http://researchspace.auckland.ac.

nz

ResearchSpace@Auckland

Copyright Statement The digital copy of this thesis is protected by the Copyright Act 1994 (New Zealand). This thesis may be consulted by you, provided you comply with the provisions of the Act and the following conditions of use: x x x Any use you make of these documents or images must be for research or private study purposes only, and you may not make them available to any other person. Authors control the copyright of their thesis. You will recognise the author's right to be identified as the author of this thesis, and due acknowledgement will be made to the author where appropriate. You will obtain the author's permission before publishing any material from their thesis.

To request permissions please use the Feedback form on our webpage. http://researchspace.auckland.ac.nz/feedback

General copyright and disclaimer In addition to the above conditions, authors give their consent for the digital copy of their work to be used subject to the conditions specified on the Library Thesis Consent Form and Deposit Licence.

Note : Masters Theses The digital copy of a masters thesis is as submitted for examination and contains no corrections. The print copy, usually available in the University Library, may contain corrections made by hand, which have been requested by the supervisor.

Characterisation and Seismic Assessment of Unreinforced Masonry Buildings

Alistair Peter Russell

A thesis submitted in partial fullment of the requirements for the degree of Doctor of Philosophy

Supervised by Associate Professor Jason M. Ingham The University of Auckland Department of Civil and Environmental Engineering New Zealand

October 2010

Preface
The research reported in this thesis was undertaken at the Department of Civil and Environmental Engineering, The University of Auckland, between March 2006 and October 2010. The contents of this thesis is the original work of the author, except where specifically acknowledged in the text, and includes nothing that is the outcome of work done in collaboration. No part of the thesis has been submitted for a degree to any other University. The opinions, conclusions and recommendations presented herein are those of the author and do not necessarily reect those of the University of Auckland or any of the sponsoring parties to this project. The thesis is approximately 85,000 words in length, including appendices, tables, references and equations, and contains exactly 140 gures.

Alistair Russell

-i-

Alistair P. Russell

Alistair P. Russell

- ii -

Abstract
This thesis describes the characterisation and seismic assessment of unreinforced masonry (URM) buildings. The research in this thesis was conducted with the primary aim of developing a deeper understanding of the response of URM buildings, with an emphasis on the response of URM walls responding in-plane, and in particular, a comprehensive inspection of the in-plane response of anged URM walls. Most previous expressions for determining the lateral strength capacity of URM walls responding in-plane have not taken into account the eect of anges, and consequently have underestimated the expected capacity. This thesis identied that in order to accurately account for the strength inherently available in URM buildings, the eect of anges should be incorporated into the seismic assessment procedure. The New Zealand URM building stock was analysed to determine typical building characteristics, and URM buildings were classied into seven typologies on the basis of building height and building footprint. Further analysis of the building stock determined that most URM buildings in New Zealand are one and two storeys in height. The characterisation of the New Zealand URM building stock was used to form the basis of the experimental programme reported in this thesis, such that specimens accurately reected construction characteristic of existing New Zealand URM buildings. Testing of URM walls showed that the presence of anges has a signicant eect on the behaviour of walls responding in-plane. Flanges increase the force and displacement capacity of in-plane loaded walls, when compared with in-plane loaded walls without anges. The results from the experimentation that was conducted in order to determine the limiting strength of the shear walls were compared with analytical results from other research, with a high level of correlation. Consequently, equations were recommended for determining the in-plane lateral strength limits of URM walls both with and without anges. Drift limits and energy dissipation characteristics were also proposed. Finally, a procedure was developed for assessing the performance of URM buildings, based on displacement based design principles. The procedure was demonstrated using an example.

- iii -

Alistair P. Russell

Alistair P. Russell

- iv -

Acknowledgements
Though writing is a solitary activity, no research is a solo venture. Firstly, I would like to thank my supervisor and friend Jason Ingham, without whom I would not have embarked on this study. I am grateful for the encouragement, support and many of the initial ideas which he provided, and not least of all for the reliable and rapid critiquing of each part of this thesis. I would like to thank my wife, Myra, without whom I would not have completed this study. The constant encouragement, support, love and patience, particularly during the long hours of writing, were sine qua non in allowing me to complete my thesis. This project would not have been possible without the nancial assistance provided by the Foundation for Research, Science and Technology (grant number UOAX0411) and by the Department of Civil and Environmental Engineering at The University of Auckland. The valuable input from Ken Elwood, John Butterworth, Mike Grith and Guido Magenes, particularly in the initial stages of formulating ideas and the later stages of analysing data, is gratefully acknowledged. The assistance of Hank Mooy, Tony Daligan, Jerey Ang, Mark Byrami and Noel Perinpanayagam in many aspects of lab work and experimentation is also sincerely appreciated. David Hopkins provided valuable feedback on the background and history of URM buildings in New Zealand and associated legislation. Patrick Cummuskey, Win Clark, Russell Green, John Buchan, Neil McLeod, Bruce Mutton, Claire Stevens, Katherine Wheeler and Richard Deakin provided invaluable data on the background, development and number of URM buildings in various jurisdictions around New Zealand. This assistance is gratefully acknowledged. Finally, many thanks to all the graduate students and sta in the retrot group at The University of Auckland, for helping in the lab and for bouncing ideas o, but most importantly for the continuous support and friendship throughout this study.

-v-

Alistair P. Russell

Alistair P. Russell

- vi -

Nomenclature

Roman ai ai af An Af b bf bw c C(T) Ch (T ) COV d dc de do du dT dVmax E ED ES f b fbt fD fdt Fi f j fm f m f j G Ge Gi Distance between inertia centre and compression edge of wall Eective oor acceleration at level i Distance between centre of ange and compression edge of wall Area of net mortared section Cross-sectional area of ange Factor to account for wall aspect ratio Width of ange Width of wall Cohesion Elastic site hazard spectrum Spectral shape factor Coecient of variation Displacement Spectral corner point diaplacement Elastic displacement Maximum displacement for one hysteretic cycle Ultimate wall displacement Design displacement Wall displacement at Vmax Youngs modulus (elastic modulus) Energy dissipated by damping Strain energy Compressive strength of bricks Direct tensile strength of bricks Damping force Diagonal tension strength of masonry Equivalent static horizontal force at level i Compressive strength of mortar Axial compressive stress Compressive strength of masonry Compressive strength of mortar Shear modulus (modulus of rigidity) Eective post-cracked shear modulus (modulus of rigidity) Permanent action at level i
- vii -

mm m/s2 mm mm2 mm2 mm mm MPa % mm mm mm mm mm mm mm MPa kNmm kNmm MPa MPa kN MPa kN MPa MPa MPa MPa MPa MPa kN

Alistair P. Russell

g g glong gshort h he I Ig Ie Ke Ki lw M me n N ND N(T,D) %NBS %NBSb %NBSnom n P PCE Qi R r R s Sa Sd T Tc Te V Vb Vbase Vcrack Vdt Vj Vmax vme Vn Vr Vs Vtc

Acceleration due to gravity Diagonal gauge length Diagonal gauge length perpendicular to applied force Diagonal gauge length parallel to applied force Height of wall Height to resultant of lateral force Moment of inertia Gross moment of inertia Eective moment of inertia Eective stiness of equivalent SDOF system Initial stiness Length of wall Base moment Eective mass of equivalent SDOF system Number of storeys Normal force on cross section Superimposed dead load at top of wall Near fault factor Percentage New Building Standard Baseline percentage New Building Standard Nominal percentage New Building Standard Proportion of gross solid area of unit Applied force Expected gravity compressive force applied to a wall or pier Imposed action at level i Earthquake return period factor Post-yield stiness ratio Reduction factor applied to displacement spectrum for damping eq Sample standard deviation Spectral acceleration Spectral displacement Translational period Corner period Eective period of equivalent SDOF system Applied lateral force Shear strength corresponding to diagonal tension failure involving cracking through bricks Base shear Base shear at rst crack Shear strength corresponding to diagonal tension failure Shear strength corresponding to diagonal tension failure involving damage in mortar joints Maximum base shear Cohesive strength of masonry bed joint Nominal shear strength Shear strength corresponding to onset of rocking Shear strength corresponding to sliding Shear strength corresponding to toe crushing
- viii -

m/s2 mm mm mm mm MPa mm4 mm4 mm4 kN/m kN/m mm kNm kg kN kN % kN kN kN m/s2 mm s s s kN kN kN kN kN kN kN MPa kN kN kN kN

Alistair P. Russell

Vu Vy Wf Wi Wt Ww x z Z Greek c m H long short V c t crack u Vmax eq n 1 2 s E n

Equivalent ultimate base shear Shear strength corresponding to eective yield Weight of ange Seismic weight at level i Total seismic weight Weight of in-plane wall Sample mean Distance from extreme compression bre to line of action of normal force (N) Earthquake hazard factor

kN kN kN kN kN kN mm -

Factor equal to 0.5 for wall xed at base-free at top Eective aspect ratio Factor to account for non-linear vertical stress distribution Shear strain Unit weight of masonry Horizontal extension Diagonal extension Vertical shortening Diagonal shortening Direct strain in compression Direct strain in tension Drift Wall drift at cracking Ultimate wall drift Wall drift at Vmax Coecient of friction Population mean Poissons ratio Equivalent viscous damping ratio Population standard deviation Axial compressive stress perpendicular to the bed joint Maximum principal stress Minimum principal stress Shear stress Earthquake imposed action combination factor Frequency Natural frequency

mm/mm kN/m3 mm mm mm mm mm/mm mm/mm % % % mm MPa MPa MPa MPa rad/s rad/s

- ix -

Alistair P. Russell

Alistair P. Russell

-x-

Contents

Contents
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . i ii

Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iv Nomenclature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v x

Table of Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

List of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvi List of Tables 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxiii 1 2 3 4 5 8

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1 1.2 New Zealand Seismicity . . . . . . . . . . . . . . . . . . . . . . . . . . . URM Building Behaviour . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2.1 1.2.2 1.3 1.4 Typical Structural Deciencies in URM Buildings . . . . . . . . In-Plane Wall Response . . . . . . . . . . . . . . . . . . . . . .

Research Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Thesis Outline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 . . . . . . . . 13

Background on Potentially Earthquake Prone Buildings 2.1 2.2 2.3 2.4

Earthquake Prone Building Legislation . . . . . . . . . . . . . . . . . . 14 URM Heritage Considerations . . . . . . . . . . . . . . . . . . . . . . . 15 New Zealand Building Codes . . . . . . . . . . . . . . . . . . . . . . . . 18 Provisions for the Seismic Upgrade of Existing Buildings . . . . . . . . 20
- xi Alistair P. Russell

Contents

2.5 3

Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 . . . . . . . . . . . . . . . . . . . . . . . . 27

Architectural Characterisation 3.1 3.2

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 Background of Building Typologies . . . . . . . . . . . . . . . . . . . . 29 3.2.1 3.2.2 3.2.3 3.2.4 European URM Typology Studies . . . . . . . . . . . . . . . . 31 North American URM Typology Studies . . . . . . . . . . . . . 37 Iranian URM Typology Studies . . . . . . . . . . . . . . . . . . 39 Previous Research into URM Typologies in New Zealand . . . . 40

3.3

New Zealand URM Building Typologies . . . . . . . . . . . . . . . . . . 40 3.3.1 3.3.2 Parameters for Dierentiating Typologies . . . . . . . . . . . . 44 Details of New Zealand URM Typologies . . . . . . . . . . . . . 46

3.4 3.5

International Comparisons with New Zealand Typologies . . . . . . . . 81 Supplementary Characteristics of New Zealand URM . . . . . . . . . . 82 3.5.1 3.5.2 3.5.3 3.5.4 Bond Pattern . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 Wall Height and Thickness . . . . . . . . . . . . . . . . . . . . 86 Age . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86 Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87

3.6 4

Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91

New Zealand URM Building Stock . . . . . . . . . . . . . . . . . . . . . . 93 4.1 4.2 4.3 4.4 4.5 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 Estimation of URM Population and Distribution . . . . . . . . . . . . . 94 Estimation of URM Population and Value . . . . . . . . . . . . . . . . 100 Estimated Vulnerability of URM Buildings . . . . . . . . . . . . . . . . 105 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 . . . . . . . . . . . . . . . . . . . 113

Diagonal Shear Testing of Wall Panels 5.1

Diagonal Shear Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114 5.1.1 5.1.2 Test Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115 Interpretation of Diagonal Shear Test
- xii -

. . . . . . . . . . . . . . 116

Alistair P. Russell

Contents

5.1.3 5.2 5.3 5.4

Derivation of Drift . . . . . . . . . . . . . . . . . . . . . . . . . 121

Construction Details . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127 5.4.1 5.4.2 Prediction of Shear Strength . . . . . . . . . . . . . . . . . . . 130 Comparison of Shear Strength with ASCE 41-06 . . . . . . . . 131

5.5 6

Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133

In-Plane Cyclic Testing of Rectangular URM Walls . . . . . . . . . . . 135 6.1 6.2 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136 Construction Details . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138 6.2.1 6.2.2 6.2.3 6.3 Wall Specications . . . . . . . . . . . . . . . . . . . . . . . . . 138 Wall Construction . . . . . . . . . . . . . . . . . . . . . . . . . 139 Material Properties . . . . . . . . . . . . . . . . . . . . . . . . . 142

Testing Details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143 6.3.1 6.3.2 6.3.3 Test Setup and Instrumentation . . . . . . . . . . . . . . . . . . 143 Test Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . 146 Predicted Flexural and Shear Strength . . . . . . . . . . . . . . 147

6.4

Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151 6.4.1 Observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151

6.5

Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153 6.5.1 6.5.2 6.5.3 6.5.4 6.5.5 6.5.6 6.5.7 Force-Displacement Response . . . . . . . . . . . . . . . . . . . 154 Energy Dissipation . . . . . . . . . . . . . . . . . . . . . . . . . 155 Bilinear Approximation . . . . . . . . . . . . . . . . . . . . . . 157 Multi-Linear Approximation . . . . . . . . . . . . . . . . . . . . 159 Ultimate Drift . . . . . . . . . . . . . . . . . . . . . . . . . . . 163 Predicted Behaviour and Measured Behaviour . . . . . . . . . . 163 Consolidation of Predictive Equations . . . . . . . . . . . . . . 165

6.6 6.7

Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166 Photos of Wall Response . . . . . . . . . . . . . . . . . . . . . . . . . . 169

- xiii Alistair P. Russell

Contents

In-Plane Cyclic Testing of Flanged URM Walls . . . . . . . . . . . . . . 173 7.1 7.2 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174 Construction Details . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179 7.2.1 7.2.2 7.2.3 7.3 Wall Specications . . . . . . . . . . . . . . . . . . . . . . . . . 179 Wall Construction . . . . . . . . . . . . . . . . . . . . . . . . . 181 Material Properties . . . . . . . . . . . . . . . . . . . . . . . . . 186

Testing Details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186 7.3.1 7.3.2 Test Setup and Instrumentation . . . . . . . . . . . . . . . . . . 186 Test Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . 190

7.4 7.5

Predicted Flexural and Shear Strength . . . . . . . . . . . . . . . . . . 190 Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192 7.5.1 Observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192

7.6

Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195 7.6.1 7.6.2 7.6.3 7.6.4 7.6.5 7.6.6 7.6.7 7.6.8 7.6.9 Force-Displacement Response . . . . . . . . . . . . . . . . . . . 195 Energy Dissipation . . . . . . . . . . . . . . . . . . . . . . . . . 199 Bilinear Approximation . . . . . . . . . . . . . . . . . . . . . . 200 Multi-Linear Approximation . . . . . . . . . . . . . . . . . . . . 201 Ultimate Drift . . . . . . . . . . . . . . . . . . . . . . . . . . . 205 Initial Stiness . . . . . . . . . . . . . . . . . . . . . . . . . . . 208 Crack Pattern Analysis . . . . . . . . . . . . . . . . . . . . . . 211 Predicted Behaviour and Measured Behaviour . . . . . . . . . . 214 Consolidation of Predictive Equations . . . . . . . . . . . . . . 216

7.7 7.8 7.9 8

General Force-Displacement Response of URM Walls With Flanges . . 217 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219 Photos of Wall Response . . . . . . . . . . . . . . . . . . . . . . . . . . 221

URM Building Assessment Procedure . . . . . . . . . . . . . . . . . . . . 227 8.1 8.2 New Zealand Existing Building Seismic Performance Criteria . . . . . . 229 Direct Displacement Based Design . . . . . . . . . . . . . . . . . . . . . 231 8.2.1 Determining Base Shear Demand . . . . . . . . . . . . . . . . . 234
- xiv -

Alistair P. Russell

Contents

8.2.2 8.3

Distribution of Base Shear . . . . . . . . . . . . . . . . . . . . . 235

Assessment Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . 236 8.3.1 8.3.2 8.3.3 Walls Responding Out-of-Plane . . . . . . . . . . . . . . . . . . 243 Walls Responding In-Plane . . . . . . . . . . . . . . . . . . . . 243 Response of Diaphragms and Connections . . . . . . . . . . . . 247

8.4 9

Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248

Summary of Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251 9.1 9.2 9.3 9.4 Architectural Characterisation Chapter 3 . . . . . . . . . . . . . . . . 253 Characterisation of New Zealands URM Building Stock Chapter 4 . . 253 Diagonal Shear Testing of Wall Panels Chapter 5 . . . . . . . . . . . 254 In-Plane Cyclic Testing of URM Walls Chapters 6 and 7 . . . . . . . 255 9.4.1 9.4.2 9.5 9.6 Rectangular Walls . . . . . . . . . . . . . . . . . . . . . . . . . 255 Flanged Walls . . . . . . . . . . . . . . . . . . . . . . . . . . . 256

URM Building Assessment Procedure Chapter 8 . . . . . . . . . . . . 258 Future Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259

10 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261 A Photographic Examples of New Zealand URM Building Typologies . 283 A.1 A.2 A.3 A.4 A.5 A.6 A.7 Typology A Photos . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 284 Typology B Photos . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285 Typology C Photos . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287 Typology D Photos . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 290 Typology E Photos . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 296 Typology F Photos . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297 Typology G Photos . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 300

B Example of Assessment Procedure Typology C . . . . . . . . . . . . . 303 B.1 B.2 Seismic Assessment of Typology C Structure . . . . . . . . . . . . . . . 304 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 316

- xv -

Alistair P. Russell

Contents

Alistair P. Russell

- xvi -

List of Figures

List of Figures
1.1 The plate boundary in New Zealand (reproduced with permission from GNS Science (2007)) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2 In-plane behaviour modes of a laterally loaded URM wall (after Marzahn, 1998; Voon, 2007) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.3 Prior cracking in the wall of a two storey URM isolated building as a potential sliding plane. . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1 Map of seismic zones [from NZSS 1900 Chapter 8:1965 (New Zealand Standards Institute, 1965)] . . . . . . . . . . . . . . . . . . . . . . . . . 19 3.1 Typologies for the Italian URM building stock (reproduced from Binda (2006a)) 3.2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 7 6 3

Typical masonry buildings in Ljubljana, Slovenia (reproduced from (Toma zevi c and Lutman, 2007)) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36

3.3

Typology in Iranian rural residential building stock; comparison of model and prototype (reproduced from (Ghannad et al., 2006)) . . . . . . . . . 39

3.4 3.5 3.6 3.7 3.8 3.9 3.10 3.11 3.12

Locations of buildings surveyed throughout New Zealand . . . . . . . . 42 Building importance levels from AS/NZS 1170.0 Table 3.1 . . . . . . . 44 Overall dimensions of a Typology A1 building . . . . . . . . . . . . . . . 47 Overall dimensions of a Typology A2 building . . . . . . . . . . . . . . . 48 Typology A buildings single storey isolated . . . . . . . . . . . . . . . 50 Overall dimensions of a Typology B building . . . . . . . . . . . . . . . 52 Typology B buildings single storey row . . . . . . . . . . . . . . . . . 53 Timber joist details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 Spikes on exterior wall for connecting internal oor diaphragm . . . . . 56
- xvii Alistair P. Russell

List of Figures

3.13 3.14 3.15 3.16 3.17 3.18 3.19 3.20 3.21 3.22 3.23 3.24 3.25 3.26 3.27 3.28 3.29 3.30 3.31 3.32 3.33 3.34 3.35 3.36 3.37 3.38 3.39 3.40 3.41 4.1

Concrete ring beam between oors . . . . . . . . . . . . . . . . . . . . . 56 Location of Typology C3 building at junction of two roads . . . . . . . . 57 Overall dimensions of a Typology C1 building . . . . . . . . . . . . . . . 58 Overall dimensions of a Typology C2 building . . . . . . . . . . . . . . . 60 Overall dimensions of a Typology C3 building . . . . . . . . . . . . . . . 62 Typology C buildings two storey isolated . . . . . . . . . . . . . . . . 64 Typology C3 buildings two storey corner . . . . . . . . . . . . . . . . 65 One storey building adjoining a two storey building . . . . . . . . . . . 67

Overall dimensions of a Typology D building . . . . . . . . . . . . . . . 67 Typology D building two storey row . . . . . . . . . . . . . . . . . . . 69 Open plan interior of upper storey in Typology E building . . . . . . . . 71 Overall dimensions of a Typology E building . . . . . . . . . . . . . . . 71 Typology E buildings three + storey isolated . . . . . . . . . . . . . . 73 Typology E buildings three + storey isolated . . . . . . . . . . . . . . 74 Overall dimensions of a Typology F building . . . . . . . . . . . . . . . 76 Typology F buildings three + storey row . . . . . . . . . . . . . . . . 78 Typology G building examples religious, institutional and industrial . 80 International comparisons with New Zealand typologies . . . . . . . . . 82 Uneven bricks made to t . . . . . . . . . . . . . . . . . . . . . . . . . . 84 English bond pattern . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84 Common (or American or English Garden Wall) bond pattern . . . . . . 84 Running bond pattern . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 Flemish bond pattern . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 Bond pattern varying over height of a single wall . . . . . . . . . . . . . 85 Date of construction on fa cade . . . . . . . . . . . . . . . . . . . . . . . 87 Typical brick dimensions . . . . . . . . . . . . . . . . . . . . . . . . . . 88 Broken bricks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 Eroded bricks in centre of wall . . . . . . . . . . . . . . . . . . . . . . . 89 Sea shells in mortar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 Proportion of population in the former Auckland Province living in the equivalent current Auckland City . . . . . . . . . . . . . . . . . . . . . . 96

4.2

Provincial estimated populations of URM buildings . . . . . . . . . . . . 98

- xviii -

Alistair P. Russell

List of Figures

4.3 4.4 4.5 4.6 4.7

Construction date of URM buildings in NZ . . . . . . . . . . . . . . . . 99 Number of URM buildings from QV according to construction date . . . 102 Number of URM buildings according to storey height . . . . . . . . . . 104 Valuation of URM building stock according to height . . . . . . . . . . . 104 Number of low rise (1 and 2 storey) buildings as a proportion of all New Zealand URM buildings . . . . . . . . . . . . . . . . . . . . . . . . 105

4.8 4.9

Estimated %NBS of URM buildings in provinces throughout New Zealand110 National estimate of potentially earthquake prone and earthquake risk URM buildings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111

4.10 5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8 5.9 5.10 5.11 5.12 6.1 6.2 6.3 6.4 6.5 6.6 6.7 6.8 6.9

Estimated %NBS of URM buildings in New Zealand . . . . . . . . . . . 112 Orientation of wall panels for ASTM procedure and modied procedure 115 Diagonal shear test setup . . . . . . . . . . . . . . . . . . . . . . . . . . 116 Gauge lengths for measuring wall panel distortion . . . . . . . . . . . . 117 Mohrs Circle interpretation of the diagonal shear test . . . . . . . . . . 120 Derivation of drift angle . . . . . . . . . . . . . . . . . . . . . . . . . . . 121 Shear strength of a wall and diagonal tension strength of masonry . . . 124 Crack patterns at failure of wall panels . . . . . . . . . . . . . . . . . . 125 Photos of cracking through bricks in wall panels . . . . . . . . . . . . . 125 Photos of cracking through mortar in wall panels . . . . . . . . . . . . . 126 Shear failure mode of Wall AP1 . . . . . . . . . . . . . . . . . . . . . . 127 Force-displacement response of wall panels . . . . . . . . . . . . . . . . 128

Maximum principal stress-drift response of wall panels . . . . . . . . . . 129 Side wall of a Typology A structure . . . . . . . . . . . . . . . . . . . . 139 Wall dimensions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140 Wall construction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141 Mortar sand gradation curve . . . . . . . . . . . . . . . . . . . . . . . . 142 Test setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144 Instrumentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145 Imposed cyclic displacement history . . . . . . . . . . . . . . . . . . . . 145 Cracking patterns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152 Force-displacement response . . . . . . . . . . . . . . . . . . . . . . . . 154
- xix Alistair P. Russell

List of Figures

6.10

Energy dissipated ED in one force-displacement loop (from (Chopra, 2007)) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156

6.11 6.12 6.13 6.14

Equivalent viscous damping ratio of Wall A2 and A4 . . . . . . . . . . . 157 Equivalent bilinear approximation (from Magenes and Calvi (1997)) . . 158 Equivalent bi-linear response of Wall A1, A2, A4 . . . . . . . . . . . . . 158 Component forces versus displacement curves (reproduced from ASCE (2007)) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160

6.15 6.16 6.17 6.18 6.19 6.20 6.21 7.1

Multi-linear approximation of Wall A1 . . . . . . . . . . . . . . . . . . . 160 Multi-linear approximation of Wall A2 . . . . . . . . . . . . . . . . . . . 161 Multi-linear approximation of Wall A4 . . . . . . . . . . . . . . . . . . . 162 Approximated response of Wall A4 . . . . . . . . . . . . . . . . . . . . . 162 Photos of Wall A1 response . . . . . . . . . . . . . . . . . . . . . . . . . 169 Photos of Wall A2 response . . . . . . . . . . . . . . . . . . . . . . . . . 170 Photos of Wall A4 response . . . . . . . . . . . . . . . . . . . . . . . . . 171 Shear stress distribution in the in-plane URM wall with and without ange (from Yi et al. (2008)) . . . . . . . . . . . . . . . . . . . . . . . . 177

7.2 7.3 7.4 7.5 7.6 7.7 7.8 7.9 7.10 7.11 7.12 7.13 7.14 7.15 7.16

Wall dimensions (continued) . . . . . . . . . . . . . . . . . . . . . . . . 184 Wall construction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185 Timber diaphragm used in Wall A3 . . . . . . . . . . . . . . . . . . . . 187 Application of axial load on anged walls . . . . . . . . . . . . . . . . . 188 Instrumentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189 Cracking patterns (continued) . . . . . . . . . . . . . . . . . . . . . . . 197 Force-displacement response . . . . . . . . . . . . . . . . . . . . . . . . 198 Equivalent viscous damping ratio of Wall A6, A7 and A8 . . . . . . . . 200 Equivalent bi-linear response of Wall A5, A6, A7, A8 . . . . . . . . . . . 201 Multi-linear approximation of Wall A3a . . . . . . . . . . . . . . . . . . 202 Multi-linear approximation of Wall A5 . . . . . . . . . . . . . . . . . . . 202 Multi-linear approximation of Wall A6 . . . . . . . . . . . . . . . . . . . 203 Multi-linear approximation of Wall A7 . . . . . . . . . . . . . . . . . . . 204 Multi-linear approximation of Wall A8 . . . . . . . . . . . . . . . . . . . 205 Average multi-linear approximations of anged walls . . . . . . . . . . . 206
- xx -

Alistair P. Russell

List of Figures

7.17

Generalised force-deformation relation for masonry elements or components (reproduced from Figure 7-1(a), ASCE (2007)) . . . . . . . . . . . 207

7.18 7.19 7.20 7.21

Flange eects on the orientation of cracking of in-plane wall . . . . . . . 212 Shear stress distribution for a shear force applied parallel to the web . . 212 Direction of principal stresses at ends of in-plane wall . . . . . . . . . . 213 Generalised force-deformation relation for masonry elements or components (reproduced from Figure 7-1, ASCE (2007)) . . . . . . . . . . . . 217

7.22 7.23 7.24 7.25 7.26 7.27 8.1

Photos of Wall A3 response . . . . . . . . . . . . . . . . . . . . . . . . . 221 Photos of Wall A3a response . . . . . . . . . . . . . . . . . . . . . . . . 222 Photos of Wall A5 response . . . . . . . . . . . . . . . . . . . . . . . . . 223 Photos of Wall A6 response . . . . . . . . . . . . . . . . . . . . . . . . . 224 Photos of Wall A7 response . . . . . . . . . . . . . . . . . . . . . . . . . 225 Photos of Wall A8 response . . . . . . . . . . . . . . . . . . . . . . . . . 226 Strength versus risk and ultimate limit state as reference point (reproduced from NZSEE (2006)) . . . . . . . . . . . . . . . . . . . . . . . . . 230

8.2 8.3 8.4 8.5 8.6 A.1 A.2 A.3 A.4 A.5 A.6 A.7 A.8 A.9 A.10 A.11

Displacement response spectra . . . . . . . . . . . . . . . . . . . . . . . 232 Eective stiness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234 URM building assessment procedure owchart . . . . . . . . . . . . . . 237 Tributary areas for axial load on each wall . . . . . . . . . . . . . . . . . 240 Two storey URM analysis . . . . . . . . . . . . . . . . . . . . . . . . . . 242 Typology A buildings single storey isolated . . . . . . . . . . . . . . . 284 Typology B buildings single storey row . . . . . . . . . . . . . . . . . 285 Typology B buildings single storey row . . . . . . . . . . . . . . . . . 286 Typology C buildings two storey isolated . . . . . . . . . . . . . . . . 287 Typology C buildings two storey isolated . . . . . . . . . . . . . . . . 288 Typology C buildings two storey isolated . . . . . . . . . . . . . . . . 289 Typology D buildings two storey isolated . . . . . . . . . . . . . . . . 290 Typology D buildings two storey row . . . . . . . . . . . . . . . . . . 291 Typology D buildings two storey row . . . . . . . . . . . . . . . . . . 292 Typology D buildings two storey row . . . . . . . . . . . . . . . . . . 293 Typology D buildings two storey row . . . . . . . . . . . . . . . . . . 294
- xxi Alistair P. Russell

List of Figures

A.12 A.13 A.14 A.15 A.16 A.17 A.18 B.1

Typology D buildings two storey row . . . . . . . . . . . . . . . . . . 295 Typology E buildings three + storey isolated . . . . . . . . . . . . . . 296 Typology F buildings three + storey row . . . . . . . . . . . . . . . . 297 Typology F buildings three + storey row . . . . . . . . . . . . . . . . 298 Typology F buildings three + storey row . . . . . . . . . . . . . . . . 299 Typology G buildings monumental, religious and institutional . . . . . 300 Typology G buildings monumental, religious and institutional . . . . . 301 Typology C building for seismic assessment . . . . . . . . . . . . . . . . 304

Alistair P. Russell

- xxii -

List of Tables

List of Tables
3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 3.10 4.1 4.2 URM typologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 Typical dimensions of a Typology A1 building . . . . . . . . . . . . . . 48 Typical dimensions of a Typology A2 building . . . . . . . . . . . . . . 49 Typical dimensions of a Typology B building . . . . . . . . . . . . . . . 52 Typical dimensions of a Typology C1 building . . . . . . . . . . . . . . 59 Typical dimensions of a Typology C2 building . . . . . . . . . . . . . . 61 Typical dimensions of a Typology C3 building . . . . . . . . . . . . . . 63 Typical dimensions of a Typology D building . . . . . . . . . . . . . . . 68 Typical dimensions of a Typology E building . . . . . . . . . . . . . . . 72 Typical dimensions of a Typology F building . . . . . . . . . . . . . . . 77 Auckland City pre-1940 potentially earthquake prone buildings . . . . . 95 Population data and URM buildings for Auckland City and Auckland Province . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 4.3 4.4 4.5 4.6 4.7 Provincial populations and estimated number of existing URM buildings 99 Number of URM buildings from QV according to construction decade . 101 URM building stock according to storey height . . . . . . . . . . . . . . 103 Baseline %NBSb for provinces . . . . . . . . . . . . . . . . . . . . . . . 108 Estimated number of potentially earthquake prone and earthquake risk URM buildings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109 5.1 5.2 5.3 Construction details of wall panels . . . . . . . . . . . . . . . . . . . . . 122 Summary of results of wall panels . . . . . . . . . . . . . . . . . . . . . 124 Default lower bound masonry properties (from ASCE (2007)) . . . . . . 132
- xxiii Alistair P. Russell

List of Tables

5.4

Comparison of wall panel shear strength with ASCE (2007) lower bound properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132

6.1 6.2 6.3 6.4 7.1 7.2 7.3 7.4 7.5 7.6 7.7

Wall specications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138 Predicted wall strengths . . . . . . . . . . . . . . . . . . . . . . . . . . . 150 Predicted wall behaviour mode . . . . . . . . . . . . . . . . . . . . . . . 150 Test results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152 Wall specications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180 Flange details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181 Predicted wall strengths from model developed by Yi et al. (2008) . . . 191 Test results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192 Eective initial stiness of anged walls . . . . . . . . . . . . . . . . . . 209 Predicted and measured strengths of anged walls . . . . . . . . . . . . 214 Predicted strengths of Walls A5 A8, neglecting the inuence of anges 215

Alistair P. Russell

- xxiv -

CHAPTER 1

Introduction

All new structural masonry in New Zealand consists of concrete blocks laid in a bed of mortar and reinforced with steel bars, similar to reinforced concrete. This is termed reinforced masonry. In many older masonry buildings, the masonry consists of clay bricks laid on a bed of mortar, but with no reinforcement. This is known as unreinforced masonry, or URM. In many parts of the world URM is still constructed today, but due to its poor performance in earthquakes and New Zealands relatively high seismicity, URM is no longer permitted in this country as a structural material for new buildings.

Unreinforced masonry is known to perform poorly when subjected to lateral forces produced by earthquakes of large magnitude (Drysdale et al., 1999; Megget, 2006; Paulay and Priestley, 1992). This has been shown numerous times in earthquakes such as the 1989 Newcastle earthquake in Australia (Page, 1996), and the 1994 Northridge earthquake in California (Klingner, 2006). In New Zealand, URM was a common building material in the later part of the 19th and early part of the 20th centuries, but since the 1931 Napier earthquake in which nearly 260 people were killed and many URM buildings in the city were destroyed (Dowrick, 1998; Scott, 1999), the use of this construction material declined (see Chapter 4 for details). The use of unreinforced masonry was also restricted by gov-1Alistair P. Russell

Chapter 1. Introduction

ernment legislation after the introduction of the building bylaw NZS 1900 in 1965 (New Zealand Standards Institute, 1965). The current masonry design standard NZS 4230:2004 refers only to reinforced masonry structural elements (Standards New Zealand, 2004b). Nevertheless many historic URM buildings still exist in New Zealand.

1.1

New Zealand Seismicity

New Zealand is seismically active and lies on the boundary of the Australian and Pacic tectonic plates. To the east of the North Island the Pacic plate is forced under the Australian plate. Under the South Island the two plates push past each other sideways, and to the south of New Zealand the Australian plate is forced under the Pacic plate, as shown in Figure 1.1. The seismicity within New Zealand varies depending on the proximity to a fault line associated with this plate boundary. For example, Wellington is considered to have a high level of seismicity, compared with Auckland, which is considered to have low seismicity.

It is estimated that New Zealand has approximately 14,000 earthquakes each year, and most are small, but that between 100 and 150 have a magnitude sucient to be felt (GNS Science, 2007). In the past 150 years, New Zealand has had approximately 15 earthquakes registering over M7.0 magnitude on the Richter scale, with centres less than 30 km deep. By comparison, the Kocaeli earthquake in Turkey in 1999 registered M7.4 and had an epicentre at a depth of 16 km. Casualties numbered nearly 16,000 and it was estimated that over 66,000 buildings collapsed or were heavily damaged (Sezen et al., 2000). Although New Zealands current methods of construction are considered to employ more advanced seismic design than those used in Turkey, New Zealands URM building stock was constructed before the current understanding of seismic design principles. Consequently, those existing buildings which were not designed to withstand the lateral forces generated in an earthquake, of which the URM building stock forms a signicant proportion, are still at risk of signicant damage and/or catastrophic structural failure in a major earthquake.
Alistair P. Russell

-2-

1.2. URM Building Behaviour

Figure 1.1: The plate boundary in New Zealand (reproduced with permission from GNS Science (2007))

1.2

URM Building Behaviour

URM buildings typically consist of foundations, URM walls and piers oriented in orthogonal directions and timber oors, acting as diaphragms, connected to walls by walldiaphragm ties. Other building components may include parapets and appendages such as awnings. URM walls are typically sti structural elements and can be categorised into in-plane and out-of-plane walls depending on the direction of earthquake motion relative to the plane of the walls. Walls oriented parallel to the motion of earthquakes are called in-plane walls, and walls perpendicular to in-plane walls are dened as out-of-plane walls. URM buildings are characterised by a limited number of storeys (typically up to three, and no greater than six in New Zealand). As a generalisation they have regular plan shapes and the external walls form part of the horizontal force resisting system.

-3-

Alistair P. Russell

Chapter 1. Introduction

1.2.1

Typical Structural Deciencies in URM Buildings

As stated above, URM has been shown to perform poorly in earthquakes. There are a number of common details and aspects of URM construction which have been identied as decient. Bruneau (1994b) summarises common failure modes of URM buildings as follows: Lack of anchorage Anchor failure In-plane failures Out-of-plane failures Combined in-plane and out-of-plane eects Diaphragm related failures. Lack of anchorage refers to the absence of positive anchorage between oors or roof elements and walls, in which exterior walls behave as cantilevers over the total building height. This cantilever eect increases the risk of out-of-plane failure as building height increases, and additionally increases the risk of global structural failure due to oor and roof collapse from de-seating. Anchor failure is similar in consequence to lack of anchorage and occurs when any connections between diaphragms and walls were poor or inadequate in the original construction, or have deteriorated over time. Out-of-plane failures occur when inertia forces are of sucient magnitude to cause displacements in an out-of-plane direction such that the wall loses integrity and collapses. This can be sudden and explosive. When adequate connections are present between walls and diaphragms, the out-of-plane excitation occurs at each oor level, and when there is inadequate connection, the walls become tall unrestrained cantilevers. Parapets are vulnerable to this type of failure, and are subject to the greatest amplication of ground motions. Failure of the diaphragm itself is not common and is mostly related to connection and anchor failures.

When a building is to be retrotted, or its seismic performance is to be improved, the rst and most obvious course of action is to restrain any unrestrained parapets and gables.
Alistair P. Russell

-4-

1.2. URM Building Behaviour

The consequences of parapet failure may not be signicant for the buildings structural integrity, but may have life safety consequences from falling to the ground below. Secondly, the connections between the walls and diaphragms must be made robust to prevent oor or roof collapse from de-seating. Magenes and Calvi (1997) note that once out-of-plane failure is prevented by proper measures (e.g. reinforced concrete ring beams or steel ties at the oor levels) the in-plane walls provide the stability necessary to avoid collapse. Bothara et al. (2010) note that URM buildings may be satisfactory in a medium earthquake risk zone if anchorage and out-of-plane failure of the walls can be prevented. For countries with high seismicity such as New Zealand (Wellington, in particular), in-plane walls still have the potential for signicant damage. Consequently, the work reported here focuses on the in-plane response on URM walls.

1.2.2

In-Plane Wall Response

Numerous researchers have reported the dierent failure modes for in-plane URM walls (see Calderini et al., 2009a,b; Cattari and Lagomarsino, 2009; Lee et al., 2008; Magenes and Calvi, 1997; Mann and M uller, 1973, 1982, 1985; Marzahn, 1998; Priestley et al., 2007). Failure modes of masonry piers subjected to seismic actions and gravity actions can be generally classied as either exure dominated or shear dominated, and are related primarily to the aspect ratio of the wall and the axial load ratio. Figure 1.2 shows in-plane behaviour modes of a laterally loaded URM wall.

Flexural response (Figure 1.2(a)) is sometimes dierentiated from rocking response (Figure 1.2(b)), but in general there is little dierence in URM and this response is characterised by the wall behaving as a vertical in-plane cantilever when subjected to lateral forces and displacements. Cracks occur in the masonry tension zone through the opening of bedjoints and shear is carried in the compression zone. Failure occurs by crushing at the wall corner (compression toe) and possibly overturning of the wall. Wall overturning is unlikely in URM buildings as drifts of up to 10% are required, and other failures will occur before this drift limit is reached. The extent of toe crushing depends largely on the ratio of applied axial stress to axial strength (axial load ratio), and when the axial
-5Alistair P. Russell

Chapter 1. Introduction

(a) Flexure

(b) Rocking/Toe crushing

(c) Sliding

(d) Diagonal shear

Figure 1.2: In-plane behaviour modes of a laterally loaded URM wall (after Marzahn, 1998; Voon, 2007)

load is low the behaviour appears more like rocking of a rigid block. On the tension side of the wall cracking may be distributed up the height of the wall, but in the absence of reinforcement to limit crack widths and spread out crack locations, cracking more often occurs at the base of the wall, either between the wall and foundation, or at the opening of a bedjoint several courses above the base. Consequently, although exural response is sometimes dierentiated from rocking and toe crushing, there is signicant overlap, and in
Alistair P. Russell

-6-

1.2. URM Building Behaviour

general dening this mode of behaviour as exural response is satisfactory. Strength and displacement limits are dicult to dene for exural response as toe crushing is the only mode which provides any practical limiting constraints. For wall integrity and gravity load carrying capacity, as well as non-structural damage limitation, a drift limit of 0.8% for exural response of non-retrotted URM walls has been suggested in the literature (Cattari and Lagomarsino, 2009; Magenes and Calvi, 1997; Priestley et al., 2007).

Shear failure is often classied as either sliding shear failure (Figure 1.2(c)) or diagonal tension failure (Figure 1.2(d)). Sliding shear occurs when entire parts of the wall displace horizontally on a sliding plane. A sliding plane can form along cracked bedjoints due to the formation of horizontal cracks subjected to reversed seismic action, or on damp proof courses. This mode of failure is more likely to occur when axial load levels and/or friction coecients are low, or the bond (cohesion) between bricks and mortar is weak. Cracks induced by deection such as foundation settlement (see Figure 1.3) can create a failure plane in much the same way as a damp-proof course. Magenes and Calvi (1997) note that sliding on horizontal bedjoints is a stable mechanism as high displacements are possible without the loss of integrity of the wall. Damage is concentrated in a bedjoint and as long as some axial load is present high energy dissipation is possible. Similar to a exural (rocking) response, a drift or displacement limit for sliding has no real meaning. Diagonal tension failure is often simply termed shear failure and is characterised by a critical

Figure 1.3: Prior cracking in the wall of a two storey URM isolated building as a potential sliding plane.
-7Alistair P. Russell

Chapter 1. Introduction

combination of principal tension and compression stresses as a result of combining shear and compression and leads to the formation and development of inclined diagonal cracks. Depending on the relative strength of mortar joints, the brick-mortar interface, and brick units, the cracks may follow the path of the bed- and headjoints, or may go through the bricks. In some instances, strength limits are given simply in terms of diagonal tension failure (ASCE, 2007), whilst in other instances equations are given which dierentiate between diagonal cracking through joints and diagonal cracking through bricks (Calderini et al., 2009b; Cattari and Lagomarsino, 2009). For walls where a shear failure dominates, a drift limit of 0.4% for non-retrotted URM walls has been suggested in the literature (Cattari and Lagomarsino, 2009; Priestley et al., 2007).

ASCE (2007) classies rocking and bedjoint sliding behaviour modes as deformation controlled actions because of the large displacements available without signicant loss of strength1 , while diagonal tension and toe crushing behaviour modes are classied as force controlled actions because the ultimate failure can be abrupt with little or no subsequent deformation.

Calderini et al. (2009a) note that it is not always easy to distinguish the occurrence of a specic type of mechanism, as many interactions may occur between them.

1.3

Research Motivation

Much research has been previously presented on in-plane wall response, but it has been identied in the literature (see Lee et al., 2008; Moon et al., 2006; Yi et al., 2006a,b, 2008) that codied equations for assessing the strength and displacement capacity of walls are overly conservative, particularly when assessing URM walls with anges (return walls)
Deformation tends to imply a change of shape, which may or may not be considered a brittle failure (see Chapter 5), and consequently where large lateral displacements are available without signicant loss of strength, it may be more correct to classify such behaviour modes as displacement controlled actions. Because this could be confusing when considering displacement based design and assessment of structures (see Chapter 8), and also because it is already dened in ASCE (2007), the denition of deformation controlled actions from ASCE (2007) is used in this thesis.
1

Alistair P. Russell

-8-

1.3. Research Motivation

at either or both ends of the wall. Consequently, one objective of this research was to investigate the response of anged URM walls, in the context of previous research into failure modes, and to determine strength and displacement limits.

Furthermore, in order to quantify the risk which New Zealands URM building stock poses, another objective of this research was to develop a comprehensive characterisation of the national URM building stock, in terms of the building properties (architectural characteristics and typical decient details) and the prevalence, distribution and nancial value of URM buildings. This characterisation was undertaken in order to provide a framework to support both other aspects of the research reported herein, and to assist the parallel research activities of others considering the seismic response and retrot of URM buildings in New Zealand.

Finally, NZSEE (2006) has provided guidelines for assessing the structural performance of buildings in earthquakes in the context of New Zealands legislative environment. Simple assessment procedures to determine if a building is potentially earthquake-prone (IEP, Initial Evaluation Procedure) as well as detailed assessment procedures are provided. It has been communicated by practitioners that whilst the IEP is useful for quickly determining if a building is potentially-earthquake prone, the detailed assessment procedures are complicated and dicult to follow. It has been identied that an assessment procedure needs to ll the gap between the existing IEP and the NZSEE detailed assessment procedures.

Consequently, a signicant emphasis in this thesis is placed on the seismic response of URM walls subjected to in-plane loading, particularly accounting for the eects of anges, and on the development of a detailed seismic assessment procedure which is straightforward to follow and implement, and is appropriate for the nancial and heritage value attached to each unreinforced masonry building.

-9-

Alistair P. Russell

Chapter 1. Introduction

1.4

Thesis Outline

As stated above, the primary aim of this doctoral investigation was to develop a deeper understanding of the seismic response of URM buildings in New Zealand. To achieve this aim, it was necessary to rst characterise the New Zealand URM building stock on both an individual building and nationwide basis. Once the characteristics of URM buildings were known, it was then possible to assess their seismic performance, focussing on the behaviour of walls responding in-plane. As such, this thesis has two main parts. The rst part constitutes the background to URM history and development in New Zealand, investigations into characterising buildings on an individual basis and characterisation of the New Zealand URM building stock on a nationwide basis. This characterisation exercise is covered in Chapters 2 4. The second part of the thesis constitutes Chapters 5 7, where the focus changes from a broad level context to the detailed experimentation and analysis of URM walls responding in-plane. The experimental programme was formed on the basis of results from the rst part of the thesis, such that wall specimens were constructed in a manner that replicated the structural characteristics of existing New Zealand URM buildings. Finally, the two main parts of the thesis are drawn together in Chapter 8, where a framework is presented for assessing the seismic performance of the in-plane loaded walls of a typical New Zealand URM building.

thesis haec divisa in partes novem est. Chapter 1 describes the aim and scope of the present work. Chapter 2 presents a brief background to the legislative environment in New Zealand for assessing and improving the seismic performance of potentially earthquake prone buildings, as well as a brief consideration of principles for assessing and retrotting buildings of an historic nature. Furthermore, this chapter traces the development of New Zealand building codes and the associated provisions for assessing existing earthquake risk buildings, and provides some background to the history of the URM building stock in New Zealand. Chapter 3 outlines general structural congurations (typologies) which apply specically to New Zealand URM buildings. Other characteristics are outlined, in terms of materials and wall geometries, and how these characteristics relate to the overall architectural typologies. Chapter 4 gives an estimate of the number of URM structures in New Zealand, and the nancial value of these buildings, both collectively and individually,
Alistair P. Russell

- 10 -

1.4. Thesis Outline

as well as an estimate of their seismic vulnerability. The results of diagonal shear tests on eight two-leaf unreinforced masonry wall panels are presented in Chapter 5. The aim of the experimentation in Chapter 5 was to investigate the diagonal tension (shear) strength of unreinforced masonry wallettes having dierent mortar properties and bond patterns, and also to provide a baseline against which to compare other retrotted masonry samples and also samples tested in-situ in existing buildings. Chapter 6 presents the in-plane cyclic response of three unreinforced masonry walls designed to replicate typical New Zealand construction in the early 20th Century, with dierent aspect ratios and dierent levels of axial load. Chapter 7 describes the results of experimentation investigating the in-plane response of walls with perpendicular anges or return walls at the wall ends. Flanges of dierent lengths and at dierent locations were investigated. Chapter 8 presents a procedure for assessing the seismic performance of URM buildings, with particular emphasis on the response of in-plane walls. Finally, Chapter 9 contains a summary of the conclusions drawn from this doctoral investigation.

This thesis does not feature a chapter containing a comprehensive literature review because the investigation reported here draws on several dierent elds, and consequently a review of previous research on each topic is presented separately in each chapter.

- 11 -

Alistair P. Russell

Chapter 1. Introduction

Alistair P. Russell

- 12 -

CHAPTER 2

Background on Potentially Earthquake Prone Buildings

This chapter presents a brief background to the legislative environment in New Zealand for assessing and improving the seismic performance of potentially earthquake prone buildings (EPBs), as well as a brief consideration of principles for assessing and retrotting buildings of historic nature. Furthermore, this chapter traces the evolution of New Zealand building codes and the associated provisions for assessing existing earthquake risk buildings, and provides some background to the history of the development of the URM building stock in New Zealand.

Much of the information included in this chapter is a summary from dierent working group reports and regulatory documents in New Zealand, particularly by the New Zealand Society for Earthquake Engineeering (NZSEE) and the Department of Building and Housing (DBH). This information is intended to provide a context for the research conducted in subsequent chapters.

- 13 -

Alistair P. Russell

Chapter 2. Background on Potentially Earthquake Prone Buildings

2.1

Earthquake Prone Building Legislation

New Zealands building stock is under the jurisdiction of the Department of Building and Housing, and legislation for structures and construction is governed by the Building Act 2004 (New Zealand Parliament, 2004). The Building Act is the legislative expression of the governments policy objective for New Zealand buildings, and seeks to reduce the level of earthquake risk to the public over time and targets the most vulnerable buildings (DBH, 2005). The country is divided into districts over which Territorial Authorities (TAs) have authority and responsibility for the consent process for the design of new buildings and the maintenance of existing buildings. The Building Act 2004 required TAs to adopt a policy on earthquake-prone buildings1 by the 31st of May 2006 (New Zealand Parliament, 2004). The rst step in the policy developed by each TA requires a preliminary assessment of their building stock to determine the number and types of buildings with the potential to be earthquake-prone. This involves both reviewing earthquake hazards within the TA district in accordance with NZS 1170.5:2004 (Standards New Zealand, 2004a) and considering the structural performance of the building stock in an earthquake taking account of age, construction, location and use in relation to the earthquake hazard in that district (DBH, 2005). The option is given for TAs to take an active or passive approach in the implementation of their policies, and the appropriateness of each depends on the size of the district under the jurisdiction of that particular Territorial Authority and the number of potentially earthquake-prone buildings in it. An active approach involves the TA carrying out an initial evaluation of buildings in its district to identify those likely to be at high risk, and then establishing priorities for further, more detailed evaluations and guidelines of required performance levels for upgrading. A passive approach is more
A building is earthquake prone for the purposes of the Building Act 2004 if, having regard to its condition and to the ground on which it is built, and because of its construction, the building (a) will have its ultimate capacity exceeded in a moderate earthquake (as dened in the regulations); and (b) would be likely to collapse causing (i) injury or death to persons in the building or to persons on any other property; or (ii) damage to any other property, (quoted from the Building Act 2004 (New Zealand Parliament, 2004)). Furthermore, a moderate earthquake is legally dened as: in relation to a building, an earthquake that would generate shaking at the site of the building that is of the same duration as, but that is one-third as strong as, the earthquake shaking (determined by normal measures of acceleration, velocity and displacement) that would be used to design a new building at the site,(quoted from the Regulations associated with the Building Act 2004 New Zealand Parliament (2005)). This denition of an earthquake-prone building and moderate earthquake is signicantly more extensive and requires a higher level of structural performance for buildings than that provided by the Building Act 1991. It encompasses all buildings, not simply those constructed of unreinforced masonry or unreinforced concrete, although small residential buildings are exempt from these provisions.
1

Alistair P. Russell

- 14 -

2.2. URM Heritage Considerations

ad hoc, where evaluations of buildings are conducted on a case-by-case basis, and usually triggered by an application to the TA under the Building Act for building alteration, change of use, extension of life or subdivision. For larger TAs, an active approach would be more appropriate as it enables the adoption of the best possible risk-reduction programme and sets and controls the level of any work required to mitigate risk. Buildings that the preliminary investigation suggests may be earthquake-prone should be subject to an initial evaluation procedure (IEP). The objective of the IEP is to identify as closely as possible all earthquake-prone buildings within a TAs jurisdiction. At the same time, this initial evaluation should limit the number of buildings that would, on a further detailed evaluation, be found to be not earthquake-prone (DBH, 2005). Where an initial evaluation indicates that a building is likely to be earthquake-prone but the precise EPB status of the building may be in doubt, it is desirable that a detailed assessment of the building is undertaken to determine more precisely whether the building falls within the Building Acts denition of earthquake-prone.

2.2

URM Heritage Considerations

An understanding of the principles involved in maintaining the character and integrity of heritage buildings is of primary importance when changing a URM building during an upgrade or strengthening. Indeed, a consideration of such principles is worth including from the beginning of any building improvement project, and as such, a brief outline of heritage conservation principles is summarised here. A more comprehensive coverage can be found in Goodwin (2008) and McClean (2007, 2009).

Robinson and Bowman (2000) state that good conservation practice balances two complimentary principles: the safe, private and public enjoyment of an historically signicant building; and the continuing practical use of a building as a property asset. If these principles are observed, the result will be the maintenance or enhancement of a buildings heritage signicance, its continuing useful life and its value as an asset for its owner and the community. Goodwin (2009) states that it is important to understand that there
- 15 Alistair P. Russell

Chapter 2. Background on Potentially Earthquake Prone Buildings

is more to a building than simply its physical material. This intangible component is comprised of concepts such as its history and social use, and spiritual signicance of the place, which are essentially what the building means to an observer or group at a point in time. When combined with the physical elements, these are collectively known as its heritage value.

There are some parts of a building which are less important than others. Some parts can be altered or removed, but others are more important to maintain. Buildings have an intrinsic logic, and retrot interventions need to follow the intrinsic logic of the building. As an example, parts of a structure which were originally designed to be visible should still be visible after the intervention. Invasiveness is the key characteristic leading to the success or otherwise of the intervention from a heritage conservation perspective. Robinson and Bowman (2000) state that for any required strengthening or stabilisation, the objective should be to minimise the adverse eects on the building fabric and the spaces around the building. There are four principles which should be considered for this purpose: Knowledge of the important characteristics of a building; The golden rule for changes to heritage buildings is as much as necessary, as little as possible. All work should involve minimum intrusion; Strengthening work should be reversible and aim to achieve structural eectiveness at reasonable cost. This approach will allow for more improved strengthening systems to be incorporated at a later date and for further adaptive reuse or restoration of original forms in the future; Strengthening systems should respect the character and integrity of a heritage building. The Department of Building and Housing also recognises the unique need for heritage buildings to be treated with special consideration. The following is quoted from the Department of Building and Housings guidance for territorial authorities for developing earthquake prone building policies (DBH, 2005), The Building Act requires TAs to state in their EPB policies how they
Alistair P. Russell

- 16 -

2.2. URM Heritage Considerations

intend to manage heritage buildings that are earthquake-prone. The age, layout, structure, type of construction and the cultural and aesthetic sensitivity of heritage buildings are such that the cost of their structural improvement is likely to be very high. These special considerations and constraints mean that TAs will need to engage fully with the owners of heritage buildings and the Historic Places Trust. TA policies should also indicate how the TA would manage the dierent needs of private and public owners of heritage buildings. In determining a suitable standard of performance improvement, TAs will need to take into account the high priority that owners and the Historic Places Trust will place on the protection of a buildings fabric, in addition to meeting its EPB policy requirements concerning the life safety of occupants. Given the importance of heritage buildings to the historical and cultural life of the nation and the local community, TAs may wish to consider special implementation measures in relation to these buildings. These could include setting an extended period in which structural improvements are to be completed or providing incentives to owners to upgrade buildings.

Finally, NZSEE (2006) notes that historical buildings of special cultural signicance should be assigned Importance level 3 unless this classication would result in signicant disruption to historical fabric. In such cases Importance Level 2 may be assigned but with the expectation of greater damage in a large (low probability) earthquake. This approach takes into account the need to balance the preservation of a heritage building for life safety and collapse prevention purposes with the consideration of the architectural impacts of the improvement measures. A minimalist intervention approach will be considered a failure if, in the event of an earthquake, the building collapses and is lost forever. Conversely, a more invasive retrot intervention which changes the buildings historic fabric permanently, while allowing the structure to withstand earthquakes which it may be subjected to, could also be considered a failure as the buildings original and historic nature is lost forever.

- 17 -

Alistair P. Russell

Chapter 2. Background on Potentially Earthquake Prone Buildings

2.3

New Zealand Building Codes

The construction of URM buildings in New Zealand peaked in the decade between 1920 and 1930 and subsequently declined (see Figures 4.3 and 4.4), with one of the most important factors in this decline being the economic conditions of the time. The Great Depression in the 1930s and the outbreak of World War II slowed progress in the construction sector signicantly, and few large buildings of any material were constructed in the period between 1935 and 1955 (Megget, 2006; Stacpoole and Beaven, 1972). Equally important in the history of URM buildings in New Zealand was the 1931 M7.8 Hawkes Bay Earthquake, and the changes in building provisions which it precipitated. The destruction of many URM buildings in Napier graphically illustrated that as a construction material, URM provided insucient strength to resist lateral forces induced in an earthquake due to its brittle nature and inability to dissipate energy. Later in 1931, in response to that earthquake, the Building Regulations Committee presented a report to the Parliament of New Zealand entitled Draft General Building By-Law (Cull, 1931). This was the rst step towards requiring seismic provisions in the design and construction of new buildings. In 1935, this report evolved into NZSS No. 95, published by the newly formed New Zealand Standards Institute, and required a horizontal design acceleration of 0.1g, and this requirement applied to the whole of New Zealand (New Zealand Standards Institute, 1935). NZSS No. 95 also suggested that buildings for public gatherings should have frames constructed of reinforced concrete or steel. The By-Law was not enforceable, but it is understood that it was widely used especially in the larger centres of Auckland, Napier, Wellington, Christchurch and Dunedin (Megget, 2006). The provisions of NZSS No. 95 were conned to new buildings only, but the draft report acknowledged that strengthening of existing buildings should also be considered, and alterations to existing buildings were required to comply with the provisions (Davenport, 2004). In 1939 and 1955 new editions of this By-Law were published, and apart from suggesting in 1955 that the seismic coecient vary linearly from zero at the base to 0.12 at the top of the building (formerly the seismic coecient was uniform up the height of the building), there were few signicant changes (Beattie et al., 2008). It was not until 1965 that much of the recent research at the time into seismic design was incorporated into legislation. The New Zealand Standard Model Building By-Law NZSS 1900 Chapter 8:1965 explicitly prohibAlistair P. Russell

- 18 -

2.3. New Zealand Building Codes

(a) North Island

(b) South Island

Figure 2.1: Map of seismic zones [from NZSS 1900 Chapter 8:1965 (New Zealand Standards Institute, 1965)]

ited the use of URM: (a) in Zone A; (b) of more than one storey or 15 ft (4.6 m) eaves height in Zone B; (c) of more than two storeys or 25 ft (7.6 m) eaves height in Zone C. These zones refer to the seismic zonation at the time, which have subsequently changed and evolved. Zone A consisted of regions of the highest seismic risk and Zone C consisted of regions of the lowest seismic risk (New Zealand Standards Institute, 1965). Details of the seismic zonation in NZSS 1900 are shown in Figure 2.1. Again, the provisions of this By-Law did not apply automatically and had to be adopted by local authorities.

The 1965 code required that buildings be designed and built with adequate ductility, although further details were not given. The next version of the loadings code was published in 1976 as NZS 4203 (Standards Association of New Zealand, 1976), and was a major advance on the 1965 code. Most importantly, the 1976 loadings code was used in conjunction with revised material codes: steel, reinforced concrete, timber and reinforced masonry, which all required specic detailing for ductility. Thus after the publication of this code in 1976, unreinforced masonry was explicitly prohibited as a building material
- 19 Alistair P. Russell

Chapter 2. Background on Potentially Earthquake Prone Buildings

throughout the whole of New Zealand.

The use of URM was implicitly discouraged through legislation from as early as 1935, and although it was still allowed in some forms after 1965, observations of existing building stock show its minimal use from 1935 onwards, especially for larger buildings. This is thought to be signicantly attributable to the exceptionally rigorous quality of design and construction by the Ministry of Works at the time (Johnson, 1963; Megget, 2006).

2.4

Provisions for the Seismic Upgrade of Existing Buildings

As building codes were being developed for the design of new buildings, attention was also given to the performance of existing buildings in earthquakes. The rst time this was addressed in legislation was Amendment 301A to the 1968 Municipal Corporations Act (New Zealand Parliament, 1968). This Act allowed territorial authorities, usually being boroughs, cities or district councils, to categorise themselves as earthquake risk areas and thus to apply to the government to take up powers to classify earthquake prone buildings and require owners to reduce or remove the danger. Buildings (or parts thereof) of high earthquake risk were dened as being those of unreinforced concrete or unreinforced masonry with insucient capacity to resist earthquake forces that were 50% of the magnitude of those forces dened by NZS 1900 Chapter 8:1965. If the building was assessed as being potentially dangerous in an earthquake, the council could then require the owner of the building within the time specied in the notice to remove the danger, either by securing the building to the satisfaction of the council, or if the council so required, by demolishing the building. Most major cities and towns took up the legislation, and as an indication of the eect of this Act, between 1968 and 2003 Wellington City Council achieved strengthening or demolition of 500 out of 700 buildings identied as earthquake prone (Hopkins et al., 2008). Auckland, in spite of having a low seismicity, took a strong interest in the legislation and this led to considerable activity in strengthening buildings (see Boardman (1983)). In Christchurch, a moderately high seismic zone, the City Council
Alistair P. Russell

- 20 -

2.4. Provisions for the Seismic Upgrade of Existing Buildings

implemented the legislation, but adopted a more passive approach, generally waiting for signicant developments to trigger the requirements. In Dunedin, now seen to be of low seismic risk, little was done in response to the 1968 legislation. Strengthening of schools, public buildings and some commercial premises was achieved. As a result, Dunedin has a high percentage of URM buildings compared with many other cities in New Zealand (Hopkins, 2009). Megget (2006) states that much of the strengthening in Wellington was accomplished with extra shear walls, diagonal bracing or buttressing and the tying of structural oors and walls together, and that many brittle hazards such as parapets and clock towers had been removed after the two damaging 1942 South Wairarapa earthquakes (M7 & M7.1) which were felt strongly in Wellington. Hopkins et al. (2008) noted that there was criticism at the loss of many older heritage buildings and at the use of intrusive retrotting measures which were not harmonious with the architectural fabric of the building. At the same time, this did provide an opportunity in many cases for the land on which the old building was situated to be better utilised with new, larger and more eciently designed structures.

A major drawback of the 1968 legislation, which endured until 2004, surviving intact with the passage of the Building Act 1991, was that the denition of an earthquake prone building and the required level to which such buildings should be improved remained tied to the 1965 code. Most territorial authorities called for strengthening to one-half or two-thirds of the 1965 code, and many buildings which were strengthened to these requirements, were subsequently found to fall well short of the requirements of later design standards for new buildings.2 This situation was recognised by the New Zealand Society for Earthquake Engineering (NZSEE), which was also concerned about the performance of more modern buildings, particularly after the observed poor performance of similarly aged buildings in earthquakes in Northridge, California (1994) and Kobe, Japan (1995). NZSEE pushed for new, more up-to-date and wide-ranging legislation. This was supported by the Building Industry Authority, later to become part of the Department of Building and Housing, and a new Building Act came into eect in August 2004 (New Zealand Parliament, 2004). This brought in new changes as to what constituted an Earthquake Prone Building. In
Wellington City Council found that in January 2008, of 97 buildings which had been previously strengthened, 61 were subsequently identied as potentially earthquake prone (Bothara et al., 2008; Stevens and Wheeler, 2008).
2

- 21 -

Alistair P. Russell

Chapter 2. Background on Potentially Earthquake Prone Buildings

particular, the denition of an earthquake prone building was tied to the current design standard of the time, and no longer to the design standard of any particular year. The legislation allowed any territorial authority that is satised that a building is earthquake prone to require the owner to take action to reduce or remove the danger. Each territorial authority was required to have a policy on earthquake prone buildings, and to consult publicly on this policy before its adoption. Policies were required to address the approach and priorities and to state what special provisions would be made for heritage buildings. The 2004 legislation applied to all types of building except small residential ones, (residential buildings were excluded unless they comprised 2 or more storeys and contained 3 or more household units).

As soon as the 1968 legislation came into eect to attempt to mitigate the eects of earthquake prone buildings, the New Zealand National Society for Earthquake Engineering set up a steering committee to provide a code of practice in an eort to assist local authorities to implement the legislation. Since the rst draft code of practice published by the NZNSEE (1972), several successive publications have been produced, each extending on the previous version. These documents have been instrumental in helping engineers and territorial authorities to assess the expected seismic performance of existing buildings consistent with the requirements of the legislation. Guidelines for assessing and upgrading earthquake risk buildings were published as a bulletin article in 1972 (NZNSEE) and then separately published the following year, which became colloquially known as the Brown Book (NZNSEE, 1973). The Brown Book provided guidelines for surveying earthquake risk buildings and for the identication of particularly hazardous buildings and features, and was found to be very helpful in many respects. The Brown Book did not establish or recommend strength levels to which earthquake prone buildings should be upgraded, and thus standards varied from one area to another. It was implicit that strengthening be to more than half the standard required in Chapter 8 of the 1965 NZSS Model Building By-Law.

In 1982, the NZSEE established a study group to examine and rationalise that situation and to produce further guidelines and recommendations. The activities of this study group culminated in the publication in 1985 of what became known as the 1985 Red
Alistair P. Russell

- 22 -

2.4. Provisions for the Seismic Upgrade of Existing Buildings

Book (NZNSEE, 1985). Again, this document was primarily a technical basis and the responsibilities of what to do with buildings still rested with local authorities. The publication was intended to promote a consistent approach throughout New Zealand for the strengthening of earthquake risk buildings. This included a recommended level to which buildings should be strengthened and the time scale to complete the requirements. The basic objective was to establish a reasonably consistent reduction of the overall risk to life which the countrys stock of earthquake risk buildings represented. Based on overseas experiences, particularly in Los Angeles in Southern California, a philosophy was accepted of providing owners of earthquake risk buildings with the option of interim securing to gain limited extension of useful life, after which the building should be strengthened to provide indenite future life. The design of interim securing systems was to be based on minimum seismic coecients which represented two-thirds of those specied in NZSS 1900, Chapter 8. For permanent strengthening measures, it was recommended that the building be strengthened to the standard of a new building, but with the design lateral forces reduced depending on the occupancy classication and type of strengthening system. This publication was widely used by territorial authorities and designers.

In 1992 the NZNSEE again set up a study group to review the 1985 publication, and this resulted in another publication, which again became colloquially known as the 1995 Red Book (NZNSEE, 1995). This document extended the approach and content of its predecessor and took into account the changing circumstances, technical developments and improved knowledge of the behaviour of URM buildings in earthquakes. In particular, earthquake risk buildings in that document were taken to include all unreinforced masonry buildings, and not just those which were dened as earthquake prone in terms of the Building Act of the time, which still referred back to the 1965 code. Another key dierence from the 1985 Red Book was that a single stage approach to strengthening was suggested, in contrast to the two stage securing and strengthening procedure of the 1985 document. The guidelines also highlighted the dierences in analysis for unsecured buildings in comparison to a building which has positive connections between oor, roof and wall elements, and cantilever elements secured or removed. Greater emphasis was placed on the assessment of the likely performance of URM buildings in their original form and with interim securing only in place, as distinct from the performance of the building with
- 23 Alistair P. Russell

Chapter 2. Background on Potentially Earthquake Prone Buildings

any strengthening work which was subsequently found to be necessary. Furthermore, material strengths were given in ultimate limit state format. Historic or heritage buildings were not given any specic or separate treatment, and the guidelines stated that the issues of risk versus the practicalities of strengthening associated with historic buildings require evaluation on a case-by-case basis. The principal problem with such buildings is that the greater the level of lateral forces that is specied for strengthening, the greater the risk of damaging the fabric that is to be preserved. (NZNSEE, 1995)

After the introduction of a new Building Act in 2004 (New Zealand Parliament) the Department of Building and Housing supported NZSEE in producing a set of guidelines, Assessment and Improvement of the Structural Performance of Buildings in Earthquakes (NZSEE, 2006). This was a major review and extension of previous guidelines, to account for the wider scope of the proposed new legislation. Prior to the enacting of the Building Act 2004, the term earthquake risk building related only to URM buildings, but now an earthquake prone building could be of any material; steel, concrete, timber or masonry. The level of risk posed by buildings constructed as recently as the 1970s was more widely appreciated, in particular the inadequate performance of reinforced concrete structures due to decient detailing. Denitions of earthquake prone and earthquake risk also changed. Essentially, earthquake prone buildings were dened as those with one-third or less of the capacity of a new building. While the Building Act itself still focussed on buildings of high risk (earthquake prone buildings), NZSEE considered earthquake risk buildings to be any building which is not capable of meeting the performance objectives and requirements set out in its guidelines, and earthquake prone buildings formed a subset of this. Moreover, NZSEE expressed a philosophical change, in acknowledgment of the wide range of options for improving the performance of structures that are found to have high earthquake risk. Some of these options involve only the removal or separation of components, and others aect a relatively small number of members. In line with performance-based design thinking, the term strengthening was replaced with improving the structural performance of, highlighting the fact that such solutions as base isolation were not strengthening but were an eective way of improving structural performance.

Alistair P. Russell

- 24 -

2.5. Conclusions

The guidelines provided both an initial evaluation procedure (IEP) and a detailed analysis procedure. The IEP can be used for a quick and preliminary evaluation of existing buildings, and takes into account the building form, natural period of vibration, critical structural weaknesses (vertical irregularity, horizontal irregularity, short columns and potential for building-to-building impact) and the design era of the building. Based on this analysis, if a territorial authority determines a building to be earthquake prone, the owner may then be required to take action to reduce or remove the danger, depending on the territorial authoritys policy and associated timeline. The level required to reduce or remove the danger is not specied in the Building Act or its associated regulations. The Department of Building and Housing suggested that territorial authorities adopt as part of their policies that buildings be improved to a level as near as is reasonably practical to that of a new building. Most territorial authorities took the view that they could not require strengthening beyond one-third of new building standard, but a significant number of territorial authorities included requirements to strengthen to two-thirds of new building standard, in line with NZSEE recommendations. In developing policies on earthquake prone buildings, most territorial authorities recognised the need for special treatment and dialogue with owners when heritage buildings were aected. It is believed by the Department of Building and Housing that the legislation has required each local community to put earthquake risk reduction on its agenda, and has left the local community to develop appropriate policies that reect local conditions and perceptions of earthquake risk (Hopkins et al., 2008).

2.5

Conclusions

The New Zealand Government, with the support of the Department of Building and Housing, has introduced legislation to mitigate the danger to life from potentially earthquake prone buildings. URM buildings in New Zealand pose a unique problem. It is known that such buildings perform poorly when subjected to lateral forces produced by earthquakes, and yet the heritage value to the community is such that retrot in order to improve the seismic performance whilst maintaining the heritage character of such buildings is a more suitable option than removal and replacement.

- 25 -

Alistair P. Russell

Chapter 2. Background on Potentially Earthquake Prone Buildings

This chapter has traced the development of building codes in New Zealand and has shown that signicant attention has been given to the assessment and improvement of existing URM buildings for many years, particularly by the New Zealand Society for Earthquake Engineering.

Alistair P. Russell

- 26 -

CHAPTER 3

Architectural Characterisation

As a precursor to analysing the structural seismic performance of New Zealands earthquake risk URM buildings, it is important to understand their geometric and aesthetic characteristics. This enables an understanding of what typical earthquake response characteristics to expect when assessing URM buildings. This chapter outlines general structural congurations (typologies) which apply specically to New Zealand URM buildings. Distinctions between typologies are made on the basis of building height and the geometry of the buildings footprint. Seven typologies are identied as one storey isolated, two storey isolated, three-and-higher storey isolated, one storey row, two storey row, three-and-higher storey row structures, and URM structures which do not have a uniform ground footprint or which do not t easily into the rst six typologies. Other characteristics are outlined, in terms of materials and wall geometries, and how these relate to the overall architectural typologies. This provides a framework for investigating suitable seismic retrot solutions for New Zealands URM building stock.

- 27 -

Alistair P. Russell

Chapter 3. Architectural Characterisation

3.1

Introduction

Within the architectural characterisation of URM buildings, the broadest and most important classication is that of the overall building conguration. The seismic performance of a URM structure depends on its general size and shape. A small, low-rise square building, when subjected to induced seismic forces, will behave dierently than a long, row-type, multi-storey building (Robinson and Bowman, 2000). In addition to this, retrot interventions which may be appropriate for one type of building may not be appropriate for another, dierent type of building. From a heritage perspective, The New Zealand Historic Places Trust recognises the importance that the conguration of a building has on its behaviour during an earthquake. It states that buildings up to three storeys high with many walls can generally withstand a reasonable amount of earthquake shaking, whereas buildings higher than three storeys with open-plan frame-type geometries are much more vulnerable, (Robinson and Bowman, 2000). And in a more general sense, Magenes (2006) notes the importance of building conguration in the overall seismic performance of masonry buildings. He states that the main factor that produces the variation of the OSR (overstrength ratio) is the geometric conguration (plan and number of storeys). The OSR arises from the analysis of the inelastic cyclic deformation and energy dissipation capacity of a structural unreinforced masonry system.

The purpose of classifying URM buildings in New Zealand into typologies has two facets. The rst of these is to provide a framework for the Seismic Retrot Solutions research project, and in particular, for the aspect of the project on URM. In order to conduct research into both the seismic assessment and seismic retrot of URM buildings it is necessary to rst know the nature of what is being assessed or retrotted. Research into seismically improving the performance of structures needs to focus not only on generic retrot science, but on providing tailored retrot solutions for New Zealand-specic structures. The second facet of classifying URM buildings is how the typologies are applied. When considering an assessment of a specic buildings expected seismic performance, the typology which that structure is classied into can be made use of in the early stages of the analysis. The performance of a prototype or generic structure in a given Typology can be applied to the specic structure under consideration and gives a basic understanding
Alistair P. Russell

- 28 -

3.2. Background of Building Typologies

of the expected seismic performance. In this context, most typological assessments will be for the purpose of initial assessment, as a rst step, and will provide some parameters from which to conduct further detailed analysis.

A systematic, consistent and convenient method is needed to accurately capture the state of the building stock on a national, regional and local level. In New Zealand the whole country is under the jurisdiction of the Department of Building and Housing, and legislation for structures and construction is governed by the Building Act 2004 (New Zealand Parliament, 2004) (see Chapter 2). The country is divided into districts over which Territorial Authorities (TAs) have authority and responsibility for new and existing structures. The Building Act 2004 required TAs to adopt a policy on earthquake-prone buildings by the 31st of May 2006 (New Zealand Parliament, 2004). The rst step in the policy developed by each TA requires a preliminary assessment of their building stock to determine the number and types of buildings with the potential to be earthquake-prone. From the point of view of an active approach in implementing such a policy, classifying buildings into typologies is a useful tool for TAs in New Zealand to assess their building stock. While it is not envisioned that a one-size-ts-all solution is viable for all URM buildings, for initial assessments and vulnerability analyses, classications of buildings into groups is a useful and necessary exercise. Also it is an achievable rst step, before having to undertake the daunting task of analysing every building under a TAs jurisdiction. It is envisioned that the use of typologies developed here will assist TAs throughout New Zealand to classify their building stock for seismic vulnerability assessments and the economic implications of their at-risk building stock.

3.2

Background of Building Typologies

The word typology is used as a classication according to a general type, and in the sphere of architectural characterisation dierent groupings of buildings can be classied correspondingly. Categorising buildings into typologies is a useful way of grouping them according to common features or elements. The study of typologies is widely used in the
- 29 Alistair P. Russell

Chapter 3. Architectural Characterisation

eld of architecture to dene buildings according to spatial or formal qualities, context or function.

Building typologies are generally used as a method of categorising the building stock on a large scale, although there are dierent levels for distinguishing typologies. Some researchers have utilised the concept of typologies on a regional level, some for urban/rural distinctions, some applied to specic towns, and others have looked at particular parts of towns. While the use of typologies in some locations is specic to that location and may not have the same applications elsewhere, there are common principles that can be applied generally.

Quatrem` ere de Quincy, the editor of the Dictionnaire dArchitecture between 1788 and 1825, rst explained the idea of typology in his seminal text Dictionnaire historique dArchitecture (1832), and drew attention to the contrast between model, which is an exact and replicable plan for a building, and typology which is a category or rule for grouping specic models (Lavin, 1992). To explain typology, Pontius (2007) drew a parallel with the concepts in biology of phenotype and genotype, where genotype is the genetic constitution of an individual organism, and phenotype is the set of observable characteristics of an individual resulting from the interaction of its genotype with the environment. Argan (1996) and Rossi (1982), in particular, investigated a path of architectural theory where instead of looking at a building as an isolated element on the landscape they shifted their focus towards a study of cities and context. For them, a buildings function could adapt to the dynamic life of the surrounding city, but the basic formal relationships within the structure and to adjacent structures remain the same. Pontius (2007) recognised the strength of categorising buildings according to typology without depending on the buildings use, because the inherent structural properties are much less likely to change over time than the function of the building.

Much previous research involving typologies in structural engineering has been on case studies villages, cities (or parts of a city) or regions. It is a way of characterising the building stock, and has been done frequently for this purpose. In Germany building
Alistair P. Russell

- 30 -

3.2. Background of Building Typologies

typologies have been used for the purpose of identifying energy consumption and demand (Neidhart and Sester, 2004). This research was performed with residential housing as the focus, and buildings were classied as one-family house, row house, small more-family house or large more-family house, and then classied according to building age. This allowed specic heat demand to be determined on a systematic basis. In Portugal, Louren co et al. (2006) investigated moisture and weathering eects on historic building stock, and a useful tool was classifying such buildings into typologies, rather then examining individual buildings. This was a case-study, and again was used with mainly residential dwellings as a focus, and clear distinctions could be made to distinguish building typologies. Common details and features within typologies allowed common defects to be detected.

The most relevant research (to seismic assessment and retrot) which has been published in the eld of structural engineering involves the use of typologies in seismically active areas. The classication of historic buildings into typologies has been used for the purpose of seismic vulnerability analysis, and in this case, many such historic buildings are constructed of unreinforced masonry. The purpose of typological analysis is to nd as much information on buildings as possible. This information used in the context of the seismicity of the region, for example, can show the seismic vulnerability of those structures. The concept of classifying buildings into typologies has been used by researchers in Italy, Portugal, Greece, Slovenia, Canada, the United States and Iran. The object of the vulnerability assessment can vary in domain from villages, to towns, to regions and even to whole countries.

3.2.1
Italy

European URM Typology Studies

As part of developing a procedure to quantify the reduction in seismic risk for historic structures, DAyala and Speranza (2003) analysed four case-study towns in the Marche region in Italy and characterised the building stock into typologies. It was found that although the typologies diered between towns in terms of several factors such as overall
- 31 Alistair P. Russell

Chapter 3. Architectural Characterisation

dimensions, foundations, and wall-oor connections, similarities could be seen in terms of the horizontal structures and masonry characteristics. Typologies were analysed and assigned vulnerability classes (low, medium, high, extreme), according to the level of seismic risk associated with the overall conguration and structural details.

Similarly Binda has characterised buildings into typologies in Italy (on various levels), for the purpose of assessing the seismic vulnerability of historic towns in earthquakes (Binda, 2005, 2006a; Binda et al., 2005). The approach was taken to analyse whole historic towns, and not single buildings, and the purpose of this was to develop a database with rescue plans and design interventions for the preservation of cultural heritage. One facet of the research focused on four sample areas in the province of Perugia (Binda, 2005; Binda et al., 2005), while another facet enabled more general conclusions to be drawn on the overall state of URM buildings in Italy as a whole (Binda, 2006a). It was noted that representative typologies of a town can be easily recognisable through similar features (number of storeys, exposure, type of fa cade, material and structural elements). On a regional scale in the case of Perugia, three typologies were identied and the structures were classied as either simple isolated buildings, row buildings, or complex buildings (Binda, 2005; Binda et al., 2005). On a more general scale, which could be applied to the whole of the historic Italian building stock, seven typologies were identied (Binda, 2006a) (see also Figure 3.1): Typology A: isolated houses and/or dwellings; Typology B: row houses; Typology C: palaces; Typology D: bell towers; Typology E: arenas; Typology F: churches and cathedrals; and within this typology, Typology F1: churches, plan based on Latin cross scheme; Typology F2: churches, with a central plan. Although Argan (1996), Pontius (2007) and Rossi (1982) remark that buildings generally
Alistair P. Russell

- 32 -

3.2. Background of Building Typologies

Figure 3.1: Typologies for the Italian URM building stock (reproduced from Binda (2006a))

retain consistent structural congurations while the function of the building may change over time, Binda et al. (2005) points out that the evolution of the building may still have a signicant eect on its seismic vulnerability: born as an isolated building, it could have become a row building or a complex one. . . The more complex the building is, the more dicult the detection of its vulnerability is; therefore, its structural evolution should
- 33 Alistair P. Russell

Chapter 3. Architectural Characterisation

be known as much as possible. It was also noted that a geometric survey may not be sucient to ascertain a comprehensive understanding of the site or building, and eort should be made to research possible modications over time. Nevertheless, a minimalist approach can still yield useful and signicant information by sampling from buildings representative of the whole, particularly with regard to materials used on an urban scale. This may also be useful to detect local and compatible materials and techniques for representative prevention and repair measures.

Valluzzi et al. (2005a) performed a simplied yet comprehensive assessment of two towns in the Treviso Province in Italy. Through the use of kinematic models describing the loss of equilibrium of structural macro-elements and classication of buildings into typologies, vulnerability assessments for the whole of each town and proposed intervention techniques were presented, without needing to analyse each individual structure. The criteria used to dene the typologies was based on dimensions, presence of contiguous constructions (isolated or row buildings) and the presence of colonnades (a row of columns supporting a roof) at ground level.

The limitations of general classications have also been reported. As part of a similar study into seismic vulnerability in Perugia, Valluzzi et al. (2005b) indicates that complex buildings in seismic areas have specic vulnerability aspects which can make it very dicult to apply or adopt common structural assessment procedures usually applied for regular buildings. Still, if there are a sucient number of complex buildings, common features and details can be made use of as Binda (2006a) notes in the case of churches.

Portugal Vicente et al. (2006) analysed traditional building stock in the old city centre of Coimbra in Portugal. A vulnerability assessment was conducted in conjunction with a GIS mapping tool, which allowed a global view of the site in question. The authors determined that without this, inadequate decisions in terms of rehabilitation and refurbishment policies are often made. In the area under consideration, the historic buildings were classied mainly
Alistair P. Russell

- 34 -

3.2. Background of Building Typologies

according to dimensions and construction materials. A relational database of building characteristics that govern the structural behaviour of historic buildings, with particular focus on their seismic vulnerability, was connected with a GIS database. This allowed the mapping of dierent damage scenarios and risk reduction actions associated with strengthening interventions. Also, the GIS tools were used to provide a global overview of building quality, current conservation status, risk scenarios and potential damage mapping for dierent seismic intensities. Vicente et al. (2006) intended that the GIS and database system could be applicable for other regions and old city centres, and if necessary easily adapted and modied for specic building features.

Greece Karantoni and Bouckovalas (1997) performed a vulnerability study considering almost all unreinforced masonry buildings in the town of Pyrgos in Greece after four signicant earthquakes in or near the town. Buildings were classied according to the masonry material (adobe, stone or clay brick) and number of storeys (one or two). Another parameter which was recorded for each building was the time period of construction. This was an important factor as it not only determined the masonry material but also the structural form of the building. Four basic time periods were distinguished: 18001850, 18501900, 19001940 and 1940present. It was determined that old adobe masonry buildings exhibited poor seismic performance, while more recent stone and brick masonry buildings exhibited a much better seismic response. Perhaps the most useful conclusion was that Prygos was thought to be a typical Greek town, and that buildings of other Greek towns would behave almost with the same manner during the actions of near-eld earthquakes. Thus it was thought that the results of this study could be applied more widely throughout the country and would allow more reliable planning of seismic scenarios for the future.

Slovenia Toma zevi c and Lutman (2007) researched typologies in Slovenia and recognised that typologies vary from region to region and from urban to rural areas. In the context of
- 35 Alistair P. Russell

Chapter 3. Architectural Characterisation

Slovenia, construction materials are usually those which are locally available limestone and slate, and sometimes brick elsewhere in the country. It was found that in cities, stone masonry houses are mostly three or four storeys (Figure 3.2), but are limited to two storeys in rural areas. The distribution of walls is usually uniform in orthogonal directions and because of the thick walls, the wall/oor area ratio is large, sometimes up to 10%. Floor structures are usually wooden and few ties are used to connect them with the walls. Roof structures are also wooden and are usually covered with ceramic tiles. The authors note that although the typology of heritage buildings varies from region to region and country to country, similar observations can be made regarding the response to earthquakes. Also, damage occurring in such buildings can be classied in a uniform way. It was stated that the two main causes of damage to URM buildings is inadequate structural integrity (particularly connections between walls at oor levels), and inadequate structural resistance, and can result in diagonal shear cracking and disintegration

Figure 3.2: Typical masonry buildings in Ljubljana, Slovenia (reproduced from (Toma zevi c and Lutman, 2007))
Alistair P. Russell

- 36 -

3.2. Background of Building Typologies

of walls and partial or total collapse of buildings.

3.2.2
Canada

North American URM Typology Studies

Nollet et al. (2005) analysed historical buildings in Old Montreal. Old Montreal has been declared an historic district in Quebec, and contains buildings of signicant architectural and cultural heritage, as well as being in a seismically active zone. Montreal is somewhat older in its original European settlement than most of New Zealand. Nevertheless, there are important similarities in terms of the principles of determining typologies and their applications to the locality in focus. Much of the early construction in Eastern Canada has greater parallel to European (particularly French) construction of the time than to other American construction, particularly in the settlement period between 1680 and 1860. In the earliest periods of European settlement in New Zealand, construction bore greater similarity to British construction practices, before its own practices developed (Ingham, 2008).

The historical buildings in Old Montreal were considered to be those built before 1929, and an inventory of such buildings was classied into typologies according to construction type (masonry, wood, steel, concrete), number of storeys, and the period of construction. The date of construction is important, although Binda et al. (2005) and Blaikie and Spurr (1992) take this further and state that the age per se 1 and evolution of the structure are important factors for evaluating the condition of the structure.

It was noted by Nollet et al. (2005) that typological classication is used in many approaches to assess the seismic vulnerability of a group of buildings, and although it is
Blaikie and Spurr (1992) state that, a common observation after recent earthquakes has been that newer buildings designed to modern day standards performed much better than older types of construction. This type of statement is often directed at old buildings constructed in the 1930s and earlier. However almost identical statements were made shortly after the 1931 Hawkes Bay earthquake. Some of the new buildings referred to are now considered old and beyond redemption. The implication of this is that age per se, not just date of construction, is an important factor even for purportedly permanent building materials.
1

- 37 -

Alistair P. Russell

Chapter 3. Architectural Characterisation

generally dened for the population of buildings under study it is sometimes used at a larger scale. The implication of this is that sometimes larger scale classications are not necessarily applicable for a smaller population. For example, the Canadian typological classication is based on the descriptions given in the report ATC-21 of the Applied Technology Council of California (ATC, 1996). It was recognised that while this may be reasonable for parts of North America, including Eastern Canada, direct application of this classication to the population of buildings in Old Montreal was questionable. It was concluded after the analysis of the inventory that there were notable dierences between structural characteristics of those buildings and the North American typologies identied in ATC-21. This conrmed the necessity to develop an adapted method for evaluation of the seismic vulnerability for historic areas of Quebec. It was acknowledged that a more general classication system was a useful starting point, from which tailored typological analyses could then be developed.

The United States Erbay and Abrams (2004, 2007) recognised a need to identify the relative impacts that rehabilitation may have on performance of large populations of buildings, instead of individual buildings, so that regional losses can be estimated and mitigated. This information is useful for public ocials, emergency response managers and other public stakeholders. In their study, the authors state that as a gross extrapolation, structural characteristics are generalised for a wide range of possible masonry buildings inherent in a region. Nevertheless, it was thought that the estimated response and loss of the entire population of buildings in the sample region was not sensitive to probable deviations of some buildings from such generalised characteristics. Within this simplied model a number of building specic parameters were varied as input to the damage model. These parameters included number of storeys, oor area, storey height, oor aspect ratio, wall area to oor area ratio, average pier height, uniform load over storey, elastic modulus of masonry, wall thickness and masonry density. The sensitivity analysis identied that hazard-loss relationships which were unacceptably scattered for individual building loss calculations, could be utilised to estimate risk/loss at a regional level (Erbay, 2004).

Alistair P. Russell

- 38 -

3.2. Background of Building Typologies

3.2.3

Iranian URM Typology Studies

Fragility curves have been reported as a useful method of estimating the probability that buildings will exceed a specic state of damage for a given seismic intensity. Karimi and Bakhshi (2006) investigated the development of fragility curves for URM buildings before and after seismic upgrade in Iran. Although the typologies which were used in the modelling consisted only of single-storey and three-storey structures, useful conclusions on the expected damage in URM structures could be drawn, which could then be extrapolated

E-W

N-S

ress, earthquake N-Sto structures Figure 15: Normal stress, earthquake in E-W and in applied on a general basis. ection direction

ulted roofs

Figure 3.3: Typology in Iranian rural residential building stock; comparison of model and Figure 17: of prototype and model prototype (reproduced from Comparison (Ghannad et al., 2006)) A study was conducted in Iran to classify existing rural houses of Iran (Bakhshi et al., 2005; Ghannad et al., 2006; Mousavi Eshkiki et al., 2006). From eld inspections, nine

4.

CONCLUSION

typologies of rural in houses recognised as representative all rural houses in of Iran. minant types of rural houses Iranwere was presented. To have a of rough estimation study classed these typologies to building material (brick/cement block, of baked andThe unbaked adobes, tests were according performed on samples taken from some ng to field investigations, arched roof (model 1)(at, and arched). brick walls withtycement block,adobe stone, walls adobe, with wooden) and geometry of roof For each uses were known as dominant vulnerable dwellings damaged in Zarand earthquake. To pology, the constituent percentage of the overall building stock and their general location e mechanisms of typical rural houses of Iran, elastic numerical models of these two were determined, as well as typicalwith decient element modelling was cone developed and the results were compared thedetails. actual Finite examples observed by the ducted on structures representative of each typology (see Figure 3.3) and the dominant failure modes were ascertained.

- 39 -

Alistair P. Russell

Chapter 3. Architectural Characterisation

3.2.4

Previous Research into URM Typologies in New Zealand

When analysing the performance of URM buildings in previous earthquakes, Blaikie and Spurr (1992) made observations on the inuence of overall conguration on building collapse. It was noted that square shaped buildings have a lower susceptibility to earthquake damage than either irregular or rectangular buildings. The same authors also reported partially on the nature of New Zealands URM building stock, by surveying buildings in Wellington City, Nelson City and the Petone suburb of Lower Hutt City. These three locations were thought to be representative of a main city, a provincial city and a suburban shopping district respectively. The data used were last updated in May 1988. It was concluded that the bulk of the sample populations earthquake risk URM building stock had one or two storeys, and that three or more storeys were found almost exclusively in Wellington City. This was extrapolated to also conclude that research into New Zealands existing URM buildings should be directed towards buildings having three or fewer storeys.

3.3

New Zealand URM Building Typologies

It has been identied that the New Zealand building stock warrants seven typologies, which are outlined in Table 3.1. Buildings are separated according to storey height, and whether they are an isolated, stand-alone building or a row building made up of multiple residences joined together in the same overall structure. A suggestion of the expected importance level of the structure is also given, according to AS/NZS 1170.0:2002 (Standards New Zealand, 2002a). The prevalence of URM structures according to the number of buildings in that typology which make up the overall building stock is shown in Table 4.3. Typology A are single storey isolated structures, Typology B are single storey row structures, Typology C are two storey isolated structures, typology D are two storey row structures, Typology E are three and higher storey isolated structures, Typology F are three and higher storey row structures and Typology G structures are religious, institutional and industrial structures. For all row-type structures (B, D and F), the dimensions
Alistair P. Russell

- 40 -

3.3. New Zealand URM Building Typologies

of each individual substructure can vary, that is, they are often heterogeneous.

Typologies are distinguished according to two factors storey height and building footprint. The storey height signicantly inuences the way the ground oor walls behave under lateral loads. This is because dierent axial loads can cause dierent failure modes of those walls. Walls of multi-storey buildings are usually thicker on the ground oor than on higher storeys, and also much thicker than single storey walls. The second distinction between typologies is whether they are row or isolated buildings. This has signicant implications for torsional eects from lateral loads. A long row building can be thought to be more torsionally stable than an isolated structure, which tends to have a ground footprint with aspect ratio of between 1:1 and 1:2. A row building could have a ground footprint aspect ratio of up to 1:10, or even higher. Also, isolated structures tend to have an open front with corresponding low stiness in that direction. This means the centre of stiness is not in the centre of the building. Pounding eects can be signicant in row buildings, particularly where oors are made from concrete or are at dierent levels. Pounding is usually not a problem in isolated structures but can occur where there is insucient clearance2 between neighbouring structures.

A prototype (mean) structure for each typology has been established, which enables a standard modelling procedure to be applied to any structure based on its typology. The information for these typologies has been obtained from building surveys throughout New Zealand. The locations of buildings and origins of photographs is shown in Figure 3.4.

Within the above typologies, further distinctions have been made. For example, Typology A buildings can be divided into those which have a dividing wall in the centre (Typology A1), and those which do not (Typology A2). Typology G buildings are generally monumental structures and those which do not t easily into the other categories, and usually for such structures unique problems are presented, and unique analyses are necessary. Nevertheless there are useful sub-classications which can also be made within this
ASCE (2007) states buildings shall be separated from adjacent structures to prevent pounding by a minimum distance. . . less than 0.04 times the height of the level under consideration above grade at the location of potential impact
2

- 41 -

Alistair P. Russell

Chapter 3. Architectural Characterisation

grouping. For example, Typology G1 buildings are religious buildings in New Zealand, Typology G2 are warehouses and factories with very large tall sides and large open spaces inside and Typology G3 are institutional buildings such as ferry terminals and train stations. Further detail on each typology is given in subsequent sections.

AS/NZS 1170.0 Table 3.1 is reproduced here as Figure 3.5 to give an indication of the relative importance levels of URM typologies (further details can be found in Chapter 3 of AS/NZS 1170.0 (Standards New Zealand, 2002a)). All structures identied fall into importance level 2 or higher, with medium to high consequences for loss of human life.

Auckland Paeroa/Thames

Taihape Bulls Palmerston North Nelson Wellington

Christchurch Ashburton Temuka Timaru Oamaru Dunedin

Figure 3.4: Locations of buildings surveyed throughout New Zealand

Alistair P. Russell

- 42 -

3.3. New Zealand URM Building Typologies

Table 3.1: URM typologies Typology Description Prevalence (rank) Importance level (AS/NZS 1170.0) 2 Description

One storey, isolated buildings

One storey URM buildings, often with an open front. Examples include convenience stores in suburban areas and small ofces in a rural town. One storey URM buildings with multiple occupancies joined with common walls in a row. Typical in main commercial districts, especially along the main street in a small town. Two storey URM buildings, often with an open front. Examples include small cinemas, a professional oce in a rural town and post oces. Two storey URM buildings with multiple occupancies joined with common walls in a row. Typical in commercial districts. Three + storey URM buildings, for example oce buildings in older parts of Auckland and Wellington. Three + storey URM buildings with multiple occupancies, joined with common walls in a row. Typical in industrial districts, especially close to a port (or historic port). Churches (with steeples, bell towers etc), water towers, chimneys, warehouses. Prevalent throughout New Zealand.

One storey, row buildings

Two storey, isolated buildings

2/3

Two storey, row buildings

Three or more storey, isolated buildings Three or more storey, row buildings

3/4

3/4

Institutional, industrial and religious

3/4/5

- 43 -

Alistair P. Russell

probability of exceedance (P) for wind, snow and earthquake loads shall be determined as given in Table 3.3.
NOTE: Guidelines for limits associated with serviceability events are given in Appendix C.

Chapter 3. Architectural Characterisation TABLE 3.1

CONSEQUENCES OF FAILURE FOR IMPORTANCE LEVELS
Consequences of failure Low Description Low consequence for loss of human life, or small or moderate economic, social or environmental consequences Medium consequence for loss of human life, or considerable economic, social or environmental consequences High consequence for loss of human life, or very great economic, social or environmental consequences Circumstances where reliability must be set on a case by case basis Importance level 1 Comment Minor structures (failure not likely to endanger human life) Normal structures and structures not falling into other levels Major structures (affecting crowds) Post-disaster structures (post disaster functions or dangerous activities) Exceptional structures

Ordinary

2 3 4 5

High

Exceptional

Figure 3.5: Building importance levels from AS/NZS 1170.0 Table 3.1

3.3.1

Parameters for Dierentiating Typologies

Storey Height Typologies are separated according to whether the buildings are one storey, two storey or three or more storeys. Buildings taller than three storeys are not common in New Zealand and exist mostly in the central business districts (CBD) of the largest city, Auckland, and the capital, Wellington, as well as some other South Island cities, such as Timaru and
--`,``,,`,`,,,```,,``,,``,,```,-`-`,,`,,`,`,,`---

Copyright Standards New Zealand Provided by IHS under license with SNZ No reproduction or networking permitted without license from IHS

Dunedin. There are large numbers of both one and two storey buildings throughout New
Licensee=UNI OF AUCKLAND/5936938001 Not for Resale, 06/05/2006 18:13:17 MDT

COPYRIGHT

Zealand, but three and higher storey buildings are few in number and a single typology to classify all URM buildings that are three storeys or higher is sucient. Moreover, the dierence in expected seismic behaviour between a three and four storey building is signicantly less than the dierence between a one and two storey building, for example. This is often because three and higher storey buildings tend to be of masonry frame construction (on at least one face of the building, and usually both the front and back faces), in contrast to solid wall construction, and have much thicker walls in the lower storeys. As a broad generalisation, in masonry frames rocking of piers between windows and openings is the expected in-plane behaviour when subjected to lateral seismic forces (Abrams, 2000), and diagonal shear failure is less likely. For walls without openings (or with small openings), and depending on the magnitude of the axial load, the expected in-plane failure mode in an earthquake is likely to be either sliding shear failure, diagonal tension (shear) failure, or global rocking of the wall itself.
Alistair P. Russell

- 44 -

3.3. New Zealand URM Building Typologies

Furthermore, another reason why buildings are distinguished according to storey height is because of the axial loads acting in the walls. One storey buildings with low axial loads are less likely to exhibit diagonal shear failure and are more likely to rock or slide. The bottom storey walls in a taller building are more likely to fail in shear because of the higher axial loads that they are subjected to. But the thickness of the ground oor walls is usually greater than the thickness of the walls higher up the building in Typology E and F structures, meaning that large lateral forces are necessary in order to generate sucient shear stresses in these walls to cause cracking or failure.

Building Footprint The second primary characteristic for separating buildings into typologies is the building footprint. That is, whether the structure is considered a stand-alone, isolated, (almost) square building, or a row building made up of multiple residences joined together with common walls. This accounts for Typologies A F, whereas those buildings with a nonuniform ground footprint (for example, many URM churches) will t into the Typology G classication.

In row structures containing walls common to neighbouring residences, pounding can increase the lateral forces on an adjoining structure during earthquake loading. This is especially likely when oor or ceiling diaphragms in neighbouring residences are at different levels. Dierent heights for the force transfer into the common wall can result in punching shear failure of the wall or diaphragm detachment and collapse. The eects of pounding are greater when concrete oor diaphragms are in the structure, compared with timber diaphragms. Conversely when diaphragms connected to shared walls are at consistent heights within the structure, the seismic resistance is greatly enhanced due to the increased stiness in one direction. Essentially square or round buildings with well distributed walls generally have a greater torsional resistance than buildings with less evenly distributed lateral force resisting walls (Robinson and Bowman, 2000). Long row structures have dierent torsional properties than isolated buildings.

A signicant dierence between isolated buildings and row buildings becomes evident at
- 45 Alistair P. Russell

Chapter 3. Architectural Characterisation

the time of upgrading the structure. An isolated structure usually contains few residences, perhaps two shops for example, or occasionally more. Row structures may contain many residents, even ten or more. An isolated structure is generally considered just that a single structure. A row structure, despite behaving in an earthquake as a single interconnected structure, may be perceived as dierent buildings. It may be more dicult to conduct earthquake strengthening on an entire row structure at one time compared with seismic retrot of an isolated structure. If seismic retrot interventions are implemented on only a part of a structure, such an intervention may be ineective.

3.3.2

Details of New Zealand URM Typologies

Typology A One Storey Isolated Buildings Typology A buildings are one storey isolated structures, and are the second most prevalent in New Zealand, out of the seven typologies identied. Equally common in both small towns and large cities, these structures are today often used as convenience stores, or small commercial premises. Typically they have an approximately square ground footprint, but may have longer side walls making the building rectangular. Sometimes there is a dividing wall through the centre of the building, usually with no piercings, but sometimes there is a doorway width opening. That dividing wall can be one or two leaves thick, but single leaf dividing walls are non-load bearing. Typology A1 refers to buildings with a dividing wall, and Typology A2 refers to those buildings without a dividing wall. Generally the external side walls have no openings. The front may be largely open, with up to 90% of the width of the front open, with glass windows, ranch-sliders or doorways, and the rear wall usually has openings in the form of doorways. The external side walls usually have no openings. A parapet is usually on top of the front wall, and often the side walls slope from the top of this parapet to the top of the rear wall, meaning that the side walls are of trapezoidal shape. Also the central wall, if present, follows the same shape as the side walls and slopes from the top of the parapet to the back wall. Sometimes there is no parapet on the front. The side walls are usually either two or three leaves thick, as are the front and back walls. The ceiling diaphragm is usually timber with a suspended ceiling underneath, and light cladding over the top, usually either tiles or corrugated iron. Typology A strucAlistair P. Russell

- 46 -

3.3. New Zealand URM Building Typologies

tures usually contain either one of two residences, and this usually corresponds to either A1 or A2 structures. In the long direction (L1 in Figure 3.6) the building is sti, but because of the large openings on the front face, the building is less sti in the L2 direction.

Figure 3.6 and Table 3.2 show typical dimensions of a Typology A1 structure, with an internal wall in the centre of the building, and Figure 3.7 and Table 3.3 show typical dimensions of a Typology A2 structure, with no centre wall. Upper and lower bound dimensions are shown, as well as typical dimensions in Typology A buildings. (The mean dimensions do not necessarily correspond with a mean building.) Photographic examples of Typology A buildings are shown in Figure 3.8.

Figure 3.6: Overall dimensions of a Typology A1 building

- 47 -

Alistair P. Russell

Chapter 3. Architectural Characterisation

Table 3.2: Typical dimensions of a Typology A1 building Dimension Lower bound structure (m) 6 4 1 0 3 2 3 0 Mean Upper bound structure structure (m) (m) 8 6 1.2 0 4 2.5 4 0.5 16 8 2 1 6 4 6 2

L1 L2 P1 P2 H1 H2 F1

Overall length Width of residence Height of parapet at front Height of parapet at rear Height from ground to bottom of parapet Height of front opening Width of front opening

G1 Depth of rear wall below front street level

Figure 3.7: Overall dimensions of a Typology A2 building

Alistair P. Russell

- 48 -

3.3. New Zealand URM Building Typologies

Table 3.3: Typical dimensions of a Typology A2 building Dimension Lower bound structure (m) 6 6 1 0 3 2 4 0 Mean Upper bound structure structure (m) (m) 8 8 1.2 0 4 2.5 6 0.5 16 12 2 1 6 4 10 2

L1 L2 P1 P2 H1 H2 F1

Overall length Width of residence Height of parapet at front Height of parapet at rear Height from ground to bottom of parapet Height of front opening Width of front opening

G1 Depth of rear wall below front street level

Further photographic examples of Typology A buildings are shown in Appendix A.

- 49 -

Alistair P. Russell

Chapter 3. Architectural Characterisation

(a) Parnell (Auckland)

(b) Ashburton

(c) Temuka (Canterbury)

(d) Seatoun (Wellington)

(e) Epsom (Auckland)

(f) Thames

Figure 3.8: Typology A buildings single storey isolated

Alistair P. Russell

- 50 -

3.3. New Zealand URM Building Typologies

Typology B One Storey Row Buildings

Typology B buildings are one storey row structures, dierentiated from Typology A structures in that they are essentially made up of multiple Typology A buildings joined together. They are slightly more prevalent throughout New Zealand (the most prevalent out of the seven typologies identied) than Typology A structures, and often occur in small provincial towns, especially along the main street which is the commercial centre of the town. Single storey row structures tend to be smaller in the overall length of the block than corresponding two storey row structures (Typology D). Generally there is a uniform front to the building which faces the street, but the location of the rear wall in each residence may vary depending on what extra rooms are appurtenant to the main structure a toilet and bathroom block, for example. The amenities block could be separated by a structural wall, or it might only be behind a non-structural party wall. Each residence can vary in width and height, and especially the height of each parapet. That is L2, H1 and P1 in Figure 3.9 can vary for each individual tenant. There can be anywhere from three to ten or more tenancies joined together in the one overall structure. Usually the overall width of the building is much longer than its length from the front to the back (L5 and L6 respectively in Figure 3.9). In the original construction the individual tenants were separated by non-pierced structural walls, see Figure 3.9. Similar to the side walls, these internal structural walls may slope from the top of the parapet at the front of the building to the back, or the top of the wall may be horizontal (refer to P1 and P2 in Figure 3.9). There are often individual parapets on the front of the building, which usually are unique for each residence, although the original parapet may not reect the current tenants. For example, there might be a single parapet over two tenants which are separated by non-structural party walls, which were not part of the original construction. Inside each individual occupancy there may be various non-structural walls, for example, in a clothing retail shop, there may be a tting room separated from the rest of the shop oor. Similar to Typology A structures, the external side walls are usually two or three leaves thick and not pierced, the frontages may be up to 90% open and there are suspended ceilings underneath roofs of either tiles or cast iron cladding. Typically Typology B structures are occupied by commercial tenants retailers, hairdressers, cafes etc. Row structures like Typology B and D structures were often built on sloping ground. The
- 51 Alistair P. Russell

Chapter 3. Architectural Characterisation

ground surface slopes away from the road, so that the rear of the building is lower than the front. Sometimes there is a basement level below ground oor at the back, even if the building is actually single storey at the front. See Figure A.2(d) in Appendix A. Figure 3.9 and Table 3.4 show common dimensions of Typology B structures. Photographic examples of Typology B structures are shown in Figure 3.10.

Figure 3.9: Overall dimensions of a Typology B building Table 3.4: Typical dimensions of a Typology B building Dimension Lower bound structure (m) 6 4 2 1 0 3 2 3 0 Mean structure (m) 8 8 4 1.5 0 4 2.5 6 0.5 Upper bound structure (m) 16 10 5 2 1 6 4 8 2

L1 L2 L4 P1 P2 H1 H2 F1

Overall length Width of residence Length of amenities block Height of parapet at front Height of parapet at rear Height from ground to bottom of parapet Height of front opening Width of front opening

G1 Depth of rear wall below front street level

Alistair P. Russell

- 52 -

3.3. New Zealand URM Building Typologies

(a) Ashburton

(b) Paeroa

(c) Thames

(d) Timaru

(e) Royal Oak (Auckland)

(f) Epsom (Auckland)

Figure 3.10: Typology B buildings single storey row Further photographic examples of Typology B buildings are shown in Appendix A.

- 53 -

Alistair P. Russell

Chapter 3. Architectural Characterisation

Typology C Two Storey Isolated Buildings

Typology C buildings are two storey isolated structures, and buildings of this type are the fourth most prevalent form of URM building throughout the country. Examples of Typology C buildings include small town cinemas, a small professional oce in a rural town, small town pubs and small hotels. They are prevalent throughout New Zealand, in both smaller rural centres and in city village areas (mainly older early settled parts of town). These structures are usually square, but on occasions they may have a trapezoidal or triangular ground footprint, particularly when the building is situated at an intersection of two streets on an acute angle. Such corner-type structures are classed Typology C3 structures, see Figure 3.14. Typology C1 structures are those with a dividing wall in the centre of the ground oor and C2 are those with an open-plan ground oor. Typical dimensions of each classication of Typology C structures are shown in Figures 3.15 3.17 and Tables 3.5 3.7. The most important feature of Typology C buildings is the oor and ceiling diaphragms and their connections to the walls. It is common for the bottom storey walls to be three leaves thick, and the upper walls to be two leaves thick. Timber joists supporting the oor are often only simply supported on the resulting ledge, with no positive anchorage. In other instances the joists may be embedded in the walls, but the depth of embedment is usually very small, providing little anchorage, see Figure 3.11. Often buildings exhibit spikes on their exterior (Figure 3.12(a)), which appear to be the xing for internal timber diaphragms, and may have been subsequent additions and not part of the original construction. In other instances these spikes appear to have also been subsequent additions, but for the purpose of tying multi-leaf walls together, and possibly cavity walls (Figure 3.12(b)). Typology C buildings occasionally have concrete oor diaphragms instead of timber, that are cast on top of the ground storey wall. The second storey masonry wall is built directly on top of the concrete. Sometimes the oor diaphragms are timber, but are embedded into a concrete ring beam cast on top of the ground storey wall (Figure 3.13). Other features of these structures are that the side walls are generally not pierced, there are window openings on the top storey at the front and back, and the front of the building on the ground oor may have large openings. The back wall is usually similar to the front, but the openings may be smaller, and sometimes in the upper storey there are no openings or just small frame windows. There are usually
Alistair P. Russell

- 54 -

3.3. New Zealand URM Building Typologies

parapets on the top of the front wall, which are sometimes very ornately decorated. As with one-storey structures, the side walls can have a horizontal top or can slope from the top of the parapet at the front to the top of the rear wall, and roofs are generally a truss structure, supporting cast iron cladding. Typology C3 structures are those which usually occur at a corner between two roads, and the structure results in an overall wedge shape, as shown in Figure 3.14(a). Sometimes C3 buildings occur between obtuse intersections, or at right-angle intersections also, see Figures 3.14(b) and 3.14(c). Typology C3 (corner) structures tend to be on at ground, and the parapets also tend to be of consistent height around the building. There are usually no piercings on the back face of Typology C3 buildings. Figures 3.15 3.17 and Tables 3.5 3.7 show typical dimensions of Typology C structures. Photographic examples of typology C structures are shown in Figure 3.18, and Figure 3.19 shows examples of two storey corner structures.

(a) Joists embedded in wall

(b) Joists seated on wall

Figure 3.11: Timber joist details

- 55 -

Alistair P. Russell

Chapter 3. Architectural Characterisation

(a)

(b)

Figure 3.12: Spikes on exterior wall for connecting internal oor diaphragm

(a)

(b)

Figure 3.13: Concrete ring beam between oors

Alistair P. Russell

- 56 -

3.3. New Zealand URM Building Typologies

(a) Acute intersection

(b) Obtuse intersection

(c) Right angle intersection

Figure 3.14: Location of Typology C3 building at junction of two roads

- 57 -

Alistair P. Russell

Chapter 3. Architectural Characterisation

Figure 3.15: Overall dimensions of a Typology C1 building

Alistair P. Russell

- 58 -

3.3. New Zealand URM Building Typologies

Table 3.5: Typical dimensions of a Typology C1 building Dimension Lower bound structure (m) 8 4 1 0 4 3 3 3 0 0 1 0.8 Mean structure (m) 15 7 2 0 6 5 5 4 0 0 2 1 Upper bound structure (m) 20 9 2.5 1 7 5 6 5 2 2 2.5 1.5

L1 L2 P1 P2 H1 H2 H3 F1 G1

Overall length Width of residence Height of parapet at front Height of parapet at rear Height from ground to bottom of second storey Height of front opening Height of second storey Width of front opening Depth of rear wall below front street level

W1 Height of windows above top of ground storey W2 Height of windows W3 Width of windows

- 59 -

Alistair P. Russell

Chapter 3. Architectural Characterisation

Figure 3.16: Overall dimensions of a Typology C2 building

Alistair P. Russell

- 60 -

3.3. New Zealand URM Building Typologies

Table 3.6: Typical dimensions of a Typology C2 building Dimension Lower bound structure (m) 8 8 1 0 4 3 3 3 0 0 1 0.8 Mean Upper bound structure structure (m) (m) 15 12 2 0 6 5 5 4 0 0 2 1 20 16 2.5 1 7 5 6 5 2 2 2.5 1.5

L1 L2 P1 P2 H1 H2 H3 F1 G1

Overall length Width of residence Height of parapet at front Height of parapet at rear Height from ground to bottom of second storey Height of front opening Height of second storey Width of front opening Depth of rear wall below front street level

W1 Height of windows above top of lower storey W2 Height of windows W3 Width of windows

- 61 -

Alistair P. Russell

Chapter 3. Architectural Characterisation

Figure 3.17: Overall dimensions of a Typology C3 building

Alistair P. Russell

- 62 -

3.3. New Zealand URM Building Typologies

Table 3.7: Typical dimensions of a Typology C3 building Dimension Lower bound structure (m) 8 2 8 1 1 4 3 3 3 0 1 0.8 Mean structure (m) 12 5 10 1.2 1.2 6 5 5 4 0 2 1 Upper bound structure (m) 18 8 16 2 2 7 5 6 5 2 2.5 1.5

L1 L2 L3 P1 P2 H1 H2 H3 F1

Overall length Width of front of building Width of rear of building Height of parapet at front Height of parapet at rear Height from ground to bottom of second storey Height of front opening Height of second storey Width of front opening

W1 Height of windows above top of lower storey W2 Height of windows W3 Width of windows

- 63 -

Alistair P. Russell

Chapter 3. Architectural Characterisation

(a) Kingsland (Auckland)

(b) Bulls

(c) Lyttelton (Christchurch)

(d) Taihape

(e) Onehunga (Auckland)

(f) Northland (Wellington)

Figure 3.18: Typology C buildings two storey isolated

Alistair P. Russell

- 64 -

3.3. New Zealand URM Building Typologies

(a) Newmarket (Auckland)

(b) Bulls

(c) Kingsland (Auckland)

Figure 3.19: Typology C3 buildings two storey corner

Further photographic examples of Typology C buildings are shown in Appendix A.

- 65 -

Alistair P. Russell

Chapter 3. Architectural Characterisation

Typology D Two Storey Row Buildings Two storey row structures are the third most prevalent of the seven typologies identied. Like Typology B structures, they are common in commercial districts and on the main streets of small towns, but they are especially common in larger centres such as Auckland, Wellington and Christchurch. Many characteristics described for Typology B structures are also applicable to typology D buildings. Some Typology D structures are mostly uniform, with consistent oor, roof and parapet levels resulting in a structure which is essentially homogeneous but divided into sections. Other structures are very much heterogeneous and have the appearance of individual and distinctive buildings joined together. There can be many variations between adjoining tenancies, both in appearance and dimensions. For example, the height of the rst oor may be dierent between adjoining tenancies (which may signicantly contribute to the eects of pounding) and parapet levels and heights can also vary. Floor diaphragms may be concrete or timber (timber is more common) and the diaphragm seating issues identied for Typology C structures are also true for Typology D structures. Openings are similar to those in Typology C structures, and the number of windows and their spacing can vary between the adjoining parts of the overall structure. In these structures, each tenancy is long compared to its width, and L1 is usually twice as long as L2 (in Figure 3.21). Photographic examples are shown in Figure 3.22. Figure 3.21 and Table 3.8 show typical dimensions of Typology D structures.

Further photographic examples of Typology D buildings are shown in Appendix A.

Whilst structures which do not t the criteria of Typology A F classications are classed as Typology G buildings, there can be further non-conformity to the above framework, particularly where buildings of two typologies are joined. Figure 3.20 shows a one storey building adjoined to a two storey building.

Alistair P. Russell

- 66 -

3.3. New Zealand URM Building Typologies

Figure 3.20: One storey building adjoining a two storey building

Figure 3.21: Overall dimensions of a Typology D building

- 67 -

Alistair P. Russell

Chapter 3. Architectural Characterisation

Table 3.8: Typical dimensions of a Typology D building Dimension Lower bound structure (m) 8 4 4 1 0 4 3 3 3 0 0 1 0.8 Mean structure (m) 12 6 4 1.2 0.5 6 5 5 4 0 0 2 1 Upper bound structure (m) 16 8 6 2 1.5 7 5 6 5 2 2 2.5 1.5

L1 L2 L4 P1 P2 H1 H2 H3 F1 G1

Length of front ground oor building Width of residence Length of ground oor rear amenities block Height of parapet at front Height of parapet at rear Height from ground to bottom of second storey Height of front opening of each residence Height of second storey Width of front opening of each residence Depth of rear wall below front street level

W1 Height of windows above top of lower storey W2 Height of windows W3 Width of windows

Alistair P. Russell

- 68 -

3.3. New Zealand URM Building Typologies

(a) Central Christchurch

(b) Timaru

(c) Devonport (Auckland)

(d) Grey Lynn (Auckland)

(e) Ponsonby (Auckland)

(f) Central Auckland

Figure 3.22: Typology D building two storey row

- 69 -

Alistair P. Russell

Chapter 3. Architectural Characterisation

Typology E Three + Storey Isolated Buildings Typology E structures are isolated buildings of three storeys or more. There are too few four, ve and six or more storey buildings in New Zealand to warrant a separate class of building for each. These structures tend to be located in the larger centres, particularly Auckland, Wellington and Dunedin. They are situated mainly around older commercial areas and also close to the ports. The most important feature of these structures is that their front and back faces have multiple and consistently spaced window openings. This means that these faces tend to behave as masonry frames and the piers between windows can rock when subjected to lateral displacements. Large openings on the walls of the upper oors are uncommon. The side walls tend not to have any openings. The ground storey walls may be up to nine leaves thick, with each subsequent wall above usually decreasing in thickness. As described for Typology C structures, the oor and ceiling diaphragms often rest on the ledge resulting from the decrease in wall thickness. Again, the oor or ceiling diaphragms are typically of timber construction, with the same seating issues as identied above. For Typology E and F structures, the upper oors are often open inside, possibly with intermediate columns supporting the roof structure above. See Figure 3.23. Photographic examples of Typology E structures are shown in Figures 3.25 and 3.26. Figure 3.24 and Table 3.9 show typical dimensions of Typology E structures.

Further photographic examples of Typology E buildings are shown in Appendix A.

Alistair P. Russell

- 70 -

3.3. New Zealand URM Building Typologies

Figure 3.23: Open plan interior of upper storey in Typology E building

Figure 3.24: Overall dimensions of a Typology E building

- 71 -

Alistair P. Russell

Chapter 3. Architectural Characterisation

Table 3.9: Typical dimensions of a Typology E building Dimension Lower bound structure (m) 12 12 1 0 4 3 3 3 3 0 0 1 0.8 1 0.8 0 1 1 0.8 Mean structure (m) 20 16 1.2 0.5 6 5 5 5 4 0 0 2 1 2 1 0 2 2 1 Upper bound structure (m) 30 20 2 1.5 7 5 6 6 5 2 2 2.5 1.5 2.5 1.5 2 2.5 2.5 1.5

L1 L2 P1 P2 H1 H2 H3 F1 B1 G1

Length of front ground oor building Width of building Height of parapet at front Height of parapet at rear Height from ground to bottom of second storey Height of front opening of each residence Height of second storey Width of front opening of each residence Width of rear opening of each residence Depth of rear wall below front street level

W1 Height of windows above top of lower storey W2 Height of lower oor windows W3 Width of windows W4 Height of upper storey windows W5 Height of top storey windows W6 Height of rear windows above top of lower storey W7 Height of rear windows W8 Height of upper storey rear windows W9 Width of rear windows

Alistair P. Russell

- 72 -

3.3. New Zealand URM Building Typologies

(a) Central Christchurch

(b) Timaru

(c) Timaru

(d) Dunedin

Figure 3.25: Typology E buildings three + storey isolated

- 73 -

Alistair P. Russell

Chapter 3. Architectural Characterisation

(a) Newtown (Wellington)

(b) Central Wellington

(c) Central Wellington

Figure 3.26: Typology E buildings three + storey isolated

Alistair P. Russell

- 74 -

3.3. New Zealand URM Building Typologies

Typology F Three + Storey Row Buildings These structures are three and higher storey buildings which, similar to Typology B and D, are row structures. These structures were often built as warehouses, granaries or storage facilities and are common both in industrial and commercial districts, especially close to a port (or historic port), for example parts of Timaru and in Aucklands waterfront. There can be a blurred distinction between Typology E and F, but generally Typology F structures are very long and may form a whole block of a street. Most characteristics of these buildings are similar to those of Typology E buildings, in particular, Typology F structures can be considered as frame type structures. The main dierence is that neighbouring (joined) residences may have dierent oor and overall heights resulting in pounding or lateral displacement incompatibility issues. This type of building might be owned by a single owner, but there are usually many dierent tenancies, both between oors and within oors. As noted in Section 3.3.1 this needs to be taken into consideration when designing a retrot for the overall structure. Typical dimensions of Typology F structures are shown in Figure 3.27 and Table 3.10 and photographic examples are shown in Figure 3.28. Figure 3.27 shows only the back of a single equivalent component Typology E structure.

Further photographic examples of Typology F buildings are shown in Appendix A.

- 75 -

Alistair P. Russell

Chapter 3. Architectural Characterisation

Figure 3.27: Overall dimensions of a Typology F building

Alistair P. Russell

- 76 -

3.3. New Zealand URM Building Typologies

Table 3.10: Typical dimensions of a Typology F building Dimension Lower bound structure (m) 12 12 1 0 4 3 3 3 3 0 0 1 0.8 1 0.8 0 1 1 0.8 Mean structure (m) 20 16 1.2 0.5 6 5 5 5 4 0 0 2 1 2 1 0 2 2 1 Upper bound structure (m) 30 20 2 1.5 7 5 6 6 5 2 2 2.5 1.5 2.5 1.5 2 2.5 2.5 1.5

L1 L2 P1 P2 H1 H2 H3 F1 B1 G1

Length of front ground oor building Width of building Height of parapet at front Height of parapet at rear Height from ground to bottom of second storey Height of front opening of each residence Height of second storey Width of front opening of each residence Width of rear opening of each residence Depth of rear wall below front street level

W1 Height of windows above top of lower storey W2 Height of lower oor windows W3 Width of windows W4 Height of upper storey windows W5 Height of top storey windows W6 Height of rear windows above top of lower storey W7 Height of rear windows W8 Height of upper storey rear windows W9 Width of rear windows

- 77 -

Alistair P. Russell

Chapter 3. Architectural Characterisation

(a) Central Christchurch

(b) Central Christchurch

(c) Central Dunedin

(d) Central Wellington

(e) Central Auckland

(f) Central Auckland

Figure 3.28: Typology F buildings three + storey row

Alistair P. Russell

- 78 -

3.3. New Zealand URM Building Typologies

Typology G Religious, Institutional and Industrial Buildings Typology G structures are characterised by structures for which a generic form is unrealistic, and those URM buildings which do not have a simple or uniform ground footprint. These are distributed all around the country. The list of typologies is comprehensive and any monumental or unique structure which does not t into Typologies A F will be a Typology G structure. It is not the purpose of these classications to dene buildings by their use, but Typology G buildings tend to be churches, warehouses and factories, or civil buildings such as a town hall, ferry building or post oce. Because of this, Typology G has been further divided into G1 for religious buildings (Figure 3.29(d)), G2 for warehouses and factories with large tall sides and large open spaces inside (Figure 3.29(e)), and G3 for institutional buildings, such as ferry terminals and train stations (Figure A.17(c)). For Typology G structures unique analyses are necessary on a case-by-case basis. Photographic examples are shown in Figure 3.29.

Further photographic examples of Typology G buildings are shown in Appendix A.

- 79 -

Alistair P. Russell

Chapter 3. Architectural Characterisation

(a) Lincoln (Canterbury)

(b) Dunedin

(c) Oamaru

(d) Seatoun (Wellington)

(e) Palmerston North

Figure 3.29: Typology G building examples religious, institutional and industrial

Alistair P. Russell

- 80 -

3.4. International Comparisons with New Zealand Typologies

3.4

International Comparisons with New Zealand Typologies

URM construction in New Zealand shares many similarities with practices elsewhere in the world. This is true particularly for British Colonies settled contemporaneously with New Zealand, and also for The United States. European and Asian construction practices extend back much further, and other types of masonry construction, such as stone and rubble, are much more prevalent there than in New Zealand.

Australia and New Zealand are considered to have had very similar building practices up until the 1931 Hawkes Bay earthquake (Page, 1996; Russell et al., 2006). URM construction typical of this era in Australia consisted of clay brick walls, supporting timber oor systems and timber roof trusses with metal or clay brick roong. The brick walls were typically cavity walls, with no rubble ll, and were supposed to be connected with metal wall ties. The bricks themselves were usually pressed bricks of a regular size. The timber oor/roof systems typically were supported with bearing supports on the inner leaf of the cavity wall without any signicant positive connection to the walls.

Comparison of New Zealand URM structures with those in California is important because of the similarities in terms of construction types, histories and seismic context. In both places, most URM buildings are pre-1935. Outside the central business districts of the main cities, most are one or two storeys high and typically have timber oors and roof framing and a two or three leaf wall construction (Blaikie and Spurr, 1992). It has been found that the use of unreinforced masonry walls with timber diaphragms for oors or roofs was commonplace in pre-1945 construction in the United States (Bruneau, 1994a; Peralta et al., 2004). Timber diaphragms typically consist of three components; sheathing, framing and chords. It was found that in California buildings constructed before 1945, sheathing was made of straight members, nailed onto the framing, rather than plywood, as would be expected in current practice. Occasionally iron wall anchors were used to connect the diaphragm to the wall, but only on about every 4th joist. Usually joist-to-wall anchors were designed without taking seismic forces into consideration.
- 81 Alistair P. Russell

Chapter 3. Architectural Characterisation

Figure 3.30 shows examples of typologies in Australia and South Africa, similar to those identied for New Zealand.

(a) Typology C structure Cape Town (South (b) Typology D structure Cape Town (South Africa), near historic British naval base Africa), near historic British naval base

(c) Typology D structure in Newcastle (Australia), (d) Typology B structure in Newcastle (Australia), (British penal colony) (British penal colony)

Figure 3.30: International comparisons with New Zealand typologies

3.5

Supplementary Characteristics of New Zealand URM

New Zealands masonry construction heritage is comparatively young, spanning from 1833 until the present time a period of less than 200 years (Ingham, 2008). Consequently, a study of New Zealands masonry building stock has a narrow scope in comparison with
Alistair P. Russell

- 82 -

3.5. Supplementary Characteristics of New Zealand URM

international norms (see for instance Binda et al., 2005; Louren co et al., 2006; Magenes, 2006). This comparatively narrow time period has the advantage of facilitating the documentation and reporting of New Zealand masonry construction practice with a greater degree of accuracy than is often possible in countries with an older and more diverse history of masonry construction (Binda, 2006b).

3.5.1

Bond Pattern

Bond pattern determines how bricks in a masonry wall are connected. This has a signicant eect on the strength of the wall and how the wall acts together as a unit. Stretcher units are bricks laid in the plane of the wall, and header units are laid across the wall, joining leaves together. In cross-section, a wall that is three bricks thick is a three leaf wall. Each leaf needs to be adequately connected with header bricks at appropriate intervals. A course is row of bricks along the length of the wall. Thus, a wall may be three leaves thick, and 30 courses high.

New Zealand URM buildings are mainly constructed with either Common bond or English bond. English bond has alternating header and stretcher courses, as shown in Figure 3.32. Common bond (sometimes referred to as American bond or English Garden Wall bond) has layers of stretchers with headers every 3 6 courses, and is shown in Figure 3.33. Running bond (only stretcher courses) often suggests a cavity wall and is shown in Figure 3.34. Flemish bond (alternating headers and stretchers in every course) is the least common and is often found in New Zealand between openings on an upper storey, for example on piers between windows. Flemish bond is shown in Figure 3.35. Common bond is the most frequently occurring bond pattern in New Zealand, but English bond also appears in New Zealand buildings often, especially in the bottom (ground) storey. Common bond can vary with headers at dierent levels in dierent buildings, and sometimes there is variation within a building. For example, at the bottom of the wall the headers may be every second course, but at the top of the third storey wall they may be every 4th course. See Figures 3.36(a) and 3.36(b): both photos are from the same building wall. Header courses may be irregular and made to t in at ends of walls and around drainpipes with half widths and other cut bricks, see Figure 3.31.
- 83 Alistair P. Russell

Chapter 3. Architectural Characterisation

Figure 3.31: Uneven bricks made to t

(a)

(b)

Figure 3.32: English bond pattern

(a)

(b)

Figure 3.33: Common (or American or English Garden Wall) bond pattern

Alistair P. Russell

- 84 -

3.5. Supplementary Characteristics of New Zealand URM

(a)

(b)

Figure 3.34: Running bond pattern

(a)

(b)

Figure 3.35: Flemish bond pattern

(a) Ground storey, header bricks every second (b) Second storey, header bricks every fourth course course

Figure 3.36: Bond pattern varying over height of a single wall

- 85 -

Alistair P. Russell

Chapter 3. Architectural Characterisation

3.5.2

Wall Height and Thickness

Most bearing walls in New Zealand are generally three leaves thick (330 mm), and parapets are two leaves thick (220 mm) with the walls increasing in thickness down the height of the building (often reducing by one leaf every 2 storeys) (Oliver, 2007). Typical storey heights generally range from approximately 3.2 m to 6.0 m (see details on individual typologies above). The Architects and Builders Pocket-Book (Kidder, 1905) from California gives guidance on wall thickness for mercantile buildings in large cities. As noted in Section 3.4, New Zealand URM construction is thought to have been similar to other British colonies at the time, and also similar to URM construction practice in the United States, especially California. In 1905 brick buildings in San Francisco were permitted to be up to six storeys high, and the top storey was prescribed to have walls 13 in. thick, the next three storeys down were to be 17 in. thick, and the next two were to be 21 in. thick. If a structure was only three storeys high, the bottom two storeys would have 17 in. thick walls and the top storey walls would be 13 in. thick (Kidder, 1905). This prescription corresponds approximately to ve or six leaf thick walls at the bottom, and three leaf walls at the top of a Typology E or F structure, which has also been observed in New Zealand. Taller buildings (typically those of Typology C, D, E & F) have tall slender window openings on the front side usually facing the street. Above the windows, lintels are generally formed with perpendicular bricks extending from the face of the wall. More sophisticated fa cades may have other decorative features around the windows and lintels. Slender wall sections between windows can act as piers which rock during lateral loading. This is seen to be a more favourable failure mode than shear failure or sliding, as residual displacements of the walls are small. But this can lead to signicant non-structural component damage from the large displacements. See Chapter 1.

3.5.3

Age

The age of existing New Zealand URM buildings is dealt with more fully in Chapter 4. URM was prohibited as a structural material in most parts of the country in 1965 with the introduction of NZS 1900 (New Zealand Standards Institute, 1965). Most URM buildings
Alistair P. Russell

- 86 -

3.5. Supplementary Characteristics of New Zealand URM

were built between about 1880 and 1930. As settlement grew in the late 19th Century, many new buildings were constructed, and the Hawkes Bay earthquake in 1931 (Dowrick, 1998) was a turning point after which few URM buildings were built. Conveniently, many URM buildings display their age on the front fa cade, as shown in Figure 3.37.

(a)

(b)

(c)

(d)

Figure 3.37: Date of construction on fa cade

3.5.4

Materials

URM buildings were constructed using bricks generally of a uniform size, which is dierent from the dimensions of both modern concrete blocks and modern clay bricks. Bricks manufactured in the early 20th Century were solid (uncored) and sometimes had frogging on one face (an imprint often with the manufacturers stamp). Typical dimensions of early New Zealand bricks are shown in Figure 3.38, but as can be seen in Figure 3.39,
- 87 Alistair P. Russell

Chapter 3. Architectural Characterisation

not all bricks were laid to a uniform quality. Bricks can be of varying qualities now due to deterioration, depending especially on proximity to sea air. Even some inner-city walls display bricks which are signicantly eroded, as shown in Figure 3.40. The condition of bricks does not seem to correspond with the colour of the bricks.

Many buildings which look like they could be constructed of URM based on characteristics such as geometry, conguration, occupancies, age and location, but appear to be made of concrete because of the nature of their external nish are often found to be brick on closer inspection, because of a concrete plaster over the bricks (see Figure 3.19(c)). This plaster can be up to 20 mm thick, although it often varies in thickness because the bricks did not form a smooth surface. In some places the plaster is eroded.

Figure 3.38: Typical brick dimensions In binding masonry walls to the foundations, dierent mortar was used from that in the walls themselves. A specication from a URM residential building in central Auckland constructed in 1919 states that ground mortar is to be two parts clean red scoria ash, one part sand, and one part hydraulic lime, nely ground together in a mortar mill, (OConnor, 1919). For mortars used between bricks, it is thought the constituent materials that were used varied considerably. Some mortars were lime based, and others
Alistair P. Russell

- 88 -

3.5. Supplementary Characteristics of New Zealand URM

Figure 3.39: Broken bricks

Figure 3.40: Eroded bricks in centre of wall

were cement based. Mortars varied from 1:3 lime:sand to 1:2:9 cement:lime:sand, to 1:2 cement:sand, by volume. Near the sea shore, poorly graded beach sand was usually used (sometimes with entire shells used in the mix, see Figure 3.41), but inland ner river sand was generally used. Lime was used as a binder in the mortar, and also to improve workability when using cement as the main binder. Cement based mortar tends to be in better condition now than lime based mortars, due to the leaching of lime particles from the mixture. This often leaves a powdery sand with no bond strength and generally occurs where water inltration has been a problem.

- 89 -

Alistair P. Russell

Chapter 3. Architectural Characterisation

Figure 3.41: Sea shells in mortar The performance of mortar in existing buildings is extremely variable. Some masonry exhibits considerable strength in terms of bond and cohesion between bricks and mortar, while others exhibit very little tensile strength at the brick-mortar interface. Mortar condition also varies according to its location in the building. Internal walls not exposed to weathering are frequently in better condition than walls exposed to external elements. Mortar in buildings close to the sea is often eroded due to salt carried from sea breezes, and also from salt in the sand from the original mortar construction.

Workmanship and mortar quality varied widely in early NZ URM construction. Public works buildings generally used good quality mortar, often cement based, while private buildings tended to use cheaper and more easily available materials for mortar (Stil, 2007). Today, the quality of workmanship in URM buildings also shows large variability. When comparing non-structural visible fa cades with non-visible internal structural walls, it is often evident that the structural components were built with lower quality control than parts which were seen. Internal walls are sometimes painted or plastered, but cracks and movement in the walls shows the often low quality of workmanship. Some URM buildings were built with great care taken, and are still in good condition today. A residential building specication in 1919 shows that there were some quality controls in place; Bricks throughout the structure are to be sound, square, well burnt, and to have
Alistair P. Russell

- 90 -

3.6. Conclusions

of an even shape and colour, and be obtained form an approved yard, (OConnor, 1919). These specications were left to the bricklayer to meet, but it is not clear how strictly quality control was met or enforced.

3.6

Conclusions

Characterisation of the URM building stock into typologies is one of the key parameters for dening and understanding the nature of URM buildings in New Zealand. The overall conguration of a building inuences its performance in an earthquake. Seven typologies have been identied to categorise congurations in the New Zealand URM building stock. Separations between typologies are made on the basis of building height and the geometry of the buildings ground footprint. When URM was commonly used in New Zealand in the late 19th Century and early 20th Century, construction practices were similar to other countries of British origin, such as Australia and The United States. New Zealand also shares common architectural characteristics and building typologies with some European countries, particularly Italy. Assessment and analysis of the structural performance of buildings within the seven identied typologies will enable targeted and cost-eective retrot solutions to be implemented for the retention of New Zealands heritage URM buildings.

- 91 -

Alistair P. Russell

Chapter 3. Architectural Characterisation

Alistair P. Russell

- 92 -

CHAPTER 4

New Zealand URM Building Stock

In addition to analysing the nature of individual buildings and classifying the national building stock into typologies, as discussed in Chapter 3, this chapter provides further background on the overall nature of the URM building stock in New Zealand. An estimate is given of the number of URM buildings in New Zealand, and their nancial value, both collectively and individually, as well as an estimate of their seismic vulnerability.

Background on the building evaluation procedures utilised in this chapter is sourced from NZSEE (2006), and similarly much of the legislative context for the work reported herein is taken from the Building Act, 2004 (New Zealand Parliament, 2004).

4.1

Introduction

For the purpose of understanding the magnitude and nature of the cumulative seismic risk posed by URM buildings in New Zealand it is useful to consider their prevalence and national distribution. Two independent methods with dierent primary data sources were used to estimate the number of URM buildings in existence throughout New Zealand
- 93 Alistair P. Russell

Chapter 4. New Zealand URM Building Stock

in 2010. Data from Auckland City Council, Wellington City Council and Christchurch City Council, in conjunction with historic population data, were utilised to determine the distribution of URM buildings throughout the country, with their associated construction dates. This analysis is presented in Section 4.2. The purpose of this procedure was to determine not only the aggregate number of URM buildings, but also an indication of their locations. The purpose of the approach taken in Section 4.3 was to evaluate the nancial value of existing URM buildings, using data provided from Quotable Value New Zealand Ltd (QV Ltd). This method also provided an estimate of the number of URM buildings. The validity of both approaches was conrmed with the close agreement in the overall aggregate number of URM buildings in existence in New Zealand. The rst method suggested there are 3879 URM buildings in New Zealand, while the second method suggested there are 3589 URM buildings nationwide. Taking the mean of both values indicates that there are approximately 3734 URM buildings in total existing in New Zealand in 2010.

4.2

Estimation of URM Population and Distribution

Several sources of data were utilised for estimating the number of URM buildings in existence throughout the country: the ocial population data of New Zealand between 1900 and 1940 (Census and Statistics Oce, 1950), a survey of potentially earthquake prone commercial buildings in Auckland City, conducted by Auckland City Council in 2008, and data provided by Wellington City Council and Christchurch City Council.

In surveying potentially earthquake prone commercial buildings in Auckland City, a total of 1335 buildings were identied to have been constructed before 1940. Although buildings with a construction date up to and including 2007 were surveyed, very few URM structures were found to have been built in Auckland City after 1940. Therefore, only pre-1940 structures were considered. Of the 1335 buildings, 28.9% were URM, 35.3% were timber, 16.3% were comprised of reinforced concrete frame and brick inll, 1.1% were reinforced masonry, 17.8% were reinforced concrete frame or shear wall buildings and 0.6%
Alistair P. Russell

- 94 -

4.2. Estimation of URM Population and Distribution

were moment resisting steel or braced steel buildings. Using the associated construction date of each building the total sample was grouped according to decade. Pre-1900 was considered as a single grouping. Table 4.1 shows the number of buildings identied in the survey according to their construction date.

Table 4.1: Auckland City pre-1940 potentially earthquake prone buildings Pre 1900 1910 1920 1930 Total 1900 1910 1920 1930 1940 URM Timber Brick inll Reinforced masonry Reinforced concrete Steel Total 6 3 4 0 1 0 15 24 21 13 0 7 0 65 16 16 4 0 7 0 45 277 341 123 10 152 5 907 63 90 74 5 71 3 304 385 417 217 15 238 8 1335 Percentage 28.9% 35.3% 16.3% 1.1% 17.8% 0.6% 100%

In order to estimate the number of URM buildings in other parts of the country, the data from Auckland City Council were extrapolated using ocial population data. In the late 19th and early 20th Century, New Zealand was divided into the following provinces: Auckland, Taranaki, Hawkes Bay, Wellington, Marlborough, Nelson, Canterbury and Otago-and-Southland. Auckland Province was made up of the area of the North Island from Taupo and north (everywhere which currently celebrates Auckland Anniversary Day) (Census and Statistics Oce, 1950). Consequently, the area over which Auckland City Council has jurisdiction in 2010 is only a part of the former Auckland Province, and the current boundaries of this jurisdiction are equivalent to that of the Eden County up until 1940. This county historically included the boroughs of Auckland City, Mt Albert, Mt Eden, Newmarket, Parnell, Onehunga, Grey Lynn, One Tree Hill, and also Ellerslie Town District. The proportion of the population of the historic Auckland Province which is made up by the current Auckland City was found using the population data from ocial New Zealand Year Books (Census and Statistics Oce, 1950). The average population of Auckland City and other parts of Auckland Province are shown in Figure 4.1.

- 95 -

Alistair P. Russell

Chapter 4. New Zealand URM Building Stock

Other Auckland City

100%

Percentage

80% 60% 40% 20% 0% pre-1900

1900 1910 109,513 84,068

1910 1920 166,260 112,096

1920 1930 245,717 147,922

1930 1940 336,570 180,297

Other Auckland City

108,660 67,278

Decade

Figure 4.1: Proportion of population in the former Auckland Province living in the equivalent current Auckland City Using the same proportional relationships shown in Figure 4.1, the number of existing URM buildings in the historic Auckland Province were estimated based on the number of existing URM buildings in Auckland City. For example, in the decade 1900 1910, Auckland City made up 43% of the population of Auckland Province. It is assumed that building prevalence was approximately proportional to population and that the rate of building demolition has been uniform throughout the former Auckland Province. There are 24 URM buildings identied from that decade now existing in Auckland City, and assuming these also make up 43% of the total number of buildings in the historic Auckland Province, then there are 55 existing URM buildings in the whole of the equivalent Auckland Province today which were built between 1900 and 1910. Similarly, an indicative URM-buildings-per-capita ratio is determined. These data are summarised in Table 4.2, clearly showing that the majority of URM buildings were constructed in the decade 1920 1930.

In addition to the data provided from Auckland City Council and extrapolated to estimate the number of URM buildings in the historic Auckland Province, similar data were provided by Wellington City Council and Christchurch City Council. This presented an assessment of the prevalence of URM buildings in the historic Wellington Province and Canterbury Province respectively. The data were provided in such a way to enable a
Alistair P. Russell

- 96 -

4.2. Estimation of URM Population and Distribution

Table 4.2: Population data and URM buildings for Auckland City and Auckland Province Pre 1900 Population of former Auckland Province Population of equivalent current Auckland City Proportion Auckland City/Province Auckland City URM buildings Auckland Province URM buildings URM buildings per 100,000 people 1900 1910 1910 1920 1920 1930 1930 1940

175,938 193,581 278,357 393,639 516,886 67,278 38.2% 6 16 9.1 84,068 43.0% 24 55 28.4 112,096 147,922 180,297 41.1% 16 40 14.4 37.5% 277 737 187.2 35.2% 63 178 34.4

breakdown of construction dates in the same age brackets as for the Auckland Province. Based on ocial provincial populations of the time, the number of URM buildings currently remaining in the historic provinces of Taranaki, Marlborough, Nelson and Westland were also estimated assuming the same ratio of URM buildings per 100,000 people as in Auckland Province. There is no evidence to suggest that the ratio of URM buildings per 100,000 people in Auckland is not valid for these provinces. Based on evidence provided in Hopkins (2009), it was considered inappropriate to assume a similar buildings per capita ratio as in Auckland for the remaining provinces of Hawkes Bay and Otago-andSouthland. When legislative guidance was introduced in 1968 (New Zealand Parliament, 1968) for assessing and upgrading earthquake prone buildings, Auckland and Wellington City Councils took a strong interest in strengthening URM buildings whilst Christchurch and Dunedin City Councils took a much more passive approach in implementing the legislation. Consequently, the rate of seismic retrot and/or demolition and reconstruction in Auckland and Wellington was signicantly dierent from that in Dunedin and Christchurch1 (see Section 2.4 for further details). Dunedin is the largest city in the former Otago-and-Southland Province and its rate of redevelopment was assumed to be characteristic of the whole province. Consequently, the number of buildings remaining
Correspondence with Dunedin City Council conrmed that data on the number of URM buildings is not recorded by the Council, but anecdotally it was suggested that gold mining in the late 19th Century throughout Otago generated signicant development and construction, and that many of the URM buildings constructed then are still in use.
1

- 97 -

Alistair P. Russell

Chapter 4. New Zealand URM Building Stock

in Otago-and-Southland was estimated using the buildings per 100,000 people ratio of Canterbury. The 1931 M7.1 earthquake in Hawkes Bay destroyed a signicant number of URM buildings in the Hawkes Bay Province, especially in Napier. As a consequence of this and the resulting awareness of the vulnerability of URM buildings, the number of remaining buildings in Hawkes Bay can be expected to be less than what would be estimated using the relationships outlined above. Nevertheless, there are no data available on the actual number of URM buildings in Hawkes Bay, and because of this, the ratio of URM buildings per 100,000 people in Hawkes Bay was estimated to be half of the corresponding value for Auckland. The estimated number of existing URM buildings in each province is shown in Table 4.3 and in Figure 4.2, and the construction date of URM buildings nationwide is shown in Figure 4.3, grouped according to the rst year in each decade. This again shows that the majority of existing URM buildings nationwide derive from the decade 1920 1930.

Otago and Southland 22%

Auckland 27%

Canterbury 22% Wellington 18% Westland 1% Nelson 3%

Taranaki 4% Hawke's Bay 2%

Marlborough 1%

Figure 4.2: Provincial estimated populations of URM buildings It is acknowledged that the data presented above are useful primarily as an initial estimation only and may not accurately represent the number of URM buildings in other regions outside of Auckland, especially in smaller towns. The number of buildings from a particular decade in Auckland captures only those buildings which still exist, rather than
Alistair P. Russell

- 98 -

4.2. Estimation of URM Population and Distribution

Table 4.3: Provincial populations and estimated number of existing URM buildings Province Auckland Taranaki Hawkes Bay Wellington Marlborough Nelson Westland Canterbury Otago and Southland Total Population URM Population URM Population URM Population URM Population URM Population URM Population URM Population URM Population URM Pre 1900 1900 1910 1910 1920 1920 1930 1930 1940 Total

175,938 193,581 278,357 393,639 516,886 16 55 40 737 178 34,486 3 37,139 2 45,973 11 46,906 6 48,546 7 51,569 5 63,273 118 65,037 72 76,968 25 77,652 0

1026 164 86 677 47 132 41 852 855 3879

132,420 189,481 199,094 261,151 316,446 27 127 169 243 111 13,499 1 33,142 3 15,042 1 15,177 3 45,493 10 15,194 3 15,985 2 48,463 7 15,714 2 18,053 34 49,153 91 14,655 27 19,149 6 59,481 19 18,676 6

145,058 166,275 173,443 206,462 234,399 7 190 211 233 211 174,664 156,688 191,130 206,835 224,069 8 179 233 233 202

Number URM Buildings

2000

50% 40% 30%

1500

1000
20%

500

10% 0%

0 pre-1900 1900 1910 1920 1930

Construction Decade

Figure 4.3: Construction date of URM buildings in NZ

all the buildings which were constructed in that time period, and the rate of demolition and redevelopment in Auckland City may not be representative of the comparable rate
- 99 Alistair P. Russell

Proportion of URM building stock

Chapter 4. New Zealand URM Building Stock

in other parts of the country. Whereas in Auckland, economic factors may have provided a stimulus for demolition of older URM structures and development of newer structures, this may have been not true in smaller towns. Smaller cities such as Whanganui, Timaru and Oamaru did not receive equivalent levels of investment and development in the 1960s and 1970s for economic reasons, and consequently many old buildings which would have otherwise been demolished in that time period still exist now (McKinnon, 2008). Moreover, legislation governing the seismic performance of existing buildings may have resulted in dierent rates of development (see Section 2.4). For example, Blenheim is in a higher seismic zone (Z = 0.33) than New Plymouth (Z = 0.18) and if a building in Blenheim which was determined to be earthquake risk and subsequently demolished was instead situated in New Plymouth, because of the lower seismicity, it may have been found to not be earthquake risk. Finally, this is not an estimation of the number of earthquake prone buildings in New Zealand, apart from the inference that many URM buildings are likely to meet the criteria of being earthquake prone.

4.3

Estimation of URM Population and Value

In addition to the above estimate of the number of URM buildings in New Zealand, data on the New Zealand building stock were obtained from Property IQ, a part of Quotable Value New Zealand Ltd (QV Ltd), which is a valuation and property information company in New Zealand. QV collects building information and conducts building valuations for rating purposes for most New Zealand Territorial Authorities. In the council valuation data, the building material and age (decade), among other data elements, is recorded. The building material refers to the wall cladding and is not a comment on the load carrying materials of the structure. It was assumed that no URM buildings were constructed in New Zealand after 1950 (Stacpoole and Beaven, 1972), and that buildings with a brick veneer but other materials for the load bearing parts of the structure (for example, timber frame buildings with a brick veneer) are recorded as mixed materials in the database. All entries for buildings constructed in New Zealand before 1950 and with brick recorded as the cladding description in the QV database were extracted. While it is acknowledged that a cladding description recorded as brick can include brick, brick veneer, adobe and
Alistair P. Russell

- 100 -

4.3. Estimation of URM Population and Value

rammed earth as the material type, it was considered that such an extraction of data would be a legitimate reection of the URM building stock in New Zealand. The building data are recorded in three groupings as follows: Separate buildings, which have a single age and a single rating valuation for the whole building. Parent/child entries, which are buildings with multiple ownership records. For example, a building with oces and apartments above will have one parent and multiple children for each use. In this case only the parent record was analysed. Orphans, which are records for buildings where the children of the parent structure are recorded but the territorial authority has stopped maintaining the parent record. In this case the orphans have the same roll and valuation number so can be grouped to determine the parent record. These records were analysed according to construction date, building height and nancial value. Table 4.4 shows the decade in which each URM structure was built. Brick buildings with mixed age are entered on the QV database as pre-1950, but their exact age is indeterminate from the data recorded. The number of URM buildings with a conrmed construction date are shown in Figure 4.4, and are grouped according to the rst year in each decade.

Table 4.4: Number of URM buildings from QV according to construction decade Decade 1870 1880 1880 1890 1890 1900 1900 1910 1910 1920 1920 1930 1930 1940 1940 1950 Mixed Total
- 101 -

URM Buildings 43 23 71 469 646 878 514 218 726 3589

Alistair P. Russell

Chapter 4. New Zealand URM Building Stock

1000 900

Number URM Buildings

800 700 600 500 400 300 200 100 0 1870 1880 1890 1900 1910 1920 1930 1940

Construction Decade

Figure 4.4: Number of URM buildings from QV according to construction date Figure 4.4 clearly shows a trend where the number of URM buildings initially increased until the end of the 1920s, and subsequently declined. This follows the increasing rate of European immigration and associated infrastructure development in New Zealand in the early 20th Century, until the 1931 M7.8 Hawkes Bay earthquake, after which URM was no longer considered a favourable building material. Further background to the history of URM buildings is addressed in Sections 2.3 and 2.4.

Table 4.5 summarises the number, total value and average value of URM buildings according to storey height. All building values are given in terms of New Zealand Dollars in 2008. In the QV database the Building Floor Area and the Building Site Cover are recorded, and an estimate of the number of storeys can be obtained by dividing the Building Floor Area by the Building Site Cover. The number of storeys is not directly recorded. The Building Floor Area is the useable oor area and does not include the roof area. In some entries, either the Building Floor Area or the Building Site Cover is not recorded, and in this case the number of storeys is shown as N/A. It could not be determined from the data whether the structure was an isolated or row structure, so interrogation of the data according to the typologies presented in Chapter 3 was not possible. It could be suggested that parent buildings with multiple children may constitute a row structure with multiple occupancies inside the structure, but this only considers ownership of the structure and does not take into account the shape of the building footprint. It must also be noted that, using this denition, one storey buildings could not with certainty be classied as typology A or B, according to Chapter 3. See, for example, Figures 3.29(d)
Alistair P. Russell

- 102 -

4.3. Estimation of URM Population and Value

and 3.29(e), where the Building Floor Area is the same as the Building Site Cover, but the structures are clearly not single storey.

For all entries the Land Value and the Improvements Value are recorded. The value of the improvements refers to the value of any structure on the site, and is independent of the value of the land on which the building is situated. All data entries were revised between July 2005 and September 2008, and all buildings are valued in New Zealand Dollars (NZ$) as at the date of valuation.

Table 4.5: URM building stock according to storey height Height 1 storey 2 storey 3 storey 4 storey 5+ storey N/A Total Number 2526 564 163 46 18 272 3589 Total Value $778,000,000 $256,000,000 $134,000,000 $54,000,000 $20,000,000 $259,000,000 $1,501,000,000 Average Value $308,000 $454,000 $822,000 $1,171,000 $1,108,000 $953,000

The aggregate of the nancial value and the average value of each building height are shown in Figures 4.6(a) and 4.6(b). There was a small number of buildings where the number of storeys was estimated as 5 or 6, and these were grouped as 5+ storeys. There is an apparent anomaly that 5+ storey buildings have on average a slightly lower nancial value than 4 storey buildings. This dierence is minor, and occurs because the nancial value of buildings greater than 3 storeys in height is not directly proportional to the number of oor levels and other factors also inuence the value on an individual building basis. Moreover, there are comparatively few 4 and 5+ storey buildings, such that the average building value is more strongly inuenced by outlier individual building values. This observation is further evidence that the typologies identied in Chapter 3 are valid, where buildings are 3 storeys or greater in height are grouped as a single typoogy.

- 103 -

Alistair P. Russell

Chapter 4. New Zealand URM Building Stock

3000

Number URM Buildings

2500 2000 1500 1000 500 0 1 storey 2 storey 3 storey 4 storey 5 + storey N/A

Building Height

Figure 4.5: Number of URM buildings according to storey height

$900,000,000 $800,000,000
$1,400,000

Rated Value (NZ$)

$700,000,000 $600,000,000 $500,000,000 $400,000,000 $300,000,000 $200,000,000 $100,000,000 $1 storey 2 storey 3 storey 4 storey 5 + storey N/A

Rated Value (NZ$)

$1,200,000 $1,000,000 $800,000 $600,000 $400,000 $200,000 $1 storey 2 storey 3 storey 4 storey 5 + storey N/A

Building Height

Building Height

(a) Total value

(b) Average value

Figure 4.6: Valuation of URM building stock according to height As shown in Table 4.5, New Zealand has in existence nearly 3600 URM buildings, with a collective nancial value (in 2009) of approximately NZ$1.5 billion. The majority of the URM building stock consists of one-storey buildings, with the caveat on how this data was determined being noted above. It is clear from Figure 4.6 that as the building height increases, the average value of the building also increases. Because the number of one-storey buildings is by far the greatest, the aggregate value of that building height is also the greatest, despite the comparatively low average value of each structure. Thus it appears that the New Zealand URM building stock is largely made up of smaller, lower value buildings, and that in particular, the combination of one- and two-storey URM buildings constitutes 86% of the entire New Zealand URM building stock (see Figure 4.7). One-storey buildings make up 70% of all buildings, but only 51% of the total value of all URM buildings, and conversely structures taller than one-storey make up only 30% of the number of buildings, but 49% of the value. The average value of the structure should determine the investment associated with seismic assessment and retrot, and
Alistair P. Russell

- 104 -

4.4. Estimated Vulnerability of URM Buildings

3+ storey 14%

1 and 2 storey 86%

Figure 4.7: Number of low rise (1 and 2 storey) buildings as a proportion of all New Zealand URM buildings thus it may be concluded that while there are comparatively fewer larger buildings, the investment associated with their seismic assessment and retrot can be justiably higher. Similarly, low rise buildings may require simplied and repeatable assessment methods and retrot interventions in order for the exercise to be nancially viable.

Finally, it must be recognised that many buildings have a worth greater than their nancial valuation, including an architectural, historic or heritage value to the community, which can be dicult to quantify (Goodwin, 2008, 2009).

4.4

Estimated Vulnerability of URM Buildings

Once the number of URM buildings and their approximate regional distribution is determined, this information can then be extended to evaluate the expected vulnerability of the URM building population. As part of the NZSEE Guidelines Assessment and Improvement of the Structural Performance of Buildings in Earthquakes (2006), an initial evaluation procedure (IEP) is provided as a coarse screening method for determining a buildings expected performance in an earthquake. The purpose of the IEP is to make an initial assessment of the performance of an existing building against the standard required for a new building, i.e., to determine the Percentage New Building Standard (%NBS).
- 105 Alistair P. Russell

Chapter 4. New Zealand URM Building Stock

A %NBS of 33 or less means that the building is assessed as potentially earthquake prone in terms of the Building Act (2004) and a more detailed evaluation will then typically be required. A %NBS of greater than 33 means that the building is regarded as outside the requirements of the Act, and no further action will be required by law, although it may still be considered as representing an unacceptable risk and seismic improvement may still be recommended (dened by NZSEE as potentially earthquake risk). A %NBS of 67 or greater means that the building is not considered to be a signicant earthquake risk. NZSEE (2006) notes that A %NBS of 33 or less should only be taken as an indication that the building is potentially earthquake prone and a detailed assessment may well show that a higher level of performance is achievable. The slight skewing of the IEP towards conservatism should give condence that a building assessed as having a %NBS greater than 33 by the IEP is unlikely to be shown, by later detailed assessment, to be earthquake prone (see NZSEE, 2006, chap. 3).

In terms of the IEP, the %NBS of a building is determined by multiplying the Performance Achievement Ratio (PAR) by the Baseline %NBS (%NBSb ). (Full details can be found in NZSEE, 2006, chap. 3). The %NBSb is determined as the product of %NBSnom and factors accounting for Near Fault Eects (A), Hazard Level (B), Building Importance Level (C), Ductility (D) and Structural Importance (E). The %NBSnom depends on soil type and building period. For determining the %NBSb for URM buildings, the following assumptions can reasonably be made in the context of the IEP (see Stevens and Wheeler, 2008): The construction date is pre-1935 The period T 0.4 s The ductility factor, = 1.5 Most URM buildings have an importance level 2 Very soft soils can be excluded. Additionally, for the purpose of estimating the vulnerability of URM buildings on a nationwide basis, Near Fault Eects were ignored, and the %NBSnom was taken as a single value for all soil types and a period of 0.4 s. The variation in %NBSnom at T 0.4 s for
Alistair P. Russell

- 106 -

4.4. Estimated Vulnerability of URM Buildings

dierent soil types was negligible, and consequently a single value of %NBSnom = 4 was obtained. Factors A F (as outlined above) are given as follows: A = 1, B = 1/Z, C = 1, D = 1.29 and E = 1/0.85. Furthermore, NZSEE (2006) states For buildings designed prior to 1935, multiply %NBSnom by 0.8 except for Wellington where the factor may be taken as 1.0.2 Taking these assumptions into account, the only factor in determining the %NBSb which varies between provinces is the seismicity of where the building is located. This parameter is determined by the Hazard Factor, Z, which for each province was evaluated by averaging the Hazard Factors from the locations in that province (see Standards New Zealand (2004a)). %NBSb is determined according to Equation (4.1), 1 Z 1 0.85

%N BS b = 4 1

1 1.29

(4.1)

where q = 1.0 for Wellington and 0.8 for the rest of New Zealand. Thus for all URM buildings, 4.9 Z

%N BS b = except in Wellington where,

%N BS b =

6.1 Z

The %NBS is obtained by multiplying the %NBSb by the PAR. The PAR is a measure of an individual buildings expected performance, independent of location, and primarily takes into account critical structural weaknesses, such as plan and vertical irregularity and pounding potential. In collaboration with Auckland City Council during 2008, 58 buildings in Auckland City were assessed using the IEP. From this analysis it was determined that the distribution of PARs in the sample was approximately normally distributed with a sample mean ( xP AR ) of 1.6 and sample standard deviation (sP AR ) of 0.41. If it is assumed that the PAR of all URM buildings in the country is normally distributed with the same mean and standard deviation as calculated for the sample population in Auckland
NZSEE (2006) states that buildings constructed prior to 1935 should be treated as having been designed to 80% of NZSS 1900 Chapter 8 (New Zealand Standards Institute, 1965) except in Wellington, where a seismic code was in place prior to 1935 and 100% of NZSS 1900 Chapter 8 is appropriate.
2

- 107 -

Alistair P. Russell

Chapter 4. New Zealand URM Building Stock

City, the distribution of %NBS for all URM buildings in each former province in New Zealand can be estimated as follows,

(%N BS ) = %N BS b sPAR

(%N BS ) = %N BS b xPAR

where (%N BS ) and (%N BS ) are the %NBS population standard deviation and population mean respectively for each province. For each province the Hazard Factor, %NBSb , and mean %NBS and standard deviation %NBS are shown in Table 4.6.

Table 4.6: Baseline %NBSb for provinces Province Auckland Taranaki Hawkes Bay Wellington Marlborough Nelson Westland Canterbury Otago and Southland Z 0.13 0.22 0.39 0.40 0.32 0.27 0.34 0.22 0.15 %NBSb 37.5 22.7 12.7 15.3 15.5 18.0 14.5 22.1 32.5 (%N BS ) 60.0 36.3 20.3 24.3 24.8 28.8 23.2 35.4 52.0 (%N BS ) 15.4 9.3 5.2 6.2 6.4 7.4 5.9 9.1 13.3

Applying the number of URM buildings determined in Section 4.2 (3879 URM buildings in total) to the normal distribution of %NBS scores, an estimate of all the %NBS scores for each of the provinces can be evaluated, as shown in Figures 4.8(a) and 4.8(b), in terms of a probability density function (pdf) and cumulative distribution function (cdf), respectively. Consequently, the number of URM buildings in each province with an estimated %NBS below 33, between 33 and 67, and above 67 can be evaluated, and thus the number of URM buildings in each province which are potentially earthquake prone, potentially earthquake risk and unlikely to be signicant, respectively, can be estimated. This information is shown in Table 4.7 and aggregated to determine the estimated overall
Alistair P. Russell

- 108 -

4.4. Estimated Vulnerability of URM Buildings

number of URM buildings in these categories throughout all New Zealand, as shown in Figures 4.9 and 4.10.

Table 4.7: Estimated number of potentially earthquake prone and earthquake risk URM buildings Potentially earthquake prone 41 59 85 622 42 94 39 338 66 4% 36% 99% 92% 90% 72% 95% 40% 8% Potentially earthquake risk 628 105 1 55 5 37 2 513 664 61% 64% 1% 8% 10% 28% 5% 60% 78% Unlikely to be signicant 357 0 0 0 0 0 0 0 126 35% 0% 0% 0% 0% 0% 0% 0% 15% 12%

Province Auckland Taranaki Hawkes Bay Wellington Marlborough Nelson Westland Canterbury Otago and Southland New Zealand

1386 36% 2010 52% 483

From these results (Figures 4.8 4.10), it can be seen that up to 36% of URM buildings currently existing in New Zealand could be potentially earthquake prone. Most of these buildings are in regions of higher seismicity, which is the most critical factor in the vulnerability of URM buildings. Bothara et al. (2008) noted from assessments conducted in Wellington that most unreinforced masonry buildings have been conrmed as potentially earthquake prone. This statement is in agreement with the results presented here, in which 92% of URM buildings in Wellington are estimated to be potentially earthquake prone. Additionally, 52% of all New Zealand URM buildings are estimated as being not earthquake prone as dened by the Building Act 2004, but can be expected to perform at a level less than 67% of the standard of a new building. NZSEE recommends that buildings with < 67%NBS should be seriously considered for improvement of their structural seismic performance. Thus up to 88% of all URM buildings in New Zealand could require seismic improvement, according to the criteria set by NZSEE (2006).

- 109 -

Alistair P. Russell

Chapter 4. New Zealand URM Building Stock

500

Number of Buildings

400

300

Auckland Taranaki Hawkes Bay Wellington Marlborough Nelson Westland Canterbury Otago & Southland

200

100

0 0 10 20 30 40 50 60 70 80 90 100

%NBS
(a) Probability density function

1200
Auckland Taranaki Hawkes Bay Wellington Marlborough Nelson Westland Canterbury Otago & Southland

1000

Number of Buildings

800

600

400

200

0 0 10 20 30 40 50 60 70 80 90 100

%NBS
(b) Cumulative distribution function

Figure 4.8: Estimated %NBS of URM buildings in provinces throughout New Zealand

Alistair P. Russell

- 110 -

4.5. Conclusions

2500

60%

Number of URM buildings

2000

2010

50%

40% 1500 1386 30% 1000 20% 483 10%

500

0 EQ Prone EQ Risk OK

0%

Figure 4.9: National estimate of potentially earthquake prone and earthquake risk URM buildings It must be recognised that the modelling exercise reported here is essentially qualitative in nature and can be expected to overestimate the number of poorly performing URM buildings, primarily because of the conservative nature of the IEP. Nevertheless, as an informative estimate of the nature of the vulnerability of New Zealands URM building stock, this analysis is considered valid.

4.5

Conclusions

Through two independent methods an estimation of the number of URM buildings in New Zealand has been established. Both approaches suggest that there are between approximately 3589 and 3879 URM buildings in existence in New Zealand currently. The majority of these URM buildings are 1 and 2 storeys in height. Trends in the age of these buildings show that construction activity increased from the early days of European settlement and reached a peak at about 1930, before subsequently declining sharply. Few URM buildings have been observed throughout the country with a construction date later
- 111 Alistair P. Russell

Proportion of URM building stock

Chapter 4. New Zealand URM Building Stock

4000 3500 3000 2500 2000 1500 1000 500 0 0 10 20 30 40

Number of URM buildings

%NBS

50

60

70

80

90

100

Figure 4.10: Estimated %NBS of URM buildings in New Zealand than 1950, and the preponderance of the existing URM building stock was constructed prior to 1940. Thus, almost all URM structures in New Zealand are between 80 and 130 years old (in 2010). Overall the URM building stock has a value of approximately $NZ1.5 billion. Collectively the value of smaller buildings is lower than the value of larger buildings, but individually the larger buildings have greater nancial value. This suggests that for seismic assessment and retrot, less time and money should be invested in smaller buildings individually, but collectively this will form the majority of the work which remains to be undertaken in making the New Zealand URM building stock safer.

An estimate of the vulnerability of all New Zealand URM buildings suggested that up to 36% of these could be assessed as potentially earthquake prone, and an additional 52% could be assessed as earthquake risk. Consequently, up to 88% of all New Zealand URM buildings could be found to require improvements to their structural performance.

Alistair P. Russell

- 112 -

CHAPTER 5

Diagonal Shear Testing of Wall Panels

The results of diagonal shear tests on eight two-leaf unreinforced masonry wall panels are presented in this chapter. The aim of the experimentation was to investigate the diagonal tension (shear) strength of unreinforced masonry having dierent mortar properties and bond patterns, and also to provide a baseline against which to compare other retrotted masonry samples and also samples tested in-situ in existing buildings. The tests presented here provide a link between in-situ as-built unreinforced masonry and laboratory built and tested retrotted masonry.

The results of the tests showed signicant variation in the diagonal tension strength of the wall panels, even when the mortar composition and other variables were consistent. Both the measured shear modulus and measured Youngs modulus of samples were consistent in all eight samples.

In this chapter, wall panels which fail due to diagonal cracking are described and analysed. The test which is used to eect this failure mode is referred to in the literature as the diagonal compression test and also the diagonal shear test. This is because in the test a diagonal compression force is applied and the diagonal cracks which form are often
- 113 Alistair P. Russell

Chapter 5. Diagonal Shear Testing of Wall Panels

referred to as shear cracks. ASTM (2007b) refers to the test as the Standard Test Method for Diagonal Tension (Shear) in Masonry Assemblages, and the diagonal cracks which form in the masonry assemblage are a result of the tension strength of the masonry being exceeded when subjected to the applied stress state. Thus the test could reasonably be referred to as either the diagonal compression test, the diagonal shear test or the diagonal tension test. Because the applied forces on a building which would lead to such a failure mode are neither exclusively tension nor compression, and because this failure mode is most often referred to as a shear failure mode, in this chapter the test which is described is referred to as the diagonal shear test.

5.1

Diagonal Shear Test

ASTM International denes a standard test method (E 519) for determining the shear strength of masonry assemblages (ASTM, 2007b), in which 1200 mm by 1200 mm masonry assemblages are rotated by 45 and a force is applied vertically between the two corners of the sample, see Figure 5.1(a) (see Gabor et al., 2006; Hamid et al., 2005; Marshall et al., 2000). In a variation of this test, the masonry assemblage rests on a plinth or a base, and the force is applied diagonally between two corners, see Figure 5.1(b). Numerous studies have been published using the latter method (see Brignola et al., 2009; Grando et al., 2003; Petersen et al., 2008, 2009; Yu et al., 2007), and in addition this method facilitates comparisons with experiments conducted in-situ, in which the masonry cannot be rotated and the force must be applied diagonally (see Brignola et al., 2006a,b; Chiostrini et al., 2000; Corradi et al., 2003, 2008; Valluzzi et al., 2002; Vestroni et al., 1995). The standard test method requires a specimen size of 1200 mm by 1200 mm because it is suggested that this is the smallest geometry that would be reasonably representative of a full-size masonry assemblage and that would permit the use of testing equipment in many laboratories.

ASTM recognises the increased shear resistance of a masonry sample subjected to additional axial load normal to bed joints, as is typically the case in a URM building, resulting from dead and live loads. Included as an annex, the standard provides a method for apAlistair P. Russell

- 114 -

5.1. Diagonal Shear Test

(a) Force applied ASTM E 519

vertically

according

to

(b) Force applied diagonally as per modied test setup

Figure 5.1: Orientation of wall panels for ASTM procedure and modied procedure plying this load, but most documented tests have been conducted without additional axial load.

5.1.1

Test Setup

The test setup is shown in Figure 5.2, and is based on the procedure found in ASTM E 519 (2007b). Each wall panel was placed on a plinth, so that the loading shoe at the bottom corner could be attached. Two high strength steel rods on each face of the wall panel connected both loading shoes located at diagonally opposite corners, and the compression force was applied with an hydraulic actuator. Each loading shoe was 225 mm long and 225 mm high. Brignola et al. (2009) noted that the stress state in the middle of the panel is not signicantly inuenced by the size of the loading shoes, see Section 5.1.2. Displacements were measured with portal gauge strain transducers located diagonally between corners, oriented both parallel and perpendicular to the direction of applied load.
- 115 Alistair P. Russell

Chapter 5. Diagonal Shear Testing of Wall Panels

Figure 5.2: Diagonal shear test setup The force, P, was applied monotonically and at a uniform rate until failure occurred. See ASTM (2007b) for full test setup details.

5.1.2

Interpretation of Diagonal Shear Test

ASTM test method E 519 provides equations for interpreting the results of each experiment. The diagonal shear stress (s ) is calculated as per Equation (5.1) and (5.2), P s = 0.707 An lw + h bw n An = 2 and the shear strain ( ) of the sample is calculated as per Equation (5.3), = c + t
Alistair P. Russell

(5.1) (5.2)

(5.3)

- 116 -

5.1. Diagonal Shear Test

where P is the applied force, An is the area of the net mortared section, lw , h and bw are the length, the height and the width (thickness) of the wall respectively, n is the proportion of gross solid area of unit (expressed as a decimal), c is the direct strain in compression and t is the direct strain in tension. ASTM denes c and t as follows, V g H g

c =

t =

(5.4)

but because the specimen is usually rotated by 45 (see Figure 5.1), the vertical shortening (V) is more appropriately dened as short , and similarly, the horizontal extension (H) is more appropriately dened as long . Both short and long are dened as positive, see Equations (5.6) and (5.7). Thus as the specimen deforms, the reduction in length parallel to the applied force between corners is positive, and the extension in length perpendicular to the applied force is also positive. short + long g

(5.5) (5.6) (5.7)

short = g gshort long = glong g

Figure 5.3 shows the distortion of the wall panel and the gauge lengths (g ) used to determine the shear strain, . Both gauges in each test (Section 5.3) were the same length,

(a) Original gauge lengths

(b) Gauge lengths after distortion of wall panel

Figure 5.3: Gauge lengths for measuring wall panel distortion

- 117 Alistair P. Russell

Chapter 5. Diagonal Shear Testing of Wall Panels

although Brignola et al. (2009) notes that if the parallel and perpendicular gauge lengths are not equal, the sum of the strain determined from each is still the same as if identical lengths are used, see Section 5.1.3.

Additionally, ASTM E 519 (2007) and ASTM E 143 (2002) provide equations for determining the shear modulus (modulus of rigidity, G) and Youngs modulus (E ) respectively, as per equations (5.8) and (5.9), P An

G = 0.707

(5.8) (5.9)

E = 2G(1 + )

where is Poissons ratio. G is calculated from the stress-strain curve as the chord modulus between 5% and 33% of the maximum shear strength (ASTM, 2004b, 2007a).

Diagonal cracks occur in the wall panel when the diagonal tension strength of the masonry, (fdt ), is exceeded by the applied tension stress, which in the diagonal shear test is 1 . Thus the diagonal tension strength of each wall panel is determined from the maximum principal stress at failure. ASTM standard E 519 assumes that the diagonal shear test produces a uniform shear stress in the specimen and that the specimen is in pure shear (see Equation (5.1)), and so the shear stress (s ) is equal in magnitude to both the maximum principal (tension) stress (1 ) and also to the minimum principal (compression) stress (2 ), see Figure 5.4. Thus according to the ASTM interpretation, P An

fdt = s = 1 = 2 = 0.707

Brignola et al. (2009) suggested that a masonry specimen, when loaded in the diagonal shear test, is in a non-uniform shear stress eld and that the principal stresses at the centre of the panel not equal. Instead, it was suggested that at the centre of the panel
Alistair P. Russell

- 118 -

5.1. Diagonal Shear Test

the principal stresses are as follows, P An P An P An

s = 1.05 1 = 0.50

(5.10) (5.11) (5.12)

2 = 1.62

Brignola et al. (2009) analysed masonry specimens as used in the diagonal shear test using nite element modelling in both the linear range and non-linear range. The inuence of boundary conditions was taken into account, where the applied force was considered as a concentrated force at two diagonally opposite corners of the masonry specimen, and also as a force applied by steel loading shoes, having a length of 1/10th of the panels side. It was determined that in the non-linear range the stress components (1 , 2 , s ) changed when compared with the corresponding stresses in the linear range due to damage and non-linear stress distributions, but that these changes inuence the minimum principal (compression) stress only and that the maximum principal (tension) stress remains unchanged from the onset of non-linear behaviour through to collapse. Moreover, it was found that the plane stress at the centre of the panel is not inuenced by the dierent boundary conditions (force applied as concentrated or with varying size steel loading shoes), and that the centre of the panel could be assumed as the control point for the behaviour of the panel, establishing a relationship between the collapse of the panel and the strength of the masonry. Consequently, the diagonal tension strength of masonry (fdt ) should be determined according to Equation (5.11). P An

fdt = 1 = 0.50

(5.11)

The stress state at the centre of the panel at failure is represented using Mohrs Circle in Figure 5.4. Both the ASTM interpretation of the diagonal shear test and the interpretation by Brignola et al. (2009) assume that the principal stresses are on planes that are 90 apart and at 45 from the horizontal and vertical surfaces of the specimen.
- 119 Alistair P. Russell

Chapter 5. Diagonal Shear Testing of Wall Panels

Figure 5.4: Mohrs Circle interpretation of the diagonal shear test Brignola et al. (2009) showed in further analysis that for poorly interlocked masonry (such as rubble stone masonry) Equation (5.11) is potentially non-conservative, and that a coecient of 0.35 would be appropriate. For solid bricks and lime mortar masonry (such as is found in New Zealand), a coecient of 0.5 is appropriate and Equation (5.11) is valid. Because of the widespread use of the equations specied by ASTM E 519, the results of walls tested in this chapter are reported according to both the equations given by ASTM and those given by Brignola et al., in Section 5.3.

Furthermore, Brignola et al. (2009) proposed that the shear modulus (G) should be evaluated in accordance with Equation (5.13), P An

G = 1.05

(5.13)

whereas ASTM denes the shear modulus as per Equation (5.8). Upon analysis of masonry specimens considering dierent boundary conditions, dierent values of Poissons ratio and dierent values of mechanical orthotropy1 , Brignola et al. (2009) determined
Mechanical orthotropy is the ratio of Ey (Youngs modulus normal to the bed joints) to Ex (Youngs modulus parallel to the bed joints).
1

Alistair P. Russell

- 120 -

5.1. Diagonal Shear Test

that Equation (5.13) was accurate in determining the actual shear modulus and that the value evaluated using the ASTM method (Equation (5.8)) was signicantly lower than the correct value. Consequently, Equation (5.13) should be used to determine the shear modulus of masonry specimens tested according to the diagonal shear test. G is calculated from the stress-strain curve as the chord modulus between 5% and 33% of the maximum shear strength as per ASTM (2004b) and ASTM (2007a).

5.1.3

Derivation of Drift

In order for results of wall panel tests presented in Section 5.3 to be interpreted meaningfully and compared with other results, it is useful to report the wall panel displacement in terms of drift angle, . Furthermore, this denition is particularly useful for interpretation by practitioners, who typically adopt drift rather than shear strain as a design parameter. The shear strain of the specimen is determined as per Equation (5.5), and the distorted shape (at an exaggerated scale) is shown in Figure 5.5(a). There are both vertical and horizontal components of shear stress and consequently, the total shear strain is made up of both horizontal and vertical components, and it is assumed that both components are equal (if the vertical and horizontal components are not equal, the total shear strain,

(a) Distortion and shear strain of a wall panel when subjected to the diagonal shear test

(b) Drift angle of distorted wall panel

Figure 5.5: Derivation of drift angle

- 121 Alistair P. Russell

Chapter 5. Diagonal Shear Testing of Wall Panels

, is still the sum of both). If the distorted shape is rotated, as shown in Figure 5.5(b), then it can be seen that the total shear strain is the sum of the vertical and horizontal components, and the drift angle, , is equal in magnitude and direction to the total shear strain. The drift is determined as per Equation (5.14). = (5.14)

Magenes and Calvi (1997) and Priestley et al. (2007) suggest that an ultimate drift of 0.4% is appropriate for diagonal shear failure of masonry. It was also reported in NZSEE (2006) that shear strains (and consequently, drift) for diagonal shear failure of masonry should not exceed 0.4%. ASCE (2007) gives a drift for the immediate occupancy (IO) limit state of 0.1%. In Figure 5.12 the response of the wall panels is shown to a maximum drift of 0.2%, as failure had occurred before this drift level was reached.

5.2

Construction Details

Eight wall panels were constructed and tested, and each was two leafs thick. The mortar Table 5.1: Construction details of wall panels Designation AP1 AP2 AP3 AP4 AP6 AP7 AP8 AP9 Mortar type (cement:lime:sand) 2:1:9 1:2:9 0:1:3 1:2:9 1:2:9 1:2:9 1:2:9 1:2:9 Bond pattern Common English Common Common Common Common Common Common
- 122 fm fj fb (MPa) (MPa) (MPa)

16.6 14.6 3.2 13.7 5.6 7.4 7.3 8.5

5.7 3.7 1.1 3.5 2.6 2.3 2.2 2.6

23.3 24.6 24.3 26.2 23.6 20.4 20.1 23.4

Alistair P. Russell

5.3. Results

strength (fj ) and bond pattern (see Section 3.5.1) were varied as shown in Table 5.1. Three mortar cubes and three prisms were constructed at the same time as each wall panel and tested on the same day as the diagonal shear test was conducted. This was to determine the compression strength of the mortar and the masonry respectively. Additionally, six half bricks were tested with the prisms and mortar cubes to determine the brick compression strength (fb ). Details of these tests can be found in ASTM (2003,
2004a, 2007a). fm refers to the compressive strength of masonry.

Wall panels AP1 AP4 were built and then tested ve months later, whilst wall panels AP6 AP9 were tested one month after being built.

5.3

Results

Table 5.2 summarises the results of each wall panel test. Each variable is evaluated as dened previously, except the mean cohesion of the masonry in the panel, c, which is dened as 1.5 fdt (fdt = 1 ), assuming a friction coecient2 , , of 0.65. E is determined according to Equation (5.9) using Poissons ratio, = 0.25, as suggested by Harry (1988) and Pande et al. (1998) for unreinforced masonry3 , and Vdt is described below.

covi Turn sek and Ca c (1970) and Turn sek and Sheppard (1980) proposed Equation (5.15) to relate the strength of a wall failing in diagonal tension, Vdt , to the diagonal tension strength of the masonry, fdt , (see Figure 5.6). fdt lw bw fm Vdt = 1+ b fdt

(5.15)

where b is a parameter that is dependent on the wall aspect ratio h/lw , and accounts for the distribution of shear stresses at the centre of the wall. Benedetti and Toma zevi c (1984)
Derakhshan et al. (2010) suggest that from eld testing of existing buildings, = 0.65 is appropriate for New Zealand URM. 3 ASCE (2007); Yi et al. (2005) suggest G = 0.4 E , which also assumes = 0.25 (see Equation (5.9).
2

- 123 -

Alistair P. Russell

Chapter 5. Diagonal Shear Testing of Wall Panels

Figure 5.6: Shear strength of a wall and diagonal tension strength of masonry proposed a value of b = 1 for h/lw 1, b = h/lw for 1 < h/lw < 1.5, b = 1.5 for h/lw 1.5.

The values of principal stress (diagonal tension strength) of the wall panels in this chapter are related to the shear strength of the side wall of a mean Typology A1 building (see Figure 6.1) with the following dimensions: lw = 8 m, bw = 230 mm, h = 4 m, fm = 0.01 MPa Table 5.2: Summary of results of wall panels Wall panel AP1 AP2 AP3 AP4 AP6 AP7 AP8 AP9 Mean COV (%)
P 1 fdt /fm (kN) (MPa)

Vdt crack (kN) % 694 416 56 388 125 124 139 111 0.15 0.08 0.02 0.12 0.02 0.01 0.02 0.01

c G E Cracking (MPa) (GPa) (GPa) through 0.75 0.45 0.06 0.42 0.14 0.14 0.15 0.12 0.93 0.93 0.80 0.95 0.92 0.84 0.78 1.02 0.90 9.1 2.32 2.33 2.01 2.36 2.31 2.11 1.96 2.56 2.25 9.0 bricks mortar mortar mortar mortar mortar mortar mortar

265 157 22 151 51 48 52 43

0.50 0.30 0.04 0.28 0.09 0.09 0.10 0.08

0.030 0.021 0.013 0.020 0.016 0.012 0.014 0.009 0.017 39

Alistair P. Russell

- 124 -

5.3. Results

(see Table 3.2), and b = 1. The results are shown as Vdt in Table 5.2.

At failure, damage in the wall panels was characterised by the cracks propagating through either the bricks or through the mortar joints in a step-wise manner. These crack patterns are illustrated in Figures 5.7(a) and 5.7(b) respectively, and photographically in Figures 5.8 and 5.9 respectively. Cracking through the bricks indicated that the mortarbrick bond was stronger than the bricks, and cracking through the mortar joints indicated

(a) Cracking through bricks

(b) Cracking through mortar

Figure 5.7: Crack patterns at failure of wall panels

(a) Side view

(b) End view

Figure 5.8: Photos of cracking through bricks in wall panels

- 125 Alistair P. Russell

Chapter 5. Diagonal Shear Testing of Wall Panels

(a)

(b)

Figure 5.9: Photos of cracking through mortar in wall panels

that the bricks were stronger than the mortar-brick bond. Figure 5.11 shows the forcedisplacement response of each wall panel. The force measured is the applied diagonal compression force, P , and the displacement is measured parallel to the direction of loading, short . Figure 5.12 shows the maximum principal stress (diagonal tension stress) and drift results of each wall, as determined from Equations (5.11) and (5.14).

Note that 1 shown in Table 5.2 is determined according to Equation (5.11), but in Figure 5.12, 1 is determined according to both Equation (5.11) (referred to as Brignola et al.) and Equation (5.1) (referred to as ASTM). Note also in Figures 5.11 and 5.12 that for greater clarity in each graph the scales on each subgure are not consistent.

Figure 5.10 shows the sudden failure in wall panel AP1. This wall panel was constructed with the strongest mortar mix, 2:1:9 (cement:lime:sand), and when the diagonal tension strength of the masonry was exceeded, the failure was brittle and sudden. In contrast, those walls in which cracking occurred exclusively around the bricks and through the mortar (AP3, AP6 AP9) exhibited a much less explosive failure. After the peak stress was reached, the two sections of wall panel on either side of the diagonal crack were able to slide past each other (see Figure 5.9) and there was some residual strength.

Alistair P. Russell

- 126 -

5.4. Discussion

(a)

(b)

(c)

(d)

Figure 5.10: Shear failure mode of Wall AP1

5.4

Discussion

In the middle of the wall panel, the stress state exists as shown in Figure 5.4, and the stresses s , 1 and 2 can be determined according to Equations (5.10), (5.11) and (5.12) respectively. Because cracking occurs as a result of the diagonal tension capacity of the masonry, fdt , being exceeded by the maximum principal (tension) stress, 1 , it is 1 which is reported instead of the maximum shear stress, s . The maximum principal stress, 1 , varied considerably between the wall panels, and this is to be expected for varying mortar compositions and bond patterns. It would be expected that those walls built with the same bond pattern and the same mortar composition would have comparable strengths, but it appears from Table 5.2 that there is considerable variation in strength between comparable wall panels. AP4 was built with 1:2:9 mortar and Common bond pattern, as were wall panels AP6 AP9, but the strength of AP4 is approximately three times that of each of AP6 AP9. Similarly, AP2 (with English bond and 1:2:9 mortar) exhibited a strength comparable to AP4, but its strength was much greater than that shown in each of AP6 AP9. AP1, with the strongest mortar mix, exhibited the greatest strength and AP3, with the weakest mortar mix, exhibited the weakest diagonal tension strength.
- 127 Alistair P. Russell

Chapter 5. Diagonal Shear Testing of Wall Panels

300 Applied Diagonal Force (kN) 250 200 150 100 50 0 0 0.5 1 1.5 2 2.5 Diagonal Displacement (mm) 3 Applied Diagonal Force (kN)

300 250 200 150 100 50 0 0 0.5 1 1.5 Diagonal Displacement (mm) 2

(a) Wall AP1

100 Applied Diagonal Force (kN) Applied Diagonal Force (kN) 80 60 40 20 0 0 0.5 1 1.5 Diagonal Displacement (mm) 2 300 250 200 150 100 50 0 0

(b) Wall AP2

0.5 1 1.5 Diagonal Displacement (mm)

(c) Wall AP3

100 Applied Diagonal Force (kN) 80 60 40 20 0 0 0.5 1 1.5 Diagonal Displacement (mm) 2 Applied Diagonal Force (kN) 100 80 60 40 20 0 0

(d) Wall AP4

0.5 1 1.5 Diagonal Displacement (mm)

(e) Wall AP6

100 Applied Diagonal Force (kN) 80 60 40 20 0 0 0.5 1 1.5 Diagonal Displacement (mm) 2 Applied Diagonal Force (kN) 100 80 60 40 20 0 0

(f) Wall AP7

0.5 1 1.5 Diagonal Displacement (mm)

(g) Wall AP8

(h) Wall AP9

Figure 5.11: Force-displacement response of wall panels

Alistair P. Russell

- 128 -

5.4. Discussion

Maximum Principal Stress "1 (MPa)

1 0.8 0.6 0.4 0.2 0 0 0.05 0.1 Drift ! (%) 0.15 0.2 ASTM Brignola et al.

Maximum Principal Stress "1 (MPa)

1 0.8 0.6 0.4 0.2 0 0 0.05 0.1 Drift ! (%) 0.15 0.2 ASTM Brignola et al.

(a) Wall AP1

Maximum Principal Stress "1 (MPa) 0.2 0.15 0.1 0.05 0 0 ASTM Brignola et al. Maximum Principal Stress "1 (MPa) 1 0.8 0.6 0.4 0.2 0 0

(b) Wall AP2

ASTM Brignola et al.

0.05

0.1 Drift ! (%)

0.15

0.2

0.05

0.1 Drift ! (%)

0.15

0.2

(c) Wall AP3

Maximum Principal Stress "1 (MPa) 0.2 0.15 0.1 0.05 0 0 ASTM Brignola et al. Maximum Principal Stress "1 (MPa) 0.2 0.15 0.1 0.05 0 0

(d) Wall AP4

ASTM Brignola et al.

0.05

0.1 Drift ! (%)

0.15

0.2

0.05

0.1 Drift ! (%)

0.15

0.2

(e) Wall AP6

Maximum Principal Stress "1 (MPa) 0.2 0.15 0.1 0.05 0 0 ASTM Brignola et al. Maximum Principal Stress "1 (MPa) 0.2 0.15 0.1 0.05 0 0

(f) Wall AP7

ASTM Brignola et al.

0.05

0.1 Drift ! (%)

0.15

0.2

0.05

0.1 Drift ! (%)

0.15

0.2

(g) Wall AP8

(h) Wall AP9

Figure 5.12: Maximum principal stress-drift response of wall panels

- 129 -

Alistair P. Russell

Chapter 5. Diagonal Shear Testing of Wall Panels

As a consequence of the variation in diagonal tension strength between wall panels, the cohesion, c, varied to a similar extent.

Because wall panels AP1 AP4 were tested ve months after they were built and wall panels AP6 AP9 were tested one month after construction, it appears that age has an eect on the diagonal tension strength of laboratory built unreinforced masonry wall panels (it was assumed at the time of testing that a masonry wall panel would have effectively reached its ultimate strength after 28 days, in a similar manner to concrete). Consequently, when considering the eect of retrots on wall panels it is important that in the laboratory, as-built and retrotted wall panels of similar ages (as well as with similar constituent material properties) are compared.

Despite large variations in the diagonal tension strength of the eight masonry wall panels reported above, the values of shear modulus and Youngs modulus were reasonably consistent between tests, with an average value of G = 0.9 GPa and an average value of E = 2.25 GPa, and with a coecient of variation4 of 9.1% and 9.0% respectively. Consequently, there appears to be no direct relationship between the diagonal tension strength of the test samples reported in this chapter and the stiness of masonry as a material. Because the shear modulus, G, (and correspondingly E ) is determined from the pre-cracked region of the stress-strain response (between 5% and 33% of the peak stress, where no cracking occurs), it would be expected that the diagonal tension strength of masonry is not directly related to the stiness, as in the pre-cracked region the masonry behaves as a homogeneous material. Thus it appears that for the laboratory built wall panels reported above, the elastic stiness is relatively independent of strength.

5.4.1

Prediction of Shear Strength

Lee et al. (2008) suggested that the diagonal tension strength of masonry can be related
to the compressive strength of masonry as fdt = 0.05fm . This is too high to correlate with the results of tests reported in this chapter. The mean of the ratio of fdt /fm of the wall
4

The coecient of variation is the ratio of the standard deviation to the mean of the sample.

Alistair P. Russell

- 130 -

5.4. Discussion

panels reported in this chapter is 0.017 with a COV of 39%. The relationship between the diagonal tension strength of masonry and the compressive strength of masonry from the tests reported in this chapter is shown in Equation (5.16),
fdt = 0.017fm

(5.16)

but because of the large scatter in the results, no specic relation between fdt and fm can

be recommended.

For the determination of the lateral strength of masonry responding in-plane and limited by diagonal tension stress, FEMA 273 (1997) allows the substitution of the diagonal tension strength fdt for the bed joint shear strength vme . FEMA 273 (1997) gives Equation (5.17) for determining the expected masonry shear strength. PCE An (5.17)

fdt = vme =

0.75 0.75 vte + 1.5

where PCE is the expected gravity compressive force applied to a wall or pier, An is the area of the net mortared section and vte is the average bed-joint shear strength as determined from a standard in-situ shear test (see Section 7.3.2.4 of FEMA 273, 1997).

Consequently, for walls failing in diagonal tension, the masonry shear strength can be determined from Equation (5.15) (in terms of force) or from Equation (5.18) (in terms of stress) using values of fdt as determined from Equation (5.17). fdt fm vdt = 1+ b fdt

(5.18)

5.4.2

Comparison of Shear Strength with ASCE 41-06

The condition of masonry can be classied as good, fair or poor, according to ASCE
- 131 Alistair P. Russell

Chapter 5. Diagonal Shear Testing of Wall Panels

Table 5.3: Default lower bound masonry properties (from ASCE (2007)) Masonry condition Good Fair Poor 495,000 psi 3.41 GPa 27 psi 0.19 MPa 330,000 psi 2.28 GPa 20 psi 0.14 MPa 165,000 psi 1.14 GPa 13 psi 0.09 MPa

Property E vme

(2007). In lieu of material tests or other data, ASCE (2007) provides default lower bound material properties to determine URM component strengths, as given in Table 5.3. As stated above, vme may be substituted for fdt . The values of shear strength may be multiplied by a factor of 1.3 to translate the lower bound strength to expected strength.

The diagonal tension strength and condition of the wall panels reported in this chapter, according to ASCE (2007) lower bound material properties, are shown in Table 5.4. Hence the diagonal tension strengths of walls panels AP6 AP9 correspond to values of strength for masonry in poor condition. The diagonal tension strength of wall panel AP3 was lower than what would be considered a lower bound value of strength for masonry in Table 5.4: Comparison of wall panel shear strength with ASCE (2007) lower bound properties Wall panel AP1 AP2 AP3 AP4 AP6 AP7 AP8 AP9 fdt ASCE (2007) condition (MPa) 0.50 0.30 0.04 0.28 0.09 0.09 0.10 0.08
- 132 -

Good + Good + Poor Good + Poor Poor Poor Poor

Alistair P. Russell

5.5. Conclusions

poor condition. The expected shear strength of masonry in good condition is 31.5 psi, or 0.242 MPa. As such, wall panels AP1, AP2 and AP4 correspond to masonry in better than good condition, according to the properties given in ASCE (2007).

The average elastic modulus E as determined from the results of the wall panel tests reported in this chapter was 2.25 GPa, as shown in Table 5.2. ASCE (2007) suggests lower bound values of elastic modulus in compression as shown in Table 5.3. The factor translating the default lower bound value of elastic modulus in compression to the expected value is 1, and as such the two values are considered to be the same. For the wall panels reported in this chapter, the elastic modulus corresponds to the elastic modulus in compression of masonry in fair condition.

5.5

Conclusions

As expected, the diagonal tension strength of unreinforced masonry wall panels varied with varying mortar strength. AP1, with the strongest mortar mix, exhibited the greatest diagonal tension strength and AP3, with the weakest mortar mix, exhibited the weakest diagonal tension strength. Clearly, there is a correlation between the mortar strength and the diagonal tension strength of the masonry.

There was no clear relationship shown between the diagonal tension strength and bond pattern. AP2 was constructed with English bond and AP4 and was constructed with Common bond, and all other variables were kept constant between these two wall panels. There was no noticeable dierence in diagonal tension strength between AP2 and AP4.

There was a signicant dierence between wall panels built with the same mortar composition. AP2, AP4, AP6, AP7, AP8 and AP9 were all built with the same constituents in the mortar (1:2:9, cement:lime:sand, by volume), but the results showed that AP2 and AP4 had a much higher diagonal tension strength when compared with AP6 AP9. The age at which AP2 and AP4 were tested was greater than the age when AP6 AP9 were
- 133 Alistair P. Russell

Chapter 5. Diagonal Shear Testing of Wall Panels

tested, and it was concluded that the age of laboratory-built samples has a signicant eect on diagonal tension strength.

Despite the variation in diagonal tension strength of the wall panels reported above, the results are useful as a baseline for comparing with retrotted masonry samples and also with in-situ masonry. The results of retrotted laboratory-built wall panels can be compared with laboratory-built as-built samples, but unless the laboratory-built as-built results are calibrated against in-situ data, only a relative performance increase can be determined. When the laboratory-built as-built results are calibrated with in-situ as-built results, the results of the tests on retrotted laboratory samples then have some applicability to real buildings.

The results of wall panels AP6 AP9 have been used as a comparison with in-situ diagonal shear tests on URM buildings in Wellington and Gisborne. It was found that the diagonal tension strengths of wall panels AP6 AP9 were similar to the results from four tests on existing URM buildings. Diagonal shear tests on retrotted masonry samples built in the laboratory with the same materials and tested at a similar age to wall panels AP6 AP9 could then be compared with in-situ results from existing buildings. Further details can be found in Derakhshan et al. (2010) and Mahmood (2011).

Finally, an equation (Equation (5.18)) for predicting the shear strength of masonry walls was suggested, based on values determined from in-situ shear tests.

Alistair P. Russell

- 134 -

CHAPTER 6

In-Plane Cyclic Testing of Rectangular URM Walls

This chapter presents the results of pseudo-static in-plane testing of three unreinforced masonry walls designed to replicate typical New Zealand construction in the early 20th Century, with dierent aspect ratios and dierent levels of axial load.

The aim of the experimentation reported in this chapter was to investigate the eects of axial load and aspect ratio on the response of rectangular URM walls without anges, and was achieved by investigating the response of three URM walls with dierent characteristics. The rst two walls (designated Wall A1 and Wall A2) both had an aspect ratio of 1:1 and the third wall (Wall A4) had an aspect ratio of 1:2. Walls A1 and A4 had low levels of axial load, and Wall A2 had a high level of axial load (determined as the highest level of axial load that could be expected in a typical New Zealand URM bearing wall). Wall A1 was tested to investigate the response of a straight, square wall with low axial load, Wall A2 was tested to investigate the response of a straight, square wall with high axial load and to determine what inuence the level of axial load had on behaviour mode, and Wall A4 was tested to investigate the response of a straight, rectangular wall with
- 135 Alistair P. Russell

Chapter 6. In-Plane Cyclic Testing of Rectangular URM Walls

low axial load and to determine what inuence aspect ratio had on failure mode.

Furthermore, the aim of the tests reported here was to acquire experimental results from URM in-plane wall tests, against which the predicted response could be compared. The response was predicted using the guidelines given in NZSEE (2006) and ASCE (2007).

The results of the tests reported in this chapter were also used to compare the response of straight (unanged) walls and the response of walls with anges, as reported in Chapter 7.

6.1

Background

Previously published experimental research that considered existing New Zealand URM wall response has focused on out-of-plane behaviour (Blaikie, 1999, 2002) and the dynamic response of half scale URM structures (Bothara, 2004; Bothara et al., 2010). Guidelines on the assessment of in-plane wall response (NZSEE, 2006) have been mainly derived from results obtained overseas (Foss, 2001; Gambarotta and Lagomarsino, 1997b; Kitching, 1999; Magenes and Calvi, 1992, 1997; Magenes and Della-Fontana, 1998).

Abrams and Shah (1992) investigated the response of URM walls with dierent aspect ratios and determined that cyclic behaviour can be represented in terms of behaviour under monotonically increasing forces. Other research into the in-plane response of URM walls has focused on the response of piers having aspect ratios (height to length) greater than 1 and on the eectiveness of dierent retrot interventions (Abrams et al., 2007; Franklin et al., 2001). The research published by Abrams et al. (2007) and Franklin et al. (2001) was reported in the context of performance criteria stipulated by ASCE (2007).

Steelman and Abrams (2007) investigated the inuence of axial compressive stress and wall aspect ratio on lateral in-plane strength of URM shear walls. The results of this study showed that lateral force capacity is largest for relatively squat walls sustaining
Alistair P. Russell

- 136 -

6.1. Background

large amounts of gravity compressive stress. This is because wall area, vertical force and frictional force are all large. In the range where bed-joint sliding governs, increases in lateral strength are nearly directly proportionate with increases in axial compressive stress. As wall aspect ratio increases, lateral force capacity decreases markedly as the bed-joint sliding failure mode transitions to either the rocking or toe compression failure mode. For slender elements with a high aspect ratio, lateral strength increases only slightly with an increase in axial stress until pre-cracking toe compression governs, when strength decreases with an increase in axial stress.

Full scale testing of URM structures has been reported by Moon (2004); Moon et al. (2007); Yi (2004) and Yi et al. (2006a,b), who investigated the response of a URM structure subjected to lateral loads. This testing enabled the observation of failure modes on a system level, and the determination of retrot intervention eectiveness. Other system level URM experimental investigations have been reported by Abrams (1997; 2000), Abrams and Costley (1996) and Paquette and Bruneau (2003, 2006), and were used to investigate the response of URM buildings with shear walls and exible diaphragms and also to determine rst-mode participation factors from dynamic testing.

Gambarotta and Lagomarsino (1997a,b) proposed damage models for mortar joints in URM and used these damage models to evaluate the in-plane response of masonry shear walls subjected to lateral forces. These models were correlated against experimental data obtained from testing of URM walls reported in Anthoine et al. (1994); Magenes and Calvi (1992) and Magenes and Calvi (1997).

The equations for use in the NZSEE guidelines for evaluating URM shear wall response were developed primarily by Magenes and Calvi (1997) and were based on damage models covi developed by both Mann and M uller (1973, 1982, 1985) and also by Turn sek and Ca c (1970) and Turn sek and Sheppard (1980). No New Zealand research has been published reporting the experimentation of URM walls responding in-plane. The results reported in this chapter aim, in part, to address this gap.

- 137 -

Alistair P. Russell

Chapter 6. In-Plane Cyclic Testing of Rectangular URM Walls

6.2
6.2.1

Construction Details
Wall Specications

Dimensions and specications for the three rectangular test walls are depicted in Figure 6.2 and listed in Table 6.1. The prex A refers to As-built walls and was included in the wall name to dierentiate from the walls tested in the lab at similar times but having retrot interventions (which are not part of this thesis). Wall A1 was 2500 mm high, 2500 mm long and 350 mm in width (3 leaves), with an aspect ratio of 1:1. After Wall A1 was constructed and tested it was decided that laboratory built walls having a height of 2000 mm were sucient to replicate existing URM walls, which would equally allow observation of the in-plane lateral response. Similarly, it was determined that the prevalence of 2 leaf walls was comparable to that of 3 leaf walls in New Zealand URM construction, and thus later walls were constructed with a width of 2 leaves (240 mm). The reduction in height and width of subsequent walls allowed a decrease in wall construction time and material requirements. Consequently, Wall A2 had an aspect ratio of 1:1, but was 2000 mm high, 2000 mm long and 240 mm in width. Wall A4 was 2000 mm high, 4000 mm long, and 240 mm wide, with an aspect ratio of 1:2.

Wall A1 was built to simulate a side wall in a single storey isolated URM structure (Typology A), as shown in Figure 6.1. A small axial load of 9 kN (10.3 kPa) was applied to the top of the wall, to simulate the axial load experienced on the side wall of a Typology A building, and on such a building, this axial load is comprised of the weight of the front Table 6.1: Wall specications Wall Age Days A1 A2 A4 23 20 72 bw mm 350 230 230 h mm lw mm
fm

fj 2.6 2.4 2.3

- 138 -

fb 27.0 26.4 25.2

fbt

fm /fm

MPa MPa MPa MPa MPa MPa 15.0 11.0 9.8 2.1 1.9 2.0 0.3 0.3 0.1 0.7 0.7 0.7

% 0.069 3.160 0.333

2500 2500 2000 2000 2000 4000

Alistair P. Russell

6.2. Construction Details

Figure 6.1: Side wall of a Typology A structure parapet and a small roof load. A wall in an actual building as shown in Figure 6.1 includes the sloping parapet on top. The test wall had a level top surface, and the weight of the sloping parapet and roof load was approximated as a uniformly distributed load along the surface. Furthermore, it was required that there be some axial load to stop the loading beam from lifting o. Wall A2 was built with an aspect ratio of 1:1 and with an axial load of 160 kN (350 kPa), which was designed to simulate the load conditions on the bottom storey of a side wall in a multi-storey URM structure (Typology E or F). On such a building, this axial load is made up of the weight of the walls above, the wooden oors and ceilings above, live load on the oors (according to Standards New Zealand (2002a)), and a small roof load. Wall A4 was built with an aspect ratio of 1:2 and with an axial load of 30 kN (33 kPa). This conguration was adopted to investigate the in-plane lateral response of a long wall, which was expected to behave in a shear failure response, rather than exhibiting rocking or exure. Furthermore, subsequent test walls having anges also had an aspect ratio of 1:2, and an axial load of 33 kPa, and this facilitated direct comparison between walls of similar dimensions with and without anges. Experimental testing of walls with anges is reported in Chapter 7.

6.2.2

Wall Construction

All three walls were constructed by experienced blocklayers under supervision (see Figure 6.3). The walls were constructed with common bond pattern and with 1:2:9 mortar
- 139 Alistair P. Russell

Chapter 6. In-Plane Cyclic Testing of Rectangular URM Walls

(a) Wall A1

(b) Wall A2

(c) Wall A4

Figure 6.2: Wall dimensions

Alistair P. Russell

- 140 -

6.2. Construction Details

(a) Wall A1

(b) Wall A4

Figure 6.3: Wall construction

(cement:lime:sand, by volume), corresponding to ASTM type O mortar, and with nominally 10 mm thick mortar joints. There were header bricks every 4th course, alternating on each side of the 3 leaf wall (see Figure 6.2). It was intentionally decided to construct the walls in a way that replicated the observed, often deteriorated, nish quality of walls
- 141 Alistair P. Russell

Chapter 6. In-Plane Cyclic Testing of Rectangular URM Walls

in real New Zealand URM buildings. Testing of Wall A1 began 23 days after construction was complete (referred to as Age in Table 6.1), testing of Wall A2 began 20 days after construction was complete and testing of Wall A4 began 72 days after construction was complete.

6.2.3

Material Properties

The walls were constructed using recycled bricks that were obtained from a demolished building. The original mortar was removed and the surfaces of the bricks were prepared for new mortar before being reused in the test wall. The bricks were estimated to be between 60 and 100 years old. These bricks were used because the manufacturing processes for making new bricks are suciently dierent to substantially alter the properties and characteristics of the bricks. In particular the dierence in porosity between currently manufactured bricks and old bricks means that the bond at the brick-mortar interface is much weaker in new bricks, when using the mortar required for this test. Within a building there is signicant variability in the bricks and reusing these bricks also introduced material variability into the test. Because this variability is natural and realistic, it was considered acceptable.

Nine mortar cubes and six prisms were constructed at the same time as each wall using randomly selected bricks, and were tested on the same day as the wall test was conducted.
100% 90% 80% 70% % PASS 60% 50% 40% 30% 20% 10% PAN 0.075 0.150 0.300 0.600 1.180 2.360 0% 4.750

SIEVE APERTURE (mm)

Figure 6.4: Mortar sand gradation curve

Alistair P. Russell

- 142 -

6.3. Testing Details

This testing was conducted to determine the compression strength of the masonry and the mortar. Additionally, twelve half bricks were selected randomly during construction and tested with the prisms and mortar cubes to determine the brick compression strength. Details of these test methods can be found in ASTM (2003, 2004a, 2007a). Standard Portland cement, hydrated lime (Calcium Hydroxide) and river sand (gradation curve shown in Figure 6.4) were used in the mortar.

Using a standard punch test (6 blows on a standard carpenters punch, with 3 mm tip diameter) as stipulated in the NZSEE guidelines (2006), the mortar in Wall A1 and A2 was classied as good quality mortar, with values of 0.3 and 0.7 for c (cohesion) and (coecient of friction) respectively. The mortar in Wall A4 had values of c = 0.1 and = 0.7.

At the conclusion of each test, an in-place shear test was conducted on an undamaged bed joint to determine the cohesive strength of the masonry bed joints (vme ). The description and results of the tests are reported elsewhere (Lumantarna, 2012), and the values are used with permission.

The material properties of all three walls are summarised in Table 6.1.

6.3
6.3.1

Testing Details
Test Setup and Instrumentation

The typical wall setup is shown in Figure 6.5. The walls were loaded laterally by means of a hydraulic actuator reacting against the laboratory strong wall. A steel channel (referred to here as the loading beam) was mortared to the top of the wall, and the lateral forces were transferred through plates welded to the underside of the loading beam. The plates extended approximately 100 mm below the loading beam, and thus the applied horizontal force was transferred into the wall through friction on the top surface and also directly into the top course of the wall. This arrangement is typical for pseudo-static wall tests
- 143 Alistair P. Russell

Chapter 6. In-Plane Cyclic Testing of Rectangular URM Walls

at The University of Auckland and is also representative of tests reported in literature (Abrams, 1997; Bruneau, 1994b; Gu et al., 2003; Marcari et al., 2003; Santa Maria et al., 2006). The wall was stabilised from moving in its out-of-plane direction by two parallel horizontal struts which were positioned perpendicular to the wall and hinged to the loading beam and a reaction frame.

Extensive instrumentation was installed on each of the three walls to capture all facets of wall response, with a typical layout shown in Figure 6.6. Instruments with a designation beginning with A F measured shear distortion on the wall, J measured exure, G measured any uplift in the bottom two corners of the wall relative to the ground, S measured sliding between the wall and the ground and between the wall and the loading beam, and DISP refers to the measurement of overall wall displacements. LCF represents the load cell measuring the horizontal force applied to the wall top and LCA refers to the load cells measuring the axial load on the wall.

The force between the hydraulic jack and the wall was recorded through a load cell,

Figure 6.5: Test setup

Alistair P. Russell

- 144 -

6.3. Testing Details

Figure 6.6: Instrumentation

Figure 6.7: Imposed cyclic displacement history

- 145 Alistair P. Russell

Chapter 6. In-Plane Cyclic Testing of Rectangular URM Walls

and accurately recorded the force being applied at that point. The control displacement (DISP1) was recorded at the opposite end of the wall, between the top of the wall (at a point attached to the bricks immediately below the bottom surface of the loading beam) and an independent frame, to eliminate any eects from exing of the strong-wall. This instrumentation reading was obtained using a portal gauge with a displacement range of 50 mm.

The axial load on each wall was eected by four post-tensioning tendons positioned approximately 400 mm from each end of the wall. Two steel box beams sat on top of the loading beam, perpendicular to the length of the wall, and the tendons passed through these box beams. The box beams were on pivots, so that the load in the tendons on either side of the wall was equal. Springs at the top of the tendons ensured that rotation was not restrained in the in-plane direction.

6.3.2

Test Procedure

The cyclic loading sequence adopted for all tests was that shown in Figure 6.7 and consisted of a series of displacement-controlled components. Each stage of loading consisted of one cycle to the selected tip displacement. This displacement-controlled pseudo-static procedure was employed to capture the non-uniformly accumulated damage in the wall, and to enable observations of damage and failure mechanisms. The loading history was based on loading histories previously used in pseudo-static wall testing at The University of Auckland, which were originally based on those proposed by Park (1989) for pseudostatic laboratory experimentation. Each wall was subjected to displacements of 0.5 mm in each direction, then 1 mm in each direction and then the displacements were increased by increments of 1 mm each cycle until 10 mm was reached. For cycles with displacements above 10 mm, the displacements were increased by increments of 2 mm. During the rst stages of testing, the potential for shear failure of the wall meant that small displacement increments were necessary in order to avoid the wall being loaded to failure at an early stage of testing. The push direction was dened as positive and the pull direction as negative.
Alistair P. Russell

- 146 -

6.3. Testing Details

In each case, wall testing was terminated when sucient strength degradation was evident. This degradation usually coincided with a marked change in the slope of the force-displacement curve. The walls were deemed to have failed when the maximum wall strength obtained in a particular cycle was less than 80% of the maximum recorded during testing. The ultimate displacement capacity, du , is the displacement corresponding to failure. In Wall A1 and A2, sucient strength degradation did not occur, and the tests were terminated because of other experimental requirements (see Section 6.4.1).

6.3.3

Predicted Flexural and Shear Strength

The New Zealand Society for Earthquake Engineering (NZSEE, 2006) provides guidelines for predicting the nominal in-plane lateral strength of URM walls, Vn . The strength limits based on sliding shear (Vs ), diagonal tension failure due to damage in mortar joints (Vj ), diagonal tension failure due to cracking through bricks (Vb ), and exural resistance (rocking, Vr ), are given below in Equations (6.1) (6.4). The maximum nominal shear strength is given as the lowest of the strength limits, as shown in Equation (6.5).

Vs =

3czbw + N 1+ 3clw bw c N

(6.1)

Vj =

clw bw + N 1 + c fbt lw bw (fbt lw bw + N ) 2.3(1 + c ) 1 N he z N

(6.2)

Vb =

(6.3)

Vr =

b 2 0.85fm w

(6.4)

Vn = MIN(Vs , Vj , Vb , Vr )

(6.5)

Vs is the limit intended to control sliding shear at the end of the member, and is derived
- 147 Alistair P. Russell

Chapter 6. In-Plane Cyclic Testing of Rectangular URM Walls

from equilibrium of forces acting over the compression zone of the cross-section. The neutral axis depth is calculated from an assumed linear variation of stress with depth, and assuming that cohesion is only eective over the neutral axis depth. Vj is the limit intended to reect crushing in mortar joints, and is based on a Mohr-Coulomb failure criterion applied over the full length of the member. Vb is the limit intended to reect shear associated with diagonal tension failure involving cracking through bricks. Vr is the rocking strength, neglecting the tension strength of bed-joints, and assumes an equivalent rectangular compression stress block depth of 0.85 times the neutral axis depth. c = M/V lw , and is the eective aspect ratio (calculated at the point of maximum moment) and z is the distance from the compression edge of the wall to the line of action of N, assumed to be lw /2.

Equations for predicting strength using ASCE (2007) are given in equations (6.6) (6.9), for bed-joint-sliding (Vs ), toe crushing (Vtc ), diagonal tension (Vdt ) and rocking (Vr ). The maximum nominal shear strength is given as the lowest of the strength limits, as shown in Equation (6.10).

lw fm Vdt = vme An 1 + he vme Vr = 0.9N lw he

Vtc = N

Vs = vme An

(6.6) lw he 1 fm
0.7fm

(6.7)

(6.8)

(6.9)

Vn = MIN(Vs , Vtc , Vdt , Vr )

(6.10)

Vs is the bed-joint sliding strength, based on An , the net mortared cross-sectional area of the wall, and vme , the cohesive strength of a masonry bed joint determined from results of an in-place shear test, Vtc is the shear strength of the wall based on the toe compresAlistair P. Russell

- 148 -

6.3. Testing Details

sive strength (toe crushing), Vdt is the limiting strength of the wall based on diagonal tension failure and Vr is the wall strength based on rocking. is a factor equal to 0.5 for a xed-free cantilever wall, or equal to 1.0 for a xed-xed pier. = 0.5 corresponds to a moment of 0 at the top of the wall and = 1 corresponds to the moment at the top of the wall being equal to the moment at the bottom of the wall1 . In the case of the walls reported here, = 0.5 was appropriate. ASCE (2007) denes bed-joint sliding and rocking responses as deformation controlled, whilst wall strengths limited by diagonal tension or toe-crushing are dened as force-controlled.

Whilst NZSEE provides equations to determine diagonal tension strength limits in terms of both damage to mortar joints (Equation (6.2)) (see Figure 5.7(b)) and cracking through bricks (Equation (6.3)) (see Figure 5.7(a)), ASCE refers only to a diagonal tension failure limit. Similarly, ASCE provides equations to determine both the rocking and toe-crushing strength limits, whilst NZSEE provides only one equation to determine the exural strength limit. The intention of the diagonal tension failure given by ASCE (2007) is that a crack will form through the masonry units (bricks). But for stair-stepped cracks (cracking through the mortar) to occur, the diagonal tension strength of the masonry as a whole (fdt ) (the strength of the masonry head- and bed-joints together) is exceeded by the maximum principal tension stress (1 ). Hence, while diagonal tension is termed a force-controlled action, sliding can occur on a plane which is along either a continuous bed-joint or from stair-stepped sliding along several bed-joints, and there is further displacement capacity, which can be considered deformation controlled. A further distinction made by ASCE (2007) for Equations (6.6) (6.9) is that the expected strength values are to be used in conjunction with displacement-controlled actions, whereas lowerbound values are to be used with force-controlled actions. The limiting wall strengths as predicted using NZSEE (2006) and ASCE (2007) are shown in Table 6.2.

Equation (5.15) is similar to Equation (6.8) for determining the strength of the wall failing in diagonal tension, except that the equation proposed by ASCE (2007) (Equation (6.8))
ASCE (2007) notes that the condition of a wall pier with full restraint against rotation at its top and bottom is often not present in actual buildings.
1

- 149 -

Alistair P. Russell

Chapter 6. In-Plane Cyclic Testing of Rectangular URM Walls

Table 6.2: Predicted wall strengths Wall Vs kN A1 A2 A4 Vj kN NZSEE Vb kN Vr kN Vn kN Vs kN 25.6 ASCE 41-06 Vtc kN 19.9 Vdt kN 97.6 Vr kN Vn kN

19.9 71.2 365.7 19.9 19.9 85.6 88.8 166.4 83.5 83.5 71.3 53.1

18.0 18.0

102.5 84.6 100.6 78.3 78.3 46.4 71.6 52.4 64.8 46.4

53.1 56.2 506.7

includes a denominator of he /lw , which for a wall such as Wall A4, where he /lw = 0.5, increases the expected shear strength by a factor of 2. Equation (5.15) uses a value of b to account for the aspect ratio and is suggested as b = 1 for h/lw 1, b = h/lw for 1 < h/lw < 1.5, b = 1.5 for h/lw 1.5, as reported in Section 5.3. Thus for long walls such as Wall A4, the expected shear strength is multiplied by a factor of 1 in Equation (5.15) to account for the aspect ratio. This modication is considered more reasonable than multiplying by lw /he , and agrees more accurately with the results presented below. Correspondingly, where Equation (6.8) predicts Vdt = 104.9 kN, Equation (5.15) predicts Vdt = 52.4 kN, which is the value shown in Table 6.2.

The behaviour mode corresponding to the nominal shear strength (Vn ) of each wall is shown in Table 6.3. The equations provided by ASCE (2007) predicted that rocking would be the governing behaviour mode for Wall A1 and A2, and that toe crushing would occur subsequently in Wall A2. Sliding shear was predicted to occur in Wall A4 rst, and at a slightly higher value of lateral force (Vdt = 52.4 kN), diagonal tension was predicted Table 6.3: Predicted wall behaviour mode Wall A1 A2 A4 NZSEE Rocking/Sliding Rocking or Mortar joint damage or Sliding Sliding/mortar joint damage
- 150 -

ASCE 41-06 Rocking Rocking followed by toe crushing Sliding/diagonal tension

Alistair P. Russell

6.4. Test Results

to occur. NZSEE (2006) predicted that for Wall A1 rocking or sliding would occur at the same level of lateral force, and for Wall A2, with a higher level of axial load, the lateral force corresponding to sliding, rocking and diagonal shear failure due to damage in the mortar joints were at similar levels. For Wall A4, sliding was predicted to occur at a slightly lower level of lateral force than the corresponding force to initiate diagonal tension failure due to damage in the mortar joints, but the small dierence would mean that a sliding/diagonal tension failure could be predicted. Because diagonal tension failure is a force-controlled (brittle) action it would be safer to predict this mechanism, as the force required to induce dierent failure mechanisms is similar. Consequently, a exure/sliding response was expected for Wall A1 and Wall A2, whilst a sliding/diagonal tension failure was expected for Wall A4.

6.4

Test Results

This section presents the test observations, followed by a description of wall behaviour and performance. General results are presented in Table 6.4 where Vn,NZSEE is the nominal shear strength as predicted by NZSEE (2006), Vn,ASCE is the nominal shear strength as predicted by ASCE (2007), Vmax is the maximum recorded lateral force, dVmax and Vmax are the corresponding displacement and drift, respectively, at Vmax , Vcrack is the lateral force when cracking was rst observed, crack is the drift corresponding to Vcrack , and du and u are the lateral wall displacement and drift, respectively, corresponding to the point at which the lateral force had degraded to 80% of Vmax , where u = du /he .

6.4.1

Observations

Wall rocking was observed for both Wall A1 and Wall A2, characterised by the formation of a horizontal crack several courses above the base of the wall. On Wall A1, the primary crack occurred rst at a drift of 0.09% on the push cycle between the 6th and 7th courses and extended to approximately the mid-point of the wall. In subsequent cycles, this crack extended towards the compression toe of the wall (in both the push and pull directions,
- 151 Alistair P. Russell

Chapter 6. In-Plane Cyclic Testing of Rectangular URM Walls

Table 6.4: Test results Wall Vn,NZSEE kN A1 A2 A4 19.9 83.5 53.5 Vn,ASCE kN 18.0 78.3 46.4 Vmax kN 24.8 96.7 62.6 dVmax mm 44.3 9.6 0.94 Vmax % 1.77 0.48 0.03 Vcrack kN 20.0 51.3 53.2 crack % 0.09 0.03 0.03 du mm 1.4 u % 0.05 Rocking Rocking Shear failure Behaviour

(a) Wall A1

(b) Wall A2

(c) Wall A4

Figure 6.8: Cracking patterns appropriately) and rocking occurred on a contact area approximately 400 mm long and across the full width of the wall (see Figure 6.19(a) and 6.19(b)). On Wall A2, the primary crack occurred rst at a drift of 0.03% on the push cycle between the 5th and 6th courses, and subsequently extended further along the length of the wall. Some damage in
Alistair P. Russell

- 152 -

6.5. Discussion

the compression toe region (toe crushing) was evident in Wall A1 (see Figure 6.19(c)) and damage was evident in the inner leaf at the centre of the wall after the test was terminated and loose bricks in the outer leaf were removed (see Figure 6.19(e)). Signicantly more compression toe damage was evident in Wall A2 (see Figures 6.20(a) and 6.20(b)), as expected with a higher level of axial load. Wall A1 was stable even after the test was terminated at a drift of 1.8% when the maximum stroke capacity of the hydraulic actuator was reached. Wall A2 became unstable at a drift of 1.8% when the lateral supports preventing the wall from moving in the out-of-plane direction were found to be insecure and the wall began to move sideways (see Figure 6.20(f)). The nal cracking pattern is shown in Figure 6.8(a) and 6.8(b) for Wall A1 and Wall A2 respectively, and shows the extent of compression toe damage in each wall.

Wall A4 failed in diagonal tension (see Figure 6.21(e)). Initial cracking was observed in the top corner of the wall where the load was applied, at a displacement of 0.5 mm (drift of 0.03%) in the push cycle, and similar cracking was observed on the pull cycle at the same displacement. At a displacement of 1.5 mm (drift of 0.08%) in the push cycle a crack suddenly occurred beginning at the top corner where the load was applied and extended through the brick-mortar interface to the bottom corner at the opposite end of the wall, and the applied force correspondingly dropped by 30%. Similarly at a displacement of -1.5 mm on the pull cycle a crack occurred from the top corner where the load was applied and extended to the opposite bottom corner of the wall. Figure 6.8(c) shows the nal cracking of Wall A4, and Figures 6.21(a) 6.21(d) show the cracks as they occurred in each direction. There was also some damage evident in the compression toe region of the wall, as shown in Figure 6.21(b).

6.5

Discussion

This section discusses wall performance in terms of lateral strength and compares the obtained values with those predicted using available equations. Further discussion is provided on the the force-displacement response, energy dissipation, methods of approxi- 153 Alistair P. Russell

Chapter 6. In-Plane Cyclic Testing of Rectangular URM Walls

mating the response and the ultimate drifts obtained during testing.

6.5.1

Force-Displacement Response

The force-displacement response of Walls A1, A2 and A4 is presented in Figure 6.9. Walls A1 and A2 exhibited symmetric maximum response in both directions of loading, reaching maximum displacements of 45 mm and 36 mm respectively before testing ceased. In neither case did the wall strength reduce below 80% of the maximum recorded strength, outlined as the failure criteria in Section 6.3.2. Testing of Wall A1 ceased because the stroke capacity of the hydraulic actuator was reached, and testing of Wall A2 ceased when the wall became unstable in the out-of-plane direction. Both Walls A1 and A2 behaved
Drift (%) 0 0.5 Drift (%) 0

!2 30 20 Base Shear (kN) 10 0 !10 !20

!1.5

!1

!0.5

1.5

75 50 Base Moment (kNm) 25 0 !25 !50 Base Shear (kN)

!1.5 120 80 40 0 !40 !80 !120 !30

!1

!0.5

0.5

1.5 200 150 100 50 0 !50 !100 !150 !200 Base Moment (kNm)

Vn,NZSEE = 19.9 kN Vn,ASCE = 18.0 kN

Vn,NZSEE = 83.5 kN Vn,ASCE = 78.3 kN

!30 !50 !40 !30 !20 !10 0 10 20 Displacement (mm)

30

40

!75 50

!20

!10 0 10 Displacement (mm)

20

30

(a) Wall A1
!0.1 80 60 40 Base Shear (kN) 20 0 !20 !40 !0.05 Drift (%) 0 0.05

(b) Wall A2
0.1 150 100 50 0 !50 !100 Base Moment (kNm)

Vn,NZSEE = 53.5 kN Vn,ASCE = 46.4 kN

!60 !80 !2 !150 !1.5 !1 !0.5 0 0.5 1 Displacement (mm) 1.5 2

(c) Wall A4

Figure 6.9: Force-displacement response

Alistair P. Russell

- 154 -

6.5. Discussion

elastically to a drift of approximately 0.1%, after which rocking ensued. Wall A4 failed at a low displacement of 1.5 mm (drift of 0.08%) when the applied lateral force in both directions was reduced below 0.8Vmax .

The force-displacement response of Wall A2 was non-symmetrical (Figure 6.9(b)). This non uniformity was because there was signicant toe-crushing after the wall reached a lateral displacement of 10 mm, and bricks in the compression toe region dislodged at a dierent rate in each direction. This observation meant that when the displacement on the pull cycle returned to zero, there was a residual force applied to the wall. This behaviour also gave the appearance that there was a residual displacement on the push cycle when the force was returned to zero.

Yi et al. (2005) note that as cracks propagate further through the length of the wall, the rocking strength increases, as the eective length on which the wall is rocking decreases. This is because the eective length is dependent on the moment acting at the rocking section. Correspondingly, as the lateral displacement demand on a wall increases, the eective length decreases, and this behaviour can be observed as the increase in peak lateral force in Figures 6.15(c) and 6.16(c).

The three lateral force-displacement response curves reported in this section are used to determine the energy dissipation characteristics of the walls, as well as to determine an approximate bilinear representation, as shown below.

6.5.2

Energy Dissipation

The energy dissipated in one cycle of a pseudo-static cyclic test, ED , is calculated as the area under the individual force-displacement loop for that cycle, as per Equation (6.11). ED = fD dd (6.11)
Alistair P. Russell

- 155 -

Chapter 6. In-Plane Cyclic Testing of Rectangular URM Walls

Figure 6.10: Energy dissipated ED in one force-displacement loop (from (Chopra, 2007)) The equivalent viscous damping ratio (eq ) can be approximated by determining the equivalent strain energy, ES , dissipated during each cycle, as shown in Figure 6.10 and Equation (6.12), where do is the average of the maximum displacements in the positive and negative directions of the cycle, and k is the secant stiness to the corresponding maximum force. The equivalent viscous damping ratio is then determined according to Equation (6.13). This relationship is strictly only true when the structure is excited at its natural frequency ( = n ), but is generally considered a satisfactory approximation at other frequencies (Chopra, 2007), and is a useful way of determining the energy dissipation characteristics in such a way that is comparable between walls tested and reported here. 1 2

ES = eq =

kdo 2

(6.12) (6.13)

1 ED 4 ES

Observation of the individual force-displacement loops indicates pinched loops in the case of Wall A1 and A2, implying little energy dissipation, but fatter loops for Wall A4. The large number of channels used on the data-logging equipment during the test of Wall A1 resulted in a slow refresh rate, and meant that the unloading portions on the forcedisplacement plot were not accurately recorded. Consequently, the energy dissipation
Alistair P. Russell

- 156 -

6.5. Discussion

Equivalent Viscous Damping Ratio, !eq

0.25 0.2 0.15 0.1 0.05 0

Equivalent Viscous Damping Ratio, !eq 0 0.25 0.5 0.75 Drift (%) 1 1.25 1.5

0.3

0.3 0.25 0.2 0.15 0.1 0.05 0

0.02

0.04 0.06 Drift (%)

0.08

0.1

(a) Wall A2

(b) Wall A4

Figure 6.11: Equivalent viscous damping ratio of Wall A2 and A4 characteristics of Wall A1 could not be accurately determined. The equivalent viscous damping ratios of Walls A2 and A4 are shown in Figure 6.11.

From Wall A4 only four cycles were available from which to determine the energy dissipation and equivalent viscous damping ratio, but it is nevertheless evident that the equivalent viscous damping ratio for the rocking response of Wall A2 was signicantly lower than for the shear response of Wall A4.

From the results of a single test (Wall A2) it was observed that the equivalent viscous damping ratio was low, and an equivalent viscous damping ratio of eq = 0.05 could be recommended for URM walls responding in exure.

6.5.3

Bilinear Approximation

For assessment purposes it is useful to idealise the force-displacement behaviour of in-plane URM walls with a bilinear approximation, where the bilinear envelope is chosen such that the areas under the curves (and energy dissipated) are the same at a selected value of maximum displacement. Magenes and Calvi (1997) note that the response of brick masonry walls is highly non-linear and a denition of yield displacement, and consequently elastic
- 157 Alistair P. Russell

Chapter 6. In-Plane Cyclic Testing of Rectangular URM Walls

Figure 6.12: Equivalent bilinear approximation (from Magenes and Calvi (1997))
Drift (%) 1 Drift (%) 0.75

30 25

0.5

1.5

75

120 100 Base Moment (kNm)

0.25

0.5

1.25

1.5

200 Base Moment (kNm)

Base Shear (kN)

20 15 10 5 0

50

Base Shear (kN)

80 60

150

100 40 20 50

25

10

20 30 Displacement (mm)

40

0 50

10 15 20 Displacement (mm)

25

0 30

(a) Wall A1
0 0.02 Drift (%) 0.04 0.06 0.08 0.1

(b) Wall A2

80 70

150 Base Moment (kNm) 60 Base Shear (kN) 50 40 30 50 20 10 0 0 0.5 1 1.5 Displacement (mm) 2 25 0 125 100 75

(c) Wall A4

Figure 6.13: Equivalent bi-linear response of Wall A1, A2, A4

Alistair P. Russell

- 158 -

6.5. Discussion

stiness, must be related to a reference stress or displacement. The approach suggested in Magenes and Calvi (1997) is to select the initial stiness as the secant stiness at 0.75Vu , and it was suggested elsewhere (Toma zevi c, 1996) that Vu = 0.9Vmax is appropriate for energy equivalence in masonry walls failing in shear. Thus the bilinear approximation is constructed such that the equivalent elastic displacement, de , is at the intersection of the line connecting the origin to 0.75Vu on the force-displacement curve, and the ultimate wall displacement, du , is dened as the displacement corresponding to a strength degradation of 0.8Vu (see Figure 6.12). Figure 6.13 shows the equivalent bilinear envelope plot of Walls A1, A2 and A4.

6.5.4

Multi-Linear Approximation

Similar to a bilinear approximation, ASCE (2007) recommends that the response of structural elements can be approximated with idealised multi-linear force-displacement curves, of which three dierent types are suggested, depending on the behaviour mode of the element (deformation- or force-controlled, see Chapter 1). The method for generating such curves is given in Section 2.8.3 of ASCE (2007). A smooth backbone curve is drawn as an envelope of the hysteretic response, which is then approximated by a series of linear segments, drawn to form a multi-linear segmented curve conforming to one of the three curves reproduced in Figure 6.14. The positive and negative branches of the multilinear force-displacement approximation is then normalised by averaging the slopes of each respective segment, and applying this stiness to a region bounded by the average of the drifts of the positive and negative contributing segments. The construction of the multi-linear approximations for Walls A1, A2 and A4 are shown in Figures 6.15, 6.16 and 6.17 respectively.

Walls A1 and A2 both t a Type 2 curve, according to Figure 6.14, and can be considered deformation controlled, despite the signicant toe crushing that occurred in Wall A2.

The response of Wall A4, failing in shear, ts a Type 3 curve according to Figure 6.14, and is considered force-controlled. A force-controlled response, as depicted by a Type 3
- 159 Alistair P. Russell

Figure 2-3 Component Force Versus Deformation Curves Chapter 6. In-Plane Cyclic Testing of Rectangular URM Walls

Figure 6.14: Component forces versus displacement curves (reproduced from ASCE (2007))

!2 30 20 Base Shear (kN) 10 0 !10 !20

!1.5

!1

!0.5

Drift (%) 0 0.5

1.5

75 50 25 0 !25 !50 Base Moment (kNm) Drift (%) 1

!30 !50 !40 !30 !20 !10 0 10 20 Displacement (mm)

30

40

!75 50

(a) Force-displacement envelope

!2 30 20 Base Shear (kN) 10 0 !10 !20 !30 !50 !40 !30 !20 !10 0 10 20 Displacement (mm) !1.5 !1 !0.5 Drift (%) 0 0.5 1 1.5 2 0 0.5 1.5 Normalized Average 2

75 50 Base Moment (kNm) 25 0 !25 !50 Base Shear (kN)

30 25 20 15 10 5 0

75

50

25

30

40

!75 50

10

20 30 Displacement (mm)

40

0 50

(b) Multi-linear force-displacement approximation (c) Average multi-linear force-displacement approximation

Figure 6.15: Multi-linear approximation of Wall A1

Alistair P. Russell

- 160 -

Base Moment (kNm)

6.5. Discussion
Drift (%) 0

!1.5 120 80 Base Shear (kN) 40 0 !40 !80 !120 !30

!1

!0.5

0.5

1.5 200 150 100 50 0 !50 !100 !150 !200 Base Moment (kNm)

!20

!10 0 10 Displacement (mm)

20

30

(a) Force-displacement envelope

!1.5 120 80 Base Shear (kN) 40 0 !40 !80 !120 !30 !1 !0.5 Drift (%) 0 0.5 1 1.5 200 150 Base Moment (kNm) Base Shear (kN) 100 50 0 !50 !100 !150 !200 !20 !10 0 10 Displacement (mm) 20 30 0 0 5 10 15 20 Displacement (mm) 25 0 30 100 80 60 100 40 20 50 200 Base Moment (kNm) 0 0.25 0.5 Drift (%) 0.75 1 1.25 1.5

120

150

(b) Multi-linear force-displacement approximation (c) Average multi-linear force-displacement approximation

Figure 6.16: Multi-linear approximation of Wall A2

curve in Figure 6.14, is representative of a brittle or nonductile behaviour where there is an elastic range followed by loss of strength and loss of ability to support gravity loads beyond the peak force.

Figure 6.18 shows both the approximated bilinear response and the multi-linear response of Wall A4, and it can be seen that the multi-linear response allows a similar equivalent elastic-plastic approximation to be made, but provides further information on the postpeak force response.

- 161 -

Alistair P. Russell

Chapter 6. In-Plane Cyclic Testing of Rectangular URM Walls

Drift (%)
!0.1 80 60 40 Base Shear (kN) 20 0 !20 !40 !100 !60 !80 !2 !150 !1.5 !1 !0.5 0 0.5 1 Displacement (mm) 1.5 2 50 0 !50 !0.05 0 0.05 0.1 150 100

(a) Force-displacement envelope

Drift (%)
!0.1 80 60 !0.05 0 0.05 0.1 150 100 40 Base Shear (kN) 20 0 !20 !40 !100 !60 !80 !2 !150 !1.5 !1 !0.5 0 0.5 1 Displacement (mm) 1.5 2 10 0 0 0.5 1 1.5 Displacement (mm) 2 50 0 !50 80 70 0 0.02

Drift (%)
0.04 0.06 0.08 0.1 Normalized 150 Average

Base Moment (kNm)

Base Moment (kNm)

60 Base Shear (kN) 50 40 30

100 75 50

20 25 0

(b) Multi-linear force-displacement approximation (c) Average multi-linear force-displacement approximation

Figure 6.17: Multi-linear approximation of Wall A4

Drift (%)
80 70 60 Base Shear (kN) 50 40 30 50 20 10 0 0 0.5 1 1.5 Displacement (mm) 2 25 0 0 0.02 0.04 0.06 0.08 0.1 150 70 80 0 0.02 Drift (%) 0.04 0.06 0.08 0.1 150

Base Moment (kNm)

100 75

50 40 30

100 75 50

20 10 0 0 0.5 1 1.5 Displacement (mm) 2 25 0

(a) Multi-linear force-displacement approximation

(b) Bilinear force-displacement approximation

Figure 6.18: Approximated response of Wall A4

Alistair P. Russell

- 162 -

Base Moment (kNm)

125

60 Base Shear (kN)

125

Base Moment (kNm)

125

6.5. Discussion

6.5.5

Ultimate Drift

Walls A1 and A2 exhibited high ultimate drifts with no corresponding drop in lateral force. In Chapter 1 it was suggested that a drift limit of 0.8% for a exural response could be considered reasonable for the purpose of limiting non-structural damage in a building. Walls A1 and A2 both surpassed this limit, although there was some damage due to toe-crushing in Wall A2 at a drift of 0.6% (see Figures 6.20(a) and 6.20(b)). Wall A2 became laterally unstable at a drift of 1.8%, and at this drift its capacity to support gravity loads was lost and the wall was in danger of collapse2 . At a drift of 0.8%, both Wall A1 and A2 were not in danger of collapse and had not lost the capacity to support gravity loads. As such, for walls responding in a exure behaviour mode a drift limit of 0.8% has been conrmed.

In Chapter 1 a drift limit of 0.4% for walls failing in shear was proposed. Wall A4 failed at a drift of 0.05%, although the test was not continued further beyond this drift limit and the gravity load carrying capacity and collapse potential could not be conrmed. Further testing would be required to conrm this drift limit for straight rectangular walls with an aspect ratio of 1:2. Wall A4 is nevertheless used as a comparison for walls tested with anges (see Chapter 7).

6.5.6

Predicted Behaviour and Measured Behaviour

The response of Wall A1 was predicted accurately by both ASCE (2007) and NZSEE (2006), where a rocking and sliding response was predicted. Rocking initiated at a force of 20 kN, and was predicted by ASCE (2007) and NZSEE (2006) to occur at 20 kN and 18 kN, respectively. The strength capacity increased with increasing displacement demand as expected (see Section 6.5.1). For straight walls with low axial load and a 1:1 aspect ratio, the available guidelines are satisfactory to predict response. Additionally, a drift limit of 0.8% was conrmed as suitable for exure dominated walls, and as such, for
The Building Act 2004 (New Zealand Parliament) denes a building as potentially earthquake prone if in a moderate earthquake. . . the building would be likely to collapse. . .
2

- 163 -

Alistair P. Russell

Chapter 6. In-Plane Cyclic Testing of Rectangular URM Walls

walls responding in exure a secant stiness to ultimate conditions can be dened.

The response of Wall A2 was predicted by NZSEE (2006) to be a mixture of rocking, sliding and diagonal tension failure due to damage in the mortar joints. ASCE (2007) predicted that rocking would occur, followed by toe crushing. ASCE (2007) successfully predicted the mode of behaviour, although the lateral force at which it was predicted to occur was less accurate than for Wall A1, where the axial load was lower. The predicted response using NZSEE (2006) was less adequate, and no diagonal tension failure due to damage in the mortar joints occurred. Nevertheless, the lateral force corresponding to failure was a sucient estimate to model the actual response. For walls with an aspect ratio of 1:1, a shear failure is unlikely, despite a high level of axial load. The axial load on Wall A2 was as high as could be expected on any wall in a New Zealand URM building (see Chapter 3), and yet a shear failure did not occur. Consequently, it is concluded that aspect ratio has a more signicant eect on determining the behaviour mode of URM walls than does axial load level. It is also concluded that the response of straight walls with an aspect ratio of 1:1 can satisfactorily be predicted by both ASCE and NZSEE.

Wall A4 was predicted by both ASCE (2007) and NZSEE (2006) to behave in a sliding mechanism, but diagonal tension failure instead occurred rst. NZSEE (2006) predicted that diagonal tension failure following sliding would occur due to damage in the mortar joints, whereas ASCE (2007) does not distinguish the type of diagonal tension failure. After cracking occurred due to diagonal tension, a plane opened on which sliding could take place (see Figure 6.21). The force corresponding to diagonal tension failure was estimated by ASCE (2007) to be signicantly higher than was observed in the actual response, but was accurately estimated using Equation (5.15), as outlined in Section 6.3.3. Consequently, Equation (5.15) should be used instead of Equation (6.8) for determining Vdt . ASCE (2007) states that lw /he should not be taken as less than 0.67 in Equation (6.8), and from the results of this test it is recommended that it also not be taken as greater than 1. This recommendation will eectively align Equation (6.8) with Equation (5.15), taking into account the parameter b, as outlined in Section 6.3.3.

Alistair P. Russell

- 164 -

6.5. Discussion

6.5.7

Consolidation of Predictive Equations

As stated above, it was concluded that the response of rectangular walls dominated by exure can satisfactorily be predicted by the guidelines available from both NZSEE and ASCE. NZSEE predicts the exural response of URM walls by consolidating the expected strength due to both rocking and toe crushing into a single equation, whereas ASCE requires separate equations for toe crushing and rocking. All three equations are adequate to predict the response, and as such, a recommendation is made here to use the predictive equation provided by NZSEE, due to the fact that a single equation can replace two separate equations. Consequently, Equation (6.14) is recommended to predict the exural response of rectangular URM walls. N he 1 N

Vr =

b 2 0.85fm w

(6.14)

The predictive equations provided by NZSEE for determining the lateral strength of rectangular URM walls failing in diagonal tension are distinguished according to whether the cracking occurs in the bricks or in the mortar joints. Equation (6.3) (diagonal tension cracking through the bricks) signicantly overestimated the strength of Wall A4 compared with Equation (6.2) (diagonal tension cracking through the mortar joints) and it has been observed in Chapter 5 and elsewhere (Derakhshan et al. (2010); Dizhur et al. (2009, 2010a,b) and Dizhur and Ingham (2010)) that for existing URM buildings in New Zealand the quality of mortar is usually poorer than the quality of bricks, and as such cracking through the mortar joints is more likely. ASCE provides a single equation for determining the lateral strength of a wall failing in diagonal tension (Equation (6.8)), and as for exure (above) it is recommended that a single equation in place of two separate equations is preferable. As stated in Section 6.5.6, Equation (6.8) is adequate for predicting the lateral strength of straight URM walls, with the modication of 0.67 lw /he 1. Consequently, Equation (6.15) is recommended to predict the diagonal tension failure strength of rectangular URM walls. fm lw Vdt = vme An 1 + he vme
- 165 -

0.67

lw he

(6.15)

Alistair P. Russell

Chapter 6. In-Plane Cyclic Testing of Rectangular URM Walls

Walls A1 and A2 were both predicted to respond in sliding and rocking by NZSEE and in rocking only by ASCE. The two predictive equations provided by NZSEE and ASCE (Equations (6.1) and (6.6) respectively) require dierent input parameters. Equation (6.6) requires the use of vme as determined from in-situ material testing. In the absence of in-situ material testing, c and can be estimated adequately (for further details see the companion study of Lumantarna (2012)). As such, if in-situ material testing is conducted, then Equation (6.17) is recommended to predict the sliding shear strength of rectangular URM walls. In the absence of material test data, Equation (6.16) is recommended to predict the sliding shear strength of rectangular URM walls. 3czbw + N 1+ 3clw bw c N (6.17)

Vs =

(6.16)

Vs = vme An

In this chapter, existing equations for determining the limiting strengths of straight rectangular walls without anges are recommended for the reason that they are already in published guidelines (NZSEE (2006) and ASCE (2007)), and it was thought most useful to keep the equations in their current published form, with the exception of Equation (6.15) where the modications are given above.

6.6

Conclusions

Magenes and Calvi (1997) state that for walls responding in a exural behaviour mode, large displacements can theoretically be obtained without signicant loss in strength, especially when the mean axial load is low compared to the compressive strength of masonry. This statement was conrmed from the response of Walls A1 and A2. For rectangular URM walls with low axial load and a 1:1 aspect ratio (Wall A1), guidelines available from both ASCE (2007) and NZSEE (2006) are satisfactory to predict response. For walls with higher axial load (Wall A2), the prediction of lateral force corresponding to cracking and the initiation of rocking is less accurate, but nevertheless satisfactory.
Alistair P. Russell

- 166 -

6.6. Conclusions

Whilst the force at which cracking/rocking initiates can be predicted, this does not predict the failure force as the walls had not failed. Walls responding in exure are deformation controlled and it was conrmed in this chapter that a drift limit of 0.8% is a satisfactory limit at which gravity load carrying capacity is not compromised and the walls are expected to not collapse. If the ultimate drift limit is dened as 0.8% and the force corresponding to rocking is determined, then a secant stiness to ultimate conditions can be dened for a wall responding in exure. Further details on the exural response of URM wall piers can be found in the companion study of Knox (2012).

Rocking is a mechanism in which energy can be dissipated after the onset of cracking. Damage in this instance is primarily related to the level of axial load, and a wall behaving in this mode can resist lateral force after the onset of rst cracking. From the results of a single test (Wall A2) it was observed that the equivalent viscous damping ratio was low, and that an equivalent viscous damping ratio of 5% is recommended for URM walls responding in exure.

For Wall A4 the guidelines available from NZSEE (2006) predicted the limiting force satisfactorily, although a sliding shear failure was predicted to occur at a lower level of lateral force than a diagonal tension failure. Because diagonal tension failure is a force-controlled (brittle) action it would be safer to predict this mechanism, as the force required to induce the two failure mechanisms was similar. The guidelines available from ASCE (2007) overestimated the force corresponding to diagonal tension failure due to a factor erroneously accounting for aspect ratio. Consequently, it is suggested that for use in Equation (6.8), the following limitation is appropriate, 0.67 < he /lw < 1.

The drift capacity of Wall A4 corresponding to failure was 0.05%, which is lower than values for ultimate drift reported in the literature, and consequently for rectangular walls with an aspect ratio greater than 1, further testing is required to determine an ultimate drift limit.

The energy dissipation of Wall A4 was higher than the energy dissipation of the walls
- 167 Alistair P. Russell

Chapter 6. In-Plane Cyclic Testing of Rectangular URM Walls

responding in exure. The equivalent viscous damping ratio of Wall A4 was close to 10%, but this was only a single test with few hysteretic cycles. This test result agrees with the 10% equivalent viscous damping ratio reported in the literature by Magenes and Calvi (1997) for rectangular URM walls failing in shear.

It was also concluded that aspect ratio is more important in changing the failure mode than the axial load. This contrasts with the ndings of Steelman and Abrams (2007), who concluded that both axial load and aspect ratio are important in determining in-plane wall failure mode.

Equations (6.14) (6.17) were recommended for predicting the response of rectangular URM walls, as follows.

Vr =

N he

lw fm 1+ Vdt = vme An he vme Vs = 3czbw + N 1+ 3clw bw c N

b 2 0.85fm w

(6.14)

0.67

lw he

(6.15)

(if no in-situ material data are available)

(6.16)

Vs = vme An

(if in-situ material data are available)

(6.17)

Alistair P. Russell

- 168 -

6.7. Photos of Wall Response

6.7

Photos of Wall Response

(a) Uplift on push cycle at displacement of 16 mm (b) Uplift on pull cycle at displacement of -20 mm

(c) Toe crushing on pull cycle at displacement of -22 mm

(d) Uplift on push cycle at 32 mm

(e) Damage in the centre of the wall at test termination

Figure 6.19: Photos of Wall A1 response

- 169 -

Alistair P. Russell

Chapter 6. In-Plane Cyclic Testing of Rectangular URM Walls

(a) Toe crushing on push cycle at dis- (b) Toe damage on pull cycle at displacement of 12 mm placement of -12 mm

(c) Toe damage on pull cycle at displacement of (d) Uplift on pull cycle at displacement of -32 mm -20 mm

(e) Wall damage at test termination

(f) Out-of-plane instability at test termination

Figure 6.20: Photos of Wall A2 response

Alistair P. Russell

- 170 -

6.7. Photos of Wall Response

(a) Shear crack on pull cycle at displacement of (b) Toe damage on pull cycle at displacement of -1.5 mm -1.5 mm

(c) Shear crack on push cycle at displacement of (d) Shear crack on push cycle at displacement of 1.5 mm 1.5 mm

(e) Wall damage at test termination

Figure 6.21: Photos of Wall A4 response

- 171 -

Alistair P. Russell

Chapter 6. In-Plane Cyclic Testing of Rectangular URM Walls

Alistair P. Russell

- 172 -

CHAPTER 7

In-Plane Cyclic Testing of Flanged URM Walls

This chapter presents the results of pseudo-static in-plane testing of six unreinforced masonry walls with anges. Flanges are walls oriented perpendicular to the direction of applied lateral force, and are sometimes also termed return walls. The six walls had varying ange lengths and plan geometries.

The aim of the experimentation reported in this chapter was to investigate the eects of anges on the response of URM walls, and was achieved by testing six URM walls with dierent ange characteristics. The rst two walls (designated Wall A3 and Wall A3a) were 2000 mm long, and Wall A3 was 2000 mm high, with an aspect ratio of 1:1, whilst Wall A3a was 1250 mm high, with an aspect ratio of 1:1.6. The remaining 4 walls (designated Wall A5, Wall A6, Wall A7 and Wall A8) were 4000 mm long and 2000 mm high with an aspect ratio of 1:2. Walls A3, A3a and A5 had anges at both ends with a ange length of 480 mm on either side of the in-plane wall, and a total ange length of 1200 mm. Walls A6 and A7 had ange lengths of 960 mm on either side of the inplane wall, with a total ange length of 2160 mm. The anges were at both ends in the
- 173 Alistair P. Russell

Chapter 7. In-Plane Cyclic Testing of Flanged URM Walls

case of Wall A6, and at one end in the case of Wall A7. Wall A8 had anges at both ends, but on a single side of the in-plane wall only, and the anges had a length of 960 mm.

Wall A3 was tested to compare directly the response of rectangular walls with aspect ratios of 1:1 (Walls A1 and A2) with the response of a anged wall, also with an aspect ratio of 1:1. Wall A3a was a retest of Wall A3 with part of the top of the wall removed, and with a higher axial load applied. Walls A5, A6, A7 and A8 were tested to investigate the response of walls with dierent ange details and also to compare directly the response of rectangular walls with an aspect ratio of 1:2 (Wall A4) against the response of walls with the same aspect ratio, but with anges.

An associated aim of the work reported in this chapter was to acquire experimental results from tests on in-plane URM anged walls, against which the predicted response could be compared. The response was predicted using an analytical model proposed by Yi et al. (2008) for determining the in-plane response of walls with anges.

7.1

Background

Yi et al. (2008) noted that previous experimental research at the structural level highlighted the eects of transverse walls (anges) on the response of in-plane walls and indicated the potential for anges to inuence pier failure modes and maximum strength (Abrams and Costley, 1994; Calvi and Magenes, 1994; Costley, 1996; Moon et al., 2006; Paquette and Bruneau, 2003; Yi et al., 2006b). Yi et al. (2008) also noted that that no experimental data were available which specically investigate anged URM walls.

Priestley et al. (2007) stated that, A proper connection between perpendicular walls may be important in existing [URM] buildings, where oor slabs [which are usually exible timber diaphragms] may be inadequate to provide any diaphragm action to redistribute the horizontal forces to the wall system and to provide adequate out-of-plane restraints. Consequently, the eect of anges when analysing the seismic performance of a URM
Alistair P. Russell

- 174 -

7.1. Background

building should not be neglected.

Shedid et al. (2008) investigated the eects of anged reinforced concrete masonry walls and determined that the presence of anges increased the available ductility in the wall. It was found that the anges acted as eective connement for the ends of the reinforced concrete masonry wall, and that the anges changed the behaviour mode and lateral strength of the in-plane wall, compared to walls without anges.

Correlated studies by Abrams and Costley (1994) and Calvi and Magenes (1994) investigated parallel shear walls with and without anges. A vertical control joint was placed along the height of full scale and reduced scale, two storey brick buildings. This joint separated the anges from the in-plane element for one of the two perforated walls. Flanges increased the attraction of in-plane web shears without increasing shear strength and thus decreased deformation capacity.

The results from Abrams and Costley (1994) and Calvi and Magenes (1994) led to experiments conducted by Moon (2004) and Yi et al. (2006a,b), which suggested that substantial ange participation could be observed for in-plane walls in each loading direction in a full scale URM test structure. Flanges were dened by Moon et al. (2006) as the portion of the out-of-plane wall that participates with the in-plane wall to resist lateral loads, and proposed the following classications: Compression anges the portion of the out-of-plane walls that resists compressive stresses generated by rocking of the adjacent pier; Global tension anges the portion of the out-of-plane wall that is lifted up by global rocking; Component tension anges the portion of the out-of-plane wall that is lifted up by local pier rocking, which can occur at both the tension and compression sides of a building. Compression anges are equivalent to the typical anges used in the design of masonry walls that are dened based on elastic shear lag eects. In contrast, the tension anges
- 175 Alistair P. Russell

Chapter 7. In-Plane Cyclic Testing of Flanged URM Walls

represent a new type of ange that is unique to URM structures, and are dened as the portions of the out-of-plane walls that are lifted up by the deformation of the in-plane walls, and thus contribute additional weight to the in-plane piers (Moon et al., 2006).

Compression anges have an eect on the rotational displacement capacity of the adjacent pier, and also tend to increase the exural strength (rocking/toe crushing) while having a negligible eect on the shear strength (sliding/diagonal tension). Consequently, the presence of compression anges tends to alter the behaviour mode from a exure dominated response to a shear dominated response. Moon et al. (2006) took the eective compression ange length to be the same as stipulated in MSJC (2008), which species the length of the ange to be the lesser of six times the ange thickness (bf ), or the total length of the ange1 .

Global and component tension anges aect the response of URM walls by providing additional weight. Component tension anges increase the level of vertical stress in adjacent piers whereas global tension anges are considered to also resist the eects of overturning moment. As such, tension anges can signicantly aect both the strength and behaviour mode of individual walls. Global rocking deformation acts to lift up the global tension ange and eectively leaves the component tension anges behind (i.e. the ange does not rock with the in-plane wall). Full details and further denitions of global and component tension anges can be found in Moon et al. (2006).

Following full scale testing of a two storey URM building (Moon, 2004; Yi, 2004) where signicant ange participation was observed, Yi et al. (2008) developed an analytical model to investigate the eects of anges on the behaviour of individual non-rectangular section URM piers. The eective pier model in which ange eects are accounted for was based on the following assumptions:
The New Zealand Standard for the design of reinforced concrete masonry structures (Standards New Zealand, 2004b) states that the eective overhanging ange length on each side of the web [in-plane wall] is the lesser of eight times the ange thickness (bf ) or the length of the ange, or if the ange is on one side only the eective overhanging ange width is the lesser of six times the ange thickness or the length of the ange. Because the New Zealand Standard refers only and explicitly to reinforced concrete masonry, whilst MSJC allows the use of clay brick masonry, the provisions stated in MSJC are used in this chapter.
1

Alistair P. Russell

- 176 -

7.1. Background

External lateral forces are applied at the shear centre of the uncracked non-rectangular pier section and are parallel to the in-plane wall, such that possible torsional behaviour is ignored; Only part of the ange participates in the response of the adjacent in-plane pier, and was assumed to be as dened in MSJC (2008) (see above). In the absence of experimental data to validate the model, Yi et al. (2008) presumed an example wall and from a pushover analysis, determined that compared to a similar wall with no anges, the lateral strength could be expected to be greater. It was also determined that the limiting drift is dierent when the ange is at dierent locations in relation to the in-plane wall. When the ange is at the toe of the wall (i.e. the ange is in compression) the ange reduces the compressive stress at the toe, and delays toe crushing failure. Conversely, when the ange is at the heel (i.e. in tension) the compressive stress in the toe increases due to the increased weight of the ange. It was determined by Yi

Figure 7.1: Shear stress distribution in the in-plane URM wall with and without ange (from Yi et al. (2008))
- 177 Alistair P. Russell

Chapter 7. In-Plane Cyclic Testing of Flanged URM Walls

et al. (2008) that the location of the anges has a signicant eect on the diagonal tension strength of the wall. If the ange is in the middle of the in-plane wall, it has no eect on the diagonal tension strength, but when the ange is positioned closer to either end the diagonal tension strength decreases rst and then increases. This is caused by the discontinuous shear stress distribution in the in-plane wall due to the existence of the ange, as shown in Figure 7.1.

The models for determining the behaviour of walls responding in-plane, when including the inuence of anges, are summarised in Equations (7.1) (7.4), which are reproduced from Yi et al. (2008).

Vs =

N + 3ai bw vme 1+ N he 3he bw vme N ai 2 N lw 3 fm lw bw Wf

(7.1)

Vtc =

lw

(7.2)

fm Vdt = bw lw bfdt 1 + fdt

1 Wf 1 Wf 1 + 1 + + Ww 6 Ww 2 Ww 2 Wf 1 + Ww 2 lw

(7.3)

lw bw ai 2

Vr =

N he

lw 2 bw 6

+ (ai af )Af 1 Af

af

(7.4)

lw bw 2

ai =

Wf af + 0.5Ww lw Wf + Ww

(7.5)

where Vs , Vtc , Vdt and Vr are the lateral strength of the URM wall corresponding to sliding, toe crushing, diagonal tension and rocking, respectively. ai is the distance between the inertia centre and the compression edge of wall, af is the distance between the centre of
Alistair P. Russell

- 178 -

7.2. Construction Details

the ange and the compression edge of wall, Af is the cross-sectional area of the ange, Ww is the weight of the in-plane wall, Wf is the weight of the ange, and is a factor to account for nonlinear vertical stress distribution and has a value = 1.3 from Yi et al. (2005).

7.2
7.2.1

Construction Details
Wall Specications

Dimensions for the six walls are depicted in Figure 7.2 and listed in Table 7.1. As in Chapter 6, the prex A refers to As-built walls and was included in the wall name to dierentiate from walls with retrot interventions that were tested in the lab at similar times (which are not part of this thesis). All in-plane walls and anges were two leaves (240 mm) thick (bf = bw = 240 mm). Wall A3 was 2000 mm long and 2000 mm high, with an aspect ratio of 1:1. At the conclusion of testing Wall A3, it was observed that damage was conned to the top nine courses, and it was decided that the uncracked material could be utilised as another wall test. The nine courses where damage had occurred in the test of Wall A3 were removed and Wall A3a was thus fteen courses high with the same length and ange properties as Wall A3, and with an in-plane wall aspect ratio of 1:1.6. Walls A5 A8 were all 4000 mm long, and 2000 mm high, with an aspect ratio each of 1:2. This repetition of geometry facilitated direct comparison between Walls A5 A8 as reported in this chapter, and also facilitated comparisons between these walls and Wall A4 as reported in Chapter 6.

Due to concerns with the application of axial load in previous tests (Walls A1 and A2), axial load was applied as gravity loads in the test of Wall A3. This loading technique was in contrast to the application of axial load through post-tensioning tendons as was adopted in all other walls. Steel and concrete blocks were used as gravity load on the top of Wall A3. The total weight of the available material was 9 kN, and this axial load was applied at the centre of the in-plane wall. See Section 7.3.1 for further details.

- 179 -

Alistair P. Russell

Chapter 7. In-Plane Cyclic Testing of Flanged URM Walls

The axial load on all other walls reported in this chapter was applied through posttensioning tendons near the ends of the wall (see Section 7.3.1). The axial load on Wall A3a was 22 kN, corresponding to 22 kPa. The axial load on Wall A5 was the same as for Wall A4 (30 kN), but because of the increased cross-sectional area of the wall due to the presence of anges, of axial force corresponded to an axial stress of 22 kPa. Initially the axial load on Wall A6 was the same as on Wall A5, but at the beginning of testing it was observed that the steel channel used to apply the lateral force was lifting o, and consequently the axial load on Wall A6 was increased to 73 kN, and corresponded to an axial stress of 41 kPa. Similarly, the axial load on Walls A7 and A8 was 76 kN and 71 kN, corresponding to an axial stress of 56 kPa and 52 kPa, respectively. See Table 7.1.

Details of the anges are listed in Table 7.2 and shown in Figure 7.2. Walls A3, A3a and A5 had anges at both ends with a length of 480 mm on either side of the in-plane wall, with a total ange length of 1200 mm. Walls A6 and A7 had ange lengths of 960 mm on either side of the in-plane wall, with a total ange length of 2160 mm. The anges were positioned at both ends in the case of Wall A6, and at one end in the case of Wall A7. Wall A8 had anges at both ends, but on a single side only, and the anges had a length of 960 mm on the side of the in-plane wall.

As stated above, the eective length of anges can be determined according to MSJC Table 7.1: Wall specications Wall Age Days A3 A3a A5 A6 A7 A8 237 252 78 29 36 40 bw mm 230 230 230 230 230 230 h mm lw mm
fm

fm MPa 0.010 0.022 0.022 0.041 0.056 0.052

fm /fm

MPa MPa MPa 18.1 18.1 10.1 9.2 11.9 9.1

- 180 -

% 0.055 0.123 0.220 0.441 0.468 0.576

2000 2000 1200 2000 2000 4000 2000 4000 2000 4000 2000 4000

0.4 0.4 0.1 0.1 0.1 0.1

0.7 0.7 0.7 0.7 0.7 0.7

Alistair P. Russell

7.2. Construction Details

Table 7.2: Flange details Wall bf mm A3 A3a A5 A6 A7 A8 230 230 230 230 230 230 lf mm 1200 both ends 1200 both ends 1200 both ends 2160 both ends 2160 one end both sides both sides both sides both sides both sides one side Flanges at Plan geometry

1200 both ends

(2008) or Standards New Zealand (2004b), but in this chapter the denition from MSJC is utilised. According to MSJC (2008), the eective ange length is 6bf on either side of the in-plane wall. For all walls reported in this chapter, bf = 240 mm, and the eective ange length on either side of the in-plane wall was 1440 mm. According to these denitions, the length of the anges on all walls reported in this chapter were considered to be less than their maximum eective length.

7.2.2

Wall Construction

All ve2 walls were constructed by experienced blocklayers under supervision (see Figure 7.3). The walls were constructed with a common bond pattern and with 1:2:9 mortar (cement:lime:sand, by volume), corresponding to ASTM type O mortar, and with nominally 10 mm thick mortar joints. There were header bricks every 4th course. It was intentionally decided to construct the walls in a way that replicated the observed, often deteriorated, nish quality of walls in real New Zealand URM buildings.

Due to a closure of the laboratory facilities for a sustained period and other factors, testing of Wall A3 was signicantly delayed such that the actual experimentation began 237 days after construction of the wall was completed. Similarly, because Wall A3a was a
2

Six walls were tested, but Wall A3a was a retest of Wall A3.

- 181 -

Alistair P. Russell

Chapter 7. In-Plane Cyclic Testing of Flanged URM Walls

(a) Wall A3

(b) Wall A3a

Figure 7.2: Wall dimensions

Alistair P. Russell

- 182 -

7.2. Construction Details

(c) Wall A5

(d) Wall A6

Figure 7.2: Wall dimensions (continued)

- 183 Alistair P. Russell

Chapter 7. In-Plane Cyclic Testing of Flanged URM Walls

(e) Wall A7

(f) Wall A8

Figure 7.2: Wall dimensions (continued)

Alistair P. Russell

- 184 -

7.2. Construction Details

retest of Wall A3, the age at testing of Wall A3a was also high, at 252 days. The age of Wall A5 at testing was 78 days, Wall A6 was tested 29 days after the completion of construction, Wall A7 was tested 36 days after construction was complete, and testing of Wall A8 commenced 40 days after construction was complete.

(a) Wall A3

(b) Wall A5

Figure 7.3: Wall construction

- 185 Alistair P. Russell

Chapter 7. In-Plane Cyclic Testing of Flanged URM Walls

7.2.3

Material Properties

The materials used in the six walls reported in this chapter are similar to those used in the walls that were reported in Chapter 6, and further details can be found in Section 6.2.3. The same sand was used in the construction of each wall, and the sand gradation curve is as shown in Figure 6.4. Nine mortar cubes and six prisms were constructed at the same time as each wall using randomly selected bricks and tested on the same day as the wall test was conducted. These cubes and prisms allowed determination of the compression strength of the masonry and the mortar. Additionally, twelve half bricks were selected randomly during construction and tested with the prisms and mortar cubes to determine the brick compression strength. Details of these tests can be found in ASTM (2003, 2004a, 2007a). Using criteria stipulated in NZSEE (2006), the properties of the mortar used in Walls A3 A8 were determined, and are shown in Table 7.1. Similar to the walls in Chapter 6, at the conclusion of each test an in-situ shear test was conducted on an undamaged bed joint to determine the cohesive strength of the masonry bed joints (vme ). The description and results of the tests are reported elsewhere (Lumantarna, 2012), and the values are used with permission.

The material properties of all six walls are summarised in Table 7.1.

7.3
7.3.1

Testing Details
Test Setup and Instrumentation

Much of the information in this section, including the general test setup, is covered in Section 6.3.1. Consequently, reference is only made here to where dierences exist. The test setup for Walls A3a A8 was consistent with that reported in Chapter 6, but the setup for Wall A3 was dierent. In order to distribute the applied lateral force as shear forces at the top of the wall, a timber diaphragm was constructed with joists oriented perpendicular to the top two courses of the in-plane wall (web) of Wall A3, as shown in
Alistair P. Russell

- 186 -

7.3. Testing Details

Figure 7.4. The joists were mortared into the wall. The steel loading beam was then attached to the diaphragm with coach screws, and the lateral load was applied as for other tests. In the case of Wall A3, the lateral load was introduced into the wall through the perpendicular joists, rather than by steel plates on the outside of the anges.

Figure 7.4: Timber diaphragm used in Wall A3 The method used to apply the axial load to Wall A3 was also dierent from the method used in each of the other tests. During the testing of Walls A1 and A2, concerns were raised that as the lateral displacement increased, the force in the tendons applying the axial load was also increasing, and consequently the axial load was not constant throughout the test. Because a plywood sheet was attached to the top of the timber diaphragm, a level surface was available on which to place heavy weights. Steel and concrete blocks were available in the laboratory and were used to eect the axial load (see Figure 7.5(a)). In this way, the axial load was kept constant throughout the test.

It was later determined that the eect of varying axial load was not signicant, and consequently for all other tests, the method for applying the axial load as used in Chapter 6 was used. The axial load on each wall (excluding Wall A3) was eected by four posttensioning tendons positioned approximately 400 mm from the outside of the anges at
- 187 Alistair P. Russell

Chapter 7. In-Plane Cyclic Testing of Flanged URM Walls

(a) Axial load on Wall A3

(b) Axial load on Wall A6

Figure 7.5: Application of axial load on anged walls each end of the wall. Two steel box beams sat on top of the loading beam, perpendicular to the in-plane wall, and the tendons passed through these box beams. The box beams were on pivots, so that the load in the tendons on either side of the wall was equal (see in Figure 7.5(b)).

Alistair P. Russell

- 188 -

7.3. Testing Details

Figure 7.6: Instrumentation

Consistent with the walls reported in Chapter 6, extensive instrumentation was installed on each of the six walls to capture all facets of wall response, with a typical layout shown in Figure 7.6. Instruments with a designation beginning with A F measured shear distortion on the wall, J measured distortions on the outside of the anges, G measured any uplift in the bottom two corners of the wall relative to the ground, S measured sliding between the wall and the ground and between the wall and the loading beam, and DISP refers to the measurement of overall wall displacements. LCF represents the load cell measuring the horizontal force applied to the wall top and LCA refers to the load cells measuring the axial load on the wall. Instruments with a designation beginning with X and Y measured deformations on the inside faces of the anges.

The force was recorded between the hydraulic jack and the wall through a load cell, and accurately recorded the force being applied at that point. The control displacement (DISP1) was recorded at the opposite end of the wall, between the top of the wall (at a point attached to the bricks immediately below the bottom surface of the loading beam) and an independent frame, to eliminate any eects from exing of the strong-wall. This displacement reading was obtained using a portal gauge with a displacement range of 50 mm.
- 189 Alistair P. Russell

Chapter 7. In-Plane Cyclic Testing of Flanged URM Walls

7.3.2

Test Procedure

The cyclic loading sequence adopted for all tests was that shown in Figure 6.7 and was as detailed in Chapter 6. During the rst stages of testing, the potential for shear failure of the wall meant that small displacement increments were necessary to avoid the wall being loaded to failure at an early stage of testing. An early sudden and brittle failure did occur in the response of Walls A3, A3a and A5 (see Section 7.5). The push direction was dened as positive and the pull direction as negative.

Also as detailed in Chapter 6, wall testing was terminated when the walls were deemed to have failed, which coincided with the maximum wall strength obtained in a particular cycle being less than 80% of the maximum recorded during testing. The ultimate displacement capacity, du is the displacement corresponding to failure.

7.4

Predicted Flexural and Shear Strength

The model proposed by Yi et al. (2008) (see Section 7.1) was used to determine the nominal lateral strength of each of the walls reported in this chapter. The predicted strengths are shown in Table 7.3. The maximum nominal shear strength Vn is given as the lowest of the strength limits, as given by Equation (7.6), and is shown in bold text in Table 7.3. The model developed by Yi et al. (2008) assumed a single ange, with the location determining af (the distance between centre of ange and compression edge of wall) and ai (the distance between inertia centre and compression edge of wall). Apart from Wall A7, all the walls reported in this chapter had anges at both ends, such that ai = lw /2. For Wall A7, ai is determined according to Equation (7.5). For symmetrical walls with anges at both ends, it was also determined that af = 2ai , so that af = lw . Vn = MIN(Vs , Vtc , Vdt , Vr ) (7.6)

For Wall A7 in the push cycle (positive) the ange was in compression (A7c ) and af = 0,
Alistair P. Russell

- 190 -

7.4. Predicted Flexural and Shear Strength

Table 7.3: Predicted wall strengths from model developed by Yi et al. (2008) Wall Vs kN A3 A3a A5 A6 A7t A7c A8 31.7 45.6 72.0 99.4 80.7 60.4 98.4 Vtc kN 35.4 57.3 Vdt kN 57.1 45.8 Vr kN 31.4 50.9 Vn kN 31.4 45.6 sliding/rocking sliding/diagonal tension diagonal tension diagonal tension Predicted failure mode

136.6 63.9 113.7 63.9 222.3 75.4 185.4 75.4 75.0 151.2 76.1 59.3 76.2 89.7

75.0 sliding/diagonal tension/rocking 59.3 sliding/diagonal tension diagonal tension

218.7 75.0 182.4 75.0

ai = lw /3. As detailed in Section 7.1 when the ange is at the toe of the wall (i.e. the ange is in compression) the ange reduces the compressive stress at the toe, and tends to increase the exural strength (rocking/toe crushing). For Wall A7 in the pull direction the ange was in tension (A7t ) and af = lw , ai = 2lw /3. In this case, the diagonal tension strength was increased due to the weight of the tension ange, and because of the discontinuous shear stress distribution at the junction of the ange and the in-plane wall (see Yi et al., 2008, for further details).

The model developed by Yi et al. (2008) predicted that for Wall A3, sliding/rocking response could be expected, and for Wall A3a the limiting strengths due to diagonal tension and sliding were similar, such that a sliding shear/diagonal tension failure was predicted. It was determined that diagonal tension would be the failure mode for Walls A5, A6 and A8. When the ange was in tension in Wall A7 it was predicted that toe crushing would limit the strength, although the limiting strengths due to diagonal tension and rocking were similar, and consequently it was determined that a shear failure or rocking response could be predicted. When the ange was in compression in Wall A7 it was predicted that diagonal tension would limit the strength, although the limiting strength due to sliding was also similar, and consequently a sliding shear/diagonal tension failure was predicted.

- 191 -

Alistair P. Russell

Chapter 7. In-Plane Cyclic Testing of Flanged URM Walls

7.5

Test Results

This section presents the test observations, followed by a description of wall behaviour and performance. General results are presented in Table 7.4, where Vn is the nominal shear strength as predicted by the model proposed by Yi et al. (2008) (see Section 7.4), Vmax is the maximum recorded lateral force, dVmax and Vmax are the corresponding displacement and drift, respectively, at Vmax , Vcrack is the lateral force when cracking was rst observed, crack is the drift corresponding to Vcrack , and du and u are the lateral wall displacement and drift, respectively, corresponding to the point at which the lateral force had degraded to 80% of Vmax , where u = du /he . In Table 7.4, A7t refers to the test of Wall A7 in the pull direction where the ange was in tension (at the heel), and A7c refers to the test of Wall A7 in the push direction where the ange was in compression (at the toe).

7.5.1

Observations

Wall A3 failed when a diagonal crack occurred at a low displacement in the push cycle of 0.4 mm. The force dropped suddenly, and the displacement increased correspondingly (see Figure 7.8(a)). The crack which developed occurred suddenly, and initiated at the point where the rst joist was in the in-plane wall (web) (see Figure 7.22(c)), then along Table 7.4: Test results Wall Vn kN A3 A3a A5 A6 A7t A7c A8 31.4 45.6 63.9 75.4 75.0 60.4 75.0 Vmax kN 54.9 72.5 66.0 66.8 75.0 61.9 66.9 1.75 1.59 1.03 0.89 1.0 1.02 0.89 Vmax /Vn dVmax mm 0.6 0.8 0.5 7.8 2.8 7.7 3.6 Vmax % 0.03 0.06 0.03 0.39 0.14 0.38 0.18
- 192 -

Vcrack kN 54.9 72.5 66.5 46.4 34.5 34.7 60.0

crack % 0.02 0.01 0.02 0.04 0.03 0.03 0.06

du mm 0.3 0.1 1.0

u %

Behaviour

0.02 Shear failure 0.01 Shear failure 0.05 Shear failure

19.0 0.95 Shear failure 14.8 0.74 Shear failure 12.8 0.63 Shear failure 19.2 0.96 Shear failure

Alistair P. Russell

7.5. Test Results

the web and into the ange, 4 courses below. The angle of cracking was shallow (< 45). When the load was then released the crack closed up and the wall returned to its original state. In the pull direction, the maximum attained force was lower than for the push cycle, and was a result of the crack which opened up during the push cycle. A similar crack occurred during the pull cycle, but the maximum lateral force which was attained was lower. In contrast to the other walls reported in this chapter, where the lateral force was applied through bearing plates on the outside of the anges, the cracks initiated at the location of the joists, and not in the top corners (see Figures 7.22(b) and 7.22(c)). Figure 7.7(a) shows the nal cracking pattern of Wall A3.

Due to safety considerations in the laboratory, further lateral force was not applied to Wall A3 beyond the point where the lateral force had degraded to less than 80% of Vmax . As the steel and concrete blocks on the top of the wall were resting on the timber diaphragm and not secured, the loss of gravity supporting capacity of Wall A3 was considered a greater safety hazard than for other walls where there were no unsecured weights on top of the wall.

Wall A3a attained a peak force of 95 kN, when a sudden and brittle crack occurred in the tension ange. The lateral force dropped suddenly, and in the subsequent cycles the wall was sliding on a plane at the bottom and punching through the ange on the pull cycle. On the push cycle, the compression ange cracked at the base and the whole ange was sliding with the main part of the wall. On the pull cycle, similarly, the tension ange was sliding with the web.

Because Wall A3a had a reduced height (1250 mm) and there were no unsecured objects on the top of the wall, the danger from collapse by testing beyond the failure strength was considered less than for Wall A3, and larger lateral displacements were applied. This continued testing allowed observation of the post-failure sliding behaviour which subsequently occurred. Thus, Wall A3a at rst failed in a force-controlled mechanism (brittle cracking) and then after this behaved in a deformation-controlled mechanism (sliding). Figure 7.7(b) shows the nal cracking pattern of Wall A3a.
- 193 Alistair P. Russell

Chapter 7. In-Plane Cyclic Testing of Flanged URM Walls

Wall A5 failed through diagonal tension cracking. Cracking occurred in the top course of the web and near the mid-point along the length of the web, and the crack propagated down towards the opposite bottom corner at approximately 45. Because the cracking was sudden and brittle, the lateral force also dropped suddenly. Lateral force continued to be applied to the wall beyond the point when failure had occurred, but after two cycles when the lateral force was less than 80% of Vmax , bricks in the top course became loose and caused a safety hazard (see Figure 7.24(e)). Consequently the test was terminated at a displacement of 1.5 mm. Figure 7.7(c) shows the nal cracking pattern of Wall A5.

Wall A6 also failed through diagonal tension cracking, although the displacement corresponding to failure was larger than for Walls A3, A3a and A5. Consequently, further hysteretic loops were obtained, allowing a proper assessment of energy dissipation and equivalent hysteretic damping. Cracking initiated in the top course of the in-plane wall, although closer to the anges than compared with the corresponding cracks in Wall A5 (see Figure 7.25). Cracking was also evident in the anges, and uplift of the global tension ange can be observed in Figure 7.25(e). Figure 7.7(d) shows the nal cracking pattern of Wall A6.

The response of Wall A7 when the ange was in tension was dierent than when the ange was in compression. Failure was due to diagonal tension cracking in both directions, although the orientation of cracks and the magnitude of the maximum lateral force was dierent. In the push cycle, when the ange was in compression, the maximum lateral force was approximately 60 kN, and occurred at a displacement of approximately 8 mm. In the pull direction, when the ange was in tension, the maximum lateral force was approximately 75 kN, and occurred at approximately 3 mm. Cracking in both directions initiated in the top course and the angle of propagation down to the bottom of the wall was steeper in the push direction. In the pull direction, cracking also occurred in the compression toe (see Figure 7.26(c)), and some cracking also initiated in the bottom corner next to the ange. Moon et al. (2006) referred to this as the component tension ange (see Section 7.1). Figure 7.7(e) shows the nal cracking pattern of Wall A7.

Alistair P. Russell

- 194 -

7.6. Discussion

The failure mode of Wall A8 was diagonal tension. Similar to Wall A6, cracking initiated in the top course of the in-plane wall, and at the opposite end from where the force was applied (next to the compression ange) the crack propagated downwards at an angle of approximately 45 (see Figure 7.27(c)). Cracking also occurred in the tension and compression anges, as can be seen in Figures 7.27(a) and 7.27(d). At the conclusion of the test, bricks in the in-plane wall became unstable, as can be seen in Figure 7.27(f). Figure 7.7(f) shows the nal cracking pattern of Wall A8.

7.6

Discussion

This section discusses wall performance in terms of lateral strength and compares the obtained values with those predicted using available equations. Further discussion is provided on the force-displacement response, energy dissipation, initial stiness, ultimate drifts obtained during testing and methods of approximating the response.

7.6.1

Force-Displacement Response

The force-displacement response of Walls A3, A3a, A5, A6, A7 and A8 is presented in Figure 7.8. Walls A3 and A3a failed at a low displacement of 0.02% and 0.01% respectively. Further force-displacement cycles are evident in the test of Wall A3a (Figure 7.8(b)), for the reasons outlined in Section 7.5.1. Wall A5 also failed at a low displacement of 0.05%. The large peak and subsequent sudden drop in strength observed in the response of Wall A3 corresponded to when the compression ange cracked. After this cracking occurred, the ange was no longer connected to the wall and did not participate in the lateral force resistance. The response of Walls A3 and A5 was not observed beyond a drift of 0.2% and 0.07% respectively, and consequently the post-peak lateral force-displacement response could not be reported.

Walls A6, A7 and A8 exhibited diagonal tension failure and sliding also, but further hys- 195 Alistair P. Russell

Chapter 7. In-Plane Cyclic Testing of Flanged URM Walls

(a) Wall A3

(b) Wall A3a

(c) Wall A5

Figure 7.7: Cracking patterns

Alistair P. Russell

- 196 -

7.6. Discussion

(d) Wall A6

(e) Wall A7

(f) Wall A8

Figure 7.7: Cracking patterns (continued)

- 197 Alistair P. Russell

Chapter 7. In-Plane Cyclic Testing of Flanged URM Walls

80 60 40 Base Shear (kN) 20 0 !20 !40

!0.2

!0.1

Drift (%) 0

0.1

0.2 150 100 Base Moment (kNm) 50 0 !50 !100 Base Shear (kN)

!0.3 100 80 60 40 20 0 !20 !40 !60 !80 !150

!0.2

!0.1

Drift (%) 0

0.1

0.2

0.3 200 150 Base Moment (kNm) Base Moment (kNm) Base Moment (kNm) 100 50 0 !50 !100 !150

Vn = 31.4 kN

Vn = 45.6 kN

!60 !80 !4 !2 0 2 Displacement (mm) 4

!100 !6

!4

!2 0 2 Displacement (mm)

!200

(a) Wall A3
!0.1 80 60 40 Base Shear (kN) 20 0 !20 !40 !100 !60 !80 !2 !150 !1.5 !1 !0.5 0 0.5 1 Displacement (mm) 1.5 2 !60 50 0 !50 !0.05 Vn = 63.9 kN Drift (%) 0 0.05 0.1 150 100 Base Moment (kNm) Base Shear (kN)

(b) Wall A3a

!1.2 !0.8 !0.4 80 Vn = 75.4 kN 60 40 20 0 !20 !40 !100 !80 !25 !20 !15 !10 !5 0 5 10 Displacement (mm) !150 25 50 0 !50 Drift (%) 0 0.4 0.8 1.2 150 100

15

20

(c) Wall A5
!0.8 !0.6 !0.4 !0.2 80 Vn (+) = 60.4 kN 60 40 Base Shear (kN) 20 0 !20 !40 !100 !60 !80 !16 !12 !8 Vn (!) = 75.0 kN !4 0 4 8 Displacement (mm) 12 !60 !150 16 50 0 !50 Drift (%) 0 0.2 0.4 0.6 0.8 150 100 Base Moment (kNm) 40 Base Shear (kN) 20 0 !20 !40

(d) Wall A6
!1.2 !0.8 !0.4 80 Vn = 75.0 kN 60 Drift (%) 0 0.4 0.8 1.2 150 100 50 0 !50 !100 !80 !25 !20 !15 !10 !5 0 5 10 Displacement (mm) !150 25

15

20

(e) Wall A7

(f) Wall A8

Figure 7.8: Force-displacement response

Alistair P. Russell

- 198 -

7.6. Discussion

teretic loops were obtained from these tests. It can be seen that for walls with an aspect ratio of 1:2 and anges with lt = 4bf on each side of the in-plane wall, loss of lateral strength was not sudden and there was some observable strength degradation after the peak lateral force was obtained.

Wall A6 reached a drift of 0.39%, corresponding to the peak lateral force, and an ultimate drift of 0.95% corresponding to 0.8Vmax . Similar to Wall A6, Wall A8 exhibited an ultimate drift of 0.96%, which was larger than the drift corresponding to Vmax (0.18%). This indicates that for an in-plane wall controlled by diagonal tension followed by sliding, there can be some residual displacement capacity beyond the suggested drift limit of 0.4%.

The response of Wall A7 was not symmetrical, as was expected due to the non-symmetrical geometry of the wall (see Figure 7.8(e)). The ultimate drift capacity in the push direction (ange in compression) was 0.63% and in the pull direction (ange in tension) was 0.74%. Overall, fat hysteretic loops were evident in both directions, and the energy dissipation characteristics are outlined in Section 7.6.2. Post-peak strength degradation was observed more in the pull direction than in the push direction.

The lateral force-displacement response of the walls, as reported in this section, is used below to determine the energy dissipation characteristics of the walls, as well as to determine an approximate bilinear representation.

7.6.2

Energy Dissipation

The method for determining equivalent viscous damping of the walls reported in this chapter is as described in Section 6.5.2. From the hysteretic response shown in Section 7.6.1 the equivalent viscous damping ratio of Walls A6, A7 and A8 only were obtainable, and are shown in Figure 7.9. From the results of these tests, consistently large equivalent viscous energy damping ratios were evident, and an average lower bound value of eq = 0.15 is recommended for anged walls responding in a diagonal tension failure (which is con- 199 Alistair P. Russell

Chapter 7. In-Plane Cyclic Testing of Flanged URM Walls

sistent with previous ndings in the literature (Magenes and Calvi, 1997)). In Figure 7.9 a dashed vertical line indicates initial cracking.

Equivalent Viscous Damping Ratio, !eq

0.25 0.2 0.15 0.1 0.05 0

Equivalent Viscous Damping Ratio, !eq 0 0.2 0.4 0.6 Drift (%) 0.8 1 1.2

0.3

0.3 0.25 0.2 0.15 0.1 0.05 0

0.1

0.2

0.3

0.4 0.5 Drift (%)

0.6

0.7

0.8

(a) Wall A6
0.3 0.25 0.2 0.15 0.1 0.05 0

(b) Wall A7

Equivalent Viscous Damping Ratio, !eq

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 Drift (%)

(c) Wall A8

Figure 7.9: Equivalent viscous damping ratio of Wall A6, A7 and A8

7.6.3

Bilinear Approximation

Section 6.5.3 describes the steps for constructing an equivalent bilinear approximation, and the same process is used in this chapter. A bilinear response was developed for Walls A5 A8, and is shown in Figure 7.10. The bilinear response for Walls A5, A6 and A8 was constructed by averaging the response of the positive and negative cycles, but in the case of Wall A7 there was a separate (normalised) bilinear response in each direction (see Figure 7.10(c)). It can be seen that Vu and du were dierent for situations when the
Alistair P. Russell

- 200 -

7.6. Discussion

ange was in tension or compression, as described in Section 7.6.1. When the ange was at the compression toe, the initial stiness of the in-plane wall was lower than when the ange was in tension (described as Flange at Heel in Figure 7.10(c)).

The bilinear plots developed here are used in Chapter 8 to inform the assessment procedure when considering in-plane URM walls with eective anges.

80 70

0.025

Drift (%) 0.05

0.075

0.1 150

80 70 Base Moment (kNm)

0.2

0.4

Drift (%) 0.6 0.8

1.2 150

50 40 30

100 75 50

50 40 30

100 75 50

20 10 0 0 0.5 1 1.5 Displacement (mm) 2 25 0

20 10 0 0 5 10 15 Displacement (mm) 20 25 0 25

(a) Wall A5
80 70 Base Moment (kNm) 60 Base Shear (kN) 50 40 30 50 20 10 0 0 2 Flange at Toe 25 Flange at Heel 0 4 6 8 10 12 14 16 Displacement (mm) 125 100 75 0 0.1 0.2 Drift (%) 0.3 0.4 0.5 0.6 0.7 0.8 150 70 60 Base Shear (kN) 50 40 30 80 0 0.2

(b) Wall A6
0.4 Drift (%) 0.6 0.8 1 1.2 150 125 100 75 50 20 10 0 0 5 10 15 Displacement (mm) 20 25 0 25

(c) Wall A7

(d) Wall A8

Figure 7.10: Equivalent bi-linear response of Wall A5, A6, A7, A8

7.6.4

Multi-Linear Approximation

Section 6.5.4 describes the steps for constructing an equivalent multi-linear approximation, and the same process is used in this chapter, except for Wall A3a. A multi-linear response
- 201 Alistair P. Russell

Base Moment (kNm)

Base Moment (kNm)

60 Base Shear (kN)

125

60 Base Shear (kN)

125

Chapter 7. In-Plane Cyclic Testing of Flanged URM Walls

Drift (%) 0.15

100

0.05

0.1

0.2

0.25

0.3 200

80 Base Shear (kN)

60 100 40 50

20

2 3 4 Displacement (mm)

(a) Average multi-linear force-displacement approximation

Figure 7.11: Multi-linear approximation of Wall A3a

Drift (%) 0 Drift (%) 0.05

!0.1 80 60

!0.05

0.05

0.1 150 100 Base Moment (kNm)

80

0.025

Base Moment (kNm)

150

0.075

0.1 160 Normalized Average 140 Base Moment (kNm) 120 100

40 Base Shear (kN) 20 0 !20 !40 !100 !60 !80 !2 !150 !1.5 !1 !0.5 0 0.5 1 Displacement (mm) 1.5 2 50 0 !50

60 Base Shear (kN)

40

80 60

20

40 20

0.5

1 1.5 Displacement (mm)

(a) Multi-linear force-displacement approximation (b) Average multi-linear force-displacement approximation

Figure 7.12: Multi-linear approximation of Wall A5

was developed for Walls A3a A8, and is shown in Figures 7.11 7.15. The multi-linear approximation of Wall A3a was constructed by simply tting a line to the envelope of the force-displacement response in the positive direction, where cracking in the compression ange occurred, following by sliding of the in-plane wall on its base course. The rigorous approach of tting a curve to the peak force of each cycle and then normalising force for each direction was not followed, and as such, the average multi-linear force-displacement approximation for Wall A3a can be seen as illustrative only. This approximation does provide useful information on the post-failure force-displacement response.

Alistair P. Russell

- 202 -

7.6. Discussion
Drift (%) 0 0.4

!1.2 80 60

!0.8

!0.4

0.8

1.2 150 100

20 0 !20 !40

50 0 !50 !100

!60 !80 !25 !20 !15 !10 !5 0 5 10 Displacement (mm) 15 20 !150 25

(a) Force-displacement envelope

!1.2 80 60 100 Base Moment (kNm) 40 Base Shear (kN) 20 0 !20 !40 !100 !60 !80 !25 !20 !15 !10 !5 0 5 10 Displacement (mm) 15 20 !150 25 10 0 0 5 10 15 Displacement (mm) 20 50 0 !50 60 Base Shear (kN) 50 40 30 50 20 25 0 25 !0.8 !0.4 Drift (%) 0 0.4 0.8 1.2 150 0 0.2 0.4 Drift (%) 0.6 0.8 1 1.2

80 70

Base Moment (kNm)

40 Base Shear (kN)

Normalized 150 Average Base Moment (kNm) 125 100 75

(b) Multi-linear force-displacement approximation (c) Average multi-linear force-displacement approximation

Figure 7.13: Multi-linear approximation of Wall A6

According to the multi-linear segmented curves provided by ASCE and reproduced in Figure 6.14, the response of Wall A3a conforms to a Type 1 curve, where some residual displacement capacity is available once the peak lateral force had degraded signicantly below 0.8Vmax .

With the available results, the response of Wall A5 corresponds to a Type 3 curve, where after the peak lateral force is attained, no further force or displacement capacity is available. This force-displacement relationship is representative of a brittle or non-ductile behaviour, with no capacity to support gravity loads beyond the peak force.

- 203 -

Alistair P. Russell

Chapter 7. In-Plane Cyclic Testing of Flanged URM Walls

Drift (%) 0 0.2

!0.8 !0.6 !0.4 !0.2 80 60

0.4

0.6

0.8 150 100

20 0 !20 !40

50 0 !50 !100

!60 !80 !16 !12 !8 !4 0 4 8 Displacement (mm) 12 !150 16

(a) Force-displacement envelope

!0.8 !0.6 !0.4 !0.2 80 60 100 Base Moment (kNm) 40 Base Shear (kN) 20 0 !20 !40 !100 !60 !80 !16 !12 !8 !4 0 4 8 Displacement (mm) 12 !150 16 10 0 0 2 4 6 8 10 12 Displacement (mm) 14 50 0 !50 60 Base Shear (kN) 50 40 30 50 20 25 0 16 Drift (%) 0 0.2 0.4 0.6 0.8 150 0 0.1 0.2 0.3 Drift (%) 0.4 0.5 0.6 0.7 0.8

80 70

Base Moment (kNm)

40 Base Shear (kN)

Flange at Toe 150 Flange at Heel Base Moment (kNm) 125 100 75

(b) Multi-linear force-displacement approximation (c) Average multi-linear force-displacement approximation

Figure 7.14: Multi-linear approximation of Wall A7 The response of Walls A6, A7 (in both directions) and A8 ts the Type 2 curve. Beyond the initial elastic portion of the response, further displacement capacity is available without the loss of ability to support gravity loads, until an ultimate displacement is reached, after which lateral force and then gravity load carrying capacity is reduced.

Similar to the construction of the bilinear approximation, the response of Wall A7 was normalised, such that the positive and negative curves were both in the same quadrant, but not averaged, as was done for each of the other walls.

The graphs used to construct the multi-linear response of Walls A5 A8 are also shown
Alistair P. Russell

- 204 -

7.6. Discussion
Drift (%) 0 0.4

!1.2 80 60

!0.8

!0.4

0.8

1.2 150 100

20 0 !20 !40

50 0 !50 !100

!60 !80 !25 !20 !15 !10 !5 0 5 10 Displacement (mm) 15 20 !150 25

(a) Force-displacement envelope

!1.2 80 60 100 Base Moment (kNm) 40 Base Shear (kN) 20 0 !20 !40 !100 !60 !80 !25 !20 !15 !10 !5 0 5 10 Displacement (mm) 15 20 !150 25 10 0 0 5 10 15 Displacement (mm) 20 50 0 !50 60 Base Shear (kN) 50 40 30 50 20 25 0 25 !0.8 !0.4 Drift (%) 0 0.4 0.8 1.2 150 0 0.2 0.4 Drift (%) 0.6 0.8 1 1.2

80 70

Base Moment (kNm)

40 Base Shear (kN)

Normalized 150 Average Base Moment (kNm) 125 100 75

(b) Multi-linear force-displacement approximation (c) Average multi-linear force-displacement approximation

Figure 7.15: Multi-linear approximation of Wall A8 in Figures 7.12 7.15, and provide further information to show how each multi-linear plot was constructed.

7.6.5

Ultimate Drift

The average multi-linear response of Walls A5 A8 is shown collectively in Figure 7.16 and the ultimate drift (shown as d) is given for each wall.

ASCE (2007) states that the procedure used to analyse an existing building shall be either a linear analysis or a nonlinear analysis. If a nonlinear static procedure is followed,
- 205 Alistair P. Russell

Chapter 7. In-Plane Cyclic Testing of Flanged URM Walls

Drift (%) 0.6 0.8 Drift (%) 0.6 0.8

80

0.2

0.4

1.2

160 140 Base Moment (kNm) 120 100

80 70 60 Base Shear (kN) 50 40 30

0.2

0.4

1.2 150 125 100 75 50

d = 0.08% 60 Base Shear (kN)

d = 0.95%

40

80 60

20

40 20

20 10 0 0 5 10 15 Displacement (mm) 20 25 0 25

10 15 Displacement (mm)

20

0 25

(a) Wall A5
80 70 60 Base Shear (kN) 50 40 30 50 20 10 0 0 5 10 15 Displacement (mm) 20 25 0 25 0 0.2 0.4 Drift (%) 0.6 0.8 1 1.2 150 70 Base Moment (kNm) 125 100 75 60 Base Shear (kN) 50 40 30 80 0 0.2

(b) Wall A6
0.4 Drift (%) 0.6 0.8 1 1.2 150 125 100 75 50 20 10 0 0 5 10 15 Displacement (mm) 20 25 0 25

d = 0.74%

d = 0.63%

(c) Wall A7 (ange in compression)

0 0.2 0.4 Drift (%) 0.6 0.8

(d) Wall A7 (ange in tension)

1 1.2 150 125 100 75 50

80 70 60 Base Shear (kN) 50 40 30 20 10 0

d = 0.96%

25 0 25

10 15 Displacement (mm)

20

(e) Wall A8

Figure 7.16: Average multi-linear approximations of anged walls

nonlinear deformation capacities for primary and secondary components, expressed in terms of storey drift ratio percentages, are required. Primary and secondary components
Alistair P. Russell

- 206 -

Base Moment (kNm)

Base Moment (kNm)

Base Moment (kNm)

7.6. Discussion

are dened as follows (ASCE, 2007): Components that aect the lateral stiness or distribution of forces in a structure, or are loaded as a result of lateral deformation of the structure, shall be classied as primary or secondary, even if they are not intended to be part of the lateral force resisting system. A structural component that is required to resist seismic forces in order for the structure to achieve the selected performance level shall be classied as primary. A structural component that is not required to resist seismic forces in order for the structure to achieve the selected performance level shall be permitted to be classied as secondary. As such, the walls tested in this chapter can be considered primary. The ultimate lateral drift for deformation controlled primary components corresponds to a drift of d in Figure 7.17, and this point is where the lateral force resistance is considered lost. The ultimate lateral drift of secondary components corresponds to a drift of e in Figure 7.17, and this point also corresponds to the point where gravity load carrying capacity of the structure as a whole is compromised.

Figure 7.17: Generalised force-deformation relation for masonry elements or components (reproduced from Figure 7-1(a), ASCE (2007))

All the walls reported in this chapter failed by diagonal tension cracking, which is a shear dominated response. Walls A6 A8 subsequently commenced sliding on the bed-joints
- 207 Alistair P. Russell

Chapter 7. In-Plane Cyclic Testing of Flanged URM Walls

which opened as a result of diagonal tension cracking. Priestley et al. (2007) suggest a drift limit of 0.4% for walls failing in a shear dominated response.

Walls A3, A3a and A5 reached an ultimate drift u of 0.02%, 0.01% and 0.05% respectively, all signicantly less than 0.4%. Although testing of Wall A3 was terminated due to concerns about the stability of the gravity blocks on top of the wall, the gravity load carrying capacity of Wall A3 was not actually compromised, and further testing would have been necessary in order to conrm the drift corresponding to collapse. Similarly, for Wall A3a, although the lateral force degraded to 0.8Vmax at a drift of 0.02%, displacement capacity and gravity load capacity was available up to a drift of 0.3%, when the test was terminated. The drift corresponding to collapse and loss of gravity load carrying capacity was not denitively obtained. The only wall where some indication that gravity load carrying capacity was compromised was Wall A5, where bricks became unstable in the top course at the centre of the in-plane wall. The wall as a whole did not lose gravity load carrying capacity. A comparison is made with Wall A2 which became laterally unstable at the conclusion of testing (see Section 6.5.5), and the gravity load carrying capacity of the whole wall was compromised.

Walls A6, A7 (in both directions) and A8 all attained an ultimate drift u in excess of 0.4%, and did not lose gravity load supporting capacity. The ultimate drift of Wall A6, A7 (ange in tension), A7 (ange in compression) and A8 was 0.95%, 0.74%, 0.63% and 0.96% respectively. From the results of Walls A6 A8, an ultimate drift limit greater than 0.4% can be suggested for walls failing in diagonal tension failure, followed by sliding (from the results of Walls A3 A5, further testing would have been required to conrm this drift limit). The lowest value of ultimate drift from Walls A6 A8 was 0.63%. Thus a value of ultimate drift of 0.6% is recommended.

7.6.6

Initial Stiness

The initial stiness (Ki ) of Walls A6 A8 before eective yielding was determined from
Alistair P. Russell

- 208 -

7.6. Discussion

Figures 7.13 7.15, and is shown for both positive and negative loading directions in Table 7.5 (where Kim refers to the measured initial stiness and Kic refers to the calculated initial stiness). ASCE (2007) suggests that laboratory tests of solid shear walls have shown that behavior can be depicted at low force levels using conventional principles of mechanics for homogeneous materials. In such cases, the lateral in-plane stiness of a solid cantilevered shear wall, Ki , can be calculated using [Equation (7.7)]. 1 he 3 3Em Ie + he Av Ge

Ki =

(7.7)

Equation (7.7) is adequate for determining the initial stiness of uncracked URM walls when subjected to low levels of lateral force. But URM walls are not homogeneous materials, and moreover there is some cracking before the eective yield point is reached. As such, modications to the input parameters are necessary in order to satisfactorily determine the initial stiness before eective yield. he is the eective height of the wall (which is the full height for a cantilever wall), Em is the masonry Youngs modulus and Gm is the masonry shear modulus, with Em and Gm taken from Table 5.2 as 2.25 GPa and 0.9 GPa

Table 7.5: Eective initial stiness of anged walls Wall Kim kN/mm Wall A6 (+) Wall A6 (-) Wall A7c (+) Wall A7t (-) Wall A8 (+) Wall A8 (-) Mean COV 34.2 27.6 29.0 45.9 44.3 38.8 36.6 21%
- 209 -

Kim /Kic 0.95 0.77 0.80 1.27 1.23 10.8 1.02 21%

Alistair P. Russell

Chapter 7. In-Plane Cyclic Testing of Flanged URM Walls

respectively, and Av is the shear area of the wall, determined from Equation (7.8). Av = bwe lwe (7.8)

For two-leaf walls as reported in this chapter, bwe should be taken as 2/3bw , because the collar joint between leaves is typically not completely lled, but because of the header bricks being located every 4th course, some connection can be reasonably assumed. lwe is the eective clear length between anges, determined from Equation (7.9) when anges are present at both ends of the wall, and from Equation (7.10) when the ange is at one end only. lwe = lw 2bf lwe = lw bf (7.9) (7.10)

Because the URM walls considered in this chapter are not homogeneous, and because cracking occurs before the eective yield point is reached, Ie (the eective moment of inertia) should be taken as a proportion of Ig (the gross moment of inertia). Ig is determined from Equation (7.11). Similarly, the shear stiness of cracked masonry should be taken as a fraction of the uncracked shear stiness. bwe lwe 3 12

Ig =

(7.11)

For determining the eective stiness, in terms of exural rigidity, ASCE (2007) suggests that for reinforced concrete walls, the cracked section stiness should be taken as half the uncracked stiness. Similarly, for reinforced masonry walls, ASCE (2007) permits the use of a cracked moment of inertia equal to 0.5Ig . For URM walls, little guidance is given on what proportion of Ig is appropriate to use to determine the cracked section stiness. Furthermore, ASCE (2007) states that the shear stiness of post-cracked masonry should be taken as a fraction of the initial uncracked masonry shear stiness value, but no specic value is given. On the basis of these suggestions, Ie = 0.5Ig and Ge = 0.2Gm are approximations that correlate the initial stiness as calculated by Equation (7.7) (Kic ) with the measured initial stiness of Walls A6 A8 (Kim ), as shown in Table 7.5. For a wall with h = 2 m, lw = 4 m and bw = 240 mm (2 leaves), which corresponds with the in-plane
Alistair P. Russell

- 210 -

7.6. Discussion

wall (web) dimensions of Walls A6 A8, Av = 469.3 103 mm2 , Ie = 290.7 109 mm4 , Ge = 0.18 GPa and Equation (7.7) gives Kic = 36.0 kN/mm. The average Kim from the experimentation of Walls A6 A8 (as shown in Table 7.5) was 36.6 kN/mm, with a COV of 21%.

From the results of 11 URM wall tests Magenes and Calvi (1997) found that the evaluation of an equivalent yield point presented a large dispersion, whilst the drift at ultimate conditions was found to be extraordinarily uniform. They also found that when ultimate conditions are considered, strength and ultimate displacement are the dominating parameters, and the determination of the initial stiness plays a lesser role. The denition of a strength and of an ultimate displacement allows an easy denition of a secant stiness to ultimate conditions. Sections 7.6.5 and 7.6.8 give details on the determination of drift and strength at ultimate conditions respectively, and Section 7.7 provides details on the general force-displacement response of wall tests reported in this chapter.

7.6.7

Crack Pattern Analysis

Conceptualised sketches of the cracking patterns of Walls A4 A7 are shown in Figure 7.18. The length of the ange lf as shown in Figure 7.18 refers to the total ange length. The actual cracking patterns are shown in Figures 6.8 and 7.7.

An explanation as to why there are dierent crack patterns associated with walls having dierent ange characteristics can be made by comparing the properties of anged URM walls with the properties of thin-walled channel sections (typically steel beams). As shear force V is applied parallel to the web of a thin-walled section, which in this case is analogous to the in-plane wall, the distribution of shear stresses can be obtained from Equation 7.12 (Gere and Timoshenko, 1997). V

web =

bw I

y dA

(7.12)

When no anges are present the shear stress varies parabolically from = 0 at the ends
- 211 Alistair P. Russell

Chapter 7. In-Plane Cyclic Testing of Flanged URM Walls

(a) Wall A4 no ange

(b) Wall A5 ange length = 5bf

(c) Wall A6 ange length = 9bf

(d) Wall A7 ange length = 9bf , one end only

Figure 7.18: Flange eects on the orientation of cracking of in-plane wall

of the web where y = lw /2, to a maximum of max = 3V /2lw bw at the centre of the web. The ratio of maximum shear stress to the shear stress at the ends of the web is innity. When anges are present, the distribution of shear stress in the anges can be obtained

Figure 7.19: Shear stress distribution for a shear force applied parallel to the web
Alistair P. Russell

- 212 -

7.6. Discussion

from Equation (7.13) (Gere and Timoshenko, 1997), V x lw 2I

f lange =

(7.13)

and the shear stress varies with x linearly along the length of the ange from = 0 at the ends of the ange where x = 0, to a maximum of V lf /2I at the web-ange junction where x = lf (for a ange on one side). For continuity of shear ow, the shear stress at the end of the web is equal to the maximum shear stress in the ange (see Figure 7.19). Assuming that the thickness of the ange and the thickness of the web are equal, the ratio of the maximum shear stress at the centre of the web to the maximum shear stress at the web-ange junction can be obtained from Equation (7.14) (Gere and Timoshenko, 1997). web,max f lange,max lw 4lf

=1+

(7.14)

For a constant lw (lw = 4000 mm for Walls A4 A8), the shear stress at the ends of the in-plane wall increases with increasing ange length, relative to the maximum shear stress at the centre of the wall. Then, for a constant tensile cracking strength of masonry fdt , the angle of cracking increases with increasing shear stress. This increase is because the angle between the direction of maximum principal tension stress and the state of pure stress decreases as the maximum shear stress increases (see Figure 7.20, where 1 , 2 and are

(a) Principal stress directions when no ange is present

(b) Principal stress directions when long ange is present

Figure 7.20: Direction of principal stresses at ends of in-plane wall

- 213 Alistair P. Russell

Chapter 7. In-Plane Cyclic Testing of Flanged URM Walls

the maximum principal (tension) stress, the minimum principal (compression) stress and shear stress, respectively). Consequently, walls with longer anges will develop steeper diagonal tension cracks near the ends of the in-plane wall when compared with walls with shorter anges (see Figure 7.18). This eect lessens with increasing ange length, and when the ange is longer than the eective length (6bf ), the dierence in crack angle is negligible.

7.6.8

Predicted Behaviour and Measured Behaviour

Table 7.4 shows the predicted nominal shear strength Vn as determined by the model developed by Yi et al. (2008), the maximum shear strength Vmax attained during testing, and the shear strength corresponding to rst cracking Vcrack . Table 7.6 shows the ratio of Vmax /Vn for the anged walls reported in this chapter. It can be seen that the average ratio of Vmax /Vn was 1.10, with a COV of 24%. For all the walls, the COV is high, but if the walls with an aspect ratio of 1:2 (Walls A5 A8) only are considered, then the average ratio is 0.97 with a COV of 7%, indicating a high level of accuracy in the model Table 7.6: Predicted and measured strengths of anged walls Wall Vn kN A3 A3a A5 A6 A7t A7c A8 Mean COV
- 214 -

Vmax kN 54.9 72.5 66.0 66.8 75.0 61.9 66.9

Vmax /Vn

31.4 45.6 63.9 75.4 75.0 60.4 75.0

1.75 1.59 1.03 0.89 1.0 1.02 0.89 1.10 24%

Alistair P. Russell

7.6. Discussion

developed by Yi et al. (2008).

There was a particularly high level of accuracy in the model when determining the limiting shear strengths of Wall A7 both when the ange was in tension and in compression. Because the model by Yi et al. (2008) was developed for walls with a single ange (at any position on the in-plane wall) it seems reasonable to expect this accuracy. Moreover, for walls with an aspect ratio of 1:2, the model was also particularly accurate, although the measured strengths of Walls A6 and A8 were both 11% lower than the predicted strength.

For Walls A5 A8, the strengths predicted by the guidelines available from NZSEE (2006) and ASCE (2007) were 53 kN and 46 kN respectively (the same strength was predicted for each of Walls A5 A8, as the inuence of anges was neglected). Walls A5 A8 all attained a higher strength than the values predicted by both NZSEE (2006) and ASCE (2007) (see Table 7.7).

Although the model developed by Yi et al. (2008) underestimated the lateral strengths and incorrectly predicted the failure mode of Walls A3 and A3a, anged walls with an aspect ratio of 1:1 (A3) and 1:1.6 are unlikely to occur in real URM buildings (piers with such aspect ratios are possible, but are likely to be bounded by windows, see the companion study of Knox (2012) for further details). Table 7.7: Predicted strengths of Walls A5 A8, neglecting the inuence of anges Wall Vmax kN A5 A6 A7t A7c A8 Mean 66.0 66.8 75.0 61.9 66.9 Vn,N ZSEE kN 53.0 53.0 53.0 53.0 53.0 1.25 1.26 1.42 1.17 1.26 1.27 Vmax /Vn,N ZSEE Vn,ASCE kN 46.0 46.0 46.0 46.0 46.0 1.43 1.45 1.63 1.35 1.45 1.46 Vmax /Vn,ASCE

- 215 -

Alistair P. Russell

Chapter 7. In-Plane Cyclic Testing of Flanged URM Walls

7.6.9

Consolidation of Predictive Equations

Vs =

N + 3ai bw vme 1+ N he 3he bw vme N 2 N ai lw 3 fm lw bw

(7.1)

Vtc =

lw

(7.2)

fm Vdt = bw lw bfdt 1 + fdt

Wf Wf 1 Wf 1 1 + 1 + + Ww 6 Ww 2 Ww 2 Wf 1 + Ww 2 lw

(7.3)

lw bw ai 2

Vr =

N he

lw 2 bw 6

+ (ai af )Af 1 Af

af

(7.4)

lw bw 2

Equations (7.1) (7.4) were proposed by Yi et al. (2008) for predicting the limiting strength of anged URM walls, where Vs , Vtc , Vdt and Vr are the lateral strengths of a anged URM wall corresponding to sliding, toe crushing, diagonal tension and rocking, respectively. The limiting strength is taken as the minimum as determined from Equations (7.1) (7.4).

Equation (7.3) proposed by Yi et al. (2008) was validated using the experimental results of Walls A5 A8, with a high level of correlation. Consequently Equation (7.3) is recommended for predicting the diagonal tension strength of anged URM walls. Equations (7.1), (7.2) and (7.4) were not validated in this testing programme, as Walls A5 A8 failed in a diagonal tension failure mode.

Alistair P. Russell

- 216 -

7.7. General Force-Displacement Response of URM Walls With Flanges

7.7

General Force-Displacement Response of URM Walls With Flanges

Figure 7.21 shows a general force-deformation response for URM walls responding inplane. Figure 7.21(a) corresponds to deformation controlled masonry elements and Figure 7.21(b) corresponds to force controlled masonry elements. Deformation controlled elements have some displacement capacity beyond the point of eective yield, but force controlled elements respond in a brittle manner, and have no further capacity beyond when the maximum force is attained. In Figure 7.21(a), point B corresponds to the point of eective yield and Vy is the corresponding shear force. Point C is the location at which the ultimate drift is reached, and this point is the limit beyond which there is no further lateral force resisting capacity of the masonry element. Point E is the location where there is no further gravity load carrying capacity, and the wall can be considered to have collapsed at this point.

The experimentation reported in this chapter conrmed that Equation (7.3) can be used for predicting the diagonal tension strength of anged URM walls. In addition, Equations (7.1), (7.2) and (7.4) have been proposed by Yi et al. (2008) for determining the maximum shear force for URM walls failing in other modes. Vy can be determined accordingly. The results reported in Section 7.6.6 suggested that the initial stiness of in-plane

(a) Deformation controlled components

(b) Force controlled components

Figure 7.21: Generalised force-deformation relation for masonry elements or components (reproduced from Figure 7-1, ASCE (2007))
- 217 Alistair P. Russell

Chapter 7. In-Plane Cyclic Testing of Flanged URM Walls

masonry walls can satisfactorily be predicted using Equation (7.7). Thus the force and drift corresponding to point B in Figure 7.21(a) can be determined.

For masonry in-plane walls subjected to lateral force, diagonal cracking occurs due to the principal tension stress 1 exceeding the diagonal tension strength of the masonry fdt . When the mortar is weaker than the bricks, the cracking can be expected to occur in a step-wise manner (see Section 5.3 and Figure 5.7(b)). Derakhshan et al. (2010); Dizhur et al. (2009, 2010a,b) and Dizhur and Ingham (2010) note that for existing URM buildings in New Zealand the quality of mortar is usually poorer than the quality of bricks, and as such cracking through the mortar joints is more likely (also see Section 6.5.7)3 . Once a stair-stepped crack is opened up, sliding can be expected to occur along the bedjoints, as was observed in the experimentation of Walls A5 A8, and deformation of the masonry wall can attain an ultimate drift of u = 0.6% in this manner, as per Section 7.6.5.

Beyond the ultimate drift u (corresponding to point C in Figure 7.21(a)) the in-plane loaded wall can be expected to have negligible lateral force resisting capacity, but the gravity load carrying capacity of the wall is not compromised until a greater drift level is reached. The drift which corresponds to loss of gravity support for primary walls responding in bed-joint sliding is equal to the drift corresponding to half a brick length. Figure 7.25(d) shows bed-joint sliding before a displacement equal to half a brick length is attained. When a section of wall on one side of a stair-stepped crack displaces more than half a brick length (usually equal to a brick width) that section of wall will suddenly displace vertically by a distance equal to the height of one brick course. At this displacement, the wall can be considered to have lost gravity load carrying capacity. Consequently the drift e corresponding to point E in Figure 7.21(a) corresponds to a displacement of half a brick length. For typical New Zealand clay bricks, this corresponds to 120 mm (see Figure 3.38).

General force-deformation relations for in-plane masonry walls responding in the manner
In the case when the mortar is stronger than the bricks, cracking due to diagonal tension will occur through the bricks, and there will not be further displacement capacity beyond the point where cracking occurs. This is a force-controlled mechanism, and is depicted in Figure 7.21(b). Further information on relative brick and mortar strengths can be found in Lumantarna (2012).
3

Alistair P. Russell

- 218 -

7.8. Conclusions

reported in this chapter can be modelled using the above parameters.

7.8

Conclusions

By comparing the results of the wall tests reported in this chapter with those reported in Chapter 6, it is concluded that walls with anges can sustain higher levels of lateral force than walls without anges, when taking into account the wall aspect ratio and axial load. Walls A5 A8 had the same in-plane aspect ratio and comparable levels of axial load as Wall A4, and the limiting strength of each of the anged walls was higher than that attained by Wall A4. The inuence of anges must therefore be taken into account when attempting to accurately assess the capacity of walls to withstand in-plane lateral forces generated by earthquakes.

All the walls reported in this chapter failed by diagonal tension cracking, which is considered a shear dominated response. Walls A6 A8 subsequently commenced sliding on the bed-joints which opened as a result of diagonal tension cracking. A limiting drift of 0.4% is suggested in the literature as the ultimate drift capacity of in-plane loaded walls when the response is shear dominated. Diagonal tension failure is normally considered a force-controlled action, but when diagonal tension cracking occurs in such a way that sliding occurs subsequently on the cracked bed-joints, this behaviour is considered a deformation-controlled action as there is residual displacement capacity beyond when the peak force is attained. The lowest value of ultimate drift from the results of testing Walls A6 A8 was 0.63%. Thus for walls failing in a diagonal tension mode followed by bed-joint sliding, a value of ultimate drift u of 0.6% is recommended. Walls failing in a brittle mode, such as sudden cracking of the ange in Wall A3a, can also be concluded to have some some residual strength and displacement capacity available. This post-peak response is not always observed in laboratory experimentation of large scale walls, as subjecting a wall specimen to larger lateral displacements can constitute a safety hazard.

Equation (7.3) proposed by Yi et al. (2008) was validated using the experimental results
- 219 Alistair P. Russell

Chapter 7. In-Plane Cyclic Testing of Flanged URM Walls

of Walls A5 A8, with a high level of correlation. Consequently Equation (7.3) is recommended for predicting the diagonal tension strength of anged URM walls.

The initial stiness of URM walls failing in diagonal tension followed by bed-joint sliding was satisfactorily modelled using Equation (7.7), with modications made to the shear modulus Gm and eective moment of inertia Ie , to account for cracking and the nonhomogeneous nature of masonry as a material. The post-cracked shear stiness Ge is recommended to be taken as 0.2Gm , and Ie is recommended to be taken as 0.5Ig .

Determining the initial stiness, the strength and the ultimate drift of URM walls responding in-plane in a behaviour mode of diagonal tension cracking followed by bed-joint sliding enabled the general force-displacement response to be modelled, as per Figure 7.21.

Finally the length of anges at the ends of in-plane URM walls can be concluded to inuence the orientation of cracking. It was observed that for walls with no anges (Wall A4, reported in Chapter 6) the cracking pattern showed cracks propagating at approximately 45 through the centre of the wall. This cracking pattern was also observed in the pull cycle of Wall A7, where there was no ange in compression. For walls with longer anges, it was observed that the greater the length of the ange, the steeper the orientation of cracking. This inuence on cracking angle was particularly evident in Wall A6 with lf = 9bf , where in this instance lf refers to the total ange length. This cracking pattern was also evident in the test of Wall A7 in the push direction, where the ange was in compression.

Alistair P. Russell

- 220 -

7.9. Photos of Wall Response

7.9

Photos of Wall Response

(a) Timber diaphragm on top of wall, prior to testing

(b) Shear crack on push cycle

(c) Shear crack originating from timber joist

(d) In-plane wall crack

(e) Flange cracking on pull cycle

(f) Flange cracking on push cycle

Figure 7.22: Photos of Wall A3 response

- 221 -

Alistair P. Russell

Chapter 7. In-Plane Cyclic Testing of Flanged URM Walls

(a) Wall A3a setup prior to testing

(b) Base course crack on pull cycle

(c) Base course crack on push cycle

(d) Wall sliding on pull cycle

(e) Flange damage at conclusion of test

(f) Flange damage at conclusion of test

Figure 7.23: Photos of Wall A3a response

Alistair P. Russell

- 222 -

7.9. Photos of Wall Response

(a) Crack on outside of ange

(b) Diagonal shear cracks on push cycle

(c) Diagonal shear cracks on pull cycle

(d) Diagonal shear cracks on pull cycle

(e) Wall damage at conclusion of test

Figure 7.24: Photos of Wall A5 response

- 223 Alistair P. Russell

Chapter 7. In-Plane Cyclic Testing of Flanged URM Walls

(a) In-plane wall shear cracking on pull cycle

(b) Cracking in the ange at the compression toe on pull cycle

(c) In-plane wall shear cracking on push cycle

(d) In-plane wall shear cracking on pull cycle

(e) Global tension ange uplift on pull cycle

(f) Wall damage at conclusion of test

Figure 7.25: Photos of Wall A6 response

Alistair P. Russell

- 224 -

7.9. Photos of Wall Response

(a) In-plane wall shear cracking next to compression toe ange (push cycle)

(b) In-plane wall shear cracking on pull cycle (ange in tension on pull cycle)

(c) Toe crushing damage at non-ange end at con- (d) Overall cracking pattern at conclusion of test clusion of test

Figure 7.26: Photos of Wall A7 response

- 225 -

Alistair P. Russell

Chapter 7. In-Plane Cyclic Testing of Flanged URM Walls

(a) Cracking on the outside of the tension ange on (b) Cracking on the inside of the compression ange push cycle on push cycle

(c) In-plane wall shear cracking next to compression (d) In-plane wall shear cracking next to compression toe ange (push cycle) toe ange (pull cycle)

(e) In-plane wall shear cracking next to compression (f) Overall cracking pattern at conclusion of test toe ange (push cycle)

Figure 7.27: Photos of Wall A8 response

Alistair P. Russell

- 226 -

CHAPTER 8

URM Building Assessment Procedure

This chapter presents an assessment procedure for analysing URM buildings. The seismic performance of a building is dependent on a number of critical components, with particularly important aspects being walls responding in-plane, walls responding out-of-plane and the response of diaphragms. The aim of this chapter is to provide a procedure for assessing the performance of the whole building, by assessing each of the components and determining the critical building elements. Because the research in this thesis is primarily associated with the response of walls subjected to in-plane loads, full details for assessing the diaphragm and out-of-plane wall response are not included in this chapter. Instead steps are allocated in the procedure for assessment of these other components, with no specic details provided for their assessment.

The need for a transparent and logical procedure for assessing the expected performance of URM buildings when subjected to seismic actions has been recognised. Such a procedure should clearly identify what seismic demand the structure is expected to be potentially subjected to, and how the structure can be analysed to determine its capacity to withstand that seismic demand. Practitioners have also expressed dissatisfaction with detailed nite element modelling for assessing low rise URM buildings, particularly because of the
- 227 Alistair P. Russell

Chapter 8. URM Building Assessment Procedure

inherent uncertainty in the material properties of URM, and because of the comparatively low nancial value of this building type. Moreover, the criteria against which the performance of the building is to be measured needs to be clearly dened and understood.

As detailed in Chapter 2, an initial evaluation procedure (IEP) is designed to be used to relatively quickly and easily assess if a building is likely to have an expected earthquake performance of < 33%NBS. The objective of the IEP is to allow Territorial Authorities to identify as closely as possible all earthquake-prone buildings within their jurisdiction. If a building is found to be potentially earthquake prone (see Chapter 2 for details) the owner has the option to conduct a detailed assessment to determine with a greater level of accuracy the expected earthquake performance. As the preponderance of URM buildings in New Zealand are likely to be one and two storeys in height, and individually of relatively low nancial value (see Chapter 4), a detailed assessment procedure which is not overly time consuming and expensive to implement is necessary. The procedure proposed in this chapter is intended to be implemented without complex software; indeed it could be termed a desktop assessment procedure.

One of the original aims of this research was to reduce the conservatism in assessment methods, in order to allow the inherent seismic resistance of a building to be accurately utilised, thus requiring less invasive and/or less expensive retrot interventions. Consequently, exclusively force-based methods for determining base-shear, relying on the initial stiness of the building, are not included here. Despite the relative familiarity among designers with tried-and-true force-based methods, the high levels of conservatism inherent in such approaches are considered to overestimate the seismic actions which URM buildings are subjected to.

A displacement-based approach as outlined in this chapter can be used to assess the expected performance of a URM building, within the criteria of The Building Act 2004, particularly for determining if the building is potentially earthquake prone. It has become widely accepted that structural damage is related to material strains, which can be related to structural displacements, and that non-structural damage of buildings is
Alistair P. Russell

- 228 -

8.1. New Zealand Existing Building Seismic Performance Criteria

related to lateral drift levels. Direct Displacement-Based Design (DDBD) is considered a more rational approach for designing structures than force-based design, or as appropriate here, for assessing a buildings expected performance.

8.1

New Zealand Existing Building Seismic Performance Criteria

The legislative requirements outlined in The Building Act 2004 (New Zealand Parliament, 2004) and its associated regulations for the seismic performance which existing buildings must meet have been covered in detail in Chapter 2. Relevant parts are repeated here for emphasis. A building is earthquake prone for the purposes of this Act if the building will have its ultimate capacity exceeded in a moderate earthquake, and a moderate earthquake is dened as an earthquake that would generate shaking at the site of the building that is of the same duration as, but that is one third as strong as, the earthquake shaking that would be used to design a new building at that site. The New Zealand Society for Earthquake Engineering (NZSEE) has interpreted these denitions so that a building which is earthquake-prone would essentially perform in a design earthquake at a level less than 33% of the standard of a new building, referred to as 33% New Building Standard, or 33%NBS (NZSEE, 2006). This denition relates to the buildings ultimate limit state capacity, as dened by current design standards. As such, an existing building must be assessed to determine whether its ultimate limit state capacity in an earthquake is greater than or less than one-third of the ultimate limit state capacity of a new building constructed on the same site. The limit of 33%NBS is prescribed by legislation, but it is the view of NZSEE that 67%NBS would be a more appropriate target. It is considered that 33%NBS corresponds to approximately 20 times the risk of the building reaching a similar condition to that which a new building would reach in a full design earthquake, and 67%NBS corresponds to a risk of 3 times (see Figure 8.1). It is the decision of the building owner as to the extent of seismic retrot necessary to elevate the performance of the building, with 33%NBS as a required minimum.

- 229 -

Alistair P. Russell

For existing buildings:

!" adjustment to material properties normally used for new buildings may be permitted to

Chapter 8. attainment URM Building Assessment Procedure of strength and similar parameters.

recognise known or measured values, or a greater or lesser confidence than normal in the

Figure 8.1: Strength versus risk and ultimate limit state as reference point (reproduced from NZSEE (2006))
Section 4 Detailed Assessment General Issues
30/06/2006

Figure 4.1: Strength versus risk and ULS as reference point

4-3

When conducting a seismic performance assessment, it is important to take into account that the legislation refers to a new building at that site. Consequently, the importance level and design working life should be considered as if a new building were being designed. A 50 year design life with importance level 2 corresponds to an annual probability of exceedance for the ultimate limit state of 1/500 for earthquakes (Standards New Zealand, 2002a). In addition to consideration of importance level, heritage conservation principles need to be taken into account, and for structures with particular heritage value, a longer design life (100 years) or a higher importance level of 3 could be considered. This assessment of heritage value needs to be balanced against the invasiveness of any required
Alistair P. Russell

- 230 -

8.2. Direct Displacement Based Design

seismic retrot, as outlined in Chapter 2. In general though, an annual probability of exceedance of 1/500 will be appropriate and corresponds to a return period factor R = 1 for the design of a new building of importance level 2.

As 1/3rd of the standard of a new building is the eective performance requirement stipulated by The Building Act 2004, one approach to assessing a buildings performance within this criteria is to simply scale the seismic demand (viz. Vbase ) that the building is subjected to by 1/3rd and to then compare that reduced demand with the assessed capacity of the building. This comparison provides a simple pass or fail criteria for determining if the building is potentially earthquake-prone. But because the determination of Vbase is not linear (see Section 8.2), it is not appropriate to simply scale the demand by 1/3rd . The approach advocated in this chapter is to assess the response of each part of the structure, determining %NBS for that component by assessing capacity/demand. Then compared with the limit of 33%NBS, the building is determined as potentially earthquake prone or not. This procedure enables targeted retrot interventions for critical components (if necessary), and moreover, this approach is considered fundamentally more rational.

8.2

Direct Displacement Based Design

Signicant research has been conducted since the early 1990s into developing the direct displacement based design procedure. The procedure was developed with the aim of providing a greater emphasis on displacements throughout all design stages. Using this method, a structure is designed to achieve a predened level of lateral displacement when subjected to a predened level of seismic intensity. DDBD utilises the concept of an equivalent linear system (ELS) dened by an equivalent damping and equivalent stiness to represent the response of a non-linear system. The periods and damping values used in displacement-based design (or assessment) are not identical to those used in force-based assessments, where force-based assessments consider the initial period and damping of the structure. A full treatment of DDBD can be found in the literature (see for example Aschheim and Black, 2000; Chopra and Goel, 2001; Priestley, 2002; Priestley and
- 231 Alistair P. Russell

Chapter 8. URM Building Assessment Procedure

Kowalsky, 2000; Priestley, 1993; Sullivan et al., 2003; Wight, 2006), and primarily in the seminal text on the topic by Priestley et al. (2007). The fundamental elements of a direct displacement-based seismic design can equally be applied as direct displacement-based seismic assessment. The basic steps of the DDBD approach as applied to a building are as follows (from Wight (2006)).

Step 1: Selection of a target displacement A target displacement is selected based on material strain damage criteria or directly by code dened drift limits. In the case of URM walls responding in-plane the drift limit is taken as either 0.6% for diagonal tension failure followed by bed-joint sliding or as 0.8% for a exure dominated failure mode, such as rocking (see Chapters 6 and 7). The building is represented by a SDOF oscillator, with an equivalent mass (me ) and equivalent height (he ). The response of the equivalent linear system can be determined from a displacement response spectra.

Step 2: Selection of Seismic Demand The seismic demand for DDBD may be expressed as a displacement response spectra generated for several levels of damping, as illustrated in Figure 8.2. Although displacement response spectra from real earthquake records have been used for validation studies of DDBD, assessment is likely to proceed through the use of simplied and smoothed codebased design spectra. Most loadings standards do not currently prescribe displacement

Figure 8.2: Displacement response spectra

Alistair P. Russell

- 232 -

8.2. Direct Displacement Based Design

spectral ordinates directly, thus requiring the designer to obtain displacement response from mapped acceleration response spectra. Conversion of a 5% damped acceleration response spectra to the corresponding 5% damped displacement response spectra is accomplished using Equation (8.1) where Sa5% and Sd5% represent the acceleration and displacement respectively at 5% damping for period T. The displacement response spectra for levels of equivalent damping eq greater than 5% can then be obtained by multiplying the 5% damped curve by R (Equation (8.2)), which is a reduction factor to account for damping, and is given in Equation (8.3)1 . Sa5% T 2 4 2

Sd5% =

(8.1) (8.2)

SdX % = Sd5% R 0.07 0.02 + eq

R =

(8.3)

Step 3: Calculation of building eective period Once the target displacement is selected in step 1, and the seismic demand established in step 2, the eective period at maximum response is obtained as shown in Figure 8.2. The designer enters the displacement response spectra with the value for the target displacement dT and reads across to the appropriate response curve and down to determine the eective period Te . The response curve that is selected is a function of the level of equivalent viscous damping eq for the system under consideration. From previously (see Chapters 6 and 7) equivalent viscous damping is taken as 15% for in-plane loaded URM walls with anges or for rectangular URM walls where force-controlled actions are critical, and 5% for rectangular URM walls where deformation-controlled actions are critical.

Step 4: Calculation of Design Base Shear Once the eective period of the substitute structure is obtained in step 3, the eective stiness Ke is obtained using Equation (8.4) (Equation (8.4) is a rearrangement of the more familiar Equation (8.5)). The eective stiness of the substitute structure is dened
Equation (8.3) is taken from Priestley et al. (2007), but there are slightly varying equations available for determining the damping reduction factor (see Eurocode 8, for example). NZSEE (2006) recommends the use of Equation (8.3).
1

- 233 -

Alistair P. Russell

Chapter 8. URM Building Assessment Procedure

Figure 8.3: Eective stiness as the secant stiness to maximum response, as shown in Figure 8.3. The design base shear force Vbase at the limit state is then obtained by multiplying the eective stiness by the target displacement (Equation (8.6)). 4 2 m e

Ke =

Te

Vbase = Ke dT

Te 2 m e = 2 Ke

(8.4)

(8.5) (8.6)

8.2.1

Determining Base Shear Demand

The steps described above can be combined to form the following expression for the base shear force demand (Equation (8.7)), where the variables dc and Tc represent the corner point spectral displacement (at 5% damping) and period respectively as shown in Figure 8.2. 4 2 me dc 2 dT Tc 2

Vbase =

(8.7)

The corner point period has been shown to be a function of moment intensity and distance to the seismic source (Faccioli et al., 2004). NZSEE (2006) states that in New Zealand, for sites where near fault eects are not an issue the displacement spectra are
Alistair P. Russell

- 234 -

8.2. Direct Displacement Based Design

well represented by a bilinear relationship pivoting around the displacement at T = 3 s and the horizontal leg beyond 3 s (see NZSEE, 2006, Section 5). Consequently, a corner period of T = 3 s can be used for displacement-based assessments of most URM buildings in New Zealand. Information on more detailed displacement response spectra can be found in NZSEE (2006). Once the corner point period is established, the corner point displacement can be obtained directly from the mapped spectral acceleration using Equation (8.1). Using the acceleration spectra provided in the New Zealand loadings standard NZS 1170.5 (Standards New Zealand, 2004a), the corner point displacement is found from Equation (8.8), Tc 2 4 2

dc =

Ch (T ) Z R N (T, D) g

(8.8)

where Ch (T ) is the spectral shape factor, Z , R, and N (T, D) are the hazard, return period and near fault factors respectively, and g is the acceleration due to gravity.

8.2.2

Distribution of Base Shear

For low rise URM buildings (typology A, B, C and D) a simple and convenient method for analysing the structure is an equivalent static analysis (ESA). In this method the base shear is distributed as equivalent static forces at each level of the structure. The New Zealand loadings standard NZS:1170.5 (Standards New Zealand, 2004a) for determining earthquake actions sets criteria for when the equivalent static method can be used. The structure must be less than 10 m tall, or the structure should not be classied as irregular and the largest translational period should be less than 2 seconds. NZSEE (2006) further notes that the equivalent static method is applicable when during a design level earthquake the lateral force resisting systems of the structure respond elastically, or the structure should have a low ductility demand ( < 2) provided that there are no significant geometric irregularities. It also states that the building may be up to 30 m tall. All these criteria are widely met in most low rise URM structures typically found in New Zealand and therefore ESA is considered an appropriate method of analysis.

- 235 -

Alistair P. Russell

Chapter 8. URM Building Assessment Procedure

Furthermore, the Commentary to NZS 1170.5 states There are limits on the applicability of the equivalent static method, but it is likely to remain the most commonly used. . . method as most buildings are less than three storeys in height. Because of the familiarity among engineers with using an ESA method for assessment, the procedure presented in this chapter using ESA is considered appropriate for assessing the performance of URM structures in New Zealand.

8.3

Assessment Procedure

The step-by-step procedure outlined in this section for assessing the expected performance of URM buildings is summarised in Figure 8.4.

Step 1: Dene Global Building Parameters

The overall building parameters will be known from the outset. The following parameters are recorded: Importance level Building design life Annual probability of exceedance Number of storeys Storey heights Building length (long direction) Building length (short direction) Wall lengths (long direction) Wall lengths (short direction) Wall thicknesses (long direction)
Alistair P. Russell

- 236 -

8.3. Assessment Procedure

1.

Dene Global Building Parameters

2.

Dene Material Properties

3.

Dene Site Characteristics

Determine displacement response spectra

4.

Collate Axial Loads

5.

Determine Wall Axial Loads

6.

Determine Seismic Weight

7.

Proportion Base Shear to Building Storeys

8.

Determine Base Shear

8a.

Base Shear for Out-of-plane Wall Response

8b.

Base Shear for In-Plane Wall Response (Deformation-controlled)

8c.

Base Shear for In-Plane Wall Response (Force-controlled)

8d.

Base Shear for Diaphragm Response

9a.

9b.

9c.

9d.

Determine Capacity of Out-of-plane Walls

Determine Capacity of In-plane Walls (Deformation-controlled)

Determine Capacity of In-plane Walls (Force-controlled)

Determine Capacity of Diaphragms and Connections

10a.

10b.

10c.

10d.

Determine %NBS of Out-of-plane Walls

Determine %NBS of In-plane Walls (Deformation-controlled)

Determine %NBS of In-plane Walls (Force-controlled)

Determine %NBS of Diaphragms and Connections

11.

Determine Critical %NBS

Figure 8.4: URM building assessment procedure owchart

- 237 -

Alistair P. Russell

Chapter 8. URM Building Assessment Procedure

Wall thicknesses (short direction) Floor areas Roof area Drift (and/or displacement) limits, as well as component damping, are dened depending on the behaviour mode as follows: For in-plane wall response controlled by diagonal tension and sliding shear, eq = 0.15 and u = 0.6%. For in-plane wall response limited by exure dominated actions, eq = 0.05 and u = 0.8%. Displacement limits for walls responding out-of-plane are given in Derakhshan (2011). Details on diaphragm damping and displacement are given in Wilson (2011).

Step 2: Dene Material Properties

If possible, material testing should be conducted to determine actual material properties. Details on material testing can be found in Lumantarna (2012). There is no substitute for actual in-situ sampling, and it is recommended that testing is conducted at three separate locations on each wall (Lumantarna, 2012). If the masonry is relatively homogeneous, samples from the building can be taken to determine c, and vme easily. For homogeneous materials, these parameters will give a reasonable estimate of the overall masonry shear strength. The following material parameters from the building are required to be recorded: m , unit weight of masonry
fm , compressive strength of masonry fmj , compressive strength of mortar

fb , compressive strength of bricks fbt , direct tensile strength of bricks

Alistair P. Russell

- 238 -

8.3. Assessment Procedure

c, cohesion , coecient of friction fdt , diagonal tensile strength of masonry vme , cohesive strength of masonry bed joint

Step 3: Dene Site Characteristics

The 5% damped acceleration response spectra can be determined from NZS 1170.5 (Standards New Zealand, 2004a), based on the location, soil type, return period factor and proximity to a fault line. This acceleration response spectra is then used to determine the displacement response spectra, as outlined previously, and for sites where near fault eects are not considered, a corner period of T = 3 s is used. The corner displacement dc is determined according to Equation (8.8), which is reproduced here. Tc 2 4 2

dc =

Ch (T ) Z R N (T, D) g

(8.8)

The following parameters are required and can be found from NZS 1170.5: Location Z Soil type R Ch (T ) N (T, D) As outlined in Section 8.1 R = 1 for most structures.

- 239 -

Alistair P. Russell

Chapter 8. URM Building Assessment Procedure

Step 4: Collate Axial Loads

Live loads on the oors and roof are determined according to NZS 1170.1 (Standards New Zealand, 2002b), and dead loads are determined from the building oor materials and occupancy characteristics. The weight of the walls (and any parapets) is determined from their volume and unit weight, as dened in Step 1. The live load is multiplied by the earthquake imposed action combination factor, E = 0.3.

Step 5: Determine Wall Axial Loads

The axial loads acting on each wall are determined from the weight of walls above, and from oor and roof loads which can be attributed to act on each wall. The tributary area for loads acting on each wall is shown in Figure 8.5, where it is assumed that lines radiating from each corner at 45 separate the boundaries between tributary areas.

(a) Long walls

(b) Short walls

Figure 8.5: Tributary areas for axial load on each wall

Alistair P. Russell

- 240 -

8.3. Assessment Procedure

The wall axial loads are recorded as both N , the axial force on the wall, and fm , the axial stress on the wall, based on the axial force and wall cross-sectional area.

Step 6: Determine Seismic Weight

The seismic weight is determined according to section 4.2 of NZS 1170.5, and is given by Equation (8.9). Wi = Gi + E Qi (8.9)

where Gi and E Qi are summed between the mid-heights of adjacent storeys. The seismic weight of each level Wi is recorded as well as the total seismic weight Wt of the entire building. The seismic weight of the bottom half of the ground storey is assumed to be lumped at the ground and does not participate in the analysis.

Step 7: Base Shear Distribution as Equivalent Static Forces

The base shear is distributed as equivalent static forces at each level of the building according to Equation (8.10).

Wi hi Fi = Ft + 0.92V n (Wi hi )
i=1

(8.10)

where Ft = 0.08V at the top level and zero everywhere else, and n is the number of storeys. The shear force applied to each level is the sum of the forces acting at the oor levels above, see Figure 8.6. This distribution of lateral forces assumes a shape function obtained from static deections due to the self-weight of the structure. When parapets are present on the top of the structure, it is appropriate to lump their mass at the roof level.

- 241 -

Alistair P. Russell

Chapter 8. URM Building Assessment Procedure

(a) Building model

(b) Equivalent static horizontal forces

(c) Storey shears

Figure 8.6: Two storey URM analysis The base shear is proportioned as a percentage to each storey. Because dierent base shears are determined in Step 8 according to limiting drifts and target displacements, it is useful to determine the proportion of base shear at each level, rather than assigning actual values in this step. For example, in a two storey building as modelled in Figure 8.6, level F2 might be 0.6V and F1 = 0.4V .

Step 8: Determine Base Shear Demand

As shown in Figure 8.4, the steps in the procedure from this point onwards separate according to the component being analysed. Because there are separate drift (and/or displacement) limits for each component, there are dierent eective periods and eective stinesses. Consequently, dierent levels of base shear demand are determined for each component. In the subsequent steps, the capacity of the component is determined, and then compared with the demand to determine the performance, dened as %NBS. Each component needs to be analysed, and the performance determined, as it is not known from the outset which component will be critical.

The seismic mass, used for determining the eective period Te in Equation (8.5) and the base shear Vbase in Equation (8.7), is determined using the seismic weight from Step 6,
Alistair P. Russell

- 242 -

8.3. Assessment Procedure

according to Equation (8.11). Wt g

m e =

(8.11)

8.3.1

Walls Responding Out-of-Plane

Steps 8a and 9a, determining the demand and capacity for walls responding out-of-plane, are not included here. Details can be found in Derakhshan (2011).

Step 10a: Determine %NBS of Walls Responding Out-of-Plane

The capacity of the critical out-of-plane wall response (Step 9a) is compared with the seismic demand on the out-of-plane wall (Step 8a and 9a), and the ratio determines the %NBS for that component.

8.3.2

Walls Responding In-Plane

Step 8b: Determine Base Shear Demand for Deformation-Controlled Walls Responding In-Plane

As detailed in Chapters 6 and 7, for in-plane wall response limited by exure dominated response, such as rocking, eq = 0.05 and u = 0.8%, and for walls responding in-plane where the response is controlled by diagonal tension followed by bed-joint sliding eq = 0.15 and u = 0.6%. The drift in the uppermost level will be critical, and the limiting drift is used to determine the target displacement. Priestley et al. (2007) state that for masonry walls responding in-plane the eective height can be assumed as he = 0.8h, with the exception of a single storey building where he = h. With the eective height and limiting drift dened, the target displacement is determined from Equation (8.12). dT = he u
- 243 -

(8.12)
Alistair P. Russell

Chapter 8. URM Building Assessment Procedure

Consequently, with the damping and target displacement establised, the buildings overall eective period is dened. The base shear is determined from Equation (8.7), and distributed to the storeys as per Step 7.

Step 9b: Determine Capacity of Deformation-Controlled Walls Responding In-Plane

For wall responding in-plane and limited by deformation-controlled actions, the capacity must be compared with the appropriate demand. The deformation-controlled capacity of the in-plane walls must be compared with the seismic demand determined in accordance with the parameters associated with deformation-controlled actions. The storey shear is divided between the number of wall lines in the direction parallel to the seismic demand force, and assigned according to lw 2 . For example, if there are two equal length parallel walls resisting the seismic demand force, the capacity of each wall is compared with half the demand on the storey.

Deformation-controlled actions are sliding shear and exure, and for walls without anges, Equations (6.14) (6.17) are used to determine the lateral force capacity. N he 1 N

Vr =

b 2 0.85fm w (if no in-situ material data are available)

(6.14)

Vs =

3czbw + N 1+ 3clw bw c N

(6.16)

Vs = vme An

(if in-situ material data are available)

(6.17)

For walls with anges, Equations (7.1) and (7.4) are used to determine the lateral force capacity. For anged walls responding in-plane, the eective length of the ange lf is the lesser of six times the ange thickness (bf ) or the total length of the ange.

Alistair P. Russell

- 244 -

8.3. Assessment Procedure

Further details on determining the capacity of perforated URM walls responding in-plane can be found in Knox (2012).

Vs =

N + 3ai bw vme 1+ 3he bw vme N lw bw ai 2 lw 2 bw 6 + (ai af )Af 1 Af af lw

(7.1)

Vr =

N he

(7.4)

lw bw 2

Step 10b: Determine %NBS of Deformation-Controlled Walls Responding In-Plane

The capacity of the critical deformation-controlled in-plane wall response (Step 9b) is compared with the deformation-controlled seismic demand on the in-plane wall (Step 8b), and the ratio determines the %NBS for that component.

Step 8c: Determine Base Shear Demand for Force-Controlled Walls Responding In-Plane

As detailed in Chapters 6 and 7, for in-plane wall response limited by force-controlled actions, such as diagonal tension cracking through the bricks and toe crushing, eq = 0.15 and u = 0.4%. The drift in the uppermost level will be critical, and the limiting drift is used to determine the target displacement. With the eective height and limiting drift dened, the target displacement is determined from Equation (8.12). dT = he u (8.12)

Consequently, with the damping and target displacement established, the buildings over- 245 Alistair P. Russell

Chapter 8. URM Building Assessment Procedure

all eective period is dened. The base shear is determined from Equation (8.7), and distributed to the storeys as per Step 7.

Step 9c: Determine Capacity of Force-Controlled Walls Responding In-Plane

For wall responding in-plane and limited by force-controlled actions, the capacity must be compared with the appropriate demand. The force-controlled capacity of the in-plane walls must be compared with the seismic demand determined in accordance with the parameters associated with force-controlled actions.

The primary force-controlled action is diagonal tension cracking through the bricks, and for walls without anges, Equation (6.15) is used to determine the lateral force capacity. lw fm 1+ Vdt = vme An he vme

0.67

lw he

(6.15)

For walls with anges, force-controlled actions are diagonal tension cracking through the bricks and toe crushing, and Equations (7.2) and (7.3) are used to determine the lateral force capacity.

Vtc =

N he

lw

ai lw

3 fm lw bw Wf

(7.2)

fm Vdt = bw lw bfdt 1 + fdt

1 Wf 1 Wf 1 + 1 + + Ww 6 Ww 2 Ww 2 Wf 1 + Ww 2

(7.3)

Alistair P. Russell

- 246 -

8.3. Assessment Procedure

For anged walls responding in-plane, the eective length of the ange lf is the lesser of six times the ange thickness (bf ) or the total length of the ange.

Step 10c: Determine %NBS of Force-Controlled Walls Responding In-Plane

The capacity of the critical force-controlled in-plane wall response (Step 9c) is compared with the force-controlled seismic demand on the in-plane wall (Step 8c), and the ratio determines the %NBS for that component.

8.3.3

Response of Diaphragms and Connections

In the same way that Steps 8a and 9a, for assessing out-of-plane loaded walls, were omitted, Steps 8d and 9d for determining the demand and capacity for diaphragms and connections2 are not included here, and details can be found in Wilson (2011).

Step 10d: Determine %NBS of Diaphragms and Connections

The capacity of the diaphragm and wall-diaphragm connections (Step 9d) are compared with the seismic demand on the diaphragm (Step 8d), and the ratio determines the %NBS for that component.

Step 11: Determine Critical Performance

The minimum %NBS from Step 10 for each of the components is used to determine the critical component in the building.

While it has not been explicitly investigated in the research presented in this thesis, it has been informally observed that the most vulnerable elements in URM buildings are likely to be the walldiaphragm connections, particularly with regard to de-seating due to the lack of positive connection between the wall and diaphragm. Although not advocated as a formal (prescriptive) step in this process, it is likely that if any retrot interventions are found to be necessary in Step 11, the rst requirement will be to secure the diaphragm to the walls.
2

- 247 -

Alistair P. Russell

Chapter 8. URM Building Assessment Procedure

Retrot

From the %NBS as determined from Step 11, the owner then has the option to retrot any components which are critical (< 33%NBS) and re-assess the buildings performance.

8.4

Conclusions

An assessment procedure for unreinforced masonry buildings has been presented in this chapter, within the framework of the existing direct displacement based design method. The use of direct displacement based design principles was chosen as it is considered that this approach is fundamentally more rational and accurate, and takes dierent failure modes into account, compared with exclusively force-based approaches. Such force-based methods are considered to overestimate the seismic demand forces on a building, resulting in added cost being borne by building owners when initiating seismic improvements.

The drift limit of = 0.6% and equivalent viscous damping of eq = 0.15 for walls responding in-plane and limited by diagonal tension cracking followed by bed-joint sliding were conrmed in previous chapters and were applied to the displacement based assessment method proposed in this chapter. Similarly, the drift limit of = 0.8% and equivalent viscous damping of eq = 0.05 for walls responding in-plane and limited by exure were conrmed in previous chapters and were applied to the assessment method. These parameters are required for determining the eective period of the building, which is subsequently used to determine the seismic base shear demand.

Further research (reported elsewhere) is required to complete the detailed steps in the procedure, particularly for URM walls responding out-of-plane, for determining the seismic response of timber diaphragms and for determining the capacity of wall-diaphragm connections. The assessment method presented here is primarily a procedural basis. Once the detailed steps have been populated with further data and nal recommendations (from other research undertaken concurrently) this procedure will form a method for assessing
Alistair P. Russell

- 248 -

8.4. Conclusions

the performance of URM buildings.

An example URM building assessment is given in Appendix B.

- 249 -

Alistair P. Russell

Chapter 8. URM Building Assessment Procedure

Alistair P. Russell

- 250 -

CHAPTER 9

Summary of Conclusions

The research reported in this thesis was conducted with the primary aim of developing a deeper understanding of the response of URM buildings, with an emphasis on the response of URM walls responding in-plane, and in particular, the in-plane response of anged URM walls. Most previous expressions for determining the lateral strength capacity of URM walls responding in-plane have not taken into account the eect of anges, and consequently have underestimated the expected capacity. It was identied that in order to accurately account for the strength inherently available in URM buildings, the eect of anges should be included. Moreover, by reducing the conservatism in the predicted strength of URM buildings, less comprehensive and invasive retrot interventions are necessary for improving the seismic performance of such buildings. It was also identied that a coherent and logical assessment procedure was necessary for assessing the performance of URM buildings when subjected to seismic actions. This was achieved with the development of the procedure presented in Chapter 8.

The rst step in developing a deeper understanding of URM buildings was to assess the New Zealand URM building stock. This assessment was achieved in two parts. The rst part classied New Zealand URM buildings into typologies and summarised the typical
- 251 Alistair P. Russell

Chapter 9. Summary of Conclusions

building characteristics. The second part was an analysis of the entire URM building stock to determine the prevalence, distribution and estimated nancial value of URM buildings.

Once the characterisation of the URM building stock was complete, the focus of the thesis changed from broad level contextualisation to detailed experimentation and analysis of URM walls responding in-plane. The experimental programme was formed on the basis of results from assessment of the URM building stock, such that walls which represented construction characteristic of existing New Zealand URM buildings were investigated.

The rst stage of experimentation constituted small scale structural testing of wall panels, with the aim of investigating shear strength and material properties. This series of testing consisted of eight wall panels tested in diagonal shear, with dierent mortar properties and bond patterns.

A series of large scale straight rectangular URM walls was subsequently tested to determine the in-plane response of URM walls. Three walls were tested, and the predicted lateral strengths using available expressions from the literature were correlated against the experimental results. This phase of experimentation was used to determine which expressions were most suitable for predicting the in-plane response of straight rectangular URM walls.

Following the testing of rectangular walls without anges, an experimental programme investigating the in-plane response of walls with anges was undertaken. The results from six tests showed that the performance of in-plane walls was better than that of walls without anges. Moreover, the lateral strength capacity of anged walls was found to be greater than could be predicted using equations where the eect of anges is not included. An analytical model accounting for ange eects which was recently proposed in the literature was correlated against the results of this experimental programme with a high level of accuracy.

Finally, the direct displacement based design approach was presented in Chapter 8 and
Alistair P. Russell

- 252 -

9.1. Architectural Characterisation Chapter 3

used to develop a procedure for assessing the seismic performance of URM buildings. A worked example was provided in Appendix B to demonstrate the method.

The important conclusions arising from this research into the in-plane seismic assessment of unreinforced masonry buildings in New Zealand are summarised in the following sections.

9.1

Architectural Characterisation Chapter 3

In this chapter New Zealand URM buildings were classied into typologies, based on their general structural conguration. Distinctions between typologies were made on the basis of building height and the geometry of the buildings footprint. Seven typologies (Typologies A G) were identied as one storey isolated, two storey isolated, three-and-higher storey isolated, one storey row, two storey row, three-and-higher storey row structures, and URM structures which do not have a uniform ground footprint or which do not t easily into the rst six typologies. Moreover, representative material properties and wall geometries were also outlined, and details of how these characteristics relate to the overall typological classications were presented.

9.2

Characterisation of New Zealands URM Building Stock Chapter 4

An analysis of the prevalence, distribution, age and nancial value, as well as an estimate of the vulnerability of New Zealands URM building stock was conducted, and the results were presented in Chapter 4. It was determined that at the time of publication there are approximately 3750 URM buildings in existence in New Zealand, with 1300 (36%) being estimated to be potentially earthquake prone and 2010 (52%) estimated to be potentially earthquake risk. The vulnerability of URM buildings was estimated using the
- 253 Alistair P. Russell

Chapter 9. Summary of Conclusions

New Zealand Society for Earthquake Engineering (NZSEE) Initial Evaluation Procedure (IEP). Although legislation in New Zealand eectively denes earthquake prone buildings as those which could be expected to perform at a level less than 33% of a new building, the approach taken by NZSEE is that any building performing at a level less than 67% of the standard of a new building could constitute an earthquake hazard. According to the criteria advocated by NZSEE, up to 88% of New Zealands URM building stock could require seismic improvement.

Trends in the age of New Zealand URM buildings showed that construction activity increased from the early days of European settlement and reached a peak in approximately 1930, before subsequently declining sharply. The preponderance of the existing URM building stock was constructed prior to 1940, and as such, almost all URM structures in New Zealand are between 80 and 130 years old (in 2010).

It was found that the majority of URM buildings in New Zealand are one and two storeys in height, with comparatively few three-and-higher storey buildings in existence. Collectively the value of smaller buildings is lower than the value of larger buildings, but individually the larger buildings have greater nancial value. This nancial analysis suggested that for seismic assessment and retrot, less time and money should be invested in smaller buildings individually, but that collectively this set of building typologies will form the majority of the work which remains to be undertaken in order to make the New Zealand URM building stock more resilient to earthquakes. It was also determined that overall the New Zealand URM building stock has a value of approximately $NZ1.5 billion (in 2010).

9.3

Diagonal Shear Testing of Wall Panels Chapter 5

In Chapter 5 the results of diagonal shear tests on eight two-leaf unreinforced masonry wall panels were presented. The aim of the experimentation was to investigate the diagoAlistair P. Russell

- 254 -

9.4. In-Plane Cyclic Testing of URM Walls Chapters 6 and 7

nal tension (shear) strength of unreinforced masonry with dierent mortar properties and bond patterns, and also to provide a baseline against which to compare other retrotted masonry samples and also samples tested in-situ in existing buildings.

The results of the tests showed signicant variation in the diagonal tension strength of the wall panels, even when the mortar composition and other variables were consistent. Both the shear modulus and Youngs modulus were consistent in all eight samples, with mean values of 0.9 GPa and 2.25 GPa respectively.

It was found that there is a clear correlation between the strength of mortar and the diagonal tension strength of the masonry. When the compression strength of the mortar is higher than the compression strength of the bricks, cracking can be expected to propagate through the bricks, and in contrast, when the compression strength of the mortar is lower than the compression strength of the bricks, cracking can be expected to propagate through the mortar bed and head joints. There was not a clear relationship between the diagonal tension strength and bond pattern. It was concluded that the bond pattern does not have a signicant eect on the diagonal tension strength of URM walls.

9.4

In-Plane Cyclic Testing of URM Walls Chapters 6 and 7

9.4.1

Rectangular Walls

In Chapter 6 results were presented of pseudo-static in-plane testing of three unreinforced masonry walls designed to replicate typical New Zealand construction in the early 20th Century, with dierent aspect ratios and dierent levels of axial load.

It was found that for rectangular walls with an aspect ratio of 1:1, a exural (rocking) response occurred. For the two walls in which rocking was the dominant behaviour mode,
- 255 Alistair P. Russell

Chapter 9. Summary of Conclusions

failure was not observed until a drift of 1.8%. It was previously suggested in the literature that 0.8% is a limiting drift for in-plane walls responding in exure, and it was conrmed in this chapter that this drift limit is suitable.

It was also observed that for a wall with an aspect ratio of 1:2, diagonal tension failure was the limiting behaviour mode, and it was concluded that aspect ratio is more important than the axial load in inuencing the failure mode.

Furthermore, the energy dissipation characteristics of walls responding in-plane were investigated, and it was found that for walls limited by exure dominated actions, an equivalent viscous damping ratio eq of 0.05 is appropriate. For walls limited by diagonal tension cracking followed by bed-joint sliding, an equivalent viscous damping ratio eq of 0.15 is appropriate.

Finally, equations available in the literature for predicting the in-plane response of rectangular URM walls (without anges) were compared with the results of the tests in this chapter. It was found that the behaviour mode and limiting strengths could be satisfactorily predicted by some of those equations. Equations (6.14), (6.15), (6.16) and (6.17) produced favourable agreement between the predicted and measured response and were recommended for assessing the in-plane seismic response of rectangular URM walls.

9.4.2

Flanged Walls

Chapter 7 presented the results of six in-plane tests on URM walls with anges. The six walls had varying ange lengths and plan geometries. It was found that for URM walls with anges, exure is less likely as a behaviour mode and shear is more likely to limit the lateral strength. It was also found that URM walls with anges are able to sustain larger lateral forces than walls without anges.

All the walls reported in this chapter failed by diagonal tension cracking, which is considAlistair P. Russell

- 256 -

9.4. In-Plane Cyclic Testing of URM Walls Chapters 6 and 7

ered a shear dominated response. Walls A6 A8 subsequently commenced sliding on the bed-joints which opened as a result of diagonal tension cracking. When diagonal tension cracking occurs in such a way that sliding occurs subsequently on the cracked bed-joints, this behaviour is considered a deformation-controlled action as there is residual displacement capacity beyond when the peak force is attained. For walls failing in a diagonal tension mode followed by bed-joint sliding, a value of ultimate drift u of 0.6% was recommended.

The results of wall testing reported in this chapter were compared with an analytical model proposed in the literature. The predicted response using this model was veried against the wall test results, with a high level of accuracy obtained, particularly for walls with an aspect ratio of 1:2. Equation (7.3) proposed by Yi et al. (2008) was validated using the experimental results of Walls A5 A8, with a high level of correlation. Consequently Equation (7.3) was recommended for predicting the diagonal tension strength of anged URM walls.

The initial stiness of URM walls failing in diagonal tension followed by bed-joint sliding was satisfactorily modelled using Equation (7.7), with modications made to the shear modulus Gm and eective moment of inertia Ie , to account for cracking and the nonhomogeneous nature of masonry as a material. The post-cracked shear stiness Ge was recommended to be taken as 0.2Gm , and Ie was recommended to be taken as 0.5Ig .

Determining the initial stiness, the strength and the ultimate drift of URM walls responding in-plane in a behaviour mode of diagonal tension cracking followed by bed-joint sliding enabled the general force-displacement response to be modelled.

Furthermore, the energy dissipation characteristics of anged walls responding in-plane were investigated, and as in Chapter 6, it was found that for walls limited by diagonal tension cracking followed by bed-joint sliding, an equivalent viscous damping ratio eq of 0.15 is appropriate.

- 257 -

Alistair P. Russell

Chapter 9. Summary of Conclusions

Finally, it was also concluded that the length of anges at the ends of in-plane URM walls inuences the orientation of cracking. It was observed that for walls with an aspect ratio of 1:2 and with no anges (reported in Chapter 6) the cracking pattern showed cracks propagating at approximately 45 through the centre of the in-plane wall. In contrast, for walls with anges, it was observed that the greater the length of the ange, the steeper (> 45) the orientation of cracking.

9.5

URM Building Assessment Procedure Chapter 8

The nal chapter in this thesis presented a procedure for assessing the performance of unreinforced masonry buildings, within the framework of the existing direct displacement based design method. The use of direct displacement based design principles was chosen as it is considered that this approach is fundamentally more rational and accurate, and takes dierent failure modes into account, compared with exclusively force-based approaches.

The assessment procedure presented in this chapter was designed to be transparent, logical and coherent, particularly with regard to determining the seismic demand on a URM building. The drift limits and equivalent viscous damping ratios as determined in the previous chapters were applied in the procedure. The focus of the method developed in this chapter was on the in-plane response of URM walls, but it was recognised that the performance of URM buildings is also controlled by the out-of-plane response of walls, the response of diaphragms and the capacity of the wall-diaphragm connections. Consequently, steps in the procedure were allocated for these components, but the details for assessing the capacity of these components were omitted, as this research has been conducted and reported elsewhere.

Alistair P. Russell

- 258 -

9.6. Future Research

9.6

Future Research

A number of issues that may require attention in the future have arisen as a result of the research reported in this thesis.

Chapter 4 assessed the population of the New Zealand URM building stock in terms of building numbers. Assessment of the URM building stock in terms of oor area would provide a direct comparison with other construction types, allowing the determination of the proportion of the overall New Zealand building stock that the URM building stock constitutes. Moreover, this proposed assessment would constitute a more direct framework for assessing URM building insurance and building replacement values.

The experimental programme reported in this thesis could be extrapolated to investigate dierent wall congurations, masonry strengths and levels of vertical compressive stress with the use of nonlinear static computational simulations. Furthermore, dynamic response simulations of URM buildings incorporating wall behaviour as observed in the experimentation conducted in this thesis would also inform the expected response of URM walls responding in-plane.

The eective ange widths used in Chapter 7 were assumed to be as given by existing design standards for the design and construction of new reinforced concrete masonry walls. Underestimating the width of the ange may be unconservative in terms of estimating web shear. Contribution to future developments on the eect of anges could be made by addressing conservative upper and lower bound values for eective ange widths.

The interaction of exible timber diaphragms and walls responding in-plane inuences the displacement demands on a URM building. If the rst action of a seismic retrot is to securely connect the masonry walls and diaphragms, the displacement demand on the building will be dierent from the displacement demand on isolated walls. Research into integrated building displacement demands would provide further information on the expected performance of walls responding in-plane.
- 259 Alistair P. Russell

Chapter 9. Summary of Conclusions

Alistair P. Russell

- 260 -

CHAPTER 10

References

Abrams, D. and Shah, N. (1992). Cyclic Load Testing of Unreinforced Masonry Walls. Advanced Construction Technology Center Report No. 92-26-10. College of Engineering, University of Illinois at Urbana-Champaign, 46 pp. Abrams, D. P. (1997). Response of Unreinforced Masonry Buildings. Journal of Earthquake Engineering, 1(1):257273. Abrams, D. P. (2000). Seismic Response Patterns for URM Buildings. TMS Journal, 18(1):7178. Abrams, D. P. and Costley, A. C. (1996). Seismic evaluation of unreinforced masonry buildings. 11th World Conference on Earthquake Engineering, Acapulco, Mexico, June 23 29, 1996. Abrams, D. P. and Costley, A. P. (1994). Dynamic Response Measurements for URM Buildings Systems. In Abrams, D. P. and Calvi, G. M., editors, Proceedings of the U.S.-Italian Workshop on Guidelines for Seismic Evaluation and Rehabilitation of Unreinforced Masonry Buildings. Department of Structural Mechanics, University
- 261 Alistair P. Russell

Chapter 10. References

of Pavia, Italy, June 22 24, 1994, Published by National Center for Earthquake Engineering Research, Bualo, N.Y. Technical report / National Center for Earthquake Engineering Research; NCEER-94-0021. Abrams, D. P., Smith, T., Lynch, J., and Franklin, S. (2007). Eectiveness of Rehabilitation on Seismic Behavior of Masonry Piers. Journal of Structural Engineering, 133(1):3243. Anthoine, A., Magonette, G., and Magenes, G. (1994). Shear-Compression Testing and Analysis of Brick Masonry Walls. 10th European Conference on Earthquake Engineering, Vienna, Austria, August 28 September 2, 1994. Argan, G. C. (1996). On the Typology of Architecture. In Nesbitt, K., editor, Theorizing a New Agenda for Architecture: An Anthology of Architectural Theory 1965 1995, pages 240246. New York: Princeton Architectural Press. ASCE (2007). Seismic Rehabilitation of Existing Buildings SEI/ASCE 41-06. American Society of Civil Engineers, Reston, Va. Aschheim, M. and Black, E. (2000). Yield Point Spectra for Seismic Design and Rehabilitation. Earthquake Spectra, 16(2):317335. ASTM (2002). Standard Test Method for Shear Modulus at Room Temperature. ASTM E 143 02. ASTM International, USA. ASTM (2003). Standard Test Method for Sampling and Testing Brick and Structural Clay Tile. ASTM C 67 03a. ASTM International, USA. ASTM (2004a). Standard Test Method for Compressive Strength of Hydraulic Cement Mortars (Using 2-in. or [50-mm] Cube Specimens). ASTM C 109/C 109M 2. ASTM International, USA. ASTM (2004b). Standard Test Method for Youngs Modulus, Tangent Modulus, and Chord Modulus. ASTM E 111 04. ASTM International, USA.
Alistair P. Russell

- 262 -

ASTM (2007a). Standard Test Method for Compressive Strength of Masonry Prisms. ASTM C 1314 07. ASTM International, USA. ASTM (2007b). Standard Test Method for Diagonal Tension (Shear) in Masonry Assemblages. ASTM E 519 07. ASTM International, USA. ATC (1996). ATC-21: Rapid Visual Screening of Buildings for Potential Seismic Hazards Training Manual. Applied Technology Council, Redwood City, California. Bakhshi, A., Bozorgnia, Y., Ghannad, M. A., Khosravifar, A., Mousavi Eshkiki, S. E., Rahimzadeh Rofooei, F., and Taheri Behbahani, A. (2005). Seismic Vulnerability of Traditional Houses in Iran. 1st International Conference on Seismic Adobe Structures, Lima, Peru. Beattie, G. J., Megget, L. M., and Andrews, A. L. (2008). The Historic Development of Earthquake Engineering in New Zealand. 14th World Conference on Earthquake Engineering, Beijing, China, Oct 12 17, 2008. Benedetti, D. and Toma zevi c, M. (1984). Sulla verica sismica di costruzioni in muratura (on the seismic assessment of masonry structures. Ingegneria Sismica, 1(0):916. [in Italian]. Binda, L. (2005). Methodologies for the vulnerability analysis of historic centres in Italy. Advances in Architecture Series, 20:279290. Binda, L. (2006a). Investigation on the masonry quality and mechanical behaviour (Clinic 1). The First International Conference on Restoration of Heritage Masonry Structures, Cairo, Egypt, April 24 27, 2006. Binda, L. (2006b). The Dicult Choice of Materials Used for the Repair of Brick and Stone Masonry Walls. The First International Conference on Restoration of Heritage Masonry Structures: Clinic 1, Cairo, Egypt, April 24 - 27, 2006. Binda, L., Cardani, G., and Saisi, A. (2005). Application of a multidisciplinary investiga- 263 Alistair P. Russell

Chapter 10. References

tion to study the vulnerability of Castelluccio (Umbria). Advances in Architecture Series, 20:311322. Blaikie, E. L. (1999). Methodology for the Assessment of Face Loaded Unreinforced Masonry Walls under Seismic Loading. Technical report, Opus International Consultants, Wellington, New Zealand. Blaikie, E. L. (2002). Methodology for Assessing the Seimic Performance of Unreinforced Masonry Single Storey Walls, Parapets and Free Standing Walls. Technical report, Opus International Consultants, Wellington, New Zealand. Blaikie, E. L. and Spurr, D. D. (1992). Earthquake Vulnerability of Existing Unreinforced Masonry Buildings. EQC Report, December 1992, Works Consultancy Services, Wellington, New Zealand. Boardman, P. R. (1983). Case Studies: Earthquake Risk Buildings. Restoration of Old Auckland Customhouse. Bulletin of the New Zealand National Society for Earthquake Engineering, 16(1):73 79. Bothara, J., Jury, R. D., Wheeler, K., and Stevens, C. (2008). Seismic Assessment Of Buildings In Wellington: Experiences And Challenges . 14th World Conference on Earthquake Engineering, Beijing, China, Oct 12 17, 2008. Bothara, J. K. (2004). A Shaking Table Investigation on The Seismic Resistance of a Brick Masonry House. Masters thesis, Department of Civil Engineering, University of Canterbury, New Zealand. Bothara, J. K., Dhakal, R. P., and Mander, J. B. (2010). Seismic performance of an unreinforced masonry building: An experimental investigation. Earthquake Engineering and Structural Dynamics, 39(1):4568. Brignola, A., Ferrini, M., Lagomarsino, S., Mangone, F., and Podest` a, S. (2006a). Experimental evaluation of shear deformability and strength parameters of masonry. Ingegneria Sismica, 3:19 31 [In Italian].
Alistair P. Russell

- 264 -

Brignola, A., Frumento, S., Lagomarsino, S., and Podest` a, S. (2009). Identication of Shear Parameters of Masonry Panels Through the In-Situ Diagonal Compression Test. International Journal of Architectural Heritage, 3(1):5273. Brignola, A., Podest` a, S., and Lagomarsino, S. (2006b). Experimental results of shear strength and stiness of existing masonry walls. 5th International Conference on Structural Analysis of Historical Constructions, New Delhi, India, November 6 8, 2006. Bruneau, M. (1994a). Seismic evaluation of unreinforced masonry buildings - a state-ofthe-art report. Canadian Journal of Civil Engineering, 21(3):512539. Bruneau, M. (1994b). State-of-the-art report on seismic performance of unreinforced masonry buildings. Journal of Structural Engineering, 120(1):230251. Calderini, C., Cattari, S., and Lagomarsino, S. (2009a). Identication of Shear Mechanical Parameters of Masonry Piers From Diagonal Compression Test. 11th Canadian Masonry Symposium, Toronto, Ontario, Canada, May 31 June 3, 2009. Calderini, C., Cattari, S., and Lagomarsino, S. (2009b). In-plane strength of unreinforced masonry piers. Earthquake Engineering & Structural Dynamics, 38(2):243267. Calvi, G. M. and Magenes, G. (1994). Experimental Research on Response of URM Building Systems. In Abrams, D. P. and Calvi, G. M., editors, Proceedings of the U.S.-Italian Workshop on Guidelines for Seismic Evaluation and Rehabilitation of Unreinforced Masonry Buildings. Department of Structural Mechanics, University of Pavia, Italy, June 22 24, 1994, Published by National Center for Earthquake Engineering Research, Bualo, N.Y. Technical report / National Center for Earthquake Engineering Research; NCEER-94-0021. Cattari, S. and Lagomarsino, S. (2009). Modelling the Seismic Response of Unreinforced Existing Masonry Buildings: A Critical Review of Some Models Proposed by Codes. 11th Canadian Masonry Symposium, Toronto, Ontario, Canada, May

- 265 -

Alistair P. Russell

Chapter 10. References

31 June 3, 2009. Census and Statistics Oce (1890 1950). The New Zealand Ocial Year Books. New Zealand Government, Wellington, New Zealand. Chiostrini, S., Galano, L., and Vignoli, A. (2000). On the determination of strength of ancient masonry walls via experimental tests. 12th World Conference on Earthquake Engineering, Auckland, New Zealand, January 30 February 4, 2000, paper no. 2564. Chopra, A. and Goel, R. (2001). Direct Displacement-Based Design: Use of Inelastic vs. Elastic Design Spectra. Earthquake Spectra, 17(1):4764. Chopra, A. K. (2007). Dynamics of Structures: Theory and Applications to Earthquake Engineering. Prentice Hall, Upper Saddle River, NJ, 3rd edition. Corradi, M., Borri, A., and Vignoli, A. (2003). Experimental study on the determination of strength of masonry walls. Journal of Construction and Building Material, 17:325337. Corradi, M., Tedeschi, C., Binda, L., and Borri, A. (2008). Experimental evaluation of shear and compression strength of masonry wall before and after reinforcement: Deep repointing. Journal of Construction and Building Material, 22(4):463472. Costley, A. C. (1996). Dynamic Response of Unreinforced Masonry Buildings with Flexible Diaphragms. NCEER-96-0001. University of Bualo, Bualo, NY, (PB97-133573, MF-A03, A15). Cull, J. E. L. (1931). Report of the Building Regulations Committee. Report to the New Zealand House of Representatives, H-21. Davenport, P. (2004). Review of Seismic Provisions of Historic New Zealand Loading Codes. New Zealand Society for Earthquake Engineering Conference, Rotorua, New Zealand, March 19 21, 2004.
Alistair P. Russell

- 266 -

DAyala, D. and Speranza, E. (2003). Denition of Collapse Mechanisms and Seismic Vulnerability of Historic Masonry Buildings. Earthquake Spectra, 19:479509. DBH (2005). Earthquake-Prone Building Provisions of the Building Act 2004: Policy Guidance for Territorial Authorities. Department of Building and Housing Te Tari Kaupapa Whare, New Zealand Government, Wellington, New Zealand. Derakhshan, H. (2011). Out-of-plane Seismic Assessment of Unreinforced Masonry Walls. PhD thesis, Department of Civil and Environmental Engineering, The University of Auckland, New Zealand (in preparation). Derakhshan, H., Dizhur, D., Lumantarna, R., Cuthbert, J., Grith, M. C., and Ingham, J. M. (2010). In-Field Simulated Seismic Testing of As-Built and Retrotted Unreinforced Masonry Partition Walls of The William Weir House in Wellington. Journal of the Structural Engineering Society New Zealand, 23(1):51 61. Dizhur, D., Derakhshan, H., Ingham, J. M., and Grith, M. C. (2009). In-Situ Out-OfPlane Testing of Unreinforced Masonry Partition Walls. 11th Canadian Masonry Symposium, Toronto, Ontario, Canada, May 31 June 3, 2009. Dizhur, D., Derakhshan, H., Lumantarna, R., and Ingham, J. M. (2010a). In-Situ Out-Of-Plane Testing of Unreinforced Masonry Wall Segment in Wintec Block F Building. New Zealand Society for Earthquake Engineering Conference, Te Papa, Wellington, New Zealand, March 26 28, 2010. Dizhur, D. and Ingham, J. M. (2010). Field testing of an earthquake-damaged unreinforced masonry building. 7th International Conference on Urban Earthquake Engineering (7CUEE) & 5th International Conference on Earthquake Engineering (5ICEE), Tokyo Institute of Technology, Tokyo, Japan, March 3 5, 2010. Dizhur, D., Lumantarna, R., Derakhshan, H., Grith, M. C., and Ingham, J. M. (2010b). In-Situ Testing of a Residential Unreinforced Masonry Building Located in New Zealand. 8th International Masonry Conference (8IMC), Dresden, Germany, July

- 267 -

Alistair P. Russell

Chapter 10. References

4 7, 2010. Dowrick, D. (1998). Damage and Intensities in the Magnitude 7.8 1931 Hawkes Bay, New Zealand, Earthquake. Bulletin of the New Zealand Society for Earthquake Engineering, 30(2):133158. Drysdale, R. G., Hamid, A. A., and Baker, L. R. (1999). Masonry Structures: Behavior and Design. The Masonry Society, Boulder, Colo., 2nd edition. Erbay, O. O. (2004). A Methodology to Assess Seismic Risk for Populations of Unreinforced Masonry Buildings. PhD thesis, University of Illinois at Urbana-Champaign. Erbay, O. O. and Abrams, D. P. (2004). A Methodology to Assess Seismic Risk for Populations of Unreinforced Masonry Buildings. 13th World Conference on Earthquake Engineering, Vancouver, British Columbia, Canada, August 1 6, 2004. Erbay, O. O. and Abrams, D. P. (2007). Modeling seismic risk for populations of unreinforced masonry buildings. Tenth North American Masonry Conference, St. Louis, Missouri, USA, June 3 6, 2007. Faccioli, E., Paolucci, R., and Rey, J. (2004). Displacement Spectra for Long Periods. Earthquake Spectra, 20(2):347376. FEMA 273 (1997). NEHRP Guidelines for the Seismic Rehabilitation of Buildings. Federal Emergency Management Agency, Washington, DC. Foss, M. (2001). Diagonal Tension in Unreinforced Masonry Assemblages. MAEC ST11: Large Scale Test of Low Rise Building System, Georgia Institute of Technology. Franklin, S., Lynch, J., and Abrams, D. P. (2001). Performance of Rehabilitated URM Shear Walls: Flexural Behavior of Piers. Mid-America Earthquake Center Report, CD Release 03-03,. Department of Civil Engineering, University of Illinois at Urbana-Champaign.

Alistair P. Russell

- 268 -

Gabor, A., Bennani, A., Jacquelin, E., and Lebon, F. (2006). Modelling approaches of the in-plane shear behaviour of unreinforced and FRP strengthened masonry panels. Composite Structures, 74(3):277288. Gambarotta, L. and Lagomarsino, S. (1997a). Damage models for the seismic response of brick masonry shear walls. Part I: The mortar joint model and its applications. Earthquake Engineering and Structural Dynamics, 26(4):423439. Gambarotta, L. and Lagomarsino, S. (1997b). Damage models for the seismic response of brick masonry shear walls. Part II: The continuum model and its applications. Earthquake Engineering and Structural Dynamics, 26(4):441462. Gere, J. M. and Timoshenko, S. P. (1997). Mechanics of Materials. PWS Pub Co., Boston, 4th edition. Ghannad, M. A., Bakhshi, A., Eshkiki, S. E. M., Khosravifar, A., Bozorgnia, Y., and Behbahnani, A. A. T. (2006). A study on seismic vulnerability of rural houses in Iran. 1st ECEES, Geneva, Switzerland, September 3 8, 2006. GNS Science (2007). The Active Earth. http://www.gns.cri.nz/what/earthact/index.html. Goodwin, C. O. (2008). Architectural Considerations in the Seismic Retrot of Unreinforced Masonry Heritage Buildings in New Zealand. Masters thesis, Department of Architecture and Planning, The University of Auckland, New Zealand. Goodwin, C. O. (2009). Identifying Heritage Value in URM Buildings. Journal of the Structural Engineering Society New Zealand, 22(2):1628, Auckland, New Zealand, September 2009. Grando, S., Valluzzi, M. R., Tumialan, J. G., and Nanni, A. (2003). Shear strengthening of URM clay walls with FRP systems. 6th International Symposium on FibreReinforced Polymer (FRP) Reinforcement for Concrete Structures (FRPRCS-6), Singapore, July 8 10, 2003.

- 269 -

Alistair P. Russell

Chapter 10. References

Gu, X. L., Ouyang, Y., Zhang, W. P., and Ye, F. F. (2003). Seismic behaviour of masonry structural walls strengthened with cfrp plates. In Tan, K. H., editor, FRPRCS-6, pages 12591268, Singapore. Hamid, A. A., El-Dakhakhni, W. W., Hakam, Z. H. R., and Elgaaly, M. (2005). Behavior of Composite Unreinforced MasonryFiber-Reinforced Polymer Wall Assemblages Under In-Plane Loading. Journal of Composites for Construction, 9(1):7383. Harry, A. H. (1988). Masonry: Materials, Design, Construction, and Maintenance. ASTM, USA. Hopkins, D. C. (2009). Earthquakes and Existing Buildings: New Zealand Experience 1968 to 2008. ATC & SEI Conference on Improving the Seismic Performance of Existing Buildings and Other Structures, San Francisco, CA, USA, Dec 9 11, 2009. Hopkins, D. C., Stannard, M., Lawrance, G., and Brewer, I. (2008). Strengthening Buildings for Earthquake: Implementing New Zealand Legislation. 14th World Conference on Earthquake Engineering, Beijing, China, Oct 12 17, 2008. Ingham, J. M. (2008). The Inuence of Earthquakes on New Zealand Masonry Construction Practice. 14th International Brick and Block Masonry Conference (14IBMAC), Sydney, Australia, February 17 20, 2008. Johnson, J. A. R. (1963). Earthquake Engineering in New Zealand, New Zealand Engineering. New Zealand Institute of Engineers, 18(9):305. Karantoni, F. V. and Bouckovalas, G. (1997). Description and analysis of building damage due to Pyrgos, Greece earthquake. Soil Dynamics and Earthquake Engineering, 16:141150. Karimi, K. and Bakhshi, A. (2006). Development of fragility curves for unreinforced masonry buildings before and after upgrading using analytical method. 1st ECEES, Geneva, Switzerland, September 3 8, 2006.
Alistair P. Russell

- 270 -

Kidder, F. E. (1905). The Architects and Builders Pocket-Book: A Handbook for Architects, Structural Engineers, Builders, and Draughtsmen. New York: John Wiley & Sons, 14th edition. Kitching, N. (1999). Small Scale Modelling of Masonry. In Masonry Research, Civil Engineering Division. The Cardi School of Engineering. Klingner, R. E. (2006). Behavior of masonry in the Northridge (US) and Tecom an-Colima (Mexico) earthquakes: Lessons learned, and changes in US design provisions. Construction and Building Materials, 20(4):209219. Knox, C. L. (2012). Seismic Assessment and Retrot of Perforated Unreinforced Masonry Shear Walls. PhD thesis, Department of Civil and Environmental Engineering, The University of Auckland, New Zealand (in preparation). Lavin, S. (1992). Quatrem` ere de Quincy and the Invention of a Modern Language of Architecture. The MIT Press, Cambridge, Massachusetts. Lee, J.-H., Li, C., Oh, S.-H., Yang, W.-J., and Yi, W.-H. (2008). Evaluation of Rocking and Toe Crushing Failure of Unreinforced Masonry Walls. Advances in Structural Engineering, 11(5):475489. Louren co, P. B., Luso, E., and Almeida, M. G. (2006). Defects and moisture problems in buildings from historical city centres: a case study in Portugal. Building and Environment, 41(2):223234. Lumantarna, R. (2012). Studies on The Material Properties of New Zealand Unreinforced Masonry. PhD thesis, Department of Civil and Environmental Engineering, The University of Auckland, New Zealand (in preparation). Magenes, G. (2006). Masonry Building Design in Seismic Areas: Recent Experiences and Prospects From a European Standpoint. 1st ECEES, Geneva, Switzerland, September 3 8, 2006.

- 271 -

Alistair P. Russell

Chapter 10. References

Magenes, G. and Calvi, G. M. (1992). Cyclic behaviour of brick masonry walls. Proceedings of the Tenth World Conference on Earthquake Engineering, Madrid, Spain, July 19 24, 1992. Magenes, G. and Calvi, G. M. (1997). In-plane seismic response of brick masonry walls. Earthquake Engineering & Structural Dynamics, 26(11):10911112. Magenes, G. and Della-Fontana, A. (1998). Simplied Non-linear Seismic Analysis of Masonry Buildings. Fifth International Masonry Conference, London, October 13 15, 1998. Mahmood, H. (2011). Seismic Assessment and Retrot of Unreinforced Masonry Walls with FRP. PhD thesis, Department of Civil and Environmental Engineering, The University of Auckland, New Zealand (in preparation). Mann, W. and M uller, H. (1973). Bruchkriterien f ur querkraftbeanspruchtes Mauerwerk und ihre Anwendung auf gemauerte Windscheiben. Die Bautechnik, pages 421 425. Mann, W. and M uller, H. (1982). Failures of shear-stressed masonry: an enlarged theory, tests and application to shear walls. Proceedings of the British Ceramic Society, 30. Mann, W. and M uller, H. (1985). Schubtragf ahigkeit von gemauerten W anden und Voraussetzungen f ur das Entfallen des Windnachweises. Verlag Ernst & Sohn. Marcari, G., Manfredi, G., and Pecce, M. (2003). Experimental behaviour of masonry panels strengthened with FRP sheets. In Tan, K. H., editor, FRPRCS-6, pages 12091218, Singapore. World Scientic Publishing Company. Marshall, O. S., Sweeney, S. C., and Trovillion, J. C. (2000). Performance testing of berreinforced polymer composite overlays for seismic rehabilitation of unreinforced masonry walls. US Army Corps of Engineers, ERDC/CERL TR-00-18.

Alistair P. Russell

- 272 -

Marzahn, G. (1998). The Shear Strength of Dry-Stacked Masonry Walls. Leipzig Annual Civil Engineering Report, Institut f ur Massivbau und Baustotechnologie, Wirtschaftswissenschaftliche Fakult at, Universit at Leipzig. McClean, R. (2007). Heritage provisions: dangerous, earthquake prone, insanitary buildings and dangerous dams policies. Sustainable Management of Historic Heritage: Guide no. 9. New Zealand Historic Places Trust Pouhere Taonga, Wellington, New Zealand. McClean, R. (2009). Toward improved national and local action on earthquake-prone heritage buildings. Historic Heritage Research Paper No.1. New Zealand Historic Places Trust Pouhere Taonga, Wellington, New Zealand. McKinnon, D. (2008). Shaking up the Past. Heritage Risk: Earthquake, re and codes. New Zealand National Heritage Conference, Wanganui. Deputy Mayors Opening Address. New Zealand Historic Places Trust - Pouhere Taonga, Wanganui, New Zealand. Megget, L. (2006). From Brittle to Ductile: 75 Years of Seismic Design in New Zealand. Bulletin of the New Zealand Society for Earthquake Engineering, 39(3):158169. Moon, F. L. (2004). Seismic Strengthening of Low-Rise Unreinforced Masonry Structures with Flexible Diaphragms. PhD thesis, Georgia Institute of Technology, Atlanta, GA, USA. Moon, F. L., Yi, T., Leon, R. T., and Kahn, L. F. (2006). Recommendations for Seismic Evaluation and Retrot of Low-Rise URM Structures. Journal of Structural Engineering, 132(5):663672. Moon, F. L., Yi, T., Leon, R. T., and Kahn, L. F. (2007). Testing of a Full-Scale Unreinforced Masonry Building Following Seismic Strengthening. Journal of Structural Engineering, 133(9):12151226. Mousavi Eshkiki, S. E., Khosravifar, A., Ghannad, M. A., Bakhshi, A., Taheri Behbahani,
- 273 Alistair P. Russell

Chapter 10. References

A., and Bozorgnia, Y. (2006). Structural Typology of Traditional Houses in Iran Based on their Seismic Behaviour. 8th US National Conference on Earthquake Engineering, San Francisco, California, USA. MSJC (2008). Building Code Requirements for Masonry Structures (TMS 402/ACI

530/ASCE 5). American Concrete Institute; Structural Engineering Institute; The Masonry Society (Masonry Standards Joint Committee), Boulder, CO, USA. Neidhart, H. and Sester, M. (2004). Identifying building types and building clusters using 3D-laser scanning and GIS data. Geo-Imagery Bridging Continents, XXth ISPRS Congress, Istanbul, Turkey, July 12 23, 2004. New Zealand Parliament (1968). Municipal Corporations Act, incorporating Amendment 301A. The Department of Internal Aairs Te Tari Taiwhenua, New Zealand Government, Wellington, New Zealand, Date of assent: 1968. New Zealand Parliament (2004). Building Act 2004. Department of Building and Housing Te Tari Kaupapa Whare, Ministry of Economic Development, New Zealand Government, Wellington, New Zealand, Date of assent: 24 August 2004. New Zealand Parliament (2005). Building (Specied Systems, Change the Use, and

Earthquake-prone Buildings) Regulations 2005. Department of Building and Housing Te Tari Kaupapa Whare, New Zealand Government, Wellington, New Zealand. New Zealand Standards Institute (1935). NZSS No. 95:1935, Model Building By-Law. Sections I to X. New Zealand Standards Institute, Wellington, New Zealand. New Zealand Standards Institute (1965). NZSS 1900:1965, Model Building By-Law. Chapter 8: Basic Design Loads. New Zealand Standrds Institute, Wellington, New Zealand. Nollet, M. J., Chaallal, O., and Lefebvre, K. (2005). Seismic vulnerability study of historical buildings in Old Montreal: Overview and perspectives. Advances in
Alistair P. Russell

- 274 -

Architecture Series, 20:227236. NZNSEE (1972). Classication of High Earthquake Risk Buildings. Bulletin of the New Zealand National Society for Earthquake Engineering, 5(2). NZNSEE (1973). Recommendations for the Classication of High Earthquake Risk Buildings. [Brown Book]. New Zealand National Society for Earthquake Engineering, Wellington, New Zealand. NZNSEE (1985). Recommendations and Guidelines for Classifying, Interim Securing and Strengthening Earthquake Risk Buildings. [1985 Red Book]. New Zealand National Society for Earthquake Engineering. NZNSEE (1995). Draft Guidelines for Assessing and Strengthening Earthquake Risk Buildings. [1995 Red Book]. New Zealand National Society for Earthquake Engineering. NZSEE (2006). Assessment and Improvement of the Structural Performance of Buildings in Earthquakes. Recommendations of a NZSEE Study Group on Earthquake Risk Buildings. New Zealand Society for Earthquake Engineering. OConnor, J. (1919). Courtville Apartments Architectural Specications. Auckland. Oliver, S. (2007). Typical storey heights for URM walls. Personal communication. Page, A. W. (1996). Unreinforced masonry structures an Australian overview. Bulletin of the New Zealand National Society for Earthquake Engineering, 29(4):242255. Pande, G. N., Middleton, J., and Kralj, B., editors (1998). Computer Methods in Structural Masonry - 4. E & FN Spon, London, UK. Paquette, J. and Bruneau, M. (2003). Pseudo-dynamic testing of unreinforced masonry building with exible diaphragm. Journal of Structural Engineering, 129(6):708 716.
- 275 -

Alistair P. Russell

Chapter 10. References

Paquette, J. and Bruneau, M. (2006). Pseudo-dynamic testing of unreinforced masonry building with exible diaphragm and comparison with existing procedures. Construction and Building Materials, 20(4):220228. Park, R. (1989). Evaluation of ductility of structures and structural assesmblages from laboratory testing. Bulletin of the New Zealand Society for Earthquake Engineering, 3(22):155166. Paulay, T. and Priestley, M. J. N. (1992). Seismic Design of Reinforced Concrete and Masonry Buildings. John Wiley & Sons, Inc. Peralta, D. F., Bracci, J. M., and Hueste, M. B. D. (2004). Seismic Behavior of Wood Diaphragms in Pre-1950s Unreinforced Masonry Buildings. Journal of Structural Engineering, 130(12):20402050. Petersen, R. B., Masia, M. J., and Seracino, R. (2008). Experimental verication of nite element model to predit the shear behaviour of NSM FRP strengthened walls. 14th International Brick and Block Masonry Conference (14IBMAC), Sydney, Australia. 17 20 February 2008. Petersen, R. B., Masia, M. J., and Seracino, R. (2009). Experimental and Numerical Investigation into the Shear Behaviour of Unreinforced Masonry Panels Retrotted with Fibre Reinforced Polymer Strips. 11th Canadian Masonry Symposium,

Toronto, Ontario, Canada, May 31 June 3, 2009. Pontius, S. E. (2007). genotype | phenotype: investigations in typology. Masters thesis, The School of Architecture and Interior Design, The University of Cincinnati. Priestley, M. (2002). Direct Displacement-Based Design of Precast/Prestressed Concrete Buildings. PCI Journal, 47(6):6679. Priestley, M., Calvi, G., and Kowalsky, M. (2007). Displacement-Based Seismic Design of Structures. IUSS Press : Fondazione Eucentre, Pavia, Italy.

Alistair P. Russell

- 276 -

Priestley, M. and Kowalsky, M. (2000). Direct displacement-based seismic design of concrete buildings. Bulletin of the New Zealand Society for Earthquake Engineering, 33(4):421 441. Priestley, N. M. (1993). Myths and Fallacies in Earthquake Engineering Conicts Between Design and Reality. Bulletin of the New Zealand National Society for Earthquake Engineering, 26(3):329341. Quatrem` ere de Quincy, A.-C. (1832). Dictionnaire historique dArchitecture. Paris. Published by Charles-Joseph Panckoucke. Robinson, L. and Bowman, I. (2000). Guidelines for Earthquake Strengthening. New Zealand Historic Places Trust - Pouhere Taonga, Wellington, New Zealand, 13 pages. Rossi, A. (1982). The Architecture of the City. Cambridge: The MIT Press. Russell, A. P., Ingham, J., and Grith, M. (2006). Comparing New Zealands Unreinforced Masonry Details with those Encountered in Other Seismically Active Countries. 7th International Masonry Conference (7IMC), London, 30 Oct 1 Nov 2006. Santa Maria, H., Alcaino, P., and Luders, C. (2006). Experimental response of masonry walls externally reinforced with carbon bre fabrics. In Proc., 8th US National Conference on Earthquake Engineering, San Francisco, California. Scott, E. F. (1999). A report on the Relief Organisation in Hastings Arising out of the [Magnitude 7.8] Earthquake in Hawkes Bay [New Zealand] on February 3, 1931. Bulletin of the New Zealand Society for Earthquake Engineering, 32(4):246256. Sezen, H., Elwood, K. J., Whittaker, A. S., Mosalam, K. M., Wallace, J. W., and Stanton, J. F. (2000). Structural Engineering Reconnaissance of the August 17, 1999 Earthquake: Kocaeli (Izmit), Turkey. A Report of the Turkey-US Geotechnical Reconnaissance Team September 3, 1999. Pacic Earthquake Engineering Research
- 277 Alistair P. Russell

Chapter 10. References

Centre, PEER Report 2000/09. Shedid, M. T., El-Dakhakhni, W. W., and Drysdale, R. G. (2008). Seismic Response of Linear, Flanged, and Conned Masonry Shear Walls. 14th World Conference on Earthquake Engineering, Beijing, China, Oct 12 17, 2008. Stacpoole, J. and Beaven, P. (1972). New Zealand Art; Architecture 1820-1970. A. H. & A. W. Reed, Wellington, New Zealand. Standards Association of New Zealand (1976). NZS 4203: Code of Practice for General Structural Design and Design Loadings for Buildings. Standards Association of New Zealand, Wellington, New Zealand. Standards New Zealand (2002a). AS/NZS 1170.0:2002, Structural Design Actions Part 0: General Principles. Standards New Zealand, Wellington, New Zealand. Standards New Zealand (2002b). AS/NZS 1170.1:2002, Structural Design Actions Part 1: Permanent, imposed and other actions. Standards New Zealand, Wellington, New Zealand. Standards New Zealand (2004a). NZS 1170.5:2004, Structural Design Actions Part

5:Earthquake actions New Zealand. Standards New Zealand, Wellington, New Zealand. Standards New Zealand (2004b). NZS 4230:2004, Design of Reinforced Concrete Masonry Structures. Standards New Zealand, Wellington, New Zealand. Steelman, J. and Abrams, D. P. (2007). Eect of Axial Stress and Aspect Ratio on Lateral Strength of URM Shear Walls. Tenth North American Masonry Conference, St. Louis, Missouri, USA, June 3 6, 2007. Stevens, C. M. and Wheeler, K. E. (2008). Implementing earthquake prone building policy under the Building Act 2004 Wellington Citys Approach. New Zealand Society for Earthquake Engineering Conference, Wairakei, Taupo, New Zealand,
Alistair P. Russell

- 278 -

April 11 13, 2008. Stil, J. (2007). Nikau Contractors. Personal Communication. Sullivan, T., Calvi, G., Priestley, M., and Kowalsky, M. (2003). The limitations and performances of dierent displacement based design methods. Journal of Earthquake Engineering, 7(Special Issue. 1):201241. Toma zevi c, M. (1996). Recent Advances in Earthquake Resistant Design of Masonry Buildings. 11th World Conference on Earthquake Engineering, Acapulco, Mexico, June 23 29, 1996. Toma zevi c, M. and Lutman, M. (2007). Heritage Masonry Buildings in Urban Settlements and the Requirements of Eurocodes: Experience of Slovenia. International Journal of Architectural Heritage, 1(1):108130. Turn sek, V. and Sheppard, P. (1980). The shear and exural resistance of masonry walls. Proceedings of the International Research Conference on Earthquake Engineering, Skopje. covi Turn sek, V. and Ca c, F. (1970). Some experimental results on the strength of brick masonry walls. Proceedings of the 2nd International Brick Masonry Conference, Stoke-on-Trent. Valluzzi, M. R., Bernardini, A., and Modena, C. (2005a). Seismic vulnerability assessment and structural improvement proposals for the building typologies of the historic centers of Vittorio Veneto (Italy). Advances in Architecture Series, 20:323 332. Valluzzi, M. R., Cardani, G., Saisi, A., Binda, L., and Modena, C. (2005b). Study of the seismic vulnerability of complex masonry buildings. Advances in Architecture Series, 20:301310. Valluzzi, M. R., Tinazzi, D., and Modena, C. (2002). Shear behavior of masonry panels
- 279 Alistair P. Russell

Chapter 10. References

strengthened by FRP laminates. Journal of Construction and Building Materials, 16:409416. Vestroni, F., Beolchini, G., Grillo, F., Martinelli, A., Ricciardulli, G. L., and Buarini, G. (1995). The design of experimental investigation of an ancient masonry building. 7th Italian Conference of Seismic Engineering, Siena, Italy, September 25 28, 1995 [In Italian]. Vicente, R., Varum, H., and Mendes da Silva, J. A. R. (2006). Vulnerability assessment of traditional buildings in Coimbra, Portugal, supported by a GIS tool. 1st ECEES, Geneva, Switzerland, September 3 8, 2006. Voon, K. C. (2007). In-plane Seismic Design of Concrete Masonry Structures. PhD thesis, Department of Civil and Environmental Engineering, The University of Auckland, New Zealand. Wight, G. D. (2006). Seismic Performance of a Post-tensioned Concrete Masonry Wall System. PhD thesis, Department of Civil and Environmental Engineering, The University of Auckland, New Zealand. Wilson, A. W. (2011). Seismic Performance of Timber Floor Diaphragms in Unreinforced Masonry Buildings. PhD thesis, Department of Civil and Environmental Engineering, The University of Auckland, New Zealand (in preparation). Yi, T. (2004). Experimental Investigation and Numerical Simulation of an Unreinforced Masonry Structure with Flexible Diaphragms. PhD thesis, Georgia Institute of Technology, Atlanta, GA, USA. Yi, T., Moon, F. L., Leon, R. T., and Kahn, L. F. (2005). Eective Pier Model for the Nonlinear In-Plane Analysis of Individual URM Piers. The Masonry Society Journal, 23(1):2135. Yi, T., Moon, F. L., Leon, R. T., and Kahn, L. F. (2006a). Analyses of a Two-Story Unreinforced Masonry Building. Journal of Structural Engineering, 132(5):653
Alistair P. Russell

- 280 -

662. Yi, T., Moon, F. L., Leon, R. T., and Kahn, L. F. (2006b). Lateral Load Tests on a Two-Story Unreinforced Masonry Building. Journal of Structural Engineering, 132(5):643652. Yi, T., Moon, F. L., Leon, R. T., and Kahn, L. F. (2008). Flange Eects on the Nonlinear Behaviour of URM Piers. The Masonry Society Journal, 26(2):3142. Yu, P., Silva, P. F., and Nanni, A. (2007). In-plane response of URM walls strengthened with GFRP grid referenced polyurea. Tenth North American Masonry Conference, St. Louis, Missouri, USA, June 3 6, 2007.

- 281 -

Alistair P. Russell

Chapter 10. References

Alistair P. Russell

- 282 -

APPENDIX A

Photographic Examples of New Zealand URM Building Typologies

- 283 -

Alistair P. Russell

Appendix A. Photographic Examples of New Zealand URM Building Typologies

A.1

Typology A Photos

Further photographic examples of Typology A buildings (one storey isolated) are given in Figure A.1.

(a) Oamaru

(b) Sumner (Christchurch)

(c) Devonport (Auckland)

(d) Seatoun (Wellington)

(e) Parnell (Auckland)

Figure A.1: Typology A buildings single storey isolated

Alistair P. Russell

- 284 -

A.2. Typology B Photos

A.2

Typology B Photos

Further photographic examples of Typology B buildings (one storey row) are given in Figures A.2 and A.3.

(a) Epsom (Auckland)

(b) Epsom (Auckland)

(c) Newmarket (Auckland)

(d) Karori (Wellington)

(e) Newmarket (Auckland)

Figure A.2: Typology B buildings single storey row

- 285 -

Alistair P. Russell

Appendix A. Photographic Examples of New Zealand URM Building Typologies

(a) Ashburton

(b) Ashburton

(c) Papanui (Christchurch)

(d) Paeroa

(e) Temuka

(f) Newtown (Wellington)

Figure A.3: Typology B buildings single storey row

Alistair P. Russell

- 286 -

A.3. Typology C Photos

A.3

Typology C Photos

Further photographic examples of Typology C buildings (two storey isolated) are given in Figures A.4 A.6.

(a) Lyttelton (Christchurch)

(b) Sumner (Christchurch)

(c) Temuka

(d) Petone (Wellington)

Figure A.4: Typology C buildings two storey isolated

- 287 -

Alistair P. Russell

Appendix A. Photographic Examples of New Zealand URM Building Typologies

(a) Oamaru

(b) Oamaru

(c) Epsom (Auckland)

(d) Onehunga (Auckland)

(e) Devonport (Auckland)

(f) Devonport (Auckland)

Figure A.5: Typology C buildings two storey isolated

Alistair P. Russell

- 288 -

A.3. Typology C Photos

(a) Newtown (Wellington)

(b) Timaru

(c) Central Christchurch

(d) Lyttelton (Christchurch)

Figure A.6: Typology C buildings two storey isolated

- 289 -

Alistair P. Russell

Appendix A. Photographic Examples of New Zealand URM Building Typologies

A.4

Typology D Photos

Further photographic examples of Typology D buildings (two storey row) are given in Figures A.7 A.12.

(a) Epsom (Auckland)

(b) Epsom (Auckland)

(c) Epsom (Auckland)

Figure A.7: Typology D buildings two storey isolated

Alistair P. Russell

- 290 -

A.4. Typology D Photos

(a) Central Christchurch

(b) Central Christchurch

(c) Central Christchurch

(d) Central Christchurch

(e) Lyttelton (Christchurch)

(f) Lyttelton (Christchurch)

Figure A.8: Typology D buildings two storey row

- 291 -

Alistair P. Russell

Appendix A. Photographic Examples of New Zealand URM Building Typologies

(a) Oamaru

(b) Oamaru

(c) Oamaru

(d) Oamaru

(e) Oamaru

(f) Dunedin

Figure A.9: Typology D buildings two storey row

Alistair P. Russell

- 292 -

A.4. Typology D Photos

(a) Newmarket (Auckland)

(b) Devonport (Auckland)

(c) Devonport (Auckland)

(d) Devonport (Auckland)

(e) Devonport (Auckland)

(f) Devonport (Auckland)

Figure A.10: Typology D buildings two storey row

- 293 -

Alistair P. Russell

Appendix A. Photographic Examples of New Zealand URM Building Typologies

(a) Petone (Wellington)

(b) Petone (Wellington)

(c) Petone (Wellington)

(d) Newtown (Wellington)

(e) Ashburton

(f) Bulls

Figure A.11: Typology D buildings two storey row

Alistair P. Russell

- 294 -

A.4. Typology D Photos

(a) Timaru

(b) Timaru

(c) Timaru

(d) Timaru

(e) Central Auckland

(f) Thames

Figure A.12: Typology D buildings two storey row

- 295 -

Alistair P. Russell

Appendix A. Photographic Examples of New Zealand URM Building Typologies

A.5

Typology E Photos

Further photographic examples of Typology E buildings (three + storey isolated) are given in Figure A.13.

(a) Timaru

(b) Central Wellington

(c) Dunedin

Figure A.13: Typology E buildings three + storey isolated

Alistair P. Russell

- 296 -

A.6. Typology F Photos

A.6

Typology F Photos

Further photographic examples of Typology F buildings (three + storey row) are given in Figures A.14 A.16.

(a) Central Christchurch

(b) Central Christchurch

(c) Central Christchurch

Figure A.14: Typology F buildings three + storey row

- 297 -

Alistair P. Russell

Appendix A. Photographic Examples of New Zealand URM Building Typologies

(a) Central Wellington

(b) Central Wellington

(c) Central Wellington

(d) Oamaru

(e) Dunedin

(f) Dunedin

Figure A.15: Typology F buildings three + storey row

Alistair P. Russell

- 298 -

A.6. Typology F Photos

(a) Timaru

(b) Timaru

(c) Timaru

(d) Petone (Wellington)

(e) Central Auckland

Figure A.16: Typology F buildings three + storey row

- 299 -

Alistair P. Russell

Appendix A. Photographic Examples of New Zealand URM Building Typologies

A.7

Typology G Photos

Further photographic examples of Typology G buildings (monumental, religious and institutional) are given in Figures A.17 A.18.

(a) Central Wellington

(b) Central Christchurch

(c) Central Christchurch

(d) Central Christchurch

Figure A.17: Typology G buildings monumental, religious and institutional

Alistair P. Russell

- 300 -

A.7. Typology G Photos

(a) Timaru

(b) Central Auckland

(c) Central Auckland

(d) Devonport (Auckland)

(e) Takapuna (Auckland)

Figure A.18: Typology G buildings monumental, religious and institutional

- 301 -

Alistair P. Russell

Appendix A. Photographic Examples of New Zealand URM Building Typologies

Alistair P. Russell

- 302 -

APPENDIX B

Example of Assessment Procedure Typology C

- 303 -

Alistair P. Russell

Appendix B. Example of Assessment Procedure Typology C

B.1

Seismic Assessment of Typology C Structure

Figure B.1: Typology C building for seismic assessment

Step 1

Dene Global Building Parameters

Importance level Building design life Annual probability of exceedance Number of storeys Height of storey 1, h1 Height of storey 2, h2 Length of building (short direction), L2 Length of building (long direction), L1 Total length of walls (short direction)
Alistair P. Russell

2 50 years 1/500 2 6m 6m 8m 12 m 8m
- 304 -

B.1. Seismic Assessment of Typology C Structure

Total length of walls (long direction) Wall thickness (short direction) Wall thickness (long direction) Floor area Roof area

24 m 230 mm 230 mm 96 m2 96 m2

Step 2

Dene Material Properties

m
fm

unit weight of masonry compressive strength of masonry

14 kN/m3 10 MPa 2.3 MPa 25 MPa 2 MPa 0.1 MPa 0.65 0.1 MPa

fmj compressive strength of mortar

fb fbt c

compressive strength of bricks direct tensile strength of bricks cohesion coecient of friction

vme Strength of masonry bed joint

Step 3

Dene Site Characteristics

Location Z Soil type Ch (T ) Ru N (T, D) dc

Tokoroa 0.21 C 0.44 1.0 1.0 207 mm

- 305 -

Alistair P. Russell

Appendix B. Example of Assessment Procedure Typology C

dc =

Tc 2 4 2

Ch (T ) Z R N (T, D) g

(8.8)

Step 4

Collate Axial Loads

Dead roof Dead oor Live roof Live oor

0.35 kPa 0.35 kPa 0 kPa 3.0 kPa

Step 5

Determine wall axial loads

Long direction Tributary area for long walls: 0.5L1 L2 0.75L2 2 32 m2 (see Figure 8.5(a))

Level 2

Groof Gwall Nwall,2 Awall,2 fm,wall,2

11 kN 39 kN 50 kN 2.76 m2 0.020 MPa 11 kN 11 kN 78 kN 0.3 27 kN 116 kN 2.76 m2

- 306 -

Level 1

Groof Goor Gwall E E Qoor Nwall,1 Awall,1

Alistair P. Russell

B.1. Seismic Assessment of Typology C Structure

fm,wall,1

0.040 MPa

Short direction Tributary area for short walls: 0.25L2 2 16 m2 (see Figure 8.5(b))

Level 2

Groof Gwall Nwall,2 Awall,2 fm,wall,2

6 kN 26 kN 32 kN 1.84 m2 0.018 MPa 6 kN 6 kN 52 kN 16 kN 80 kN 1.84 m2 0.044 MPa

Level 1

Groof Goor Gwall Qoor Nwall,1 Awall,1 fm,wall,1

Step 6

Determine Seismic Weight

Wi = Gi + E Qi

Dead loads: Level 2 Gparapet Groof Gwall G2 0 kN 34 kN 386 kN 420 kN

- 307 Alistair P. Russell

Appendix B. Example of Assessment Procedure Typology C

Level 1

Goor Gwall G1

34 kN 840 kN 874 kN

Live loads: Q2 E Q2 Q1 E Q1 Seismic weight: W2 W1 Wt 420 kN 960 kN 1380 kN 0 kN 0 kN 288 kN 86 kN

Step 7

Base Shear Distribution as Equivalent Static Forces

Wi hi Fi = Ft + 0.92V n (Wi hi )
i=1

(8.10)

W2 h2 F2 W1 h1 F1 V2 V1

420 kN 12 m 0.51V 960 kN 6m 0.49V 0.51V V

Alistair P. Russell

- 308 -

B.1. Seismic Assessment of Typology C Structure

Step 8

Determine Base Shear Demand

m e =

Wt g

(8.11)

m e

141,000 kg

Step 8a

Determine Base Shear Demand for Walls Responding Out-of-Plane

If the drift is assumed as 0.4% and the damping is assumed as 0.15, and if he = 9 m (mid-height of the top storey), then,

dT R dc Vbase V2 V1 alevel,2

36 mm 0.64 207 mm 195 kN 99 kN 195 kN 0.11g

The mass of the out-of-plane responding wall in the top storey when the direction of earthquake motion is parallel to the short direction is determined as follows:

lw hw bw m wall mass

12 m 6m 230 mm 14 kN/m3 lw hw bw m /g = 23,600 kg

- 309 Alistair P. Russell

Appendix B. Example of Assessment Procedure Typology C

hw /bw

26

Consequently,

Foutof plane Aoutof plane Foutof plane

25.6 kN 72 m2 0.35 kPa

The mass of the out-of-plane responding wall in the top storey when the direction of earthquake motion is parallel to the long direction is determined as follows:

lw hw bw m wall mass hw /bw

8m 6m 230 mm 14 kN/m3 lw hw bw m /g = 15,700 kg 26

Consequently,

Foutof plane Aoutof plane Foutof plane

17.0 kN 48 m2 0.36 kPa

Details for determining the capacity of walls responding out-of-plane are omitted here. Consequently, Step 10a is also omitted.

Alistair P. Russell

- 310 -

B.1. Seismic Assessment of Typology C Structure

Step 8b Determine Base Shear Demand for Deformation-Controlled In-Plane Walls

u he dT R dc Vbase V2 V1

0.8% 9.6 m 77 mm 1 207 mm 344 kN 175 kN 344 kN

Step 9b Determine Capacity of Deformation-Controlled Walls Responding In-Plane

Consider top storey (i = 2) Long direction: Vr Vs MIN(Vs ,Vr ) Short direction: Vr Vs MIN(Vs ,Vr ) 184 kN 99 kN 99 kN 381 kN 152 kN 152 kN

Consider bottom storey (i = 1) Long direction: Vr Vs MIN(Vs ,Vr ) Short direction:

- 311 Alistair P. Russell

701 kN 234 kN 234 kN

Appendix B. Example of Assessment Procedure Typology C

Vr Vs MIN(Vs ,Vr )

403 kN 181 kN 181 kN

Step 10b Determine %NBS of Deformation-Controlled Walls Responding In-Plane

In the short direction there is 1 line of wall resistance (the open front of the building is not considered to contribute any lateral force resistance, and the capacity is compared with full the demand on the storey,

Consider top storey (i = 2) Vcapacity,short Vdemand,short %NBSshort 99 kN 176 kN 56%

Consider bottom storey (i = 1) Vcapacity,short Vdemand,short %NBSshort 181 kN 344 kN 53%

In the long direction there are 2 lines of wall resistance, and the capacity is compared with half the demand on the storey,

Consider top storey (i = 2) Vcapacity,long Vdemand,long

Alistair P. Russell

152 kN 88 kN
- 312 -

B.1. Seismic Assessment of Typology C Structure

%NBSlong

172%

Consider bottom storey (i = 1) Vcapacity,long Vdemand,long %NBSlong 234 kN 172 kN 136%

Step 8c

Determine Base Shear Demand for Force-Controlled In-Plane Walls

u he dT eq R dc Vbase V2 V1

0.4% 9.6 m 38 mm 0.15 0.64 132 mm 441 kN 224 kN 441 kN

Step 9c

Determine Capacity of Force-Controlled Walls Responding In-Plane

Consider top storey (i = 2) Long direction: Vdt Short direction: Vdt 129 kN 223 kN

- 313 -

Alistair P. Russell

Appendix B. Example of Assessment Procedure Typology C

Consider bottom storey (i = 1) Long direction: Vdt Short direction: Vdt 162 kN 264 kN

Step 10c Determine %NBS of Force-Controlled Walls Responding In-Plane

In the short direction there is 1 line of wall resistance (the open front of the building is not considered to contribute any lateral force resistance, and the capacity is compared with full the demand on the storey,

Consider top storey (i = 2) Vcapacity,short Vdemand,short %NBSshort 129 kN 224 kN 57%

Consider bottom storey (i = 1) Vcapacity,short Vdemand,short %NBSshort 162 kN 442 kN 37%

In the long direction there are 2 lines of wall resistance, and the capacity is compared with half the demand on the storey,

Consider top storey (i = 2)

Alistair P. Russell

- 314 -

B.1. Seismic Assessment of Typology C Structure

Vcapacity,long Vdemand,long %NBSlong

223 kN 112 kN 200%

Consider bottom storey (i = 1) Vcapacity,long Vdemand,long %NBSlong 264 kN 221 kN 120%

Step 8d Determine Base Shear Demand for Diaphragms and Connections

Take the base shear as determined from Step 8b and 8c

Vbase V2 V1

441 kN 224 kN 441 kN

In the long direction there are 2 lines of wall resistance, lw Fdiaphragm 12 2 = 24 m 441/24 = 18 kN/m

In the short direction there are 2 lines of wall resistance, lw Fdiaphragm 8 2 = 16 m 441/16 = 28 kN/m

The capacity of the diaphragm connections is compared against a demand force of 28 kN/m.

Details for determining the capacity of the diaphragms and connections are omitted here. Consequently, Step 10d is also omitted.

- 315 -

Alistair P. Russell

Appendix B. Example of Assessment Procedure Typology C

Step 11

Determine Minimum %NBS of All Components

In this example the minimum %NBS is determined from Steps 10b and 10c only. %NBSshort %NBSlong 37% 120%

B.2

Conclusion

In terms of the in-plane response of URM walls, the capacity of this buildings is greater than 33%NBS, and on that basis is not considered to be potentially earthquake prone.

The critical wall responding in-plane is the rear wall (short direction) of the building and the lateral force capacity is limited by the diagonal tension strength (deformationcontrolled). This component achieved 37%NBS. This means that the building could be considered potentially earthquake risk (33%NBS < performance < 67%NBS), but not potentially earthquake prone. The overall performance of the building would need to be determined by taking into account the performance of the walls responding out-of-plane, the performance of the diaphragms, and the performance of the wall-diaphragm connections.

By rationally determining seismic demand and by accounting for the eects of anges, the seismic performance of this example building (at a particular site in Tokoroa, on soil class C), based on the behaviour of walls responding in-plane, has been determined to be 37%NBS.

Alistair P. Russell

- 316 -