Solution to the PEER framing equations accounting for dependence of earthquake repair costs on the number of damaged components

Abstract

An analytical framework is presented to account for repair cost dependence on the number of damaged components in assembly-based seismic performance assessments. The analytical framework is investigated specifically in comparison to prevailing methodologies, which are based on Monte Carlo simulation, in order to highlight their relative advantages and disadvantages. An illustrative example considering exterior glazing in a three-storey building is set out to demonstrate an application of the proposed method. Implementation of a fully analytical framework is found to be hindered by the correlation of demand parameters, for which Monte Carlo simulation remains an attractive solution. Adoption of first-order second-moment assumptions within the proposed framework increase computational speed but decrease accuracy in the assessment of building-level decision variables. A full first-order second-moment framework is set out with application to error propagation and parameter studies accounting for dependence on the number of damaged components.

Keywords:

• Seismic performance assessment
• probabilistic methods
• Monte Carlo simulation
• buildings
• damaged components
• repair cost