Boulder transport by high energy: Numerical model-fitting experimental observations

Abstract

A well-controlled experimental study was conducted to validate a numerical model of boulder transport by tsunamis. The model can simulate boulder transport due to sliding, rolling, and saltation. The boulders were modeled by cubic/rectangular blocks (flatness number varying from one to three) with different densities. The tsunami overland flow was simulated by the dam-break flow. The numerical model was capable of reproducing experimental observations fairly well. The difference between the modeling and experimental results was mainly attributed to secondary motions like swing and rotation due to both the variable friction and fluid forces applied to the blocks. The impact force generated when the flow front hit a block was modeled by a simple formula. Though it can increase the initial velocity of the block, its effect on the maximum transport distance was not significant. The maximum transport distance was sensitive to the lengths of axes and density of the block, density of the fluid, and coefficients of drag and friction. The pre-transport angle of a block affected the maximum transport distance. Placing the long axis of the block perpendicular to the fluid flow produced the maximum transport distance, while the minimum transport distance occurred when the long axis was parallel to the flow. An interaction effect was unavoidable if many boulders are transported together. The experimental results revealed that collisions of blocks of the same weight at the beginning of transport lead to different maximum transport distances. The higher the number of blocks, the greater the collision effect was. Even though the collision effect was very local and marginal, the effect on the maximum transport distance became more significant with an increasing number of blocks.