You are on page 1of 10

13th World Conference on Earthquake Engineering

Vancouver, B.C., Canada August 1-6, 2004 Paper No.257

Marc BOUCHON1, Nebojsa ORBOVIC2, Bernard FOURE3


The test program consisted of mechanical tests performed on reinforced concrete walls subjected to alternating shear loads. The aim of the program was to define crack geometry. Tests were carried out on 3 reinforced concrete walls with low slenderness ratios and varying percentages of steel. Each wall was subjected to a sequence of three increasing loads. One of the three walls was loaded to failure. The main results of the program were measurements of displacement as a function of horizontal force, cracking states during the loading cycle, and deformation of rebars. The measurements obtained were used to define a relationship between crack geometry and diagonal elongation, and to verify the suitability of the formula given in the CEB-FIP-1978 model code as a means of assessing crack spacing and opening.


The French Institute for Radiological Protection and Nuclear Safety (IRSN) conducts expert appraisal assignments and research in the nuclear safety fields and provides technical support for the French Nuclear Safety Authority. In this field, earthquakes are one of the key external events to be considered when assessing the behavior of civil engineering structures. French nuclear facilities are mostly built using low-rise reinforced concrete walls. Requirements relating to the behavior of nuclear buildings demand more than simple structural stability and can also apply to the degree of leaktightness of civil structural elements, in particular those forming the outer shell of the buildings. Within this context, IRSN launched a test program on three low-rise reinforced concrete walls, focusing on the study of crack geometry (length, width, and spacing) as a

Senior Engineer, Institut de Radioprotection et de Sret Nuclaire (IRSN) B.P.17 92262 FONTENAY AUX ROSES CEDEX FRANCE IRSN 2 Senior Engineer, Institut de Radioprotection et de Sret Nuclaire (IRSN) B.P.17 92262 FONTENAY AUX ROSES CEDEX FRANCE IRSN 3 Consulting Engineer, 15 Chemin des Beauvilliers 78380 BOUGIVAL FRANCE

function of alternating loads applied to represent the effects of the horizontal components of an earthquake. The mechanical tests performed at the CEBTP (French Building and Public Works Test Center) were coupled with measurements of air flow rates at low pressure through cracked walls performed by IRSN.

The three test specimens were similar to those used in earlier studies (Four and Bouchon, 1989), consisting of a reinforced concrete wall, 0.10 m thick, 1.50 m long, and 0.75 m high (see Figure 1). The wall was framed by 0.20 m x 0.18 m beams and columns. The dimensions of the beams used to secure the wall to the test setup were as follows: 0.29 m x 0.18 m for the top beam, and 0.34 m x 0.18 m for the bottom beam. Column dimensions were chosen to balance the bending loads so that the wall was subjected to predominantly shear forces, and 0/20 mm aggregate concrete was used, with no downscaling, in an attempt to reproduce the same irregular profile in cracks along their length and through the wall thickness. The average mechanical characteristics of the concrete (4 measurements for fcj, 3 for ftj and 1 pour Eij) obtained from tests performed on test specimens, are shown in Table 1.

14 holes: dia. 33 mm, spacing 150 mm

Figure 1: Shapes and dimensions of test structure

Wall No. 1 No. 2 No. 3

Age (days) 69 71 77

Compression fcj (MPa) 31.7 36.4 28.6

Tension ftj (MPa) 3.05 3.25 3.05

Modulus Eij (MPa) 27,000 28,000 26,000

Table 1

As the test structures were not very thick, reinforcement consisted of a single layer of deformed rebars at mid thickness (0.10 m 0.10 m mesh). The same arrangement was adopted for all three test structures, but using varying percentages of rebars, as follows: 6 bars (0.3%) for wall no. 1, 8 bars (0.5%) for wall no. 2, and 10 bars (0.8%) for wall no. 3. Tensile strength tests on the test specimens yielded the results shown in Table 2.

Steel (mm) 6 8 10 12

e1 3 (10 ) 2.5 2.6 2.6 2.4


fe (MPa) 585 615 620 565


e2 3 (10 ) 4.9 5.1 5.1 4.8


fr (MPa) 610 635 655 595

fr fe
1.04 1.03 1.06 1.05

r 3 (10 ) 22 25 20 20
(****) (****)

Table 2 ((*)Ultimate elastic deformation (a=Ea.) at 0.85 fe approx. (**)0.2% offset yield strength (***) Deformation at a=fe : 3 (****) e2=2.10 +fe/Ea Highly scattered values

The columns are reinforced with 4 12 deformed rebars confined by 3 mm dia. hoops 0.10 m apart.

Figure 2

Detail A (horizontal section) (3) 8 horizontal bars spacing 100 mm (4) 7 hoops spacing 100 mm (1) 2 x 2 bars

Detail B (vertical section) (2) 15 vertical bars spacing 100 (3) 8 bars spacing 100 mm


A constant vertical force yields a mean compression stress of 1 MPa in the wall. An alternating horizontal force was applied in the mean plane of the walls, at three increasing values F1, F2 and F3. This force was related to the measurement of the horizontal displacement, which was held constant at each load level during the measurements. These were performed for each load level in positions +Fi and -Fi and for the unloaded wall. The first level F1 was chosen to be slightly above the value at which shear cracks appear in the wall (flexural cracks appear in the posts at an earlier stage). The last plateau F3 was close to the ultimate load (approximately 0.85 Fu). Load F2 was close to the mean value between F1 and F3. Only the third wall was loaded beyond the F3 level to ultimate state. Wall ultimate state occurred for the negative value of load Fu = 853 kN, corresponding to a shear stress of 5.68 MPa. The lowest value (760 kN) obtained in the positive direction was probably underestimated owing to the momentary loss of relation. For each test structure, the horizontal force values considered during the tests are shown in Table 3, along with the corresponding shear stresses.
Specimen Plateau F1 F1 (KN) Test No. 1 No. 2 No. 3 304 304 304 (MPa) 2.0 2.0 2.0 Plateau F2 F2 (KN) Test 400 450 490 (MPa) 2.7 3.0 3.3 Plateau F3 F3 (KN) Test 550 645 687 (MPa 3.7 4.3 4.6 ultimate Fu Fu (KN) Test
(*) (*)




Table 3

For each horizontal force level, the variables measured were as follows: horizontal force, horizontal displacement at the top, vertical displacement at both ends, variations in diagonal lengths, crack width (using an optical crack measuring device) at the intersecting points of a grid plotted on the wall (0.10 m x 0.10 m grid to match the reinforcement grid) and rebar deformation (particularly on wall no. 3, which was loaded to failure). Cracks were plotted as loading progressed and photographed during the test phases. Crack width was measured to the nearest 0.02 mm. Below this value, cracks were considered as hair cracks with a theoretical width of 0.01 mm.


Crack density Crack density was characterized by the total length of the cracks observed on one side of the wall, reduced to the unit area. Three categories of cracks were analyzed: "open" cracks due to the effect of negative forceFi, closed cracks that existed under the force +Fi and all cracks together. The following results were obtained (see Table 4).

Test No. 1

Percentage of rebars 0.3%

Load plateau -F1 -F2 -F3 -F1 -F2 -F3 -F1 -F2 -F3

No. 2


No. 3


Mean shear (Mpa) 2.0 2.7 3.7 2.0 3.0 4.3 2.0 3.3 4.6

Crack density (length in m/area in m) 2.3 4.2 9.6 0.8 5.3 12.4 2.7 9.4 16.8

Table 4

Even when reduced to the same shearing level, crack density values increased with the percentage of rebars, or in other words, they increase as average crack spacing decreased.

Crack width under load For all cracks, width distribution was characterized by the values given in Table 5.
Test No. 1 Level -F1 -F2 -F3 -F1 -F2 -F3 F1 -F2 -F3 Number of measurements 26 44 101 14 52 128 35 109 187 minimum 0.01 0.01 0.01 0.01 0.01 0.01 0.03 0.01 0.01 Widths (mm) average 0.045 0.088 0.157 0.024 0.101 0.126 0.057 0.163 0.110 maximum 0.19 0.35 0.60 0.12 0.34 0.62 0.20 0.35 0.60

No. 2

No. 3

Table 5

Owing to the accumulation of very low values, opening distribution was not Gaussian. Load F3, even saw the appearance of two-peak histograms, no doubt corresponding to the "open" and "closed" crack categories.
Residual width of cracks after loading

A factor of 0.4 was applied to crack widths determined under load to obtain a good approximation of residual widths after rather high loading, but with no notable yielding in rebars (see Table 6).



opening ( Fi = 0) opening ( Fi )
0.62 0.65 0.39 0.58 0.14 0.44 0.61 0.25 0.14 0.42 0.60 0.35 0.32

No. 1

No. 2

No. 3


F1 F2 F3 F1 F2 F3 F1 F2 F3 average F1 F2 F3

Table 6

Predicting crack spacing and width Based on the analytical formulas of the CEB-FIP 1978 model code, which appear more reliable than those of the 1980 model code, average crack spacing can be determined as follows: sm =2 c+ s +k1k2 4bh ; 10 k1 = 0.4 (deformed rebars) k2 = 0.1875 (intermediate values between 0.25 for pure tension and 0.125 for simple bending).

( )

Test Cover c (mm) Spacing s (mm) Effective thickness b (mm) Effective width h (mm) Average spacing sm (mm) calculated Average spacing sm (mm) observed

No. 1 47 100 90 90 243 Few cracks

No. 2 46 100 100 100 231 175

No. 3 45 100 100 100 205 130

Table 7

Agreement with the test was average, given that the cracks were shear and not tensile or flexural cracks. The 1990 model code gives the following formula for calculating spacing for two orthogonal layers of rebars following directions x (horizontal) and y (vertical), and where is the angle of the cracks with respect to y: s'm=( cos/smx +sin/smy)-1 This common value ensures compatibility with the crack width values estimated along the x and y axes. The average rebar elongation is computed as follows:

sm =

s ,max

1 2

sc s ,max

> 0.4

s ,max

; deformed rebars: 1 = 1

No actual load repetition: 2 = 1. Stress in rebar just after the first cracks open: sc The main application difficulty was in assessing stresses s,max and sc due to shear in a lowrise RC wall. These values could be determined from the measurements carried out for test no. 3 using gauges bonded to the rebars. It was deduced from the values that the sc/s,max ratio (first crack stress/maximum stress) was approximately 0.35 on average, and that s,max was approximately 320 Mpa on average for horizontal rebars and 300 Mpa for vertical rebars. 3 According to the CEB-FIP 1978 formula, this gives sm= 1.4x10 for horizontal bars and 3 1.3x10 for vertical bars. Average crack width is calculated by: wm =sm sm For cracks slanting with respect to the orthogonal rebars, this would give: wmx= (s'm/ cos ) smx wmy= (s'm/ sin ) smy wm=1/2 (wmx cos + w my sin) = 1/2 (smx + smy) s'm A more or less equivalent result can be obtained by directly calculating: w'mx =smx smx and w'my =smy smy wm=1/2 (w'mx cos + w'my sin)

The experimental value of 130 mm obtained for the spacing of the main cracks in test no. 3 was taken and converted into spacing along the x and y axes as a function of the average angle = 51.5 smx = lf/cos 210 mm, smy = lf/sin 165 mm whence: w'mx 0.29 mm, w'my 0.21 mm, wm=1/2 (w'mx cos + w'my sin) 0.17 mm This value is higher than the average measured value of 0.11 mm. This could be due to the fact that deformation measurements, which are located around the central part of the wall, overestimate the value of s,max valid on average for the whole wall.

CONCLUSION This test program provided data for estimating crack geometry in low-rise RC walls subjected to alternating shear loads. The main conclusions are as follows: - crack density increases with the percentage of rebars, i.e. when average crack spacing and width decrease, - residual crack width after unloading is on average about 40% of the width under load, - the formulas for predicting crack width, which were initially valid for flexural cracks in beams, are not entirely satisfactory for shear cracks in low-rise RC walls w. Further studies will be required on this subject. In addition, it should be borne in mind that the mechanical tests were coupled with air flow rate measurements for various pressure values.


1. Bouchon M. and Four B. Essai de voiles en bton arm percs d'une ouverture, soumis des cisaillements alterns dans leur plan- AFPS 2nd national symposium, April, Vol. II, CM 29-39; 2. Lafolie F., Bouchon M., Four B. and al.- Behavior of reinforced concrete walls subjected to alternating dynamic loads- 10th WCEE, July 1992, Madrid, Spain ; 3. Four B., Durand J-P, Essais de voiles en bton arm, CEBTP nR112-6-231.