Seismic assessment of vulnerable reinforced concrete structures

Cheng Song, Purdue University


There are approximately 17,000 reinforced concrete (RC) buildings built before 1980 in regions of high seismicity in California (Comartin et al., 2011). These structures were not built to meet requirements that we now know are necessary for survival during strong ground motion. Studying all these buildings in detail to determine which need to be fixed will have a prohibitive cost. Efficient screening methods are needed to identify the most vulnerable structures in this large inventory of older buildings. This study focused on low- to mid-rise RC frames (with 6 or fewer stories) without shear walls. The ability of six seismic vulnerability indicators (SVI) and nonlinear dynamic analysis (NDA) to rank existing RC structures with respect to their seismic vulnerability was studied. The SVIs include: 1) column index (Hassan and Sozen, 1997), 2) ratio of moment capacities of columns and beams, 3) ratio of column shear capacity to plastic shear demand, 4) axial load ratio for 1st-story columns, 5) ratio of building initial fundamental period to number of stories, 6) ratio of building base-shear demand to base-shear strength. This study had two steps to evaluate the seven screening methods listed above (six SVIs and NDA). First, the six SVIs were evaluated using numerical analysis of hypothetical frames. The hypothetical frames were proportioned to have configurations common in pre-1980s reinforced concrete buildings in the west coast (ATC, 2011). The second step was to evaluate the SVIs and NDA against observations from 18 cases surveyed in Erzincan, Turkey, in 1992 and Kathmandu, Nepal, in 2015. The reliability of each method was evaluated using the error between estimated damage and observed damage. The evaluation described shows column index (CI; Hassan and Sozen, 1997) produced the best correlation between observed frequency of damage and estimated vulnerability. At the same time, CI is the index that requires the least amount of both a) information about the structures, and b) computational effort. It was also observed that buildings with lower CI have higher frequency of severe damage. This suggests retrofit work should start in the buildings with lowest CI.




Pujol, Purdue University.

Subject Area

Civil engineering

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server