Skip to content

Latest commit

 

History

History
39 lines (26 loc) · 1.91 KB

README.md

File metadata and controls

39 lines (26 loc) · 1.91 KB

1D Euler equation solver written in Python

This code can solve 1D shock tube problem using FVM method. Currently the following Riemann solvers are implemented:

  • Godunov (Exact Riemann solver)
  • Roe
  • AUSM
  • HLL
  • HLLC

Note that the Godunov Riemann solver is rewritten from this repository: ToroExact. Other approximate Riemann solvers mainly refers to Toro's book: Riemann Solvers a n d Numerical Methods for Fluid Dynamics and this MATLAB code: Approximate Riemann Solvers.

And the code supports two reconstruction methods:

  • 0th-order
  • MUSCL-TVD
  • 5th-order WENO

Three time-advancement techniques are implemented:

  • 3th-order Runge-Kutta
  • 4rd-order Runge-Kutta
  • Euler time-advancement

On the other hand, the code also supports WENO 5th-order finite-difference discretization using characteristic decomposition method to compute the flux. To make the code operate in finite-difference mode, just set the argument FD to True when the Euler1D object is instantiated.

solver = Euler1D.Euler1D("./HLLCTest_WENO/SOD-WENO.txt", FD = True)

Below is the result of SOD shock tube problem (t = 0.25) solved by Godunov, Roe, HLL, and HLLC method. The computational domain consists 300 cells and 0-order reconstruction is used.

SOD

Shu-Ohser problem is solved on 500 cells using the MUSCL-TVD method and the 5th-order WENO method. HLLC Riemann solver gives the flux on the interface. It can be seen that the 5-th order WENO method gives a much better resolution of the high frequency oscillation near the discontinuity.

Shu-Osher

The initial condition, the location of discontinuity, and the numerical methods can be specified in the control files. The examples of control files can be found in *Test directories.