pinn_CDnozzle is the PINN code to solve for compressible flow in a Convergent-Divergent nozzle.It includes four main examples.

Divergent nozzle.py is the example for solving the steady state flow in divergent nozzle.In this example, only the boundary conditions at the inlet of the expanding nozzle are provided, not at the outlet, and PINN will randomly give a supersonic or subsonic solution.

Inverse problem.py is the inverse problem example for solving the flow in a CD nozzle. In this case take $50$ solution points of pressure uniformly as known and demand PINNs to deliver profiles of $\rho$, $Ma$ and $T$, and as well as the specific heat ratio $\gamma$ in the equation of state. log4_50.xlsx is the data used in the example.

Steady state flow with shock.py is the example for solving steady state flow in a CD nozzle. The original code would get a solution for a mixed state of subsonic and supersonic flow with the position of the excitation wave at x = 1.875. By providing different boundary conditions at the outlet, solutions can be obtained for different cases, including subsonic, supersonic, or mixed states with shock.

Unsteady flow.py is a typical example for solving unsteady flow in a CD nozzle. The size of the neural network used is 4*50 and has an additional distribution of training points. The original code would get a unsteady solution for the shock wave at x = 1.875, within $t\in \left [ 0,15 \right ] $.