): FLOSS2=EXP(-MIU*(ALFAT+BETAP*LPA2))
3300 REM ANCHORAGE FRICTION LOSSES 3310AFLOSS=.02*PJ 3320 REM ANCHORING LOSSES 3330 TW1=AFLOSS/5700: TW2=AFLOSS/3530 3340ZL1=(3*EPTENDON*19.64)/TW1 : ZJ1=SQR(ZLI): ALOSSl=2*ZJl*TWl 3345 ZL2-=(3*EPTENDON*19.64)/TW2 : ZJ2=SQR(ZL2): ALOSS2=2*ZJ2*TW2 3350 REM EFFECTIVE PRESTRESS AND COMPONENTS AFTER ALL LOSSES 3360PE1=(PJ*FLOSS1)-ELOSS1-AFLOSS-ALOSS1 : PE2=(PJ*FLOSS2)-ELOSS2-
AFLOSS-ALOSS2 3365 REM LOSSES 3370LOSSES1=(PJ-PE1)/PJ*100 : LOSSES2=(PJ-PE2)/PJ*100 3380PE1X=PE1*COS(THETA1) : PE1Y=PE1*SIN(THETA1) 3390 PE2X=PE2*COS(THETA2) : PE2Y=PE2*STN(THETA2) 3395 REM GOSUB 9500 3400 REM COMPUTE At, Ut 3410 AT=XS*YS : UT=2*(XS+YS) 3420 REM 3430 RFM SPANDREL STRENGTH *************************************** 3440 REM 3450 MY = ( 1(.6*(AST*FLY + AP2*FPY)) / ( BSP*DSPEFF*FC1) )*DSPEFF*(
AST*FLY + AP2*FPY) 3460 REM TRANSVERSE STRENGTH OF SPANDREL
**************************
3470 IF S=0 THEN W=0 : GOTO 3490 3480 W=AWS*FWY/S 3490 REM ADDTIONAL TRANSVERSE STRENGTH OF SPANDREL
********************
3495 W0=446 3500 REM COMMON EQUATIONS 3502 EF ALS=0 THEN ALFA= 10000/200000! ELSE ALFA=ALS*FLY/200000! 3504 BETA=UT*(W+W0)/50000! : DR=DSPEFF/DEEFF 3506 DELTA=ALFA*BETA*DR 3510 REM CORNER OR EDGE COLUMN? 3520 IF COL$="CORNER " GOTO 3800 3530 REM ANALYSIS AT EDGE COLUMN
**************************************
3540 GAMMA=(75/DS)A3 3550 X1=GAMMA*DELTA 3570 REM 3580 REM EVALUATE VI 3590 IF DSPEFF<=(1.2*DEEFF) THEN M1=(600/(2*BSP+C2))A.9*M1 3600 IF DR<=1.2 THEN V1=9.350001+3.49*M1*.001/LC ELSE
V1=2.84+5.62*(M1+MM)*.001/LC 3610 REM FORMULA FOR SI 3620 SIP=1.98-1.88*LOG(Xl)/Z 3630 REM COMPUTE CONSTANTS
290
Appendix J
3640 K6=1-(D1-DS)/(2*L1): K7=(V1*1000-N1*PE1Y) 3650 REM IF ALFA<0 OR BETA<0 OR SIP<=0 THEN 6000 3660K1=200000!*(ALFA*BETA/ABS(SIP))A.5*(SIP-K6*MC2/MY) 3670K2=(K6*MCl)-Ml-K7*(Cl-(BSP-DEEFF)/2)+(Nl*PElX*El) 3680 K3=AT/(SQR(2)*UT) 3690 K4=(DSPEFF/(2*UT))A.5 3700 K5=(Cl-BSP)/2 3710 REM ANALYSIS OF Vu 3720K8=2*K3*K4*K1-K4*K2+K3*K7 : K9=K3-K4*K5 3730 VU=K8/K9 3740 GOTO 1500 3750 REM 3800 REM ANALYSIS AT CORNER COLUMN
************************************
3840 GAMMA=(C2/C1)A3*(75/DS)A.5 3850 X1=GAMMA*DELTA 3870 REM EVALUATE VI 3880 IF DSPEFF<=(1.2*DEEFF) THEN M1=(400/(BSP+C2))A.25*M1 3890 IF DR<= 1.2 THEN Vl=-3.66+5.04*(Ml+MM)*.001/LC ELSE Vl=-
.5+4.7*(Ml+MM)*.001/LC 3900 REM FORMULA FOR SI 3910 IF C2>C1 THEN SIP=3.5-1.76*L0G(X1)/Z ELSE SIP=1.86-1.58*L0G(X1)/Z
3920 REM 3930 REM COMPUTE CONSTANTS 3940 K6=T-(D1-DS)/(2*L1): K7=(V1*1000-N1*PE1Y) 3950 REM IF ALFA<0 OR BETA<0 OR SIP<=0 THEN 6000 3960K1=200000!*(ALFA*BETA/ABS(SIP))A.5*(SIP-
K6*(MC2+N2*PE2X*L1)/MY) 3970K2=(K6*MCl)-Ml-K7*(Cl-(BSP-DEEFF)/2)+(Nl*PElX*El)
3980 K3=AT/(SQR(2)*UT) 3990 K4=(DSPEFF/(2*UT))A.5 4000 K5=(Cl-BSP)/2 4010 REM ANALYSIS OF Vu 4020 K8=K3*K4*K1-K4*K2+K3*K7 : K9=K3-K4*K5 4030 VU=K8/K9 4040 REM 4100 GOTO 1500
6000 REM ***.** ***************************************************:i::i:*:!:::::i!*****
6010 REM ERROR MESSAGE 6020 CLS 6030 LOCATE 10,10:PRINT "ERROR DURING COMPUTATION !!!' 6040 LOCATE 12,10:IF ALFA<0 THEN PRINT "_ = ";ALFA;" A NEGATIVE
VALUE'" ELSE IF BETA<0 THEN PRINT "6 = ";BETA;" A NEGATIVE VALUE!" ELSE IF SIP<0 THEN PRINT "_ = ";SIP;" A NEGATIVE VALUE!"
ELSE PRINT "_ = ";SIP;" A ZERO!" 6050 LOCATE 14,10:PRINT "[PRESS ESC TO PROCEED]"
291
Appendix J
6100 KY$=INPUT$(1): IF KY$=CHR$(27) THEN CHAIN "ROO.BAS",230,ALL ELSE GOTO 6000
6500 REM SAVE RESULTS? 6510 CLS:LOCATE 10,10:INPUT "DO YOU WISH TO SAVE THE RESULTS (y/n)
";SAVERE$ 6520 IF SAVERE$="y" OR SAVERE$="Y" THEN RECORDL%=RECORD% :
GOTO 7000 ELSE GOTO 8000 7000 REM TO SET RESULTS FILE1 7010 OPEN "R",#2,"PUNS.DAT" 7020 FIELD #2,1 AS FL$,10 AS ML$,10 AS CL$,4 AS ALFA$,4 AS DELTA$,4 AS
DR$,4 AS Vl$,4 AS BETA$,4 AS SIP$,4 AS GAMMA$,4 AS MY$,4 AS Kl$,4 AS K2$,4 AS K3$,4 AS K4$,4 AS K5$,4 AS K6$,4 AS VUTL$,4 AS VU$
7030 REM 7040 LSET ML$=MODEL$ : LSET CL$=COL$ : LSET ALFA$=MKS$(ALFA):
LSET DELTA$=MKS$(DELTA) 7050 LSET DR$=MKS$(DR) : LSET V1$=MKS$(V1) : LSET
BETA$=MKS$(BETA): LSET SIP$=MKS$(SIP) 7060 LSET GAMMA$=MKS$(GAMMA): LSET MY$=MKS$(MY): LSET
K1$=MKS$(K1) : LSET K2$=MKS$(K2) 7070 LSET K3$=MKS$(K3): LSET K4$=MKS$(K4) : LSET K5$=MKS$(K5):
LSET K6$=MKS$(K6) 7080 LSET VUTL$=MKS$(VUT): LSET VU$=MKS$(VU) 7090 LSET FL$=CHR$(255) 7100 PUT#2,RECORDL% 7110 CLOSE #2 7200 REM TO SET RESULTS FELE2 7210 OPEN "R",#2,"TUNS.DAT" 7220 FIELD #2,1 AS FL$,4 AS K7$,4 AS PE1$,4 AS PE2$,4 AS PE1X$,4 AS
PE1Y$,4 AS PE2X$,4 AS PE2Y$,4 AS PLOSSl$,4 AS PLOSS2$,4 AS DSPEFF$,4 AS DEEFF$
7230 REM 7240 LSET K7$=MKS$(K7): LSET PE1$=MKS$(PE1): LSET PE2$=MKS$(PE2) :
LSET PE1X$=MKS$(PE1X) 7250 LSET PE1Y$=MKS$(PE1Y) : LSET PE2X$=MKS$(PE2X) : LSET
PE2Y$=MKS$(PE2Y) 7260 LSET PLOSSl$=MKS$(PLOSSl): LSET PLOSS2$=MKS$(PLOSS2): LSET
DSPEFF$=MKS$(DSPEFF): LSET DEEFF$=MKS$(DEEFF) 7270 LSET FL$=CHR$(255) 7300 PUT#2,RECORDL% 7310 CLOSE #2 7400 REM TO SET RESULTS FILE3 - FOR PARAMETRIC STUDY 7410 OPEN "R",#2,"PARA.DAT" 7420 FIELD #2,1 AS FL$,4 AS AT$,4 AS UT$,4 AS ALFA$,4 AS BETAS,4 AS
MY$ 4 AS PE2X$,4 AS PE1X$,4 AS PE1Y$,4 AS Vl$,4 AS VUS,4 AS DSPEFF$,4 AS DEEFF$,4 AS El$,4 AS DELTA$,4 AS GAMMA$,4 AS Nl$,4 AS N2$,4 AS SIP$
7430 LSET AT$=MKS$(AT): LSET UT$=MKS$(UT): LSET ALFA$=MKS$(ALFA): LSET BETA$=MKS$(BETA)
292
Appendix J
7440 LSET MY$=MKS$(MY): LSET PE2X$=MKS$(PE2X) : LSET PE1X$=MKS$(PE1X)
7450 LSET PE1Y$=MKS$(PE1Y) : LSET V1$=MKS$(V1) : LSET VU$=MKS$(VU) 7460 LSET DSPEFF$=MKS$(DSPEFF) : LSET DEEFF$=MKS$(DEEFF): LSET
E1$=MKS$(E1) 7470 LSET FL$=CHR$(255): LSET DELTAS =MKS$(DELTA): LSET
GAMMA$=MKS$(GAMMA) 7480 LSET N1$=MKS$(N1) : LSET N2$=MKS$(N2): LSET SIP$=MKS$(SIP) 7500 PUT#2,RECORDL% 7510 CLOSE #2 8000 CHAIN "ROO.BAS",130,ALL 9000 END 9100 REM SUBROUTINE FOR LTNE-DRAWING(LONG) 9110 PRINT STRING$(80,196) 9120 RETURN 9200 REM (SHORT) 9210 PRINT STRTNG$(50,196) 9220 RETURN
J.2 BASIC Program for the Australian Standard Procedure, AS 3600-1988 [SAA,
1988]
10 REM SAVE"TAN.BAS",A 20 REM 100 CLS: 103 LOCATE 1,5:PRTNT "DATA" 104 GOSUB 9200 105 LOCATE 3,5:PRTNT "MODEL ";MODEL$;"AT "COL$;"COLUMN" 106 GOSUB 9200 110 LOCATE 5,5:PRTNT "SHEAR HEAD : ";SHEAR$:LOCATE 5,30:PRINT
"CLOSED TIES: ";TIE$ 120 REM 130 LOCATE 6,5:PRINT "Mel = ";MC1 :LOCATE 6,30:PRINT "s = ";S 140 LOCATE 7,5:PRINT "Bsp = ";BSP LOCATE 7,30:PRINT "Fc = ";FC1 150 LOCATE 8,5:PRTNT "Dsp = ";DSP :LOCATE 8,30:PRINT "x = ";XS 160 LOCATE 9,5:PRTNT "DI = ";D1 -.LOCATE 9,30:PRINT "y = ";YS 170 LOCATE 10,5:PRINT "Ds = ";DS :LOCATE 10,30:PRINT "Fwy = ";FWY 180 LOCATE 11,5:PRINT "De = ";DE :LOCATE 11,30:PRINT "DI = ";DL 190 LOCATE 12,5:PRINT "CI = ";C1 LOCATE 12,30:PRINT "Dt = ";DT 200 LOCATE 13,5:PRINT "C2 = ";C2 :LOCATE 13,26:PRINT "Vu.test = ";VUT 210 LOCATE 14,5:PRINT "Aws = ";AWS 215 GOSUB 9200 220 LOCATE 16,5:PRINT "Pj = ";PJ :LOCATE 16,30:PRTNT "nl = ";N1 230 LOCATE 17,5:PRINT "dpi = ";DP1 :LOCATE 17,30:PRINT "n2 = ";N2 240 LOCATE 18,5:PRINT "dp2 = ";DP2 :LOCATE 18,30:PRINT "_tot = ";ALFA
293
Appendix J
250 LOCATE 19,5:PRTNT "Apl = ";AP1 :LOCATE 19,30:PRENT "u = ";MIU 260 LOCATE 20,5:PRINT "Ap2 = ";AP2 :LOCATE 20,30:PRINT "6p = ";BETA 270 GOSUB 9200 280 LOCATE 22,5:PRINT "PRESS [ESC] TO PROCEED " 290 KY$=TNPUT$(1): IF KY$=CHR$(27) THEN 300 ELSE GOTO 10 299 REM 300 REM PRESTRESSING FACTOR **************************** 310 IF PJ>0 THEN 320 ELSE GOTO 400 320 GOSUB 4000 400 REM ANALYSIS *************************************** 410 REM SETTING LIMITS TO ZERO 420 VUMIN=0 : VUMX=0 : DR=DSP/DE 430 REM TO DETERMINE 8h - Cl.9.2.1 ********************* 440 IF C1>=C2 THEN BETAH=C1/C2 ELSE BETAH=C2/C1 450 REM TO DETERMINE a,yl - Cl.9.2.1 ******************* 460A=Cl+DE/2 470 Y1=YS+DL+DT: AS=AWS/S 480 IF Y1=0 THEN RXY=1 : GOTO 500 ELSE ASMTN=.2*Y1/FWY 490 RXY=AS/ASMIN 500 REM CRITICAL PERIMETER ***************************** 510 IF COL$="CORNER " THEN U=C1+C2+DE : BE=C2+DE/2 ELSE
U=2*C1+C2+2*DE : BE=C2+DE 540 REM SHEAR HEAD"? ************************************ 550 IF SHEAR$="AT COLUMN " THEN 600 560 REM NO SHEAR HEAD ********************************** 570FCV=.17*(l+2/BETAH)*SQR(FCl) : FCVMX=.34*SQR(FC1) 580 IF FCV>FCVMX THEN FCV=FCVMX 590 VUO=U*DE*(FCV+.3*SIGCP): GOTO 700 600 REM WITH SHEAR HEAD ******************************** 610 VU0=U*DE*(.5*SQR(FC1)+.3*SIGCP): VUOMX=.2*U*DE*FCl 620 IF VUO>VUOMX THEN VUO=VUOMX 700 REM NO MOMENT TRANSFER ie. MC1=0 - Cl.9.2.3 ******** 710 IF MC 1=0 THEN VU=VUO : GOTO 2000 800 REM MOMENT TRANSFER ie. MC1>0 - Cl.9.2.4 *********** 810 IF TIE$="NONE " THEN 820 ELSE IF TIE$="IN TORSION STRIP"
THEN 850 ELSE GOTO 900 820 REM NO CLOSED TIES - C1.9.2.4(a) ******************* 830VU=VUO/(1+U*MC1/(8*VUT*A*DE)) 840 GOTO 2000 850 REM TORSION STRIP - C1.9.2.4(b) ******************** 860 VUMTN= 1.2* VUO/( 1 +U*MC 1/(2* VUT* AA2)) 870 R=DS/A 880 GOTO 1000 900 REM SPANDREL BEAM - C1.9.2.4(c) ******************** 910VUMIN=1.2*VUO*(D1/DS)/(1+U*MC1/(2*VUT*A*BSP)) 920 R=D1/BSP 1000 REM CHECK FOR ASMIN - C1.9.2.4(d) ***************** 1010 IF AS<=ASMIN THEN VU=VUMIN : GOTO 1100
294
Appendix J
1020 IF AS>ASMIN THEN VU=VUMIN*SQR(RXY) 1100 REM CHECK FOR Vumax ******************************* 1110IFR>1THENR=1/R 1120 VUMX=3*VUMIN*SQR(R) 1130 IF VU>VUMX THEN VU=VUMX 1140 REM 2000 REM RESULTS 2010 CLS:LOCATE 1,10:PRINT "RESULTS PRINTOUT" 2015 GOSUB 9100 2020 LOCATE 3,10:PRINT "MODEL ";MODEL$;" - ";COL$;"COLUMN " 2025 LOCATE 4,10:PRINT "AUSTRALIAN STANDARD AS3600-1988 TO
DETERMINE Vu" 2030 GOSUB 9100 2040 LOCATE 6,10:PRTNT "FEATURES OF MODEL :" 2060 LOCATE 7,10:PRTNT "SHEAR HEAD - ";SHEAR$ 2070 LOCATE 8,10:PRINT "CLOSED TIES - ";TIE$ 2080 GOSUB 9100 2090 LOCATE 10,10:PRTNT "Sh *= ";BETAH:LOCATE 10,40:PRTNT "a = ";A 2110 LOCATE 11,10:PRTNT "u = ";U:LOCATE 11,40:PRINT "fcv = ";FCV 2120 LOCATE 12,10:PRTNT "Vuo = ";VUO:LOCATE 12,40:PRTNT "Vumin =
";VUMTN 2130 LOCATE 13,10:PRTNT "Vumax = ";VUMX:LOCATE 13,40:PPJNT "dl/d =
";DR 2140 IF PJ>0 THEN LOCATE 14,10:PRINT "_cp = " ;SIGCP:LOCATE
14,40:PRINT "Pj = ";PJ 2150 IF PJ>0 THEN LOCATE 15,10:PRINT "Pe = ";PE/1000:LOCATE
15,40:PRINT "losses (%) = ";LOSSES 2160 GOSUB 9100 2170 LOCATE 17,10:PRINT USTNG "Vu.test = ######.##";VUT 2180 LOCATE 18,10:PRTNT USTNG "Vu,pred =######.##";VU 2190 LOCATE 19,10:PRINT USING "Ratio Vu,test/Vu,pred = ##.###" ;VUT/VU 2200 GOSUB 9100 2210 REM 2220 LOCATE 21,10:PRINT "[PRESS ESC TO END]" 2230 KY$=TNPUT$(1): IF KY$=CHR$(27) THEN CLOSE #2: GOTO 6500 ELSE
GOTO 2000 2240 REM 2250 REM 4000 REM SUBROUTINE FOR PRESTRESSING FACTOR 4010 REM PRESTRESS LOSSES 4020 REM ELASTIC LOSSES 4030EPCONC=2400A(1.5)*.043*SQR(FC1): EPTENDON= 166667! : El=DPl-DS/2
: E2=DP2-DS/2 4040SIGCPl=(Nl*PJ*1000/(C2*DS))+(Nl*PJ*1000*El*DS/2)/(C2*DSA3/12):
SIGCP2=(N2*PJ* 1000/(C1*DS))+(N2*PJ* 1000*E2*DS/2)/(Cl *DSA3/12) 4050 ELASTICl=.5*(EPTENDON/EPCONC)*SIGCPl :
ELASTIC2=.5*(EPTENDON/EPCONC)*SIGCP2
295
Appendix J
4060 EL0SS1=ELASTIC1*( API): ELOSS2=ELASTIC2*(AP2): ELOSS=(ELOSS l+ELOSS2)/2
4100 REM FRICTION LOSSES 4110 LPA=(Cl+DE/2)/1000 4120 FLOSS=EXP(-MIU*( ALFA+BETA*LPA)) 4150 REM ANCHORAGE FRICTION LOSSES 4160 AFLOSS=.02*PJ*1000 4200 REM ANCHORING LOSSES 4210 TW1=AFLOSS/5700 : TW2=AFLOSS/3530 : TW=(TWl+TW2)/2 4220 ZL=(3*EPTENDON*19.64)/TW : Z=SQR(ZL): ALOSS=2*Z*TW 4250 REM EFFECTIVE PRESTRESS FORCE PER WIRE AFTER ALL LOSSES 4260 PE=(PJ* 1000*FLOSS)-ELOSS-AFLOSS-ALOSS 4270 LOSSES=(PJ* 1000-PE)/(PJ* 1000)* 100 4300 REM EFFECTIVE NORTH-SOUTH & EAST-WEST SLAB SECTIONAL
AREAS 4310 AREANS=353O*80 : AREAEW=5700*80 4320 REM EFFECTIVE PRESTRESS 4330 PSIGNS=24*PE/AREANS : PSIGEW=32*PE/AREAEW 4340 REM AVERAGE PRESTRESS 4350 SIGCP=(PSIGNS+PSIGEW)/2 4360 RETURN 6500 REM SAVE RESULTS? 6510 CLS:LOCATE 10,10:TNPUT "DO YOU WISH TO SAVE THE RESULTS (y/n)
";SAVERE$ 6520 IF SAVERE$="y" OR SAVERE$="Y" THEN RECORDL%=RECORD% :
GOTO 7000 ELSE GOTO 8000 7000 REM TO SET RESULTS FILE 7010 OPEN "R",#1,"TANS.DAT" 7020 FIELD #1,1 AS FL$,10 AS ML$,10 AS CL$,10 AS SH$,16 AS TI$,4 AS
BETAH$,4 AS A$,4 AS U$,4 AS FCV$,4 AS VUO$,4 AS VUMTNS,4 AS VUMX$,4 AS DR$,4 AS VUTL$,4 AS VU$,4 AS SIGCP$,4 AS LPA$
7030 REM 7040 LSET ML$=MODEL$ : LSET CL$=COL$ : LSET SH$=SHEAR$ : LSET
TI$=TIE$ 7050 LSET BETAH$=MKS$(BETAH) : LSET A$=MKS$(A) : LSET U$=MKS$(U):
LSET FCV$=MKS$(FCV) 7060 LSET VUO$=MKS$(VUO) : LSET VUMIN$=MKS$(VUMIN) : LSET
VUMX$=MKS$(VUMX) : LSET DR$=MKS$(DR) 7070 LSET VUTL$=MKS$(VUT): LSET VU$=MKS$(VU): LSET
SIGCP$=MKS$(SIGCP): LSET LPA$=MKS$(LPA) 7080 LSET FL$=CHR$(255) 7100 PUT# 1 ,RECORDL% 7110 CLOSE #1 8000 CHAIN "RAN.BAS",130,ALL 9000 END 9100 REM SUBROUTINE FOR LINE-DRAWING(LONG) 9110 PRINT STRING$(80,196) 9120 RETURN
296
Appendix J
9200 REM (SHORT) 9210 PRINT STRING$(50,196) 9220 RETURN
J.3 BASIC Program for the ACI Method, American Concrete Institute ACI318-
89 [ACI, 1989]
10 REM SAVE"TCI.BAS",A 20 REM 100 CLS: 103 LOCATE 1,5:PRINT "DATA" 104 GOSUB 9200 105 LOCATE 3,5:PRINT "MODEL ";MODEL$;"AT "COL$;"COLUMN" 106 GOSUB 9200 110 REM 120 LOCATE 6,5:PRINT "CLOSED TIES: ";TIE$ 130 LOCATE 7,5:PRINT "Mel = ";MC1 :LOCATE 7,30:PRTNT "De = ";DE 135 LOCATE 8,5:PRINT "Mc2 = ";MC2 :LOCATE 8,30:PRTNT "CI = ";C1 140 LOCATE 9,5:PRTNT "Bsp = ";BSP :LOCATE 9,30:PRINT "C2 = ";C2 150 LOCATE 10,5:PRINT "Dsp = ";DSP :LOCATE 10,30:PRTNT "F'c = ";FC1 160 LOCATE 11,5:PRINT "DI = ";D1 :LOCATE 11,26:PRINT "Vu,test = ";VUT 170 LOCATE 12,5:PRTNT "Ds = ";DS 180 GOSUB 9200 190 LOCATE 14,5:PRINT "Pj = ";PJ :LOCATE 14,30:PRINT "nl = ";N1 200 LOCATE 15,5:PRINT "dpi = ";DP1 :LOCATE 15,30:PRINT "ri2 = ";N2 210 LOCATE 16,5:PRFNT "dp2 = ";DP2 :LOCATE 16,30:PRINT "_tot = ";ALFA 220 LOCATE 17,5:PRINT "Apl = ";AP1 :LOCATE 17,30:PRINT "u = ";MIU 230 LOCATE 18,5:PRINT "Ap2 = ";AP2 :LOCATE 18,30:PRINT "Bp = ";BETA 240 GOSUB 9200 250 LOCATE 20,5:PRINT "PRESS [ESC] TO PROCEED " 260 KY$=TNPUT$(1): IF KY$=CHR$(27) THEN 300 ELSE GOTO 10 270 REM 300 REM ANALYSIS *************************************** 310 REM PRESTRESSING FACTOR 320 IF PJ > 0 THEN 350 ELSE GOTO 400 350 GOSUB 4000 355 IF (Cl/DS) < 4 THEN 400 360 REM SHEAR STRENGTH FOR PRESTRESSED SLAB WHERE MC1=0 370 IF PJ>0 THEN VC=(.29*SQR(FC1)+.3*FPC)*B0*DE 380 GOTO 400 390 REM 400 REM PRELIMINARIES FOR ALL COLUMN POSITIONS 410 REM COLUMN OPTIONS 420 IF COL$="CORNER " THEN 450 ELSE IF COL$="EDGE " THEN 600 ELSE
GOTO 700
297
Appendix J
430 REM 450 REM CORNER COLUMN-PRELIMINARY 460BO=C1+C2+DE 470Bl=C2+DE/2 480 B2=Cl+DE/2 490 ALPHAS= 1.66 500 AC=(C1+C2+DE)*DE 510 CAB=(C1+DE/2)A2*DE/(2*AC) 520 CBD=(C2+DE/2)A2*DE/(2*AC) 530 CCD=Cl+DE/2-CAB 540 CAC=C2+DE/2-CBD 550 JC1 = (DE*(C1+DE/2)A3/12) + ((C1+DE/2)*DEA3/12) +
((C2+DE/2)*DE*CABA2) + (C1+DE/2)*DE*((C1 + DE/2)/2-CAB)A2 560 JC2 = (DE*(C2+DE/2)A3/12) + ((C2+DE/2)*DEA3/12) +
((C1+DE/2)*DE*CBDA2) + (C2+DE/2)*DE*((C2 + DE/2)/2-CBD)A2 570Gl=CCD-Cl/2 580 G2=CAC-C2/2 585 IF JC1>JC2 THEN JC=JC1 ELSE JC=JC2 586 IF G1>G2 THEN G=G1 ELSE G=G2 590 GOTO 800 600 REM EDGE COLUMN-PRELIMINARY 610BO=2*C1+C2+2*DE 620B1=C2+DE 630B2=Cl+DE/2 640 ALPHAS=2.5 650 AC=(2*C1+C2+2*DE)*DE 660 CAB=(C1+DE/2)A2*DE/AC 670 CCD=Cl+DE/2-CAB 680 JC = (2*DE*(C1+DE/2)A3/12) + (2*(C1+DE/2)*DEA3/12) +
((C2+DE)*DE*CABA2) + 2*(C1+DE/2)*DE*((C1 + DE/2)/2-CAB)A2
690G=CCD-Cl/2 695 GOTO 800 700 REM INTERIOR COLUMN-PRELIMINARY 710 B0=2*(C1+C2)+4*DE 720B1=C2+DE 730B2=C1+DE 740 ALPHAS=3.33 750 AC=2*DE*(C1+C2+2*DE) 760 G=0 770JC=DE*(Cl+DE)A3/6+(Cl+DE)*DEA3/6+2*DE*(C2+DE)*(Cl*DE)A2/6 780CAB=Cl/2 790CCD=Cl/2 800 REM OPTIONS FOR REINFORCED AND PRESTRESSED SLABS 802 IF PJ > 0 AND (Cl/DS) > 4 THEN GOTO 900 ELSE GOTO 808 808 REM ALL COLUMNS IN REINFORCED SLABS 810 BETAC=C1/C2 : DR=DSP/DE 820 IF BETAC<1 THEN BETAC=1/BETAC 830 REM SHEAR STRENGTH AT NO MOMENT TRANSFER ie.Mcl=0
298
Appendix J
840 VC 1 =(. 166+.33/BETAC)*SQR(FC 1 )*BO*DE 850VC2=(ALPHAS*DE/BO+.166)*SQR(FC1)*BO*DE 860 VC3=.33*SQR(FCl)*BO*DE 870 REM MINIMUM Vc 880 IF VC1>VC2 THEN VCO=VC2 ELSE VC0=VC1 890 IF VCO>VC3 THEN VC=VC3 ELSE VC=VCO 900 REM CHECK ZERO MOMENT TRANSFER 910 IF MC1=0 THEN VU=VC : GOTO 2000 920 REM MOMENT TRANSFER ie.Mcl>0 930 REM COLUMN POSITION 940 IF COL$="CORNER " THEN 950 ELSE GOTO 1100 950 REM CORNER COLUMN-PUNCHING SHEAR CALCULATIONS 960GAMMAVl=l-l/(l+(2/3)*SQR(Bl/B2)) 970 GAMMA V2= 1 -1/( 1 +(2/3)*SQR(B2/B 1)) 975 IF GAMMAV1>GAMMAV2 THEN GAMMAV=GAMMAV1 ELSE
GAMMAV=GAMMAV2 980VUSTRSA=(VUT/AC)+(GAMMAVl*(MCl-VUT*Gl)*CAB/JCl)-(
GAMMAV2*(MC2-VUT*G2)*CAC/JC2) 990 VUSTRSB=(VUT/AC)+(GAMMAV 1 *(MC 1 -VUT*G 1 )*CAB/JC1 )-(G
AMMAV2*(MC2-VUT*G2)*CBD/JC2) 1000 VUSTRSD=(VUT/AC)+(GAMM AV 1 *(MC 1 -VUT*G 1 )*CCD/JC 1 )-
(GAMMAV2*(MC2-VUT*G2)*CBD/JC2) 1010 REM CRITICAL SHEAR STRESS 1020 IF VUSTRSA>VUSTRSB THEN VUSTRSO=VUSTRSA ELSE
VUSTRSO=VUSTRSB 1030 IF VUSTRSO>VUSTRSD THEN VUSTRS=VUSTRSO ELSE
VUSTRS=VUSTRSD 1040 GOTO 1200 1100 REM EDGE & INTERIOR COLUMN-PUNCHING SHEAR CALCULATIONS 1110 GAMMAV=l-l/(l+(2/3)*SQR(Bl/B2)) 1120 VUSTRSAB=(VUT/AC)+(GAMMAV*(MC1-VUT*G)*CAB/JC) 1130 VUSTRSCD=(VUT/AC)-(GAMMAV*(MC1-VUT*G)*CCD/JC) 1140 IF VUSTRSAB>VUSTRSCD THEN VUSTRS=VUSTRSAB ELSE
VUSTRS=VUSTRSCD 1150 REM 1200 REM CHECK MAXIMUM SHEAR STRESS 1210 VUSTRSMX=.85*VC/(BO*DE) 1220 IF VUSTRS>VUSTRSMX THEN VUSTRS=VUSTRSMX 1230 REM PUNCHING SHEAR ' 1240 VU=VUSTRS*BO*DE 1250 REM 2000 REM RESULTS 2005 CLS:LOCATE 1,10:PRINT "RESULTS PRINTOUT" 2006 GOSUB 9100 2010 LOCATE 3,10:PRINT "MODEL ";MODEL$;" - ";COL$;"COLUMN " 2015 LOCATE 4,10:PRINT "AMERICAN CONCRETE INSTITUTE ACI318-89 TO
DETERMINE Vu" 2020 GOSUB 9100
299
Appendix J
2030 LOCATE 6,10:PRINT "FEATURE OF MODEL :" 2060 LOCATE 7,10:PRTNT "CLOSED TIES - ";TIE$ 2070 GOSUB 9100 2080 LOCATE 9,10:PRINT "Be = ";BETAC:LOCATE 9,40:PRINT "Ac = ";AC 2090 LOCATE 10,10:PRINT "bo = ";BO:LOCATE 10,40:PRINT "g = ";G 2100 LOCATE 11,10:PRINT "_v = ";GAMMAV:LOCATE 11,40:PRFNT "Jc =
";JC 2110 LOCATE 12,10:PRTNT "vu = ";VUSTRS:LOCATE 12,40:PRINT "vumax =
";VUSTRSMX 2120 LOCATE 13,10:PRINT "Vc = ";VC:LOCATE 13,40:PRTNT "dl/d = ";DR 2130 IF PJ>0 THEN LOCATE 15,10.PRTNT "fpe = ";FPC:LOCATE 15,40:PRTNT
"Pj = ";PJ 2140 IF PJ>0 THEN LOCATE 16,10:PRINT "Pe = ";PE/1000:LOCATE
16,40:PRINT "losses (%) = ";LOSSES 2145 IF PJ>0 AND (C1/DS)>4 THEN LOCATE 17,10:PRINT "Cl/DS = ";C1/DS;" >
4 : effect of prestress is considered" ELSE IF PJ>0 AND (C1/DS)<4 THEN LOCATE 17,10:PRINT "Cl/DS = ";C1/DS;" < 4 : effect of prestress is ignored"
2150 GOSUB 9100 2160 IF PJ>0 THEN LOCATE 19,10:PRTNT USING "Vu.test
######.##" ;VUT ELSE LOCATE 15,10:PRINT USING "Vu,test ######.##" ;VUT
2170 IF PJ>0 THEN LOCATE 20,10:PRINT USING "Vu,pred ######.##";VU ELSE LOCATE 16,10:PRINT USTNG "Vu,pred ######.##";VU
2180 IF PJ>0 THEN LOCATE 21',10:PRINT USING "Ratio Vu,test/Vu,pred = ##.###" ;VUT/VU ELSE LOCATE 17,10:PRINT USING "Ratio Vu,test/Vu,pred = ##.###" ;VUT/VU
2190 GOSUB 9100 2200 REM 2210 REM 2220 LOCATE 23,10:PRINT "[PRESS ESC TO END]" 2230 KY$=TNPUT$(1): IF KY$=CHR$(27) THEN 6500 ELSE GOTO 2000
2240 REM 2250 REM 4000 REM SUBROUTINE FOR PRESTRESSING FACTOR 4010 REM PRESTRESS LOSSES 4020 REM ELASTIC LOSSES 4030EPCONC=2400A(1.5)*.043*SQR(FC1): EPTENDON= 166667! : El=DPl-DS/2
: E2=DP2-DS/2 4040SIGCPl=(Nl*PJ*1000/(C2*DS))+(Nl*PJ*1000*El*DS/2)/(C2*DSA3/12):
SIGCP2=(N2*PJ* 1000/(C 1 *DS))+(N2*PJ* 1000*E2*DS/2)/(C 1 *DSA3/12) 4050 ELASTIC l=.5*(EPTENDON/EPCONC)*SIGCPl :
ELASTIC2=.5*(EPTENDON/EPCONC)*SIGCP2 4060 ELOSSl=ELASTICl*(APl): ELOSS2=ELASTIC2*(AP2) :
ELOSS=(ELOSS l+ELOSS2)/2 4100 REM FRICTION LOSSES 4110LX=(Cl+DE/2)/1000 4120 FLOSS=EXP(-MIU*(ALFA+BETA*LX))
300
Appendix J
4150 REM ANCHORAGE FRICTION LOSSES 4160 AFLOSS=.02*PJ*1000 4200 REM ANCHORING LOSSES 4210 TW1=AFLOSS/5700 : TW2=AFLOSS/3530 : TW=(TWl+TW2)/2 4220ZL=(3*EPTENDON*19.64yrW : Z=SQR(ZL) : ALOSS=2*Z*TW 4250 REM EFFECTIVE PRESTRESS FORCE PER WIRE AFTER ALL LOSSES 4260 PE=(PJ* 1000*FLOSS)-ELOSS-AFLOSS-ALOSS 4270 LOSSES=(PJ* 1000-PE)/(PJ* 1000)* 100 4300 REM EFFECTIVE NORTH-SOUTH & EAST-WEST SLAB SECTIONAL
AREAS 4310 AREANS=3530*80 : AREAEW=5700*80 4320 REM EFFECTIVE PRESTRESS 4330 PSIGNS=24*PE/AREANS : PSIGEW=32*PE/AREAEW 4340 REM AVERAGE PRESTRESS 4350 SIGCP=(PSIGNS+PSIGEW)/2 4360 IF PSIGNS < 1! THEN PSIGNS=1! 4370 IF PSIGNS > 3.5 THEN PSIGNS=3.5 4380 IF PSIGEW < 1! THEN PSIGEW=1! 4390 IF PSIGEW > 3.5 THEN PSIGEW=3.5 4400 REM AVERAGE PRESTRESS 4410 FPC=(PSIGNS+PSIGEW)/2 4420 IF FC1>35 THEN FC1=35 4450 IF COL$="CORNER " THEN B0=C1+C2+DE : GOTO 4500 4440 B0=2*C1+C2+2*DE 4500 RETURN 6500 REM SAVE RESULTS? 6510 CLS:LOCATE 10,10:INPUT "DO YOU WISH TO SAVE THE RESULTS (y/n)
";SAVERE$ 6520 IF SAVERE$="y" OR SAVERE$='Y" THEN RECORDL%=RECORD% :
GOTO 7000 ELSE GOTO 8000 7000 REM TO SET RESULTS FILE 7010 OPEN "R",#3,"TACI.DAT" 7020 FIELD #3,1 AS FL$,10 AS ML$,10 AS CL$,16 AS TI$,4 AS BETAC$,4 AS
AC$,4 AS BO$,4 AS G$,4 AS GAMMAV$,4 AS JC$,4 AS VUSTRS$,4 AS VUSTRSMX$,4 AS VC$,4 AS DR$,4 AS VUTL$,4 AS VU$,4 AS FPC$,4 AS LX$
7030 REM 7040 LSET ML$=MODEL$ : LSET CL$=COL$ : LSET TI$=TIE$ 7050 LSET BETAC$=MKS$(BETAC): LSET AC$=MKS$(AC): LSET
BO$=MKS$(BO) : LSET G$=MKS$(G) : LSET VC$=MKS$(VC) 7060 LSET GAMMAV$=MKS$(GAMMAV): LSET JC$=MKS$(JC) : LSET
VUSTRS$=MKS$(VUSTRS) : LSET DR$=MKS$(DR): LSET VUSTRSMX$=MKS$(VUSTRSMX)
7070 LSET VUTL$=MKS$(VUT): LSET VU$=MKS$(VU) : LSET FPC$=MKS$(FPC) : LSET LX$=MKS$(LX)
7080 LSET FL$=CHR$(255) 7100 PUT#3,RECORDL% 7110 CLOSE #3
301
Appendix J
8000 CHAIN "ACI.BAS",230,ALL 9000 END 9100 REM SUBROUTINE FOR LINE-DRAWING(LONG) 9110 PRINT STRING$(80,196) 9120 RETURN 9200 REM (SHORT) 9210 PRINT STRING$(50,196) 9220 RETURN
J.4 BASIC Program for the British Method, British Standard Institute BS
8110:1985 [BSI, 1985]
10 REM SAVE"TSI.BAS",A 20 REM 100 CLS: 103 LOCATE 1,5:PRTNT "DATA" 104 GOSUB 9200 105 LOCATE 3,5:PRINT "MODEL ";MODEL$;"AT "COL$;"COLUMN" 106 GOSUB 9200 110 REM 120 LOCATE 6,5:PRINT "CLOSED TIES: ";TIE$ 130 LOCATE 7,5:PRINT "Mel = ";MC1 -.LOCATE 7,30:PRINT "CI = ";C1 135 LOCATE 8,5:PRINT "Bsp = ";BSP :LOCATE 8,30:PRTNT "C2 = ";C2 140 LOCATE 9,5:PRINT "Dsp = ";DSP :LOCATE 9,30:PRINT "Fc = ";FC1 150 LOCATE 10,5:PRINT "DI = ";D1 :LOCATE 10,30:PRTNT "Aslab = ";ASLAB 160 LOCATE 11,5:PRINT "Ds = ";DS :LOCATE 11,28:PRINT "Vu.test = ";VUT 170 LOCATE 12,5:PRINT "De = ";DE 180 GOSUB 9200 190 LOCATE 14,5:PRINT "Pj = ";PJ :LOCATE 14,30:PRINT "nl = ";N1 200 LOCATE 15,5:PRINT "dpi = ";DP1 :LOCATE 15,30:PRINT "n2 = ";N2 210 LOCATE 16,5:PRINT "dp2 = ";DP2 :LOCATE 16,30:PRINT "_tot = ";ALFA 220 LOCATE 17,5:PRINT "Apl = ";AP1 :LOCATE 17,30:PRINT "u = ";MIU 230 LOCATE 18,5:PRTNT "Ap2 = ";AP2 :LOCATE 18,30:PRTNT "fip = ";BETA 240 LOCATE 19,5:PRINT "_ = ";DELTA :LOCATE 19,5:PRINT "fpu = ";FPU 250 GOSUB 9200 260 LOCATE 22,5:PRINT "PRESS [ESC] TO PROCEED " 270 KY$=INPUT$(1): IF KY$=CHR$(27) THEN 300 ELSE GOTO 10 280 REM 300 REM ANALYSIS *************************************** 310 REM PRESTRESSING FACTOR 320 IF PJ>0 THEN 350 ELSE GOTO 400 350 GOSUB 4000 360 REM PUNCHING SHEAR FOR PRESTRESSED SLAB 370 IF PJ>0 THEN GOTO 2000 ELSE GOTO 400 400 REM PUNCHING SHEAR FOR NON-PRESTRESSED SLAB
<*
302
Appendix J
405 REM CONVERT CYLINDER fc TO CUBE STRENGTH feu 406 FCU=3.5969+.99686*FC1 410 REM COLUMN OPTIONS-PRELIMINARY 420 IF COL$="CORNER " THEN 450 ELSE GOTO 500 430 REM 450 REM CORNER COLUMN-PRELIMINARY 460UC=C1+C2+3*DE 470BV=C2+1.5*DE 480 GOTO 550 490 REM 500 REM EDGE COLUMN-PRELIMINARY 510 UC=2*C1+C2+6*DE 520 BV=C2+3*DE 540 REM CONCRETE SHEAR STRESS, vc 550 RHO=100*ASLAB/(BV*DE) 560 KHO=400/DE : RATIO=DSP/DE 570 GAMMAM= 1.25 580 IF RHO>3 THEN RHO=3 590 IF KHO<l THEN KHO=l 600 IF FCU>40 THEN FCU=40 610 IF FCU<25 THEN VC=(-79*RHOA(l/3)*KHOA(l/4))/GAMMAM ELSE
VC=(.79*RHOA(l/3)*KHOA(l/4)*(FCU/25)A(l/3))/GAMMAM 620 REM 630 REM CHECK FOR MAX VC 640 VCMX=.8*SQR(FCU) 650 IF VCMX>5 THEN VCMX=5 660 IF VC>VCMX THEN VC=VCMX 670 REM 680 REM CHECK MOMENT TRANSFER 690 VT=VC*UC*DE 700 REM NO MOMENT TRANSFER, Mcl=0 710 IF MC 1=0 THEN VU=VT : GOTO 2000 ELSE GOTO 720 720 REM MOMENT TRANSFER, Mcl>0 770VEFF=1.25*VT 780 REM 790 REM PUNCHING SHEAR Vu 800 VU=VEFF 810 GOTO 2000 820 REM 2000 REM RESULTS 2005 CLS:LOCATE 1,10:PRINT "RESULTS PRINTOUT" 2006 GOSUB 9100 2010 LOCATE 3,10:PRTNT "MODEL ";MODEL$;" - ";COL$;"COLUMN " 2015 LOCATE 4,10:PRTNT "BRITISH STANDARD INSTITUTE BS8110:1985 TO
DETERMINE Vu" 2020 GOSUB 9100 2030 LOCATE 6,10:PRTNT "FEATURE OF MODEL :" 2060 LOCATE 7,10:PRTNT "CLOSED TIES - ";TIE$
303
Appendix J
2070 GOSUB 9100 2080 LOCATE 10,10:PRTNT "bv = ";BV:LOCATE 10,40:PRTNT "Uc = ";UC 2090 LOCATE 11,10:PRINT "_m = ";GAMMAM:LOCATE 11,40:PRINT "dl/d
= ";RATIO 2100 LOCATE 12,10:PRTNT "vc = ";VC:LOCATE 12,40:PRTNT "vcmax =
";VCMX 2110 LOCATE 13,10:PRINT "Vt = ";VT:LOCATE 13,40:PRTNT "Veff = ";VEFF 2120 LOCATE 14,10:PRINT "Steel ratio = ";RHO:LOCATE 14,40:PRINT "feu =
";FCU 2130 IF PJ>0 THEN LOCATE 15,10:PRTNT "fcp = ";FCP:LOCATE 15,40:PRTNT
"Pj = ";PJ 2140 IF PJ>0 THEN LOCATE 16,10:PRINT "Pe = ";PE/1000:LOCATE
16,40:PRINT "losses (%) = ";LOSSES 2150 GOSUB 9100 2160 LOCATE 18,10:PRINT USING "Vu.test = ######.##" ;VUT 2170 LOCATE 19,10:PRINT USING "Vu,pred = ######.##";VU 2180 LOCATE 20,10:PRTNT USING "Ratio Vu,test/Vu,pred = ##.###";VUT/VU 2190 GOSUB 9100 2200 REM 2210 REM 2220 LOCATE 23,10:PRINT "[PRESS ESC TO END]" 2230 KY$=TNPUT$(1): IF KY$=CHR$(27) THEN 6500 ELSE GOTO 2000 2240 REM 2250 REM 4000 REM SUBROUTINE FOR PRESTRESSING FACTOR 4010 REM PRESTRESS LOSSES 4020 REM ELASTIC LOSSES 4030EPCONC=2400A(1.5)*.043*SQR(FC1): EPTENDON= 166667! : El=DPl-DS/2
: E2=DP2-DS/2 4040SIGCPl=(Nl*PJ*1000/(C2*DS))+(Nl*PJ*1000*El*DS/2)/(C2*DSA3/12):
SIGCP2=(N2*PJ* 1000/(C 1 *DS))+(N2*PJ* 1000*E2*DS/2)/(C 1 *DS A3/12) 4050 ELASTICl=.5*(EPTENDON/EPCONC)*SIGCPl :
ELASTIC2=.5*(EPTENDON/EPCONC)*SIGCP2 4060ELOSS1=ELASTIC1*(AP1) : ELOSS2=ELASTIC2*(AP2) :
ELOSS=(ELOSS l+ELOSS2)/2 4100 REM FRICTION LOSSES 4110LX=(Cl+DE/2)/1000 4120 FLOSS=EXP(-MIU*(ALFA+BETA*LX)) 4150 REM ANCHORAGE FRICTION LOSSES 4160 AFLOSS=.02*PJ*1000 4200 REM ANCHORING LOSSES 4210 TW1=AFLOSS/5700 : TW2=AFLOSS/3530 : TW=(TWl+TW2)/2 4220ZL=(3*EPTENDON*19.64)/TW : Z=SQR(ZL) : ALOSS=2*Z*TW 4250 REM EFFECTIVE PRESTRESS FORCE PER WIRE AFTER ALL LOSSES 4260 PE=(PJ* 1000*FLOSS)-ELOSS-AFLOSS-ALOSS 4270LOSSES=(PJ*1000-PE)/(PJ*1000)*100 4300 REM EFFECTIVE NORTH-SOUTH & EAST-WEST SLAB SECTIONAL
AREAS
304
Appendix J
4310 AREANS=3530*80 : AREAEW=5700*80 4320 REM EFFECTIVE PRESTRESS 4330 PSIGNS=24*PE/AREANS : PSIGEW=32*PE/AREAEW 4340 REM AVERAGE PRESTRESS & OTHER STRESSES 4350 FCP=(PSIGNS+PSIGEW)/2 : FPE=PE/AP1 4360 REM CONVERT CYLINDER fc TO CUBE STRENGTH feu 4370 FCU=3.5969+.99686*FC1 4380 FT=.24*SQR(FCU) 4400 REM SECTION PROPERTIES 4410AG=C2*DS 4420IG=(C2*DSA3)/12 4430 FPT=PE/AG+(PE*El*DS/2)/IG 4450 REM DECOMPRESSION MOMENT 4460 MO=.8*FPT*IG/(DS/2) 4500 REM ANALYSIS OF VU 4510 IF COL$="CORNER " THEN 4550 ELSE GOTO 4600 4550 REM CORNER COLUMN-PRELIMINARY 4560UC=C1+C2+3*DE 4570BV=C2+1.5*DE 4580 GOTO 4700 4600 REM EDGE COLUMN-PRELIMINARY 4610 UC=2*C1+C2+6*DE 4620 BV=C2+3*DE 4700 REM CONCRETE SHEAR STRESS, vc 4710 RHO=100*(ASLAB+AP1)/(BV*DE) 4720 KHO=400/DE : RATIO=DSP/DE 4730GAMMAM=1 4740 IF RHO>3 THEN RHO=3 4750 IF KHO<l THEN KHO=l 4760 IF FCU>40 THEN FCU=40 4770 IF FCU<25 THEN VC=(.79*RHOA(l/3)*KHOA(l/4))/GAMMAM ELSE
VC=(.79*RHOA(l/3)*KHOA(l/4)*(FCU/25)A(l/3))/GAMMAM 4800 REM CHECK FOR MAX VC 4810 VCMX=.8*SQR(FCU) 4820 IF VCMX>5 THEN VCMX=5 4830 IF VC>VCMX THEN VC=VCMX 4900 REM ANALYSIS OF VU 4910VCO=.67*BV*DS*SQR(FTA2+.8*FCP*FT) 4920VCR=(1-.55*FPE/FPU)*VC*BV*DE+MO*(VUT-PE*STN(DELTA))/MC1 4930 IF VCO<VCR THEN VU=VCO ELSE VU=VCR 5000 RETURN 6500 REM SAVE RESULTS? 6510 CLS:LOCATE 10,10:INPUT "DO YOU WISH TO SAVE THE RESULTS (y/n)
";SAVERE$ 6520 IF SAVERE$="y" OR SAVERE$="Y" THEN RECORDL%=RECORD% :
GOTO 7000 ELSE GOTO 8000 7000 REM TO SET RESULTS FILE 7010 OPEN "R",#3,"TSI.DAT"
305
Appendix J
7020 FIELD #3,1 AS FL$,10 AS ML$,10 AS CL$,16 AS TI$,4 AS BV$,4 AS UC$,4 AS GAMMAM$,4 AS RATIO$,4 AS VC$,4 AS VCMX$,4 AS VT$,4 AS VEFF$,4 AS RHO$,4 AS VUTL$,4 AS VU$
7030 REM 7040 LSET ML$=MODEL$ : LSET CL$=COL$ : LSET TI$=TIE$ 7050 LSET BV$=MKS$(BV): LSET UC$=MKS$(UC) : LSET
GAMMAM$=MKS$(GAMMAM) 7060 LSET RATIO$=MKS$(RATIO): LSET VC$=MKS$(VC): LSET
VCMX$=MKS$(VCMX): LSET VT$=MKS$(VT) 7070 LSET VUTL$=MKS$(VUT): LSET VU$=MKS$(VU) : LSET
VEFF$=MKS$(VEFF): LSET RHO$=MKS$(RHO) 7080 LSET FL$=CHR$(255) 7100 PUT#3,RECORDL% 7110 CLOSE #3 8000 CHAIN "BSI.BAS",230,ALL 9000 END 9100 REM SUBROUTINE FOR LTNE-DRAWING(LONG) 9110 PRINT STRING$(80,196) 9120 RETURN 9200 REM (SHORT) 9210 PRINT STRING$(50,196) 9220 RETURN
306
PAPERS PUBLISHED BASED ON THIS THESIS
Chiang, C L . and Loo, Y.C. (1993), "Punching Shear Tests of Half-Scale Post-
Tensioned Concrete Flat Plate Models with Spandrel Beams," Proc. 13th
Australasian Conference on Mechanics of Structures and Materials (ACMSM
13), Wollongong, July 1993, V.l, pp.185-192.
Loo, Y.C. and Chiang CL. (1993), "Methods of Punching Shear Strength Analysis of
Reinforced Concrete Flat Plates - A Comparative Study", International Journal
of Structural Engineering and Mechanics, Techno-Press, V.l, No.l, October
1993, pp.75-86.
Loo, Y.C. and Chiang CL. (1994), "Performance of Analytical Methods For Punching
Shear Strength of Reinforced Concrete Flat Plates", Paper No. 124,
Australasian Structural Engineering Conference-1994 (ASEC), to be held in
Sydney Hilton, Sydney, 21-23 September 1994.
307