13th World Conference on Earthquake Engineering

Vancouver, B.C., Canada

August 1-6, 2004
Paper No. 1093

EXPERIMENTAL BEHAVIOR OF MASONRY STRUCTURAL WALLS

USED IN ARGENTINA.

Francisco ZABALA1, José Luis BUSTOS2, Alberto MASANET3, Jorge SANTALUCÍA4

SUMMARY

Confined masonry is extensively used in seismic regions of Argentina. Experimental data about confined
masonry built using local practice are very scarce and this lack of knowledge affects the seismic safety and
the design practice of masonry structures. Some test results from real scale masonry models are presented
in this paper. The models were designed based on the typical layout of a three story residential building
and were built using the common practice in the Region of Cuyo in Argentina. Tests were performed for
different displacement levels, with displacement control for lateral loads and load control for vertical
loads. Six models of confined masonry walls with different column reinforcement were tested. Two of
them had horizontal reinforcement placed at the joints. Behavior of the models during the test, as well the
initial stiffness, strength, failure modes and plastic strain capacity are described in the paper. On the basis
of the experimental results some conclusions are drawn in relation to strength and stiffness estimations by
the Argentine building code, the influence of the vertical and horizontal reinforcement in the actual
strength of the wall and the performance under severe earthquakes.

1
Director, Instituto de Invest. Antisísmicas, UNSJ, San Juan, Argentina. Email: fzabala@unsj.edu.ar
2
Professor, Instituto de Invest. Antisísmicas, UNSJ, San Juan, Argentina. Email: jlb@unsj.edu.ar
3
Professor, Instituto de Invest. Antisísmicas, UNSJ, San Juan, Argentina.
4
Professor, Instituto de Invest. Antisísmicas, UNSJ, San Juan, Argentina.
 INTRODUCTION

Confined masonry structures are convenient from an economic point of view. They are commonly used in
the seismic regions of Argentina for buildings up to three floors. Normally this kind of building has
masonry walls in two main perpendicular directions, joined by reinforced concrete slabs. Seismic action,
represented by lateral forces applied to each floor and to the roof, is resisted by a mechanism of walls,
coupled by lintels and sills, and connected by the slabs. Slabs are assumed to behave as non-deformable
diaphragms, being able to distribute the lateral forces to the walls.

It is possible to estimate the theoretical flexural strength of a confined masonry wall in a simple way, by
considering the amount of columns reinforcement, vertical load supported by the wall and yield stress of
the reinforcement steel. On the other hand, the cracking shear load and the maximum shear strength of
these walls are more uncertain, since they depend on several factors like: individual brick strength, mortar
and workmanship qualities, vertical load, amount of columns reinforcement and amount of horizontal
reinforcement embedded in the masonry. In addition, the manufacturing conditions of the bricks and the
walls are very variable, causing high dispersion of the resulting mechanical properties.

DESCRIPTION OF THE TESTS

In order to obtain better knowledge about the seismic behavior of confined masonry walls used in the
seismic region of Argentina tests were performed on six real scale model walls at the Earthquake Research
Institute of the National University of San Juan (IDIA). The walls were built with handmade solid ceramic
bricks, 18 cm wide. Wall confinement was provided by reinforced concrete columns, with nearly square
sections, 20cm wide by the thickness of the wall. The design of the tested models was based on the typical
building layout used by the San Juan Provincial Institute of Housing (IPV). The model dimensions are
shown in Figure 1. The materials used for the walls were characterized by testing brick piers under simple
compression and small masonry probes under diagonal compression, according to the Inpres-Cirsoc 103
Code [1]. A summary of the results of material characterization tests are presented in Table 1 [2]-[3]

Table 1. Summary of tests on brick piers and small walls.

Mortar σm σc τm
(mixing ratio C:L:S) [MN/m2] [MN/m2] [MN/m2]
Normal strength 4,1 2,7 0,22
(1:1:5)
Intermediate strength 5,0 2,9 0,28
(1:1/2:4)
High strength 8,7 6,1 0,31
(1:0:3)

• [σm]: Mean compression strength of piers.

• [σc]: Characteristic compression strength of piers.
• [τm]: Mean shear strength from diagonal shear tests.
• Solid ceramic bricks: mean compression strength 8.2 MN/ m2, characteristic compression strength
4.5 MN/ m2.
• Mean elasticity modulus measured in piers under simple compression: E= 1600MN/ m2
 3.00m
0.20 2.60 0.20 0.20

0.20

2.60m 3.00m

0.20

0.50

0.50
3.80 0.70

Figure 1. Model dimensions.

The walls were tested under a prescribed constant vertical load and allowing free rotation of the upper
end. The vertical load was applied through a stiff steel beam by means of two vertical servo-controlled
actuators. The tests were performed by applying cycles of lateral displacements at the wall head. An
outline of the test setup and its instrumentation is presented in Figure 2. The instrumentation consisted of
one displacement transducer controlling the horizontal displacement of the wall head (L.V.D.T.1), two
vertical displacement transducers at both sides of the model (L.V.D.T 2 and 3), two diagonal displacement
transducers (L.V.D.T. 4 and 5), three load cells mounted in series with the hydraulic jacks and a number
of strain gages applied to some columns reinforcement bars.
 Figure 2.

ANALYSIS OF RESULTS

Table 2 summarizes the main features of the six tested walls: columns reinforcement, horizontal masonry
reinforcement, vertical load, theoretical flexural capacity, estimated shear capacity using the expressions
of the Inpres-Cirsoc 103 Code, and the maximum strength measured during the tests [4]. Values of
theoretical flexural capacity, which vary with vertical load and columns vertical reinforcement, have been
calculated considering the point of application of the horizontal load and the nominal yield strength of
steel bars (420 MN/ m2).

In walls 1 and 2, the amount of column reinforcement causes the flexural capacity to be slightly larger
than the shear capacity estimated using the expressions of Inpres-Cirsoc 103 Code. In the case of the walls
3 and 4 the flexural capacity is several times larger than the shear capacity. In the third group of walls (5
and 6) the amount of vertical reinforcement was decreased and horizontal reinforcement embedded in the
masonry was added, in order to ensure the shear capacity to be larger than the flexural capacity.

Table 2. Tested walls features and results.

 Theoretical Estimated Maximum
Wall Vertical Horizontal Vertical load. flexural shear capacity. measured
reinforcement reinforcement capacity (1) (2) strength
[kN] [kN] [kN] [kN]

4 φ10
1 (3.12 cm2) - 100 142 109 118

4 φ 10
2 (3.12 cm2) - 100 142 109 93

4 φ 16
3 (8.05cm2) - 200 342 138 207

4 φ 16
4 (8.05cm2) - 200 342 138 235

2φ6
4φ8 each 2 mortar
5 100 105 109+ 72 (3) 157
(2.01 cm2) joint.
(3.1 cm2/m)
2φ6
4φ8 each 2 mortar
6 100 105 109+ 72 (3) 169
(2.01 cm2) joint
(3.1 cm2/m)

Notes:
(1) Considering the horizontal load applied at the horizontal actuator level, the applied vertical load and
σs= 420 MN/ m2 (yield stress of the steel)
(2) Vur= (0.3 σ +0.6 τmo)[1]. Where σ = compression stress, τmo = diagonal shear strength of small
masonry probes. τmo= 0.3 MN/ m2
(3) Additional strength due to horizontal masonry reinforcement.

The models 1 to 4 developed the crack pattern presented in the sequence of Figure 7. This pattern includes
diagonal cracking of the masonry panel and partial separation of the confinement columns. These walls
clearly show a shear failure, but sustained their strength for a displacement up to 20 mm. Figure 6
summarizes the envelope curves of all tests. None of these walls reached their theoretical flexural capacity
and the final state was controlled by the columns shear strength. This is due to the fact that, under large
displacements, diagonal cracking of masonry extended to columns. Compression failure never occurred
nor the emptying of the cracked masonry panel (See Figures 8 and 9).
 250

200 Wall 1
Wall 2
150 Wall 3
P[kN]

Wall 4
Wall 5
100
Wall 6

50

0
0 5 10 15 20 25 30 35 40 45
Def.[mm]
Figure 6. Load-displacement envelope curves.

The walls 1and 2 showed similar initial stiffness and cracking strength to those measured on walls 3 and
4. The latter sustained higher values of lateral load due to a combined truss frame mechanism in which
diagonal compression of masonry is limited only by the shear strength of the confinement column, since
vertical reinforcement bars do not reach yielding in tension. Flexural and shear capacities of the plastic
hinge in the column are larger in these walls than in wall 1 and 2. In addition, greater columns-panel
separation was observed in walls 3 and 4.

Walls 5 and 6, having a shear capacity clearly larger than the flexural capacity, reached, by hardening of
the vertical reinforcement bars, strength values substantially larger than the theoretical flexural capacity.
Under the applied displacement cycles, with increasing amplitude, it was observed that these walls (Figure
10) maintain their strength and their energy dissipation ability for larger displacement amplitudes than
walls 1 to 4. Bending-induced horizontal cracking was observed and the separation between column and
panel did not occur (Figure 12). The final state is controlled again by the shear strength of the column at
the joints with the confinement beams. Figure 11 shows all the displacement cycles applied to wall 6.
Each displacement amplitude was applied two times. It is possible to observe the stiffness degradation
between the first and the second cycle with the same displacement amplitude.

The initial secant stiffness, measured at one per thousand drift, was higher for walls 5 and 6 than for walls
1 to 4. This is because the horizontal masonry reinforcement prevents the separation between column and
the masonry panel. Table 3 shows the estimated stiffness by simple theory and the measured stiffness for
0.5 and 1 per thousand drift.
 MURO2 carga-deformación

Ca
rg
a
La
t.[t
n] -30 -20 -10 0 10 20 30

-
Desp.[mm]

MURO2 carga-deformación

Car
ga
Lat.
[tn]

-30 -20 -10 0 10 20 30

Desp.[mm]

M URO2 c a rg a -d e s p la z a m ie n t o

Carga Lat.[tn]

- 30 -2 0 -1 0 0 10 20 30

-
D es p .[m m ]

MU RO2 ca rg a -d e s p la z a m ie n to
Carga Lat.[tn]

-3 0 -2 0 -1 0 0 10 20 30

-
D e s p .[m m ]

MURO2 c a r g a -d e fo rm a c ió n
Carga Lat.[tn]

-30 -2 0 -10 0 10 20 30

-
D e s p .[ m m ]

MURO2 c a r g a -d e fo r m a c ió n

10

5
Carga Lat.[tn]

0
-3 0 -2 0 -1 0 0 10 20 30

-5

-10
D e s p .[ m m ]

Figure 7. Damage sequence for wall 2.

 Figure 8. Final cracking for wall 4.

Figure 9. Shear failure at column in wall 1.

 200

150

100

C arga Lat.[kN ]
50

0
-35 -25 -15 -5 5 15 25 35
-50
D esp.[mm]
-100

-150

-200

200

150

100

C arga Lat.[kN ]
50

0
-35 -25 -15 -5 5 15 25 35
-50
D esp.[mm]
-100

-150

-200

200

150

100

Carga Lat.[kN]
50

0
-35 -25 -15 -5 5 15 25 35
-50
Desp.[mm ]
-100

-150

-200

200

150

100
Carga Lat.[kN]

50

0
-35 -25 -15 -5 5 15 25 35
-50
Desp.[mm]
-100

-150

-200

200

150

100
Carga Lat.[kN]

50

0
-35 -25 -15 -5 5 15 25 35
-50
Desp.[mm]
-100

-150

-200

200

150

100
Carga Lat.[kN]

50

0
-35 -25 -15 -5 5 15 25 35
-50
Desp.[m m ]
-100

-150

-200

Figure 10. Damage sequence for wall 6.

 WALL 6
20

15

10
Lateral force [KN]

0
-40 -30 -20 -10 0 10 20 30 40

-5

Desp.[mm]
-10

-15

-20

Figure 11. Wall 6 tests.

Figure 12. Final state of the wall 6.

 Table 3. Estimated and measured secant stiffness
0 0
Wall K e[kN/mm] K 0,5 /00[akN/mm] K 1 /00[kN/mm]

1 52 39

1
2 3
31 23
h h = 39
+
3EmJm GmAm
3 50 37

4 52 45

1
5 3
72 47
h h = 74
+
3EmJt GmAm
6 83 54

h : wall height.
Em=1600 MN/m2.Young modulus.
Gm=500 MN/m2. Shear modulus
Am= Wall section without transformation of confinement columns sections.
Jm= Inertia moment without transformation of confinement columns sections.
Jt=. Inertia moment with transformation of confinement columns sections.
K 10/00 : Secant stiffness for 1 0/00 drift.

CONCLUSIONS

The Inpres-Cirsoc building code allows a reasonable estimation of the wall strength, based on measured
strength in diagonal shear test of small masonry probes, only in the case of lightly reinforced columns, not
providing larger flexural capacity than shear capacity (See Table 2). For larger reinforcement ratios of
columns, the wall strength is controlled by the shear strength of the confinement columns and beam joints.
The code should require the capacity design of columns and joints reinforcement, considering the
maximum expected shear force induced by the compressed masonry strut, arising from the cracking
pattern of the panel. For the used brick type, a compression failure of this strut is not likely to occur and
therefore the wall strength becomes controlled by the vertical reinforcement of the columns. The amount
of transverse reinforcement in critical zones of the confinement columns and beams normally used in
practice is insufficient in order to sustain this shear force.

Placing horizontal reinforcement bars with enough cross section ratio (0.18 %) so as to control the
diagonal cracking and to increase shear strength, together with low ratios of vertical steel, proved to
radically change the failure mechanism of the tested walls and increased the initial stiffness and plastic
strain capacity. The Inpres-Cirsoc Code should include the horizontal reinforcement in computing the
shear strength of the wall. It must be noted that the reinforcement steel placed in the tested walls is
slightly larger than the minimum recommended by the code (2 φ 4.2 c/50 cm).

The secant stiffness of walls without horizontal reinforcement, for a drift level of 0.001, can be estimated
in a simple way by considering the uncracked gross section of the wall as homogeneous. On the other
hand, for walls with horizontal reinforcement steel in which the confinement column do not separate from
the panel, it is possible to estimate the initial stiffness by considering the transformed and cracked sections
of the confinement columns.

REFERENCES

[1]. Inpres-Cirsoc 103. “Normas Argentinas para construcciones sismorresistentes”. Parte III.
Construcciones de Mampostería.1983

[2]. Juan J. Muñoz. “Estudio del comportamiento sismorresistente de edificios de mampostería”. Trabajo
final. Facultad de Ingeniería. U. Nacional de San Juan.1999

[3]. Francisco Zabala y otros. “Informe Final Proyecto PID 0646”. Facultad de Ingeniería. Universidad
Nacional de San Juan.

[4] Zabala, F, Bustos, J.L, Masanet, A,R y Santalucía J,R (2002). “Aspectos del diseño de muros de
mampostería encadenada bajo cargas sísmicas”. Jornadas Sul-Americanas de Engenharia Estrutural.