Modified Cloud Analysis: An efficient tool for seismic fragility assessment

Abstract

Modified Cloud Analysis (MCA) is proposed as an alternative to the original Cloud Analysis for handling the collapse-inducing records. Collapse and non-collapse part of the data can be treated as mutually exclusive and collectively exhaustive events. In this way, the original cloud analysis can be applied to the non-collapse portion of the records. The estimation of the collapse probability can be improved through using a generalized regression model (logistic regression). MCA, which relies on ﾓcarefulﾔ record selection, employs un-scaled records and can be carried out with analysis effort (in the order of 15-30 records) comparable with code-compatible requirements (7 records). Focusing on fragility assessment for a 3D structural model of an existing five-story RC building located in LﾒAquila (central Italy), the present work provides a multi-faceted discussion of the Modified Cloud Analysis and its various applications in modern probabilistic performance-based earthquake engineering. The structure in question has collapsed due to a soft-story mechanism in the LﾒAquila 2009 earthquake. It is shown that (1) information-theory concepts lend themselves perfectly to measuring the sufficiency of the adopted IMﾒs in the MCA procedure as an interface variable in the seismic risk assessment; (2) with some required adjustments, the MCA can be effectively used for performance-based code-compatible safety checking based on time-history analysis; (3) MCA can be used in a one-to-one time-history/structural model realization coupling to efficiently propagate modelling uncertainties using various simulation techniques.