As a final project for my first year programming course, my team built a basic wind tunnel simulator in Python using a finite volume approach with a Newton-Krylov solver regime for the steady-state, compressible Euler equations.
The integrated mesher is far from perfect; it can only handle simple, smooth objects, in part due to its architecture and in part due to limited access to compute power. Future improvements we would like to make include bug fixes to the drag force and coefficient computation, a better mesher, and parallel processing to allow a much finer spatial discritization. Although our program has remaining challenges, I learned a lot from my first attempt at finite element analysis.