Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
DOCTOR OF PHILOSOPHY
in
CIVIL ENGINEERING
Approved by:
___________________________________________ __ _________________
Luis E. Suárez, Ph.D. Date
Member, Graduate Committee
_____________________________________________ _________________
Felipe J. Acosta, Ph.D. Date
Member, Graduate Committee
_____________________________________________ _________________
Ricardo R. López, Ph.D. Date
Member, Graduate Committee
_____________________________________________ _________________
Daniel A. Wendichansky, Ph.D. Date
President, Graduate Committee
_____________________________________________ _________________
Basir Shafiq, Ph.D. Date
Representative of Graduate Studies
_____________________________________________ _________________
Ismael Pagán Trinidad, M.S.C.E. Date
Chairperson of the Department
UMI Number: 3280290
Copyright 2007 by
Velez Velez, Edgardo M.
This research presents a study of the lateral capacity of the reinforced concrete
houses constructed in Puerto Rico. These houses are built with a reinforced concrete roof
slab and three reinforced concrete walls oriented in the same direction. The lateral loads
applied in the direction perpendicular to the reinforced concrete walls are resisted by the
frame action of the walls and roof acting in the weak direction of the walls and by
The experimental part of the study was focused in testing six full scale one story
specimens comprised each of three reinforced concrete walls monolithic with the roof
slab subjected to cyclic loading in the weak direction of the assembly. Five of the
specimens contained concrete block walls in one of the bays. The concrete block walls
were constructed under different configurations of the steel reinforcement and openings.
The results demonstrated that the lateral resistances of the specimens were governed by
different failure modes; among of them are: plastic hinges at the R/C wall ends,
punching shear failure of R/C walls and roof slab, masonry corner crushing, toe crushing
and shear failure of the piers. The lateral capacity and the energy dissipation of the
specimens increased significantly when partition wall was included inside the frame.
The analytical part of the research included the development and the validation of
simple analytical models of the full scale specimens. Once validated, the analytical
models of each component of the house were used to construct a full house computer
model. The full house computer model was used to perform seismic response predictions
of the typical residential house subjected to a four earthquake motions in its weak
ii
direction. The results showed that this structure can withstand the seismic demand
imposed by Uniform Building Code 1997 for Seismic Zone 3 and Soil Type Sd in the
elastic regimen. Severe damage is expected when the earthquake records where scaled to
peak ground acceleration (PGA) of 1.0g; 2.77 times the PGA established by UBC-1997.
iii
RESUMEN
típicamente construidas con una losa de techo y tres paredes de hormigón armado
perpendicular a estas paredes son resistidas por la acción de marco de las paredes y el
techo en su eje débil y las paredes divisorias, las cuales son construidas con paredes de
bloques.
monolíticamente con la losa de techo, las cuales fueron sometidas a cargas cíclicas
laterales en su dirección débil. Cinco de las muestras fueron construidas con una pared
de bloques de concreto en una de las luces. Las paredes de bloques fueron construidas
falla, tales como, articulaciones plásticas en los extremos de las paredes de hormigón
para realizar las predicciones de respuesta sísmica de estas estructuras sometidas a cuatro
estructuras pueden resistir la demanda sísmica impuesta por Código UBC-1997 para una
Zona Sísmica 3 y suelo tipo Sd en el rango elástico. Se espera daño severo en estas
estructuras cuando los registros de los terremotos se escalaron a 1.0 g, 2.77 veces la
v
To God
To my parents, Adelina and Manuel
To my sister and brother, Vanessa and Emmanuel
To my fiancée, Igdali
vi
ACKNOWLEDGMENTS
The work presented in this research would not have been possible with out the
my mentor, his advice was not only limited to technical field; also help me to grow in the
professional and personal areas. I would also like to express my appreciation to the
graduate committee members, Dr. Ricardo López, Dr. Felipe Acosta and Dr. Luis Suárez,
I would like to thank Luis Montejo for provide me with the artificial earthquake
the Civil Engineering Department of the University of Puerto Rico at Mayagüez for their
Cuadrado, Jaffet Martínez, Juan C. Morales, Augusto Poitevin, and John Vera for their
Rolando García, Jorge Ayala, Elvis Ramos and Iván Santiago, for their friendship,
support and valuable collaboration during the construction, and carrying out of the
I would like to thank the staff of the Civil Engineering Department: Monserrate
Cruz, Miguel Bonilla, Miguel Lugo, Edgar Matías, Samuel Del Valle, Miguel Báez, Justo
vii
I would like to thank my parents, sister, brother and fiancée for their
comprehension and unconditional support in all difficult moments that I cannot share
with them.
At last but not less important to GOD for giving me wisdom, patience and
viii
TABLE OF CONTENTS
CHAPTER 1 INTRODUCTION
1.1 BACKGROUND 1
1.2 JUSTIFICATION OF THIS RESEARCH 5
1.3 OBJECTIVES OF THIS RESEARCH 5
REFERENCES 152
xi
LIST OF TABLES
xiii
LIST OF FIGURES
CHAPTER 1 INTRODUCTION
Figure 1-1: Fault Map of Puerto Rico and neighboring Islands. 2
Figure 1-2: Typical one and two story residential houses constructed
in Puerto Rico. 4
xiv
CHAPTER 4 EXPERIMENTAL RESULTS
Figure 4-1: Load-displacement hysteresis curve for Specimen 1. 51
Figure 4-2: Lateral load normalized with respect to the Weff of
Specimen 1. 51
Figure 4-3: Strain distributions for Specimen 1 loaded in the push
direction. 52
Figure 4-4: Details of joint between the roof slab and the exterior
wall 3. 53
Figure 4-5: Hypothesized plastic mechanisms for Specimen 1. 53
Figure 4-6: Failure patterns of Specimen 1. 54
Figure 4-7: Load-displacement hysteresis curve for Specimen 2. 57
Figure 4-8: Lateral load normalized with respect to the Weff of
Specimen 2. 57
Figure 4-9: Strain distributions for Specimen 2 loaded in the pull
direction. 58
Figure 4-10: Failure patterns of Specimen 2. 59
Figure 4-11: Load-displacement hysteresis curve for Specimen 3. 62
Figure 4-12: Lateral load normalized with respect to the Weff of
Specimen 3. 62
Figure 4-13: Strain distributions for Specimen 3 loaded in the pull
direction. 63
Figure 4-14: Failure patterns of Specimen 3. 64
Figure 4-15: Load-displacement hysteresis curve for Specimen 4. 67
Figure 4-16: Lateral load normalized with respect to the Weff of
Specimen 4. 67
Figure 4-17: Strain distributions for Specimen 4 loaded in the push
direction. 68
Figure 4-18: Failure patterns of Specimen 4. 69
Figure 4-19: Load-displacement hysteresis curve for Specimen 5. 72
Figure 4-20: Lateral load normalized with respect to the Weff of
Specimen 5. 72
Figure 4-21: Strain distributions for Specimen 5 loaded in the positive
direction. 73
Figure 4-22: Failure pattern of Specimen 5. 74
Figure 4-23: Load-displacement hysteresis curve for Specimen 6. 78
Figure 4-24: Lateral load normalized with respect to the Weff of
Specimen 6. 78
Figure 4-25: Strain distributions for Specimen 6 loaded in the positive
direction. 79
Figure 4-26: Failure pattern of Specimen 6. 80
Figure 4-27: Hysteresis envelopes for the six specimens. 84
xv
CHAPTER 5 SIMPLE ANALYTICAL MODELS
Figure 5-1: Diagonal Strut Model. 86
Figure 5-2: Behavior of infilled frames under lateral load. 91
Figure 5-3: Assumed punching shear mechanism. 94
Figure 5-4: Compression strut analogy in perforated infill panel. 98
Figure 5-5: Failure of the unreinforced walls with piers. 99
Figure 5-6: Distribution of the interaction forces in confined masonry. 100
Figure 5-7: Analytical model for Specimen 1. 105
Figure 5-8: Element used for R/C walls and roof slab. 105
Figure 5-9: Moment-rotation relationship for the rotational springs. 106
Figure 5-10: Experimental and analytical results of Specimen 1. 106
Figure 5-11: Analytical model for Specimen 2. 108
Figure 5-12: Multi-linear backbone for the compression link element. 109
Figure 5-13: Experimental and analytical results of Specimen 2. 111
Figure 5-14: Analytical Model for Specimen 3. 114
Figure 5-15: Experimental and analytical results of Specimen 3. 116
Figure 5-16: Analytical model for Specimen 4. 117
Figure 5-17: Multi-linear backbone for the tension link element. 118
Figure 5-18: Experimental and analytical results of Specimen 4. 120
Figure 5-19: Analytical model for Specimen 5 and Specimen 6. 122
Figure 5-20: Experimental and analytical results of Specimen 5. 125
Figure 5-21: Experimental and analytical results of Specimen 6. 128
xvi
APPENDIX A SUMMARY OF THE RESIDENTIAL HOUSES DETAILS
SELECTED FOR THE SPECIMENS CONSTRUCTION
Figure A-1: Construction sequence of the residential houses studied. 161
Figure A-2: Typical details of concrete wall. 162
Figure A-3: Typical details of concrete block wall. 163
Figure A-4: Typical joint details between foundation and concrete
wall. 164
Figure A-5: Typical joint details between foundation and concrete
block wall. 165
Figure A-6: Typical joint details between roof and interior concrete
wall. 166
Figure A-7: Typical joint details between roof and exterior concrete
wall. 167
Figure A-8: Typical details of roof slab. 168
Figure A-9: Typical details of foundation slab. 169
xvii
1
CHAPTER 1
INTRODUCTION
1.1 BACKGROUND
Island is located in the limit between the North America and the Caribbean plates. These
two plates are constantly moving creating a lateral, left strike slip fault zone. Figure 1-1
shows the eight seismic activity zones: Puerto Rico Trench to the north, Slope faults in
the North and South of Puerto Rico, Northeast of “Zona del Sombrero”, Mona Canyon to
the west, Mona Passage, Virgin Islands and Anegada depressions to the east, “Muertos”
Throughout the years, several strong earthquakes have shaken Puerto Rico since
the beginning of its colonization. In August 15, 1670, an earthquake caused destruction
in the San German and San Juan municipalities. In May 2, 1787, another earthquake
caused damage and destruction in most of the Island. This earthquake demolished the
Arecibo church along with the El Rosario and La Concepción monasteries and damaged
the churches at Bayamón, Toa Baja and Mayagüez. It also caused considerable damage
to the castles of San Felipe del Morro and San Cristobal, breaking cisterns, walls and
guard houses. In November 18, 1867, 20 days after the Island was devastated by
magnitude of 7.5 on the Richter scale. The epicenter was located in the Anegada Passage,
between Puerto Rico and St. Croix, Virgin Islands. The earthquake produced a tsunami
2
that ran inland almost 150 meters (490 feet) in the low parts of the coast of Yabucoa. This
earthquake caused damage in numerous buildings on the Island, especially in the eastern
zone. In October 11, 1918, a 7.5 magnitude (on the Richter scale) earthquake, whose
epicenter was located northwest of Aguadilla in the Mona Canyon, was accompanied by
a tsunami which got up to 6 meters (19.5 feet) high. The earthquake killed about 116
people and caused more than 4 million dollars of damage. Numerous houses, factories,
public buildings, chimneys, bridges and other structures suffered severe damage. The
destructive earthquakes. Since the last strong earthquake occurred, 89 years have passed
and thus currently there is a high risk that a severe shaking may occur.
increase when the structural system used to withstand the seismic loads have not been
designed following an appropriate design philosophy. The 1990 Census of Puerto Rico
reported that 75.20 percent of the residential houses were built using a combination of
concrete walls and concrete roof. A large number of these houses are constructed with
reinforced concrete walls oriented primarily in one direction, as seen in Figure 1-2. This
orientation is referred to as the strong direction. Masonry walls are then oriented
perpendicular to the concrete walls. In the strong direction, the reinforced concrete walls
provide adequate structural capacity for resisting the inertial loads generated during a
seismic event. In the weak direction, the masonry walls are the only system available to
resist the inertial loads in addition to the reinforced concrete walls and roof acting in their
weak direction. The design philosophy used in the Island and the construction methods
solid panels, panels with window and/or door openings, and without the seismic details
established by the design codes, such as the Uniform Building Code 1997. The in-plane
capacity of these masonry walls has not been clearly established, nor the lateral capacity
of residential houses in the weak direction. Since the only reliable way to obtain the
this research is focused in carrying out experimental tests to full scale model of the
residential houses and determining their structural capacity in the weak direction when
a) One Story
b) Two Story
Figure 1-2: Typical one and two story residential houses constructed in Puerto Rico.
5
1.2 JUSTIFICATION OF THIS RESEARCH
Typically, residential reinforced concrete walls are infilled with masonry walls in
their weak direction, which are commonly used as partition walls. Normally, designers
consider the masonry concrete block walls as non structural elements during the design of
these residential houses. Therefore, interaction of the masonry infill walls with its
surrounding elements during the seismic events is neglected. In seismic areas such as
Puerto Rico, ignoring the composite action is not always on the safe side, since the
interaction between the infill panel and the surrounding elements under lateral loads
changes dramatically the stiffness and the dynamic characteristics of the composite
structure and consequently, its response to seismic loads. In order to obtain information
about how resistant are these structures in their weak direction, this investigation is
focused in conducting tests to full scale specimens to investigate the behavior of the
infilled frames under cyclic loading and study the possible scenarios of failure.
The goal of this research is to experimentally establish the lateral capacity in the
weak direction of the typical residential houses constructed in Puerto Rico. Specifically,
develop and calibrate simple analytical models of the full scale model,
LITERATURE REVIEW
2.1 INTRODUCTION
the last five decades to investigate the in-plane seismic response of masonry infilled
frames. The experimental investigations have been carried out to examine the responses
behavior of the infilled frames during experimental tests and seismic events. The first
part of this section presents a series of field observations of the seismic response of
masonry infilled frames during various earthquakes, while the second part presents
Miranda and Bertero (1989) concluded that the good performance of the low- rise
reinforced concrete building during the 1985 Mexico City earthquake was due to the
presence of masonry infills. The masonry infills that were adequately placed in the low-
rise frame structures resulted in significant increases in strength and stiffness which help
Flanagan et al. (1996) investigated the performance of masonry infills during the
Northridge earthquake. They found that the collapse of the masonry infills during the
8
earthquake was due to diagonal cracking, cracking around the infill perimeter, and corner
crushing. However, in spite of the damage of the masonry infill, the buildings remained
Tezcan and Ipek (1996) found that the masonry walls of three and four story
houses collapsed immediately during the 1995 Dinar, Turkey earthquake. The masonry
houses constructed using bearing walls made of hollow core brick tiles collapsed due to
insufficient wall rigidities, improper wall thicknesses, and wall openings. However,
when the bearing walls were constructed from solid brick walls or stones, and they were
one or two-story high, the masonry buildings survived the earthquake with minor cracks.
Sezen et al. (2000) analyzed the role of infill walls in the response of moment-
frame buildings during the 1999 Kocaeli, Turkey earthquake. Their found that many
buildings collapsed due to soft stories. Irregular placement of infill masonry walls
produced stiffness discontinuities, which concentrated the deformation in the first story,
Humar et al. (2001) investigated the performance of buildings during the 2001
Bhuj earthquake in the Kachchh region of the province of Gujarat in India. They
concluded that the presence of masonry throughout the height of the buildings prevented
the collapse of many buildings even though such infills were neither reinforced nor
ductile reinforced concrete frames under quasi-static loads simulating the effects of
strong earthquake ground motions. The experimental tests were conducted in bare frame
infilled with clay blocks and bare frame infilled with concrete blocks. The experimental
results showed that throughout all deflection ranges, the infilled frames dissipated two
Dawe and Seah (1989) examined the effect of a doorway in the panel on the
stiffness and strength of the infilled frames. They concluded that if the opening must
interrupt the compression diagonal, it is preferable that it placed so that as much panel
material as possible is between it and the point of load. This gives a diagonal strut effect
an opportunity to develop in that part of the panel. They recommend that the best
Abrams and Paulson (1991) carried out experimental tests of one-quarter scale
three-story test structures consisted of two perforated flanged walls, which resisted lateral
forces parallel with their primary plane. The walls of the first model were perforated
with a symmetrical pattern of window openings, whereas walls of the second model
consisted of an asymmetrical pattern of door and window openings. For both structures,
the experimental results showed that the ultimate limit state was characterized by
first frame was designed to resist wind loads and the second frame was designed to resist
strong earthquake forces. The experimental tests were performed using twelve one-half-
scale, single-story, and single bay frames. The tests focused in investigating the strength
of infill panels with respect to the bounding frame, the panel aspect ratio, and the
distribution of vertical load. The experimental results indicated that the specimens with
strong frames and strong panels exhibited a better response that those with weak frames
and weak panels in terms of load resistance and energy dissipation capability. They also
indicated that infill panels can be potentially used to improve the performance of the
Mosalam et al. (1997) carried out several experimental tests of gravity load
designed steel frames with semi rigid connections, infilled with unreinforced masonry
walls subjected to cyclic lateral loads. The experimental tests were conducted to evaluate
the effects of the relative strength of the concrete blocks and the mortar joints, the
number of bays and the opening configuration in the infill on the performance of single
story reduced scale infilled frames. The experimental results demonstrated that the
compressive strength of the concrete blocks determines the mode of failure, such as
corner crushing or mortar cracking of the infill panels. Also the ultimate load for the two-
bay specimen was about double the capacity of the single-bay specimen, and the presence
of openings reduced solid infill stiffness values by 40 percent while the presence of a
door opening in one wall reduced the peak capacity by about 20 percent.
Negro and Colombo (1997) performed pseudo dynamic tests to a full scale four-
nonstructural masonry panels in framed buildings. The tests were conducted on the same
full-scale four-story frame, with three different infill configurations. The first test was
conducted on the bare frame. The second test was performed by infilling the frames with
hollow brick masonry in all four stories, developing a uniform infill distribution. In the
third test, a soft-story effect was created by excluding the infill at the first story. The
results of the pseudodynamic tests demonstrated that irregular distributions of infills can
result in unacceptably high ductility demands in the frame. However, the uniformly
infilled structure showed that a regular distribution of infills may result in a globally
irregular behavior of the frame. The regular distribution of infills corresponded to the
formation of a series of partial story-level sideway mechanisms after the failure of the
plane seismic behavior of steel frames with unreinforced masonry infills having large
window openings. Five large-scale steel-frame masonry infill specimens were subjected
to in-plane lateral deformations at the floor level. Test parameters included the pier
width of the infill between the steel column and the window opening, and the number of
imposed drift of 0.20 percent, the effective stiffness deteriorated to about 30 percent of
the initial stiffness. Amplitudes of drift larger than 0.75 percent produced excessive
splitting and crushing of the bricks in the masonry infill. The stiffness continued to
deteriorate uniformly until no stiffness remained. This occurred at 2.0 percent drift.
12
Buonopane and White (1999) carried out pseudo dynamic tests to a half-scale
specimen of a two-story, two-bay reinforced concrete frame infilled with masonry. The
second-story infill included window openings. The experimental results showed that the
difference in crack patterns and the associated hysteretic behavior between the two stories
suggests different strut mechanisms to capture each type of behavior. The first-story
infill behavior was characterized by a main diagonal strut at low force levels. For higher
force levels, bed joint sliding reduces the effectiveness of main diagonal struts. In the
second story, the window openings forced compressive struts to form at steep angle to
pass below the openings. The bed joint cracking occurred above or below the window
capacity and deformation behavior of five infilled specimens subjected to monotonic load
and a substantial drift (9%). The five half scaled models were all single-story, non
ductile reinforced concrete (RC) frames of single, double or triple-bay construction. The
experimental results demonstrated that the modes of failure, such as shear or corner
crushing of the infill frames, are driven by the compressive and shear strength of the
panels.
Lee and Woo (2002) studied the response of masonry-infilled RC frames with
non-seismic detailing under the simulated earthquake ground motions. The tests were
performed for 1:5 scale models composed of two bay and three stories. The experimental
results demonstrated that the masonry infill benefits the buildings because the amount of
increase in earthquake inertial force appears to be relatively small when compared with
13
the increase in the strength of the masonry infill. Also, the masonry infill appears to have
The analytical investigations were made using different approaches: the macro-
models and micro-models. The macro-models are based on simple analytical models
such as equivalent frame and equivalent strut. The micro-models are based on the theory
of elasticity, equilibrium, and energy approach, plastic analysis and lately the finite
element method.
2.3.2.1 Macro-models
Stafford (1967) developed the equivalent strut concept to predict the lateral
stiffness and strength of multi-story infilled frames. The equivalent strut concept was
developed by considering that the frame members are rigidly connected to each other and
that the infills which are not bonded to the frame and are made of homogenous and
isotropic material. In the equivalent strut concept, the structure is modeled as a braced
frame where the infill walls are replaced by equivalent pin-jointed diagonal strut. As a
the effective width of the diagonal strut based on the relative stiffness of the infill and
frame. Also, he concluded that the lateral load that produces a compressive failure of the
infill depends on the relative stiffness of the columns to the infill, and it is independent of
Liauw (1972) presented the concept of the equivalent frame for the analysis of
infilled frames with or without opening. This concept was developed by transforming the
14
infilled framed in to an equivalent frame whose members have the properties of the
composite sections of the actual structure. The analytical results obtained by the
equivalent frame concept were compared with the experimental results obtained from an
elastic model experiment. The comparison between the experimental and analytical
results showed a good agreement when the openings are more than 50 percent of the full
infill area. When the openings are less than the 50 percent of the full infill area, the
Sobaih and Abdin (1988) used a concept of equivalent strut for the linear analysis
implemented in the computer program SAPF (Seismic Analysis of Plane Frame). The
program was used to simulate 13 cases of bare and infilled frames in order to investigate
the effect of different factors such as the presence and continuity of infill panels, the
height of the structure, infill material, panel rectangularity ratio, and width of the
equivalent strut. The results showed that infill panels increase the stiffness of the
structure and the stresses on columns, but decrease the lateral displacement of the frame.
diagonal strut approach for the analysis and design of steel frames with concrete or
masonry infillings walls subjected to in-plane forces. The method takes into account the
elastic and plastic behavior of infilled frames considering the limited ductility of infill
materials. The method provides a rational basis for predicting the lateral strength and
stiffness of infilled frames as well the infill diagonal cracking load. Various governing
factors such as the infill aspect ratio, the shear stress at the infill-frame interface and
relative beam and column strength are accounted for in this development.
15
To represent masonry infill panels in nonlinear analysis of frame structures, an
equivalent strut integrated with a smooth hysteretic model was proposed by Reinhorn et
al. (1995) and Madan et al. (1997). The model is based on an equivalent diagonal strut
with a hysteretic force-deformation that includes the strength and stiffness degradation as
well as pinching resulting from opening and closing of masonry gaps. The equivalent
strut model was implemented in the computer program IDARC Version 4.0. The
macromodeling approach does not permit to study local effects such as frame-infill
interaction within the individual infilled frame subassemblies. However, the proposed
approach allows for the evaluation of the nonlinear force-deformation response of the
and the lateral load capacity of concrete masonry-infilled steel frames failing in corner
crushing mode. The method consisted in replacing each masonry panels by three struts
infill. In order to determine the bending moments and shearing forces in the frame
members, a single diagonal strut is connected between the two loaded corners, and the
other two struts are located off-diagonal at the points of maximum field moments in the
beams and the columns. They concluded that three struts do not fail simultaneously,
which is the case in actual infill panels, because the crushing starts at the corners and
Perera (2005) proposed a damage model based on the equivalent strut for the
characterization of masonry walls subjected to lateral cyclic loads. The strut element is
formulated by using the concepts and principles of continuum damage mechanics. For
this, the axial force versus deformation relation is formulated through the effective stress
concept and the strain equivalence principle. Using this approach, a scalar damage
The damage variable considers the progressive decrease of the effective width of the
diagonal compression strut, due to the cracking occurring in the infill panel by tension
effects.
2.3.2.2 Micro-models
The nonlinear behavior of a steel frame and the concrete infill under monotonic
load using a finite element model was presented by Kwan and Liauw (1984). The infilled
frame structure was modeled using elements of interface, panel, and frame. The
analytical results showed that in the elastic range stress concentration occurs at all the
four corners. As cracks propagate, the stresses at the tensile corners are reduced, while
those near the compressive corners are increased. When the corners yield in
compression, the frame moments increase significantly, forming a plastic hinge and
A simple iterative finite element method was proposed by Achyutha et al. (1986)
to investigate the infilled frames with openings, and with or without stiffeners around the
openings. The method takes into consideration the separation, slip, and frictional loss at
the interface of the infill and the frame. The bounding frame members were represented
by prismatic beam elements having three degrees of freedom at each node. The
17
continuum infill panel was modeled by two-dimensional four-node rectangular plane
stress elements having two degree of freedom at each node. The interface between the
frame and infill was represented by short stiff beam elements having three degrees of
freedom at each node. The analytical results demonstrated that for cases of window
opening area greater than 50% of the solid infilled area the lateral stiffness of the infill
panels with openings can be neglected when compared to that of solid infilled frames.
May and Naji (1991) presented a nonlinear finite element program to simulate the
behavior of steel frames infilled with concrete panels subjected to monotonic and cyclic
loading. The infilled frame was modeled using panel elements, frame elements and
interface elements. The program was validated by using experimental results of the tests
conducted by May and Ma (1984), and Liauw and Kwan (1982). For the analysis of the
infilled frame under monotonic loading, the analytical results represented the different
modes of behavior observed experimentally. When the infilled frame was subjected to
cyclic loading, the analytical results showed a good agreement with the test results for the
first cycle of loading. For other cycles, the hysteretic loops predicted by the program
Mehrabi and Shing (1997) investigated the lateral load resistance of masonry-
infilled frames R/C frames using finite element models. The finite element models
considered the fracture behavior of R/C frames, masonry units, mortar joints, and frame-
panel interface. The tension and compression behavior of the concrete frame and
masonry units were modeled with smeared crack element, while the fracture of the mortar
joints, the separation of the frame-panel interface, and the shear cracks in the concrete
columns were modeled using interface elements. The models were validated with results
18
of experimental tests conducted by Mehrabi (1994) on half-scale, single-story, R/C
frames infilled with concrete masonry units. The analytical results showed an acceptable
correlation with the experimental results, thus allowing the use of the analytical models to
frames.
Ghosh and Amde (2002) presented a new finite element model for infilled frames,
in which the interface between the frame and the infill and the mortar joints surrounding
the blocks of masonry were simulated by using a non associated interface model based on
test data on masonry joints. The cracking in tension and plasticity in compression of the
infill were modeled by using smeared crack model and the plasticity model, respectively.
The finite element model was validated by comparison with the results of the
experimental tests carried out by other researchers. The results obtained by the finite
element model showed that the numerical models were capable of providing detailed
EXPERIMENTAL METHODOLOGY
3.1 INTRODUCTION
The response of infilled frames subjected to in-plane lateral loading has been
investigated by various researchers (Stafford 1967, Klinger 1977, Abrams and Paulson
1991, Mehrabi et al. 1996, Lee and Woo 2002) during the last decades. The results of
these investigations revealed that the failure mechanisms are governed by the relative
strength and stiffness of the infill with respect to the surrounding elements. However, all
these investigations were carried out in infill bounded by steel or reinforced frames
composed by beams and columns, neglecting the behavior of the infill panels surrounded
by reinforced concrete walls acting in their weak direction. The infill panels surrounded
by reinforced concrete walls are part of the structural system of the residential houses
This section presents the experimental program performed in six full scale
experimentally the overall lateral capacity of these structures composed of infilled frames
surrounded by reinforced concrete walls. The first part of this section presents the
selection and the construction of the test specimens, while the last part includes the
experimental setup, instrumentation, load pattern, and data acquisition system used
The behavior of the structural and non structural elements of the residential
houses was experimentally investigated using six specimens. Figure 3-1 shows a plan
view of a typical concrete house indicating the locations of the different specimens
studied. The construction details used in the specimens were based on the result of a
houses. During the survey study, special attention was given to the connections between
the footing and the roof with the reinforced concrete wall or with the masonry wall.
Also, the amount of horizontal and vertical reinforcement of the concrete and masonry
walls was considered. The results of this survey study are summarized in Appendix A.
The first specimen was a typical reinforced concrete one story wall-slab frame
with two bays. This specimen was set as control model to obtain information about the
overall lateral capacity of the slab-wall interaction. The dimensions of the structural
elements, steel reinforcement patterns, and connection details between the footing and the
roof with reinforced concrete wall for this specimen are shown in Figure 3-2. This
specimen has a width of 4 feet -1.25 inches, a clear height of 8.0 feet and a length of 24.0
feet center to center between the exterior walls. The thickness of the floor slab was 4.0
inches, while the concrete wall and roof slab thickness was 5.0 inches. The concrete in
2'-7"
3'-6"
5'-0"
Bedroom Bedroom
5'-0"
Specimen 1
5'-11"
Cl oset
Cl oset Specimen 4
9'-6"
Bathroom
2'-0"
Specimen 4
Cl oset
Specimen 4
36'-0"
5'-0"
5'-0"
Bedroom
2'-0"
Kitchen
4'-0"
Specimens 2&3
5'-0"
3'-0"
5'-0"
2'-0" 2'-0"
Living Room Dining Room
3'-0"
5'-0"
Specimens 5 & 6
3'-3" 5'-0" 3'-3" 1'-6" 5'-0" 1'-6"
LEGEND
Figure 3-1 : Plan view of a typical one story residential house constructed in Puerto
Rico.
12'-0" 12'-0"
Load Direction
3'-6" 7'-0" (Hydraul ic
#3@9" Act uator)
1'-6" #3@9"
5"
#3@9"
1'-6"
#3@9" #3@12"
#3@9"
#3@9" 5"
8'-0"
5'-0" Is olat ed by
A 4@14" A polyethylene
#3@9" # 3 @ 24" sheet
1'-6"
1'-6"
10"
#3 @ 12"
1'-10"
4'-1.5"
3 @ 12"
SE CTION A-A
Figure 3-2: Details of dimensions, reinforcement and foundation for Specimen 1-Control Model.
22
23
The sequence of construction of the control model was made following the
current construction practice as it is observed in the photos of Figure 3-3. The soil was
simulated by placing a fill material below the floor slab. The fill material consisted of
reinforced concrete with No. 3 bars spaced each 10 inches in both ways. After the fill
material was poured, the steel reinforcement of the floor slab and the concrete wall
dowels were placed. The floor slab and the fill material were isolated by a polyethylene
sheet placed between them. The roof slab was constructed after the reinforced concrete
walls were finished. Due to construction process of each structural element, construction
joints were generated at the top and the bottom of the reinforced concrete walls.
The second specimen was constructed in the same fashion as the control model,
with the difference that the specimen strip was 6 feet -1.50 inches wide and one bay was
reinforcement, and foundation for the specimen are shown in Figure 3-4. The purpose of
this specimen was to obtain the capacity of the system with an unreinforced infill panel
The construction details of the unreinforced masonry wall are shown in Figure
3-5. The concrete block wall was built by a professional mason after the frame elements
were completed. The construction details between the concrete walls and the infill panels
allow that the only the load transfer mechanism between these wall elements and the
masonry wall are the mortar adhesion and the strength of the masonry unit. The masonry
units, which consisted of a concrete block with dimensions of 6.0 in x 8.0 in x 16.0 in,
24
a) Fill material before concrete pouring. b) Fill material after concrete pouring.
1'-6"
#3@9" #3@12"
#3@9" #3@9"
5"
Unreinforced C oncrete
Bl ock Wall
Wall 1 Wall 2 Wall 3
5'-0" Is olat ed by
A B polyethylene
#3@9" # 3 @ 24" sheet
1'-6"
1'-6"
10"
2"
1'-10" 4@14" 3" #3 @ 12"
#4 @ 14"
3#4 1'-6"
4" 6" 1'-6"
Figure 3-4: Details of dimensions, reinforcement and foundation for Specimen 2-Unreinforced Masonry Wall.
25
26
of the hollow blocks, mortar was applied into the face shells only (Figure 3-5b), while the
grout was placed in each three cells of the masonry unit (Figure 3-5c). The joints were
filled partially with mortar of 3/8 inches of thick, as it is done in practice. After the block
wall was built, it was plastered with a mortar layer thickness of 1/4 inches approximately.
The plastered surface helps to detect cracks pattern in the masonry walls.
Details of the dimensions, reinforcement, and foundation for this specimen are
shown in Figure 3-6, while the construction process is presented Figure 3-7. In general,
the Specimen 3 was built following a construction process similar to that used in the
second specimen, with the difference that the masonry wall was connected to the
In the vertical direction, the masonry wall was reinforced with No. 3 bars spaced
every 16.0 inches. Vertical dowels (No. 3 @ 16.0 inches), were hooked into the floor and
roof slab. The steel reinforcement bars were extended along the entire masonry wall;
accomplishing with the minimum splice length of 16.0 inches or two rows of blocks, as
indicate in the structural notes of the construction drawings. Because the dowels placed
at the roof slab complicated the block wall construction, it was necessary to change the
running bond pattern to stack pattern in the last two rows of blocks (Figure 3-7f).
12'-0" 12'-0"
Load Direction
3'-6" 7'-0" (Hydraul ic
C Act uator)
#3@9" 1'-6" #3@9"
5"
1'-6"
#3@9" #3@12" C
#3@9" #3@9"
5"
Reinforced C oncret e
Bl ock Wall
5'-0" Is olat ed by
#3@9" # 3 @ 24" B polyethylene
1'-6"
1'-6"
sheet
10"
3"
2"
1'-10" #3 @ 12"
4@14"
Figure 3-6: Details of dimensions, reinforcement and foundation for Specimen 3-Reinforced Masonry Wall.
28
29
Vertical
Reinforcement
Masonry/Wall
Connector
(dove tail)
Horizontal
Reinforcement
(duro-wall)
Hooked
bars
Stack
Pattern
truss No. 9 placed at every two rows of blocks (Figure 3-7b). The infill panel was
connected to the reinforced concrete wall using metal ties, such as dove tails (Figure 3-
7b). The dove tails were placed every two rows, at same level of the duro-wall truss. In
this specimen, the construction details allow that the load transfer mechanism is
distributed between mortar adhesion, the strength of masonry unit, the duro-wall and the
opening, is shown in Figure 3-8. This specimen was constructed to address the behavior
of the reinforced concrete wall with a non centered opening. The reinforced panel was
built following the same construction process used in Specimen 3. Figure 3-9 shows the
construction details of the panel confined in one side by the reinforced concrete wall and
in the other side by a reinforced concrete column. The concrete column was reinforced
longitudinally with two No. 4 bars, which were anchored to floor slab dowels, as it can be
observed in Figure 3-9d. During the construction process of the column, ties No. 3 @
12.0 inches were used to provide stability to the longitudinal bars. Following the current
construction practice, the concrete mixture used in this column was that same used for the
grout mixture.
For the masonry wall, the vertical reinforcement was extended from the floor slab
dowels to the roof slab; this means that the vertical reinforcement was not anchored to the
roof slab. In the horizontal direction, the block wall was connected to the reinforced
12'-0" 12'-0"
Load Direction
3'-6" 7'-0" (Hydraul ic
C Act uator)
#3@9" 1'-6" #3@9"
5"
D D
1'-6"
#3@9" 2'-8" C
#3@12"
6"
#3@9" #3@9"
5"
#3@12"
3'-0"
#3@9" # 3 @ 24" B polyethylene
1'-6"
1'-6"
sheet
2"
10"
1'-10" 4@14" 3" #3 @ 12"
1'-6"
#3@12" door opening or bond
#3@9" Bl ock Wall beam for wi ndow opening
6"
#4 @ 14"
3#4 1'-6" 6'-1 1/2"
Figure 3-8: Details of dimensions, reinforcement and foundation for Specimen 4-Reinforced Masonry Wall with
Door Opening.
32
Stacked
Pattern
The behavior of the unreinforced masonry wall with a centered opening, such as
window opening, was investigated with Specimen 5. The window’s opening dimensions
and construction process of this specimen are shown in Figure 3-10 and Figure 3-11,
respectively. The unreinforced masonry wall was built following the same process used
the concrete block cells around the window opening were bonded by a concrete beam as
shown in Figure 3-11c. The bond beam was longitudinally reinforced with two No. 4
bars and transversally reinforced with ties of No. 3 bars spaced each 12.0 inches. Similar
to the column in the Specimen 4, this beam was constructed using similar grout mix.
Figure 3-12 and Figure 3-13 show the typical details of Specimen 6, which was
constructed to investigate the response of a reinforced infill panel with centered opening
subjected to in-plane loadings. The horizontal and vertical steel reinforcement ratios and
construction sequence of the reinforced masonry wall used in this specimen was similar
to the reinforced infill panel of Specimen 3. Also, the blocks around the opening were
bonded by a concrete beam. The bond beam was reinforced and constructed in the
similar way as that presented in Specimen 5. Unlike the reinforcement steel beam of the
Specimen 5, the vertical steel of the bond beam was fully anchored to the roof slab
5'-0"
1'-6"
#3@9" #3@12"
6"
#3@9" #3@9"
2#4 5"
5'-0"
C C
Window Opening
#3@12"
Wall 1 Wall 2 Wall 3
5'-0"
Is olat ed by
# 3 @ 24" A B polyethylene
#3@9"
sheet
1'-6"
1'-6"
10"
2"
3"
#3 @ 12"
1'-10" 4@14"
#4 @ 14"
3#4 1'-6"
Figure 3-10: Details of dimensions, reinforcement and foundation for Specimen 5-Unreinforced Masonry Wall with Window
34
Opening.
35
5'-0"
1'-6"
#3@9" A
#3@12"
6"
#3@9" #3@9"
5"
5'-0"
2#4
Window Opening C C
#3@12"
Wall 1 Wall 2 Wall 3
5'-0"
Is olat ed by
# 3 @ 24" polyethylene
#3@9"
B
sheet
1'-6"
1'-6"
10"
2"
3"
#3 @ 12"
1'-10" 4@14"
#3@16"
1'-6"
6"
#4 @ 14"
3#4 1'-6"
4" 6" 1'-6"
6'-1 1/2" As Shown Ties #3@12" 2#4
SE CTION A-A SE CTION B-B SECTION C-C
Figure 3-12: Details of dimensions, reinforcement and foundation for Specimen 6-Reinforced Masonry Wall with Window
36
Opening.
37
Roof Dowels
(Hooked bars)
Floor Dowels
(Hooked bars)
a) Floor and roof slab dowels details. b) Details of horizontal and vertical steel.
The set-up used to perform the cyclic load test to full scale specimens is shown in
Figure 3-14. The lateral load (V) was applied to the roof slab by a hydraulic actuator
with capacity of 110 kips and a stroke of + 6.0 inches. Two channels C9.5 x 11 were
used to transmit the lateral load applied by the hydraulic actuator to the roof slab. In
order to avoid any force concentration at the frame joints, the channels were connected to
the roof slab by sixteen A 325 bolts (with a diameter of 1.0 inches) spaced at 3.0 feet
39
each. The connection between the channel system and the actuator was designed to allow
for specimen uplift. In the push direction, the lateral load was transmitted to the channels
system via two steel rollers, as can be observed in Figure 3-14c. These rollers permit
vertical movement of the specimens. In the other direction, the specimen was pulled by
four threaded bars (with a diameter of 1.0 inch) anchored to actuator head and connected
to the channels system. Eight threaded bars with a diameter of 1.25 inch were used to
anchor the floor slab specimen to the strong floor. The bars were used without nuts in
order to restraint the horizontal movement and allow for the uplift of the base.
3.3.2 INSTRUMENTATION
The typical instrumentation used during the experimental tests conducted to full
The strain gages were installed in each test to measure the strain in the reinforcing
bars. These strain gages were placed in zones of maximum stress, such as the top and
bottom of the reinforced concrete walls, and at the roof slab ends, as shown in Figure
3-15 and Figure 3-16. For the control model, the strain gages were placed to identify the
plastic hinge at the member ends. For Specimens 2 to 6 additional strain gages were
installed in the reinforced concrete walls to investigate the force distribution produced by
1) and a LVDT 1 shown in Figure 3-17 were used to measure the lateral displacement of
40
St eel Plate
Hydraulic Actuator
(110 Ki ps)
St eel Rol lers
(φ = 2.25 " )
C 9 X 11.5
ST 4.0" x 4.0"x 1/4"
WF 14 X 90
Bol ts (φ = 1" ) @ 3.0'
Connect ing
Fixture Supporti ng Bars
ST 3.0" x 3.0"x 3/8"
St eel Plate
12"x 12"x 1/2"
b) Actual view of the channel c) View of roller fixture that connects the channel
connection system. system and the hydraulic actuator.
Figure 3-14: Set-up used to carry out the cyclic load tests.
SG18 (Top)
SG16 (Top) SG20 (Bott om) SG22 (Top)
C.L
SG17 (Top)
1'-6"
SG15 (Top) SG19 (Bott om) SG21 (Top)
PLAN VI EW
2 1/4"
C.L
4'-1 1/2"
SG17, SG18
SG15, SG16 SG19, SG20 SG21, SG22
8'-0"
2'-3" 1'-6"
Fi ll Material
2'-3"
SG17 (Top) SG21 (Top)
SG19 (Bott om)
C.L PLAN VI EW
9 3/4"
6'-1 1/2"
9 3/4" 2'-3" 2'-3" 9 3/4" SG17, SG18
SG19, SG20 SG21, SG22
SG16 SG15 SG14 Wall 3 SG9, SG10 SG11, SG12, SG13 SG14, SG15, SG16
SG13 SG12 SG11 Wall 2
SG10 SG9 Wall 1
8'-0"
2'-3" 2'-3" 9 3/4"
Fi ll Material
frames without infill panels, while extensometer’s 4, 5, 6, and 7, showed in Figure 3-17
and Figure 3-18, were used to measure the distortion of the infill panels with or without
opening. The potentiometers P1, P2, P3, P4, P5 and P6 were used to identify the concrete
wall base uplift. The lateral load applied to the specimens was measured directly from
The typical load pattern applied to the first specimen under load control is shown
in Figure 3-19. The cyclic lateral load was applied in increments of 1.0 kip; for each
increment of load three cycles were made. After the maximum load was reached, the
For the Specimens 2, 3, 4, 5 and 6, the cyclic lateral load was applied under a dual
control technique, monitoring the applied load and displacement and making the
necessary adjustments during the tests. The load increment was established at 2.5 kip,
while the displacement increment was 1/8 inch for a drift range of + 1.0 inch, and 1/4
inches for a drift range greater than 1.0 inch. Three fully reversed displacement/load
cycles were applied in order to consider the effect of the repeated load on the stiffness
Hydraulic Actuator
LVDT
EXT. 1
EXT. 3 EXT. 5
EXT. 2 EXT. 4
P1 P2 P5 P6 P3 P4
LVDT 1
8.0" 15.5"
LE GEND
POTE NTIOMETE R
EXTE NSOMET ER
LVDT
Figure 3-17: Typical extensometers and potentiometers location for all specimens.
EXT. 7
EXT. 7
EXT. 6
EXT. 6
15.5" 15.5"
3
Lateral load (V), kips
2
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13
-1
-2
-3
-4
Cycles
Figure 3-20 shows the data acquisition system used during the experimental tests.
To measure the data of the instruments, a personal computer (Pentium III 600 MHz)
input channels was used. The analog signal was conditioned using a DEWE-RACK with
were used to read the strain data, the other 14 modules, DQP-VOLTAGES, were used to
read the data from the potentiometers, LVDT’S, extensometers, and hydraulic actuator
load cell.
46
During the experimental tests, the program used to acquire and process the data
program that works through modules where the user programs one or several tasks using
various modules, as shown in Figure 3-21. The DASYLab worksheet reads, processes,
The experimental data was processed as follows: (1) During the experimental
tests, the values were collected at a constant rate of 5 samples/seconds. The function of
The DQPV module converts the data in volts or milivolts to displacement (inches),
strains (µε) or load (kip) using the calibration constant of each instrument. To display the
values of load and displacement during the test, the Dig.-DQPV module was used.
EXPERIMENTAL RESULTS
4.1 INTRODUCTION
This chapter presents the experimental results of the tests conducted to the six full
scale specimens, as well as a discussion and interpretation of the results from each tested
specimen. In the first part, the behavior and the different failure modes of each specimen
observed during the test are presented in details. In the last part, interpretations of the
As were presented in Chapter 3, the six full scale specimens of the slice of a
typical house were aligned in the East-West direction. This means, when the specimen
goes toward East (pulled), the sign convention adopted for the lateral displacement and
lateral load are positives. On the other hand, when the specimen was loaded in the other
direction, toward the West (pushed), the lateral displacement and the lateral load are
negatives.
For each specimen, three curves are presented: 1) The hysteresis curve, 2) The
lateral capacity of the specimen in term of its effective self weight (denoted Weff which is
the weight excited in the dynamic analysis), and 3) Strain distribution of the longitudinal
bars placed in the reinforced concrete walls. The hysteresis curve shows information
49
about maximum resistances in both directions, degradation of the stiffness and strength,
and the residual strength after the maximum lateral load was reached. The lateral load-
lateral displacement relationship of each specimen normalized by its effective weight can
give an idea about the capacity of each specimen in terms of its inertial loads, which are
generated seismic events. The third curve illustrates the bars yielding sequence during
the tests.
The control model was a typical reinforced concrete wall slab frame with one
single-story and two bays. The experimental results of this specimen revealed
information of the overall lateral capacity of the slab-wall interaction. The symmetrical
lateral load-lateral displacement curve of the control model with a severe pinching is
shown in Figure 4-1. The specimen started to exhibit strength degradation at lateral load
of 2.0 kips due to flexural cracks between the interior concrete wall and floor slab. The
maximum lateral load reached by the control model was 4.0 kips in both directions (push
and pull). The pinching of hysteretic loops was due to opening and closing of concrete
wall cracks. Figure 4-2 shows the ratio between lateral load applied to the specimen and
its effective weight (Weff). The specimen effective weight was calculated taking into
account the roof slab weight plus the tributary weight of the upper part of three reinforced
concrete walls. For lateral load greater than the 20.0 percent of Weff, the specimen started
to shown strength and stiffness degradation. The specimen had the capacity of resisting a
walls of the control model when loaded in the push direction are shown in Figure 4-3.
For a lateral load of 2.0 kips, the strains at yielding were not detected. When the lateral
load was incremented to 4.0 kips or 43.0 percent of Weff and a lateral displacement was
2.17 inches or 2.20 percent of drift, the dowel bars placed at the top of wall 3 remained
unyielding. However, all other bars placed in the R/C walls yielded.
Figure 4-4 illustrates steel reinforcement discontinuity between the slab bottom
bars and the R/C wall bars, which did not permit transfer of the positive moment capacity
of the slab to the wall. When the specimen was pushed, positive moment (the section top
fibers are in compression) was developed at the joint between the roof slab ends and wall
The specimen exhibits a ductile behavior, where five plastic hinges developed in
the three reinforced concrete walls, as shown in Figure 4-5. The deformed shape and the
failure pattern of the control model, such as flexural cracks between the floor-wall or roof
slab joints, and the tensile crack of the concrete walls are shown in Figure 4-6.
Specimen 2 was a typical reinforced concrete wall slab frame with one bay
infilled with an unreinforced concrete block wall. Figure 4-7 shows the unsymmetrical
mainly due to the effect of the specimen self weight. When the vertical component of the
diagonal strut developed in the infill panel exceeds the tributary specimen self weight
acting on the interior and exterior concrete walls, the specimen started to uplift. The
51
Push-Pull Push-Pull
5
2
Lateral load, kips
0
-8 -6 -4 -2 0 2 4 6 8
-1
-2
-3
-4
-5
Late ral displace me nt, in
0.5
W eff
h
h/2 0.4
0.3
0.2
Lateral load/Weff
0.1
0
-8 -6 -4 -2 0 2 4 6 8
-0.1
-0.2
-0.3
-0.4
-0.5
Lateral displacement, in
Figure 4-2: Lateral load normalized with respect to the Weff of Specimen 1.
52
V=%Weff 16000 16000
14000 14000
Wall 1 Wall 2 Wall 3
12000
Micro strains, µ ε
12000
12000 12000
10000 10000
S.G6 8000 S.G7 S.G8 8000
6000 6000
S.G13 S.G14
4000 4000
2000 2000
0 0
-25 -20 -15 -10 -5 0 5 10 15 20 25 -25 -20 -15 -10 -5 0 5 10 15 20 25
Figure 4-3: Strain distributions for Specimen 1 loaded in the push direction.
53
Wall 3
Figure 4-4: Details of joint between the roof slab and the exterior wall 3.
Mp Mp
Mp Mp Mp
Mp Mp
Mp Mp Mp
in the push direction, while the uplift of the foundation of wall 2 (interior wall) was
observed at 20.0 kip in the pull direction. Concrete that surround the vertical bars at the
top of the wall 3 was lost at lateral load of -20.0 kips and a lateral displacement of -0.18
in (or 0.18/98.5 =0.18 percent of drift) when the specimen was pushed, showing a first
lost in stiffness and strength. In the opposite direction, the stiffness and strength
degradation were observed for a lateral load of 22.0 kips and 0.10 inch displacement
The overall capacity of the specimen was governed by the punching shear failure
of the interior and exterior concrete walls, and the corner crushing of the infill panel. The
punching shear failure of the interior wall was detected at 32.50 kips and at a
displacement of 0.62 inches (0.62 percent of drift) in the pull direction. The corner
crushing of the masonry panel and punching shear failure of the wall 3 occurred at -28.52
kips and a -1.64 inches of displacement (1.66 percent of drift) in the other direction.
After the maximum load capacity was reached in both directions, the specimen showed a
residual strength of at least 56 percent of the ultimate load (Vu=28.52 kips) obtained in
Figure 4-8 shows the lateral load-lateral displacement curve normalized by Weff of
Specimen 2. For this specimen, its effective weight included the weight of the roof slab
plus the tributary weight of the upper part of three reinforced concrete walls and masonry
wall. For lateral load lesser than 0.66Weff, the specimen showed a linear behavior. The
maximum lateral resistances were 1.73Weff in the negative direction and 1.93Weff in the
positive direction.
56
Figure 4-9 shows the strains profile of the No. 3 bars of the reinforced concrete
walls of the Specimen 2. The low levels of strains at the R/C walls were an indicative
that the maximum lateral load was supported by the infill panel. The yielding was
observed in the longitudinal reinforcement at the top of the interior wall due to punching
shear failure.
The failure patterns of second specimen; such as; the lost of concrete that
surround the vertical bars at the top of the wall 3, corner crushing of the masonry wall
and punching shear failure of the interior and exterior reinforced concrete walls are
shown in Figure 4-10. At the end of the test, a contact length of 36 inches between the
interior wall and masonry wall was observed, whereas for the exterior wall, the contact
The Specimen 3 was built following a similar construction process used in the
Specimen 2, with the difference that the masonry wall was reinforced in the vertical and
horizontal directions with No. 3 bars spaced each 16 inches and duro-wall type truss
placed each two rows of blocks, respectively. The infill panel was connected to the floor
and roof slab using hooked bars and to the R/C walls via metal ties, such as dove tail.
displacement curve was due to the specimen self weight. Because the self weight of the
Specimen 2 and the Specimen 3 are the same, the foundation uplift occurred at the same
Push-Pull Push-Pull
50
40
Punching
Wall 1 Wall 2 Wall 3 30 shear failure
20
Lateral load, kips
Residual
10 Strength
0
-8 -6 Residual-4 -2 0 2 4 6 8
Strength -10
Concrete cover loss of
-20
Punching shear dowels at top of wall 3
failure and -30
masonry crushing
-40
-50
Lateral displacement, in
3
W eff
h h/2
2
1
Lateral load/Weff
0
-8 -6 -4 -2 0 2 4 6 8
-1
-2
-3
Lateral displacement, in
Figure 4-8: Lateral load normalized with respect to the Weff of Specimen 2.
58
V=%Weff 16000 16000
14000 14000
Micro strains, µ ε
12000 12000
Micro strains, µ ε
10000 10000
8000 8000
6000 6000 S.G9 S.G10
S.G1 S.G2
4000 4000
2000 2000
0
0
-40 -30 -20 -10 0 10 20 30 40
-40 -30 -20 -10 0 10 20 30 40
Wall width, in Wall width, in
1.78Weff 1.93Weff 1.78Weff 1.93Weff
12000
Micro strains, µ ε
12000
10000 10000
8000 8000
S.G14 6000
6000 S.G15 S.G16
S.G6 S.G7 S.G8 4000
4000
2000 2000
0
0
-40 -30 -20 -10 0 10 20 30 40
-40 -30 -20 -10 0 10 20 30 40
Wall width, in Wall width, in
Figure 4-9: Strain distributions for Specimen 2 loaded in the pull direction.
59
28 inches
36 inches
16 inches
b) Corner crushing of masonry wall and punching shear failure at the exterior wall.
(at every 9 inches, the same spaces of the vertical steel) at the top of wall 3 began to
appear. This occurred at a load of -30.0 kips and 1.74 inches displacement in the push
direction. For the same loading direction, the maximum lateral capacity was given by
few radial cracks due to punching load at top of wall 3 and the corner crushing of the
masonry wall, at lateral load of -33.57 kips and -3.47 inches of displacement. In the other
load direction, the strength and stiffness degradation began at 30.0 kips and 0.35 inches
displacement, due to growth of radial cracks at the top of the interior wall followed by
sign of corner crushing and propagation of heavy horizontal crack between the roof slab
and the last row of blocks. The maximum resistance was determined by the punching
shear failure of the interior wall accompanied the failure of the last row of concrete
blocks at 41.40 kips and 1.20 inches of displacement. Once the punching shear failure of
central wall happened, the roof slab started to move relative to masonry wall, failing
concrete blocks. The concrete block near to the reinforced concrete wall 3 failed by
crushing mode, while other block failed due to connection between the roof slab and the
masonry wall, which was made using No. 3 hooked bars. After the maximum load
capacity was reached, a severe strength and stiffness degradation were observed in the
For third specimen, the overall lateral capacity in terms of Weff is shown Figure
4-12. For lateral load lesser than Weff, the system showed stable hysteretic loops without
strength and stiffness degradation. The specimen resisted lateral loads in the order of
concrete walls when the specimen was loaded in the pull direction is shown in Figure
4-13. Strains in the bars showed that great part of the lateral load was resisted by the
masonry wall. In Specimen 2 the yield strains were observed at the top of center bars
placed in walls 3, at lateral load level of 1.93Weff or 32.5 kips, however for the Specimen
3 yield strains were not detected for these bars at the same lateral load. At ultimate load,
strains measured in the center bars of the interior wall exceeded the yield strains.
Figure 4-14 shows the crack pattern of the specimen observed during the test. At
the end of test a contact zone of 27 inches between the interior wall and masonry wall
was observed, while for the exterior wall the contact zone was 14 inches.
Specimen 4 was a reinforced concrete block wall with a non centered door
opening. The reinforced infill panel was built following the same construction process
used in the Specimen 3, with the difference that the vertical bars were extended to entire
Figure 4-15 shows the lateral load-lateral displacement relationship for the
Specimen 4. Although the foundation uplift occurred only in the negative direction, at a
load of -12.0 kips and -0.06 inches of displacement, the non symmetrical behavior shown
in the hysteresis curve occurred mainly because the specimen behaved like a reinforced
concrete wall-slab frame in the positive direction and as a masonry-infilled frame in the
other direction. At the beginning of test, in the pull direction, the specimen exhibited
very stiff behavior, until reaching its maximum lateral load capacity at
62
Push-Pull Push-Pull
50
Punching
40 shear failure
Wall 1 Wall 2 Wall 3
30
Severe strength
20
Degradation
Lateral load, kips
10
0
-8 Residual -6 -4 -2 0 2 4 6 8
-10
Strength
-20
3
W eff
h h/2
2
1
Lateral load/Weff
0
-8 -6 -4 -2 0 2 4 6 8
-1
-2
-3
Lateral displacement, in
Figure 4-12: Lateral load normalized with respect to the Weff of Specimen 3.
63
V=%Weff 16000 16000
14000 14000
Micro strains, µ ε
12000 12000
12000
Micro strains, µ ε
12000
10000 10000
8000 8000
2000 2000
0
0
-40 -30 -20 -10 0 10 20 30 40
-40 -30 -20 -10 0 10 20 30 40
Wall width, in Wall width, in
1.93Weff 2.45Weff 1.93Weff 2.45Weff
Figure 4-13: Strain distributions for Specimen 3 loaded in the pull direction.
64
27 inches
18 inches
14 inches
b) Corner crushing and radial cracks due to punching load at the exterior wall.
roof slab and masonry wall of 62.50 psi (V/Mortared Area=13.0 kips/(208.0 in2)=62.50
psi). The time of maximum lateral resistance coincided with the formation of a
horizontal crack between the roof slab and masonry wall interface, and the vertical failure
of the masonry units placed at the last row, adjacent to wall 3. At the end of the test, the
specimen showed residual strength of 6.55 kips at a 3.0 inches of lateral displacement.
The premature failure of the last row (at the top in the right corner) of masonry
concrete blocks in the positive direction produced the crushing of the loaded corner when
the load was applied in the negative direction. Because the compression zone at the
corner was lost due to the previous failure of the masonry unit, the compression diagonal
strut with a contact zone that varies from 8.0 inches to 12 inches, moved from the wall-
slab joint to two feet below the joint, producing flexural and radial cracks at the exterior
wall and the masonry crushing. The flexural and radial crack in the exterior wall
continued growing until the maximum lateral load was reached at -25.0 kips and -1.30
inches of lateral displacement, in the push direction. The vertical cracks spaced at each 9
inches, at top of wall 3, were detected at a lateral load of -20.0 kips and a lateral
The lateral capacity of the specimen in terms of its Weff is shown in Figure 4-16.
When the specimen was loaded in the positive direction, the maximum lateral capacity
was 0.80Weff, after that, a residual strength of 0.35Weff was observed at the end of test.
In the opposite direction, the lateral load capacity was 1.56Weff and the residual strength
was 1.0Weff.
66
Figure 4-17 shows the strain distributions of the longitudinal bars placed in the
R/C walls when the specimen was loaded in the negative direction. For a lateral load of -
1.25Weff or -20.0 kips (80 percent of ultimate load) and -1.0 inches of lateral
displacement, the maximum strains observed were approximately a half of the yielding
strains. The yielding strains were observed at the base of wall 1 and at the ends of the
wall 2, at lateral load of -1.62Weff or -25.0 kips and a lateral displacement of -1.30 inches.
For this level of resistances, strains at yielding were not detected in the bars placed in
wall 3. The yielding did not occur at this level of load due to several reasons: 1) The
premature failure of the masonry wall corner moved the strut action outside slab-wall
joint (two feet below approximately) and 2) The discontinuity of the bars at the roof slab
Figure 4-18 shows failure mode of the Specimen 4. Figure 4-18a shows the
horizontal and vertical cracks detected in the last rows of the masonry units, at lateral
load of 13.0 kips in the pull direction. When the specimen was loaded in the push
direction, the internal compression force of the principal diagonal strut (Strut 1)
generated crushing in the loaded corner at a lateral load of -20.0 kips, developing a
crushing zone between the 6.0 inches and 8 inches, as shown in Figure 4-18b. For a
lateral load of -25.0 kips, a secondary strut (Strut 2) was formed at 24 inches from the
roof slab-wall joint with a contact length of 12.0 inches, inducing some masonry crushing
and horizontal tensile and punching shear failure in the exterior wall, as shown in Figure
4-18c. Figure 4-18d shows the detachment of the dove tail from the concrete wall
Push-Pull Push-Pull
50
40
Wall 1 Wall 2 Wall 3
30
20 Loss of the
Lateral load, kips
mortar adhesion
10
0
-8 Residual -6 -4 -2 0 2 4 6 8
-10
Strength
-20
-30
Masonry -40
corner crushing
-50
Lateral displacement, in
3
W eff
h h/2
2
1
Lateral load/Weff
0
-8 -6 -4 -2 0 2 4 6 8
-1
-2
-3
Lateral displacement, in
Figure 4-16: Lateral load normalized with respect to the Weff of Specimen 4.
68
V=%Weff 16000 16000
14000 14000
Wall 1 Wall 2 Wall 3 12000 12000
Micro strains, µ ε
Micro strains, µ ε 10000 10000
8000 S.G1 S.G2 8000
S.G9 S.G10
6000 6000
4000 4000
2000 2000
0 0
-40 -30 -20 -10 0 10 20 30 40 -40 -30 -20 -10 0 10 20 30 40
Wall width, in Wall width, in
-1.25Weff -1.62Weff -1.25Weff -1.62Weff
Micro strains, µ ε
12000 12000
10000 10000
8000 8000
6000 S.G14 6000 S.G15 S.G16
S.G6 S.G7 S.G8
4000 4000
2000 2000
0 0
-40 -30 -20 -10 0 10 20 30 40 -40 -30 -20 -10 0 10 20 30 40
Wall width, in Wall width, in
-1.25Weff -1.62Weff -1.25Weff -1.62Weff
Figure 4-17: Strain distributions for Specimen 4 loaded in the push direction.
69
6-8 inches
24 inches
c) Corner crushing and crack patterns in the exterior wall at -25.0 kips.
Strut 1
Strut 2
Specimen 5 was built to investigate the behavior of the unreinforced masonry wall
with a centered opening, such as window opening. The unreinforced concrete block wall
was constructed in the same way that concrete block wall of the Specimen 2.
Figure 4-19 shows the hysteresis curve for the Specimen 5. The first substantial
change in stiffness was due to a horizontal crack at the base of the interior pier (pier
adjacent to interior R/C wall) in the horizontal mortar joint, at a lateral load of 22.5 kips
and 0.08 inches of lateral displacement in the pull direction. The interior pier failed by
toe crushing, while the exterior produced a punching shear failure in the roof slab. The
toe crushing of the interior pier occurred at the joint between the column and bond beam
at a lateral load of 30.42 kips and lateral displacement of 0.69 inches when the specimen
was loaded in the pull direction. The punching shear failure and masonry corner crushing
occurred at the loaded corner of the exterior pier at a lateral load of -28.53 kips and -1.48
The hysteresis curve in term of the Weff is shown in Figure 4-20. For lateral load
lower than 1.10Weff, the specimen showed stable hysteresis loops without strength and
stiffness degradation. When the specimen was loaded in the positive direction, the
maximum lateral load was 1.94Weff, whereas in the other direction was -1.82Weff. For a
0.35Weff and 0.90Weff when the specimen was loaded in the positive and negative
directions, respectively.
Figure 4-21 shows the strains in the vertical bars of the reinforced concrete walls
when the specimen was loaded in the positive direction. For a lateral load of 22.5 kips
71
(75 percent of the ultimate load) or 1.43Weff, the measured strains in the bars were lower
than the yielding strains. Due to the strut action developed at the interior pier, the strain
pattern observed at the top of the R/C wall 2 was very similar to the strain patterns
observed in the interior wall of the second and third specimens, like a punching shear
pattern, as shown in Figure 4-21d. These strains exceeded yielding strain at ultimate load
condition.
Figure 4-22 illustrates the typical cracks pattern of Specimen 5 observed during
the experimental test. For the first cycles of load at -12.5 kips and lateral displacement of
0.0635 inches, hair line cracks started to appears at the bottom of the right corner of the
interior pier (bond beam-column joint) in the push direction. When the specimen was
loaded in the other direction, a similar crack pattern appears at the bottom of the exterior
pier in the left corner. These crack patterns demonstrated that when specimen was loaded
in the positive or negative directions, the two piers worked in parallel (one in tension and
the other in compression). Once the mortar adhesion between the roof slab and the last
row of concrete blocks of the pier was lost, the piers started work individually (only in
compression). Horizontal cracks continued growth at the base of interior and exterior
piers at lateral load of ±15.0 kips as it is shown in Figure 4-22a. In general, the
horizontal cracks were indicative of the bond loss between the concrete block and the
mortar at the pier base where maximum bending moment occurs. For a lateral load of
±22.5 kips, continuous horizontal cracks were detected at the base of the interior and
Push-Pull Push-Pull
Int. Ext. 50
Pier Pier
40
Wall 1 Wall 2 T oe cruching
Wall 3 30 of the interior pier
20
Lateral load, kips
10
0
-8 -6 -4 -2 0 2 4 6 8
-10
-20
Punching shear
failure of roof slab -30
and masonry
corner crushing of -40
exterior pier -50
Lateral displacement, in
3
W eff
h h/2
2
1
Lateral load/Weff
Residual
strength
0
-8 -6 -4 -2 0 2 4 6 8
Residual
strength -1
-2
-3
Lateral displacement, in
Figure 4-20: Lateral load normalized with respect to the Weff of Specimen 5.
73
V=%Weff 16000 16000
14000 14000
Wall 1 Wall 2 Wall 3 12000
Micro strains, µ ε
12000
10000 10000
Micro strains, µ ε
10000 10000
8000 8000
S.G6 S.G7 S.G8
6000 6000
4000 S.G14 4000 S.G15 S.G16
2000 2000
0 0
-40 -30 -20 -10 0 10 20 30 40 -40 -30 -20 -10 0 10 20 30 40
Wall width, in Wall width, in
1.43Weff 1.94Weff 1.43Weff 1.94Weff
Figure 4-21: Strain distributions for Specimen 5 loaded in the positive direction.
74
a) Horizontal cracks at the base of the interior and exterior piers at 15.0 kips.
b) Begin of interior pier toe crushing. c) Loaded corner crushing of exterior pier.
Strut
Strut
d) Interior pier stage at the test end. e) Exterior pier stage at the test end.
exterior piers, which were given by the toe crushing and punching shear failure of the
roof slab followed by the corner crushing, respectively. For the interior pier, a heavy
Figure 4-22d and Figure 4-22e shows the crack patterns at the end of the test due
In the last specimen, the effect of the centered opening; such as window opening,
in a reinforced infill panel was investigated. The panel with the opening was constructed
similar to the panel of the previous specimen, with the difference that it was reinforced
Figure 4-23. The specimen shows a little change in the lateral stiffness at lateral load of
22.50 kips and 0.09 inches of lateral displacement, in the pull direction. In the other
direction, the first stiffness change was detected at lateral load of -21.31 kips and a lateral
displacement of -0.30 inches. Similar to the Specimen 5, this stiffness change coincided
with horizontal cracks at the base of the piers, in the horizontal mortar joints. The
reinforced steel patterns used in the infill panel and the connection details between the
wall-slab frame and the perforated masonry panel avoided the rocking mode failure of the
piers that was observed in Specimen 5. The failure of the interior pier was given by a
severe diagonal cracking, typical of the shear failure, at a lateral load of 36.31 kips and a
lateral displacement of 1.32 inches, in the positive direction. Few diagonal cracks were
76
observed at lateral load of 33.0 kips. When the specimen was loaded in the other
direction, the exterior pier exhibited corner crushing followed by a shear failure. The
masonry corner crushing happened at lateral load of -31.62 kips and a lateral
displacement of -1.12 inches. The shear failure occurred at a lateral load of -32.50 kips
and -2.97 inches of lateral displacement. The shear transmitted by the dowels from the
roof slab to the piers generated a few vertical cracks which deteriorated the pier corner.
After the maximum lateral loads were reached in both directions, the specimen showed
The overall lateral capacity in terms of the Weff of Specimen 6 is shown in Figure
4-24. The strength and stiffness degradation began at a lateral load greater than 1.40Weff.
The maximum lateral load capacity was 2.34Weff, when the specimen was loaded in the
positive direction. When the specimen was loaded in the negative direction, the
maximum lateral capacity was -2.10Weff. At the end of test, when the lateral
displacement was ±5.50 inches, the specimen showed a residual strength of the 0.81Weff
Figure 4-25 illustrates the strains measured in the longitudinal bars of the
reinforced concrete walls when the specimen was pulled. Unlike Specimen 5, in which
yield strains were observed at the top of center bars placed in wall 2, at lateral load of
1.94Weff or 30.5 kips (84 percent of the ultimate load), yield strains were not detected for
the these bar, at the same lateral load. Near the ultimate load, for a lateral load of
2.14Weff or 33.5 kips (92 percent of the ultimate load) strains measured in the center bars
5, horizontal cracks started to appear at the base of the interior and exterior piers, at
lateral loads of ±17.50 kips, as it is shown in Figure 4-26a. Figure 4-26b illustrates the
failure mechanism of the interior pier, which was given by the shear failure at 36.31 kips,
when the specimen was pulled. For exterior pier, the mechanism of failure was given by
the masonry corner crushing and the shear failure at lateral loads of -31.62 kips and -
32.50 kips, respectively. The diagonal cracks developed were indicative of the high
tensile stress due to the formation of compression strut (Strut 1) as shown in Figure
4-26c. For the exterior pier, once the crushing of the masonry corner occurred due to
internal force of the diagonal compression strut (Strut 2), a third compression strut (Strut
3) formed outside the compression zone of the second strut, producing the shear failure.
The six specimens were constructed taking into account different parameters of
the construction practice used in Puerto Rico. Among them are: 1) Construction details
of wall-slab frames, 2) The inclusion of the infill panels in these frames, 3) The steel
reinforcement and connection details of the infill panels, and 4) The opening effects in
the infill panel. The experimental results presented in the previous subsections revealed
valuable information about these parameters, which are summarized in the following
sections.
78
Push-Pull Push-Pull
Int. Ext. 50
Pier Pier
40 Shear failure of
Wall 1 Wall 2
interior pier
Wall 3 30
20
Lateral load, kips
10
0
-8 -6 -4 -2 0 2 4 6 8
-10
-20
3
W eff
h h/2
2
1
Residual
Lateral load/Weff
strength
0
Residual
-8 -6 -4 -2 0 2 4 6 8
strength
-1
-2
-3
Lateral displacement, in
Figure 4-24: Lateral load normalized with respect to the Weff of Specimen 6.
79
V=%Weff 16000 16000
14000 14000
Wall 1 Wall 2 Wall 3 12000
Micro strains, µ ε
12000
10000 10000
8000 8000
S.G3 S.G4 S.G5 S.G12 S.G13
6000 S.G11 6000
4000 4000
2000 2000
0 0
-40 -30 -20 -10 0 10 20 30 40 -40 -30 -20 -10 0 10 20 30 40
Wall width, in Wall width, in
1.94Weff 2.14Weff 1.94Weff 2.14Weff
Micro strains, µ ε
10000 10000
8000 8000
S.G6 6000 S.G7 S.G8 S.G14 6000 S.G15 S.G16
4000 4000
2000 2000
0 0
-40 -30 -20 -10 0 10 20 30 40 -40 -30 -20 -10 0 10 20 30 40
Wall width, in Wall width, in
1.94Weff 2.14Weff 1.94Weff 2.14Weff
Figure 4-25: Strain distributions for Specimen 6 loaded in the positive direction.
80
a) Horizontal cracks at the base of the interior and exterior piers at ±17.5 kips.
Strut 1 Strut 2
Strut 3
specimens. Table 4-1 summarized the secant stiffness, maximum lateral load, the lateral
displacement at maximum load, and the failure modes for each tested specimen. The
secant stiffness was calculated as the slope of a line connecting the extreme points of a
small-amplitude displacement cycle in which the load in the cycle is equal to 35 percent
extrapolate the experimental results obtained in the first specimen (which has a width of
49.5 inches), to an equivalent specimen with a width of 73.50 inches (the width of
Specimens 2 to 6). Assuming that these results are proportional to width, the equivalent
lateral stiffness and strength are 5.68 kips/in (73.5/49.5×3.82 kips/in) and 5.94 kips
The construction details used in the Island demonstrated that the reinforcement
steel details used in the joints between the roof slab and the exterior R/C walls did not
allow the development of reversal moment generated when the specimen was subjected
The placement of the longitudinal steel reinforcement at the middle of the R/C
walls permitted the crack growth at the wall ends, producing a hysteresis loops with
severe pinching.
82
4.3.2 INFILL PANEL
The inclusion of the infill panel in the wall-slab frame increased significantly the
stiffness and the maximum lateral resistances. The unreinforced infill panel with a
centered opening (Specimen 5) increased the lateral stiffness and the lateral capacity of
the equivalent frame 106 times and 5.12 times, respectively. A significant increase in
stiffness and strength were achieved when the solid infill panel was reinforced and
connected to the surrounding elements, as case of the Specimen 3, where stiffness and
lateral capacity increased the equivalent frames 177 times and 7 times, respectively.
The unreinforced solid infill panel was investigated in the second specimen,
whereas the reinforced solid infill panel attached to bounding elements was investigated
in the third specimen. The experimental results of the third specimen showed that the
steel reinforcement and the connection details such as, hooked bars and dove tail had
minor effect in the initial stiffness; however these details had considerable effect in the
maximum lateral load capacity and the displacement ductility. The initial stiffness of the
Specimen 3 was 11 percent higher than the initial stiffness of the Specimen 2. The
maximum lateral capacity and the ductility displacement of the third specimen were 25
Similar behavior was observed in the experimental results of the Specimens 5 and
6. The Specimen 5 consisted of an unreinforced infill panel with centered opening, while
in the Specimen 6 the panel was reinforced and tied to the surrounding elements. The
initial stiffness in the last specimen was only 5 percent higher than the initial stiffness of
83
the Specimen 5. The lateral capacity and displacement increased 19 percent and 91
percent, respectively. Although the horizontal and the vertical steel ratio and the
connection details increased the lateral capacity and lateral displacement, these allowed
The connection details of the infill panels with bounding elements used in the
Island showed the following: 1) The hooked bars used to connect the roof slab with the
reinforced masonry wall, as case of the Specimen 3 and Specimen 6, were able to
transmit the load from the roof to masonry wall, reducing the strain (due to punching
load) in the R/C walls. These hooked bars produced the failure of concrete block unit of
last row of reinforced masonry wall of the Specimen 3. 2) When roof slab and masonry
wall were connected with mortar as was in the Specimen 4, the stiff behavior, typical of
the infilled frame, was lost in the first cycles of load, changing the behavior like to the
bared frames, as was Specimen 1. 3) In the Specimens 3, 4 and 6 was observed the
detachment of the dove tail from R/C walls and not from the concrete block units.
The initial stiffness obtained in the Specimen 5, which was constructed with an
unreinforced panel with a centered opening was 34 percent lesser than the calculated for
Specimen 2, which was constructed with solid unreinforced panel. When the infill panels
were reinforced and constructed with and without opening, as were Specimens 3 and 6,
Push-Pull
50
40
30
Lateral Load, kips
20
10
0
-8 -6 -4 -2 -10 0 2 4 6 8
-20
-30
-40
-50
Lateral displacement, in
Specimen 1 Specimen 2 Specimen 3
Specimen 4 Specimen 5 Specimen 6
5.1 INTRODUCTION
The main objectives in performing experimental tests to full scale specimens are
not only to capture the collapse mechanisms of these types of structures subjected to
lateral cyclic loading, but also to use the results to propose and validate simple analytical
models which simulate the experimental behavior observed. These validated simple
analytical models can be used later to predict the behavior of a similar structure subjected
analytical modeling of the masonry infill panel using a macro-model approach. Then, the
inelastic program DRAIN-2DX (Prakash et. al. 1993) in conjunction with the analytical
This section presents detail of the analytical models used to determine the lateral
stiffness and strength of the solid infill panel as well as the perforated panel.
Figure 5-1 illustrates the diagonal compression strut concept proposed by Stafford
Smith (1967) and Mainstone (1971), which is used to determine the initial lateral stiffness
of an infilled frame.
86
Compression Strut
α
rinf
w
Contact hinf hcol
Length
αl
lm
l beam
strut which represents an infill panel depends of the contact length between the frame and
the infill. This contact length depends on the relative stiffness of the infill panel and the
frame elements. Using the approach of a beam on an elastic foundation, Stafford Smith
proposed the following expression between the contact length, α, and the relative
stiffness:
α π
= (5-1)
hcol 2λ hcol
Emtinf sin(2θ )
λ=4 (5-2)
4 E f I col hinf
Where:
Stafford Smith (1967) used a triangular distribution of the contact stress and
calculated the average strain along the diagonal using theory of elasticity and the finite
distributions were in reasonable agreement with the test results. Based on this, the width
of the equivalent diagonal strut to obtain the same diagonal strain has been established.
He also developed a set of empirical curves where the relative stiffness (λhcol) is related
to the effective width of an equivalent diagonal strut. These curves are commonly used
to evaluate the effective width of the strut and, thereby, the lateral stiffness of the infilled
frames.
effective width of an equivalent strut following a very similar approach to the Stafford
Smith. For the reinforced concrete frames infilled with masonry, the relation proposed is:
Where:
307 (1998a) and 308 (1998b) to assess the effective width of a diagonal compression
strut.
Paulay and Priestley (1992) pointed out that a high value of w will result in a
stiffer structure, and therefore potentially higher seismic response. They suggested a
1
w= rinf (5-4)
4
Drysdale et al. (1999) proposed that the vertical (αh) and the horizontal (αl)
contact lengths and the effective strut width (w) are given by:
π 4 E f I col hinf
αh = 4 (5-5)
2 Emtinf sin(2θ )
π 4 E f I b Linf
αl = 4 (5-6)
2 Emtinf sin(2θ )
1
w= α 2 h + α 2l (5-7)
2
Where:
equivalent width:
wtinf Em
Ks = (5-8)
rinf
89
The lateral stiffness of the infill panel based on the compression strut properties is
(1996), Buonopane and White (1999), demonstrated experimentally that the equivalent
strut concept underestimates the initial stiffness of the infilled frames. Fardis (1996), and
Buonopane and White (1999) found that when the infill panel was modeled as a shear
beam, a best prediction for the initial stiffness can be made based in their experimental
results.
According with Tomazevic (1999), and Gulkan and Sozen (1999) the initial in-
plane stiffness of the infill panel should be calculated by taking into account shear and
flexural deformations of the cross-sectional area of the panel. FEMA 356 (2000)
recommends the following expressions to calculate the initial stiffness of the masonry
panel.
1
K inf ( fixed − fixed ) = 3
(5-10)
h heff
eff
+
12 Em I g Av Gm
1
K inf ( fixed − free ) = 3
(5-11)
h heff
eff
+
3 Em I g Av Gm
Where:
The previous expressions are commonly used to model the elastic behavior of an
infill frames. However the experimental observations indicate that the structures with
infill panels exhibit highly inelastic behavior. The most important factors contributing to
the non-linear behavior on infilled frames arise from the material non-linearity. These
factors depend of the inelastic behavior of the infill panel, surrounding frame and panel-
frame interfaces.
Klingner (1977), Reinhorn et al. (1995), and Crisafulli et al. (2000) modeled the
nonlinear behavior of infill frame subjected to lateral cyclic loading by replacing the infill
panel by a non linear compression strut. Tomazevic (1999) indicated that the infill
frames exhibit three different stages when they are loaded to its collapse load, as shown
in Figure 5-2. In the first stage, infill panel remain attached to the surrounding elements,
and the initial lateral stiffness of the infill panel is calculated as shear beam model; taking
into account shear and flexural deformations of the panel. In the second stage, the
separation of the infill panel with the bounding elements occurs, and the stiffness of the
frame, as was shown in Figure 5-1. In this stage, the width (w) of the equivalent strut is
obtained following the Paulay and Priestley (1992) approach. In the last stage, the
masonry infill is modeled by a compression strut, which supports the windward column
at 2/3 of the filler wall’s height. In this particular case, the width (w) of the strut should
rin hcol/3
f
rin
f
f
in
r
5*
ri nf
hcol hcol
0. 2
3*
0 .1
2hcol/3
a) Shear beam model b) Diagonal braced action c) Off diagonal braced action
To assess the secant stiffness of the infill panel, Fardis (1996) concluded that the
Mainstone (1971) equivalent strut width shall be used. Reinhorn et al. (1995) developed
the hysteresis loops can be considered by the selection of the proper values of the nine
parameters included in the model. Some of these parameters are empirical, such as the
For reinforced masonry infill panel, Moghaddam and Dowling (1987) and Dawe
and Seah (1989) reported that horizontal reinforcement does not have a significant
influence in increasing the lateral stiffness of the infilled frames. Due to the fact that
experimental results of the specimens demonstrated that the initial stiffness of the
unreinforced and reinforced masonry panels with or without opening are very similar, the
expression presented here to calculate the stiffness of the unreinforced masonry panel can
strength of infilled frames strongly depends on the type of the failure mechanism
developed, which is dependent of the interaction between the reinforced concrete wall
horizontal load that can sustain the compression strut is as follows: First, the infill panel
is analyzed based on the possible failure modes observed in the experimental tests or in
seismic events and a maximum lateral resistance is established (Vinfill). FEMA 306
(1998) indicates that the strength capacity of an infill panel is a complex phenomenon. It
compression failure, diagonal tension failure of the panel, and general shear failure of the
panel. Second, the strength of the frame elements that surround the infill panels is
established (Vframe) based on different failure scenarios, such as shear or bending failures.
Finally, the compression struts are used as fuse between the frame and the infill panel,
and the maximum lateral capacity is given by the lesser of infill panel capacity (Vinfill) or
The overall lateral capacity of the second specimen which was constructed with
unreinforced solid panel was governed by the masonry corner crushing and the punching
shear failure of the reinforced concrete walls. For the third specimen, built with a
reinforced solid panel, the lateral resistances were limited by masonry corner crushing,
failure of the masonry unit due to the dowel action, and punching shear failure of the
reinforced concrete walls. Based in these failure modes, the infill panel and slab-wall
93
frame were analyzed and the maximum horizontal capacity of the diagonal strut was
established.
The diagonal (Rc) and the horizontal (Vc) compression failure forces can be
2
Rc = α htinf f 'm sec(θ ) (5-12)
3
FEMA 306 (1998) proposed a modified version of the compression failure of the
equivalent diagonal strut of the method suggested by Stafford-Smith and Carter (1969).
The shear force (horizontal component of the diagonal strut capacity) is calculated as:
1
Vc − FEMA = wtinf f 'm cos(θ ) (5-14)
2
Figure 5-3 shows the assumed punching shear mechanism in which the interaction
between the compression strut and the R/C walls was analyzed analogous to a flat plate
loaded by an interior column. The dimensions of the interior column were defined by
masonry wall thickness and the contact length. ACI-318-02 establishes that the nominal
punching shear strength Vp, when no shear reinforcement is used, is the smallest of:
⎛ 4 ⎞
Vp = ⎜ 2 + ⎟ f 'c bo d (5-15a)
⎝ βc ⎠
⎛ α ⎞
Vp = ⎜ s + 2 ⎟ f 'c bo d (5-15b)
⎝ bo d ⎠
94
V p = 4 f 'c bo d (5-15c)
Where:
Roof Slab
B B
Hypothesized compression
Contact Length, α strut mechanism
Steel Reinforcement
Concrete Wall
A
Concrete
Block Wall
Roof Slab
d/2
Concrete
Wall
Steel Reinforcement
Equivalent
Column Area d/2
d/2 t1
Concrete d/2
Block Wall
d Concrete
Block Wall
d/2 t1 d/2
nonlinear behavior of the integral infilled frames, in which the infill and the frame are
bonded or connected together. They found that the collapsed shear load is given by the
expression:
H u = σ c tα c h + Fb (5-16)
Where:
Hu = collapse shear.
σc = crushing stress of infill material.
t = thickness of infilled panel.
α ch = length from loaded corner to plastic hinge on column.
Fb = force induced from the shear stress at the infill/beam
interface.
In this research, the first term of the expression presented by Kwan and Liauw
(1984) is based in the experimental failure observed, such as the punching shear load or
the masonry corner crushing. The second term includes the shear capacity of the hooked
bars anchored in the grouted cell of the concrete blocks. The UBC 1997 Section
2108.1.5.2 establishes that the shear capacity of the bolt embedment in the masonry is
Bs = 0.25 Ab f y (5-17b)
Where:
of load is less than 12 bolt diameters, the value of Bs in Equation (5-17a) shall be reduced
FEMA 302 (1997) proposes a similar equation than the one proposed by the UBC
1997 to determine the shear capacity of the bend-bar anchor bolts. It is given by the
lesser of:
Bv = 0.54 Ab f y (5-18b)
Where:
FEMA 302 (1997) indicates that when the anchor bolt edge distance, lbe, is less
than 12 bolt diameters, the value of Bv in Equation (5-18a) shall be reduced by linear
Tomazevic (1999) indicates that the amount of shear that can be carried by a
dowel action of a single vertical bar Hdrv, embedded in the mortar or grout, can be
calculated by:
Where:
The effect of the centered and non centered openings in the lateral stiffness and
reinforced infill panel with non centered door opening was investigated. In Specimens 5
and 6, the effect of the centered opening was evaluated in the unreinforced and reinforced
Mallick and Garg (1971) investigated the effect of openings on the lateral
stiffness of infilled frames. They found that the composite action between the frame and
the infill is adversely affected as opening position is moved towards the compression
diagonal. The experimental results showed that an infilled frame with an opening on
either of the loaded diagonal reduces the stiffness by 60 to 70% as compared with that of
a similar infilled frame with a solid panel. They suggested that due to the nature of
Liauw (1972) proposed the equivalent frame concept to analyze the infill frame
with opening. The infilled frame was transformed into an equivalent frame whose
members have the properties of the composite section of the actual structure. The
analytical results obtained by the method were compared with results from the elastic
model experiment using infills of different rigidity. The comparison showed a good
agreement between the analytical and experimental results when the opening is more than
50 percent of the full infill area. The method is in the conservative side when the opening
FEMA 356 (2000) indicates that it is possible to evaluate the local effect of the
perforated infill panel using multiple compression struts as shown in Figure 5-4.
98
However, the theoretical work and the experimental data for determining the multiple
strut placement and strut properties are not sufficient to establish reliable guidelines and
the use of this approach requires judgment on a case-by-case basis. FEMA 306 (1998)
recommends that infilled panels with openings should be investigated also as assemblies
of subcomponents of the appropriate materials. FEMA 273 (1997) specifies that if the
beams that surround the piers are sufficiently stiff in bending, the piers can be assumed to
be fully restrained against rotation at their top and bottom. If the opening in a perforated
wall is relatively large, the wall system will deflect as a cantilevered shear element.
Fx2 Level 2
Fx1 Level 1
Tomazevic (1999) indicated that the window pier works as confined masonry in
seismic events due to the restrain imposed by the reinforced concrete elements that
surround it. He proposed that the initial stiffness of the pier should be determined
considering the flexural and shear deformations, similar to infilled frames. For the secant
experimental results or based on a damage index. The damage index can be found by
statistical analysis and correlation between damage to the walls and lateral stiffness.
99
5.2.4 STRENGTH OF THE MASONRY INFILL PANEL WITH OPENING
Generally, rocking, sliding, diagonal tension and toe crushing are the failure
modes that govern the lateral capacity of the unreinforced masonry piers. In the case of
the reinforced masonry pier, the failure modes are defined by diagonal shear, sliding
unreinforced masonry panel with a centered opening, was defined by the toe crushing of
the piers. In Specimen 6, the reinforced masonry built with the same centered opening,
the failure mechanisms observed were the masonry corner crushing and the shear failure
(diagonal cracking). Similar to infilled frames, diagonal strut will be used as fuse
Figure 5-5a illustrates the typical configuration of a wall divided into four piers
by openings, while Figure 5-5b shows the maximum shear force that can be transmitted
V P
P1
P1 P2 P3 P4 V1
V1 V2 V3 V4
h a/2
1 2 3 4
a/2
V1 V2 V3 V4
V1
P1 P2 ` P3 P4
P1
lw
moments about the toe reaction P1 (for Pier 1) and considering the pier self-weight as
P1(lw − a )
V1 = (5-20)
h
P1
a= (5-21)
0.85 f ' mt
Where:
P1 = toe reaction.
h = height of the pier.
lw = length of the pier.
a = compression contact length at ultimate.
Pauley and Priestley (1992) to estimate lateral capacity in confined masonry. Figure 5-6
shows the shape and distribution of the interaction forces between the R/C columns and
masonry panels.
Vi
Xvz
Hi
h
Yh
Hi Xvs
l Vi
the equilibrium of the moments of interaction forces, the following expression between
the lateral load Hi and the resultant of vertical interactions forces Vi was proposed:
Vi l α
Hi = (5-22)
h
Where:
Hi = lateral load.
Vi = the resultant of vertical interaction forces.
l = length of the pier.
h = height of the pier.
α = parameter which depends on the assumed shape and
distribution of the interaction forces, (α = 1.25).
The diagonal tension cracks are considered to occur when the maximum principal
tensile stress in the piers are larger than the diagonal tensile strength of the masonry.
The shear capacity of the reinforced masonry pier can be calculate considering the
contribution of masonry and the horizontal reinforcement. The ultimate shear strength is
given by:
Vu = Vm + Vhs (5-23)
Where:
infilled frames using three types of frames, square and rectangular single panels, three
story single-bay square panels, and single story three bays square panels. The parameters
102
that they investigated analytically were: boundary separation, interface friction, aspect
ratio (h/l), frame stiffness, rigidity of the frame joints and relative beam stiffness. Based
in the diagonal tensile failure, they proposed that maximum shear load can be calculated
Where:
Priestley and Calvi (1991) suggested that the horizontal force to induce diagonal
cracking can be estimated using the relationship for the tensile stress in a disk loaded
along a diameter:
π
Vm = td m ft cos(θ ) (5-25)
2
Where:
Drysdale and Hamid (1984) concluded that the tensile stress of the masonry can
be vary between from 2.0√f’m to 3.8√f’m. Bennett et al. (1996) indicated that it is
possible get values of cracking strength in between 1.3√f’m to 2.9√f’m. However, they
suggested a high value of 5.0√f’m, when the masonry compressive strength is computed
using the net area. The previous works suggest that the tension capacity of the masonry
2.5√f’m is adopted as cracking strength. This value is recommended by the UBC 1997 as
Av f y d
Vhs = (5-26)
s
Where:
Av = steel area.
fy = yielding strength.
d = effective depth.
s = vertical spacing of the horizontal reinforcement.
Figure 5-7 shows a computer model built in the inelastic program DRAIN-2DX
(Prakash et. al. 1993). The computer model takes into account the failure mechanism
observed in the control model test, which was governed by the plastic hinges at R/C wall
ends. In the model, the roof slab and the reinforced concrete walls were modeled using
plastic hinge beam–column elements (Type 2 element in the elastic mode, yielding was
not permitted) connected between them with a bilinear inelastic rotational springs at the
element ends, as shown in Figure 5-8. Figure 5-9 illustrates the backbone curve of the
rotational springs used to simulate the continuous opening and closing of the flexural
cracks generated in the reinforced concrete frame element ends during the experimental
tests, as consequence of cyclic loadings. The rotational springs were used with option of
inelastic unloading with gap. Because severe flexural crack patterns were concentrated
on the R/C wall ends and not on the floor slab ends, the floor slab was modeled using
104
only plastic hinge beam-column element (Type 2 element in the inelastic mode, yielding
was allowed).
The Bilinear Moment-Rotation Curve for the inelastic spring was calculated
with an unconfined ultimate strain of 0.004 was used. For the steel bars, the elasto-
plastic stress-strain curve was utilized, as allowed by the different design codes, such as
ACI-318-02 and UBC 1997. The moment-curvature relationship of the R/C element was
calculated using the commercial software CSI Section Builder Version 8.10. In the
also taken into account. The average bond stress permitted by the ACI-318-02 for steel
The computer model was restrained in the horizontal direction whereas the
specimen base uplift with certain vertical friction was permitted (1.0 percent of the lateral
load). The friction was caused by the horizontal restraint elements, such as vertical
threaded bars acting in shear placed between the specimen foundation and strong floor,
were simulated using three vertical elements acting in parallel. To perform the non linear
cyclic analysis, the specimen self weight was first applied; next, the specimen was
laterally loaded with the same lateral displacement histories obtained from the
experimental test.
general, the maximum lateral capacity was predicted with a reasonable agreement.
105
Because of the constraints of DRAIN-2DX, it was not possible to model exactly the
14 15
16
17 18
Nodes 15 and 17 are slaved to master
Rotational Connection node 16 in X and Y directions. Plastic Hinge Beam-Column
Element (Type 04) Element (Type 02)
3'-6" 5'-0" 7'-0" 5'-0" 3'-6"
11 12 14 15 17 18 19 20 21 23 24 26 27
10 28
13 16 22 25
3 Zero Length 9
6
Element
2 5 8
29 30 31 32 33 34 35 36
1 4 7
37 38 39 40 41 42 43 44 45 46 47
10" 3'-2" 4'-0" 3'-0" 1'-0" 1'-0" 3'-0" 4'-0" 3'-2" 10"
Node i Node j
Elastic beam
Inelastic Rotational Spring
Figure 5-8: Element used for R/C walls and roof slab.
106
Moment
My(+) 1 Ki Ke
1
Ke
1
Rotation
My(-)
Where:
My(+), My(-)=Pos itive or negative yield moment
Ke=elastic stiffness
Ki=inelastic stifness
Push-Pull
5
4
3
2
Lateral load, kips
1
0
-8 -6 -4 -2 -1 0 2 4 6 8
-2
-3
-4
-5
Lateral displacement, in
Experimental DRAIN-2DX
The punching shear failure of R/C walls and corner crushing of the unreinforced
masonry wall were failure mechanisms achieved by the second specimen. Figure 5-11
shows the computer model for the second specimen. The frame elements were modeled
following the same approach that was used in the first specimen, with the differences that
the moment and curvature and moment and rotation relationships for nonlinear springs
were computed for the width specimen of 73.5 inches in place of 49.5 inches, as it was
To simulate the punching shear failure and certain degree of masonry crushing it
was necessary to model the infill panels with four compression struts. Two inelastic
compression struts (Strut 1 and Strut 2) were placed in the diagonal of the frame and the
other two elastic compression struts were placed off diagonal (Strut 3 and Strut 4) at the
end of punching shear failure and masonry crushing zone. In general, lateral capacity and
the nonlinear behavior were governed mainly by the inelastic diagonal struts. The off
diagonal struts were used to simulate behavior of the upper part of the interior and
exterior R/C walls once the maximum lateral load was attained.
Figure 5-12 illustrates the multi-linear backbone curve of the hysteretic model of
the compression strut. The most important parameters to define the multi-linear
backbone curve are: the axial force at cracking (Fcr = γFu) and at ultimate strength (Fu),
the initial stiffness to cracking (K1), and the secant stiffness to ultimate strength (K2 =
αK1). The falling post-ultimate branch does not attain the horizontal axis, but reaches a
plateau corresponding to a residual strength (Fr = rFu) at very large deformations. The
108
parameters γ , α, β, and r where obtained based in the experimental results, specifically in
For the computer model of the Specimens 2 to 6, the initial elastic stiffness of the
the elastic stiffness of the compression strut (K1) was determined as follows:
In general, the compression strut is used as fuse between the infill panel and
surrounding frame elements. The compression strut strength takes into account the
failure mechanisms observed during the tests. The horizontal component (Fux) of the
ultimate compression capacity (Fu) of the strut was established as the lesser of the
resistances given by the punching shear failure (Vpunching shear) or masonry crushing
(Vmasonry crushing).
48 8'-4.5"
Strut 3
St rut 4
Wall 1 Wall 2 Wall 3
Compression Link
Elements (Type 09)
4 7
10" 3'-2" 4'-0" 3'-0" 1'-0" 1'-0" 3'-0" 4'-0" 3'-2" 10"
i = node number
Fu
αK1
βK1
Fcr=γFu 1
K1 K1
Fr=rFu
u1 u2 u3
Axial deformation
Int ermediate loading-unloading slope.
Table 5-1 presents the material properties of the R/C walls and the unreinforced
masonry panel used to construct the multi-linear backbone of the compression struts. The
different parameters used to define the multilinear backbone curve, such as initial, secant
and negative stiffness, cracking, ultimate and residual strength of the diagonal struts and
the elastic properties of the off diagonal struts, are summarized in Table 5-2.
and flexural deformations of the infill panel. The secant stiffness for the Strut 1 was
obtained using the Mainstone 1971 approach. For the Strut 2, secant stiffness was
calculated using an equivalent width w=1/8 rinf (Tomazevic 1999-Stage 3). The masonry
corner crushing load predicted by FEMA 306 (1998) is less than the punching shear load
predicted by ACI318-02 for the interior wall. However, during the test crushing was not
detected at the load predicted by FEMA 306 (1998). The diagonal strut capacity was
established by the punching shear mechanisms. For the off diagonal strut, the stiffness
was calculated using an equivalent width w =1/8 rinf (Tomazevic 1999-Stage 3).
Figure 5-13 shows a comparison between the analytical and experimental results
for the second specimen. The computer model predicted the maximum lateral load in the
positive and negative directions with a difference of 0.01 percent and 10 percent with
Push-Pull
50
40
30
20
Lateral load, kips
10
0
-8 -6 -4 -2 -10 0 2 4 6 8
-20
-30
-40
-50
Lateral displacement, in
Experimental DRAIN-2DX
Figure 5-14 shows the computer model of the Specimen 3. In the model, the
reinforced infill panel attached to the surrounding elements was modeled using four
compression struts. The inelastic compression struts (Strut 1 and Strut 2) were placed in
the diagonal of frame to simulate the principal mechanism of failure, which were the
punching shear of R/C walls, failure of the last row of concrete block due to dowel action
and some masonry corner crushing. In order to simulate the post ultimate behavior of the
specimen, two additional compression struts were placed off diagonal (Strut 3 and Strut
The material properties used to construct the third analytical model are presented
in Table 5-3 while the most important parameters used to define the compression struts
The initial stiffness of the diagonal struts was calculated taking into account the
shear and flexural deformations of the infill panel, considering fixed-free end condition.
For the diagonal compression Strut 1, secant stiffness was calculated based in the best
fitting of the experimental results. The secant stiffness of the inelastic compression Strut
2 was computed considering a strut with an equivalent width, w of 3/16 of the diagonal
length, an average value between w=1/4 rinf (Paulay and Priestley 1992) and w=1/8 rinf
The maximum shear load of Strut 1 was established considering only the
punching shear load. The experimental results showed that it is possible to get failure of
punching shear simultaneously with masonry corner crushing when both capacities are
very close. However, due to loss in the contact zone between the masonry wall and R/C
113
walls it is not possible to get the punching shear failure in R/C walls after masonry
crushing occurs. The shear capacity of the dowel action was include only in the
considered that shear capacity of the hooked bars at the interface between the roof slab
and reinforced wall is occurring only in one loading direction because during the test the
last rows of the masonry concrete block wall failed when the specimen was pulled.
For the compression Strut 2, the lateral capacity was established considering the
punching shear load and the shear capacity of the hooked bars according to FEMA 302
(1997). The shear capacity of the hooked bars predicted by FEMA 302 (1997) is close to
the average values of the capacities predicted by UBC 1997 and Tomazevic (1999). Due
to severe strength and stiffness degradation observed at the test end, the residual strength
The computer model simulated the overall lateral behavior of the Specimen 3 with
reasonable agreement, as illustrated in Figure 5-15. The analytical model predicted the
maximum lateral load in the positive and negative directions with a difference of 0.01
48 8'-4.5"
Strut 3
St rut 4
Wall 1 Wall 2 Wall 3
Compression Link
Elements (Type 09)
4
7
10" 3'-2" 4'-0" 3'-0" 1'-0" 1'-0" 3'-0" 4'-0" 3'-2" 10"
i = node number
Push-Pull
50
40
30
20
Lateral load, kips
10
0
-8 -6 -4 -2 -10 0 2 4 6 8
-20
-30
-40
-50
Lateral displacement, in
Experimental DRAIN-2DX
Figure 5-16 shows the computer model constructed to simulate the failure
between a wall-slab frame and reinforced masonry panel with a non-centered opening
was simulated using three struts; two in compression (Strut 1 and Strut 3) and one in
tension (Strut 2). The inelastic compression Strut 1 was used to simulate the masonry
corner crushing. The inelastic tension Strut 2, with a multi-linear backbone shown in
Figure 5-17, simulates the loss of mortar adhesion between roof slab and the masonry
wall. This inelastic tension strut was placed in the model to simulate the experimental
behavior observed. However, the use of the inelastic tension strut it is not recommended
for practical cases because to the mortar adhesion was loss in the first cycles of load as
was observed in the test. The third strut was used to simulate the formation of the
secondary strut observed experimentally, which formed two feet below the wall-slab
4 7
10" 3'-2" 4'-0" 3'-0" 1'-0" 1'-0" 3'-0" 4'-0" 3'-2" 10"
Axial Force
Fu
βK1
K1
K1 Fr=rFu
u1 u2
Axial deformat ion
Intermediate loading-unloading slope.
The material properties used in the construction of the analytical model are
presented in Table 5-5 while the most important parameters used to define the struts
were determined considering shear and flexural deformations of the infill panel,
considering a fixed-free end condition. The secant stiffness of the inelastic compression
Strut 1 was calculated considering a strut with an equivalent width, w of 3/16 of the
diagonal length. The horizontal capacity of the Strut 1 was governed by the masonry
crushing load. For the tension Strut 2, the lateral capacity was established based on the
5-18 demonstrate that the simplified computer model simulated reasonably well the
Push-Pull
50
40
30
20
Lateral load, kips
10
0
-8 -6 -4 -2 -10 0 2 4 6 8
-20
-30
-40
-50
Lateral displacement, in
Experimental DRAIN-2DX
Figure 5-19 shows the computer model constructed to simulate the failure
mechanism observed during the test conducted on Specimen 5. The lateral capacity when
the specimen was loaded in the positive direction was governed by the toe crushing of the
interior pier. In the negative direction, the lateral capacity was governed by punching
shear failure of the roof slab and masonry crushing of the loaded corner in the exterior
pier. These failure mechanisms were simulated using six struts and an elastic panel.
Strut 1 and Strut 6 were used to simulate post ultimate behavior of the specimen once
failure of the piers occurred. These compression struts produce a short column effect on
the R/C walls. The inelastic compression Strut 2 and Strut 4 were placed in parallel with
the inelastic tension Strut 3 and Strut 5 in order to model the non linear behavior of the
exterior and interior piers, respectively. The inelastic tension struts, Strut 3 and Strut 5,
represent the mortar adhesion between the roof slab and masonry piers.
The material properties and the parameters that define the behavior of the struts
are summarized in Table 5-7 and Table 5-8, respectively. The initial stiffness of Strut 2
to Strut 5 was calculated considering shear and flexural deformations of the pier. Due to
the rocking mode of the piers activated during the tests, the secant stiffness of the
inelastic compression struts was calculated based on the best fitting of the experimental
results. The cracked stiffness of the other compression struts, Strut 1 and Strut 6, was
shear that can be transmitted by the pier in the rocking mode. In the pier rocking, the toe
reaction, P, was established as the lesser of the masonry corners crushing at ultimate
122
3'-6" 5'-0" 7'-0" 5'-0" 3'-6"
19 28 Crushing
Zone
=0.58 ft
49 51 8'-4.5"
Strut 6 Strut 1
48 52
Wall 1 Elastic
Panel
4 7
10" 3'-2" 4'-0" 3'-0" 1'-0" 1'-0" 2'-6" 5'-0" 2'-8" 10"
Description Values
Compressive masonry strength f'm,psi=2281.00
Compressive strength of R/C walls f'c, psi=3950.00
Compressive strength of roof slab f'c, psi=3384.00
Compressive strength of the bond column f'c, psi=1000.00
Effective depth of the R/C wall d, in=2.50
Effective depth of the roof slab d, in=4.06
Height of the R/C walls h, in=96.00
Width of the R/C walls w, in=73.50
Height of the masonry pier hm,in=60.00
Length of the masonry pier lm,in=39.50
Modulus of elasticity of concrete Ec, ksi=3582.39
Modulus of elasticity of masonry Em, ksi=1710.75
Moment of inertia of the R/C walls Igross, in4=765.63
Angle of the diagonal struts (radians) θ=0.99
Thickness of block wall t1, in=5.63
Mortared thickness t, infill, in=2.00
123
Table 5-8: Struts properties of Specimen 5.
Description Values
Compression strut stiffness
Diagonal length of Strut 1and Strut 6 rinf, in=53.44
Equivalent width of the Strut 1, w=1/8 rinf w, in=6.68
Elastic stiffness of the Strut 1 and Strut 6 K1, kip/in=427.69
Lateral stiffness of the pier considering shear and flexural deformations K pier, kip/in=240.60
Axial stiffness of the Strut 2 and Strut 4 K1, kip/in=795.74
Secant stiffness factor for Strut 2 α2=0.07
Secant stiffness for the Strut 2 K22, kip/in=55.70
Secant stiffness factor for Strut 4 α4=0.07
Secant stiffness for the Strut 4 K24, kip/in=55.70
Negative stiffness factor for Strut 2 β2=-0.03
Negative stiffness for the Strut 2 K32, kip/in=-25.00
Negative stiffness factor for Strut 4 β4=-0.03
Negative stiffness for the Strut 4 K34, kip/in=-25.00
Tension strut stiffness
Lateral stiffness of the pier considering shear and flexural deformations K pier, kip/in=240.60
Axial stiffness of the Strut 3 and Strut 5 K1, kip/in=795.74
Negative stiffness factor for Strut 3 and Strut 5 β=-0.02
Negative stiffness for the Strut 3 and Strut 5 K2, kip/in=-12.73
Maximum toe pier reaction, P
Compression contact length observed experimentally
Average contact length for Strut 2 (Exterior pier) a, in=9.55
Contact length for Strut 4 (Interior pier) a, in=9.44
Punching shear load at roof slab (Exterior Pier) P1, kips=44.39
Compression capacity of the bond column (Interior Pier-Priestley 1992) P2, kips=45.66
Maximum shear in the rocking mode, V
For the exterior pier (Strut 2) V1, kips=21.15
For the interior pier (Strut 4) V2, kips=22.80
Compression struts strength
Axial capacity of Strut 2 Fu-1, kips=38.46
Reduction factor (linear behavior is assumed) γ1=0.65
Cracking load for Strut 2 Fcr-1, kips=25.00
Percent of Residual Strength (Fr=rFu) for Strut 2 r=0.010
Axial capacity of Strut 4 Fu-2, kips=41.46
Reduction factor (linear behavior is assumed) γ2=0.65
Cracking load for Strut 4 Fcr-2 kips=26.95
Percent of Residual Strength (Fr=rFu) for Strut 4 r=0.010
Tension strut strength
Maximum horizontal load of the Strut 3 and Strut 5 Vu-1, kips=5.000
Axial capacity of the tensions struts Fu-1, kips=8.961
Reduction factor (linear behavior is assumed) γ1=1.000
Cracking load for Strut 1 Fcr-1, kips=8.961
Percent of Residual Strength (Fr=rFu) for Strut 1 r=0.001
124
and the punching shear load of the roof slab. Knowing the maximum toe reaction P, the
maximum shear load, V, was calculated. During the tests, it was observed that the
compression contact length, a, varies from 8.76 inches to 10.0 inches. For the other
inelastic compression Strut 4, the horizontal capacity was established in a similar way,
where the maximum toe reaction P was given by the compression capacity of the bond
column. For the interior pier, the compression contact length observed experimentally
As it can be seen in Table 5-8, the horizontal capacity of the interior pier is very
close to horizontal capacity of the exterior pier. The tension capacity of the Strut 3 and
Strut 5 take into account the mortar adhesion between the pier and roof slab.
Figure 5-20 shows a comparison between the experimental and analytical results.
The computer model underestimated the maximum lateral load by 6.0 percent when the
specimen was loaded in positive direction and by 7.0 percent when the specimen was
40
30
20
Lateral load, kips
10
0
-8 -6 -4 -2 -10 0 2 4 6 8
-20
-30
-40
-50
Lateral displacement, in
Experimental DRAIN-2DX
Figure 5-19 shows the computer model constructed to simulate the experimental
behavior observed in the Specimen 6. In this specimen, the ultimate lateral capacity of
the interior pier and exterior pier was governed by the shear failure (diagonal cracking).
Six struts and an elastic panel were used to simulate the ultimate and post ultimate
behavior observed during the tests. The compression struts, Strut 1 and Strut 6, were
used to simulate post ultimate behavior of the specimen once failure of the piers
occurred. These struts produce a short column effect on wall 2 and wall 3. The inelastic
compression struts, Strut 2 and Strut 4, placed parallel with inelastic tension Strut 3 and
Strut 5 were used to model the non linear behavior of the exterior and interior piers,
respectively. The horizontal components of the inelastic tension Strut 3 and Strut 4
simulate the maximum yielding shear that the pier can transmit due to its flexure
126
capacity. Table 5-9 and 5-10 summarize the material properties used in the model and
Due to the small difference obtained in the initial stiffness between Specimen 5
and Specimen 6 in the experimental tests, which was less than 5.0 percent, the elastic
stiffness of Struts 2 to 5 were calculated in the same way. The secant stiffness of the
inelastic compression struts was calculated based on the best fitting of the experimental
results. The cracked stiffness of the other compression struts, Strut 1 and Strut 6, were
Description Values
Compression strut stiffness
Diagonal length of Strut 1and Strut 6 rinf, in=53.44
Equivalent width of the Strut 1, w=1/8 rinf w, in=6.68
Elastic stiffness of the Strut 1 K1, kip/in=427.69
Lateral stiffness of the pier considering shear and flexural deformations K pier, kip/in=240.60
Axial stiffness of the Strut 2 and Strut 4 K1, kip/in=795.74
Secant stiffness factor for Strut 2 α2=0.04
Secant stiffness for the Strut 2 K22, kip/in=31.83
Secant stiffness factor for Strut 4 α4=0.04
Secant stiffness for the Strut 4 K24, kip/in=31.83
Negative stiffness factor for Strut 2 β2=-0.02
Negative stiffness for the Strut 2 K32, kip/in=-11.94
Negative stiffness factor for Strut 4 β4=-0.02
Negative stiffness for the Strut 4 K34, kip/in=-11.94
Tension strut stiffness
Lateral stiffness of the pier considering shear and flexural deformations K pier, kip/in=240.60
Axial stiffness of the Strut 3 and Strut 5 K1, kip/in=795.74
Negative stiffness factor for Strut 2 β=-0.02
Negative stiffness for the Strut 2 K2, kip/in=-12.73
Shear capacity of masonry pier
Shear capacity of masonry (Diagonal Cracking-Priestley and Calvin 1992) Vm, kips=24.06
Shear capacity of bond column (ACI 2002) Vbc, kips=1.95
Shear capacity of horizontal steel reinforcement; duro-wall type 9 Vs, kips=3.02
Total shear capacity, Vu, kips=29.03
Compression struts strength
Axial capacity of Strut 2 Fu-1, kips=52.79
Reduction factor (linear behavior is assumed) γ1=0.65
Cracking load for Strut 2 Fcr-1, kips=34.32
Percent of Residual Strength (Fr=rFu) for Strut 2 r=0.010
Axial capacity of Strut 4 Fu-2, kips=52.79
Reduction factor (linear behavior is assumed) γ2=0.65
Cracking load for Strut 4 Fcr-2 kips=34.32
Percent of Residual Strength (Fr=rFu) for Strut 4 r=0.010
Tension strut strength
Yielding moment of the pier My, kip-ft=18.00
Yielding shear (Vy=2My/hm) Vy, kips=7.20
Maximum horizontal load of the Strut 3 and Strut 5 Vu-1, kips=7.20
Axial capacity of the tension struts Fu-1, kips=10.362
Reduction factor (linear behavior is assumed) γ1=1.000
Cracking load for Strut 1 Fcr-1, kips=10.362
Percent of Residual Strength (Fr=rFu) for Strut 1 r=0.001
128
For Strut 2 and Strut 4, the maximum horizontal load capacity was calculated
taking into account the maximum shear capacity of the reinforced piers. The shear
capacity of the pier was established considering the contribution of the masonry, bond
Figure 5-21 shows a comparison between the experimental results and analytical
results. The computer model predicted with a reasonable agreement the maximum lateral
load when the specimen was loaded in positive direction. In the other load direction, the
Push-Pull
50
40
30
20
Lateral load, kips
10
0
-8 -6 -4 -2 -10 0 2 4 6 8
-20
-30
-40
-50
Lateral displacement, in
Experimental DRAIN-2DX
6.1 INTRODUCTION
The previous chapters presented the experimental tests conducted to full scale
observed, simple analytical models were proposed and validated. Using these
experimental and analytical results as starting point, this chapter is focused on predicting
the seismic response of the typical house subjected to a wide range of earthquake ground
motions in its weak axis. First, the analytical model of a typical house is presented. The
next section presents the selection of the earthquake records, which take into account the
geological conditions of Puerto Rico as well as the design spectrum established by UBC
1997 for a Seismic Zone 3 and Soil Type Sd. Finally, using the results of the non-linear
time history analysis, the lateral capacity of the house and the expected damage level is
established.
Figure 6-1 shows a plan view of the typical residential houses constructed in the
Island. The dimensions of the structural and non structural elements and the orientation
of the reinforced concrete walls and masonry block walls were obtained from the survey
Figure 6-2 shows the computer model constructed to predict the seismic response
of the typical houses subjected to seismic loading in its weak direction. The computer
130
model was constructed using the inelastic program DRAIN-2DX (Prakash et. al 1993).
In the computer model, the roof slab was considered as a stiff in-plane diaphragm, in
which the inertial loads generated in the center of mass during seismic events (Node 16)
are transmitted from it to the reinforced concrete walls and masonry block walls via rigid
links. The lateral stiffness of the reinforced concrete walls oriented in the strong
direction of the houses was modeled using elastic supports. The masonry concrete block
walls, placed in the weak direction of the house, were modeled using inelastic
translational springs (Type 9). In the analytical model, the solid and the perforated
component simulates the behavior of each specimen tested. For example, Component 1
Figure 6-3 shows the multi-linear backbone curves of each component. These
skeleton curves were defined using the analytical results of the calibrated computer
The computer model of house has the following limitations: 1) Because of the
constraints of DRAIN-2DX, it was not possible to model exactly the stiffness and the
strength degradation observed during the tests. 2) Only the in-plane behavior of the solid
and perforated infill panels was considered in the computer model. The out-of-plane
behavior of the masonry wall was neglected. This consideration required special
hysteretic rules, in which the capacity of the elements in one direction (in-plane) is
coupled to the capacity of the element in the other direction (out-of-plane), like an
2'-7"
(Specimen 5)
3'-6"
(Specimen 5)
5'-0"
5'-0"
9'-6"
Component 3
(Specimen 4)
Component 5
36'-0" (Specimen 4)
5'-0"
Component 2
5'-0"
(Specimen 2)
5'-0"
3'-0"
Component 1
(Specimen 5)
3'-3" 5'-0" 3'-3" 1'-6" 5'-0" 1'-6"
5 6 7
Elastic
Spring
(R/C walls)
4 8
14 17
16 21
16
Component 7
Component 4
(Specimen 5)
(Specimen 5)
Component 6
(Specimen 4)
9
13 20
16
3
Component 3
(Specimen 4) 10
19
16
16
Center
of Mass Component 5
(Specimen 4)
2 12
Non-linear spring
(Bas ed in Components
Component 2
Rigid Link behavior)
(Specimen 2)
11 15 18
16
1
Component 1
(Specimen 5)
u2 u1 u1 u2
Lateral displacement, in
-18.00
K1=333.33 kips/in Intermediate loading-unl oading slope
K2=10.75 kips/in -28.00
K3=-7.50 kips/in
u1=-0.054 in
u2=-0.984 in
u2 u1 u1 u2
Lateral displacement, in
K1=230.04 kips/in
K2=-3.18 kips/in
u1=0.053 in
12.10 u2=1.968 in
6.00 K2
u2 u1 u1 u2
K1 Lateral displacement, in
-13.00
K2 -20.00 Intermediate loading-unloading slope
K1=108.33 kips/in K3
-26.20
K2=12.22 kips/in
K3=-5.74 kips/in
u1=-0.120 in
u2=-1.200 in
research to determine the seismic vulnerability of a typical house loaded in their weak
direction. The artificial records selected are based in the works done by Irizarry (1999)
spectra for Puerto Rico’s main municipalities of Mayagüez, Ponce and San Juan. The
proposed design earthquakes and design spectra were developed considering the geologic
conditions of the Island, which is located in the limit between the plates of the North
America and the Caribbean. The parameters used in the development of the artificial
records were obtained from worldwide strong motion records; among them are:
earthquake magnitude, focal depth, epicenter distance, site’s geology and fault-structure
type. She recommended that Mayagüez and Ponce are exposed to the same high seismic
Puerto Rico. The UBC-1997 locates the Island in a Seismic Zone 3, and it allows the
designers to assume a soil profile Type Sd when the soil profile at the site of the structure
develop an artificial earthquake record whose response spectrum will be compatible with
compatible with a target spectrum. For this research, Montejo supplied the artificial
earthquake record compatible with the Uniform Building Code 1997 for a Seismic Zone
Figure 6-4 shows the four earthquake records selected and Figure 6-5 shows a
comparison between the ground response spectrum of each record and the UBC-1997
(Zone 3-and soil Sd) design spectrum. The earthquakes response spectra correspond to a
peak ground acceleration of 0.36g and a 5 percent of damping ratio. The value of 0.36g
corresponds to the peak ground acceleration established by UBC-1997 for the assumed
soil properties.
135
0.6
Acceleration, g 0.4
0.2
0
0 10 20 30 40 50 60 70 80 90 100
-0.2
-0.4
0.4
Acceleration, g
0.2
0
0 10 20 30 40 50 60 70 80 90 100
-0.2
-0.4
0.4
Acceleration, g
0.2
0
0 10 20 30 40 50 60 70 80 90 100
-0.2
-0.4
0.4
Acceleration, g
0.2
0
0 10 20 30 40 50 60 70 80 90 100
-0.2
-0.4
1.2
Acceleration, g 1
0.8
0.6
0.4
0.2
0
0 0.5 1 1.5 2 2.5 3
Period, sec
UBC(1997-Sd) Chi Chi
0.8
0.6
0.4
0.2
0
0 0.5 1 1.5 2 2.5 3
Period, sec
UBC(1997-Sd) Mayaguez
0.8
0.6
0.4
0.2
0
0 0.5 1 1.5 2 2.5 3
Period, sec
UBC(1997-Sd) SanJuan
0.8
0.6
0.4
0.2
0
0 0.5 1 1.5 2 2.5 3
Period, sec
UBC(1997-Sd) SdM
In the non-linear time history analysis of the house, each earthquake is scaled
Figure 6-6 shows the base shear-lateral displacement (node 16-in the weak
direction of the house) curve of the house for earthquakes scaled at 0.36g. The structure
remained in the elastic regimen: it showed that it is capable to resist the seismic loading
generated by the selected earthquake records. The maximum damage expected in the
panels with a non-centered opening, as is the case of the Components 3, 5 and 6, is the
loss of the mortar adhesion between the roof slab and the masonry concrete block wall.
When the earthquakes were scaled 2.0 times the peak ground acceleration
the Mayagüez and San Juan earthquakes, as shown in Figure 6-7. The maximum
response was observed for the San Juan earthquake, where lateral load and lateral
displacement were 3.0 percent and 30.0 percent higher than those induced by the
Mayagüez earthquake, respectively. For the San Juan earthquake, the expected damage
are the growth of radial cracks at the interior and exterior reinforced concrete walls and
the roof slab due to punching shear load, and the beginning of corner crushing of the
Figure 6-8 shows that the house reached its maximum lateral capacity (V=150.0
kips, V=-163.0 kips) when the San Juan, Mayagüez and UBC-1997 compatible
earthquake records were scaled to a PGA=1.0 g, that is 2.77 times the peak ground
inches. At this stage, severe damage is expected in the house, such as punching shear
failure of the roof slab and the interior and exterior reinforced concrete walls, and the toe
crushing of the interior piers. Also it is expected that 57 percent of the components
All house components reached its maximum lateral capacity when the
earthquakes were scaled to 1.08g. Figure 6-9 shows the seismic response of the house at
Figure 6-10a shows the capacity-demand ratio of the house for the selected peak
ground accelerations, while the percent of drift is shown in Figure 6-10b. These damage
levels are shown in Figure 6-11 and summarized in Table 6-2. It is possible to conclude
that for a peak ground acceleration of 0.36g, the house can sustain the Mayagüez
earthquake with a safety factor of 1.66. For earthquakes with peak ground acceleration of
0.72 g, the safety factor is approximately equal to 1.10. The maximum lateral capacity of
the house (C/D=1.0) was reached when the earthquake records were scaled to 1.0g.
139
175
125
75
25
-25
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
-75
-125
-175
Lateral displacement, in
125
75
25
-25
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
-75
-125
-175
Lateral displacement, in
125
75
25
-25
-2 -1.5 -1 -0.5 0n 0.5 1 1.5 2
-75
-125
-175
Lateral displacement, in
Figure 6-6: Seismic response of the house for the selected earthquakes scaled to 0.36g.
140
175
125
75
25
-25
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
-75
-125
-175
Lateral displacement, in
125
75
25
-25
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
-75
-125
-175
Lateral displacement, in
125
75
25
-25
-2 -1.5 -1 -0.5 0n 0.5 1 1.5 2
-75
-125
-175
Lateral displacement, in
Figure 6-7: Seismic response of the house for the selected earthquakes scaled to 0.72g.
141
175
125
125
75
25
-25
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
-75
-125
-175
Lateral displacement, in
125
75
25
-25
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
-75
-125
-175
Lateral displacement, in
125
75
25
-25
-2 -1.5 -1 -0.5 0n 0.5 1 1.5 2
-75
-125
-175
Lateral displacement, in
Figure 6-8: Seismic response of the house for the selected earthquakes scaled to 1.0g.
142
175
125
125
75
25
-25
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
-75
-125
-175
Lateral displacement, in
125
75
25
-25
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
-75
-125
-175
Lateral displacement, in
125
75
25
-25
-2 -1.5 -1 -0.5 0n 0.5 1 1.5 2
-75
-125
-175
Lateral displacement, in
Figure 6-9: Seismic response of the house for the selected earthquakes scaled to 1.08g.
143
4
Level 1 Level 2 Level 3
3.5
3 T he maximum lateral
capacity of the house was
2.5
reached.
C/D
1.5
0.5
0
0 0.25 0.5 0.75 1 1.25 1.5
PGA
3.5
3
Percent of Drift
2.5
1.5
0.5
0
0 0.25 0.5 0.75 1 1.25 1.5
PGA
Table 6-2: Summary of the seismic response prediction of the typical houses.
Damage PGA, g Percent of Description
Level Drift
The structure remains in the elastic regimen. The damage expected in
1 0.36 0.03 masonry panel with non-centered opening, such as door openings, is
the loss of the mortar adhesion between the roof slab and the masonry
concrete block wall.
Radial cracks in the roof slab and the exterior and interior reinforced
concrete walls due to punching shear load. In the roof slab, the
2 0.72 0.58 punching shear load is generated by the toe reaction of the piers.
In the interior and exterior R/C walls the punching shear load is due to
compression load of the compression strut mechanism.
3 1.0 1.16 Severe damage is expected: punching shear failure of the R/C walls,
and roof slab, and toe crushing of piers.
145
CHAPTER 7
7.1 INTRODUCTION
The goal of this research was to establish experimentally the lateral capacity in
the weak direction of typical residential houses constructed in Puerto Rico. This goal
was accomplished by performing several tasks. First, in order to determine the typical
construction details of one-story residential houses used in the Island, a field survey study
was carried out. Second, using the results of the survey study, six full scale specimens
were selected, constructed, and tested. Each specimen simulated a slice of the typical
residential house where the behavior of the wall-slab frames and the masonry concrete
block walls under different configurations of openings and steel reinforcement were
validated using the experimental results and the failure mode of each specimen. Finally,
using the validated analytical model of each component of the house, a full house
computer model was constructed and used for seismic response predictions of the typical
Based on the results obtained during the research, a number of conclusions and
recommendations are presented next. The conclusions are subdivided into experimental
results, simple analytical models, seismic response predictions and general observations.
order to establish the lateral behavior and find the different failure modes of the houses
among them were wall-slab frame behavior, infilled frame behavior, unreinforced and
reinforced solid masonry panels, and unreinforced and reinforced perforated panel. The
following conclusions and findings are drawn based on the experimental results
1) Wall-slab frame vs. Infilled frame. The wall-slab frame exhibited a ductile
behavior, with plastic hinges developed at the bottom and top of the reinforced
concrete walls. During the test it was observed that the steel reinforcement details
used in the joints between the roof slab and the exterior R/C walls did not allowed
for the development of the full reversal moments, which is generated by the cyclic
opening in one side of the one story wall-slab frame increased the lateral strength
and lateral stiffness 5.12 times and 106 times, respectively. When the
unreinforced perforated panel is replaced by the reinforced solid panel the lateral
strength increased 7 times while the lateral stiffness increased 177 times. In both
cases, the inclusion of the infill panels significantly improved the lateral strength,
wall-slab frame with an unreinforced solid panel were governed by the masonry
corner crushing and punching shear failure of the reinforced concrete walls which
147
was caused by the internal strut action developed in the unreinforced infill panel.
Also, the lost of concrete that surrounds the vertical bars at the top of the wall 2
and wall 3 was other damage observed during the test. When the infill panel was
reinforced and connected to the wall-slab frame, the collapse mechanism was
governed by failure of the last row of the concrete block wall and the punching
shear failure of the R/C walls. The failure of the last row of concrete blocks was
produced by the hooked bars, which were used to connect the roof and floor slab
to the reinforced masonry wall. When the wall-slab frame was constructed with a
percent, depending of the load direction, was obtained compared with the
3) Unreinforced and Reinforced Perforated Panel. For the case that the wall-slab
frame was built with a reinforced masonry wall with a non-centered opening, the
mortar adhesion between the roof slab and the reinforced masonry wall was loss
in the first cycles of the load application. During the tests, two compression struts
were detected. The masonry corner crushing was caused by the formation of first
compression strut at the diagonal of masonry wall. Horizontal and radial cracks
were detected in the exterior wall 3 due to the reaction of the second strut which
was formed two feet below roof slab- exterior wall joint. The lateral capacity of
the reinforced perforated panel was 25.5 percent lesser than the obtained for the
panel with a centered opening, the lateral capacity was governed by the strength
of the interior and exterior piers. During the test it was observed that the pier
resisted the lateral load due to the external equilibrium produced the toe piers
reaction, which was developed when the rocking mode was activated. The failure
modes observed during the test were the toe crushing of the interior pier and
punching shear failure of the roof slab due to the vertical corner reaction of the
exterior pier. If the perforated panel is reinforced and attached with bond beams
to the roof slab, the failure modes were given by the piers shear failure. An
the load direction, was obtained compared with the unreinforced perforated panel.
infill panel reduced the lateral stiffness 33.66 percent and 37.0 percent compared
computer model of the full scale specimens. The computer models were
developed using the macro-model approach, where the different failures modes
observed experimentally between the wall-slab frames and the infill panels were
seismic behavior showed that when the resistances of the perforated and solid
houses, this has sufficient intrinsic strength to withstand the seismic loads
generated by earthquakes with PGA = 0.36 g (such as the one prescribe in UBC
1997) in the elastic regimen (Level 1). The maximum story drift of the typical
earthquakes were scaled 2.0 times the peak ground acceleration established by
UBC-1997 (i. e. PGA = 0.72g). Among the expected damage (Level 2) are the
growth of some radial cracks at the interior and exterior reinforced concrete walls
and the roof slab due to punching shear load, and the beginning of corner crushing
of the interior and exterior piers. At this damage level, the maximum story drift
When the earthquake records were scaled to a PGA = 1.0 g (2.77 times the
house, such as, punching shear failure of the roof slab and the interior and exterior
reinforced concrete walls, and the toe crushing of the interior pier. For these
numerical simulation of the seismic response, the single story residential house
150
with masonry wall oriented in its weak direction, withstood the UBC-1997
1) The findings presented here took into account the real current construction
2) The numerical predictions of the seismic response of the house provide useful
structural systems.
3) The findings may be incorporated into tools for emergency management agencies
to use not only for natural hazards planning but also for disaster mitigation and
loss assessment.
1) Perform experimental tests in two stories houses, where vertical irregulaties of the
hysteretic rules of the frames elements and multilinear compression link elements.
151
2) The new analytical tools should consider how the in-plane capacity of the infill
REFERENCES
Achyutha, H., Jagadish, R., Rao, P. S. and Shakeebur Rahman, S. 1986. Finite Element
Simulation of the Elastic Behaviour of Infilled Frames with Openings.
Computers & Structures, 23(5): 685-696.
Al-Chaar, G., Issa, M. and Sweeney, S. 2002. Behavior of Masonry Infilled Nonductile
Reinforced Concrete Frames. Journal of Structural Engineering, ASCE, 128(8):
1055-1063.
American Concrete Institute 2002. Building Code Requirements for Structural Concrete
(ACI 318-02) and Commentary (ACI 318R-02).
American Society of Testing and Material (ASTM). 2002. ASTM C476-02. “Standard
Specification for Grout for Masonry”. Annual Book of ASTM Standards,
September.
American Society of Testing and Material (ASTM). 2003. ASTM C144-03. “Standard
Specification for Aggregate for Masonry Mortar”. Annual Book of ASTM
Standards, June.
American Society of Testing and Material (ASTM). 2003. ASTM C1019-03. “Standard
Test Method for Sampling and Testing Grout”. Annual Book of ASTM
Standards, August.
American Society of Testing and Material (ASTM). 2003. ASTM C404-03. “Standard
Specification for Aggregate for Masonry Grout”. Annual Book of ASTM
Standards, August.
American Society of Testing and Material (ASTM). 2003. ASTM A615/A 615M-03a.
“Standard Specification for Deformed and Plain Billed-Steel Bars for Concrete
Reinforcement”. Annual Book of ASTM Standards, August.
153
American Society of Testing and Material (ASTM). 2003. ASTM C140-03. “Standard
Test Methods for Sampling and Testing Concrete Masonry Units and Related
Units”. Annual Book of ASTM Standards, October.
American Society of Testing and Material (ASTM). 2004. ASTM C270-03b. “Standard
Specification for Mortar for Unit Masonry”. Annual Book of ASTM Standards,
January.
American Society of Testing and Material (ASTM). 2004. ASTM C39/C 39M-04a.
“Standard Test Method for Compressive Strength of Cylindrical Concrete
Specimens”. Annual Book of ASTM Standards, November.
Bennett, R. M., Flanagan, R. D., Adham, S., Fischer, W. L. and Tenbus, M. A. 1996.
Evaluation and Analysis of the Performance of Masonry Infills during the
Northridge Earthquake. The National Science Foundation.
CSI Section Builder Program. Version 8.10. Analysis and Design of Concrete, Steel and
Composite Sections. Computers and Structures, Inc., Berkeley, California.
Crisafulli, F. J., Carr, A. J. and Park, R. 2000. Analytical Modeling of Infilled Frame
Structures – A General View. Bulletin of the New Zealand Society for
Earthquake Engineering. 33 (1): 30-47.
Dawe, J. L. and Seah, C. K. 1989. Behaviour of masonry infilled steel frames. Canadian
Journal of Civil Engineering, 16: 865-876.
154
Department of Commerce of the U.S.A. Economic and Statistics Administration. 1990.
Puerto Rico Census: Detailed Characteristics of Residences.
http://www.censo.gobierno.pr/Censo_Poblacion_ViviendaCH_2)_53_1990_TB17
.pdf
Drysdale, R. G. and Hamid, A.A. 1984. Tension Failure Criteria for Plain Concrete
Masonry. Journal of Structural Engineering, 110(2): 228-243.
Drysdale, R. G., Hamid, A.A. and Baker, L. R. 1999. Masonry Structures. Behavior and
Design. The Masonry Society, Colorado, 888 pp.
El-Dakhakhni, Wael W., Elgaaly, M. and Harmid, A. A. 2003. Three-Strut Model for
Concrete Masonry-Infilled Steel Frames. Journal of Structural Engineering,
ASCE, 129(2): 177-185.
FEMA 273, NEHPR Guidelines for the Seismic Rehabilitation of Buildings, FEMA,
October 1997.
FEMA 274, NEHPR Commentary on the Guidelines for the Seismic Rehabilitation of
Buildings, FEMA, October 1997.
FEMA 302, NEHPR Recommended Provisions for Seismic Regulations for New
Buildings and Other Structures, FEMA, 1997.
FEMA 306, Evaluation of Earthquake Damaged Concrete and Masonry Wall Buildings,
Basic Procedures Manual, FEMA, 1998.
FEMA 307, Evaluation of Earthquake Damaged Concrete and Masonry Wall Buildings,
Technical Resources, FEMA, 1998.
FEMA 308, Repair of Earthquake Damaged Concrete and Masonry Wall Buildings,
Technical Resources, FEMA, 1998.
FEMA 356, Prestandard and Commentary for the Seismic Rehabilitation of Buildings,
FEMA, November 2000.
155
Flanagan, R. D., Bennet, R. M., Adham, S. A. and Fischer, W. L. 1996. Masonry Infill
Performance during the Northridge Earthquake. Paper presented at the 1996
North American Masonry Conference, South Bend, Indiana.
Humar, J. M., Lau, D. and Pierre, J. R. 2001. Performance of Buildings during the 2001
Bhuj Earthquake. Canadian Journal of Civil Engineering, 28: 979-991.
Irizarry, J. 1999. Design Earthquake and Design Spectra for Puerto Rico’s Main Cities
based on Worldwide Strong Motion Records. MS Thesis, Civil Engineering
Department, University of Puerto Rico at Mayagüez.
Mainstone, R.J. 1971. On the Stiffnesses and Strengths of Infilled Frames. Proceedings
of the Institution of Civil Engineers. Supplement IV. 57-90.
156
Mallick, D. V. and Garg, R. P. 1971. Effect of Openings on the Lateral Stiffness of
Infilled Frames. Proceedings of the Institution of Civil Engineers 49: 193-209.
May, I. M. and Ma, S. Y. A. 1984. Computer Aided Analysis and Design of Shear Wall
Panel in Frames using Finite Element Method. Proc. Int. Conf. on Computer
Aided Analysis and Design of Concrete Structures, Split, Yugoslavia, pp 1033-
1047.
May, I. M. and Naji, J. H. 1991. Nonlinear Analysis of Infilled Frames under Monotonic
and Cyclic Loading. Computers & Structures, 38(2): 149-160.
Miranda, E. and Bertero, V. V. 1989. The Mexico Earthquake of September 19, 1985 –
Performance of Low-Rise Building in Mexico City. Earthquake Spectra 5(1):
121-143.
Moghaddam, H. A. and Dowling, P. J. 1987. The State of the Art in Infilled Frames.
ESEE Research Report No. 87-2. Imperial College of Science and Technology,
Civil Engineering Department, London.
Mosalam, K. M., White, R. N. and Gergely, P. 1997. Static Response of Infilled Frames
using Quasi-Static Experimentation. Journal of Structural Engineering, ASCE,
123(11): 1462-1469.
Rai, D. C. and Goel, S. C. 1996. Seismic Strengthening and Unreinforced Masonry Piers
with Steel Elements. Earthquake Spectra, 12(4): 845-862.
Reinhorn, A. M., Madan, A., Valles, R. E., Reichmann, Y. and Mander, J. B. 1995.
Modeling of Masonry Infill Panels for Structural Analysis. Report No. NCEER-
95-0018, National Center for Earthquake Engineering Research, State University
of New York, Buffalo, New York.
Schneider, S. P., Zagers, B. R. and Abrams, D. P. 1998. Lateral Strength of Steel Frames
with Masonry Infills Having Large Openings. Journal of Structural Engineering,
ASCE, 124(8): 896-904.
Sezen, H., Elwood, K. J., Whittaker, A. S., Mosalam, K. M., Wallace, J. W. and Stanton,
John F. 2000. Structural Engineering Reconnaissance of the August 17, 1999,
Kocaeli (Izmit), Turkey, Earthquake. Report No. PEER 2000/09, Pacific
Earthquake Engineering Research Center, University of California, Berkeley,
California.
Stafford Smith, B. 1967. Methods for Predicting the Lateral Stiffness and Strength of
Multi-Storey Infilled Frames. Building Science, 2: 247-257.
Stafford Smith, B. and Carter, C. 1969. A Method of Analysis for Infilled Frames.
Proceedings of the Institution of Civil Engineers. 44:31-48.
158
Suarez, L. E. and Montejo, L. A. 2005. Generation of artificial earthquakes via the
wavelet transform. International Journal of Solids and Structures 42:5905–5919.
A.1 INTRODUCTION
The usefulness of the experimental results found in this research is based in that
the selected and experimentally investigated specimens simulate the common methods
and typical construction details of the residential houses in the Island. A way to obtain
these construction details of these houses is through a field survey. This field survey
included reviews of the construction drawings and visits to the field which were focused
in two areas: identify the process that generally is followed during the construction of
these residences and obtain their typical constructions details, such as steel reinforcement
patterns, connection types, and structural element dimensions. Based in the results of the
survey study, six full scale specimens were selected and constructed.
structural element dimensions, typical connection details between floor/roof slab and
houses randomly selected throughout Puerto Rico. During the field visits, special
attention was paid to the construction sequence of the houses, which was followed in the
construction of the full scale specimens and showed in Figure A-1. Once the
construction sequence was obtained, the next step was to obtain the steel reinforcement
160
details of the twenty residences and to select the construction details to be used during the
Figure A-2 to A−9 show the eight typical details previously mentioned, while
Table A-1 to A−8 present a summary of the results obtained using the construction
drawings. For each table, the average and the mode values of each construction details or
steel reinforcement amount of the twenty houses were calculated. In general, the steel
reinforced amount and construction details selected for full scale specimens lie between
the average and modes values. These specimens were constructed in the Structural
Mayagüez.
Finally, Table A-9 presents a comparison between the field survey results and the
selected properties for the full scale specimens that were constructed.
161
e) Concrete walls oriented in the strong f) Typical slab concrete wall frames.
direction.
As-H
As-V
Concrete Wall
As-H Concrete
Block Wall
As-V
Concrete Wall
Ld wall As-J
T slab Ld footing
T key L hook
B footing B key
Figure A-4: Typical joint details between foundation and concrete wall.
Concrete
Block Wall
As-J
Ld wall
T slab Ld footing
T key L hook
B footing B key
Figure A-5: Typical joint details between foundation and concrete block wall.
Ld roof T roof
As-J Ld wall
Figure A-6: Typical joint details between roof and interior concrete wall.
As-B
Ld dowel
Concrete Wall
T wall
Figure A-7: Typical joint details between roof and exterior concrete wall.
As-Temp
As-B
Concrete Wall
T wall
Concrete Wall
As-T
T slab
As-Hk
x
As-B
MATERIAL PROPERTIES
B.1 INTRODUCTION
In order to characterize the different failure modes of the full scale specimens and
properties of the materials used in the construction of the full scale specimens. These
materials are representatives of the real material used in the common construction
practices of the residential houses in Puerto Rico. The investigated materials were the
same as the ones that were used in the construction of the wall-slab frames and masonry
components. The mechanical properties determined were the compressive strength of the
reinforced concrete and masonry components and the tensile strength of the
reinforcement steel bars. These properties were determined through experimental tests,
following the specifications presented by the American Society of Testing and Material
(ASTM). The following sub-sections present in detail the experimental tests performed
to the wall-slab frames and masonry components materials as well their experimental
In order to approximate the strength of the materials used during the construction
process of the residential houses in Puerto Rico, compressive cylinder strength of 3.0 ksi
and tensile yield stress of the reinforcement bars of 60.0 ksi were selected in the assembly
of wall-slab frames. Two other important criteria in the concrete mix design were
specified: the aggregate size and good workability. Coarse aggregate with a nominal
maximum size of 3/4 inches was used. Good workability was required to achieve proper
consolidation due to difficult placement of the concrete into vertical elements, such as
reinforced concrete walls. A slump of 5.0 inches was specified during the concrete
mixture design. The mixture was supplied by a concrete supplier of the locality and the
The mechanical properties of the reinforcement steel bars and the compressive
ASTM A615/A 615M-03a (Standard Specifications for Deformed and Plain Billet-Steel
Bars for Concrete Reinforcement) and ASTM C39 (Standard Tests Method for
Chapter 3. These phases were the construction of the fill material, floor slab, reinforced
concrete walls and roof slab. From each batch of the concrete used in the construction of
each specimen phase, at least 3 cylinders of 6.0 inches diameter by 12.0 inches high were
cast. The compression tests of the cylinders were performed at approximately the same
173
date of corresponding specimen test (approximately six months). All cylinders were air
The masonry components such as mortar, grout, concrete block units, and
masonry assemblies were constructed with the same materials and proportions that were
used in the construction of the masonry concrete block walls of Specimens 2 to through
6.
B.2.2.1 Mortar
The ASTM C270-03b (Standard Specification for Mortar for Unit Masonry)
covers mortar for use in the construction of non-reinforced and reinforced unit masonry
materials, well graded sand, and sufficient water to produce a plastic, workable mixture.
The mortar is used as bonding agent that holds the concrete blocks, reinforcement, and
connectors together to act as a complete assembly. The mortar mixture design can be
specifications. These two methods cover four types of mortar: M, S, N, and O. Table
B-1 and Table B-2 summarize the proportion specification requirements and property
specification requirements for the different types of mortar, respectively. Although the
ASTM C270-03b recommend proportions for the mortar materials based on the mortar
type, for the mortar used in the construction of the masonry concrete block wall (MCBW)
of Specimens 2 through 6, the proportions of the mortar materials were based on the
current construction practice of residential houses in Puerto Rico. Table B-3 presents the
174
proportions of the materials for mortar, such as Portland cement, manufactured sand, and
The Portland cement and manufactured sand were obtained from hardware stores
of the locality. Figure B-1 shows a gradation analysis performed to the manufactured
sand and a comparison with the upper and lower limits established by ASTM 144-03
During the construction of the masonry concrete block walls (MCBW) for
(Preconstruction and Construction Evaluation of Mortars for Plain and Reinforced Unit
Masonry ). From each batch of the mortar used in the construction of MCBW, at least of
3 cylinders of 2.0 inches diameter by 4.0 inches high were cast. The compression tests of
the cylinders were performed at approximately the same time of the corresponding
specimen test. All cylinders were air cured in the laboratory under the same conditions as
B.2.2.2 Grout
The ASTM C 476-02 “Standard Specification for Grout for Masonry” covers two
types of grout, fine and coarse grout, for use in the construction of masonry structures.
The fine grout is made with fine aggregate and the coarse grout is made with a
combination of coarse and fine aggregates. Grout is a mixture composed of sand, gravel,
Portland cement and sufficient water to produce fluid mixture with slump ranges of 8.0 to
11.0 inches. The grout is used to bond masonry units to the steel reinforcement or to
bond together adjacent masonry units. It is placed in the cores of hollow masonry units.
175
Table B-1: Proportion specification for masonry mortar
(adopted from ASTM C 270-03b).
Portland Proportions by Volume (Cementitious Materials)
Cement Mortar Cement Masonry Cement
Mortar Type or Hydrated Lime
Blended or Lime Putty
Cement M S N M S N
M 1 - - - - - - 1/4
Cement-Lime S 1 - - - - - - over 1/4 to 1/2
N 1 - - - - - - over 1/2 to 1 1/4
O 1 - - - - - - over 1 1/4 to 2 1/2
M 1 - - 1 - - - -
M - 1 - - - - - -
Mortar S 1/2 - - 1 - - - -
Cement S - - 1 - - - - -
N - - - 1 - - - -
O - - - 1 - - - -
M 1 - - - - - 1 -
M - - - - 1 - - -
Masonry S 1/2 - - - - - 1 -
Cement S - - - - - 1 - -
N - - - - - - 1 -
O - - - - - - 1 -
The total aggregate shall be equal to not less than 2 ¼ and not more than 3 times the sum of the
volumes of the cement and lime used.
100
80
Percent Passing
60
40
20
0
0.20 0.15 0.10 0.05 0.00
Particle diameter (in)
Figure B-1: Particle size distribution curve for manufactured sand for mortar.
When the proportions design is used, the proportion by volume will be established
following the ASTM C 476-02. Table B-4 presents these grout proportions. For strength
requirements design, the grout is sampled following the ASTM C1019-03 (Standard Test
Method for Sampling and Testing Grout). When the strength requirements are used in
the grout mixture design, the grout can be proportioned to have a compressive strength
equal to or exceeding the specified compressive strength of the masonry, f’m, but not less
Table B-5 presents the proportions of the Portland cement, gravel, manufactured
sand, and water used in the grout mixture for Specimens 2 through 6. Instead of use the
strength or ASTM C476-02 requirements to grout mixture design, the proportions of the
materials were obtained from the current construction practice of residential houses in the
Island.
Similar to the mortar aggregates, the gravel and manufactured sand were obtained
from hardware stores of the locality. Grading analysis was carried out to gravel and
manufactured sand. Figure B-2 presents the particle size distribution of the manufactured
sand and their limits established by the ASTM C404-03 (Standard Specification for
Aggregate for Masonry Grout). Figure B-3 presents the particle size distribution of the
100
80
Percent Passing
60
40
20
0
0.20 0.15 0.10 0.05 0.00
Particle diameter (in)
Figure B-2: Particle size distribution curve for manufactured sand for grout.
100
80
Percent Passing
60
40
20
0
0.60 0.50 0.40 0.30 0.20 0.10 0.00
Figure B-3: Particle size distribution curve for gravel for grout.
179
The grout was sampled and tested using the ASTM C 1019-03 (Standard Test
Method for Sampling and Testing Grout). The grout tested according to ASTM C 1019-
03 was the same one used in the constructions of the MCBW of the Specimens 2 to 6.
Two main tests were performed to the grout: slump test and the compression tests. The
grout slump was 8.0 inches. For compression tests, at least 3 square molds with sides of
3.5 inches by 3.5 inches and height of 7.0 inches were cast.
Figure B-4 presents the typical dimensions of the grout specimens and the molds
used during the sampling of grout. The grout specimens were air cured in the laboratory
under the same conditions of the MCBW of the specimens. The compression tests of
grout specimens were carried out approximately at the same time of specimen tests.
The physical and mechanical properties of the concrete block units were obtained
according to the ASTM C140-03 (Standard Test Methods for Sampling and Testing
Concrete Masonry Units and Related Units). These tests methods cover the sampling and
testing of concrete masonry units for dimensions, compressive strength, absorption, unit
weight, and moisture content. Six units were selected to determine the properties of the
concrete block, three units to determine the physical properties and the other three to
determine the mechanical properties. For each masonry unit, the width, height, length,
and minimum thicknesses of face and shell webs were measured. Also, the average net
area, unit weight, moisture content, and water absorption were determined for each
concrete block unit. Figure B-5 presents the typical average dimension of the concrete
block.
180
3 1/2"
Concrete block
Grout Specimen
Paper towel
7"
Nonabsorbent Block
L=15.50"
tfs=1.02" tw=1.05"
A A
H=7.75" W=5.68"
L=15.50"
B
L=15.75"
SIDE VIEW TOP VIEW
The compressive strength of the concrete block units was determined following
the requirements of the ASTM C140-03. In order to achieve uniform stress distribution
on the bearing surfaces of the concrete block unit during the compressive test, the top and
bottom of units were capped with Hydrostone Gypsum. Figure B-6 presents the concrete
block units capped according to the ASTM C1552-03a (Standard Practice for Capping
Concrete Masonry Units, Related Units and Masonry Prisms for Compression Testing).
Figure B-7 presents the set-up used to perform the compression tests of the
capped concrete masonry units, which consisted of: (1) Forney Machine with load
capacity of 3000,000 pounds and (2) Potentiometer with maximum deformation of 1.0
inch to measure the compressive deformation of the concrete block unit. The
compression load was applied at constant rate of 1000 pounds per second. Using this
loading rate, the failure load was obtained within 1 to 2 minutes, which is a requirement
the center of each face shell of the concrete block unit. These displacement
measurements were converted into axial strains dividing these by the gage lengths of the
potentiometer. The compression stress was calculated dividing the compression load by
Forney Machine
(300,000 lb)
Steel plate
t = 1.0"
Concrete block
Potentiometer (1.0")
Gage length
of 4.0"
Hydrostone Gypsum
Capping
Figure B-7: Set up used to determine the compressive properties of the concrete block
units.
184
B.2.2.4 Masonry Assemblies
Masonry assemblies consisted of concrete block unit, mortar, and grout. The
compression capacity of the masonry assemblies was obtained by prism tests. The prism
tests were made according to the ASTM C1314-03b (Standard Test Method for
Compressive Strength of Masonry Prisms). This test method covers procedures for
masonry prism construction and testing, and procedures for determining the compressive
For each specimens (2 to 6), six prisms were constructed, three prisms ungrouted
and three prisms fully grouted. For each prism unit, the length and width at the edges of
the top and bottom faces were measured. Figure B-8 presents the average length and
width of the prism unit which were calculated by averaging the four measurements of
each dimension.
The prism units were capped and tested following the similar procedure to use in
the concrete block units. Figure B-9 presents the set up used to perform the compression
tests of the capped prism unit, which consisted of: (1) Forney Machine and (2)
The ASTM 1314-03b requires that the load should be applied at uniform rate not
less than 1 minute and not more than 2 minutes. This requirement was reached applying
the compression load at constant rate of 1000 pounds per second. The axial displacement
of prism unit was measured using two potentiometers, mounted at the center of each face
shell of the prism units. These two displacement measurements were converted into
185
L=15.50"
W=5.68"
L=15.75" 13.00"
L=15.50"
A A
W=5.68"
L=15.75" 13.00"
Forney Machine
(300,000 lb)
Steel plate
t=1.0"
Concrete block
Potentiometer
(4.0")
Gage length
of 8.0"
Hytrostone Gypsum
Capping
Figure B-9: Set up used to determine the compressive properties of the ungrouted and
grouted prisms under monotonic loads.
187
axial strains dividing these by the gage length of the potentiometer. For ungrouted prism,
the compression stress was determined by dividing the compression load by the average
net area, while the gross area was used to determine the compression stress in the grouted
prism.
Figure B-6 shows the average compressive cylinders strength of concrete used in
the wall-slab frame construction. The concrete mix was provided by two different
concrete mix suppliers. The first provider supplied the concrete mixture for the
Figure B-10 shows the tensile Stress-Strain curve for a typical No. 3 bar. The
tensile stress and strain at yield were 60.0 ksi and 2176 micro strains, respectively.
70
60
50
Stress σ, ksi 40
30
20
10
0
0 2000 4000 6000 8000 10000 12000
Strain ε, in/in
B.3.2.1 Mortar
Table B-7 summarizes the average compressive strength of the mortar used in
strength was attained following the proportions used in the construction practice in the
Island. Because the compression stresses of the mortar for all specimens were greater
than 2,500 psi, these were classified as mortar type M (see Table B.2).
The average compression stress of the grout used during the construction of the
specimen’s concrete block walls are shown in Table B-8. Only the compression stress of
the grout obtained for Specimens 4 and 6 were lesser that the minimum compression
stress (2,000 psi) allowed by the Section 2104.4 of Uniform Building Code 1997.
Table B-9 and Table B-10, summarize the average values of the physical and the
mechanical properties of the concrete block units, respectively. The averages values
presented in the tables are based on the results of three samples. The masonry units used
in Specimens 4 to 6 showed higher compressive strengths than the one observed in the
Specimens 2 and 3.
Figure B-11 shows the compression stress-strain curve for the concrete block
units. The concrete block showed linear behavior until its maximum compression stress
of 1707 psi and strain (ε) of 0.001828 in/in. After the maximum compression stress was
reached, the block unit started to unloaded, losing its capacity to sustain load. Figure
1.5
Stress,ksi
0.5
0
0 0.005 0.01 0.015 0.02 0.025 0.03
Strain ε, in/in
Figure B-11: Compression Stress Strain Curve for concrete block units.
191
Figure B-12: Failure mechanism of the concrete block units under compression load.
The compression stress for ungrouted and grouted masonry prisms was
determined following the procedure presented in the ASTM C1315-03: (1) Masonry
prisms strength was determine by dividing each prism’s maximum compressive load by
its cross sectional area. (2) Multiply the compression stress of the masonry prism by the
correction factors due to the slenderness ratio (height/width) of the prisms, which are
presented in Table B-11. (3) Averaging the corrected compression stresses of the three
Figure B-13 shows a comparison between compression stress-strain curve for the
concrete block, and the ungrouted and grouted masonry prisms. The overall behavior of
masonry prisms was very similar. Unlike to the concrete block which failed in crushing,
the prisms failed in shear, showing a great loss in capacity and failing to sustain
compression load. The shear failures observed in the ungrouted and grouted masonry
1.5
Stress,ksi
0.5
0
0 0.005 0.01 0.015 0.02 0.025 0.03
Strain ε, in/in
Grouted masonry prism Ungrouted masonry prism Concrete block
Figure B-13: Compression Stress Strain Curve for concrete block, ungrouted and grouted
masonry prisms.
193