Fuzzy-random fragility model to estimate non-structural damage to reinforced-concrete frames

Damage has to be quantified to assess the seismic performance of buildings but, in fact, its definition is vague by nature. The limit states associated with non-structural damage especially are verbal and qualitative rather than analytical and quantitative. Therefore, it is proper to consider the fuzziness together with the randomness. A probabilistic model to include fuzziness in the fragility computation is formulated herein, consistent with the time-invariant first-order second-moment reliability method. Such a fuzzy-random model is characterized in terms of moments, distribution, and percentiles, in comparison with the classical reliability model. The fuzziness causes the fragility mean value to increase at lower seismic intensity, but to decrease at higher intensity. The fuzziness also causes the fragility dispersion to decrease in most cases. As application, the model is implemented to compute the fuzzy-random fragility curves of non-structural damage to a masonry-infilled reinforced-concrete frame. All results by the proposed model are reasonable.