The projects aim is to create an open-source, web-based platform that leverages Physics-Informed Neural Networks to simulate wind flow in urban environments
steps through the process of solving a 2D flow for the Lid Driven Cavity (LDC) example using physics-informed neural networks (PINNs) from NVIDIA’s Modulus Sym.
generate a 2D geometry using Modulus Sym’ geometry module;
set up the boundary conditions;
select the flow equations to be solved;
interpret the different losses and tune the network; and
steps through the process of solving a 2D flow for the Lid Driven Cavity (LDC) example using physics-informed neural networks (PINNs) from NVIDIA’s Modulus Sym.
generate a 2D geometry using Modulus Sym’ geometry module;
set up the boundary conditions;
select the flow equations to be solved;
interpret the different losses and tune the network; and
do basic post-processing.