Documente Academic
Documente Profesional
Documente Cultură
By
SHIH-WEI CHUANG
M.E., Cornell University, U.S.A, 1995
Ph.D. Thesis
A thesis submitted to the School of Geosciences, Minerals
and Civil Engineering, University of South Australia in
partial fulfilment of the requirements for the degree of
Doctor of Philosophy
OF
MAY 2004 iA
I •)
LiBRARY
Preliminaries
This thesis presents theoretical, numerical, and experimental work performed by the
author. All references to, and use, works by other researchers are fully acknowledged
throughout the text. The remaining work described herein, to the best of my knowledge
and belief, is original.
The work contained in this thesis has not been previously submitted, either in part or in
whole, for a degree at this or any other university.
Signed:
Date:
U
Preliminaries
ACKNOWLEDGEMENT
The work presented in this thesis would not have been possible without contribution
of many people. I would like to express my appreciation and thanks to my supervisor
Dr. Yan Zhuge for her guidance, patience, and enthusiastic support during the course
of this study. I would also like to express my appreciation to my associate supervisor
Mr. Peter McBean for his invaluable advice and continuous encouragement
throughout this study.
Financial supports for this study were provided under Department for Administrative
and Information Services at South Australia Government and University of South
Australia presidential scholarship. The supports are gratefully acknowledged.
The testing programme was time consuming and presented many technical
difficulties. Special thanks extended to Mr Ray Yerbury, Mr. Colin Sharp and Mr
Gerald Lewis from University of South Australia's team of technical staff. Their
advice and help during the setting up and testing was vital to the success of the
programme.
I would also like to thank my friends and all my colleagues at the School of
Geosciences, Minerals and Civil Engineering, for their friendship and support.
ill
Preliminaries
ABSTRACT
In the last decade, Australian civil engineers have not paid enough attention to
earthquake resistant design of structures. It was commonly believed that the
earthquake risk was insignificant thus earthquake resistant design was not considered
for most buildings. Earthquake resistant design of structures has become an important
issue in Australia following the Newcastle earthquake in 1989.
Masonry is one of the most commonly used materials throughout Australia for the
construction of low rise buildings. Even though the history of past earthquakes has
shown that masonry buildings suffered the maximum damage and accounted for the
maximum loss of life, they continue to be popular. It was evident from the Newcastle
earthquake that most unreinforced masonry structures were seriously damaged.
Therefore it is important to retrofit and strengthen existing masonry structures to resist
the potential earthquake damages.
For the last twenty years, several seismic retrofitting techniques for masonry
structures have been developed and practiced, but rarely validated with experiments
and numerical modelling. Further more, the research has been carried out mainly in
America and Japan where the risk of major earthquake is high. In Australia, although
unreinforced masonry is one of the most popular types of construction, research into
seismic retrofitting of masonry structures is rare.
The purpose of this research is to develop a new, cheaper and high strength seismic
retrofitting technique for masonry structures. An innovative retrofitting technique is
presented for improving the seismic resistance of unreinforced masonry walls using
cable systems. In this thesis, the experimental results from four unreinforced masonry
walls and eight unreinforced masonry walls retrofitted with cable systems are
presented. Seven unreinforced masonry walls retrofitted with FRP were also
conducted in this research for comparison purpose. All walls were tested under
combined constant gravity load and incrementally increasing in-plane lateral
displacement reversals. The results showed that both the strength and ductility of
iv
Preliminaries
tested specimens were significantly enhanced with this new technique. Seismic
retrofitting of unreinforced masonry walls with cables proved to be an effective and
reliable strengthening alternative.
In this thesis, a basic mechanical model has been introduced first using the simple
truss model. Then, two nonlinear finite element models based on this basic
mechanical model have been developed to validate the experimental results. One
model is developed for unreinforced masonry walls retrofitted by cable system and
the other model is developed for unreinforced masonry walls retrofitted by FRP. All
of the models have been found between the analytical and experimental results are
reasonably good agreement. The model takes into account the material nonlinearities
as well as damage due to progressive cracking. Behaviour of the masonry is modelled
using the theory of plasticity and cracking is modelled using smear cracking
approach. The model is generated using ABAQUS Finite Element program. The
validity of the model is established by comparison with the experimental results. It is
shown that both of the numerical models are capable of predicting not only the load
carrying capacity, but also the failure mode and, ductility of the retrofitted masonry
walls.
This thesis reviews the literature on all seismic retrofitting techniques and compares
their advantages and disadvantages to identify the most effective and economic
retrofitting method for unreinforced masonry buildings. It also presents the basic
concepts of seismic retrofitting and summarizes the findings from recent experimental
and analytical research activities on the seismic retrofitting of unreinforced masonry
buildings and provides some retrofit strategies on the most common failure modes. It
is expected that this thesis will provide some guidelines to assist Australian engineers
to retrofit unreinforced masonry buildings.
v
Preliminaries
LIST OF PUBLICATIONS
This thesis is the result of three years and three months of research work since March
2001 at the School of Geosciences, Minerals and Civil Engineering, University of South
Australia, Australia. During this period the following papers and reports have been
published or submitted for publication.
Chuang, S.W. and Zhuge, Y. (2002). " The development of seismic retrofitting
techniques for existing masonry and concrete structures ", Progress Report No.1,
Department for Administrative and information Services Building Management, South
Australia, Australia.
Chuang, S.W. and Zhuge, Y. (2003). "Seismic retrofitting of unreinforced masonry walls
by FRP strips", Proceedings of 2003 Pacific Conference on Earthquake Engineering,
Christchurch, New Zealand.
Chuang, S.W. and Zhuge, Y. (2003). " The development of seismic retrofitting
techniques for existing masonry and concrete structures ", Progress Report No.2,
Department for Administrative and information Services Building Management, South
Australia, Australia.
vi
Preliminaries
Chuang, S.W., Zhuge, Y. and McBean P.C. (2004). "Numerical analysis of Unreinforced
Masonry Walls retrofitted by Cable System ", 13th International Brick/Block Masonry
Conference, Amsterdam, Holland, (accepted for publication).
vii
Preliminaries
TABLE OF CONTENTS
INTRODUCTION
1.1 General 1-1
1.2 Statement of the research problems 1-2
1.3 Research objectives 1-2
1.4 Thesis organization 1-4
viii
Preliminaries
ix
Preliminaries
4. EXPERIMENTAL PROGRRAM
4.1 Introduction 4-1
4.2 Test specimens 4-1
4.3 Wall dimension 4-2
4.4 Footing and top beam 4-3
4.5 Fabrication of wall specimens 4-4
4.5.1 Footing forms 4-4
4.5.2 Reinforcement cages 4-5
4.5.3 Beam construction 4-5
4.5.4 Wall construction 4-6
4.6 Retrofitting details 4-7
4.6.1 The detailed design for the cable system 4-7
4.6.2 The detail design of FRP 4-10
4.7 Material properties 4-11
4.8 Test setup 4-12
4.9 Loading history 4-15
4.10 Instrumentation and data acquisition system 4-16
4.11 Summary 4-20
x
Preliminaries
xi
Preliminaries
xii
Preliminaries
xlii
Preliminaries
REFERENCES
APPENDIX
xiv
Preliminaries
LIST OF FIGURES
xv
Preliminaries
4.2 The details of the reinforcement and dimensions for the footing 4-4
4.3 The forms the reinforced cages 4-5
4.4 A reinforced concrete foundation beam after concrete had been casted 4-6
4.5 Wall construction 4-6
4.6 Wall retrofitted using cable systems 4-7
4.7 The detailed of the cable system design 4-8
4.8 Connection details between cable system and masonry (TOP) 4-9
4.9 Connection details between cable system and masonry (BOTTOM) 4-10
4.10 The test specimen wall fixed with FRP 4-11
4.11 Thetestsetup(1) 4-13
4.12 The test setup (2) 4-13
4.13 Hydraulic jack (1) 4-14
4.14 Hydraulicjack(2) 4-14
4.15 The testjig 4-15
4.16 The loading history 4-16
4.17 Strain gauges 4-17
4.18 The locations of strain gauges for FRP 4-18
4.19 Kyowa displacement transducers 4-18
4.20 The locations of displacement transducers 4-19
5.1 Hysteretic behaviour of Wall-i (Unretrofitted wall) 5-3
5.2 The effective height H0 5-5
xvi
Preliminaries
xvii
Preliminaries
xviii
Preliminaries
xix
Preliminaries
xx
Preliminaries
LIST OF TABLES
xxi
Preliminaries
NOTATION
A b net mortared area of the bed joints above and below the test brick
B1 damping coefficient used to adjust one-second period spectral response for the
effect of viscous damping
B5 coefficient used to adjust short-period spectral response for the effect of viscous
damping
D wall length
Fa factor to adjust spectral acceleration in the short-period range for site class
G shear modulus
xxii
Preliminaries
G1 fracture energy
H thickness of soil
N average SPT blow count in soil within the upper 100 feet of soil
P axial load
xxiii
Preliminaries
QE action due to design earthquake loads calculated using forces and analysis models
T0 period at which the constant acceleration region of the design response spectrum
begins at a value = 0.2 T5
xxiv
Preliminaries
VbJS lateral strength of wall or pier based on bed-joint shear strength, pounds
the maximum horizontal shear that can be resisted by a rocking failure under
static in-plane loading
Y years
oç shear ratio
c initial cohesion
xxv
Preliminaries
h,, height (in feet) above the base to the roof level
a coefficient which takes into account the vertical stress distribution at the
compressed toe
t wall thickness
xxvi
Preliminaries
dilatancy angle
a factor equal to 0.5 for fixed-free cantilever wall, or equal to 1.0 for fixed-fixed
pier
9 rotation angle between the principal stresses and the material axes
a parameter which takes into account the boundary conditions at the typical test
layouts
a knowledge factor
1 n(S natural logarithm of the spectral acceleration parameter ("i" = "s" for short-period
or "i" =1 for 1 second period) at the desired probability of exceedance
xxvii
Preliminaries
natural logarithm of the spectral acceleration parameter ("i" = "s" for short-
period or "1" =1 for 1 second period) at a 1O%/50 year exceedance rate
natural logarithm of the spectral acceleration parameter ("i" = "s" for short-
period or "i" =1 for 1 second period) for the BSE-2 hazard level
xxviii
Chapter 1: Introduction
CHAPTER 1
INTRODUCTION
1.1 General
Australian civil engineers have not paid enough attention to earthquake resistant
design of structures until Newcastle earthquake in 1989. It was commonly believed
that the earthquake risk was insignificant thus earthquake resistant design was not
specially considered for most buildings. Earthquake resistant design of structures has
only become an important issue in Australia following Newcastle earthquake in 1989.
The Newcastle earthquake in 1989 led to the creation of a new set of guidelines for
earthquake resistant design. This new code (AS1 170.4, 1993) has resulted in the need
to systematically retrofit structures that no longer comply with the new guideline.
Un-reinforced masonry (URM) is one of the oldest and most widely used construction
methods in Australia. Inherent advantages of URM include: aesthetics, heat and sound
insulation, fire resistance, economical considerations and sound understanding of its
mechanical properties and contribute to its continuing appeal. Although the history of
past earthquakes has shown that URM buildings suffered both severe damage and
accounted for the maximum loss of life, they continue to be popular. In Australia the
majority of URM buildings were constructed with little or no seismic design
requirements. This has resulted in a large inventory of buildings with a poor capacity
to dissipate energy through inelastic deformation in an earthquake event. Further
more, most of the historical buildings throughout Australia are unreinforced masonry,
highlighting the need to improve their performance by retrofitting and strengthening
to resist potential earthquake damages.
1-1
Chapter 1: Introduction
Masonry buildings consist of several structural element types. However, the one most
likely to be subjected to earthquake damages is the load-bearing wall. These elements
are designed primarily to resist vertical loads. However, they are often subjected to in-
plane or out-of-plane loads resulting from lateral loads such as earthquakes. The in-
plane shear resistance in load-bearing URM walls is provided by the shear bond
strength of the mortar and the friction shear due to the vertical load. The aging and
often deteriorated mortar joints have little shear capacity. Under severe earthquake
loads the shear capacity of the mortar is often exceeded, which can cause the failure
of the wall. Effective retrofitting techniques are needed to strengthen such walls.
From the review of seismic retrofitting techniques for URM, it can be seen that
adhesive techniques can be applied easily to practice. Therefore, the proposed
research will focus on adhesive techniques by using different materials such as steel
rods, Fibre Reinforced Polymer (FRP) and cables to attach on to the masonry walls to
1-2
Chapter 1: Introduction
retrofit or strengthen them. The literature review has also revealed that steel rods and
FRP have been used by other researchers/engineers; however, using cables will be an
innovation of the current project. It is expected that using cables to strengthen the
seismic resistance of URM will become one of the effective retrofitting methods.
Cable consists of a number of wires or strands and has high tensile strength, lightness
and high corrosion resistance. These materials can absorb tensile stress and increase
overall element stiffness, ductility and bearing capacity. Using cable system for
seismic retrofitting applications has other advantages such as architectural versatility,
durability, and no loss of valuable space. Furthermore, it does not add a significant
mass to the existing building and can remain unchanging dynamic properties for
structures. Therefore, Cable would be seen to be particularly suitable for seismic
retrofitting materials.
An innovative retrofitting technique is presented in this thesis using the cable system.
This cable system consists of adding diagonal cables tied to the walls. These cables
are connected to the wall through the steel plates and chemical bolts to transfer the
load to the foundation. This method appeared to be an efficient upgrading technique
worth investigating.
1-3
Chapter 1: Introduction
walls were retrofitted with FRP to compare the results with cable system; All walls
were tested under combined constant gravity load and incrementally increased in-
plane lateral displacement reversals.
The thesis is presented in 10 chapters. The introduction and the objectives of the
proposed research project and the proposed retrofitting method were described in
Chapter 1. An extensive review of the literature on relevant and recent research of
seismic retrofitting techniques for masonry structures can be found in Chapter 2. The
overall behaviour of a retrofitted masonry wall cannot be anticipated without
clarifying the characteristic behaviour of each of its constituent part. Such knowledge
is prerequisite to the determination of the lateral load capacity of retrofitted walls.
Chapter 3 reviews the characteristics, behaviour, and failure mode of unreinforced
masonry walls and cables under reversed cyclic loading for the purpose of designing
retrofitted walls. Chapter 4 describes the design of the test specimens and the
expected behaviour of the retrofitted walls. The details of the test specimens and
testing procedure are also described in Chapter 4. Chapter 5 reports the behaviour of
the test specimens observed during the test and the experimental results. Chapter 5
also compares and discusses the observed behaviour and results between the
retrofitted and unretrofitted walls. An overview of the numerical models, the failure
criterion, the constitutive law and the behaviour after cracking for unreinforced
masonry are given in Chapter 6. A basic mechanical model for the retrofitted wall
with cable system is developed in Chapter 7. The finite element models for the
retrofitted walls by cable system and FRP are proposed in Chapter 8. The validation
of the numerical modelling results and some discussions are also described in Chapter
8. The development of the performance-based design procedure for URM walls
retrofitted with the cable system is presented in Chapter 9. Chapter 10 presents a
summary of research findings and conclusions.
1-4
Chapter 2: Review of the related literature
CHAPTER 2
REVIEW OF THE RELATED LITERATURE
2.1 Introduction
The basic concept of seismic retrofitting is to upgrade the structural strength and improve
the inelastic deformation capacity or ductility of the structure. In the last two decades,
several seismic retrofitting techniques for masonry structures have been developed and
practiced, but rarely validated with experiments and numerical modelling. According to the
literature search, the seismic retrofitting techniques for masonry structures offer the
following advantages (Mayes, 1989, FEMA, 1997, Cesteffi-Guidi, 1983):
All of the techniques could strengthen the structure to some extent. However, one could
not say which one is better than others because different techniques are suitable for
different cases. But it has been found out that the adhesive techniques can be applied easily
in practice. Therefore, this literature review will focus on the adhesive technique and
compare several materials such as FRP, reinforced bars, steel strips and steel plates to
2-1
Chapter 2: Review of the related literature
provide some useful information for practising engineers involved in the design and
assessment of seismic retrofitting of URM buildings.
Older masonry buildings are often unreinforced and seismically vulnerable and should be
retrofitted to provide higher strength and ductility. Common causes of seismic inadequacy
are: insufficient design loads, structural damage sustained in a previous earthquake, flawed
design or construction, and changes in building occupancy (Badoux and Jirsa, 1990). For
seismic retrofitting, one could follow the general procedure to evaluate the building and
identify the reasons for poor seismic capacity of the building. Then based on this
evaluation, a suitable scheme to strengthen the building could be developed. In this
literature review, the basic concept and the general procedures of seismic retrofitting will
be reviewed and discussed first, followed by an extensive review and comparison of
various retrofitting techniques. It is expected that the review and discussion will provide
useful information to practising engineers on how to choose a suitable technique to retrofit
the building effectively.
The basic concept of seismic retrofitting is to upgrade the ultimate strength of overall
building and improve the inelastic deformation capacity (ductility). There are two ways to
improve the seismic capacity. One is to reduce the earthquake force and the other is to
increase the capacity of the existing building to resist the earthquake load. There are also
two ways to upgrade the existing building. One is to change the structural system or to
change the load paths to reduce the load to the specific element and the other one is to
upgrade the individual element strength.
The general procedures for seismic retrofitting can be described as follows (Wyllie, 1983):
2-2
Chapter 2: Review of the related literature
The criteria for seismic strengthening selected must be suitable to accomplish the goal of
the complying with current building codes, or regulations such as AS3826-1998. The
criteria for retrofitting in conjunction with the importance of the structure, seismic activities
and seismic intensities expected at the site. If the goal is to comply with current building
codes or regulations, then those codes or regulations generally become the criteria, unless
the engineer feels that a more severe criterion is appropriate and the owner agrees.
Based on the seismic evaluation, a retrofitting technique could be selected to conduct the
seismic retrofitting analysis, to develop and select the strengthening schemes. The design
of retrofit scheme for a building is generally constrained by the existing structural and
architectural configuration. To design a successful retrofit scheme, the engineer needs to
identify the structural deficiencies of the building and understand the local and global
mechanical characteristics of the building. Among these characteristics, at least the
following factors should be considered: global and local strength and stiffness supplies,
global and local deformation and energy dissipating capacities, existing structural
irregularities and the dynamic properties of the building.
A key point of all seismic retrofitting techniques is the connection between old and new
materials.
Finally, the retrofitted structure is re-evaluated to confirm that the seismic capacity satisfies
the required capacity. Sometimes after local retrofitting, the weight increase and the
2-3
Chapter 2: Review of the related literature
frequency change of the building will affect the dynamic behaviour and also increase the
earthquake force. The foundation design should also be checked carefully to see if it is
good enough for the new structure.
Conventional construction can cause very high floor accelerations in stiff buildings and
large interstory drifts in flexible structures. These two factors cause difficulties in insuring
the safety of the building components and contents (Mayes, 1989) (Figure 2.1). Installing
base isolation into conventional construction can reduce both interstory drift and floor
accelerations and provide a viable economic solution to the difficult problem of reducing
non-structural earthquake damage (Mayes, 1989) (Figure 2.2).
The concept of base isolation relies on decoupling the structure from the damaging
component of the seismic ground motion, which is practically obtained by supporting the
2-4
Chapter 2: Review of the related literature
structure on a highly flexible layer. In this manner, the fundamental period of the
construction is shifted in the high-periods region, where the energy content of the
earthquake is usually lower. The base isolation technique is particularly suitable to low to
mid-rise buildings and applied mainly to strategic buildings where feasibility of the
structure must be guaranteed in emergency situations and the contents represent an
economical or cultural value to be safeguarded. Although the cost of installing a base
isolation system has been reduced recently, for example, the cost of installing a new hybrid
base isolation system is about 3% of the total construction cost (Takenaka, 2001), the
technique is still quite expensive compared to many other retrofitting techniques which will
be discussed later in the thesis.
• Installing Damper
Damping devices can reduce seismic induced vibrations and improve the overall behaviour
of the structure by increasing its internal damping through the energy dissipated by the
inelastic deformation of these special devices. Consequently, the dynamic structural
response is considerably reduced, particularly in the original members of the structure.
There are two classes of damper. One is passive damper such as fluid viscous damper,
friction damper, steel-hysteretic damper and visco-elastic damper and the other one is
active and hybrid damper such as active mass driver (AMD), hybrid mass damper (HMD)
and active variable stiffness (AVS) system.
In passive dampers, the modified properties of the structure (period and/or damping
capacity) do not vary with the seismic ground motion. Figure 2.3 shows a typical example
of friction damper system, which is installed in the building's diagonal bracing system
(Cheung et al, 2000). Friction damper system's function is similar to that of a shock-
absorbing system in a car. In general the passive dampers have proven economical for
retrofits, as an example, the cost comparison for retrofitting a three-story reinforced
concrete structure with unreinforced masonry infill walls (total area was 6,243m2) was
made by Cheung et al. It was found that the total cost was 60% less using passive dampers
than the conventional methods which inclñde the strengthening of existing shear walls,
addition of new shear walls and partial foundation upgrading.
2-5
Chapter 2: Review of the related literature
In active and hybrid damper, the structural characteristics are modified just as a function of
the specific seismic input (Mele and De Luca, 1994). Apparently, it requires outside power
sources and gets less attention from structural engineers who worry about scarce electricity
in the event of an earthquake. Figure 2.4 shows a typical AIVID system.
aI-,d
Sensor
11th
6th
2-6
Chapter 2: Review of the related literature
Removing heavy items can reduce the weight of the whole building. Therefore, it can
reduce the earthquake force (FEMA-273, 1997).
Changing structural system is to upgrade the whole building's strength and ductility to
resist earthquake force. It usually combines to use with upgrading the individual element
strength.
The designer can determine the force path in the retrofitted building and adjust the strength
and stiffness as needed. They can use the following ways to change the structural system to
achieve a higher strength of the building:
Upgrading the element strength is the way to strengthen key individual elements such as
beam, column, and wall element separately. There are several retrofitting techniques to
upgrade unreinforced masonry walls as described as below:
2-7
Chapter 2: Review of the related literature
flexural strength and ductility, and in-plane shear strength. The following jacketing
techniques are the most common ways to retrofit masonry walls.
1 Confining masonry walls with new reinforced concrete columns (Comti Euro-
International Du Beton, 1997)
2 Confining masonry walls with steel sections (Comti Euro-International Du Beton, 1997)
3 Jacketing masonry walls by reinforced concrete (R.C. sheath) (Benjamin, 1995), steel
plate, fibre reinforced polymer (FRP), steel rings, closed stirrups and prestressed
diametral bars (Cestelli-Guidi, 1983)
2-8
Chapter 2: Review of the related literature
2.3.3.2 Additional thin surface treatments (plaster with wire mesh and
cement mortar, e.g. shotcrete, coatings, grouted steel bars)
• Shotcrete overlays
Application of reinforced shotcrete to the surface of a masonry wall is a common method
for enhancing both in-plane and out-of-plane strength of the walls. The shear area of the
wall is increased and the height-to-thickness (hit) ratio is lowered. The thickness of the
shotcrete overlay varies with seismic demand forces, but generally ranges from 70mm to
150mm (Abrams, 1998). Commonly the shotcrete overlay is assumed to resist all of the
lateral force applied to the wall with the brick masonry being neglected all together.
Reinforcement embedded in the shotcrete layer substantially improves both shear and
flexural capacities. However, shotcrete is expensive because considerable materials and
labour are involved with placement (Abrams, 1998).
2.3.3.3 Repointing
2-10
Chapter 2: Review of the related literature
material or by a concrete casting and therefore to restore the physical and visual integrity of
masonry (Cestelli-Guidi, 1983, Tinazzi et a!., 2000). When reinforcement is used, the
technique is referred to as "structural re-pointing". The technique could be used when the
mortar is deeply eroded or has fallen out or small cracks formed in the joints.
University of Missouri-Rolla has conducted a series of research using FRP laminates and
rods for seismic retrofitting of masonry structures. In particular, using FRP Structural re-
pointing technique and combining with FRP laminates on particular surface has proved to
be the most effective retrofitting technique (Tinazzi et al., 2000)(Figure 2.7). This is
achieved by placing deformed FRP rods into the joints, which are bonded to the masonry
wall by a suitable paste. It has been proved from their experimental testing that the method
has shown to dramatically improve the shear and bending moment capacities of the
masonry under static or cyclic in-plane and out-of-plane loads. Also the residual load
bearing capacity remains remarkable even after high damage levels were introduced
(Tinazzi et al., 2000).
.r4
Figure 2-7. Structural re-pointing technique and combining with FRP laminates
(Tinazzi et al., 2000).
2-11
__
The advantages of using steel bars to retrofit URM walls are cost-effective, little disruption
to users of the structure, avoiding any noticeable alteration to the original masonry finishes,
simple and quick installation and no obvious additional weight of the structure. The
technique mainly used to improve both in-plane and out-of-plane flexural load capacity of
the masonry walls.
Several researchers have successfully used reinforcing bars or rods to enhance the strength
and ductility of URM structures (Cesteffi-Guidi, 1983, Plencnik et al., 1984, FEMA, 1992,
Pierepiekarz, 2000). Different techniques have been used among these investigators. For
example, some investigators used the centre coring technique, which involves grouting
reinforcing bars in cores drilled through the wall height in the vertical direction with
adequate anchorage (FEMA, 1992)(Figure 2.8). This technique consists of removing 100
mm diameter vertical cores from the top of the wall into the footing at regular intervals
(about 0.9 m to 1.5 m apart) and placing reinforcing steel and grout in the cored holes. The
reinforcement could be used with and without post-tensioning. This technique provides a
reinforced vertical beam to resist flexural stresses (Plencnik et al., 1984, FEMA, 1992).
The advantages of the technique include the preservation of the aesthetic appearance of a
building and significant construction cost savings (50% compared to the conventional
shotcrete approach) (Pierepiekarz, 2000). However, the method is criticized as well since it
is irreversible.
.. I.-
I
-
—
rekntorcernent
4- rt.
2-12
Chapter 2: Review of the related literature
Grouted steel bars with diagonal, vertical and horizontal configurations have been used by
several investigators to increase tensile strength in these directions (Garrity, 1995,Manzouri
et a!., 1996). A retro-reinforcement technique that install stainless steel reinforcement
horizontally into existing masonry structures to increase the tensile strength and ductility of
the structures was developed jointly by the University of Bradford and Proprietary
Reinforcement Engineering (P.R.E) (Garrity, 1995)(Figure 2.9). The main procedures for
the techniques are: grooves up to 10 mm thick are cut into the mortar bed joints with a
depth of 50 — 70 mm; remove dust; inject a layer of grout; push the steel bar into the layer
of grout and then a second layer of grout is injected over the reinforcing bar. The major
limitation of the technique is that it requires a relatively thick layer of mortar joints.
- reinforced masonry
spandrel beam
reinforced masonry
edge beam
2-13
Chapter 2: Review of the related literature
2.3.3.5 Tying
Figure 2-10. Tension tie for wall to floor retrofit (Mercer, 1998)
Grout injection technique has been used to repair cracked/damaged masonry. It should be
noted that the selection of grout materials is important, because corrosion protection and
compatibility with the existing materials have to be ensured. Injection of an incompatible
grout mix has been known to cause considerable damage (Gigla and Wenzel, 1997).
Epoxy materials have excellent penetration and bond characteristics and are commonly
used for concrete retrofits. However, they are not suitable for masonry partially because
the high strength epoxy materials are not structurally compatible with soft, relatively weak
masonry, and large inclusions of injected epoxy could significantly change stress
2-14
Chapter 2: Review of the related literature
distributions under loads (Manzouri et al., 1996). Cement based grout injection is one of
the most widely used retrofit methods for unreinforced masonry. The grout injection
procedure involves drilling small injection holes (around 10 mm) into mortar joints;
pumping a fluid cement-based grout mixture into cracks, voids, or cavities within masonry.
Low injection pressures should be maintained to prevent damage while strengthening
already fragile materials. The method is most cost-effective for repairing isolated cracks.
When steel bars are used, the technique is referred to as "reinforced injection" and it
provides an effective way to restore composite action and the integrity of previously
damaged masonry walls. Supplementary injection anchors are used as a repair system for
historical masonry for transferring additional tensile and shear forces that cannot be
sustained by the masonry alone (Gigla and Wenzel, 1997). The anchors perform as un-
tensioned steel reinforcement or as prestressed tendons. Common anchor materials are
standard reinforcing steel, threaded rods or special prestres sing bars with roll-formed
sections. After clearing and pre-wetting the drill hole, the anchor element is inserted,
centred with spacers and grouted (Figure 2.11). Although reinforced injection is an
effective technique to increase the strength and ductility of masonry walls, it requires a
similar level of effort and cost as traditional techniques such as reinforced shotcrete or
composite fibre overlays. The main advantage of the technique is that they are least
disruptive to the building's appearance and occupants (Manzouri et al., 1996).
——I
anchor length
2-15
Chapter 2: Review of the related literature
Door and window opening in URM walls may be enlarged to alter the aspect ratio of an
adjacent pier. By removing a portion of masonry above or below an opening, the height to
length aspect ratio of the adjacent piers will be increased to such an extent that rocking
behaviour may govern their response. Although this approach will weaken a perforated
masonry wall, it will also increase its inelastic deformation capacity if a ductile rocking
mechanism can be invoked. Furthermore, if the method is used, excessive diagonal tensile
stresses can be relieved for a relatively stocky pier, thus lowering its vulnerability to non-
ductile "X" cracking (FEMA, 1997).
More recently, steel strips and new technologies in composites such as FRP laminates have
been adapted to attach to the surface of masonry walls to strengthen them. The techniques
will enhance strength and ductility of unreinforced masonry walls.
Fibre reinforced polymers (FRP) consist of high resistance fibre impregnated with
polymeric resins and have high tensile strength, lightness and corrosion insensitivity. FRP
can be adhered to the surface of a masonry wall with epoxy to provide strength across
potentially weak planes. Because of the high tensile strength of the material, a fairly thin
sheet can provide substantial strength to the masonry (Figure 2.12) (Korany et al., 2001).
2-16
Chapter 2: Review of the related literature
There are three common types of FRP being used, which are Glass, Aramid, and Carbon
Fibre Reinforced Polymers (GFRP, AFRP and CFRP, respectively) and FRP are
commercialised in different shapes: rods, tendons, laminates and three-dimensional
components. All of them have been successfully used to enhance the strength and ductility
for seismic retrofitting of unreinforced masonry structures (Ehsani et al., 1999, Kolsch,
1998, Tumialan et a!., 2001).
Using FRP in seismic retrofitting applications has advantages like rapid application,
durability, low cost of installation, no loss of valuable space and no additional weight and
therefore can remain unchanging dynamic properties for the structures. However, it is
found out from the research that the inelastic deformation capacity of the wall may be
limited as a result of de-bonding of the FRP sheets which occurs suddenly with little
ductility (Abrams, 1998, Ehsani et a!., 1999). Therefore, FRP materials should be placed
strategically on those piers that are likely to attract damage first (high demand to capacity
ratio) (Abrams, 1998). The aesthetic appearance of the building may be a concern as well
as the natural appearance of the masonry wall is replaced with the FRP sheets. In general,
due to high cost and no complete knowledge of their mechanical behaviour, FRP has not
become popular in Australia yet for seismic retrofitting of masonry walls.
Figure 2-12 Retrofitting masonry walls with FRP (Korany et al., 2001)
2-17
Chapter 2: Review of the related literature
More recently, an entirely new method of retrofitting concrete using Sprayed FRP (SFRP)
coatings have been developed (Banthia and Boyd, 2000). The investigators argued that
FRPs with continuous fibres are highly anisotropic. Properties in the direction of fibre
alignment are generally much superior to those in the direction perpendicular to fibre
alignment. In order to overcome such disadvantage, a new technique which consists of
spraying polymer and short, randomly distributed fibres concurrently on the surface of
concrete to be retrofitted using a spray gun is developed (Banthia and Boyd, 2000)(Figure
2.13). It was found from the testing results that sprayed composites with randomly
distributed short fibres performed equally well better than wraps with continuous fibres.
However, the method has not been applied to masonry structures yet.
A steel strip system for seismic retrofitting of low-rise concrete block masonry and
reinforced concrete walls has been developed by Taghdi et a! (2000). This steel strip
system consists of diagonal and vertical strips that were attached using through-thickness
bolts. Stiff steel angles and anchor bolts were used to connect the steel strips to the
2-18
Chapter 2: Review of the related literature
foundation and the top-loading beam (Taghdi et al., 2000) (Figure 2.14). The width of each
diagonal steel strip was 220 mm with a thickness of 3.81 mm. The steel strip width was
chosen to ensure yielding of the gross section in tension prior to net section fracture at the
bolt locations. Two 80 x 3.81 mm vertical steel strips were added as boundary elements on
each side of the walls. Through-thickness 9.5 mm and 15.9 mm diameter structural steel
bolts were used to fasten the vertical and diagonal steel strips to the wall. The experimental
testing proved the technique was effective in significantly increasing the in-plane strength,
ductility and energy dissipation capacity of low-rise unreinforced walls. The testing results
indicated that the steel strips contributed to the energy dissipation for about 50% of the
total. However, the method may not be suitable for brick masonry. Brick can be difficult
to drill and anchor into. Most brick is hard and brittle. Old, red clay brick is often very soft
and is easily over-drilled. Both of these situations can cause problems in drilling and
anchoring.
2-19
Chapter 2: Review of the related literature
investigate the possibility of using the SMAs for the realisation of mechanical and seismic
protection systems for cultural heritage structures based on devices having intrinsically
energy dissipation capabilities (Renda et al., 2000). A series of experiments on URM
shear walls retrofitted with SMAs device under the in-plane cyclic loading have been
conducted (Renda et al., 2000) (Figure 2.15). It was found that the pre-stressing capability
and the energy dissipation of SMAs give an effective improvement of earthquake
resistance of the URM structures. The results showed that the SMAs devices contributed to
the energy dissipation for about 30% of the total. However, this new technique has a
similar problem as for steel strip retrofit; the fragility of the masonry does not allow a
simple fixing of the SMA devices.
Figure 2-15 Full-scale model and SMAs based protection system (Renda et al., 2000)
2-20
Chapter 2: Review of the related literature
The use of adhesive techniques has proved to be one of the most exciting and effective
technologies for external strengthening of masonry structures. Many researchers have used
different materials for adhesive technique to strengthen masonry structures and the results
seem quite successful to improve the strength and ductility of the structures. But which
one is the best? Apparently there is no simple answer for the question. FRP has the
advantages of the ease of application and lightweight; better durability than steel (as they
are not subject to corrosion). Therefore these make FRP as one of the most suitable
materials for seismic retrofitting of URM. The reason why FRP is not too popular yet is
partially due to their high cost. The initial cost of FRP materials is about 10 times more
than steel (ASBI, 2000), however the maintenance cost could be reduced as the steel
requires protection from corrosion which increases its total cost. Another reason for FRP
being less popular is that many practising engineers have not obtained enough knowledge
on FRP; especially their long-term behaviour requires to be better understood. At the
present, FRP is perfectly suitable for heritage structures because the money is not an issue
for the case (Tumialan et a!., 2001).
When the applied force at the top of the wall reaches a certain amount, the overturning
moment will equal the resisting moment. Even the slightest additional applied force will
cause the wall to turn over. Since we cannot allow walls to overturn, we must install
anchors to hold them down to the foundation. Hold-down anchors are the most common
way to attach a wall to its foundation to resist uplift and overturning, but the concrete
foundation must be strong enough to hold the anchors. If the concrete is weak or
deteriorating and drilling holes for anchors is likely to cause cracks or crumbling, the
foundation should be replaced.
Hold-down anchors can take many forms. Some look like straps that emerge from the
foundation and nail to edge or face of the wall (Figure2.16). Creating a connection between
the foundation and shear wall need not involve bolts. This connector relies on an embedded
strap similar to the wall anchors you would see used in concrete tilt-up construction. Nails
2-21
Chapter 2: Review of the related literature
create the connection between the strap and wall. The strap may also deteriorate due to its
reduced cover by concert.
Others connect foundation bolts or threaded rods to the walls via bolts or nails (Figure
2.17). These ase called hold-down anchor bolts. The depth that the bolt is embedded in the
concrete is important. The hold-down anchors create a path for forces to travel out of a
shear wall and into the foundation.
2-22
Chapter 2: Review of the related literature
Studs
to framing
size as required!
Sill Plate
Bolt to foundation
Concrete foundation
Hold-down anchor bolts are manufactured in a variety of types and sizes. These types of
anchors can be divided into two generic categories: expansion-type anchors and adhesive
or chemical-type anchors. (BIA, 1986)
1. Expansion type anchors: Two different types of expansion anchors are most
commonly used in seismic retrofit applications: the wedge anchor and the sleeve
anchor (see Figure 2.18). These anchors develop their strength by means of
expansion into the base material. Wedge anchors develop their hold by means of a
wedge or wedges that are forced into the base material when the bolt is tightened.
The wedges create large point bearing stresses within the hole; therefore, this
anchor requires a solid base material to develop its full capacity.
2-23
Chapter 2: Review of the related literature
Sleeve anchors develop their strength by the expansion of a cylindrical metal sleeve
or shield into the base material as the bolt is tightened. The expansion of the sleeve
along the length of the anchor provides a larger bearing surface than the wedge
anchor, and is less affected by irregularities and voids in the base material than is
the wedge anchor.
//
A) AJN NOR B) ANeJOR
Drop-in and self-drilling anchors (see Figure 2.19) are two other types of expansion
anchors available, but are typically not recommended for use in seismic retrofit
applications. The reason for this is due to the embedment and setting characteristics
of the two anchors. Both anchors are produced to allow shallow embedment depths
and are expanded or set by an impact setting tool. The combination of shallow
embedment and high stresses imparted by the expansion tend to cause cracking or
splitting in concrete. Depending on the extent of cracking or splitting, the anchor
could experience a reduction in load-carrying capacity or undergo complete failure
during installation.
2-24
Chapter 2: Review of the related literature
There ase several considerations that should be examined when contemplating the
use of expansion-type anchors. These are: 1) Expansion anchors should not be used
to resist vibratory loads. Vibratory loads tend to loosen expansion anchors. 2)
Specific torques are required to set expansion anchors. Excessive torque can reduce
anchor strength or may lead to failure as excessive torque is applied. 3) Expansion
anchors require solid, hard embedment material to develop their maximum
capacities.
2. Chemical (adhesive) anchors are a threaded rod that is set in epoxy. For most
retrofit work, should use chemical anchors. For hold-down, wedge anchors do not
generally perform as well as adhesive anchors during the cyclic uplift loading
caused by earthquakes. Therefore, chemical anchors were adopted in this study.
Two basic types of adhesive anchors are currently available. The major difference
between the two is that one anchor is manufactured as a pre-mixed, self-contained
system, whereas the second type requires measurement and mixing of the epoxy
materials at the time of installation. The more popular self-contained types use a
double glass vial system (see Figure 2.20) to contain the epoxy. The outer vial
contains a resin and the inner vial contains a hardener and aggregate. The glass vial
is placed in a re-drilled hole and a threaded rod or bar is driven into the hole with a
2-25
Chapter 2: Review of the related literature
rotary hammer drill, breaking the vials and mixing the adhesive components. The other
type of adhesive anchor requires that the epoxy components be hand-measured and
mixed before the epoxy is placed into a pre-drilled hole. A threaded rod or bar is then
set into the epoxy mixture, as shown in Figure 2.21.
A) EPOXY CAPSULE
B) THREADED ROD
C) INStÔ1LIJED ANCHOR
MIX
PLACE
2-26
Chapter 2: Review of the related literature
There are special requirements and limitations that should be considered when
contemplating the use of adhesive anchors. They are: 1) Specially designed mixing and/or
setting equipment may be required. 2) Dust and debris must be removed from the pre-
drilled holes to insure proper bond between the adhesive and base material. 3) The adhesive
mixture tends to fill small voids and irregularities in the base material. 4) Large voids may
cause reductions in anchor capacities. This is especially true with the self-contained
adhesive anchors since a limited volume of epoxy is available to fill the voids and provide
a bond to the anchor. 5) The adhesive bond strength is reduced at elevated temperatures
and may also be adversely affected by some chemicals.
2.4 Conclusions
Several seismic retrofitting techniques for URM structures have been reviewed and
discussed in this chapter. In order to retrofit existing buildings, it is important to understand
the seismic performance of the building and therefore to identify the reasons for poor
seismic capacity of the building. Then based on this evaluation, a suitable retrofitting
technique/scheme could be developed. Seismic retrofitting of masonry buildings is
relatively new for many structural engineers. A good retrofitting solution requires
consideration of technical, economic and social aspects. Each retrofitting technique has its
own advantages and disadvantages, when a technique is appropriate for one building, it
may not necessarily be appropriate for another. The selected method must be consistent
with aesthetics, function and the strength, ductility and stiffness, and the cost requirements.
According to the literature review, the following comments could be made: different
seismic retrofitting techniques maybe applied to different cases. For the adhesive
technique, FRP seems to provide the best solution. However, due to the high cost of FRP,
it has not become a popular retrofitting method yet. Research is still required to better
defme non-linear hysteretic behaviour of plain and retrofitted unreinforced masonry walls
so that design parameters (strength and deformation capacity) can be quantified for various
rehabilitation methods (Abrams, 1998).
2-27
Chapter 3: Unreinforced masonry walls and cable system
CHAPTER 3
3.1 General
The overall behaviour of a retrofitted unreinforced masonry wall cannot be evaluated without
understanding the characteristic behaviour of each constituent part. This chapter reviews the
characteristics, behaviour and failure modes of unreinforced masonry wall and cables, under
reversed cyclic lateral loading.
Typical Unreinforced masonry (URM) buildings are likely to be composed of several load-
bearing masonry walls arranged in orthogonal planes, with relatively flexible floor
diaphragms.
Two types of failure are commonly observed in load bearing URM wall subjected to seismic
load. These are in plane-failure characterized by a diagonal tensile crack pattern and the out -
of -plane failure, where cracks are primarily along the mortar bed joint; some inclined cracks
may also be developed. The exact crack pattern will depend on the wall's boundary
condition.
These two types of failure mode are shown in Figures 3.1 and 3.2.
3-1
Chapter 3: Unreinforced masonry walls and cable system
———7
--I /
/ S.
/
/
-p
—i-
/
/
/ /
The reasons for causing the out-of-plane failure of URM elements are excessive deflections
of diaphragms and insufficient connections between them such as parapet walls, veneers, and
unanchored load bearing walls during earthquake events. However, out-of-plane failure could
be prevented easily by proper measures (e.g. reinforced concrete ring beams, steel ties at the
floor levels). To provide the stability for in-plane walls to avoid collapse becomes a very
important issue. Therefore, this thesis will focus on the in-plane failure of URM walls.
A low-rise wall is a wall with aspect ratios (i.e., with a ratio of height to horizontal length) of
less than two, when subjected to large shear force reversals the wall may exhibit non-ductile
behaviour. The flexural capacity of these walls cannot be attained and diagonal tension,
compression or sliding-shear failures develop prematurely. These modes of shear failure may
be brittle. Therefore, inelastic deformability and hence dissipation of earthquake induced
seismic energy is reduced. Wall with aspect ratios of less than 1.0 will induce shear failure
and greater than 1.0 will induce flexural failure. Thus, the aspect ratio of a shear wall is a
major parameter that dictates its behaviour (Taghdi, 1998).
3-2
Chapter 3: Unreinforced masonry walls and cable system
Reversed cyclic loading applied to low-rise shear walls produces crossing of diagonal shear
cracks created by lateral load reversals. These intersecting diagonal cracks reduce the
capacity of diagonal compressive struts that form a major part of the mechanism of shear
resistance. Moreover cyclic loading may cause shear walls to slide along their construction
joints. This sliding does not occur in monotonically loaded walls. Therefore, while
monotonic force-deformation relationships may be taken as representative of the envelope of
hysteresis loops in flexure dominant members, the same may not be true at later stages of
loading for low-rise walls where cyclic loading effects often result in strength degradation
(Doostdar, 1994). Up to certain values, the axial compression will decrease the shear capacity
because the high axial compression will crush the bricks and mortars (Zhuge, 1996).
Generally, axial compression tends to increase the strength of walls. Also, in presence of
axial compressive stress, the inclination of diagonal shear cracks tends to be greater than 45
degree (as would be predicted by Mohr circle). Higher axial compression results in increased
shear capacity, as well as increased flexural capacity except for compression-controlled
failures (Drysdale et a!, 1994).
A construction joint is the weakest point in a low-rise shear wall subjected to inelastic load
reversals. Due to the interaction of flexure and shear, sliding along the construction joint may
take place before the wall reaches its capacity. This sliding, accentuating the problem,
destroys the interface along the construction joint. This will severely lower the strength of the
wall and result in a poor energy dissipation mechanism. Walls with low aspect ratio are more
likely to develop sliding shear along their construction joints (Taghdi, 1998).
The capability of unreinforced masonry walls to resist lateral loads is limited by the strength
of both masonry units and bed joint. As mentioned in Section 3.3.3, axial loads affect the
behaviour of low-rise walls. At lower axial loads, two failure modes are possible for the wall
with aspect ratios of less than 1.0: (1) when earthquake forces act in the plane of the walls,
3-3
Chapter 3: Unreinforced masonry walls and cable system
sliding may take place along bed joints or (2) a horizontal crack may develop along the base
and the wall may rock, i.e. overturn around its toe. Localized cracks will appear at the
comprised corner due to the overturning moment. These cracks increase until complete toe
crushing occurs. This toe crushing does not affect the overall stability of the building when
deformations are small, which is usually the case for non-slender walls. The rocking failure is
the typical ultimate behaviour of a wall when subjected to lateral in-plane force combined
with low vertical loads (Taghdi, 1998).
Above a certain level of axial load, lateral sliding is suppressed and diagonal cracks form
within the masonry bricks due to an increase in bed joint friction. Under seismic loading,
cracks will occur in directions perpendicular to principal tensile stress. These in-plane shear
cracks are commonly expressed by double-diagonal (X) shear cracking. Such unreinforced
masonry walls are generally incapable of withstanding severe repeated load reversals,
suffering from low energy-dissipation capacity and severe strength degradation
characteristics (Taghdi, 1998).
The failure modes of URM walls under in-plane lateral loads can be summarized into two
groups: flexure bending and shear failure (Zhuge, 1996).
The flexure bending failure is caused because horizontal load or displacement demand
increase, bed joints crack in tension and shear is carried by the compressed masonry; final
failure is obtained by overturning of the wall and simultaneous crushing of the compressed
corner (Magenes and Calvi, 1997).
The shear failure is caused because peak resistance is governed by the fonnation and
development of inclined diagonal cracks, which may follow the path of bed joints and head
joints or may go through the bricks, depending on the relative strength of mortar joints,
brick-mortar interface, and bricks. At low levels of vertical load andlor low friction
coefficients, potential sliding planes can form along the cracked bed joints due to the
formation of tensile horizontal crack in the bed joints (Magenes and Calvi, 1997).
3-4
Chapter 3: Unreinforced masonry walls and cable system
.1
!'
!1
I 1T/
'I
Figure 3-3 The failure modes of URM under in-plane lateral loads
(Zhuge, 1996)
It should be noted that not all the above failure modes will involve collapse of the wall panels
and final failure may be the combination of several failure modes.
Figure 3.4a shows the deformation vs. force curve of an axially loaded cable. As shown, the
cable buckles elastically when subjected to compressive strains, while it can develop its
yielding strength under tension. Figure 3.4b illustrates the behaviour of the same cable
subjected to several cycles of deformation. As shown, all inelastic tension elongation
accumulates every time the cable yields in tension. Figure 3.5a shows the deformation vs.
3-5
Chapter 3: Unreinforced masonry walls and cable system
force curve of an axially loaded prestressed cable. Figures 3.4a and 3.5a are similar, except
that there is an initial state of stress and strain in the prestressed cable, which is accounted for
in Figure 3.5a by shifting the origin of the force vs. displacement axes. Because of this shift,
the cable can now "resist" compressive force (reduction of its initial tension can be
interpreted as the cable developing a compressive force) (Teran-Gilmore et a!. 1995).
Figure 3-4 Axial displacement vs. axial load behaviour of cable with no prestress (Teran-
Gilmore et al. 1996)
F F
C '9
I
/
&
ooy 8
3-6
Chapter 3: Unreinforced masonry walls and cable system
initial I
prestr& /
Figure 3-5 Axial displacement vs. axial load behaviour of cable with prestress (Teran-
Gilmore et al. 1996)
The yielding of a prestressed cable implies a reduction in its initial prestress and thus a
reduction in its capacity to "resist" compression. Assume the cable yields following the path
OABC in Figure 3.5b: the cable remains elastic in OA; once it reached point A, it yields and
follows AB; finally, the cable unloads up to its original prestressed deformation and reaches
point C. Figure 3.5c, which shows the behaviour of the prestressed cable beyond point C,
illustrates schematically the loss of prestress in the cable by comparing the initial prestress in
the cable with the remaining prestress after the cable yields and unloads to its initial position
(i.e., to equal to 0). As shown, there is an important reduction in the cable's capacity to
"resist" compression due to this loss of prestress.
3-7
Chapter 3: Unreinforced masonry walls and cable system
The use of high levels of prestress, so that the cables may moderately yield in tension at
relatively small drifts, has been discussed by Pincheira and Jirsa (1992). Under these
circumstances, the cables dissipate some energy through their non-linear behaviour.
Pincheira and Jirsa (1992) noted that letting cables yield is more effective than keeping them
elastic provided they do not become slack in the process. Although this is an attractive
option, it should be considered carefully since cables with high levels of prestress are likely
to yield and lose some prestress during small and moderate seismic events that usually occur
prior to the safety level ground motion.
Ductility is defined as the ratio of a specific deformation at any instant to that at the
beginning of yield. Ductility in structural walls can be exhibited only if the constituent
material itself is ductile (Magenes et al. 1997).
Materials with good post-yield behaviour are usually ductile, and have a relatively large
hysteretic ioop indicating good energy absorption per cycle such as cable. Materials
exhibiting brittle behaviour, such as unreinforced masonry, have poor energy absorption
characterised by small pinched and degrading hysteresis loops and are therefore non-ductile.
Although masonry can carry considerable compressive stresses, their performance in tension
is weak. But the cable is a good ductile material with good tensile strength. Therefore,
detailing of this kind of composite structure becomes extremely important to achieve an
overall ductile behaviour.
Yielding of cables of laterally loaded cantilever wall is a major source of energy dissipation.
A retrofitted unreinforced wall will increase energy dissipation due to distributing the
cracking to entire wall if the following failure modes (diagonal tension or diagonal
compression caused by shear, sliding shear along construction joints) are prevented.
Although cables are good in durability, the issue wasn't a concern of this thesis and will not
be discussed in details.
3-8
Chapter 3: Unreinforced masonry walls and cable system
In order to preclude the above types of failures that are typically observed in URM walls and
to make the walls behave in a ductile manner, diagonal cables will be used to cover the areas
where principal stresses develop. The diagonal cables will extend between the corners of the
wall. These cables will cover the diagonal tension zone. It will strengthen diagonal tension
resistance and will prevent diagonal tension failure. Finally, the wall will be secured against
sliding using very stiff angles and high strength masonry anchors.
3-9
Chapter 4: Experimental program
CHAPTER 4
EXPERIMENTAL PROGRAM
4.1 Introduction
The purpose of this research is to develop a new high strength seismic retrofitting
technique for masonry structures. An innovative retrofitting technique with cable
system is presented in this thesis. In this chapter, the set-up of the experimental
program will be discussed.
To simulate practical situations, solid clay bricks (medium sandstock, size of brick:
230x1 lOx76mm) which are commonly used as load bearing walls in Australia were
chosen from commercially available clay bricks and payers to build the test wall
panels. The mortar used had a mix design of 1:1:6 (cement: lime sand: sand, by
volume) ratio and a thickness of 10mm. To maintain consistency, all the wall panels
were built by the same bricklayer and were cured in air for 28 days, before testing.
Each test wall panel was built on a reinforced concrete beam which was bolted on to
the reaction floor. The beams acted as a foundation for the walls.
The type of FRP used for this experiment was Sika CarboDur S (width 30mm,
thickness 1.2mm and tensile strength 2800 MPa) and epoxy adhesive for bonding
Sika CarboDur Strips was Sikadur-30. The FRP was fixed for one side of the wall
only.
4-1
Chapter 4: Experimental program
The type of cable used for this experiment was Rostan typical grade 316 stainless
steel wire rope (19 single strands, diameter 10mm, and breaking load 71 KN) and
ending for cable was seafast threaded swage terminals (RF1513M1O1O). The
anchorage for the connection plate to the foundation beam was Ramset Chemset
Anchor (M l6xl9Omm, design tensile and shear load 8.5KN per anchor and Chemset
800 series). The cable was fixed on one side of the wall only.
Walls with aspect ratios of greater than 1.0 have shear strength higher than their
flexural strength. These walls should fail in shear in a non-ductile manner. The
proposed retrofitting system aims to improve the performance of walls by increasing
shear strength above their flexural strength and by increasing ductility and energy
dissipation capability, therefore all specimens were chosen with an aspect ratio of 1.0
to ensure that most of the unretrofitted walls would exhibit shear-induced damage.
Therefore, the in-plane failure would be dominated. The dimension of the wall is
940mm long x 940mm high x 110 wide (11 courses high and 4 bricks in each course)
(see Figure 4.1). The wall was constructed on top of a concrete foundation beam to
simulate a house footing.
4-2
Chapter 4: Experimental program
II II II
II II
II II II
940mm [4 courses)
Figure 1 Dimension
A stiff reinforced concrete footing was built to provide full fixity at the base. To
achieve full foundation fixity, they were over-designed, both in terms of strength and
stiffness. The specifications for the reinforcement and dimensions are shown in Figure
4.2. Each footing was provided with two holes. These holes were located to match
with the holes in the strong floor. High-strength bolts were used to fix the foundation
to the strong floor. Also, a stiff steel top beam was placed at the top of each wall
specimen. Its function was to transfer vertical compressive stress uniformly to the
wall.
4-3
Chapter 4: Experimental pro gram
Y2Oiiini bars
Figure 4-2 The details of the reinforcement and dimensions for the footing
The footing forms and the reinforced cage for footing were built first.
The forms consisted of a wooden base with steel sides connected by means of wing
nuts. This allowed for easy removal of the sides to access the lifting points cast into
the beam. The ends of the forms were wood and attached to the steel sides with wing
nuts. The forms had significant reinforcement along the length of the sides to
counteract the bowing of the sides that would occur due to the force of the concrete.
The forms were completely oiled before each pour to ensure the forms would separate
easily from the concrete.
4-4
Chapter 4: Experimental program
Each cage consisted of four 20 mm diameter steel reinforced bars, two top and two
bottom. Shear ligatures were placed at locations of the greatest shear, namely the ends
of the beams and the points where the beam was to be fixed to the floor. The beams
were fixed to the floor by two 50mm nuts cast into the beam. The vertical
reinforcement bars that would form the reinforcement for the wall were welded to one
of the bottom bars and then bent to ensure that the bar would exit the top of the beam
in the middle. Figure 4.2 contains beam design diagrams.
The forms were oiled and the reinforced cages were placed in the middle, as shown in
Figure 4.3. A concrete mix was purchased from a local supplier and the concrete
strength specified was 30 MPa. A poker vibrator was used to ensure that the concrete
was adequately compacted. Figure 4.4 shows a reinforced concrete foundation beam
after concrete had been casted. Both slump and compressive tests were performed on
the concrete to ensure that it met the specifications required.
4-5
Chapter 4: Experimental pro gram
Figure 4-4 a reinforced concrete foundation beam after concrete had been casted
Once the foundation beams had been given 28 days to cure a professional bricklayer
was employed to construct the walls. Twelve walls were built as shown in Figure 4.5.
To simulate a realistic situation, solid clay bricks such as those that are commonly
used to build load-bearing walls were chosen.
4-6
Chapter 4: Experimental program
The detailed design of two retrofitting methods, namely cable and FRP will be
discussed in this section.
Retrofit was accomplished by adding two 10mm diameter cables (wire ropes) on only
one side of the wall face, as shown in Figure 4.6. Ideally cables should be added on
both sides of the wall to prevent an eccentric stiffness and strength distribution that
may cause twisting of the retrofitted walls and provide simultaneous retrofit against
out-of-plane failures (although this was not tested in this investigation). The details of
the cable system design are shown in Figure 4.7. The cable diameter was chosen to
ensure that the wall would fail earlier than the cable. The anchor was designed for
transferring the load from cable to foundation and the cable should fail first.
Connection details between cable system and masonry are shown in Figure 4.8 and
Figure 4.9.
4-7
Chapter 4: Experimental program
Steel Plates
Steel block
See detaile
nei
From both
4-8
Chapter 4: Experimental program
lay
Figure 4-8 Connection details between cable system and masonry (TOP)
4-9
Chapter 4: Experimental pro gram
Figure 4-9 Connection details between cable system and masonry (BOTTOM)
Retrofit was accomplished by adding two (width 30mm, thickness 1.2mm and tensile
strength 2800 MPa) FRP strips on only one side of the wall face, as shown in Figure
4.10. FRP should be added on both sides of the wall to prevent an eccentric stiffness
and strength distribution that may cause twisting of the retrofitted walls, and provide
simultaneous retrofit against out-of-plane failures (although this was not tested in this
investigation). Epoxy adhesive, Sikadur-30, was used to attach FRP Strips on the
masonry wall.
4-10
Chapter 4: Experimental program
In order to ensure that the results could be as free from potential errors as possible all
materials used were tested to check if they possessed the properties as required. These
materials included the clay bricks, the concrete and the mortar.
Each of the three materials was tested for their compressive strengths. A standard
concrete compressive strength test was performed after the samples cured for 28 days.
An average compressive strength of 30.8 MPa was obtained from the testing.
A standard mortar compressive strength testing was also performed after the samples
were collected and cured for 28 days. The compressive strength of these samples
returned an average of 7.6 MPa.
Finally, a standard brick compressive strength testing was performed. The bricks
failed at an average of 20.3 MPa.
4-11
Chapter 4: Experimental pro gram
All the details of these testing results and material properties can be found in
Appendix A. These material properties are used in the ABAQUS numerical models
in this thesis.
The testing equipment was designed and set up to house the wall. The testing
equipment was a quite complex. It consisted of a jig to house the wall and apply a
roof load, initially a hydraulic jack to apply the displacement, in planeload cell• to
measure the side load when displacement was applied and a series of transducer and
strain gauge to measure wall movement and micro strain in the FRP.
The Nuts had been cast into the concrete foundation to allow the wall to be bolted to
the strong floor effective fixing it in place. A series of chocks were set up between the
jig and the wall. Once bolted in, it will further ensure that there were no sideway
movements of the wall when the load was applied. A beam was set up at the top
course of brick to ensure that the wall did not move out of plane. The equivalent of an
axial pressure of O.21MPa was applied as appoint load onto a ball bearing car, onto a
beam that sat on the top of the wall, allowing the wall to move in plane as the load
was applied but maintaining the roof load. Figure 4.11 and Figure 4.12 show the test
set up.
One hydraulic jack was used to apply lateral loads to the specimen and another one
used to apply a vertical load. A steel reaction frame supported the jacks. The vertical
load was applied uniformly to the top of the wall through a stiff spreader beam. The
walls were braced by using two lateral steel beams to prevent the out-of-plane
deflection.
4-12
Chapter 4: Experimental program
4-13
Chapter 4: Experimental pro gram
The load was applied via a hydraulic jack (figure 4. l3and figure 4.14) that was
support by the test jig (figure4. 15) and was attach to the wall so that the wall could be
moved at accurate increment of displacement in both directions.
4-14
Chapter 4: Experimental program
In this testing program, axial loads were applied to all specimens in addition to the
lateral loads, to create a more realistic loading condition. As discussed in Chapter
3.3.3, the behaviour of the masonry walls are considerably affected by the presence of
gravity loads. For example, the rocking capacity of URM walls depends directly on
the magnitude of the applied axial load. If bearing loads are not applied on such a
masonry wall, its lateral capacity is theoretically zero (if the self weigh of the wall is
neglected). Identical axial loads were applied to all specimens of this testing program.
This simulated identical tributary floor areas, and made it convenient to compare the
results between specimens. A realistic axial load of 0.2 MPa was applied to the wall
to simulate an applied roof and second story onto the wall.
4-15
Chapter 4: Experimental program
Load history
4
E
E
0)
E
0)
0
(0
.ao
(I)
a 1 0
(0
-2
N
x0
-4 .
-6
-8
Cycle No.
The specimens were well instrumented for displacement, rotation and strain
measurements. Strain gauge as shown in figure 4.17 was applied to the FRP and
cables to measure the amount of micro strain that the FRP material was experiencing
during the cycle of displacement. A computer records all the deflection of the wall
due to cycle load until failure. This micro strain data could be use to determine the
amount of force that was being taken by the FRP and cables for each configuration.
Figure 4.18 illustrates the locations of strain gauges for FRP.
4-16
Chapter 4: Experimental program
4-17
Chapter 4: Experimental pro gram
II II
LI
Strain gauge
I
I
II
I II
IL
II II
II II II
4-18
_________________ ________
II
1.1 L..2
ii
I II II IP
I II II II I
I II II II I
I I
I II I I
II II
II II
4-19
Chapter 4: Experimental program
4.11 Summary
During the testing, each wall panel was subjected to a constant vertical load. The
magnitude of vertical compressive stress for each wall, together with the geometry of
the all specimens, and retrofitting methods were used are indicated in Table 4.1.
Retrofitting
Specimens Size (m) Methods Roof load
Wall 1 0.94x0.94 URM 2OKN
4-20
Chapter 5: Observed behaviour and test results
CHAPTER 5
OBSERVED BEHAVIOUR AND TEST RESULTS
5.1 General
In this chapter, results and observed behaviour of tested walls are presented. Cracking loads,
ultimate loads, crack patterns, and failure modes are identified. Interpretation and discussion of
the recorded data are also presented in this chapter.
In this chapter, discussions on the general behaviour of retrofitted and unretrofitted wall
specimens, including the comparisons of predicted and experimentally measured behaviour, are
presented. The Comparisons between the retrofitted and unretrofitted wall specimens in terms
of ultimate capacity, load-deformation characteristic, stiffness, shape of hysteretic curve,
energy dissipation and ductility are also covered. The summaries of test results are presented in
Table 5.1.
5-1
Chapter 5: Observed behaviour and test results
Retrofitting
Specimens Methods Cracking Load Ultimate Load
Wall 1 URM 14.7KN 23.8KN
Wall 2 CABLE 17.1KN 39.13KN
Wall 3 CABLE 13.7KN 52.35KN
Wall 4 CABLE 14.9KN 47.72KN
Wall 5 FRP 25.5KN 56.79 KN
Wall 6 FRP 20.9 KN 60.39KN
Wall 7 FRP 24.72 KN 63.51 KN
Wall 8 CABLE N/A N/A
Wall 9 FRP N/A N/A
Wall 10 URM N/A N/A
Wall 11 URM N/A 9.2KN
Wall 12 4 FRP strips N/A 28.6KN
Wall 13 FRP N/A 40.5KN
Wall 14 Cable 10.9KN 67.7KN
WaIl 15 Cable 11.2KN 62.2KN
Wall 16 FRP N/A N/A
Wall 17 Cable 18.4KN 39.7KN
Wall 18 Cable 8.7KN 46.5KN
Wall 19 URM 10.5KN 21.2KN
Note: The reason why some of the loads are not available is attached in Appendix D.
The un-retrofitted walls failed in a combination of the sliding developed in one direction,
and the rigid-body rocking (with some small amount of sliding) developed in the other
direction, at an ultimate force of 23.8KN, as evidenced by the unsymmetric hysteresis
loop shown Figure 5.1.
5-2
Chapter 5: Observed behaviour and test results
URM
z
Ct
0
U)
Displacement (mm)
As it can be seen from the hysteretic relationship of URM wall in Figure 5.1, the
hysteretic energy dissipated in a rocking mode response is generally smaller and the
hysteretic energy dissipated in a sliding mode response is higher.
The wall exhibited relatively large deformations with minor strength increasing before
failure. Rocking and sliding could only develop as a consequence of cracking along the
bed joint. Despite the low strength of this wall, which indicates a certain strength
deficiency, its sliding friction and rocking behaviour noticeably dissipates energy.
In hindsight, even though calculations prior to testing predicted that rocking would
develop at a lateral load of lOkN, assuming cracking along the base of the wall, cracking
above the base joint should not be surprising; the base joint was fully mortared in
compliance with construction requirements and was thus stronger, forcing the crack to
occur in the weaker joints above. However, the reason cracking occurred above the third
course of masonry rather than the first is due to the metal bracket to prevent the wall from
moving.
5-3
Chapter 5: Observed behaviour and test results
The lateral load necessary to produce rocking at the crack level can be calculated using
the diagram of Figure 5.3 and the rocking equations suggested by Magenes et. al. (1998).
The maximum horizontal shear which can be resisted by a rocking failure under static in-
plane loading may be approximated introducing a proper stress distribution for the
masonry in compression and neglecting the tensile strength of bed joints. With reference
to Figure 5.3, equilibrium leads to the following expression:
ratiooç:
MH ço'H
vVD D D
Considering typical test layout, the parameter assumes a value of 1 when the wall is
fixed on one end and free to rotate on the other.
P 20000
p=—=
Dt 940x110
=0.2MPa
—
M — H0 — ço' H — lx(94O — 2x76 —10) —
—0 8
D— 940 940
5-4
Chapter 5: Observed behaviour and test results
r
H02L
= 940x110 — 0.2
)= 12.8kN (Experimental result 18.15KN)
0.8 2 0.85x16.5)
ii
Lii J
I II
I
I II II
I II
I I
5-5
Chapter 5: Observed behaviour and test results
v
—
KX>
D/2 D/2
Figure 5-3 Assumptions for rocking strength evaluation of a wall falling with crushing at
the base corner (Magenes et. al. 1998)
The strength of a wall undergoing sliding, under seismic excitation, along a horizontal
joint is sometimes expressed as:
=,iP
where J.1. represents the sliding coefficient of friction of the masonry joint, and cohesion is
neglected invoking the fact the joint is already cracked in tension due to flexure.
These calculated forces are less than the experimentally obtained values for lateral
loading in the sliding and rocking directions. However, the reasons for the higher
experimental values maybe because of the different levels of axial load and different
boundary conditions, different aspect ratios of the walls for the experiment.
5-6
Chapter 5: Observed behaviour and test results
In general, all retrofitted wall specimens exhibited superior behaviour when compared
with those unretrofitted wall specimens. The effective of the cable system and FRP
retrofit solutions is demonstrated in this section.
The retrofitted walls eventually tended to slide or rock in all cases as the lateral load
gradually increased. However, this sliding or rocking would only appear after cracking of
the masonry, at or near the base, parallel to the direction of applied load. After a
horizontal crack had formed above the second row of blocks in Wall 1, the unretrofitted
wall would have rotated as rigid bodies. However, presence of the cable system or FRP
prevented development of this rigid body rotation due to spreading cracks more evenly
over the entire wall.
The testing results of three out of eight walls are discussed in this section. The load-
deflection envelopes for three walls are compared with the un-retrofitted wall in Figure
5.4 and the comparison between the ultimate lateral load resistance of the unretrofitted
wall and retrofitted walls is presented in Table 5.2. It can be seen from the envelope and
the table, the dramatic increase of the ultimate strength for walls retrofitted with cable is
evident. The improvement of the ultimate lateral load resistance of the retrofitted walls
with cable system is about twice the capacity of unretrofitted wall.
5-7
Chapter 5: Observed behaviour and test results
Table 5-2 The summary of test results for the retrofitted walls by cable system
The
Retrofitting The Cracking ultimate
Specimens Methods Load Load Failure mode
60
50
z
0
a)
20
10
0
0 5 10 15 20
Displacement (mm)
5-8
Chapter 5: Observed behaviour and test results
Variation in stiffness with lateral drift is plotted in Figure 5.5 for wall 4 and wall 1. The
stiffness at every drift level was calculated using the two-peak value of each cycle. Only
the first cycle of each drift was considered. The stiffness of each cycle was computed
using the peak-to-peak method. The curves indicated the sensitivity of the stiffness of the
specimen with respect to the top horizontal displacement. As can be seen from the figure
presented, stiffness of retrofitted walls improved compared to the unretrofitted wall.
U)
a)
15
10
5
0- I I I I I
0 2 4 6 8 10 12 14 16
Displacement (mm)
Figure 5-5 Variation in stiffness with lateral drift for wall 4 and wall 1
The testing results of three out of seven walls retrofitted with FRP are discussed in this
section. The summaries of test results are presented in Table 5.3. The load-deflection
envelopes for three walls are compared with the un-retrofitted wall in Figure 5.6. It can
be seen from the envelope, the dramatic increase of the ultimate strength for walls
retrofitted with FRP is evident. The improvement of the ultimate lateral load resistance of
•
the retrofitted walls with two FRP strips strengthening is respectively three times the
capacity of un-reinforced wall.
5-9
Chapter 5: Observed behaviour and test results
Table 5-3 The summary of test results for the retrofitted walls by FRP
The The
Retrofitting Cracking ultimate
Specimens Methods Load Load Failure mode
From Figure 5.6, it shows that URM wall is strongly nonlinear at low level of load due to
the low tensile strength of bed and head joints. It also shows that URM wall decreased
both strength and stiffness as the damage due to cracking increases. However, the URM
walls retrofitted by the FRP system slowed down the cracking propagation and increased
both strength and ductility.
5-10
Chapter 5: Observed behaviour and test results
Wall (FRP)
70
60
z
V 40
0
-J 30
ci)
V
20
10
0
0 10 20 30 40
Displacement (mm)
Figure 5-6 The load-deflection envelops of four walls
The amount of hysteretic energy per cycle dissipated by a specimen at a specific drift
level is equal to the area enclosed by a hysteresis ioop of the lateral load versus top
displacement curve. The hysteretic energy dissipated during the first cycle of each drift
level is as a function of lateral drift, for all walls. This makes it possible to compare the
relative energy dissipation capability of the retrofitted and unretrofitted walls.
As it can be seen from the hysteretic relationship of URM wall in Figure 5.1, the
hysteretic energy dissipated in a rocking mode response is generally smaller and the
hysteretic energy dissipated in a sliding mode response is higher. The hysteretic
relationships of retrofitted walls are shown in Figures 5.7, 5.8, and 5.9, respectively. It
can be seen from these figures, the hysteretic energy dissipations are much higher than
URM
5-11
Chapter 5: Observed behaviour and test results
wall because the effectiveness of cable system in distributing the cracks over the entire
wall increased the whole wall energy dissipation capacity.
Wall 2(CABLE)
H.
Displacem ent(m m)
Figure 5-7 Hysteretic behaviour of Wall-2 (wall retrofitted with cable)
wall 3(CABLE)
/
E -1 () J
1'
I,,,
0)
Displacement(m m)
5-12
Chapter 5: Observed behaviour and test results
WaIl 4 (CABLE)
z
0 I I
10 15 20 25—C
Cl)
Displacement (mm)
The above figures indicate that the retrofitted URM walls exhibit good strength, ductility,
stiffness, and dissipation of energy compared with the unretrofitted wall.
The hysteretic relationships of wall specimens retrofitted with FRP are shown in Figures
5.10, 5.11, and 5.12, respectively. They indicate that the retrofitted URM walls exhibit
approximately symmetrical stable hysteretic behaviour with significant increase in
ductility, stiffness, and dissipation of energy.
5-13
Chapter 5: Observed behaviour and test results
Wall 5 (FRP)
0
-J
ci)
C,)
Displacement (mm)
Wall 6 (FRP)
z
-a
c'j
0
-J
a)
D
C/)
Displacement (mm)
5-14
Chapter 5: Observed behaviour and test results
WaIl 7 (FRP)
D
0
-J
ci)
-o
Cl)
Displacement (mm)
All wall specimens retrofitted using the proposed FRP system exhibited superior
behaviour when compared with the un-retrofitted wall specimen. The above figures
indicate that the retrofitted URM walls exhibit good strength, ductility, stiffness, and
dissipation of energy compared with the un-retrofitted wall. From the hysteretic
relationship of URM wall, it shows that the wall behaved in a combination of rocking and
sliding as evidenced by the unsymmetrical hysteresis loops. It also shows that the flexural
response was occurred because very large displacements were obtained without
significant loss in strength. From the hysteretic relationship of retrofitted walls, they
show that the walls behaved in diagonal shear cracking response as evidenced by the loss
in strength at post-peak response of hysteresis ioops.
These figures also indicate the hysteretic energy dissipations are much higher than URM
wall because the effectiveness of FRP system in distributing the cracks over the entire
wall increased the whole wall energy dissipation.
5-15
Chapter 5: Observed behaviour and test results
The forces carried by cable and the forces carried by the whole retrofitted wall are shown
in Figures 5.13, 5.14, and 5.15, respectively. They indicate that the cable carried about
the 50% of the force acted on the whole wall after major cracks occurred. In these figures,
they also show that the cables start to carry a small amount of the force at the very
beginning stage. The results proved that that URM wall retrofitted by cable system can
effectively increase the load capacity after the onset cracking of the wall.
45
40
35
z 30
a)
25
0 5 10 15 20
Displacement (mm)
Figure 5-13 Comparison of forces carried by the cable and the whole wall (Wall 2)
5-16
Chapter 5: Observed behaviour and test results
60
50
40
z
0 30
-J
U)
U)
20
10
0
0 2 4 6 8 10 12 14 16 18 20
Displacement (mm)
Figure 5-14 Comparison of forces carried by the cable and the whole wall (Wall 3)
Wall 4-cable
50
40
z
30
-j0
2 10 12 14 16
Displacement (mm)
Figure 5-15 Comparison of forces carried by the cable and the whole wall (Wall 4)
5-17
Chapter 5: Observed behaviour and test results
The forces carried by FRP and the forces carried by the whole retrofitted wall are shown
in Figures 5.16, 5.17 and 5.18, respectively. They indicate that the FRPcarried about the
50% of force acted on the whole wall even after early loading stage, which is different
from the cable system. This is because FRP is fully attached to the wall. The results
proved that the URM wall retrofitted by the FRP system becomes a whole new structure
to carry the force together.
Wall 5 (FRP)
z
V
0 — Force on whole wall
-J — Force on FRP
w
V
Cl)
Displacement (mm)
Figure 5-16 Comparison of forces carried by the cable and the whole wall (Wall 5)
5-18
Chapter 5: Observed behaviour and test results
Wall 6 (FRP)
z
— Force on whole wall
-j0 —Force on FRP
0
(I)
0 2 4 5 8 10 12 14 15 18 20
Displacement (mm)
Figure 5-17 Comparison of forces carried by the cable and the whole wall (Wall 6)
Wall 7 (FRP)
-5 0 5 10 15 20 25 30 35
Displacement (mm)
Figure 5-18 Comparison of forces carried by the cable and the whole wall (Wall 7)
5-19
Chapter 5: Observed behaviour and test results
For the ultimate failure patterns of all tested walls, refer to Appendix B. The following
series of photograph shows typical examples of the failure modes of the test specimens.
Figure 5.19 shows the initial cracking occurred at a displacement of 0.5mm along the
mortar joint near the steel plate restraint.
• .
•- ...., —
5-20
Chapter 5: Observed behaviour and test results
The brick cracks occur at the displacement of 5mm as shown in Figure 5.20 and Figure
5.21 during the test.
5-21
Chapter 5: Observed behaviour and test results
Figure 5.22 shows the final failure pattern of URM wall. It was due to a combination of
rocking above the steel plate restraint and then sliding along the mortar joint that were
above the lateral restraint on the left and the right hand side of the wall.
5-22
Chapter 5: Observed behaviour and test results
Figure 5.23 shows the initial cracking occurred at a displacement of 1.0mm along the
mortar joint near the steel plate restraint.
The crack of bricks occurs at the displacement of 14mm as shown in Figure 5.24.
5-23
Chapter 5: Observed behaviour and test results
The in-plane diagonal shear cracks occur at the displacement of 14mm during the test as
shown in Figure 5.25.
Figures 5.26 and 5.27 show the final failure pattern of wall retrofitted with cable. The
wall behaved in a combination of rocking, sliding and diagonal cracking. Generally, the
presence of the cable system prevented development of the rigid body rotation observed
in the unretrofitted wall and allowed cracks to spread more eventually over the entire wall.
The retrofitted wall obviously increased the load canying capacity and changed the load
paths and cracking patterns. In the retrofitted walls, they show the effectiveness of cable
system in distributing the cracks over the entire wall as opposed to cracks concentrating
only at certain locations for URM walls.
5-24
Chapter 5: Observed behaviour and test results
—
-V
- i -
--
:1
I..
-4
4
*;
-J L'
I q —
—--— -
* -j
5-25
Chapter 5: Observed behaviour and test results
Figure 5.28 shows the initial cracking occurred at a displacement of 0.8mm along the
mortar joint near the steel plate restraint.
..-/i
U
L
0 0
I.
The crack of bricks occurs at the displacement of 9mm as shown in Figure 5.29.
1'
Ic-a
j.,.
5-26
Chapter 5: Observed behaviour and test results
The in-plane diagonal shear cracks occur at the displacement of 9mm during the test as
shown in Figure 5.30 and Figure 5.31.
5-27
Chapter 5: Observed behaviour and test results
The rocking failure was also observed at the final stage after diagonal cracking as
illustrated in Figure 5.32
The lift-up crack opened wider and wider after the rocking failure occurred as illustrated
in Figure 5.33.
5-28
_______ _________________
Figure 5.34 and 5.35 show the final failure pattern for test specimen, behaved in a
combination of rocking, sliding and diagonal cracking.
I
•
• —. f— ——Orr'— - if
--e---W' -;
-
0
—'---
-
—
1
tu
5-29
Chapter 5: Observed behaviour and test results
Figure 5.36 shows that the FRP debonding as the ultimate value of displacement during
the test and Figure 5.37 shows the cracking bricks at the rear of the wall.
5-30
Chapter 5: Observed behaviour and test results
The in-plane diagonal shear cracks occur at the displacement of 9mm during the test as
shown in Figure 5.38.
Figures 5.39 and 5.40 show the final failure pattern for test specimen, behaved in a
combination of rocking, sliding and diagonal cracking.
5-31
Chapter 5: Observed behaviour and test results
Figure 5.41 sbows the crack initiating at the first cycle of 0.7mm of displacement on the
front of the wall.
5-32
Chapter 5: Observed behaviour and test results
The crack of bricks occurs at the displacement of I 3mm as shown in Figure 5.42 during
the test.
The Figure 5.43 shows that the FRP debonding as the ultimate value of displacement
during the test.
5-33
Chapter 5: Observed behaviour and test results
The in-plane diagonal shear cracks occur at the displacement of 9mm during the test as
shown in Figure 5.44.
Figures 5.45 and 5.46 show the final failure crack pattern for test specimen, behaved in a
combination of rocking, sliding and diagonal cracking.
5-34
Chapter 5: Observed behaviour and test results
if
-
5-35
Chapter 5: Observed behaviour and test results
Figures 5.48 and 5.49 show that the crack of bricks occurs during the testing and the FRP
debonding at the middle of the wall.
5-36
Chapter 5: Observed behaviour and test results
The in-plane diagonal shear cracks occur at the displacement of 9mm during the test as
shown in Figure 5.50.
5-37
Chapter 5: Observed behaviour and test results
The Failure mode for the test wall 7 is quite similar to other walls that have been
discussed, behaved in a combination of rocking; sliding and diagonal cracking. Figures
5.51 and 5.52 show that the wall fail in plane diagonal shear at the ultimate value of
displacement during the test.
ii—
iF
-: II. -
-
J[
-_'
U
.7'
C)
>/ 1
5-38
Chapter 5: Observed behaviour and test results
The summaries of test results are presented in Table 5.1. All the test wall specimens had
the same height-to-thickness ratio, and under the same constant axial compression. As it
is shown in the Table, the cable retrofitting technique can not prevent the wall from
cracking, but it increases more than twice times of the ultimate load capacity of the wall
and ductility. The improvement of the ultimate lateral load resistance of the walls
retrofitted with FRP strips is about three times the capacity of un-retrofitted wall. The
cracking loads of the walls retrofitted with FRP were also increased compared to URM
wall.
The testing results indicate although both retrofitting techniques could significantly
enhance the ductility, stiffness, dissipation of energy and the ultimate load carrying
capacity of URM walls, FRP retrofitting technique is more effective than cable to
improve the ultimate load capacity of the URM walls. However, due to high cost and no
complete knowledge of their mechanical behaviour (there are some concerns that the
inelastic deformation capacity of the wall may be limited as a result of de-bonding of the
FRP sheets which occurs suddenly with little ductility) (Abrams 1998, Ehsani et al.
1999), there are some limitations to apply FRP as a seismic retrofitting method.
Therefore, the cable retrofitting technique could provide an alternative way for seismic
retrofitting of URM walls. The cable retrofitting technique also has advantages of
aesthetic appearance of the wall and low cost. However, drilling and anchoring to brick
must be careful to construct as most brick is hard and brittle.
5-39
Chapter 5: Observed behaviour and test results
Experiments conducted in this study show that both FRP and cable system retrofitted to
low-rise masonry walls are effective in significantly increasing their in-plane strength,
ductility, and energy dissipation capacity. The improvement of the ultimate lateral load
resistance of the retrofitted walls is 2 and 3 times over the capacity of unreinforced wall
for cable and FRP retrofitting techniques respectively. Seismic retrofitting of
unreinforced masonry walls with FRP or cable system proved to be an effective and
reliable strengthening alternative. However as discussed earlier, each retrofitting
technique has its own advantages and disadvantages. Although FRP is more effective in
increasing the ultimate load carrying capacity of URM walls than cable, the high cost
may be a major concern. The initial cost of FRP materials is about 10 times more than
steel (ASBI, 2000). Another concern for FRP is that many practising engineers have not
obtained enough knowledge on FRP; especially their long-term behaviour requires to be
better understood. On the other hand, although cable system could be a low cost
alternative, the fragility of the masonry does not allow a simple fixing of the cable
system.
Note that, although no out-of-plane tests were conducted within the scope of this research,
it's believed that it is preferable to use the proposed cable or FRP system on both sides of
the wall, to provide greater out-of-plane strength and minimize out-of-plane
displacements. Seismic retrofitting of unreinforced masonry walls with cable system or
FRP proved to be an effective and reliable strengthening alternative.
5-40
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
CHAPTER 6
NUMERICAL MODELLING STRATEGIES FOR
UNREINFORECED MASONRY WALLS
6.1 Introduction
In this chapter, the numerical modelling strategies for unreinforced masonry walls are
reviewed and presented.
6-1
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
Table 6-1 summarizes the physical properties that define the natural characteristics of
various problems (Moaveni, 1999)
modulus of elasticity, E
a truss
HI modvlus of elasticity, £
an elastic plate
Load
On the other hand, there are parameters that produce disturbances in a system.
Examples of these parameters include external forces, moments, and temperature
difference across a medium, and pressure difference in fluid flow.
The system characteristics as shown in Table 6.1 dictate the natural behaviour of a
system, and they always appear in the homogeneous part of the solution of a
governing differential equation. In contrast, the parameters that cause the disturbances
appear in the particular solution. It is important to understand the role of these
parameters in finite element modelling in terms of their respective appearances in
stiffness or conductance matrices and load or forcing matrices. The system
6-2
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
Structural analysis methods, as shown in Figure 6.1, may be divided into two major
categories: (1) analytical (classical) and (2) numerical methods. (Nagy, 1989) The
development of analytical methods goes back to several centuries. This method may
be further subdivided into the exact (closed formed), and the approximate solution
groups. Unfortunately, analytical techniques can only be used for simple structures
which severely limits their utilization, although they serve useful purpose in teaching
strength of material and engineering mechanics• courses, Most real structures have
complex geometrical shapes, as well as loading and boundary conditions for which
numerical methods usually give better results. Numerical techniques are relatively
new, but the popularity rapidly increased with the availability of digital computers.
Numerical methods can also be further subdivided into the following groups: (a) finite
differences, (b) finite element, and (c) boundary integral techniques. (Nagy, 1989)
The finite difference method is based on replacing the governing partial differential
equations and the equations of boundary conditions by the corresponding finite
difference equations. This reduces the problem to a set of simultaneous algebraic
equations. The finite difference solution gives a point-wise approximation to the
governing equations. It requires a uniform, square mesh, thus creates difficulties in
describing irregular shaped structure.
Finite
Element
6-3
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
The finite element technique is currently the most popular numerical method. It is also
based on solving simultaneous algebraic equations, but it gives a piece-wise
approximation to the governing equations. Both of these methods are based on
discretization of the structure, however in the finite element method different element
type can be utilized and the mesh size can be varied, thus it provides a closer
approximation of the structure boundary. The difference between FDM and FEM is
show in Figure 6.2. (Nagy, 1989)
( -- I'
—
I II
It
/
—
-TA A
—---
Finite Difference Finite Element
Model Model
Figure 6-2 The difference between FDM and FEM (Nagy, 1989)
The boundary integral technique is still more or less in a research stage, no general-
purpose codes are in existence. By contrast, many general-purpose programs based on
the finite element method exist on the market.
The primary differences between classical methods and finite elements are the way
they "view" the structure and the ensuing solution procedure. Classic methods
consider the structure as a continuum whose behavior is governed by partial or
ordinary differential equations. The finite element method considers the structure to
be an assembly of small finite-sized particles. The behaviour particles and the overall
structure are obtained by formulating a system of algebraic equations that can be
readily solved with a computer.
The finite element method (FEM) is a computer based procedure that can be used to
analyze structures and continua. FEM is based on the idea of building a complicated
6-4
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
object with simple blocks, or dividing a complicated object into small and
manageable pieces. (See Figure 6.3)
21
Actual Structure Finite Element Mode!
In other words, the concept of finite element method is to replace the infinite number
of interconnected points of continuum with discrete numbers. This is accomplished
by assembling several or many simple shaped small (finite) elements in such a way
that the overall structure is approximated. This procedure is similar to putting the
jigsaw puzzle pieces together to obtain the total picture. The main difference is that
the jigsaw puzzle pieces have complex shapes, while the finite elements are made
simple.
In this study the simple truss model and the finite element method are adopted to
simulate the structural behavior. The finite element method is a powerful numerical
method for the analysis of structures, and it has been extensively used for analyzing
the behaviour of masonry structures in the last decade. The proposed numerical
models of URM walls retrofitted by cable system and FRP are generated using
ABAQUS finite element program.
The model considers both material nonlinearity and progressive cracking. It is based
on a simple yield surface with isotropic hardening and associated flow when the state
of stress is predominantly compressive, and uses damaged elasticity (smeared
cracking) to account for cracking, the onset of cracking is defined by a so-called
"crack detection surface". Due to the limitation of ABAQUS, the reduction of
stiffness caused by inelastic straining, which is very important for masonry is
6-5
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
neglected. Therefore, the model is not capable of predicting cyclic response; besides
the use of associated flow usually leads to over-prediction of volume strain. In spite of
these limitations, the model has proven to be able to reasonably predict the masonry
behaviour in monotonic loading or the envelope of cyclic response, as long as proper
material definition is provided. However, the main objective of this research is to
prove a new retrofitting technique with experimental testing. These ABAQUS models
are for limit-state examination purpose only.
Masonry is a composite material that consists of units and mortar joints. A detailed
analysis of masonry, hereby denote micro-modeling, must then include a
representation of units, mortar and the unitlmortar interface. This approach is suited
for small structural elements with particular interest in strongly heterogeneous states
of stress and strain. The primary aim of micro-modeling is to closely represent
masonry from the knowledge of the properties of each constituent and the interface.
The necessary experimental data must be obtained from laboratory tests in the
constituents and small masonry samples. Several attempts to use interfaces for the
modeling of masonry were carried out in the last decade with reasonably simple
models, see Anthoine (1992) and Lourenco (1996) for references. In particular,
gradual softening behavior and all failure mechanisms, namely tensile, shear and
compressive failure, have not been fully considered.
In large and practice-oriented analysis the knowledge of the interaction between units
and mortar is, generally, negligible for the global structural behavior. In these cases a
different approach can be used, hereby denoted macro-modeling, where the material
is regarded as an anisotropic composite and a relation is established between average
masonry strains and average masonry stresses. This is clearly a phenomenological
approach, meaning that the material parameters must be performed in masonry tests
of sufficiently large size under homogeneous states of stress. A complete macro-
model must reproduce an orthotropic material with different tensile and compressive
strengths along the material axes as well as different inelastic behavior for each
6-6
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
material axes. A reduced number of orthotropic material models specific for masonry
has been proposed by many researchers, see Anthoine (1992) and Dhanasekar et al.
(1984, 1985c) for references.
Masonry is a material which exhibits distinct directional properties due to the mortar
joints which act as planes of weakness. In general, the approach towards its numerical
representation can focus. on the micro-modeling of masonry as components, such as
unit (brick, block, etc.) and mortar, or the macro-modeling of masonry as a composite,
see also Rots (1991). Depending on the level of accuracy and the simplicity desired, it
is possible to use the following modeling strategies, see Figure 6.4.
Li Ii I
(a) (b)
(c) (d)
Figure 6.4 Modelling strategies for masonry structures: (a) masonry sample ;( b)
detailed micro-modelling; (c) simplified micro-modelling; (d) macro-modelling
(Lourenco, 1996)
6-7
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
• Macro-modeling — unit, mortar and unit-mortar interface are smeared out in the
continuum.
In the first approach Detailed micro-modeling, Young's modulus, Poisson's ratio and,
optionally, inelastic properties of both unit and mortar are taken into account. The
interface represents a potential crack/slip plane with initial dummy stiffness to avoid
interpenetration of the continuum. This enables the combined action of unit, mortar
and interface to be studied under a magnifying glass. In the second approach
LI Simplified micro-modeling, each joint, consisting of mortar and the two unit-mortar
interface, is lumped into an "average" interface while the units are expanded in order
to keep the geometry unchanged. Masonry is thus considered as a set of elastic blocks
bonded by potential fracture/slip lines at the joints. Accuracy is lost since Poisson's
effect of the mortar is not included. The third approach LlMacro-modeling does not
make a distinction between individual units and joints but treats masonry as a
homogeneous anisotropic continuum. One modelling strategy cannot be preferred
over the other because different application fields exist for micro- and macro-models.
Micro-modeling studies are necessary to give a better understanding about the local
behavior of masonry structures. This type of modeling applies notably to structural
details, but also to modem building systems like those of concrete or calcium-silicate
blocks, where window and door openings often result in piers that are only a few
block units in length. These piers are likely to determine the behavior of the entire
wall and individual modeling of the blocks and joints is then to be preferred. Macro-
models are applicable when the structure is composed of solid walls with sufficiently
large dimensions so that the stresses across or along a macro-length will be essentially
uniform. Clearly, macro-modeling is more practice oriented due to the reduced time
and memory requirements as well as a user-friendly mesh generation. This type of
modeling is most valuable when a compromise between accuracy and efficiency is
needed. Therefore, the macro-modelling is adopted to simulate the masonry behavior
in this study as the major objective is to study the retrofitting effect of the wall.
6-8
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
The proposed numerical model is based on ABAQUS concrete model and is modified
to simulate the special characterises of masonry. The model employs oriented
damaged elasticity (smeared cracking) and isotropic compressive plasticity to
represent the inelastic behaviour of masonry. The model is intended to simulate the
masonry behaviour for relatively monotonic loadings under fairly low confining
pressures. The model adopts a smeared crack approach in the sense that it does not
track individual "macro" cracks. Constitutive calculations are performed
independently at each integration point. The presence of cracks enters into these
calculations by the way in which the cracks affect the stress and material stiffness
associated with the integration point. The anisotropy introduced by cracking is
assumed to be important in the simulations for which the model is intended, so the
model includes consideration of this anisotropy.
Due to the limitations of ABAQUS, the model makes no attempt to include prediction
of cyclic response or of the reduction in• the elastic stiffness caused by inelastic
straining. Nevertheless, it is likely that--even in such cases--the stress .trajectories will
not be entirely radial and the model must predict the response in such cases in a
reasonable way. An isotropically hardening "compressive" yield surface forms the
basis of the model for the inelastic response when the principal stresses are
dominantly compressive. In tension once cracking is defined to occur (by the "crack
detection surface" of the model), the orientation of the cracks is stored and oriented,
damaged elasticity is then used to model the existing cracks. Stress components
associated with an open crack are not included in the definition of the crack detection
surface for detecting additional cracks at the same point, and it only allowed cracks to
form in orthogonal directions at a point.
Since ABAQUS is an implicit, stiffness method code and the material calculations
used to define the behaviour of the masonry are carried out independently at each
integration point in that part of the model that is made of masonry, the solution is
known at the start of the time increment. The constitutive calculations must provide
values of stress and material stiffness at the end of the increment, based on the current
6-9
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
estimate of the kinematic solution for the response at the spatial integration point
during the increment that provides the (logarithmic) strain at the end of the increment.
Once cracks exist at a point, the component forms of all vector and tensor valued
quantities are rotated so that they lie in the local system defined by the crack
orientation vectors (the normals to the crack faces). The model ensures that these
crack face normal vectors will be orthogonal, so that this local system is rectangular
This use of a local system simplifies the computation of the damaged
elasticity used for the components associated with existing cracks.
The model, thus, consists of a "compressive" yield/flow surface to model the masonry
response in predominantly compressive states of stress, together with damaged
elasticity to represent cracks that have occurred at a material calculation point, the
occurrence of cracks being defined by a "crack detection" failure surface that is
considered to be part of the elasticity. The details of the model are discussed in the
following sections.
Most masonry walls subjected to in-plane loads are in a state of biaxial stress.
Therefore, it is necessary to consider a biaxial stress-strain model for masonry subject
to in-plane loading. The state of stress of masonry is shown in Figure 6.5.
The stress-strain relations of masonry depend on the state of stress. The typical
unaxial stress-strain curve for masonry under compression is shown in Figure 6.6.
6-10
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
Lln on
a.
o-It
(+)
Figure 6.6 The typical unaxial stress-strain curve for masonry under compression
(Zhuge, 1996)
6-11
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
The compressive strength of masonry in the direction normal to the bed joints has
been traditionally regarded as the sole relevant structural material property, at least
until the recent introduction of numerical methods for masonry structures. A test
frequently used to obtain this uniaxial compressive strength is the stacked bond
prism, see Figure 6.7 (a), but it is still somewhat unclear what are the consequences
in the masonry strength of using this type of specimens, according to Mann and
Betzler (1994). It is commonly accepted that the real uniaxial compressive strength of
masonry in the direction normal to the bed joints can be obtained from the so-called
RILEM test, see Wesche and Ilanzis (1980), shown in Figure 6.7 (b). The RILEM
specimen is however relatively large and costly to execute, especially when
compared to the standard cube or cylinder tests for concrete. Since the pioneering
work of Hilsdorf (1969) it has been accepted by the masonry community that the
difference in elastic properties of the unit and mortar is the precursor of failure.
Uniaxial compression of masonry leads to a state of triaxial compression in the
mortar and of compression/biaxial tension in the unit. Mann and Betzler (1994)
observed that, initially, vertical cracks appear in the units along the middle line of the
specimen, i.e. continuing a vertical joint. Upon increasing deformation additional
cracks appear, normally vertical cracks at the small side of the specimen, that lead to
failure by splitting of the prism. Examples of load-displacement diagrams obtained in
500 X 250 X 600 [mm3] prisms of solid soft mud bricks are shown in Figure 6.7 (c).
Increasing strength leads to a more brittle behavior. The quite high value of the
compressive fracture energy G which equals approximately 45 [Nmm/mm2 can
6-12
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
be explained by the continuum fracture energy and the 600 [mm] height of the
specimen, see Vonk (1992). Atkinson and Yan (1990) collected information about
the compressive stress-strain relation for masonry, but these results are not energy
based.
(a)
hj
h 5/tb ,,'iii1 32
ii4 b
fl_c) ----i
b -. 0.0 4.0 6.0 S.0 10.0
(,nrnj
(b) (c)
Figure 6.7 Uniaxial compressive behaviour of masonry upon loading normal to the
bed joints: (a) stacked bond prism; (b) schematic representation of RILEM test
specimen; (c) typical experimental stress-displacement diagrams for 500 X 250 X 600
[mm3] prisms of solid soft mud bricks, Binda et al. (1988). Here, is the mortar
compressive strength.
Uniaxial compression tests in the direction parallel to the bed joints have received
substantially less attention from the masonry community. However, masonry is an
anisotropic material and, particularly in the case of low longitudinal compressive
strength of the units due to high or unfavorable perforation, the resistance to
compressive loads parallel to the bed joints can have a decisive effect on the load
bearing capacity. According to Hoffmann and Schubert (1994), the ratio between the
uniaxial compressive strength parallel and normal to the bed joints ranges from 0.2 to
0.8. These ratios were obtained for masonry samples of solid and perforated clay
6-13
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
units, calcium silicate units, lightweight concrete units and aerated concrete units.
The current model adopts the uniaxial compressive strength normal to the bed joints.
For tensile loading perpendicular to the bed joints, failure is generally caused by
failure of the relatively low tensile bond strength between the bed joint and the unit.
As a rough approximation, the masonry tensile strength can be equated to the tensile
bond strength between the joint and the unit.
In masonry with low strength units and greater tensile bond strength between the bed
joint and the unit, e.g. high-strength mortar and units with numerous small
perforations, which produce a dowel effect, failure may occur as a result of stresses
exceeding the unit tensile strength. As a rough approximation, the masonry tensile
strength in this case can be equated to the tensile strength of the unit.
For tensile loading parallel to the bed joints a complete test program was set-up by
Backes (1985). The specimen consists of four courses, initially laid down in the usual
manner; see Figure 6.8 (a). A special device attached to the specimen turns it 90° in
the intended direction of testing shortly before the test time, see Figure 6.8 (b). The
load is applied via steel plates attached to the top and bottom of the specimen by
special glue (unspecified in the paper). The entire load-displacement diagram is
traced upon displacement control. Two different types of failure are possible;
depending on the relative strength of joints and units, see Figure 6.9. In the first type
of failure cracks zigzag through head upon increasing deformation. The post-peak
response of the specimen is governed by the fracture energy of the head joints and the
post —peak mode II behavior of bed joints. In the second type of failure cracks run
almost vertically through the units and head joints. A typical stress-displacement
diagram shows progressive softening until zero. The post-peak response is governed
by the fracture energy of the units and head joints. The current model adopts the
uniaxial tensile strength normal to the bed joints.
6-14
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
(h)
Figure 6.8 Test set-up for tensile strength of masonry parallel to the bed joints,
(Backes, 1985): (a) building the test specimen; (b) test specimen before 90°rotation
and testing.
0,
061)
I fnun)
6-15
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
MatesMi
6-16
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
It is further noted that the strength envelope shown in Figure 6.10 is of limited
applicability for other types of masonry. Different strength envelops and different
failure modes are likely to be found for different materials, unit shapes and geometry.
A comprehensive program to characterize using full scale specimens, see Ganz and
Thurlimann (1982) for hollow clay units masonry, Guggisberg and Thurlimann
(1987) for clay and calcium-silicate units masonry and Lurati et al. (1990) for
When the principal stress components are dominantly compressive, the response of
the masonry is modelled by an elastic-plastic theory, using a simple form of yield
surface written in terms of the first two stress invariants. Associated flow and
isotropic hardening are used. The cracking and compression responses of masonry
that are incorporated in the model are illustrated by the equivalent uniaxial response
of a specimen shown in Figure 6.11. The biaxial effect has been considered in the
model. The uniaxial compressive strength will be discussed in Section 6.4.2.2.
= In
Ineia*tlc
6-17
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
The current model uses the classical concept of plasticity theory: a strain rate
decomposition into elastic and inelastic strain rates,
de = ded + (6-1)
where de is the total mechanical strain rate, dee! is the elastic strain rate (which
includes crack detection strains--this elastic strain will be further decomposed when
we describe the elasticity), and is the plastic strain rate associated with the
"compression" surface.
We assume that the elastic part of the strain is always small, so this equation can be
integrated as
S= + (6-2)
(6-3)
otherwise, =0.
In the definition of dec, c0 is a constant that is chosen so that the ratio of ef1' in a
monotonically loaded biaxial compression tests to in a monotonically loaded
6-18
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
uniaxial compression test is a value specified as part of the failure surface data.
The equation defining c0 from and the other constants in the compression surface
is derived next.
(6-4)
Since
ac7 3
and
— 3 S
— 2 q
then
(de:' = + — (6-5)
(e:' = + — iJ (6-6)
6-19
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
the total plastic strain components that would occur in monotonically loaded biaxial
and uniaxial compression tests. In biaxial compression when both nonzero principal
1
S11 so the flow rule gives
(6-7)
Using this equation and Equation 6-5 then defines from and the other constants as
—
2 (-)
The constant a0 is chosen from the ultimate stress reached in biaxial compression to
the ultimate stress reached in uniaxial compression. The derivation of a0 can be found
in Section 6.5.2.
6-20
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
which means they grow only when the load is increased. Around peak load an
acceleration of crack formation takes place and formation of macro crack starts. The
macro cracks are unstable, which means that the load has to decrease to avoid an
uncontrolled growth. In a deformation controlled test the macro crack growth results
in softening and localization of cracking in a small zone while the rest of the
specimen unloads.
For tensile failure this phenomenon has been well identified, see e.g. Hordijk (1991).
For shear failure, a softening process is also observed as degradation of the cohesion
in Coulomb friction models. For compressive failure, softening behavior is highly
dependent upon the boundary conditions in the experiments and the size of the
specimen, Van Mier (1984) and Vonk (1992). Experimental concrete data provided
by Vonk (1992) indicated that the behavior in uniaxial compression is governed by
both local and continuum fracturing processes.
Figure 6.12 shows some classical models for the post-peak tensile behaviour, where
is the uniaxial tensile strength and cr is the strain at the stress level of
c ,
Figure 6.12 Post-peak tensile stress-strain models (Born and Sorace, 1993)
6-21
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
Model (e) is commonly used to analyze ductile material and would not be sufficiently
accurate in representing the true material and would not be sufficiently accurate in
representing the true material behaviour of masonry.
Models (a) and (b) are the most commonly implemented methods in the analysis of
brittle material. Model (a) is called the "tension cut-off model", represents a brittle
failure, consisting of a total stress released in the direction orthogonal to the crack.
Research has shown that the tensile softening effects have generally a limited effect
on the global response of plain concrete and masonry structures (Rots, 1998; Rots
and Blauwendraad, 1989). At the same time, the presence of softening mechanical
behaviour is very burdensome for computation. For tension softening, the bilinear
model (d) is adopted in the current model.
Following crack detection the damaged elasticity could be used to model the failed
material. The elasticity is written in the form
(6-9)
Let a represent a cracked direction, with corresponding direct stress a— and direct
6-22
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
summation is implied by repeated indices with a bar over them. If the fracture energy
concept is used, the strains are related to the user-specified stress/displacement
definition for the tension stiffening behaviour bye = u / c, where c is the characteristic
length associated with the integration point.
open open
where is the stress corresponding to e (as defined in the tension
C
open
= max(C— I
(6-11)
open
(6-12)
D (6-13)
We also assume no Poisson effect for open cracks: for >0, D— =0 for
The shear terms in the elasticity associated with existing crack directions are
D =G,
ajkrfl
6-23
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
I .0
the user in the shear retention data. Here where (and) are
\ax/ + \/313/ ,
0
" ' otherwise
L0
6-24
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
6.5.1 Introduction
The failure of masonry under in-plane lateral load has been investigated for many
years. Unlike other materials (concrete or steel), masonry is a non-homogeneous
material. Therefore, it is not possible to attribute failure of masonry to a single cause.
Many types of failure are possible and the one that gives the lower bound is the
critical one. Furthermore, it has been extensively observed that the strength of
masonry under multiple stress conditions is a function of the state of stress and
cannot be predicted using simple tensile, compressive, and shearing stress
independently of each other. Because masonry is a material, which exhibits
directional properties due to the mortar joints acting as a plane of weakness, a
realistic failure model should consider this effect as well. Depending upon the stress
state acting on the joints, failure may occur in the joints alone, or in some form of a
combined mechanism involving the mortar and masonry unit. Therefore, to develop a
rational failure criterion, all the above factors must be taken into account.
To formulate a failure criterion for masonry, it is necessary to allow for all the
possible modes of failure. According to the experimental results from the present and
other research, the failure modes of URM under in-plane lateral loads are illustrated
in Figure 6.14.
6-25
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
Figure 6.14 Modes of failure of solid clay units masonry under biaxial loading,
(Dhanasekar et al. 1985c)
For uniaxial tension, failure occurred by cracking and sliding of the head and bed
joints. The influence of the lateral tensile stress on the tensile strength is not known
because no experimental results are available. A lateral compressive stress decreases
the tensile strength, which can be explained by the damage induced in the composite
material, by microslip of the joints and microcracking of the units. In the tension
compression loading cases failure occurred either by cracking and sliding of the
joints alone or in a combined mechanism involving both units and joints. Similar
types of failure occurred for uniaxial compression but a smooth transition is found to
other type of failure mode in biaxial compression. In biaxial compression failure
6-26
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
For principal stress ratios<< 1 and>> 1, the orientation played a significant role and
failure occurred in a combined mechanism involving both joint failure and lateral
splitting. The increase of compressive strength under biaxial compression can be
explained by friction in the joints and internal friction in the units and mortar.
For concrete, the failure envelope seems to be largely independent of the loading
path, Nelissen(1972), which confirms the presence of single failure mode, i.e.
continuous crack growth at the microlevel. Presently, it is not known if the masonry
failure envelope obtained by Page is valid for nonproportional loading, particularly
because different failure modes can be triggered. Another point is that experimental
data about the softening of masonry under biaxial stress state.
The common method to predicting the failure of masonry is the biaxial failure
criterion (defined in terms of principal stresses). Figure 6.15 shows a typical failure
surface of masonry, but it didn't consider the effect of joint. Von Mises yield surface
is used to predict a crushing type of failure, whereas a tension cut-off surface is used
for tension-tension or tension-compression type of failure.
Tension
Compression
1
1'rn
Crushing Failure
Regime
Tensile
Regime Mises yield envelope
Failure critenon:
Concrete
— 0- 02 ÷ — =
Compression
6-27
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
A simplified failure envelope for masonry has been developed by Zhuge (1996). This
failure envelope, which combines both biaxial and Coulomb shear failure models, is
capable of predicting joint failure through the use of the ubiquitous joint
model(Zienkiewicz and Pande, 1977).
The failure envelope has been shown in Figure 6.16 (Zhuge, 1996) with various stress
ratios = 5, 10, 15) .To illustrate the failure criteria discussed in this study, the
It can be seen from this figure that when 45 0 <0 900 (a low ratio of It), failure
is controlled by the shear failure criterion [straight line], which agrees well with the
test results of Page (1982), showing that the crack developed along the bed joints due
to shear slips in this range. When the bed joint angle 0 decreased, the shear strength of
the wall increased as increased. If the shear strength exceeds the principal tensile
stress 0 <450, the failure will be controlled by the biaxial failure criterion (principal
tensile stress).
6-28
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
'C
In multiaxial stress states the failure criteria can be generalized through the concept of
surfaces of failure and of ultimate strength in stress space. These surfaces are defined
below and are fitted to experimental data. Typical surfaces are shown in Figure 6.17.
6-29
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
ufli*xi&I tension a
/
corn
/ / A
/ bI Ia I
/
tø iii Ion
/.
com
/
rtace
/
bIaxIal compresalon
(6-14)
1
p= ——trace (a) , (6-15)
(6-16)
6-30
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
is the size of the yield surface on the q -axis at p =0, so that is the yield stress
in a state of pure shear stress when all components of a are zero except
a12 = = The hardening is measured by the value of the relationship
is user-specified.
This simple surface is a straight line in (p - q) space (Figure 6.16) and provides a
good match to experimental data over a fairly wide range of pressure stress values (up
to four to five times the maximum compressive stress that can be carried by the
masonry in uniaxial compression). This form of the surface means that, as the
hardening changes, the surfaces in (p - q) space are similar, so--for example--
the ratio of flow stress in biaxial loading to flow stress in uniaxial loading is the same
at all flow stress levels. This does not appear to be contradicted by any experimental
data, and it means that only one constant (a0) is needed to define the shape of the
surface. -
(6-17)
In biaxial compression p =2/3 abC and q = abC, where abC is the magnitude of each
nonzero principal stress. Therefore, on =0,
(1 2a0
(618)
abC 3
6-31
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
The value of / = is specified by the user as part of the failure surface data.
(6-19)
Cracking dominates the material behaviour when the state of stress is predominantly
tensile. The model uses a "crack detection" plasticity surface in stress space to
determine when cracking takes place and the orientation of the cracking. Damaged
elasticity is then used to describe the post failure behaviour of the masonry with open
cracks.
Numerically the "crack detection" plasticity model was used for the increment in
which cracking takes place and subsequently damaged elasticity was used once the
crack's presence and orientation have been detected. As a result there is at least one
increment in which the crack detection "plastic" strains were calculated. Since these
are really just the outcome of a numerical device to treat cracking, they are recast as
elastic strains in the direction of cracking and as plastic strains in the other directions.
(This means that we retain the stresses calculated for equilibrium purposes, as well as
the strain decomposition of Equation 6-1.)
f, =0 (6-20)
where is the failure stress in uniaxial tension and is a constant that is defined
from the value of the tensile failure stress, in a state of biaxial stress when the
6-32
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
The stress measures p and q are defined in the same way as p and q, except that
all stress components associated with open cracks (that is, if a or /3 is a crack
direction in which the direct strain >0 or >0) are not included in these
,8,8
This surface has a simple mathematical form but matches plane stress data quite well.
The hardening is introduced in the particular form shown in Equation 6-15 so that, as
—+ 0, the surface becomes q — = 0, which in (p - q) space is the cone containing
the principal axes of stress. This means that, as the tension stiffening is exhausted in a
plane stress test, the stress point will drop back onto the nearest principal stress axis.
The value of is obtained as follows. The failure surface data include a definition of
a ratio that states that, in a plane stress test cracking would occur when one
principal stress has the value = — is the magnitude of the ultimate stress in
uniaxial compression) and the other nonzero principal stress has the value =
Another value provided as part of the failure surface data is which defines
= (6-21)
= (i — (6-22)
and
6-33
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
(6-23)
Therefore, with =
f1 =0 (6-24)
so
(6-25)
1+ (1—f)
The crack detection model uses the assumption of associated flow, so if f, =0 and
def1 = (6-26)
The major reason for the nonlinear behaviour of masonry is progressive cracking of
the material due to the tensile strength of masonry being lower that its compressive
strength.
1. Smeared-cracking model
2. Discrete-cracking model
3. Fracture-mechanics model
6-34
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
The particular cracking model to be selected from the three alternatives depends upon
the purpose of the analysis. In general, if overall load-deflection behavior is desired,
without regard to completely realistic crack patterns and local stresses, the smeared-
crack model is probably the best choice. If detailed local behavior is of interest,
adaptations of the discrete-cracking model are useful. For the special class of
problems in which fracture mechanics is the appropriate tool, a specialized fracture
model may prove necessary. For most structural engineering applications, the
smeared-cracking modeling is generally used.
In this approach, the cracked concrete is assumed to remain a continuum; i.e., the
cracks are smeared out in a continuous fashion. It is assumed that the concrete
becomes orthotropic or transversely isotropic after the first cracking has occurred,
one of the material axes being oriented along the direction of cracking. Such
formulations easily allow for a gradual build-down or sudden drop of strength in the
direction of tension. Also, shear-strength reserves becasue aggregate interlocking can
be accounted for by retaining a positive shear modulus. This shear capacity also
means that secondary cracking does not necessarily appear perpendicular to the first
direction of cracks.
In the smeared-cracking model, a crack is not discrete but implies an infinite number
of parallel fissures across that part of the finite element (Fig 6.18). After cracking has
occurred, the cracked masonry becomes an orthotropic material and new incremental
relationships must be derived. This is accomplished by modifying the tan gent-
material-stiffness or the tangent-elasticity matrix [Di]. For instance, for plane stress
the stress-strain relationships in the cracked directions (n and t axes in Fig.6. 18)
become
Ao = [D,1 (6-27)
Ar [Ay
6-35
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
000
[D]=o E 0 (6-28)
00/3G
As (6-29)
E
In Eq. (6-28) the modulus of elasticity E of the masonry is reduced to zero in the
direction normal to the cracked (n axis). Further, a reduced shear modulus is
assumed on the cracked plane to account for the aggregate interlocking. For stiffness
calculations it is necessary to transform the I matrix into global coordinates, and
this is accomplished by using the well-known transformation rule for stress and strain
tensors
6-36
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
The smear crack model is used in the ABAQUS concrete model. Constitutive
calculations are performed independently at each integration point of the finite
element model. The presence of cracks enters into these calculations by the way in
which the cracks affect the stress and material stiffness associated with the
integration point.
a) Cracking Failure
Once tensile failure occurs, the stress normal to the crack is reduced to a very small
value. It is assumed that this is released, with the other stress parallel to the crack
remaining unchanged. The shear modulus is assumed to have a small residual value
to simulate frictional resistance due to aggregate interlock and this can be achieved
through an aggregate interlock factor 13'.
b) Crushing Failure
Once crushing failure occurs, all local stresses are released completely and the
material is assumed to lose all its capacity to transmit further loads. All the terms of
the constitutive matrix are reduced to approximately zero (a very small value is
adopted to avoid numerical difficulty).
6-37
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
Figure 6-19 [D1 } matrices for different modes of failure (Zhuge, 1996)
6-38
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
c) Shear failure
When shear sliding type of failure occurs, the shear stress must be reduced to that of
the allowable shear stress and not allowed to increase with further loading. To
simulate the sliding deformation upon further application of load, the shear modulus
G is reduced to a small value by multiplying it by the aggregate interlock factor
Stiffness in other directions is kept unaltered (see Figure 6.19).
6-39
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
(a) (b)
The success fracture-mechanics theory has been used in solving various types of
cracking problems in metals, ceramics, and rocks leads to its use in finite-element
analysis of masonry structures. If one accepts the fact that masonry is a notch-
sensitive material, a cracking criterion based on tensile strength may be dangerously
unconservative and the use of fracture-mechanics theory will provide a more rational
approach to concrete cracking. In its current state of development, however, the
practical applicability of fracture mechanics to masonry is still questionable, and
much remains to be done (Kesler et al., 1972) At present, this area is being very
actively studied by several researchers (e.g., Hillemier and Hilsdorf, 1997; Bazant and
Cedoline, 1980).
6-40
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
Cracks are irrecoverable: they remain for the rest of the calculation (but may open and
close). No more than three cracks can occur at any point (two in a plane stress case,
one in a uniaxial stress case). Following crack detection, the crack affects the
calculations because a damaged elasticity model is used.
q
U
auilace
I
Figure 6.21 Masonry failure surface in the p-q plane (ABAQUS, 2000)
6-41
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
For the purpose of developing the model, it is assumed that the material loses strength
through a softening mechanism and that this is dominantly a damage effect, in the
sense that open cracks can be represented by loss of elastic stiffness (as distinct from
the nonrecoverable straining that is associated with classical plasticity effects, such as
what is used for the compressive behaviour model). The model neglects any
permanent strain associated with cracking; that is, we assume that the cracks can close
completely when the stress across them becomes compressive.
Various researchers have proposed three basic crack direction models (Rots and
Blaauwendraad, 1989): fixed orthogonal cracks; the rotating crack model; and fixed
multidirectional (nonorthogonal) cracks. In the fixed orthogonal crack model the
direction normal to the first crack is aligned with the direction of maximum tensile
principal stress at the time of crack initiation. The model has memory of this crack
direction, and subsequent cracks at the point under consideration can only form in
directions orthogonal to the first crack. In the rotating crack concept only a single
crack can form at any point (aligned with the direction of maximum tensile principal
stress). Thus, the single crack direction rotates with the direction of the principal
stress axes. This model has no memory of crack direction. Finally, the
multidirectional crack model allows the formation of any number of cracks at a point
as the direction of the principal stress axes changes with loading. In practice, some
limitation is imposed on the number of cracks allowed to form at a point. The model
has memory of all crack directions.
The multidirectional crack model is the least popular, mainly because the criterion
used to decide when subsequent cracks form (to limit the number of cracks at a point)
is somewhat arbitrary: the concept of a "threshold angle" is introduced to prevent new
cracks from forming at angles less than this threshold value to existing cracks. The
fixed orthogonal and rotating crack models have both been used extensively, even
though objections can be raised against both. In the rotating crack model the concept
of crack closing and reopening is not well-defined because the orientation of the crack
6-42
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
can vary continuously. The fixed orthogonal crack model has been criticized mainly
because the traditional treatment of "shear retention" employed in the model tends to
make the response of the model too stiff. This problem can be resolved by
formulating the shear retention in a way that ensures that the shear stresses tend to
zero as deformation on the crack interfaces takes place (this is done in the ABAQUS
model, as described later). Finally, although the fixed orthogonal crack model has the
orthogonality limitation, it is considered superior to the rotating crack model in cases
where the effect of multiple cracks is important (the rotating crack model is restricted
to a single crack at any point).
The fixed orthogonal cracks model is adopted in the current model so that the
maximum number of cracks at a material point is limited by the number of direct
stress components present at that material point of the finite element model (for
example, a maximum of three cracks in three-dimensional, axisymmetric, and plane
strain problems or a maximum of two cracks in plane stress problems). Once cracks
exist at a point, the component forms of all vector and tensor valued quantities are
rotated so that they lie in the local system defined by the crack orientation vectors (the
normals to the crack faces). The model ensures that these crack face normal vectors
are orthogonal so that this local system is rectangular Cartesian. Crack closing and
reopening can take place along the directions of the crack surface normals. The model
neglects any permanent strain associated with cracking; that is, we assume that the
cracks can close completely when the stress across them becomes compressive.
As the masonry cracks, its shear stiffness is diminished. This effect is defined in the
* SHEAR RETENTION option by specifying the reduction in the shear modulus as a
function of the opening strain across the crack. This option also provides for the
definition of a reduced shear modulus for closed cracks. This reduced shear modulus
will also have an effect when the normal stress across a crack becomes compressive.
The new shear stiffness will have been degraded by the presence of the crack.
6-43
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
The modulus for shearing of cracks is defined as pG, where G is the elastic shear
modulus of the uncracked masonry and p is a multiplying factor. The shear retention
model assumes that the shear stiffness of open cracks reduces linearly to zero as the
crack opening increases:
p=O (6-31)
where is the direct strain across the crack and E is the value given on the data line
of the *SHEAR RETENTION option. The model also assumes that cracks that
subsequently close have a reduced shear modulus:
close
= for < 0, (6-32)
6.7 Conclusion
In this study, the finite element method is adopted for numerical modelling to
simulate masonry behaviour. Masonry is a composite material that consists of units
and mortar joints. Macro-modelling is adopted to model this composite material due
to the practical purpose. To develop an accurate finite element macro-modelling, it is
very important to give an accurate composite material behaviour and composite
failure criteria.
6-44
Chapter 6: Numerical modelling strategies for unreinforced masonry walls
6-45
Chapter 7: A basic mechanical model for the retrofitted walls
CHAPTER 7
A BASIC MECHANICAL MODEL FOR THE
RETROFITTED WALLS
7.1 General
The results of the experimental study carried out in this investigation demonstrate that
the proposed cable system is an effective seismic retrofitting strategy. Analysis of this
retrofitted wall with cable system will be conducted in this chapter in an effort to
explain the mechanism of load resistance. Simple analysis tools, suitable for the
design office environment will be developed and the results will be compared with
those obtained from the experimental research. A simple truss model based on the
theory of plasticity is presented and verified by the experimental data.
A simplified truss model is introduced for the analysis of retrofitted walls to obtain
their load-deflection relationships. An indeterminate truss with five members is used
to represent retrofitted walls as shown in Figure 7.1. The masonry strut in tension is
presented as member '1' in Figure 7.1. The width of this vertical tension strut is
assumed to be one tenth of wall length. The rigid horizontal member in the plane of
the floor or roof, or the upper rigid top beam in the case of wall specimens is labelled
'2'. The masonry strut in compression is referred to as member '3'. The width of this
vertical compression strut is assumed to be one tenth of wall length. The diagonal
member in tension and compression are labelled as member '4' and '5', respectively.
7-1
Chapter 7: A basic mechanical model for the retrofitted walls
The cable in tension is labelled as member '4'. The section properties of this element
were taken equal to the cable section properties. Member '5' is a masonry diagonal
compressive strut with an effective width taken as one-thirty two of the wall length.
Members '1' and '3', as well as '4' and '5' swap places when the horizontal load
reverse.
Figure 7.1 A simple truss model for the retrofitted wall with cable system
7-2
Chapter 7: A basic mechanical model for the retrofitted walls
F5
(
1
Rax Rbx
Ray Rby
t,F4 PS
P3
8 Jr
a
8
b
Ray
I
Rby
Rbx
(b) A free body for point "a" (c) A free body for point "b"
7-3
Chapter 7: A basic mechanical model for the retrofitted walls
Fx = 0
Z Fx=O
Then, from the equilibrium of forces in the horizontal direction in the Figure 7.2 (a),
we obtain
Fx = 0
Substituting equations (7-1) and (7-2) into this equation (7-3), we find
From equation (7-4), it shows that the force V is contributed from F4 (cable) and F5
(masonry). Or on the other word, the stress-strain relationship for the retrofitted wall
can be obtained from combing the stress-strain relationship of the masonry and cable
as shown in Figure 7.3. Therefore, this structure can be simplified by dividing it into
two static structures as shown in Figure 7.4. The displacement A is caused by the
force V. Because element 2 is a rigid element, it will cause the two new structures
(Figure 7.4(b) and (c)) having the same amount of displacement.
7-4
Chapter 7: A basic mechanical model for the retrofitted walls
0
E
0 0
E
From Figure 7.5, it shows that force and displacement relationships in elastic analysis
for two simplified trusses.
7-5
Chapter 7: A basic mechanical model for the retrofitted walls
cos 0
M=Acose
4E5
=
p5
F4 144 P4
LI
(a) Masonry wall (b) cable
Figure 7.5 Force and displacement relationship in elastic analysis for two trusses
A4=A5 (7-5)
F4L4 = F5L5
(7-6)
A4E4 A5E5
From which
F5L5 A4E4
F4= (7-7)
A5E5 L4
Substituting equation (7-7) into equation (7-4) and solving for F5, we find
7-6
Chapter 7: A basic mechanical model for the retrofitted walls
The theory of plasticity is adopted to estimate the strength of the retrofitted low-rise
walls. This theorem is based on the following assumptions:
analysis.
Eventually, the masonry (F5) element will reach the yield stress This occurs when
the force F5 become equal to A5. With further increase in the load V, the force
F4 (cable element) will increase also, but the force F5 (masonry element) remains
constant because the masonry (F5) element has become plastic. Finally, the cable
element (F4) will also become plastic, and then the structure can not carry any
additional load. This load is called the ultimate load Vi,.
When the masonry (F5) element yields, the corresponding value of the load V is
A4E4
= A5 (1+ x ) cos 8 (7-9)
A5E5
7-7
Chapter 7: A basic mechanical model for the retrofitted walls
When the cable element (F4) also yields, the corresponding value of the load V is
called the ultimate load obtained by setting F4 equal to A4:
A5)cosO (7-10)
Upon reaching this force level, a complete plastic collapse mechanism forms. The
ultimate strength of the retrofitted wall is given by equation (7-10). The resulting
load-deflection envelope for a retrofitted wall can be constructed using these values.
Note that gravity loads were not considered in the above analysis. If the cables are
added as part of a seismic retrofitting scheme, the gravity loads would already be
resisted by the existing masonry wall. This can be accounted for in the truss model by
reducing the compressive strength of Members 5 by an amount corresponding to
gravity-induced stresses. If the gravity loads are applied after the cables are installed,
as was the case in this experimental program, both the existing masonry wall and the
new cable systems will resist the entire gravity load together. A correction can be
made in the above procedure.
The force-displacement relationships from the simplified truss model described in this
chapter have been compared with experimentally obtained lateral force-horizontal
displacement relationships of the retrofitted wall.
V=(a4A4+a5A5)cosO (7-11)
7-8
Chapter 7: A basic mechanical model for the retrofitted walls
a4 =E4e4 (7-12)
And
v= A4E4 A5E5
A4cosO+ A5cos8 (7-14)
L4 L5
Based on the experimental specimens, the following parameters have been adopted
into this equation.
For cable,
For masonry, the stress-strain relationship which was shown in Figure 7.3 has been
simplified to a bi-linear curve (figure 7.6) based on the discussion in Section 7.2.
7-9
Chapter 7: A basic mechanical model for the retrofitted walls
/ L
A5 = 10 = 3231 mm2 (one-thirty two of the wall length X width of the wall)
940
L5 = cos45 = 1330 mm (masonry diagonal length)
These results are presented in Figs 7.7 and 7.8 for the whole wall and force carried by
cable, respectively. The model provides good correlations with test data and can be
considered as effective analytical tools. In these figures, cable force has been
considered 30% loss when the force transfers from the masonry to cable through the
steel plate connections.
7-10
__
35
z
0 2 4 6 8 10 12 14
Displacement (mm)
25
20 --- - - - -.
z
0
0 15 . - - - -. - -
10
-- - -- Experiment
- . - -. -
— Simple truss model
5 ---.
0 2 4 6 8 10 12 14
Displacement (mm)
7-11
Chapter 7: A basic mechanical model for the retrofitted walls
The effective width of the masonry considered in calculating the strength and stiffness
of compression strut was taken equal to one-thirty two of the length of the wall for
diagonal element. For the effective width of diagonal struts, a range of values were
recommended by other researchers in studies of other types of structures. For
example, Drysdale et al (1994) recommended, diagonal strut widths equal to one-
tenth to one-third of the diagonal length of the wall for truss models used in analysis
of infihled frames. The New Zealand Code specifies a diagonal strut width equal to
one quarter of infill length. Paulay (1992) proposed a width equal to one-quarter of
the diagonal length for design of infihled frames. A parameter study was conducted in
this study to determine a reasonable diagonal strut width for use in truss models. The
results showed that the resultant force displacement curve can be matched well with
the experimental results when the diagonal strut was equal to one-thirty two of the
wall length. Therefore, one-thirty two of the length of the wall has been adopted in
this study.
7.5 Conclusions
A basic mechanical model was developed to predict strength and ductility of low-rise
unreinforced masonry walls retrofitted with cable system. A step by step calculation
procedure is also provided to conduct inelastic analysis and design of the retrofitted
wall with the model described. It was found that this simple truss model captured
accurately the lateral force-horizontal displacement relationships for the retrofitted
wall and the cable. The model provided analytical results that are in good agreement
with those obtained experimentally. The model described in this chapter allows the
formulation of a simple analysis calculation that engineers may employ in retrofitting
existing low-rise walls.
7-12
Chapter 8: Analytical study for Retrofitted walls
CHAPTER 8
ANALYTICAL STUDY FOR RETROFITTED WALLS
8.1 Introduction
In this chapter, two nonlinear finite element models have been developed to exam the
experimental results. One model is developed for unreinforced masonry walls
retrofitted by cable system and the other model is developed for unreinforced masonry
walls retrofitted by FRP. It has been found that both models agree reasonably well
with the experimental results. The models take into account the material nonlinearities
as well as damage due to progressive cracking. Compressive behaviour of masonry is
modelled using the theory of plasticity and cracking is modelled using smeared crack
model. The models are generated using ABAQUS finite element program. The results
of the numerical modelling are also discussed in this chapter. It is shown that the
numerical models are capable of predicting not only the load carrying capacity, but
also ductility of the retrofitted masonry wall. The main objective of this research is to
prove a new retrofitting technique with experimental testing. These models are for
limit-state examination purpose.
Masonry is a composite material. It consists of bricks and mortar joints. The macro-
modeling does not make a distinction between individual units and joints but treats
masonry as a homogeneous anisotropic continuum. Masonry can be assumed to be a
homogeneous material if a relation between average stresses and strains in the
composite material is established. Errors are induced by this assumption due to mortar
8-1
Chapter 8: Analytical study for Retrofitted walls
joints acting as a plane of weakness. For the numerical analysis, bilinear plane stress
continuum elements with full gauss integration were utilized.
8.3 The finite element model for the retrofitted wall by cable system
In this section, a nonlinear finite element model has been developed to validate the
experimental results from unreinforced masonry walls retrofitted by cable system.
Reasonably good agreement has been found between the analytical and experimental
results. The model takes into account material nonlinearities as well as damage due to
progressive cracking. Behaviour of the masonry is modelled using the theory of
plasticity and cracking is modelled using smear cracking approach. The model is
generated using ABAQUS finite element program. The validity of the model is
established by comparison with experimental results. It is shown that the numerical
model is capable of predicting the load carrying capacity of the URM wall and the
retrofitted walls.
8-2
Chapter 8: Analytical study for Retrofitted walls
The selection of element for masonry walls, steel plates and cables are discussed in
this section.
8.3.1.1 Masonry
1 2
x
4node reduced
element
Figure 8.1 The geometry and node locations for CPS4R (ABAQUS 2000)
A more sophisticated element, such as the eight-node isoparametric element, was not
used because it has been shown from previous research (Bazant and Cedolin, 1980;
Au, 1987) that the use of higher order elements was not warranted for the analysis of
brick masonry (where nonlinearity is mainly due to progressive cracking and not non-
linear material characteristics), provided a relatively fine element mesh was adopted.
The four-node doubly curved general-purpose shell element, S4R, was used to model
the steel plates that used to connect between the cable and masonry wall. The shell
8-3
Chapter 8: Analytical study for Retrofitted walls
element has four nodes with six degrees of freedom at each node-translations and -
rotations in the nodal x, y and z directions. The geometry and node locations for this
element type are shown in Figure 8.2
Y
a
x
1 2
4-node reduced
integration
Figure 8.2 The geometry and node locations for S4R (ABAQUS 2000)
8.3.1.3 Cable
The two-dimensional truss element, T2D2, was used to model the cable. The 2-node
straight truss element uses linear interpolation for position and displacement and has a
constant stress. The truss element is long, slender structural member that can transmit
only axial force. No compressive option was used to make sure the cable only takes
tension. The element has two degrees of freedom at each node and translations in the
nodal x and y directions. The geometry, node locations, and the coordinate system are
shown in Figure 8.3.
2
2
Y
1
x end 1
2- oleniertt
Figure 8.3 The geometry and node locations for T2D2 (ABAQUS 2000)
8-4
Chapter 8: Analytical study for Retrofitted walls
The connector element, connection type BEAM, was used to model connection
between the steel plate and masonry wall. The geometry, node locations, and the
coordinate system are shown in Figure 8.4.
The connector element, connection type BEAM, was also used to model the
connection between the steel plate and cable.
2
1
b
a
I
e"3
Figure 8.4 The geometry and node locations for connector element (BEAM)
(ABAQUS 2000)
8-5
Chapter 8: Analytical study for Retrofitted walls
masonry, ABAQUS requires input data for material properties as follow (please
see details in Appendix A):
8-6.
Chapter 8: Analytical study for Retrofitted walls
The model is capable of predicting failure for masonry materials. Both cracking and
crushing failure modes are accounted for. The two input strength parameters, i.e.,
ultimate uniaxial tensile and compressive strengths, are needed to define a failure
surface for the masonry. Consequently, a criterion for failure of the masonry due to a
multiaxial stress state can be calculated.
A typical two-dimensional failure surface for masonry that is adopted in this study is
shown in Figure 8.6. In masonry, cracking occurs when the principal tensile stress in
any direction lies outside the failure surface. After cracking, the elastic modulus of the
masonry element is set to zero in the direction parallel to the principal tensile stress
direction. Crushing occurs when all principal stresses are compressive and lie outside
the failure surface; subsequently, the elastic modulus is set to zero in all directions
(ABAQUS, 2002), and the element effectively disappears.
biaxial
tension
8-7
Chapter 8: Analytical study for Retrofitted walls
Steel plates were added at the connection between cable and masonry wall in this
finite element model. Properties such as elastic modulus and yield stress, for the steel
plate used in this FEM study follow the design material properties used for the
experimental investigation. The steel plate for the finite element models was assumed
to be a linear elastic material. Material properties for the steel plate for all models are
as follows:
8.3.2.4 Cable
The type of cable used for this experiment was Ronstan typical grade 316 stainless
steel wire rope (19 single strands, diameter 10mm, and breaking load 71 KN) and
ending for cable was seafast threaded swage terminals (RF1513M1O1O). Material
properties for the cable for all models are as follows:
Breaking load = 71 KN
Poisson's ratio,v =0.3
8.3.3 Geometry
The dimensions of the walls were 940mm X 940mm X 110mm. Figure 8.7 illustrates
typical dimensions for all walls before cable reinforcing. Figure 8.8 shows the finite
element model for the URM wall before cable reinforcing. Figure 8.9 shows the
actual wall retrofitted with cable. Figure 8.10 shows the finite element model for the
8-8
Chapter 8: Analytical study for Retrofitted walls
retrofitted wail with cable. Figure 8.11 shows the details of connection modelling
(section A-A).
i11, (-
- II
F"''' II II
Li II II
I
C'
LIII
II
II
II
II
II
II It II
940mm (4 courses)
I I ii I I I I I I I I I I I
Figure 8.8 The finite element model for the URM wall before cable reinforcing
8-9
Chapter 8: Analytical study for Retrofitted walls
I
Section A:A
Figure 8.10 The finite element model for the retrofitted wall with cable
8-10
Chapter 8: Analytical study for Retrofitted walls
Cable
constraint
bolt
Masonry
In Figure 8.11, it shows the masonry wall and steel plate are connected by four
BEAM connector elements. There are 1 mm between the masonry wall and steel
plate. It also shows the steel plate and cable are connected by a BEAM connector
element which is 1mm long.
All walls were tested under combined constant gravity load and incrementally
increased in-plane lateral displacement reversals, as shown in Figure 8.12. The finite
element models were loaded and constrained at the same locations as the walls, but
the models were only loaded monotonically.
8-11
Chapter 8: Analytical study for Retrofitted walls
Roof load
Lateral
Displacement
Fix to
strong floor
Figure 8.12 Loading and boundary condition of the retrofitted wall model
As an initial step, a finite element analysis requires meshing of the model. In other
words, the model is divided into a number of small elements, and after loading, stress
and strain are calculated at integration points of these small elements (Bathe 1996).
An important step in finite element modelling is the selection of the mesh density. A
convergence of results is obtained when an adequate number of elements is used in a
model. This is practically achieved when an increase in the mesh density has a
negligible effect on the results (Adams and Askenazi 1998) after the convergence
study. Figure 8.13 shows meshes for the retrofitted wall model.
8-12
Chapter 8: Analytical study for Retrofitted walls
8.4.1.1 Masonry
For this macro-modelling, the same four-node bilinear two-dimensional plane stress
element, CPS4R, was used to model the masonry. Please see the section 8.3.1.1. for
details.
8.4.1.2 FRP
The membrane element, M3D4R, was used to model the FRP. This type of element is
used to represent thin surfaces that offer strength in the plane of the element but have
no bending stiffness. The element has three degrees of freedom at each node and
translations in the nodal x, y and z directions. The geometry and node locations for
this element type are shown in Figure 8.14.
Y 3
x
4- node reduced
integration element
Figure 8.14 The geometry and node locations for M3D4R (ABAQUS 2000)
A surface-based Tie constraint was used to model connection between FRP and
masonry wall. This constraint type, TIE, makes two surfaces to be tied together. The
TIE constraint can make the translational and rotational motion as well as all other
active degree of freedom equal for a pair of surfaces. The surface-based tie algorithms
are shown in Figure 8.15.
8-14
Chapter 8: Analytical study for Retrofitted walls
8.4.2.1 Masonry
The same masonry model was used in this model. Please see the section 8.3.2.1 for
details.
8.4.2.2. FRP
The type of FRP used for this model was Sika CarboDur S (Width 30mm, thickness
1.2mm and tensile strength 2800 MPa) and epoxy adhesive for bonding Sika
CarboDur Strips was Sikadur-30. Material properties for FRP for all models are as
follows:
8-15
Chapter 8: Analytical study for Retrofitted walls
8.4.3 Geometry
The dimensions of the walls were 940mm X 940mm X 110mm. Figure 8.16 shows
the actual wall retrofitted with FRP. Figure 8.17 shows the finite element model for
the retrofitted wall with FRP.
H±il+HikJXW
Figure 8.17 The finite element model for the retrofitted wall with FRP
8-16
Chapter 8: Analytical study for Retrofitted walls
All walls were tested under combined constant gravity load and incrementally
increased in-plane lateral displacement reversals, as shown in Figure 8.18. The finite
element models were loaded and constrained at the same locations as the walls, but
the models were only loaded monotonically.
Roof load
Lateral
Displacement
Fix to
strong floor
Figure 8.18 Loading and boundary condition of the retrofitted wall model
8-17
Chapter 8: Analytical study for Retrofitted walls
There are two types of nonlinearities which need to be considered. One is "material
nonlinearity" which is caused by the nonlinear stress-strain relation or progressive
local failure, the other is "geometric nonlinearity" caused mainly by the large
displacement of the structure. Only material nonlinearity is considered in this
research, as geometric nonlinearity is not applicable to unreinforced masonry under
in-plane lateral load since only small deflections occurred.
8-18
Chapter 8: Analytical study for Retrofitted walls
The finite element models generated in ABAQUS are usually nonlinear and can
involve from a few to thousands of variables. In terms of these variables the
equilibrium equations obtained by discretizing the virtual work equation can be
written symbolically as
(8.1)
where FN is the force component conjugate to the variable in the problem and
tiM is the value of the Mth variable. The basic problem is to solve Equation 8.1 for•
the tiM throughout the history of interest.
In nonlinear analysis, the total load applied to a finite element model is divided into a
series of load increments called load steps. At the completion of each incremental
solution, the stiffness matrix of the model is adjusted to reflect nonlinear changes in
structural stiffness before proceeding to the next load increment.
The basic formalism of Newton's method is as follows. Assume that, after iteration i,
an approximation to the solution has been obtained. Let be the difference
between this solution and the exact solution to the discrete equilibrium equation
Equation 8.1. This means that
(8.2)
8-19
Chapter 8: Analytical study for Retrofitted walls
Expanding the left-hand side of this equation in a Taylor series about the approximate
solution u7' then gives
+ +... = 0 (8.3)
= FN (8.4)
where
= aFN
KNP
F1N =F NI M
(8.5)
and all entries in are sufficiently small. Both these criteria were checked by
default in an ABAQUS/Standard solution. ABAQUS/Standard also records peak
values in the force residuals, incremental displacements, and corrections to the
incremental displacements at each iteration.
8-20
Chapter 8: Analytical study for Retrofitted walls
In this study, Newton's method is adopted to solve nonlinear problem. For the
masonry elements, convergence criteria were based on force and displacement, and
the convergence tolerance limits were initially selected by the ABAQUS program.
8.7 Validation of the numerical modelling results with the test results
This section compares the results from ABAQUS finite element analyses with
experimental results. The following comparisons are made: load-deflection plots; first
cracking loads; loads at failure and crack patterns at failure. The development of crack
patterns for each wall and summaries of the maximum force occurring in the cable for
the finite element models are also discussed. The data from the finite element analyses
were collected at the same locations as the experimental tests for the masonry wall.
Figure 8.20 shows that the load-deflection plot from the finite element analysis agrees
well with the experimental data for the URM wall. The load-deflection plot from the
finite element analysis is slightly stronger than that from the experimental results.
This is possibly due to the relative homogeneity of the finite element models when
compared to the relative non-homogeneity of the actual walls that contain two
different materials and neglect the weaker mortar plane effect. The first cracking load
8-21
Chapter 8: Analytical study for Retrofitted walls
for the finite element analysis is 15.2 KN, which is higher than the load of 14.7 KN
from the experimental results by only 4%. Lastly, the final load of 24.4 KN from the
model is higher than the ultimate load of 23.8 KN from the experimental data by only
3 %.
30
0
-J
G)
210
U)
0
0 1 2 3 4 5 6
Displacement (mm)
Figure 8.21 shows that the load-deflection plot from the finite element analysis agrees
well with the experimental data for the retrofitted wall with cable. The load-deflection
plot from the finite element analysis is stiffer than that from the experimental results.
This is possibly due to the fatigue of the material. The result from the finite element
model is for the monotonic loading, but the results from the experiment are a
consequence of cyclic loading. The first cracking load for the finite element analysis
is 17.5 KN, which is higher than the average first cracking load of 15.2 KN from the
experimental results by 15%. Lastly, the final load of 49.15KN from the model is
higher than the average ultimate load of 46.4 KN from the experimental data by 6 %.
8-22
Chapter 8: Analytical study for Retrofitted walls
50
40
z
0
0
0 2 4 6 8 10 12
Displacement (mm)
Figure 8.22 shows the finite element results of load-deflection curve for the URM
wall before and after retrofitting with cable. The load-deflection curve for the
retrofitted wall is much stiffer than the load-deflection curve for the unretrofitted wall.
In this figure, it also shows that the retrofitted wall significantly increases the strength
and ductility. These results are similar to the experimental results. The improvement
of the ultimate lateral load resistance of the retrofitted walls with cables is 2 times the
capacity of unreinforced wall. The improvement of the, ductility of the retrofitted
walls with cables is about six times the capacity of unreinforced wall.
8-23
Chapter 8: Analytical study for Retrofitted walls
50
40
z 30
0
a)
V 20
Cl)
UR R
10
RETROFITTING)
If
0
0 2 4 6 8 10 12
Displacement (mm)
Figure 8.22 F.E.M. results of load-deflection plot for URM wall before and after
retrofitting
The first cracking load from the finite element analysis is the load step where the first
signs of cracking occur for masonry in the model. Loads at first cracking from the
model and the experimental results are compared in Table 8.1.
Table 8.1 Comparisons between experimental and ABAQUS first cracking loads
The first cracking loads from the finite element analyses and the experimental data are
within 15% for URM wall and retrofitted walls. In all cases, the first cracking load
from ABAQUS is higher than that from the experimental data. This is possibly due to
8-24
Chapter 8: Analytical study for Retrofitted walls
the relative homogeneity of the finite element models when compared to the relative
heterogeneity of the actual walls that contain a number of microcracks.
Table 8.2 compares the ultimate loads for the URM wall and retrofitted walls from
experiments and from the finite element simulations.
For the actual retrofitted walls, there was no evidence that the cable failed before
overall failure of the walls. This is confirmed by the finite element analyses. In figure
8.23, the forces carried by the cable from ABAQUS are compared with the tensile
strengths for the cable measured from experimental tests. In this figure, it shows 2KN
carried by the cable at the beginning stage due to the pretension of the
cable.
w 30
25
w
— 20
0)
I..
CD
0
5
CD
0
-J
0 2 4 6 8 10 12
Displacement (mm)
8-25
Chapter 8: Analytical study for Retrofitted walls
Figure 8.24 shows that the load-deflection plot from the finite element analysis agrees
well with the experimental data for the retrofitted wall with FRP. The load-deflection
plot from the finite element analysis is stiffer than that from the experimental results.
This is possibly due to the fatigue of the material. The result from the finite element
model is for the monotonic loading, but the results from the experiment are a
consequence of cyclic loading. The first cracking load for the finite element analysis
is 25.3 KN, which is higher than the average first cracking load of 23.7 KN from the
experimental results by 6%. Lastly, the final load of 65.0 KN from the model is higher
than the average ultimate load of 60.2 KN from the experimental data by 7 %.
Wall (FRP)
70
60
,4o
20
10
0
0 2 4 6 8 10 12 14
Displacement (nun)
8-26
Chapter 8: Analytical study for Retrofitted walls
Figure 8.25 shows the finite element results of load-deflection curve for the URM
wall before and after retrofitting with FRP. The load-deflection curve for the
retrofitted wall is much stiffer than the load-deflection curve for the unretrofitted wall.
In this figure, it also shows that the retrofitted wall significantly increases the strength
and ductility. These results are similar to the experimental results. The improvement
of the ultimate lateral load resistance of the retrofitted walls with FRP is 3 times the
capacity of unreinforced wall. The improvement of the ductility of the retrofitted
walls with cables is about six times the capacity of unreinforced wall.
70
10
0- I
0 2 4 6 8 10 12 14
Displacement (mm)
Figure 8.25 F.E.M. results of load-deflection plot for URM wall before and after
retrofitting
8-27
Chapter 8: Analytical study for Retrofitted walls
The first cracking load from the finite element analysis is the load step where the first
signs of cracking occur for masonry in the model. Loads at first cracking from the
model and the experimental results are compared in Table 8.3.
Table 8.3 Comparisons between experimental and ABAQUS first cracking loads
The first cracking loads from the finite element analyses and the experimental data are
6% for URM wall and retrofitted walls. In all cases, the first cracking load from
ABAQUS is higher than that from the experimental data. This is possibly due to the
relative homogeneity of the finite element models when compared to the relative
heterogeneity of the actual walls that contain a number of microcracks.
Table 8.4 compares the ultimate loads for the URM wall and retrofitted walls from
experiments and from the finite element simulations.
8-28
Chapter 8: Analytical study for Retrofitted walls
In figure 8.26, the forces carried by FRP from ABAQUS are compared with the
tensile strengths for FRP measured from experimental tests. From this figure, it shows
that the forces carried by FRP are linear from ABAQUS and are nonlinear from the
experiments. The reason for that is due to FRP debonding during the experiment. The
current model is neglecting this effect.
35
30
20
15
n 10
0
0 1 2 3 4 5 6 7 8 9 10
Displacement (mm)
8-29
Chapter 8: Analytical study for Retrofitted walls
Figure 8.27 shows the finite element results of load-deflection curve for the retrofitted
walls by cable system and FRP. The load-deflection curve for the retrofitted wall by
FRP is which is agreed well with the testing stiffer than the load-deflection curve for
the retrofitted wall by cable system.
70
60
50
30
20
10
0
0 2 4 6 8 10 12 14
Dispalcement (mm)
Figure 8.27 F.E.M. results of load-deflection plot for the retrofitted walls by cable
system and FRP
8-30
Chapter 8: Analytical study for Retrofitted walls
The first cracking load from the finite element analysis is the load step where the first
signs of cracking occur for masonry in the model. Loads at first cracking from the
models for the retrofitted wall by cable system and by FRP are compared in Table 8.5.
Table 8.5 Comparisons between the retrofitted walls by cable system and FRP
for first cracking loads
The first cracking loads for the retrofitted wall by FRP are much higher than the first
cracking loads for the retrofitted wall by cable system. These results are similar as the
experimental results.
Table 8.6 compares the ultimate loads for the retrofitted walls by cable system and
FRP from experiments and from the finite element simulations.
Table 8.6 Comparisons between the retrofitted walls by cable system and FRP
for ultimate loads
The ultimate load for the retrofitted wall by FRP is much higher than the ultimate
loads for the retrofitted wall by cable system. These results are similar as the
experimental results.
8-31
Chapter 8: Analytical study for Retrofitted walls
Figure 8.28 compares the forces carried by the cable and by FRP from the finite
element simulations. The forces carried by FRP are much higher than the forces
carried by cable system. These results are similar as the experimental results
35
Displacement (mm)
Figure8.28 Comparing the forces carried by the cable and by FRP from ABAQUS
8.9 Conclusion
The main objective of this research is to prove a new retrofitting technique with
experimental testing. These models are for limit-state examination purpose only. Non-
linear finite element models were developed to predict strength and ductility of URM
wall, URM walls retrofitted with cable system and walls retrofitted with FRP. The
experimental results obtained from URM masonry walls retrofitted with the cable
system and FRP are compared with those obtained from analytical solutions. The
analytical results were obtained using the finite element program ABAQUS. The
results from these models also show good agreement with the experimental results.
8-32
Chapter 8: Analytical study for Retrofitted walls
The limitations of this model are that this current model couldn't be predicted well for
the retrofitted walls and couldn't model the FRP debonding effect. Due to the
limitations of ABAQUS, the model makes no attempt to include prediction of cyclic
response or of the reduction in the elastic stiffness caused by inelastic straining.
8-33
Chapter 9: Performance-Based Design for retrofitted wall
CHAPTER 9
PERFOEMANCE-BASED DESIGN
FOR RETROFFITED WALLS
9.1 Introduction
Experience in recent California earthquakes, including the 1999 Loma Prieta and 1994
Northridge events, indicates that modern code provisions in zones of high seismicity are
relatively reliable in avoiding life-threatening building damage. Both events, having large
magnitudes and epicentral locations close to large population centers, resulted in fewer
than 100 fatalities. However, the economic losses associated with these earthquakes, $7
billion for Loma Prieta and more than $15 billion for Northridge, were unacceptably
large, given that events of this size can be experienced relatively frequently in zones of
high seismicity (Hamburger, 1996). As a result, public officials and members of the
insurance and financial industries are calling for building construction practices that can
better limit damage in future earthquakes.
9-1
Chapter 9: Performance-Based Design for retrofitted wall
The performance-based design is used to design the retrofitted wall. The concept of
performance-based design and the performance-based design procedure are presented in
this Chapter. A numerical design example, using this performance-based design and
Australian Standard, is also presented.
9-2
Chapter 9: Performance-Based Design for retrofitted wall
DESIGNATION DESCRIPTION
9-3
Chapter 9: Performance-Based Design for retrofitted wall
94
Chapter 9: Performance-Based Design for retrofitted wall
The primary objective of current building code provisions for earthquake resistive design
is to avoid building collapse and loss of life. Recently, as the reliability of design and
construction practice in meeting this objective has improved, demand for building
designs capable of meeting other objectives, including minimization of damage,
continuous occupancy, and business interruption, has increased. Programs for the
development of performance-based design procedures have been initiated to address this
concern. However, significant barriers must be overcome to successfully implement these
concepts. Performance-based earthquake engineering implies design, evaluation, and
construction of engineered facilities whose performance under common and extreme
loads responds to the diverse needs and objectives of owner-users and society. PBEE is
based on the premise that performance can be predicted and evaluated with sufficient
confidence for the engineer and client jointly to make intelligent and informed decisions
based on building life-cycle considerations rather than on construction costs alone
(Hamburger, 1996).
In the United States, the new. generation of seismic codes for bridges is moving towards
performance-based design as a rational and cost-effective approach to seismic design.
Current seismic codes are based on accelerations and forces, which result in stronger and
more rigid structures, which in turn attract higher forces. Instead, the performance-based
design approach uses displacements and deformations along with the proper detailing to
allow the deformations to develop, permitting structures to be somewhat more flexible. In
addition, the necessary details to meet given performance criteria usually cost less to
construct than those required to meet the criteria of force-based approaches — the money
is spent where it is needed (Jackson, 2001).
9-5
Chapter 9: Performance-Based Design for retrofitted wall
objectives for each building. For such a choice to be useful, it is necessary to define the
various behavior alternatives in terms meaningful to the layperson. Important parameters
include: the potential for loss of life, the cost of repairing sustained damage, and the
amount of time the building is out of service for repair or, in extreme cases, replacement.
While these parameters are meaningful to the public, and therefore can serve as a basis
for selecting among building performance alternatives, they are not useful as a basis for
design. An engineer cannot design for such performance specifications as a business
interruption of two weeks or a repair cost that is 20% of replacement value. Therefore, as
a prerequisite to practical implementation of performance-based engineering,
corresponding relationships must be established among parameters that are meaningful to
building users and design professionals. Since a building can experience a wide spectrum
of potential behavior states ranging from a complete absence of damage and earthquake
effects to complete collapse, establishing corresponding relationships is not a trivial task.
9-6
Chapter 9: Performance-Based Design for retrofitted wall
higher performance
less loss
Operational Level 4
Backup 'utility services
'maintain functions; very little
damage. (Si ÷NA)
Several conceptual frameworks for PBEE have been developed in recent professional
efforts (SEAOC Vision 2000, FEMA 273, ATC-40). They differ in details but not in
concepts. The Figure 9.2 illustrates a global framework that identifies processes,
concepts, and major issues.
9-7
Chapter 9: Performance-Based Design for retrofitted wall
The SEAOC's Vision 2000 document recommends that buildings be constructed, based
on their intended occupancies and uses, to meet the performance objectives indicated in
Figure 9.3. Each diagonal line in the figure indicates design performance levels and
earthquakes recommended for the design of buildings of different occupancies and uses.
Individual, informed building users, of course, could select more stringent performance
objectives, if desired. In the figure, each combination of an earthquake return period and
performance level, indicated by a red diamond, represents a specific design performance
objective. The intent is that ordinary buildings provide a low risk that life be endangered
as a result of the performance of the building in any earthquake likely to effect it; and that
for frequent earthquakes, the building user not be burdened with extensive repairs or loss
9-8
Chapter 9: Performance-Based Design for retrofitted wall
of use; that buildings required for emergency response and essential public function have
a low risk of being damaged beyond a level that would permit their use, and; that
facilities housing systems and materials that would pose a hazard to many persons if
released, have a low risk of damage resulting in such release.
(50% 50
w
C- (20% - :5Q
w
=
U)
Because of lack of proper reinforcement in the past, older facilities don't meet newer
standards, and this is where retrofit comes into the picture. To start a Performance-based
Seismic Rehabilitation Program, owners need to review their inventory and understand
which facilities may be at risk. If there is risk, the owner must prioritize and determine
9-9
Chapter 9: Performance-Based Design for retrofitted wall
which facilities are most important in order to schedule the order of retrofits and ensure
that those structures will not suffer damage in the future.
It is important to note that any step taken improves the reliability index. It is the owner's
choice how far to go, depending on the level of risk, the performance level selected, and
the amount to be spent.
The performance-based seismic rehabilitation design procedure (see Figure 9.4) includes
the steps indicated below (NEHRP, 2000).
9-10
Chapter 9: Performance-Based Design for retrofitted wall
I
• Size elements, components, and connections
(Chapters 2,5 through 9. and 11)
I
• Review for economic acceptability
I
6,IA Section 1.3.6i:Recfesign 6.2A Section 1.3.6.2: Prepare 6.18 Seclion1.3.6.1: Redesign 6.28 Section 1.3.6.2: Prepare
Unacceptable Construction Documents of Unacceptable Construction Documents
Rehabilitation Acceptable Rtihabiiilation Rehabilitation olAcceptable
Rehabilitation
Return to 4A to • Develop construction Return to4B to revise
reconsider documents analysis and or to • Develop construction
Rehabittation Objective • Segin rehabilitation 2 to reconsider documents
I
or to SA to revise Rehabilitation Ob(ective • Begin rehabilitation
• Exercise quality control
corrective measures • Exercise quality
9-11
Chapter 9: Performance-Based Design for retrofitted wall
(For more details about equation 9-1 and Type 3 curve, please refer to Appendix C.5)
QI
The practical design guideline has been provided in this section. It involves the following
steps.
9-12
Chapter 9: Performance-Based Design for retrofitted wall
Step2 Determine the lateral seismic forces using a performance-based design code, such
as the FEMA 273.
Step3 Use Eqs. 9.2 and 9.3 to find lower bound lateral strength of wall (original
unretrofitting capacity).
QCL = (9-2)
=
QCL =
-
(For more details about equation 9-2 and 9-3, please refer to Appendix C.5)
Step 6 Check the Drift requirement in Table 9.2. If not safe, the design must be
modified. (For more details about Table 9.2, please refer to Appendix C.2)
Step 7 Design the connections between cable and the wall. (Appendix C-2)
Some useful practical information has been provided in Appendix C including cable
system, chemical anchor and FRP.
9-13
Chapter 9: Performance-Based Design for retrofitted wall
Table 9.2 Structural Performance level for URM walls (FEMA-274, 1997)
A sample design of the cable system developed in this research program is presented in
this section. The design problem involved the single-story building shown in Figure 9.6
and Figure 9.7. The wall is to be retrofitted using the cable system, as shown in Figure
9.8. The building is 9 m x 9 m in plan, and has 3m clear height. It is enclosed by four
unreinforced end walls in each direction. The length of each wall is 3 m. A 300 mm thick
reinforced (0.25%) concrete slab was used at the roof, for parking. The walls are 110mm
thick, each. The building is assumed to be in site class A (hard rock with average shear
wave velocity, v5 >5.000ftlsec) and Earthquake Hazard level is assumed using values of
=0.25 and S1 =0.1 taken from MCE spectral response acceleration contour maps
(FEMA356).The life safety performance level is selected for the target building
performance level and the basic safety objective is selected for the rehabilitation
objective for this building.
9-14
Chapter 9: Performance-Based Design for retrofitted wall
j3m
*
9m
9-15
Chapter 9: Performance-Based Design for retrofitted wall
j3m
9m
The equivalent static seismic base shear, V, in the East-West direction must be
determined first. This requires the computation of weight, W. The weight of the roof is
computed as:
W1= yxV=24x(O.3x9x9)=593KN
(i = 24 KN/m3, specific weight for concrete)
W2 =yxV=19x(!x3x3xO.11x9)=75KN
9-16
Chapter 9: Performance-Based Design for retrofitted wall
W3=0.2x9x9x2=32KN
height to the roof level and = 0.02, 13 = 0.75 for URM buildings. (Please refer to
0.75
T = 0.02 x 3 = 0.046 second
Determine the lateral seismic forces from Equation 9-4 (Please refer to Appendix C.4.1
for details)
V=C1C2C3CmSaW (9-4)
C1 =1.5 (T<0.lOsecond.)
9-17
Chapter 9: Performance-Based Design for retrofitted wall
where Fa =0.9 and are site coefficients determined respectively from Appendix
C.4.1, Tables C-13 and C-14.
where
=0.25 and S1 =0.1 taken from MCE 10%5oyears spectral response acceleration
contour maps (FEMA356)
and where and B1=1 for 5% damping (form Appendix C.4.1, Table C-b)
for 0<T<T0
0.046
= — = 0.2 x — = 0.15 (9-9)
+0.41
V U
4
9-18
Chapter 9: Performance-Based Design for retrofitted wall
KQCL QUF
U0
K 0.75
The retrofitted cable configuration is 450 from the horizontal level so that the minimum
needed breaking load for cable is:
F=
B =59KN
cos45°
Using this value to choose the suitable size of cable from the cable company catalogue.
Then the connections between the wall and cable should be designed to make sure the
cable is broken before the connections fail.
The total horizontal earthquake base shear (V) in the East-West direction shall be
determined from the following equation
V= (9-10)
9-19
Chapter 9: Performance-Based Design for retrofitted wall
R = structural response factor = 2.75 (The selection of the structural response factor, R,
requires an explanation because this type of the walls is not included in the structural
response factor table of AS 1170.4. Observations and results of this experimental study
show that walls retrofitted by means of cables are ductile structural members with
hysteresis curves representative of element between concentrically-braced frames ( R =
4.0) and unreinforced masonry shear walls (R = 1.5). Therefore, an average value of R,
2.75 was chosen.)
1.25a 1.25*0.15
C = earthquake design coefficient = 2 = 2
=1.17
T3
Gg=G+(PcQ
where
This requires the computation of weight, Gg .The weight of the roof is computed as:
W1= yxV=24x(0.3x9x9)=593KN
(y = 24 KN/m3, specific weight for concrete)
9-20
Chapter 9: Performance-Based Design for retrofitted wall
W2 =yxV= 19x(!x3x3xO.11x9)=75KN
(y = 19 KN/m3, specific weight for masonry)
W3= 0.4 x 9 x9 x 2 = 64 KN
Gg=593 +75+64=732 KN
V=
2.75
V U
4
Knowing that the original (unretrofitted) capacity, V is 10 KN (using the same value
as section 9.5.1)
9-21
Chapter 9: Performance-Based Design for retrofitted wall
The retrofitted cable configuration is 450 from the horizontal level so that the minimum
needed breaking load for cable is:
F=
B
cos45°
=59KN
Using this value to choose the suitable size of cable from the cable company catalogue.
Then the connections between the wall and cable should be design to make sure the cable
is broken before the connections fail.
As indicated in 9.5.1 and 9.5.2, the two designs received the similar results. That is
because Australian Standard is only based on the Basic Safety Objective to design the
buildings to satisfy the Life Safety Performance level and the same objective and
performance level were selected when the performance-based design was used.
The difference between Performance-Based Design and Australian Standard is that one
has choices to decide the rehabilitation objective and target building performance level
when the performance-based design was adopted and such options don't exist in the
current Australian Standard.
If the different rehabilitation objective and target building performance level are selected,
the response spectrum acceleration will be different so that the lateral seismic force will
be different.
Buildings continue to be designed for the minimum levels of life safety protection offered
by the Australian Standard design approach, however, Performance-Based Design offers
9-22
Chapter 9: Performance-Based Design for retrofitted wall
owners/operators the opportunity to limit business interruption economic loss and other
consequences for less severe but more probable earthquake hazards.
Although performance-based earthquake engineering design is a great idea, but there are
some difficulties that need to conquer to implant PBD.
Despite statements to the contrary following damaging earthquakes, the public generally
believes that building codes provide adequate protection against earthquake-induced
losses. The public is generally unaware of the performance standards inherent in the
building codes or the likely behavior of buildings in earthquakes. Therefore, they are
unlikely to discriminate in their selection of a building based on potential earthquake
performance. Thus, there is limited market value added to a building designed and
constructed for such enhanced performance.
The exceptions are those buildings developed specifically for the occupancy of a business
or institution knowledgeable in earthquake risk management. Most large corporations and
institutions today have an in-house risk management group, responsible for protecting the
organization against large catastrophic losses. In recent years, many such organizations
with properties in regions of high seismicity have attempted to implement performance-
9-23
Chapter 9: Performance-Based Design for retrofitted wall
based design concepts in the development of new properties. In some cases, however,
more conventional design approaches were adopted because the additional cost inherent
in the development of buildings with enhanced earthquake performance characteristics
could not be justified by management.
9-24
Chapter 9: Performance-Based Design for retrofitted wall
selected system is less than appropriate for a building, for fear that the architect will find
another structural engineer who will support the original design concept, regardless of
whether or not it will perform adequately.
Even if the necessary design procedures were available today and the engineer could
design structures with predictable and acceptable seismic performance characteristics,
this would not assure that the buildings themselves will meet the intended seismic
performance objectives. Buildings are a complex collection of systems-structural,
mechanical, fire protection, electrical, architectural, and others. The performance of the
individual components of the nonstructural systems can be as important to the overall
earthquake performance of a building as is the performance of the structure, particularly
for enhanced performance objectives that address building operability. Yet the structural
engineer has little participation in the design of these systems for seismic resistance and
is often unaware of any details for installing these systems. Consequently, poor
earthquake performance of these systems is common, even in buildings that have been
designed to very stringent structural standards.
An example of such behavior is the performance of the Olive View Hospital, in Sylmar,
California, during the 1994 Northridge Earthquake (see front and back covers). The Olive
View Hospital replaced a previous facility, destroyed by the 1971 San Fernando
Earthquake. It was designed to criteria that were substantially in excess even of those
enforced by the State of California for construction of hospitals, since the 1971 event.
This building was located within a few kilometers of the epicenter for the Northridge
Earthquake, with recorded accelerations at the roof of the structure exceeding 2g. The
building structure behaved very well and could be said to meet the performance
requirements for the Operational performance level. However, utility piping within the
building failed, resulting in extensive water damage to the facility and forcing its closure.
As a result, the building's performance, as opposed to the structure's performance, was
below the Immediate Occupancy level. While this performance was still excellent, it did
not meet the intent of the design-to provide immediate post-earthquake health-care
service.
9-25
Chapter 10: Summary and Conclusions
CHAPTER 10
SUMMARY AND CONCLUSIONS
10.1 Summary
Many existing unreinforced masonry low-rise shear walls, built before the 1989
Newcastle Earthquake, were not designed to resist earthquakes. As a result, their
seismic behaviour may not be adequate, and may need to be retrofitted. A new
retrofitting procedure has been developed in the current research project. The
retrofitting strategy includes a cable system, designed to raise both strength and
ductility to a level compatible with today's standards.
Cable consists of a number of wires or strands and has high tensile strength, lightness
and high corrosion resistance. These materials can absorb tensile stress and increase
overall element stiffness, ductility and bearing capacity. Using cable system for
seismic retrofitting applications has other advantages such as architectural versatility,
durability, and no loss of valuable space. Furthermore, it does not add a significant
mass to the existing building and can remain unchanging dynamic properties for
structures. Therefore, Cable would be seen to be particularly suitable for seismic
retrofitting materials.
Four unreinforced masonry walls and eight unreinforced masonry walls retrofitted
with cable systems were designed, constructed and tested under combined gravity
load and in-plane deformation reversals. Seven unreinforced masonry walls retrofitted
with FRP were also conducted in this research for the comparison purpose. The
testing wall specimens had a thickness of 110 mm, an aspect ratio of 1.0 and a height
10-1
Chapter 10: Summary and Conclusions
of 1000 mm. All walls were tested under combined constant gravity load and
incrementally increased in-plane lateral displacement reversals.
The cable system consists of diagonal cables (19 single strands, Diameter 10mm),
attached to the walls using through-thickness bolts. Stiff steel angles and anchor bolts
were used to connect the cable to the foundation. The cable was fixed on one side of
the wall only.
The type of FRP used for this experiment was Sika CarboDur S (Width 30mm,
thickness 1.2mm and tensile strength 2800 MPa) and epoxy adhesive for bonding
Sika CarboDur Strips was Sikadur-30. The FRP was fixed for one side of the wall
only.
Analyses of walls were conducted using the nonlinear F.E.M model (ABAQUS). The
results computed using F. E. M. were compared with recorded experimentally.
The comparisons showed good agreement. Furthermore, the performance-based
design procedure was also applied to design URM walls retrofitted with the cable
system.
10.2 Conclusions
Experiments conducted in this study show that the cable system, proposed to retrofit
low-rise masonry walls under in-plane lateral loads, is one of the most effective ways
in significantly increasing the in-plane strength, ductility and energy-dissipation
capacity. For the particular wall specimens considered in this thesis, the addition of
cable system increased the lateral load resistance of the wall by approximately 25 KN,
making the retrofitted walls 2 times stronger that the unreinforced masonry.
Experiments also indicate that the cable carried about the 50% of force acted on the
whole wall after major cracks occurred. The results proved that that URM wall
retrofitted by cable system can effectively increase the load capacity after the onset
cracking of the wall.
10-2
Chapter 10: Summary and Conclusions
For the addition of FRP, the improvement of the ultimate lateral load resistance of the
retrofitted walls is 3 times the capacity of un-reinforced wall. Experiments also
indicate that the FRP carried about the 50% of force acted on the whole wall even
after early loading stage, which is different to cable system. This is because FRP is
fully attached to the wall. The results proved the URM wall retrofitted by FRP system
becomes a whole new structure to carry the force together.
Seismic retrofitting of unreinforced masonry walls with FRP or cable system proved
to be an effective and reliable strengthening alternative. However as discussed earlier,
each retrofitting technique has its own advantages and disadvantages. Although FRP
is more effective in increasing the ultimate load carrying capacity of URM walls than
cable, the high cost and no complete knowledge of its mechanical behaviour may be a
major concern as well as the aesthetic appearance of the building. On the other hand,
although cable system could be a low cost alternative, the fragility of the masonry
does not allow a simple fixing of the cable system.
In this study, the finite element method is adopted for numerical modelling to
simulate the masonry behaviour. Masonry is a composite material that consists of
units and mortar joints. Macro-modelling is adopted to model this composite material
due to the practical purpose.
Two nonlinear F.E.M models were developed to predict the strength and ductility of
the walls by cable system and FRP and validate the experimental results. One model
is developed for unreinforced masonry walls retrofitted by cable system and the other
model is developed for unreinforced masonry walls retrofitted by FRP. It has been
found that the analytical and experimental results are agreed well.
The proposed numerical models are based on ABAQUS concrete model and are
modified to simulate the special characterises of masonry. The models take into
account the material nonlinearities as well as damage due to progressive cracking.
The model employs oriented damaged elasticity (smeared cracking) and isotropic
compressive plasticity to represent the inelastic behaviour of masonry. The model is
intended to model the masonry behaviour for relatively monotonic loadings under
fairly low confining pressures. Due to the limitations of ABAQUS, the model makes
10-3
Chapter 10: Summary and Conclusions
The fixed orthogonal cracks model is adopted so that the maximum number of cracks
at a material point is limited by the number of direct stress components present at that
material point of the finite element model (for example, a maximum of three cracks in
three-dimensional, axisymmetric, and plane strain problems or a maximum of two
cracks in plane stress problems). Once cracks exist at a point, the component forms of
all vector and tensor valued quantities are rotated so that they lie in the local system
defined by the crack orientation vectors (the normals to the crack faces). The model
ensures that these crack face normal vectors are orthogonal so that this local system is
rectangular Cartesian. Crack closing and reopening can take place along the directions
of the crack surface normals. The model neglects any permanent strain associated
with cracking; that is, we assume that the cracks can close completely when the stress
across them becomes compressive.
An isotropically hardening "compressive" yield surface was adopted to model for the
inelastic response when the principal stresses are dominantly compressive. For
tension softening, the bilinear model is adopted in the current model. The effect of
cracked shear retention is also considered in the models.
The ultimate shear strength of the retrofitted wall can be simply obtained by the
F.E.M model. A F.E.M approach, described in this thesis, gives analytical results that
are in good agreement with experimental results. The global lateral force-
displacement behaviour of all retrofitted walls can also be captured with a reasonable
degree of accuracy using the F.E.M. model described in this thesis.
10-4
Chapter 10: Summary and Conclusions
If the different rehabilitation objective and target building performance level are
selected, the response spectrum acceleration will be different so that the lateral
seismic force will be different.
Buildings continue to be designed for the minimum levels of life safety protection
offered by the Australian Standard design approach, but Performance-Based Design
offers owners/operators the opportunity to limit business interruption economic loss
and other consequences for less severe but more probable earthquake hazards.
Although performance-based earthquake engineering design is a great idea, but there
are some difficulties such as the economic and competitive pressures faced both by
the building designer and the building developer that need to conquer to implant PBD.
As the conclusion, the results from both of experiments and numerical modelling
showed that both the strength and ductility of tested specimens were significantly
enhanced with this new technique. Seismic retrofitting of unreinforced masonry walls
with cable system proved to be an effective and reliable strengthening alternative. The
limitations of this model are that this current model couldn't be predicted well for the
retrofitted walls and couldn't model the FRP debonding effect. Due to the limitations
of ABAQUS, the model makes no attempt to include prediction of cyclic response or
of the reduction in the elastic stiffness caused by inelastic straining.
10-5
Chapter 10: Summary and Conclusions
10.3 Reconimendations
• The improvement of the ultimate lateral load resistance of the retrofitted walls
by FRP is higher than cable system, but the wall retrofitted by cable system is
cheaper than the wall retrofitted by FRP.
• The anchor bolts used to attach the cable system to the masonry walls must not
be fastened (torque) to a level that could damage the masonry wall.
10-6
Reference
REFEREN CE
R-1
Reference
Au., S. (1987). "Concentrated loads on solid masonry." PhD Thesis, The University of
Newcastle.
Arya, S. K., ASCE, A. M. and Hegemier, G. A. (1982). "Finite element method for
interface problems." Journal of Structural Division, 108 (ST2), 327-342.
Banthia N and Boyd AJ. (2000). "Sprayed fibre-reinforced polymers for repairs." Canadian
Journal of Civil Engineering, 27 (5), 907-9 15.
Badoux M. and Jirsa J.O. (1990). "Steel bracing of RC frames for seismic retrofitting."
Journal of Structural Engineering, ASCE, 116 (1), 55-74.
Bailey, K. A. (1983). "The analysis of plain, brick masonry under in-plane loads." PhD
Thesis, University of the West Indies.
Benjamin B. (1995). "Reinforced concrete shear for repair of damaged masonry walls."
JABSE Sym. San Francisco, Extending the of structures, 373-378.
R-2
Reference
Beolchini, G. C., Grub, F. and Valente, G. (1997). "Analyses of repair and maintenance of
a masonry wall", Structural studies; repairs, and maintenance of historical buildings, 3,
367-376.
Binda, L., Fontana, A. and Frigerio, G. (1995). "Mechanical behaviour of brick masonries
derived from unit and mortar characteristics." Proc. mt. Brick and Block Masonry Conf.,
J.W. de Courcy, Elsevier Applied Science, London, UK, 378-402.
Backes, H. P. (1985). "On the behaviour of masonry under tension in the direction of the
bed joins." Dissertation, Aachen University of Technology.
Bruneau, M. (1995). "Performance of masonry structures during the 1994 Northridge (Los
Angeles) earthquake." Canadian Journal of Civil Engineering, 22, 378-402.
BIA. (1986). "Anchor Bolts for Brick Masonry." Technical Notes on Brick Construction
44, Brick Institute of America, Reston, Virginia, 1986.
Casabonne, C. (2000). "Masonry in the seismic areas of the Americas, recent investigation
and developments." Proceeding of structure engineering and material, 319-327.
Calvi, G.M. and Magenes G. (1991). "Experimental evaluation of seismic strength of old
masonry structures." Proceeding of the International Brick/Block Masonry Conference,
Calgary, .Canada, 319-327.
R-3
Reference
Cook, R. D., Malkus, D. S., and Plesha, M. E. (1989). Concept and applications of finite
element analysis, John Wiley & Sons, mc, Canada.
Chen, W.F. (1982). Plasticity in reinforced concrete. United States of America, McGraw-
Hill Book Company.
Cheung M., Foo S. and Granadino J. (2000). Seismic retrofit of existing buildings:
innovative alternatives. Proceedings of mt. Conf. on the Seismic Performance of
Traditional Buildings, Turkey.
Dhanasekar, M., Page, A.W. and Kleeman, P.W. (1984). "Finite element model for the in-
plane behaviour of brick masonry." Ninth Australasian Conference on the Mechanics of
Structures and Materials, ACMSM 9. Sydney, 262-267.
R-4
Reference
Dhanasekar, M., Page, A.W. and Kleeman, P.W. (1985a). "Biaxial stress-strain relations
for brick masonry." Journal of the Structural Engineering, ASCE, Vol.111, No.5, 1085-
1100.
Dhanasekar, M., Page, A.W. and Kleeman, P.W. (1985b). "The failure of brick masonry
under biaxial stress." Proc. Instn. Civ. Engrs. Part 2, No. 79, 295-3 13.
Drysdale, R. G., Hamid, A.A., and Baker, L.R. (1994). Masonry structures, behaviour and
design, Prentice-Hall, Englewood Cliffs, N.J.
Ehsani, M. R., Saadatmanesh, H., and Al-Saidy, A. (1997). "Shear behavior of URM
retrofitted with FRP overlays." J. Comp. for Constr., ASCE, 1(1), 17-25.
R-5
Reference
FEMA. (1992). "NEHRP guidelines for the seismic rehabilitation of buildings." Report No.
FEMA-172, Federal Emergency Management Agency, Building Seismic Safety Council,
Washington, D.C.
FEMA. (1997a). "NEHRP guidelines for the seismic rehabilitation of buildings." Report
No. FEMA-273, Federal Emergency Management Agency, Building Seismic Safety
Council, Washington, D.C.
FEMA. (1997b). "NEHRP commentary on the guidelines for the seismic rehabilitation of
buildings." Report No. FEMA-274, Federal Emergency Management Agency, Building
Seismic Safety Council, Washington, D.C.
Ganz, H. R and Thurlimann, B. (1982). "Tests on the biaxial strength of masonry." Report
No. 7502-3, Institute of Structural Engineering, ETH Zurich, Zurich, Switzerland.
R-6
Reference
Gigla, B., and Wenzel, F. (1997). "The repair of historic masonry structures by injection
thj
anchors." International Brick/Block Masonry Conference, Tongji University,
Shanghai, China, 1154-1167.
Hamburger, R.O. (1997). "A framework for performance —based earthquake resistive
design" (http://nisee.berkeley.edu/lessonsIhamburger.html)
Hamid, Ahmad A., Chia-Calabria, Catherine and Harris, Harry G. (1992). "Flexural
behavior of joint reinforced block masonry walls." ACI Structural Journal, 89 (1), 20-26.
Hamilton III, H. R. and Dolan, C. D. (2001). "Flexural capacity of glass FRP strengthened
concrete masonry walls." Journal of composites for construction, 170-179.
R-7
Reference
Hiraishi.H et al. (1998). "New framework for performance based design of building
structures —Design flow and social system-" Wind and Seismic Effects, 234-247
Horton, R. T. and Tadros, M. K. (1990). "Deflection of Reinforced Masonry Members."
ACI Structural Journal, 87 (4), 453-463.
Kachlakev, D., Miller, T. and Yim, S. (2001). "Finite element modelling of reinforced
concrete structures strengthened with FRP laminates." Final Report, Oregon Department of
transportation Research Group, Washington.
Korany, Y., Drysdale, R., and Chidiac, 5. (2001). Retrofit of Unreinforced Masonry
Buildings: The State-of-The-Art. The 9th Canadian Masonry Sym, Fredericton, New
Brunswick, June, CD Paper RETFITO8.
Kolsch, H. (1998). "Carbon fiber cement matrix (CFCM) overlay system for masonry
strengthening." J. of Comp. for Constr., ASCE, 2(2), 105-109.
R-8
Reference
Lourenco, P. B. (1995). "An orthotropic continuum model for the analysis of masonry
structures." Report 03-21-1-31-27, Deift University of Technology, Deift, The Netherlands.
Lotfi, H. R and Shing, P. B. (1993). "Analysis of masonry walls with smeared and discrete
crack models." Proceedings of the Symposium on Structural Engineering in Natural
Hazards Mitigation, 1179-1184.
Lurati, F and Thurlimann, B. (1987). "Test in concrete masonry walls." Report No. 8401-3,
Institute of Structural Engineering, ETH Zurich, Zurich, Switzerland
Magenes, G. and Calvi, G. M. (1997). "In-plane seismic response of brick masonry walls."
Earthquake engineering and structural dynamics, Vol 26, 1091-1112.
Manzouri, T., Schuller, M. P., Shing, P.B.,and Amadei, B. (1996). "Repair and retrofit of
unreinforced masonry structures." Earthquake Spectra., 12(4), 903-922.
Mayes RL. (1989). "Design of structures with seismic isolation." The Seismic Design
Handbook, 413-438.
Meichers R.E. (1990). Newcastle Earthquake Study, The Inst. Of Eng., Australia.
R-9
Reference
Mele E and De Luca A. (1994). State of the art report on base isolation and energy
dissipation. STESSA, Proceedings mt Workshop, organised by the European Convention
for Constructional Steelwork, 631-658.
Mercer J. C. (1998). "Simple low cost retrofit procedures for historic unreinfroced masonry
buildings." Proc. of the 1998 ASCE Structures Congress.
Page, Adrian W. (1978). "Finite element model for masonry." Journal of Structural
Division, 104 (ST8), 1267-1285.
Page A.W. (1992). "The design, detailing and construction of masonry — the lessons from
the Newcastle Earthquake." Australian Civil Eng. Transactions, Vol. CE 34, No. 4, 1992,
343-353.
Pierepiekarz M. (2000). New research can help building owners who are shaken up by
quake hazards, Seattle Daily Journal of Commerce online edition.
R-10
Reference
Pauley, T. and Priestley, M.J.M. (1992). Seismic design of reinforced concrete and
masonry buildings, Wiley, New York.
Rai, D. C., Goel, S. C. and Holmes, W. T. "Seismic strengthening of URM buildings with
steel bracing." 697-705.
Renda V. Tirelli D, Magonette G and Molina J. (2000). "Contribution of the joint research
centre to the charac.terisation of shape memory alloys and full-scale validation tests of
retrofitting techniques for masonry shear walls typical of the architectural heritage", mt
Millennium Congress — Archi, Paris, France.
R-11
Reference
Saliba, J. F., Al-Allad, Riad S. amd Sawaya, George E. (1996). "Use of finite elements in
masonry structures." The Masonry Society Journal, 9-16.
Shing, P. B., Noland, J. L., Ekiamerus, E. and Spaeh, H. (1989). "Inelastic behavior of
concrete masonry shear walls." Journal of Structural Engineering, 115 (9), 2204-2225.
Shing, P. B., Schuller, M., Soskere, V. S. and Carter, E. (1990). "Flexural and shear
response of reinforced masonry walls." ACI Structural Journal, 87 (6), 646-656.
Spyrakos, C. C. (1996) Finite element modeling in engineering practice, Algor, mc, the
United States of America.
Stasa, F. L. (1985). Applied finite element analysis for engineers, CBS, the United States
of America.
R-12
Reference
Northridge lessons; Part 4: moving the Blue toward performance based engineering."
nisee.
Soroushian, P., Obaseki, K. and Choi, Ki-Bong. (1988). "Nonlinear modeling and seismic
analysis of masonry shear walls." Journal of Structural Engineering, 114 (5), 1190-1119.
Taghdi, M., Bruneau, M. and Saatcioglu, M. (2000). "Analysis and design of low-rise
masonry and concrete walls retrofitted using steel strips." Journal of Structural
Engineering, 126 (9), 1026-1032.
Takenaka Corp. (2001). "Hybrid base isolation system used in super high-rise office
building for the first time in the world." News Release, www.takenaka.co.jp.
Tena-Colunga, A., EERI, M. (1996). "Some retrofit options for the seismic upgrading of
old low-rise school buildings in Mexico." Earthquake Spectra., 12(4), 883-902
Taghdi, M. (1998), "Seismic retrofit of low-rise masonry and concrete walls by steel
strips." PhD Thesis, Univ. of Ottawa.
Taghdi, M., Bruneau, M., and Saatcioglu, M. (2000). "Seismic retrofitting of low-rise
masonry and concrete walls using steel strips." J. Struct. Engrg., ASCE, 126(9), 1017-
1025.
R-13
Reference
Taghdi, M., Bruneau, M., and Saatcioglu, M. (2000). "Analysis and design of low- rise
masonry and concrete walls retrofitted using steel strips." J. Struct. Engrg., ASCE, 126(9),
1026-1032.
Tinazzi, D., Tumialan, G., Myers, J. J., and Nanni A."Field evaluation of unreinforced
masonry walls strengthened with FRP composites subjected to out-of-plane loading."
Center for Infrastructure Engineering Studies, Univ. of Missouri -Rolla.
Tinazzi, D., Arduini, M., Modena, C, and Nanni A.(2000). "FRP-Structural repointing of
masonry assemblages." Proc., 3rd Inter. Conf on Advanced Composite Materials in Bridges
and Strucutres, Ottawa, Canada, 585-592.
Tomazevic, M., Lutman, M. and Petkovic, L. (1996). "Seismic behavior of masonry walls:
Experimental: simulation." Journal of Structural Engineering, 1040-1047.
R-14
Reference
Tumialan, G., Tinazzi, D., Myers, J. J. and Nanni, A. "Field evaluation of unreinforced
masonry walls strengthened with FRP composites subjected to out-of-plane loading."
Tumialan, G., Huang, P. C., Silva, P., and Nanni A.(2001) "Strengthening of masonry walls
by FRP structural repointing." Non-Metallic Reinforcement for Concrete Structures-
FRPRCS-5, Cambridge
Tumialan, G., Micelli, F., and Nanni A.(2001) "Strengthening of masonry structures with
FRP composites ." Structures 2001, Washington D.C., May 21-23,2001.
Weaver, Jr., William and Johnston, Paul R. (1984). Finite elements for structural analysis,
Prentice-Hall, mc, the United States of America.
R-15
Reference
Zhuge, Y., Thabbiratnam, D. P., and Corderoy, J. (1998). "Nonlinear dynamic analysis of
unreinforced masonry." J. Struct. Engrg., ASCE, 124(3), 270-277.
Zienkiewicz, 0. C., and Taylor, R. L. (2000). The finite element method Volume 1 The
basis, Bitterworth-Heinemann 2000, England.
R-16
Appendix A
Table Al
Diameter Average Height Area Load
Diameter
Specimens mm mm mm mm2 N MPa
1 100.8 100.4 202 8011.85 241000 31
100.0
Table A2
Diameter Average Height Area Load
Diameter
Specimens mm mm mm mm2 N MPa
1 101.2 101.0 204 8011.85 61000 7.60.
100.8
101.2
Table A3
Length Width Height Area Core Load Load •
. Diameter Bearing.
Area
Specimens mm mm mm mm2 mm mm2 N MPa
1 230 110 600 25300 40 21530.09 484000 22.480
2 230 110 600 25300 40 21530.09 484000 19.972
3 230 110 600 25300 40 21530.09 484000 18.579
A-i
Appendix A
Table A4
Elastic Poisson's Yield Compressive Tensile Breaking
Modulus ratio stress strength strength load
Material MPa MPa MPa MPa KN
Masonry 5,940* 0.09 7.92 0.2
Cable** 107,500 0.3 . 71
Steel*** .
A-2
________
Appendix B
I II
I Ii II II Ii I
I II II II I
I II II II II I
I II II II I
I. F I I I II I
I II II II
II II II
H' III II
•II
111111!!
I! II
ii II II
B-i
Appendix B
I II II Ii I
I ii II I
I I I
I I
I I
I I I I I II II I
I II II I
I II
I I
B-2
__H_______
Appendix B
I ii II I
B-3
Appendix B
I II I
I II I
II I
I II I
I II
I II
II
II II 1.
Appendix B
I II I
I I I
I I
I I
I I
I II
I I
I II II It
B-5
Appendix B
I II I II I
I II II II II J
I II II
I I
I I I
I I
I ii I
I II II
I I
I ill • it it
B-6
Appendix B
I II II I
I I
I I I
I II I
I II
II
I [I IL__
B-7
Appendix B
I I
II
I ii II I
I I I
II
I I
IL
B-8
Appendix B
I II II II 1
I Ii II II
I I
II
ii
I
B-9
_______
Appendix B
I hi ii Ii I
I II II II
II I
I I
ii
I hi II I
t. ii flu
H' lit II
H
I
F I I I I
B-1O
Appendix B
I II II Ii I
ii II II I
• I I
I I
II
II
II II
II
ii II
B-li
Appendix B
I ill II 1:1
lii I
iii
Lit 2' HI
I I L ii Ii 1
•—1t I
1:1
I I III
B-12
____________________ __________
Appendix B
I II II II I
I II II II II I
I II
I 11 I I I
I II I
I II
H I'
B-13
Appendix B
I ii I
I II I
I
I
II
II I
I I
I II
B-14
Appendix C
APPENDIX C - PERFORMANCE-BASED
DESIGN GUIDELINES
The design professional shall review initial considerations with the building owner or
code official to determine any restrictions that exist on the design of rehabilitation
measures. Initial considerations shall include structural characteristics of the building,
seismic hazards including geologic site hazards known to be present at the site, results of
prior seismic evaluation, building use and occupancy requirements, historic status,
economic considerations, societal issues, and local jurisdictional requirements.
Partial Life Safety for BSE-1 Life Safety br BSE-1 Life Safety br 855-1
Partial Collapse Prevention Cotlapse Prevention for BSE-2 (Collapse Prevention for BSE-2
for BSE-2
and
Partial tite Safety < 500 yearS Immediate occupancy at any return
Collapse Prevention 2.500 years period, or
limrted Safety at any return panod Damage Control at any return period, or
tile Safety al'. 500 years
General goal: To make the General goal: To provide General goal: To provide a low
building better than it was a tow risk of endangerment nsk of endangerment of life
before rehabilitation of life safety for any event safety for any event likely to
likely to affect Ihe building affect the building, and to further
protect building features and/or
contents against damage
Earthquake demands are a function of the location of the building with respect to
causative faults, the regional and site-specific geologic characteristics, and the ground
motion hazard level(s) selected in the Rehabilitation Objective.
50%/50 year 72
C-2
Appendix C
They are termed a Basic Safety Earthquake 1 (BSE-1) and Basic Safety Earthquake 2
(BSE-2). The BSE-1 and BSE-2 earthquakes are typically taken as 1O%/50 and 2%/50
year events, respectively, except in regions near major active fault.
The Rehabilitation Objective selected as a basis for design will determines, to a great
extent, the cost and feasibility of any rehabilitation project, as well as the benefit to be
obtained in terms of improved safety, reduction in property damage, and interruption of
use in the event of future earthquakes. Table C. 1 presents a matrix indicating the broad
range of Rehabilitation Objectives that may be used in the NHERP Guideline. Each cell
in this matrix represents a single Rehabilitation Objective. The goal of a rehabilitation
project may be to satisfy a single Rehabilitation Objective—for example, Life Safety for
the BSE-1 earthquake—or multiple Rehabilitation Objectives—for example, Life Safety
for the BSE-1 earthquake, Collapse Prevention for the BSE-2 earthquake, and Immediate
Occupancy for an earthquake with a 50% probability of exceedance in 50 years. A
C-3
Appendix C
o
0)
In
C 0 C),,
C
Cu
E
co
—
C
C)
C)
OS-s It 0-a
50%/5O year a b C d
I
,.
20%/SO year e 1 g h
-o
a-
BSE—1 J k
'Cu
(—iQ%/50 year)
a BSE-2
(—2%/50 year)
m n 0
Notes:
1. Each cell in the above matrix represents a discrete Rehabilitation
Objective.
2. The Rehabilitation Objectives in the matrix above may be used to
represent the three specific Rehabilitation Objectives defined in
St tions 1.4.1, 1.4.2, and 1.4.3, as follows:
= Basic Safety Objective (BSO)
Jc+ p i-any of a, e, i, I>, f, j, or n = Enhanced Objeciives
o alone or n alone or m alone = Enhanced Objective
k alone or p alone Limited Objectives -
c, g, d, li, 1= Limited Objectives
In general, Rehabilitation Objectives that expect relatively low levels of damage for
relatively infrequent earthquake events will result in more extensive rehabilitation work
C-4
Appendix C
and greater expense than objectives with more modest goals of controlling damages.
Figure C.2 schematically presents the relationship between different Rehabilitation
Objectives and probable program cost.
cost
2% (BSE-2)
Increasing
performance
Figure C-2 Surface showing relative costs of various Rehabilitation Objectives (FEMA-
274, 1997)
A desirable goal for rehabilitation is to achieve the Basic Safety Objective (BSO). In
order to achieve this objective, building rehabilitation must be designed to achieve both
the Life Safety Performance Level (3-C) for BSE-1 earthquake demands and Collapse
Prevention Level (5-E) for BSE-2 earthquake demands. Once a desired Building
Performance Level for a particular ground shaking severity (seismic demand) is selected,
the result is a Rehabilitation Objective. With the exception of the Basic Safety Objective
(BSO), there are no preset combinations of performance and ground shaking hazard. The
Basic Safety Objective is met when a building can satisfy two criteria: (1) the Life Safety
Building Performance Level, which is the combination of the Structural and
Nonstructural Life Safety Performance Levels, for the Basic Safety Earthquake 1 (BSE-
1), and (2) the Collapse Prevention Performance Level, which only pertains to structural
C-5
Appendix C
performance, for the stronger shaking that occurs less frequently as defined in the Basic
Safety Earthquake 2 (BSE-2).
Any Rehabilitation Objective intended to provide performance superior to that the BSO is
termed an Enhanced Objective. An Enhanced Objective must provide better than BSO-
designated performance at either the BSE-l or BSE-2, or both. Enhanced performance
can be obtained in two ways:
Any Rehabilitation Objective intended to provide performance inferior to that of the BSO
is termed a Limited Objective. A Limited Objective may consist of either Partial
Rehabilitation or Reduced Rehabilitation. Limited Rehabilitation Objectives should be
permissible if the following conditions are met:
• The rehabilitation measures do not create a structural irregularity or make an
existing structural irregularity more severe;
• The rehabilitation measures do not result in a reduction in the capability of the
structure to resist lateral forces or deformations;
• The rehabilitation measures do not result in an increase in the seismic forces,
unless this component's behavior is still acceptable considering overall structural
performance;
• All new or rehabilitated structural elements are detailed and connected to the
existing structure, as required by the NEHIRP Guidelines;
C-6
Appendix C
Immediate
Collapse Prevention Life Safety Occupancy Operational
Level Level Level Level
Overall Damage Severe Moderate Light Very Light
General Little residual stiffness Some residual No permanent drift. No permanent drift;
and strength, but load- strength and stiffness Structure structure substantially
bearing columns and left in all stories. substantially retains retains original
walls function. Large Gravity-load-bearing original strength and strength and stiffness.
permanent drifts, elements function. No stiffness. Minor Minor cracking of
Some exits blocked, out-of-plane failure of cracking of facades, facades, partitions,
Infills and unbraced walls or tipping of partitions, and and ceilings as well
parapets failed or at parapets. Some ceilings as well as as structural elements.
incipient failure, permanent drift, structural elements. All systems important
Building is near Damage to partitions. Elevators can be to normal operation
collapse. Building may be restarted. Fire are functional.
beyond economical protection operable.
repair.
Nonstructural Extensive damage. Falling hazards Equipment and Negligible damage
components mitigated but many contents are generally occurs. Power and
architectural, secure, but may not other utilities are
mechanical, and operate due to available, possibly
electrical systems are mechanical failure or from standby sources.
damaged, lack of utilities.
Comparison with Significantly more Somewhat more Much less damage Much less damage
performance intended damage and greater damage and slightly and lower risk, and lower risk.
for buildings risk, higher risk
designed, under the
NEHPJ' Provisions, for
the Design
Earthquake
C-7
Appendix C
The Structural Performance Levels are the Immediate Occupancy Level (S-i), the Life
Safety Level (S-3), and the Collapse Prevention Level (S-5). Table C.3 relates these
Structural Performance Levels to the limiting damage states for common vertical
elements of lateral-force-resisting systems.
H
Unreinforced Primary Extensive Extensive cracking Minor (<1/9
Masonry Infill cracking and some crushing width) cracking of
and
Walls crushing; but wall remain in masonry infills and
portions of face place. No falling veneers at a few
course shed. units. Extensive corner openings.
crushing and spalling
of veneers at corners
of openings.
Secondary Extensive Same as primary. Same as primary.
crushing and
shattering; some
walls dislodge.
Drift 0.6% transit or 0.5% transit; 0.3% 0.1% transit;
permanent. permanent. negligible
permanent.
Table C.3 Structural Performance level for URM walls (FEMA-274, 1997)
Inherent in the concept of Performance Levels and Ranges is the assumption that
performance can be measured using analytical results such as story drift rations or
strength and ductility demands on individual components or elements.
The structural Performance Ranges are the Damage Control Range (S-2) and the Limited
Safety Range (S-4). Specific acceptance criteria are not provided for design to these
intermediate performance ranges. The engineer wishing to design for such performance
needs to determine appropriate acceptance within the Damage Control Range may be
obtained by interpolating the acceptance criteria for performance within the Limited
Acceptance criteria for performance within the Life Safety and Collapse Prevention
Performance Levels.
C-8
Appendix C
Structural Performance Level S-3, Life Safety, means the post-earthquake damage state
in which significant damage to the structure has occurred, but some margin against either
partial or total structural collapse remains. Some structural elements and components are
severely damaged, but this has not resulted in large falling debris hazards, either within or
outside the building. Injuries may occur during the earthquake; however, it is expected
that the overall risk of life-threatening injury as a result of structural damage is low. It
should be possible to repair the structure; however, for economic reasons this may not be
practical. While the damaged structure is not an imminent collapse risk, it would be
prudent to implement structural repairs or install temporary bracing prior to reoccupancy.
Structural Performance Level S-5, Collapse Prevention, means the building is on the
verge of experiencing partial or total collapse. Substantial damage to the structure has
occurred, potentially including significant degradation in the stiffness and strength of the
lateral-force-resisting system, large permanent deformation of the structure, and—to a
more limited extent—degradation in vertical —load-carrying capacity. However, all
significant components of the gravity-load-resisting system must continue to carry their
gravity load demands. Significant risk of injury due to falling hazards from structural
debris may exist. The structure may not be technically practical to repair and is not safe
for reoccupancy, as aftershock activity could induce collapse.
C-9
Appendix C
Structural Performance Range S-2, Damage Control, means the continuous range of
damage states that entail less damage than that defined for the Life Safety level, but more
than that defined for the Immediate Occupancy level. Design for Damage Control
performance may be desirable to minimize repair time and operation interruption; as a
partial means of protecting valuable equipment and contents; or to preserve important
historic features when the cost of design for Immediate Occupancy is excessive.
Acceptance criteria for this range may be obtained by interpolating between the values
provided for the Immediate Occupancy (S-i) and Life Safety (S-3) levels.
Structural Performance Range S-4, Limited Safety, means the continuous range of
damage states between the Life Safety and Collapse Prevention levels. Design parameters
for this range may be obtained by interpolating between the values provided for the Life
Safety (S-3) and Collapse Prevention (S-5) levels.
C-1O
Appendix C
and lighting. Occupant contents and furnishings (such as inventory and computers) are
also included.
C-li
Appendix C
In some cases, the decision may be made to rehabilitate the structure without addressing
the vulnerabilities of nonstructural components. It may be performed without interruption
of building operation. In some cases, it is possible to perform all or most of the structural
rehabilitation from outside occupied building areas, while extensive disruption of normal
operation may be required to perform nonstructural rehabilitation. Also, since many of
the most severe hazards to life safety occur as a result of structural vulnerabilities, some
municipalities may wish to adopt rehabilitation ordinances that require structural
rehabilitation only.
C-12
Appendix C
N-C Life Safety 1-C 2-C Life Safety 3-C 4-C 5-C 6-C
C-l3
Appendix C
Many building owners may wish to achieve this level of performance when the building
is subjected to moderate levels of earthquake ground motion. In addition, some owners
may desire such performance for very important buildings; under sever levels of
earthquake ground shaking. This level provides most of the protection obtained under the
Operational Level, without the cost of providing standby utilities and performing rigorous
seismic qualification of equipment performance.
C-14
Appendix C
This Building Performance Level consists of the Structural Collapse Prevention Levels
with no consideration of nonstructural vulnerabilities, except that parapets and heavy
appendages are rehabilitated. Buildings meeting this performance level may pose a
significant hazard to life safety resulting from failure of nonstructural components.
However, because the building itself does not collapse, gross loss of life should be
avoided. Many buildings meeting this level will be complete economic losses.
This level has sometimes been selected as the basis for mandatory seismic rehabilitation
ordinances enacted by municipalities, as it results in mitigation of the most severe life-
safety hazards at relatively low cost.
Earthquake hazard is a representation of the demand that will be placed upon a structure
at a given site for an intensity of ground shaking. The selection of an appropriate intensity
of ground shaking is a critical step in the Performance Based Seismic Design process.
Earthquake hazard can be defined using probabilistic or deterministic approaches. For
example, if a known fault source exists proximate to the site that has historically
produced earthquakes of a characteristic magnitude; a scenario event can be defined as
the earthquake hazard for consideration. For this scenario, a faulting mechanism,
hypocenter, length of rupture, magnitude and duration can be selected and deterministic
methods used to estimate the intensity of ground shaking at a given site. Alternatively, a
ground shaking intensity can be probabilistically selected as the shaking associated with
an event that has a prescribed probability of exceedence in a prescribed period of time or
an average return period. For example, building codes have historically been based on
ground shaking intensities corresponding to a 10% probability of exceedence in a 50-year
period of time. The selection of an appropriate earthquake hazard permits an owner,
architect or other design professional to implement a rational strategy for earthquake
design that reflects the unique needs of an owner and the seismic climate of the site.
C-15
Appendix C
As a minimum, at least one site visit should be performed to obtain detailed information
regarding building configuration and condition, site and geotechnical conditions, and any
issues related to adjacent structures, and to confirm that the available construction
documents are generally representative of existing conditions. If the building is a historic
structure, it is also important to identify the locations of historically significant features
and fabric. Care should be taken in the design and investigation process to minimize the
impact of work on these features. The following sections are focused to discuss about
obtaining as-built information for masonry wall.
C-16
Appendix C
of a building system are those representing the composite. Thus, permissible values are
given in this section for compressive strength and elastic modulus of the masonry
assemblage, flexural tensile strength at the unit-mortar interface, and shear strength and
shear modulus of vertical components such as piers, panels, for estimating stiffness and
strength of masonry wall.
Expected masonry compressive strength, fme' shall be measured using one of following
three methods. (FEMA 273, 1997)
1. Test prisms shall be extracted from an existing wall and tested per Section 1.4.B.3
of the Masonry Standards Joint Committee's Building Code Requirements for
Masonry Structures (MSJC, 1995a).
2. Prisms shall be fabricated from actual extracted masonry units, and a surrogate
mortar designed on the basis of a chemical analysis of actual mortar samples. The
test prisms shall be tested per Section 1 .4.B.3 of the Specification for Masonry
Structures (MSJC, 1995b).
3. Two flat jacks shall be inserted into slots cut into mortar bed joints and
pressurized until peak stress is reached.
For each of three methods, the expected compressive strength shall be based on the
net mortared area.
If the masonry unit strength and the mortar type are known, values may be taken
from Tables C.5 and C.6 of MSJC (1995a) for clay or concrete masonry constructed
after 1960. The fme value shall be obtained by multiplying the table values by a factor
that represents both the ratio of expected to lower bound strength and the height-to-
thickness ratio of the prism.
In lieu of material tests, default values for masonry prism compressive strength shall
be taken to not exceed 900 psi for masonry in good condition, 600 psi for masonry in
fair condition, and 300 psi for masonry in poor condition.
C-17
-
Appendix C
Masonry Condition1
Shear Strength3 - .
Fully grouted masonry with a lay-up other Ihan running bond 27 psi 20 psi 13 psi
Partially grouted or ungrouted masonry with a lay-up other 11 psi 8 psi 5 psi
than running bond
L Masonry condition shall be classified as good. fair, or poor as defined in this standard.
Table C-5 Default values for masonryprism compressive strength (MSJC, 1995a).
Property Factor
Compressive Strength (frne) 1
Elastic Modu'us in
Fiexural Tensile Strength 13
Shear Strength 1
Table C-6 The factor that represents both the ratio of expected to lower bound strength
and the height-to-thickness ratio of the prism. (MSJC, 1995a)
C-18
Appendix C
strength. Deformations between the two flat jacks shall be measured to infer
compressive strain, and thus elastic modulus.
In lieu of prism test, values for the modulus of elasticity of masonry in compression shall
be taken as 550 times the expected masonry compressive strength, fme.
Expected flexural tensile strength, fie' for out-of-plane bending shall be measured using
one of the following three methods:
1. Test samples shall be extracted from existing wall, and subjected to minor-axis
bending using the bond-wrench method.
2. Test samples shall be tested in situ using the bond-wrench method.
3. Sample wall panels shall be extracted and subjected to minor-axis bending in
accordance with ASTME 519.
In lieu of material test, default values of masonry flexural tensile strength for walls or
infill panels loaded normal to their plane shall be taken to not exceed 20 psi for masonry
in good condition, 10 psi for masonry in fair condition, and zero psi for masonry in poor
condition. For masonry constructed after 1960 with cement-based mortars, default values
of flexural tensile strength can be based on values from Table 8, 3.10.5.1 of BSSC
(1995).
Flexural tensile strength for unreinforced masonry (URM) walls subjected to in-plane
lateral forces shall be assumed to be equal to that for out-of-plane bending, unless testing
is done to define expected tensile strength.
For URM components, expected masonry shear strength, v me' shall be measured using
the in-place shear test. Expected shear strength shall be determined in accordance with
Equation C-i.
C-19
Appendix C
]
Vme = 1.5 (C-i)
where
The 0.75 factor on the v te term may be waived for single wythe masonry, or if the collar
joint is known to be absent or in very poor condition.
Values for the mortar shear strength, v te' shall not exceed 100 psi for the determination
Of Vte in Equation C-i.
Average bed-joint shear strength, v fe' shall be determined from individual shear strength
test values, in accordance with Equation C-2.
Vtest
— PD+L (C-2)
= Ab
where V test is the load at first movement of a masonry unit, A b is the net mortared area
of the bed joints above and below the test brick, and PD+L is the estimated gravity stress
at the test location.
In lieu of material tests, default values of shear strength of URM components shall be
taken to not exceed 27 psi for running bond masonry in food condition, 20 psi for running
bond masonry in fair condition, and 13 psi for running bond masonry in poor condition.
These values shall also be used for masonry in other than running bond if fully grouted.
For masonry in other than running bond and partially grouted or ungrouted, shear
strength shall be reduced by 60% of these values. For masonry constructed after 1960
C-20
Appendix C
with cement-based mortars, default values of shear strength can be based on values in
BSSC (1995) for unreinforced masonry.
The number and location of material tests shall be selected to provide sufficient
information to adequately define the existing condition of materials in the building. Test
locations shall be identified in those masonry components that are determined to be
critical to the primary path of lateral-force resistance.
A visual inspection of masonry condition shall be done in conjunction with any in situ
material tests to assess uniformity of construction quality. For masonry with consistent
quality, the minimum number of tests for each masonry type, and for each three floors of
construction or 3000 square feet of wall surface, shall be three, if original construction
records are available that specify material properties, or six, if original construction
records are not available. At least two tests should be done per wall, or line of wall
elements providing a common resistance to lateral forces. A minimum of eight tests
should be done per building.
Test should be taken at locations representative of the material conditions throughout the
entire building, taking into account variations in workmanship at different story levels,
variations in weathering of. the exterior surfaces, and variations in the condition of the
interior surfaces due to deterioration caused by leaks and condensation of water andlor
the deleterious effects of other substances contained within the building.
For masonry with perceived inconsistent quality, additional tests shall be done as needed
to estimate material strengths in regions where properties are suspected to differ.
Nondestructive condition assessment tests may be used to quantify variations in material
strengths.
C-21
Appendix C
An increased sample size may be adopted to improve the confidence level. The relation
between sample size and confidence shall be as defined in ASTM E22.
If the coefficient of variation in test measurements exceeds 25%, additional tests shall be
done. If the variation does not reduce below this limit, use of the test data shall be limited
to the Linear Static Procedures.
If mean values from in situ material tests are less than the default values, additional tests
shall be done. If the mean continues to be less than the default values, the measured
values shall be used, and shall be used only with the Linear Static Procedures.
The size and location of all masonry shear and bearing walls shall be determined. The
orientation and placement of the walls shall be noted. Overall dimensions of masonry
components shall be measured, or determined from plans, including wall heights, lengths,
and thickness. Locations and sizes of window and door openings shall be measured, or
determined from plans. The distribution of gravity loads to bearing walls should be
estimated.
The type and condition of the mortar and mortar joints shall be determined. Mortar shall
be examined for weathering, erosion, and hardness, and to identify the condition of any
repointing, including cracks, internal voids, weak components, and/or deteriorated or
eroded mortar. Horizontal cracks in bed joint. Vertical cracks in head joints and masonry
units, and diagonal cracks near openings shall be noted.
C-22
Appendix C
The examination shall identify vertical components that are not straight. Bulging or
undulations in walls shall be observed, as well as separation of exterior wythes, out-of-
plumb walls, and learning parapets or chimneys.
Connections between masonry walls, and between masonry walls and floors or roofs,
shall be examined to identify details and condition. If construction drawings are
available, a minimum of three connections shall be inspected for each general connection
type (e.g., floor-to-wall, wall-to-wall). If no deviations from the drawings are found, the
sample may be considered representative. If drawings are unavailable, or significant
deviations are noted between the drawings and constructed work, then a random sample
of connections shall be inspected until a representative pattern of connections can be
identified.
• Nondestructive tests
The location and number of nondestructive tests shall be in accordance with the
requirements.
• Supplemental tests
Ancillary tests are recommended, but not required, to enhance the level of confidence in
masonry material properties, or to assess condition. These are described in the
Commentary to this section.
C-23
Appendix C
To account for uncertainly in the collection of as-built data, a knowledge factor, K, shall
be selected from Table C-7 considering the selected Rehabilitation Objective, analysis
procedure, and data collection process. Knowledge factors shall be applied on a
component basis as determined by the level of knowledge obtained for individual
components during data collection. A knowledge factor, K, equal to 0.75, representing a
minimum level of knowledge of the structural system, shall be used if a visual
examination of masonry structural components is. A knowledge factor, K, equal to 1.00,
shall be used only with a comprehensive level of knowledge of the structural system.
This section provides basis engineering information for assessing attributes of structural
walls, and included stiffness assumptions, strength acceptance criteria, and deformation
acceptance criteria for Immediate Occupancy, Life Safety, and Collapse Prevention
Performance Levels.
Masonry walls shall be categorized as primary or secondary elements. Walls that are
considered to be part of the lateral-force system, and may or may not support gravity
loads, shall be primary elements. Walls that are not considered as part of the lateral-force-
C-24
Appendix C
resisting system, but must reniain stable while supporting gravity loads during seismic
excitation, shall be secondary elements.
• Stiffness
The lateral stiffness of masonry wall and pier components shall be determined based on
the minimum net sections of mortared and grouted masonry in accordance with the
guidelines of this subsection. The lateral stiffness of masonry walls subjected to lateral
in-plane forces shall be determined considering both flexural and shear deformations.
For linear procedures, the stiffness of a URM wall or pier resisting lateral forces parallel
with its plane shall be considered to be liner and proportional with the geometrical
properties of the uncracked section.
For nonlinear procedures, the in-plane stiffness of URM walls or piers shall be based on
extent of cracking.
Story shears in perforated shear walls shall be distributed to piers in proportion to the
relative lateral uncracked stiffness of each pier.
Stiffness for existing enhanced and new wall shall be determined using the same
principles of mechanics.
The basic procedure to select Rehabilitation Method shall include the following:
C-25
Appendix C
An analysis of the structure shall be conducted to determine the distribution of forces and
deformations induced in the structure by the design ground shaking and other seismic
hazards corresponding with the selected Rehabilitation Objective(s). The analysis shall
address the seismic demands and the capacity to resist these demands for all elements in
the structure that either:
• Linear analysis, including Linear Static Procedure (LSP), and Linear Dynamic
Procedure (LDP), including:
- Response Spectrum Analysis, and
- Linear Time-History Analysis, or
C-26
Appendix C
Linear Procedures
Linear procedures shall be permitted for buildings that do not have an irregularity. For
buildings that have one or more of the irregularities, linear procedures shall not be used
unless the earthquake demands on the building comply with the demand capacity ratio
(DCR) requirements.
The results of the linear procedures can be very inaccurate when applied to buildings with
highly irregular structural systems, unless the building is capable of responding to the
design earthquake(s) in a nearly elastic manner.
The methodology presented in this section shall be used to determine the applicability of
linear analysis procedures based on four conditions of irregularity. The determination of
irregularity shall be based on the configuration of the rehabilitated structure. A linear
analysis to determine irregularity shall be performed by either an LSP or an LDP. The
results of the analysis shall be sued to identify the magnitude and uniformity of
distribution of inelastic demands on the primary elements and components of the lateral-
force-resisting system.
The magnitude and distribution of inelastic demands for existing and added primary
elements and components shall be defined by demand-capacity ratios (DCRs) and
components in accordance with Equation (C-3):
DCR = QUD
(C-3)
QCE
C-27
Appendix C
where:
Q UD = Force due to the gravity and earthquake loads calculated
If the Linear Static Procedure (LSP) is selected for seismic analysis of the building,
the design seismic forces, their distribution over the height of the building, and the
corresponding internal forces and system displacements shall be determined using a
linearly elastic, static analysis in accordance with this section. Buildings shall be
modeled with linearly elastic stiffness and equivalent viscous damping values
consistent with components responding at or near yield level. The pseudo lateral load
shall be used to calculate internal forces and system displacements due to the design
earthquake.
Period Determination
. Method 1—Analytical
C-28
Appendix C
• Method 2—Empirical
(C-4)
Where:
h,, = Height (in feet) above the base to the roof level
• Method 3—Approximate
For unreinforced masonry buildings with single span flexible diaphragms, six stories or
less in height, use of Equation (C-5) to approximate the fundamental period shall be
permitted.
C-29
Appendix C
T= (C-5)
The pseudo lateral load in a given horizontal direction of a building shall be determined
using Equation (C-6). This load shall be used to design the vertical elements of the
lateral-force-resisting system.
V=C1C2C3CmSaW (C-6)
where:
C-30
Appendix C
c m
= Effective mass factor to account for higher mode mass participation effects
obtained from Table C-8. Cm shall be taken as 1.0 if the fundamental period, T, is
greater than 1.0 second.
Sa = Response spectrum acceleration, at the fundamental period and damping ratio of the
building in the direction under consideration. The values of S a shall be obtained
from the procedure.
W = Effective seismic weight of the building including the total dead load and applicable
portions of other gravity loads listed below:
1. In areas used for storage, a minimum 25% of the floor live load shall be
applicable. The live load shall be permitted to be reduced for tributary area as
approved by the code official. Floor live load in public garages and open
parking structures is not applicable.
2. Where all an allowance for partition load is included in the floor load design,
the actual partition weight or minimum weight of 10 psf of floor area,
whichever is greater, shall be applicable.
3. Total operating weight of permanent equipment.
C-31
Appendix C
4. Where the design flat roof snow load calculated in accordance with ASCE 7
exceeds 30 psf, the effective snow load shall be taken as 20 % of the design
snow load. Where the design flat roof snow load is less than 30 psf, the effective
snow load shall be permitted to be zero.
Seismic hazard due to ground shaking shall be based on the location of the building with
respect to causative faults, the regional and site-specific geologic characteristics, and
selected Earthquake Hazard Level. Assessment of seismic hazard due to earthquake-
induced geologic hazards shall be performed.
Seismic hazard due to ground shaking shall be defined as acceleration response spectra or
acceleration time-histories on either a probabilistic or deterministic basis.
Unless otherwise approved, the site-specific procedure shall be used where any of the
following conditions apply.
1. The building is located Type E soils and the mapped BSE-2 spectral response
acceleration at short periods (S exceeds 2.0g.
The seismic hazard due to ground shaking shall be defined for any Earthquake Hazard
Level using approved spectral response acceleration contour maps of 5%-damped
response spectrum ordinates for short-period (0.2 second) and long-period (1 second)
response.
C-32
Appendix C
1. If the desired Earthquake Hazard Level corresponds with one of the mapped
Earthquake Hazard Levels, obtain spectral response acceleration parameters
directly from the maps. Values between contour lines shall be interpolated.
2. If the desired Earthquake Hazard Level does not correspond with the mapped
levels of hazard, then obtain spectral response acceleration parameters from the
available maps, and modify them to the desired hazard level, either by logarithmic
interpolation or extrapolation.
3. Obtain design spectral response acceleration parameters by adjusting the mapped
or modified spectral response acceleration parameters for site class effects.
4. If the desired Earthquake Hazard Level is the Basic Safety Earthquake 2 (BSE-2),
obtain spectral response acceleration parameters.
5. If the desired Earthquake Hazard Level is the Basic Safety Earthquake 1 (BSE-1),
obtain the spectral response acceleration parameters.
6. Using the design spectral response acceleration parameters that have been
adjusted for site class effects, develop the general response spectrum.
Parameters and shall be obtained by interpolating between the values shown on the
map for the response acceleration contour lines on either side of the site, or by using the
value shown on the map for the higher contour adjacent to the site.
C-33
Appendix C
1. The values of S and taken from approved 1O%'50 year spectral response
acceleration contour maps and modified for site class. Values between contour
lines shall be interpolated.
2. Two-thirds of the values of the parameters for the BSE-2 earthquake hazard level.
For probabilities of exceedance, EY' between 2%150 years and I O%/50 years, when the
mapped BSE-2 short-period response acceleration parameter, is less than 1 .5g, the
where:
in(S1) = Natural logarithm of the spectral acceleration parameter ("i" = "s" for
C-34
Appendix C
I n(S ilO / 50) = Natural logarithm of the spectral acceleration parameter ("i" = "s" for
short-period or "i" =1 for 1 second period) for the BSE-2 hazard level
ln(P R) = Natural logarithm of the mean return period corresponding to the
— (C-8)
ln(1—PEY)
probabilities of exceedance between 2%/50 years and lO%/50 years shall be determined
from Equation (C-9):
si —s
— 110/50
C9
Where Si, 5j10/50 and are so defined above and n shall be obtained from Table C.9
C-35
Appendix C
A general horizontal response spectrum as shown in Figure C-3 shall be developed using
Equation (C-b), (C-i 1), and (C-12) for spectral response acceleration, S a versus
a
(C-b)
Sa for (C-li)
Sa (C- 12)
B1) (C-13)
C-36
Appendix C
T0 = 0.2 T5 (C-14)
8s B1
0.8 0.8
5 1.0 1.0
10 1.3 1.2
20 1.8 1.5
30 2.3 1.7
40 2.7 1.9
3.0 2.0
I. Damping coefficients shall be based on linear interpolation for
effective 'viscous damping values other than those given.
Sa = Sxs _L
Bs I Ts + '0.41
0
Sa =
Sa = IB1T
I
U) Sxi/Bi
Ts 1.0
Period, T
Use of spectral response accelerations calculated using Equation (C-l0) in the extreme
short period range (T<T0) shall only be permitted in dynamic analysis procedures and
only for modes other than the fundamental mode.
C-37
Appendix C
• Damping Ratios
A 5% damped response spectrum shall be used for the rehabilitation design of all
buildings and structural systems.
For probabilities of exceedance greater than I O%150 years and when the mapped short-
period response acceleration parameter, S is less than 1.5g. the modified mapped short-
parameter at a one-second period, S1. shall be determined from Equation (C-9), where
the exponent n is obtained from Table C-li.
C-38
Appendix C
For probabilities of exceedance greater than 1 O%/50 years and when the mapped short-
period response acceleration parameter, S is greater than or equal to 1 .5g, the modified
Region S1
C-39
Appendix C
The design short-period spectral response acceleration parameter, S and the design
spectral response acceleration parameter at one-second, S shall be obtained from
Equations (C-15) and (C-16), respectively, as follows:
(C-15)
(C-16)
where .Fa and F,, are site coefficients determined respectively from Tables C-13 and C-
14, based on the site class and the values of the response acceleration parameters S and
S1 for the selected return period.
:ij
C-40
Appendix C
Values of as a of Site
Class and Mapped Spectral
Response Acceleration at
One-Second Period S1
• Site classes
3. Class C: Very dense soil and soft rock with 1,200 2,SOOft/sec or with
either standard blow count N >50 or undrained shear strength >2,000 psf
C-41
Appendix C
4. Class D: Stiff soil with 600 ftlsec<v5 1,200 ftlsec or with 15<N 1,000
The parameters, , N and v5 are, respectively, the average values of the shear wave
velocity, Standard Penetration Test (SPT) blow count, and undrained shear strength of the
upper 100 feet of soils at the site. These values shall be calculated from Equation (C-17),
below:
(C-17)
where:
C-42
Appendix C
and
(C-19)
Where v data are available for the site, such data shall be used to classify the site. If
such data are not available, N data shall be used for cohesionless soil sites (sands,
gravels), and data for cohesive soil sites (clays). For rock in profile classes B and C,
classification shall be based either on measured or estimated values of Classification
of a site as Class A tock shall be based on measurements of v either for material at the
site itself, or for rock having the same formation adjacent to the site; otherwise, Class B
rock shall be assumed. Class A or B profiles shall not be assumed to be present if there is
more than 10 feet of soil between the rock surface and the base of the building.
For unreinforced masonry buildings with flexible diaphragms for which the fundamental
period is calculated using Equation (C-5), it shall be permitted to calculate and distribute
the pseudo lateral loads as follows:
1. For each span of the building and at each level, calculated period from Equation
(C-5).
2. Using Equation (C-6), calculate pseudo lateral load for each span.
3. Apply the lateral loads calculated for all spans and calculate forces in vertical
seismic-resisting elements using tributary loads.
4. Diaphragm forces for evaluation of diaphragms shall be determined from the
results of step 3 above and distributed along the diaphragm span considering its
deflected shape.
5. Diaphragm deflection shall not exceed 6 inches for this method of distribution of
pseudo lateral loads to be applicable.
6.
C-43
Appendix C
Some existing buildings have substantial strength and stiffness, but some of their
components may not have adequate strength, toughness, or deformation capacity to
satisfy the Rehabilitation Objectives. An appropriate strategy for such structures may be
to perform local modifications of components that are inadequate while retaining the
basic configuration of the building's lateral-force-resisting system. Local modifications
that can be considered include improvement of component connectivity, component
strength, and/or component deformation capacity. This strategy tends to be the most
economical rehabilitation approach when only a few of the building's components are
inadequate. Local strengthening allows one or more under-strength elements or
connections to resist the strength demands predicted by the analysis without affecting the
overall response of the structure. This could include measures such as cover plating steel
beams or columns, or adding plywood sheathing to an existing timber diaphragm. Such
measures increase the strength of the element or component and allow it to resist more
earthquake-induced force before the onset of damage.
C-44
Appendix C
confinement jacket around a reinforced concrete column to improve its ability to deform
without spalling or degrading reinforcement splices. Another measure is reduction of the
cross-section of selected structural components to increase their flexibility and response
displacement capacity.
Such irregularities are often, but bit always, caused by the presence of a discontinuity in
the structure, as for example, termination of a perimeter shear wall above the first story,
Simple removal of the irregularity may be sufficient to reduce demands predicted by the
analysis to acceptable levels. However, removal of discontinuities may be inappropriate
in the case of historic buildings, and the effect of such alterations on important historic
features should be considered carefully.
C-45
Appendix C
Partial demolition can also be an effective corrective measures for irregularities, although
this obviously has significant impact on the appearance and utility of the building, and
this my not be an appropriate alternative for historic structures. Portions of the structure
that create the irregularity, such as setback towers or side wings, can be removed.
Expansion joints can be created to transform a single irregular building into multiple
regular structures: however, care must be taken to avoid the potential problems associated
with pounding.
Global stiffening of the structure may be effective rehabilitation strategy I the results of a
seismic evaluation show deficiencies attributable to excessive lateral deflection of the
building, and critical components do not have adequate ductility to resist the resulting
deformations.
Construction of new braced frames or shear walls within an existing structure is effective
measures for adding stiffness.
C-46
Appendix C
• Mass Reduction
• Seismic isolation
C-47
• Appendix C
C-48
Appendix C
Elements and components that provide the capacity of the structure to resist collapse
under seismic forces induced by ground motion in any direction shall be classified as
primary. Other elements and components shall be classified as secondary.
The Type 1 curve depicted in Figure C-9 is representative of ductile behaviour where
there is an elastic range (point 0 to point 1 on the curve) followed by a plastic range
(points 1 to 3) with non-negligible residual strength and ability to support gravity loads at
point 3. The plastic range includes a strain hardening or softening range (points 2 to 3).
Primary component actions exhibiting this behaviour shall be classified as deformation-
controlled if the strain-hardening or strain-softening range is such that e>2g; otherwise,
they shall be classified as force-controlled. Secondary component actions exhibiting Type
1 behaviour shall be classified as deformation-controlled for any e/g ratio.
The Type 2 curve depicted in Figure C-9 is representative of ductile behaviour where
there is an elastic range (point 0 to point 1 on the curve) and a plastic range (points 1 to 2)
followed by loss of strength and loss of ability to support gravity loads beyond point 2.
Primary and secondary component actions exhibiting this type of behaviour shall be
classified as deformation-controlled if the plastic range is such that e 2g; otherwise,
they shall be classified as force-controlled.
C-49
Appendix C
• Deformation-Controlled Actions
where
Q E = Action due to design earthquake loads calculated using forces and analysis models.
The following component gravity forces, Q G' shall be considered for combination with
seismic loads.
When the effects of gravity and seismic loads are additive,
QG =1.1(QD+QL+Qs) (C-20)
QG =O.9QD (C-21)
C-50
Appendix C
where:
Q = Effective live load effect (action), equal to 25% of the unreduced design live load
but not less than the measured live load
Q = Effective snow load effect (action), equal to either 70% of the full design snow
load or, where conditions warrant and approved by the regulatory agency, not less
than 20% of the full design snow load, except that where the design snow load is 30
pounds per square foot or less, Q = 0.0
Evaluation of components for gravity and wind forces, in the absence of earthquake
forces, is beyond the scope of this document.
• Force-Controlled Actions
Force-controlled design actions, Q UF' shall be calculated using one of the following
methods:
1. Q UF shall be taken as the maximum action that can be developed in a component
based on a limit-state analysis considering the expected strength of the
components delivering load to the component under consideration, or the
maximum action developed in the component as limited by the nonlinear response
of the building.
2. Alternatively, Q UF shall be calculated in accordance with Equation (C-22).
C-51
Appendix C
± (C-22)
Q =Q
123
where:
J Force-delivery reduction factor, greater than or equal to 1.0, taken as the smallest
DCR of the components in the load path delivering force to the component in
question, calculated in accordance with Equation (2-1). Alternatively, values of J
equal to 2.0 in Zones of High Seismicity, 1.5 in Zones of Moderate Seismicity,
and 1.0 in Zones of Low Seismicity shall be permitted when not based on
calculated DCRs. J shall be taken as 1.0 for the Immediate Occupancy Structural
Performance Level. In any case where the forces contributing to Q UF are
delivered by components of the lateral force resisting system that remain elastic, J
shall be taken as 1.0.
The basic approach for calculating force-controlled actions for design differs from that
used for deformation-controlled actions because nonlinear deformations associated with
forced-controlled actions are not permitted. Therefore, force demands for force-
controlled actions should not exceed the force capacity (strength).
Ideally, an inelastic mechanism for the structure will be identified, and the force-
controlled actions, Q UF for design will be determined by limit analysis using that
mechanism. This approach will always produce a conservative estimate of the design
actions, even if an incorrect mechanism is selected. Where it is not possible to use limit
(or plastic) analysis, or in cases where design forces do not produce significant nonlinear
response in the building, it is acceptable to determine the force-controlled actions for
design using Equation (C-22).
Coefficients C1, C2 and C3 were introduced in Equation (C-6) to amplify the design
base shear to achieve a better estimate of the maximum displacements expected for
buildings responding in the inelastic range. Displacement amplifiers, C1, C2 and C3 are
C-52
Appendix C
divided out of Equation (C-22) when seeking an estimate of force level present in a
component when the building is responding inelastically.
• Deformation-Controlled Actions
where:
K = Knowledge factor
Q CE' the expected strength, shall be determined considering all coexisting actions on the
component under the design loading condition.
• Force-Controlled Actions
Force-controlled actions in primary and secondary components and elements shall satisfy
Equation (C-24):
C-53
Appendix C
KQCL (C-24)
where:
Q CL' the lower-bound strength, shall be determined considering all coexisting actions on
the component under the design loading condition.
Expected lateral strength of existing URM walls or pier components shall be based on
expected bed-joint sliding shear strength, or expected rocking strength, in accordance
with Equations C-25 and C-26, respectively. The strength of such URM walls or piers
shall be the lesser of:
QCE (C-25)
QCE (C-26)
where:
C-54
Appendix C
CE
= Expected vertical axial compressive force per load combinations in Equations C-
20 and C-21
VbJS = Lateral strength of wall or pier based on bed-joint shear strength, pounds
a = Factor equal to 0.5 for fixed-free cantilever wall, or equal to 1.0 for fixed-fixed pier
Expected lateral strength of newly constructed wall or pier components shall be based on
the NEHERP Recommended Provisions (BSSC, 1995), with the exception that capacity
reduction factors shall be taken as equal to 1.0.
Lower bound lateral strength of existing URM walls or pier components shall be limited
by diagonal tension stress or toe compressive stress, in accordance with Equation C-27
and C-29, respectively. The lateral strength of URM walls or piers shall be the lesser of
Q CL values given by these two equations.
f/h eff is larger than 0.67 and less than 1.00, then:
where:
C-55
Appendix C
A h eff , L and a are the same as given for Equations C-25 and C-26, and:
fa = Upper bound of vertical axial compressive stress from Equation C-20, psi
21, pounds
For determination of the bed-joint shear strength, may be substituted for the
diagonal tension strength, f' dt' in Equation C-25.
If the Linear Static Procedures are used, lateral forces from gravity and seismic effects
shall be less than the lower bound lateral strength, Q CL' as required by Equation C-22.
Lower bound vertical compressive strength of existing URM walls or pier components
shall be limited by masonry compressive stress per Equation C-29.
where:
C-56
Appendix C
If the Linear Static Procedures are used, vertical forces from gravity and seismic effects
shall be less than the lower bound lateral strength, Q CL' as stated in Equation C-22.
A. Linear procedures
If the linear procedures are used, the product of expected strength, Q CE' of those
components classified as deformation-controlled, multiplied by m factors given in Table
C-16 for particular Performance Levels and K factors , shall exceed the sum of unreduced
seismic forces, Q E' and gravity forces, Q G' per Equation 3-19. The limiting behavior
mode in Table C-15 shall be identified from the lower of the two expected strengths as
determined from Equations C-25 and C-26. For determination of m factors from Table C-
16, the vertical compressive stress, fae' shall be based on an expected value of gravity
compressive force given by the load combinations given in Equations C-20 and C-2 1.
Linear Static Procedure—rn Factors for UFiM In-Plane Walls and Piers
in Factors
Noic:
Table C-15 m factors for URM in-plane walls and piers (FEMA-273, 1997)
B. Nonlinear procedures
C-57
Appendix C
1.0
eff
Note: is pennittvd
Table C-16 Simplified force-deflection relation for URM in-plane walls (FEMA-293,
1997)
C-58
Appendix C
Level shall be considered at approximately 75% of the e value in the table. Drift
percentages based on these criteria are given in Table C-17.
If the Nonlinear Dynamic Procedure is used, nonlinear force-deflection relations for wall
and pier components shall be established based on the information given in Table C-17,
or on a more comprehensive evaluation of the hysteretic characteristics of those
components.
C-59
Appendix D
D-1
Appendix E
TYPICAL GRADE 316 STAINLESS STEEL WIRE ROPE BREAKING LOAD SPECIFICATIONS
1.2mm 3/64 in. 1.08 110 243 0.81 83 183 0.85 87 192
1.5mm 1/16 in. 2.11 215 474 1.44 147 324 2.01 205 452
2.0mm 5/64 in. 3.14 320 705 2.26 230 507 2.37 242 534
2.5mm 3/32 in. 4.90 500 1102 3.82 390 860 . 3.71 378 833
3.0mm 1/8 in. 7.06 720 1587 5.00 510 1124 5.34 544 1200
4.0mm 5/32 in. 12.60 1285 2833 8.89 907 2000 9.49 1089 2400
4.8mm 3/16 in. 18.90 1930 4255 12.60 1280 2822 14.10 1437. 3168
5.0mm — 19.60 1999 4407 13.90 1417 3124 14.80 1509 3327
5.6mm 7/32 in. 24.20 2470 5445 17.20 1750 3858 18.10 1850 4080
6.0mm — 28.00 2876 6340 20.00 2039 4495 21.40 2181 4810
6.4mm 1/4 in. 34.00 3440 7584 22.00 2280 5027 25.90 2642 5825
7.0mm 9/32 in. 35.00 3549 7824 36.00 2784 6138 29.10 2966 6540
8.0mm 5/16 in. 46.00 4640 10229 36.00 3630 8003 3874 8540
9.5mm 3/8 in. 65.00 6580 14506 51.00 5150 11354 57.20 5830 12850
10.0mm — 71.00 7250 15984 56.00 5670 12500 . 59.30 6045 13330
11.0mm 7/16 in. 86.00 8770 19335 68.00 6950 15322 76.70 7820 17240
12.0mm — 102.00 10401 22930 80.00 8158 17985 85.40 8700 19180
12.7mm 1/2 in. 119.00 12101 26678 90.00 9150 20172 107.00 10900 24030
14.0mm 9/16 in. 139.00 14174 31248 - 109.00 11115 24504 117.00 11930 26300
SWAGING DIMENSIONS
Correct installation of a swage fitting requires that the shank of the fitting be formed down onto the wire with specialised dies and presses in accordance
with the following dimensions:
Wire mm 2.5 3.0 4.0 5.0 5.6 6.0 7.0 8.0 10.0 11.0 12.0 14.0 16.0 19.0 22.0 26.0
Diameter in. 3/32 1/8 5/32 7/32 1/4 9/32 5/16 3/8 7/16 1/2 9/16 5/8 3/4 7/8 1
Code mm 2.5M 03M 04M 05M . 06M 07M 08M 1OM — 12M 14M 16M 19M 22M 26M
Code in. 03 04 05 06 07 08 09 10 12 14 16 18 20 24 28 32
LenglhotWire mm 32.3 38.4 45 55 60.5 70 78.5 88 110 122 140 157 176 210 245 286
lnsideSwage in. 1.27151 1.77 2.17 2.38 2.76 3.09 3.46 4.33 4.80 5.51 6.18 6.93 8.27 9.65 11.26
Diameter 00 mm 5.54 6.35 7.54 9.12 10.85 12.55 16.13 17.86 19.84 21.44 24.99 8.17 34.52 40.46 46.02
Before swaging in. 0.218 0.250 0.297 0.359 0.427 0.494 0.563 0.635 0.703 0.781 0.844 0.984 1.109 1.359 1.593 1.812
DiameterOD mm 4.83 5.56 6.35 7.95 9.53 11.13 12.70 14.30 15.88 17.48 19.05 22.23 25.40 31.75 36.50 41.28
After swaging in. 0.190 0.219 0.250 0.313 0.375 0.438 0.500 0.563 0.625 0.688 0.750 0.875 1.000 1.250 1.437 1.625
1 x 19 Grade 316
WR6119-1.5M 1.5 305 (1000) 180 (397) 0.011 (0.024)
*
rigging, Balustrading
WR6119-14M 14.0 305 (1000) 14174 (31254) 0.971 (2.141)
* Standard imperial diameter wire. Refer imperial diameter fitting tables. ** Wire rope PVC coated.
Wire Cutters - -
CLA6O6 4 (5/32) H.W.C.9 - - 310 (11.0)
i 'ii u ii g :1; i i i II ru
Threaded swage terminals (2.5-26mm, 3/32"-l" wire) have rolled right Ronstan's Extra Long threaded swage terminals can be used to lengthen
hand UNF threads for greater strength and are manufactured from Grade cables to ensure adequate thread engagement is maintained.
316 Stainless Steel for maximum corrosion resistance. WIRE DIAMETER
Metric: 3mm — 26mm Imperial: 1/8" —1"
32mm CUSTOM0 - M45 46.0 (113/16) - - 529 (20 13/16) 205 (81/16) - - 6600 (232.8)
CUSTOMQ - - - - -
36mm CUSTOM - M48 46.0 (113/16) - - 595 (23 7/16) 231 (93/32) - - 8820 (311.1)
Imperial Swage - -
3/32 RF217,218,219 3/16" UNF 4.0 (5/32) 30 (13/16) 78 (31/8) (1/4) 10 (0.4)
1/8 in. RF151O-0404 RF1576-0404 1/4" UNF 4.0 (5/32) 39 (19/16) 132 (53/16) 75 (215/16) 7.0 (9/32) 30 (1.1)
5/32 in. RF151O-O504 RF1576-O5O4 1/4" UNF 4.0 (f3/4) 141 (51/2) 75 (215/16) 7.0 (9/32) 30 (1.1)
5/32 in. RF151O-0505 RF1576-0505 5/16" UNF 5.3 (7/32) 45 (13/4) 155 (61/8) 89 (31/2) 8.7 (11/32) 45 (1.6)
3/16 in. RF1576-O6O4 1/4" UNF 4.0 (5/32) 44 (13/4) 139 (51/2) 75 (215/16) 7.0 41 (1.4)
3/16 in. RF151O-0605 RF1576-0605 5/16" UNF 5.3 (7/32) 55 (23/16) 164 (61/2) 89 (31/2) 8.7 (11/32) 56 (2.0)
3/16 in. RF151O-06O6 RF1576-O6O6 3/8" UNF 6.3 (1/4) 55 (23/16) 179 (71/16) 105 (41/8) 10.5 (13/32) 81 (2.9)
7/32 in. RF1 510-0706 RF1 576-0706 3/8" UNF 6.3 (1/4) 60 (23/8) 185 (71/4) 105 (41/8) 10.5 (13/32) 91 (3.2)
1/4 in. RF151O-O8O7 - 7/16" UNF 70 (23/4) 221(811/16) 124 (47/8) 11.9 (15/32) 150 (5.3)
1/4 in. RF151O-08O8 RF1 576-0808 1/2" UNF 8.5 (1 1/32) 70 (23/4) 241 (91/2) 143 (55/8) 13.9 (9/16) 195 (6.9)
9/32 in. RF1 51 0-0908 RF1 576-0908 1/2" UNF 8.5 (11/32) 78 (31/16) 248 (93/4) 143 (53/8) 13.9 (9/16) 220 (7.8)
5/16 in. RF151O-1008 RF1576-1008 1/2" UNF 8.5 (11/32) 88 (31/2) 261 (10 1/4) 143 (55/8) 13.9 (9/16) 240 (8.5)
5/16 in. RF151O-1009 - 9/16" UNF 10.4 (13/32) 88 (31/2) 276 (107/8) 159 (61/4) 15.6 (5/8) 310 (10.9)
5/16 in. RF151O-1O10 RF1576-1010 5/8" UNF 11.5 (7/16) 88 (31/2) 306 (12 1/16) 191 (71/2) 16.7 (21/32) 410 (14.5)
3/8 in. RF151O-121O RF1576-1210 5/8" UNF 11.5 (7/16) 109 (45/16) 335 (133/16) 191 (71/2) 16.7 (21/32) 460 (16.2)
7/16 in. RF1511-1412 RF1577-1412 3/4" UNF 12.0 (15/32) 122 (413/16) 262 (10 5/16) 105 (41/8) 21.2 (27/32) 460 (16.2)
1/2 in. RF1511-1614 RF1577-1614 7/8" UNF 140 (51/2) 297 (1111/16) 118 (45/8) 24.0 (15/16) 670 (23.6)
9/16 in. RF1511-1816 RF1577-1816 1" UNF 18.0 (23/32) 157 (63/16) 341 (13 3/8) 130 (51/8) 25.5 (1) 1040 (36.7)
5/8 in. RF1511-2016 RF1577-2016 1" UNF 18.0 (615/16) 355 (14) 130 (1) 1210 (42.7)
3/4 in. RF1511-2420 RF1577-2420 11/4" UNF 25.5 (1) 219 (85/8) 419 (16 4/8) 150 (515/16) 33.5 (15/16) 2320 (81.8)
7/8 in. RF1511-2820 RF1577-2820 11/4" UNF 25.5 (1) 245 (95/8) 462 (183/16) 150 (515/16) 33.5 (15/16) 2720 (95.9)
1 in. RF1511-3222 RF1577-3220 13/8" UNF 29.5 (15/32) 281 (11 1/16) 586 (23 1/8) 236 (9 5/16) 38.5 (11/2) 4770 (168.2)
EXTRA LONG Imperial_Swage
1/8 in. RF151OLO4O4 UNF 4.0 (5/32) 234 75 (2 5/1 6r7.0 (9/32) 30 (1.1)
3/16 in. RF1510L0605 RF1576-O6O5 5/16" UNF 5.3 (7/32) 55 (23/16) 316 (12 7/16) 89 (31/2) 8.7 (11/32) 56 (2.0)
*Recommended for balustrade applications only. B.L is below the rated B.L of 3/16" and 5mm wire
O Available as a completed cable assembly only
Appendix E
Features
— Two component epoxy technology with
one component dispensing system.
— Excellent resistance to alkali and moisture
in new concrete.
— Tough and durable for long term high
strength.
— No shrinkage during curing.
— Suitable for diamond cored holes.
- Non flammable.
- Low odour.
,,,,,
Anchor Thread Hole Installation Structural Limits Rec. Working Load (kN) See Safety Factors p11 & 12 No.01
Size Size 0 Embedded Fixture Fixture Tight Minimum Minimum Minimum Fixings
(mm) Depth Clearance Thickness Torque Edge Anchor Structural 2OMPa 3OMPa 4OMPa Per
(mm) 0 (mm) (Nm) Distance Spacing Thickness Cartridge
(mm) (mm) (mml (mm) Tensile •Shear Tensile •Shean Tensile •Shear
16 M16 18 125 19 40 70 65 100 160 15,8 19.8 19.3 19.8 22.3 19.8 34
20 M20 24 150 25 75 140 80 120 190 25.3 30.0 31.0 30.0 35.7 30.0 13
M20 24 **170 25 55 140 80 120 220 28.6 30.0 35.1 30.0 40.5 30.0 11
24 M24 26 160 30 105 230 95 145 200 29.2 43.4 35.8 43.4 41.3 43.4 15
M24 26 **210 30 55 230 95 145 270 38.3 43.4 47.0 43.4 54.2 43.4 11
• For shear loads acting toward the edge(s) of the concrete, the above edge distances and spacings are not applicable, please consult Ramset Technical Consultant. This table
should be read in conjunction with the Rarnset Engineers Design Manual.
Note: See pages 56 — 59 for anchors and accessories.
10 90 33 10 12 90 81
20 20 23 • 12 14 . 110 53
25 15 20 16 18 125 34
30 12 17 20 24 150 13
40 7 10 24 26 160 15
Temperatures refer to base material not air temperature. Shetf life 6 months.
® Ramset
Concrete Results"
Chemical Anchoring
ChemSet" Injection 800 Series
20
Product
ChemSet" Injection 801 is a heavy true epoxy
injection anchor.
ChemSei" Injection 802 is a heavy duty, true epoxy
injection anchor.
Installation
Installation temperature limits:
— Substrate: 5°C to 40°C.
Mortar: 18°C to 35°C.
Load should not be applied to anchor until the
chemical has sufficiently cured as specified.
Setting Times
1. Dnll recommended diameter and depth hole.
2. Clean hole with hole cleaning brush.
Gel lime Loading lime Gel lime Loading lime
Remove all debris using hole blower. - (mins) (Ins) (mins) (Iris)
Hole must be dry.
3. Insert mixing nozzle to bottom of hole. 40°C — — I — 7 I 10
Fill hole to 3/4 the hole depth slowly,
ensuring no air pockets form. 3o:c — 3 I 10 12 I 17
40 24 90 33
6. Attach fixture.
5°C - 75 36 120 66
0
Nate Cartridge temperature minimum 15°C.
® Ramset
Chemical Anchoring
Injection 800 Series
* Note: For shear loads acting towards an edge or where these minimum dimensions are not achievable, please use the simplified strength limit state designprocess to verify capacity.
Preferred h = otherwise,
h = Le - t
h 6 * dh
ENGINEERING PROPERTIES
20.4
0,
°' 40
=
0,
0,
1::
0
0 10 20 30 40 50 60 70 80 90 100
Table lb Absolute minimum edge distance and anchor spacing values, em and am (mm)
Anchor size, db M8 M1O M12 M16 M20 M24
em, am 25 30 35 50 60 75
Preferred h = otherwise,
h = Le - t
h 6 * dh
[Checkpoint
LI Anchor size determined, absolute minima compliance achieved, effective depth (h) calculated.
® Ramset
Concrete
20 Chemical Anchoring
Injection 800 Series
Strength Limit State Design
Table 2a Reduced characteristic ultimate concrete tensile capacity, (kN), = 0.6, = 32 MPa
Anchor size, db M8 M10 M12 M16 M20 M24
Drilled hole dia., dh (mm) 10 12 14 18 24 26
* Anchor size, db
Edge distance, e (mm)
M8 M10 M12 M16 M20 M24
fl
25 0.85
30 0.98 0.83
35 1 0.91 0.81
40 1 0.88
50 1 0.85
60 0.96 0.83
65 1 0.87
75 0.96 0.85
80 1 0.88
100 1
n n
*
a
Anchor size, db
Anchor spacing, a (mm)
M8 M10 M12 M16 M20 M24
25 0.76
30 0.81 0.75
35 0.86 0.79 0.74
40 0.92 0.83 0.78 .
I
II Anchor size, dh
25 0.52
30 0.63 0.50
35 0.73 0.58 0.49
40 0.83 0.67 0.56
50 1 0.83 0.69 0.52
60 1 0.83 0.63 0.50
15 1 0.78 0.63 0.52
100 1 0.83 0.69
120 1 0.83
150 1
Checkpoint
Step 3b Reduced characteristic ultimate bolt steel tensile capacity, ONif (kN)
Not appropriate for this product.
Check N* / ONur 1,
if not satisfied return to step 1
® Ramset
Concrete
20 Chemical Anchoring
Injection 800 Series
Strength Limit State Design
Table 4a Reduced characteristic ultimate concrete edge shear capacity, DV (kN), 0q = 0.6, =32 MPa
Anchor size, db M8 M10 M12 M16 M20 M24
Edge distance, e (mm)
25 1.6
30 2.2 2.4
35 2.7 3.0 3.2 3.6
50 4.6 5.1 5.5 6.2 6.9
60 6.1 6.7 7.2 8.2 9.0 9.8
75 8.5 9.3 10.1 11.4 12.6 13.7
125 18.3 20.1 21.7 24.6 27.1 29.5
200 40.6 43.8 49.7 54.9 59.7
300 80.5 91.3 100.9 109.7
400 140.5 155.4 168.9
500 . 217.2 236.1
600 310.3
Note: Effective depth. h must be 6 x drilled hole diameter. dh for anchor to achieve tabled shear capacities.
Anchor spacing
0.20 0.40 0.60 0.80 1.00 120 1.40 1.60 1.80 2.00 2.25 2.50
Edge distance, a / e
Number of anchors, ii
2 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
3 0.72 0.78 0.80 0.83 0.86 0.88 0.91 0.93 0.95 0.96 0.98 1.00
4 0.57 0.64 0.69 0.74 0.79 0.82 0.86 0.89 0.92 0.94 0.97 1.00
5 0.49 0.57 0.63 0.69 0.74 0.79 0.83 0.87 0.90 0.93 0.97 1.00
6 0.43 0.52 0.59 0.66 0.71 0.77 0.81 0.85 0.89 0.93 0.96 1.00
1 0.39 0.48 0.56 0.63 0.69 0.75 0.80 0.84 0.88 0.92 0.96 1.00
8 0.36 0.46 0.54 0.61 0.68 0.74 0.79 0.84 0.88 0.92 0.96 1.00
9 0.34 0.44 0.52 0.60 0.67 0.73 0.78 0.83 0.87 0.91 0.96 1.00
10 0.32 0.42 0.51 0.59 0.66 0.72 0.17 0.82 0.81 0.91 0.96 1.00
15 0.26 0.37 0.47 0.55 0.63 0.70 0.76 0.81 0.86 0.90 0.95 1.00
20 0.23 0.35 0.45 0.54 0.61 0.68 0.75 0.80 0.85 0.90 0.95 1.00
Checkpoint
Threaded Stud
8.9 14.1 21.0 39.7 59.9 86.8
Grade 5.8 Carbon Steel
Threaded Stud
12.7 23.9 34.7 64.6 100.8 145.2
A4/316 Stainless Steel
Step 5b Reduced characteristic ultimate bolt steel shear capacity, OV5f (kN)
Not appropriate for this product.
Check / OVur 1,
if not satisfied return to step 1
® Ramset
Concrete
20
STEP
® Ramset
Concrete
*
PRODUCT Allows for close anchor spacing between
Chemset Injection 800 Series is a chemical anchor fixing parts.
*
system based on epoxy mortar. The two parts are Integrated system, no need to mix
*
dispensed and mixed in one action through a static Metered system reduces waste.
*
mixing nozzle, which allows accurate mixing with no For a partially used cartridge, simply change
mess. nozzle and re-use.
*
High strength.
*
DESCRIPTION Excellent chemical resistance.
*
Chemset 800 Series injection mortars are tough, Low shrinkage — allows for large anchors to
corrosion resistant products suitable for anchoring be installed.
*
jobs close to edges where there is a need to avoid Low odour.
*
bursting stress on the surrounding substrate. Non-flammable for transport and storage
INSTALLATION INSTRUCTIONS:
Read the aPrecautionso section of these instructions 4. Reinforcing bars, internally threaded sockets,
prior to use. - threaded rods or studs to be used should be cleaned
and free from oil, grease, flaking rust or debris.
Setting and technical data provided applies to: Threaded rods or studs should be chisel ended to
a Holes drilled o ith Rotary Hammers using drill prevent them unthreading from the cured mortar.
bits toleranced in accordance a ith Dd\18035,
and a here holes have been cleaned using a 5. Ensure that holes are dry. d holes have been left for
brush and air pump. a prolonged period since drilling, re-cleaning in
o Holes cored a ith diamond coring eauipment that accordance El ith 12 a 3oabove is recommended.
have been cleaned using a brush and air pump.
a AID ays El ear safety glasses or goggles.
® Ramset
Concrete
I
8. open the orange valve.
(Turn so arroo points
fora ard). the handle 13.Do not touch the anchor until mixture
so the mortar is dispensed has gelled and do not load anchor
out of the until an until curing is complete as per Chart
even, uniform light grey No.1.
colour is achieved.
(Approximately 5-6 trigger pulls should be adeauate).
The initial mortar floc is unsuitable for fastening and
must not be used. This initial floa should be disposed
of into empty packaging or similar materials.
Ramsew Fasteners (Aust) Pty aimited does not give any representation, guarantee or a arranty, express or implied that the information
is or a ill be complete or accurate or that it has been independently verified. To the extent permitted by statute, Ramseto Fasteners
(Aust) Pty aimited, its servants and its agents a ill not be liable (a hether in negligence or other tort, by contract or under statute) in
respect of any loss or damage (including consenuential loss or damage) arising by any reason of or in connection a ith the provision of
the information or by purported reliance on it.
20
I-
C) -
10 CC
10
5-
.= 5=
= =
C#30 Cl)
C)
40
C) p'. 1 11111 1 11
- - =
—
ill
- CC c...
C) '?'.'' ———\- -8(U——- - -—
I-
20 E ————s — ————— — ——
C) — — — C) S
————— ————— — ——
—— —
5= — — — — — — — — — — — — —
=
Cl)O
0 10 2030 40 50 60 70 80 90100110120 0 5 10 15 20 2530 35 40 45 50 55 60 65 70
Minutes Hours
Note: For ambient temperature beloo the mortar must be conditioned to a minimum of See precautions, page 2.
ANCHOR STUDS
CHART 2 - SOLoD SuBSTRATE uNSTALLATnON DETAOLS
Anchor Thread Drill Hole Hole Maximum Maximum Tightening Minimum Minimum Minimum
Sine Sine Sine Cleaning Depth Fixture Fixture Tornue Edge Anchor Structural
(mm) Brush in Clearance Thickness (Nm) Distance Spacing Thickness
Sine Substrate n (mm) (mm) (mm) (mm)
(mm) (mm) (mm)
8 M8 10 13 80 11 15 8 30 50 100
10 M10 12 13 90 13 25 15 40 60 120
12 M12 14 13 110 15 30 30 50 70 140
16 M16 18 20 125 19 40 70 65 100 160
20 M20 24 20 150 25 75 140 80 120 190
24 M24 26 26 160 30 105 230 95 145 200
CHART n aER
10 12 90 62 132 1 1
12 14 110 41 86 2 1.5
16 18 125 26 54 2 2
20 24 150 10 22 6 5
24 26 160 11 24 6 5
For shear loads acting ton ards an edge(s) of the concrete the above capacities are not applicable.
Contact a Ramset technical consultant for advice.
DESCRIPTION Sika CarboDur is a heavy-duty strengthening system for reinforced ôoncrete and tim-
ber. It consists of two components: Sikadur-30 adhesive for bonding reinforcement
and Sika CarboDur-laminates.
USES The Sika CarboDur System can be used to strengthen reinforced concrete and timber
structures due to:
— Loading increases.
— Damage to structural parts.
— Serviceability improvements.
— Change in structural system.
. — Design or construction defects.
STORAGE AND When not exposed to direct sunlight, Sika CarboDur-laminates have unlimited shelf
SHELF LIFE life. Sikadur-30, when stored in the original sealed containers w ithin t he temperature
range of +5°C to ÷25°C will ke ep for a minimum of three years.
Mixing
Sikadur-30 is supplied in factory proportioned units com-prising the correct quantities
of Part A (Resin) and Part B (Hardener). Thoroughly stir both components separately
using a slow running drill/stirrer with a helical paste mixer (max. speed 600rpm).
Decant all Part B into Part A and mix thoroughly together until a uniform colour is
achieved (typically 3 mins). A streaky colouration is indicative of inadequate or incom-
plete mixing. Apply immediately. Small units may be hand mixed provided an even
colour is achieved.
Application
If any necessary patching work needs to be done on the surface, this must be done
with Sikadur-41, on the day preceding the actual bonding operations.
Page 1 of 4
Apply the well mixed Sikadur-30 adhesive carefully to the prepared substrate with a
spatula to form a first layer of at least 1 mm. Place the Sika CarboDur-laminate on a
table and clean it with Sika® Colma Cleaner. Apply the Sikadur-30 adhesive across the
width of the laminate, ensuring a total coverage. The adhesive should be minimum 1
mm deep at each side, to minimum 2 mm deep at the centre. This is best done using
a plastic spatula shaped on site to achieve this profile.
Within the open time of the adhesive, depending on temperature, place the Sika
CarboDur-laminate onto the concrete surface. Using a roller, press the laminate into
the epoxy material until the adhesive is squeezed out on both sides of the laminate.
Remove surplus epoxy adhesive. Samples should be made up on site to check the
adhesive used in respect of curing rate and final strength. Measure the compressive
bending and adhesive strength after curing. As a final check, test the laminates for
drum miness by tapping lightly.
There is no need for mechanical equipment to press strips onto the substrate nor is it
necessary to provide clamps or supportive devices to keep overhead strips in place.
Once cured the top of the laminates can be painted with a coating material such as
Sikagard-62, Sikagard-670W or Sikagard-680S.
Cleaning
Clean tools immediately with Sika® Colma Cleaner. Wash hands and skin thoroughly in
warm soapy water. Cured material can only be removed mechanically.
Colour Black
- . . -.
- Sika® CárboDur
Page 2 of 4
________
100
— +35C
80
60
40
20
L
Hours
0 50 100 150 200 250 300 350
Hours .0
0 0 150 200 250 300 350
20
::
;;
Hours
0 50 100 150 200 250 300 350
Sika® CarboDur
Page 3 of 4
DESIGN NOTES • A Sika CarboDur-laminate has no plastic deformation reserve.
• Maximum bending resistance of a strengthened section is reached when laminate failure occurs
during steel yield and before concrete failure.
• Mode of failure is influenced by the laminate cross section.
• To limit crack widths and deformation the yield point should not be reached in the reinforcing bars
under service conditions.
• Any shear cracks which occur must be prevented from causing displacement on the strengthened
surface and shearing of the laminate.
• Stress and deformation calculations can be made by the normal methods.
• When assessing the condition of the structure, look at the dimensions, quality of existing
construction materials, climatic conditions and agreed conditions of service.
• Design aspects to be verified in a Sika CarboDur Strengthening System:
Bearing safety — non strengthened structure (total safety factor g=1 .2)
— strengthened structure (mode of failure mentioned above, check on strains)
— shearing of laminate must not occur
— anchorages
Fatigue resistance — check concrete and steel stresses
Serviceability — deformation (with average strains, elasti behaviour of the structure and
time-based strain changes in concrete)
— steel stresses (no plastic deformation in service conditions)
— limit crack widths (by limiting the steel stresses to less than the yield point
under service conditions).
A full Sika Design Manual is available. Contact our Technical Department for details.
IMPORTANT NOTES • Do not apply Sikadur-30 to surfaces with standing water. Maximum moisture content of the con-
crete 10%.
• Always mix a full kit to avoid mix ratio error.
• Only mix as much material as can be applied within the stated potlife.
• Do not dilute the product with solvent as this will affect the cure and in-service performance.
• Constant exposure to service temperatures >50°C may affect the performance of the product.
• Sika CarboDur can be firerated if required using standard fire rating materials.
• The temperature at which the Sikadur-30 is stored at during the 24 hours before it is mixed will
govern its potlife when mixed.
.
• For shear strengthening, refer to SikaWrap technical data sheets.
• Compressive strengths, etc., of epoxy resins must be qualified by the testing method, eg. Test Standard
or size of specimen under test and the rate at which the test piece is loaded under test, as these
factors will affect the result markedly. Faster loading rates will generally give higher ultimate loads
and vice versa. Also, a specimen at lower temperature will show higher strengths and vice versa.
• Sikadur-30 Parts A and B are a water pollutant and should not be discharged into drains, waterways
or soil.
HANDLING • Avoid contact with the skin, eyes and avoid breathing its vapour.
PRECAUTIONS • Wear protective gloves when mixing or using.
• If poisoning occurs, contact a doctor or the Poisons Information Centre.
• If swallowed, do NOT induce vorniting. Give a glass of water.
• If skin contact occurs, remove contaminated clothing and wash skin thoroughly.
• If in eyes, hold eyes open, flood with water for at least 15 minutes and see a doctor.
IMPORTANT The information, and, in particular, the recommendations relating to the application and
NOTIFICATION end-use of Sika products, are given in good faith based on Sika's current knowledge
and experience of the products when properly stored, handled and applied under normal
conditions. In practice, the differences in materials, substrates and actual site conditions are
such that no warranty in respect of merchantability or of fitness for a particular purpose, nor
any liability arising out of any legal relationship whatsoever, can be inferred either from this
information, or from any written recommendations, or from any other advice offered. The
proprietary rights of third parties must be observed. All orders are accepted subject to our
current terms and conditions of sale. Users should always refer to the most recent issue of the
Technical Data Sheet for the product concerned, copies of which will be supplied on request.
PLEASE CONSULT OUR TECHNICAL DEPARTMENT FOR FURTHER INFORMATION.
Sikadur®-30
Thixotropic adhesive for bonding reinforcements
Description Sikadur-30 is a thixotropic adhesive mortar based on a 2-component solvent
free epoxy resin. Sikadur-30 is used primarily to bond structural
reinforcements to other substrates. It can also be used to bond and fill a wide
variety of building and construction materials.
Storage and Shelf Life Minimum shelf life is approximately two (2) years. Store under controlled
conditions in original containers (minimum 5°C, maximum 35°C temperature
range).
Sikadur®-30
Page 1 of 4
— —...- -
—.——- -j
Mixing (continued) thoroughly until a uniform colour is achieved (typically 3 mins.). A streaky
colouration is indicative of inadequate or incomplete mixing. Apply
immediately. Small units may be hand mixed provided an even colour is
achieved.
Application When using Sikadur-30 as a structural reinforcement adhesive, apply well
mixed epoxy to the substrate, at maximum 1mm thickness, using a spatula,
trowel, or float. Adhesive should then be applied to the Sika CarboDur
laminate or steel plate in a roof shape (minimum 1mm thickness at the sides,
2mm in the centre).
Steel plates must be fixed to the concrete substrate using specially prepared
supports.
Sika CarboDur laminates are pressed into the substrate with hard rubber
rollers by hand. It is important that laminates are evenly pressed onto the
substrate until the adhesive is squeezed from both sides of the laminate.
Cleaning Uncured material may be cleaned from application tools, etc. by using Sika
Colma Cleaner (flammable solvent). Cured material can only be removed
mechanically.
Technical and Physical Data
Appearance Part A White paste
Part B Black paste
Part A + B Light grey when mixed
Mix Ratio A: B 3 : 1 (parts by weight & volume)
Density 1.77 kg/L (A + B)
Pot Life * 40 minutes (at 35°C)
Open Time* 30 minutes (at 35°C)
Sag Flow* 3 - 5mm (at 35°C)
Shrinkage 0.04%
Glass Transition Point* 62°C
Static 12,800 MPa
Adhesive Strength* Concrete: 4MPa (Concrete Failure)
Steel 33MPa (Sandblasted Substrate)
Shear Strength* 15 MPa (Concrete Failure)
Coefficient of Expansion 9x1 per °C (-10°C to 40°C)
Packaging 5 kg tins Part A 3.75 kg
Part B 1.25 kg
*10 F.I.P Federation Internationa le de Ia Precontrainte
Note: The values given may vary according to amount of air extrained during mixing.
40.at 0
0 '50 200 250 3 0 3
I
+35.C
•33•C
,
0 50 100 1150 200 250 300 350
Sikadur®-30
Page 2 of 4
Important Notes • Do not apply Sikadur-30 to surfaces with standing water. Maximum
moisture content of the concrete 10%.
• Always mix a full kit to avoid mix ratio errors.
• Only mix as much material as can be applied within the stated potlife.
- It is important to entrain as little air as possible when mixing Sikadur-30.
Air entrainment will result in lower compressive strength development.
• Do not dilute the product with solvent as this will affect the cure and
service performance.
- Constant service temperature >50°C may affect the performance of this
product.
- Sikadur-30 can be heated up to 70°C during curing to speed up the
setting time.
If in doubt, consult our Technical Department.
• Compressive strengths etc. of epoxy resins must be qualified by the
testing method e.g. Test Standard or size of specimen under test and the
rate of which the test piece is loaded while under test, as these factors
will effect the result markedly. Faster loading rates will generally give
higher ultimate loads and vice versa. Also, a specimen at lower
temperature will show higher strengths and vice versa.
Handling Precautions • Avoid contact with the skin, eyes and avoid breathing it's vapour.
• Wear protective gloves when mixing or using.
• If poisoning occurs, contact a doctor or Poisons Information Centre.
• If swallowed, do NOT induce vomiting. Give a glass of water.
• If skin contact occurs, remove contaminated clothing and wash skin
.
thoroughly.
• If in eyes, hold eyes open, flood with water for at least 15 minutes and
see a doctor.
• For more detailed information refer to Material Safety Data Sheet.
Important Notification The information, and, in particular, the recommendations relating to the
application and end-use of Sika's products, are given in good faith based on
Sika's current knowledge and experience of the products when properly
stored, handled and applied under normal conditions. In practice, the
differences in materials, substrates and actual site conditions are such that no
warranty in respect of merchantability or of fitness for a particular purpose,
nor any liability arising out of any legal relationship whatsoever, can be
inferred either from this information, or from any written recommendations, or
from any other advice offered. The proprietary rights of third parties must be
observed. All orders are accepted subject of our terms and conditions of
sale. Users should always refer to the most recent issue of the Technical
Data Sheet for the product concerned, copies of which will be supplied on
request.
PLEASE CONSULT OUR TECHNICAL DEPARTMENT FOR FURTHER
INFORMATION.
Page 3 of 4
}
r --—
A U 0
®
Sika Australia Pty Limited
ABN 12 001 342 329
www.sika.com.au
55 Elizabeth Street
Wetherill Park NSW 2164
Tel: (02) 97251145
Fax: (02) 9725 2605
Sikadur®-30
Page 4 of 4