Physics-based parameterized fragility models for coastal residential buildings considering the effects of neighboring structures

Coastal infrastructure continues to suffer widespread damage from hurricane events leading to significant economic losses.Among the most vulnerable structures are wood-framed residential buildings, which have been repeatedly exposed to storm surge and wave actions in past events.Acknowledging this vulnerability, researchers have developed fragility models using physics-based simulations to assess the damage incurred by these buildings in coastal communities.However, the influence of neighboring structures in residential building portfolios has been overlooked in current physics-based fragility derivation procedures, despite past hurricane-induced storm surge events highlighting their potential impact on demand modification across the portfolio due to wave deformation phenomena.This oversight limits the models' accurate representation of portfolio performance in risk assessments.To address this gap, this study develops closed-form parameterized fragility functions for residential buildings that account for the effects of neighboring structures.Using computational fluid dynamics (CFD) simulations, this study captures the modification in demand exerted by nearby buildings by incorporating multiple structures into the numerical analysis domain.To overcome the high computational cost of full-scale CFD simulations, an active learning algorithm is leveraged to guide the sample selection process, focusing only on critical parameter combinations of hazard and structural properties.Findings reveal that the probability of building failure varies based on proximity to the coastline, with structures closer to the coast being more directly affected and those farther inland experiencing greater influence from neighboring structures.These fragility models enhance the accuracy of portfolio-level risk assessments while enabling rapid vulnerability evaluations through their closed-form functional relationships.The improved regional vulnerability estimates are expected to better inform risk mitigation decisions for structures as well as inform the design of portfolio layouts that minimize the demands imposed by neighboring buildings.