Experimental Quantification of Tsunami Debris-Damming Loads and Failure Probability for an Elevated Structure Using a Kriging-Assisted Sequential Design of Experiments

This study employed a nondeterministic kriging (NDK)-assisted sequential design of experiments (SDoE) strategy to quantify the tsunami-driven debris-damming loads on elevated coastal structures, including associated uncertainties. The experiments were performed in the large wave flume at the O.H. Hinsdale Wave Research Laboratory of Oregon State University, at a 1:20 scale to mimic tsunami-driven debris damming at an isolated column structure. The test combinations consisted of 12 wave conditions, 2 structure arrangements, and 6 debris shapes at two densities. During the SDoE process, trial combinations for experiments were selected based on a criterion that prioritized exploration of combinations with a high likelihood of large debris-damming loads on the middle column. These configurations represented global maxima or local maxima with high uncertainty. After the convergence of the SDoE process, the logarithmic values of maximum damming loads estimated by the final NDK model differed by only 4.2% and 5.1% from the actual experimental loads for the two structure arrangements. Furthermore, the trained NDK model was used to calculate the structural failure probabilities of a coastal structure in Crescent City, California, under tsunami-induced hydrodynamic and debris damming forces. The probabilistic failure assessment, performed using a Monte Carlo simulation, showed that the anticipated failure probabilities for tsunami hydrodynamic loads under debris-damming effects do not meet the general target probabilities in Chapter 1 of ASCE 7, especially under debris conditions with 12.2-m shipping containers. The results of this study showed the applicability of the SDoE method to large-scale physical experiments on tsunami-driven debris damming. Furthermore, the results offer a robust foundation for improving the design of elevated coastal structures for tsunami debris damming to meet the required level of resilience while accounting for significant effects of larger debris, such as 12.2-m shipping containers.