Seismic Structural Damage Prediction Considering 3D Topographic and Basin Effects on the Distribution of Seismic Hazards: A Demonstration in Hong Kong

Abstract

This study demonstrates a performance-based regional structural damage prediction methodology through a case study on the Tuen Mun-Yuen Long Basin in Hong Kong, incorporating 3D topographic, basin, and site amplification effects into fragility mapping. Regional-scale spectral element method simulations, accounting for soil nonlinearity, are conducted to quantify the spatial distribution of spectral acceleration amplifications (SaAmp). The 2%-in-50-years uniform hazard spectrum for Hong Kong was disaggregated to identify dominant magnitude-distance scenarios governing the fragility of short-period structures. An input ground motion was selected to match a short-period conditional mean spectrum, ensuring that the seismic hazard controls structural fragility. Two prototypical low-rise reinforced concrete frame structures were analyzed in ETABS to determine their modal properties and base shear threshold values corresponding to three structural component damage states defined in FEMA P-58-1. Structural fragility was subsequently mapped based on the probability of exceeding these damage states under 2%-in-50-years design spectral demand, spatially amplified using the regional SaAmp datasets. The resulting fragility maps reveal pronounced spatial variability in structural damage potential governed by local geological conditions. The prototypical three-story residential structures located above 20–30 m deep basin deposits exhibited collapse probabilities exceeding 50% due to double resonance between the soil and structure. The prototypical single-story warehouse structures remained resilient, with negligible damage probability. The study demonstrates that accounting for the seismic demands that govern structural fragility, together with spatially-distributed 3D site effects, strongly influences the expected level of structural damage. Neglecting these factors may lead to a substantial underestimation of seismic consequences.