Evaluating tsunami damage of wood residential buildings in a coastal community considering waterborne debris from buildings

Abstract

Tsunamis are one of the most severe natural hazards that could result in human life loss and extreme damage to structures. During tsunami events, many buildings could experience significant structural damage, and some of them could be washed away. Even with many studies working on data-driven approaches using various statistical methods, the tsunami-induced physical damage to the coastal communities could not be carefully captured, especially with coastal communities with various residential building types and geographic terrains. Numerical analyses that treat buildings as single nodes ignore the effects of neighboring buildings. Nevertheless, the integrity of buildings in communities could be affected by other failed buildings since failed buildings could generate debris and impose additional loadings on neighboring buildings due to waterborne debris. In the present study, a coastal community, including multiple types of wood residential buildings, is selected to evaluate the performance of coastal communities subjected to tsunami-induced damage. Finite element models of typical buildings are established, and the inundation depth map for the community is used to consider tsunami loads. A physics-based approach considering building-generated debris is proposed for damage assessments of the coastal community. Parametric analyses are conducted to investigate the behaviors of buildings under different tsunami loads considering inundation depths, flow velocities, and debris mass. The effects of the waterborne debris generated from the upstream collapsed buildings are considered to update the damage maps of the community to show failed buildings at different flow directions. The failure rate map of the entire community with the consideration of building-generated debris impacts is further obtained. Fragility curves of single buildings are also obtained with the inundation depth treated as a random variable.