Dr. Cheng has worked extensively on the multi-scale and constitutive modeling of heterogeneous soils (e.g., GRS), Bayesian parameter estimation, and 3D micro-CT image analysis. His research aims to understand the physical processes where the micro-mechanics is important and relevant length/time scales are not well defined. To this end, he develops multi-scale modeling and Bayesian inference tools to bridge geomechanical models at micro and macro scales. His most recent work involves developing a computational framework for concurrent multi-model simulations of granular matter, in all physical states. 

Dr. Cheng has developed and coupled several open-source software packages, including YADE, MercuryDPM, OomphLib, and LB3D, for solving geotechnics and geophysics problems, such as soil-geosynthetics interaction (FEM-DEM) and wave propagation in granular media (LBM-DEM). He has developed a Bayesian calibration software "GrainLearning" which has many users in the industry and academia. Dr. Cheng is the main developer of the coupled MercuryDPM-OomphLib code, with a vision to implement a unified multi-scale framework for modeling solid/fluid-like behavior of granular materials, incorporating state-of-the-art​ open-source codes from soil mechanics, fluid mechanics, and beyond.

In the Thermal and Fluid Engineering Department, Dr. Cheng coordinated and gave lectures in three 5EC MSc courses. He taught the dynamics of particle-fluid systems in Multiphase Flows, soil elastoplasticity in Granular Matter, and image analysis in Advanced Programming in Engineering.

In the Civil Engineering Department, he teaches an MSc course GeoRisk Management (5EC). Students learn how to use probability theory and stochastic modeling to integrate numerical models (e.g., FEM) and observation data within a probabilistic framework for risk assessment. This multidisciplinary course covers several topics, including geomechanics, flow in porous media, finite element method, and stochastic modeling. Students practice the knowledge learned in the lectures via the final project "Dyke Risk'', by implementing a small Matlab soler for probability-based dike stability assessment.

A full list of courses I teach and contribute can be found here.