Flow-structure interaction has raised significant attention among geotechnical community because it is important for prevention of high-speed flowing geo-disasters. Physical modeling serves as a fundamental tool with which to reveal the disaster-causing mechanism, but such modeling is challenging to perform especially in 1 g model tests because of the complex rheology of flows, transient nature of impact processes and strong nonlinear response of structures. The author thus presents a critical review of the state-of-the-art practices of modeling flow-structure interaction using the centrifuge and discusses several key issues. Hierarchical scaling is useful to guarantee flow similarity in microscale and macroscale, while the scaling of the Brazil nut effect and particle–fluid interaction still questionable in current stage. Flow similarity and structure similarity are equally crucial in the modeling of the flow–structure interaction, which indicates that a coupled modeling method should be encouraged. The Coriolis effect complicates flow regimes and exaggerates or lessens the impact effect of simulated flowing disasters, and it is thus important to investigate the mechanism and magnitude of such an influence. To address the above issues and facilitate in establishing basic scaling laws, robust numerical tools are recommended for experiment design and exploring fundamental mechanisms.

流固耦合在岩土工程界引起了极大的关注，因为它对于防止高速流动的地质灾害十分重要。物理建模是揭示灾难机理的基本工具，但由于流变复杂，冲击过程具有瞬态特性以及结构强烈的非线性响应，因此要在1 g模型测试中进行建模尤其具有挑战性。因此，作者对使用离心机对流-结构相互作用建模的最新实践进行了严格的评论，并讨论了几个关键问题。分级缩放有助于确保微观和宏观尺度上的流相似性，而在当前阶段，巴西坚果效应的缩放和粒子与流体的相互作用仍然值得怀疑。在流固耦合建模中，流相似性和结构相似性同样重要，这表明应鼓励采用耦合建模方法。科里奥利效应使流态变得复杂，并夸大或减小了模拟流动灾害的影响，因此，研究这种影响的机理和程度非常重要。为了解决上述问题并促进建立基本的缩放定律，建议使用健壮的数值工具进行实验设计和探索基本机制。因此，研究这种影响的机理和程度很重要。为了解决上述问题并促进建立基本的缩放定律，建议使用健壮的数值工具进行实验设计和探索基本机制。因此，研究这种影响的机理和程度很重要。为了解决上述问题并促进建立基本的缩放定律，建议使用健壮的数值工具进行实验设计和探索基本机制。