主办单位：苏州科技大学土木工程学院

土木工程学院青年教师发展委员会

土木工程学院关工委

报告内容简介：

For large-scale geophysical flows, such as debris flows and submarine landslides, flow velocity and runout distance are critical parameters for risk assessment and hazard management. However, understanding and predicting the dynamic processes of these flows remain challenging, primarily due to scaling issues when attempting to replicate similar behaviours in small-scale experiments or larger-scale field tests. Imperfect scaling arises because geophysical flows are multiscale phenomena, involving interactions between solid particles and interstitial fluid across varying microscopic (particle-level) and macroscopic (flow-level) spatial and temporal scales. This seminar introduces three approaches to bridging different scales in geophysical flows. The first is up/down scaling, where key parameters are appropriately scaled by maintaining essential dimensionless numbers constant. Examples demonstrating this approach include runout scaling in mudflows and immersed granular flows. The second approach involves micro-informed constitutive modelling, in which macroscopic models are developed based on a fundamental understanding of microscopic mechanics. Illustrative examples include modelling grain segregation and basal resistance of granular flows over micro-topography (roughness). Finally, the seminar discusses the potential of using artificial neural networks to address scaling challenges in complex geophysical flow problems.

主讲人简介：

郭颂怡博士，香港大学助理教授，拥有剑桥大学工程学博士学位、麻省理工学院工程学硕士学位，以及谢菲尔德大学一级荣誉工程学学士学位。她曾先后就职于明尼阿波利斯的Itasca咨询集团和洛杉矶的CDM公司。其研究聚焦于滑坡灾害预测与防治、可持续地质材料、以及与油气开采和碳封存相关的能源岩土工程领域。曾获美国岩石力学协会（ARMA）“未来领袖”称号，并荣获2022-23年度香港大学工程学院最佳教学奖。现任International Journal of Rock Mechanics and Mining Sciences期刊编委，同时担任Natural Hazards期刊责任主编。