This National Science Foundation (NSF) Engineering grant award to Cornell University, with a total funding of $412,236, will deploy artificial intelligence (AI) strategies to model reactive flow and rock weathering processes across multiple spatial and temporal scales. The key objectives are to: (1) construct a database of virtual reactive flow experiments; (2) train a deep convolutional neural network to identify microstructural features linked to field variable variations; (3) enhance homogenization theory with inclusion-specific characteristic times; (4) train a deep CNN to adapt the homogenization scheme based on microstructure changes; and (5) solve coupled thermo-hydro-chemo-mechanical boundary-value problems of geomechanics using the adaptive homogenization method. This work aims to improve the safety and sustainability of underground geological storage facilities and enhance understanding of chemical weathering processes. The project will also provide undergraduate research opportunities, a diversity/equity/inclusion seminar series, and develop new modeling tools for computational geomechanics. A subaward to Georgia TECH Research Corp will focus on developing a homogenization scheme informed by AI to simulate the mechanical behavior of cracked materials.
Generated 12/3/24, 8:32 AM