Five full-scale reinforced concrete wall specimens with openings typical of the first story backside exterior walls of row houses in Taiwan were tested using lateral cyclic loading to study their seismic behavior. The effects of location and size of the window opening were studied. Test results showed that the wall with the window opening on the side of the wall web showed a higher average lateral strength than that with the opening placed around the center of the wall web. The increase of the opening length reduced more the lateral strength than the increase of the opening height. A critical wall segment tended to show a lower lateral strength when the edge of the segment is bordered by a door than by a boundary column. To estimate the lateral strength of the wall specimens, a detailed and three simplified models are proposed. Comparison with the test results shows that ignoring variation in the axial force among the wall segments and columns does not change the average prediction but increases the scatter of prediction. Simply summing the lateral strengths of all the wall segments and columns together further increases the scatter and greatly reduces the degree of conservatism. Summing the lateral strengths of all the web segments requires the least computational effort but greatly increases the degree of conservatism.

Introduction

Low-rise reinforced concrete (RC) row houses (Fig. 1) are a common building type in Taiwan. During the Chi-Chi Earthquake in 1999, many row houses collapsed due to a weak first story along the street direction (Fig. 2). Today, row houses are built stronger than before with higher material strengths, better reinforcement detailing and larger member dimensions. However, the weak-story problem still exists. This is because walls with openings are usually used for row houses along the street direction (Fig. 1). However, the strengths of such walls are typically ignored in design because the behavior of such walls is not wellunderstood. Moreover, ignoring the strengths of such walls seems to be on the safe side in design. However, this makes the weak-story check unable to identify the weak story caused by abrupt change of such walls between adjacent stories. Previous experiment studies on squat walls with openings have shown that when the opening area is small, the behavior of the wall may not be significantly affected [1,2]. When the opening area is large, the initial stiffness and strength of the wall can be significantly reduced [1,2]. However, the ductility may be increased [2]. The reduction in capacity of squat walls depends on the size, shape and location of the opening [3,4]. For example, walls with openings close to the edge or boundary of a wall tend to have lower stiffness and strength than those with openings farther away from the edge or boundary [5]. Slender walls with openings have also been studied [6,7].