Godunov's scheme

Godunov's scheme. It is a conservative numerical method for solving hyperbolic partial differential equations. Developed by the Soviet mathematician Sergei K. Godunov in 1959, it fundamentally relies on solving local Riemann problems at the interfaces between computational cells. This approach provides a robust framework for accurately capturing discontinuities like shocks and contact surfaces in fluid dynamics.

Overview

The scheme was introduced by Sergei K. Godunov while working on problems in Computational fluid dynamics. Its core innovation was the direct use of the Riemann problem solution to compute fluxes, providing a physically conservative update. This method was a significant advancement over existing techniques for the Euler equations (fluid dynamics), as it naturally handled the propagation of nonlinear waves. The original formulation is formally first-order accurate and obeys the Lax–Wendroff theorem, ensuring convergence to the correct Weak solution.

Mathematical formulation

The method discretizes the domain into control volumes, applying the integral form of the conservation laws. For a system like the Euler equations (fluid dynamics), the state at each cell interface is found by solving the Riemann problem with data from adjacent cells. The flux is then computed from this intermediate state, often using an exact or approximate Riemann solver like the HLL Riemann solver. This update procedure ensures properties like conservation of mass, momentum, and energy, aligning with the Rankine–Hugoniot conditions across discontinuities.

The Riemann problem

The Riemann problem is an initial value problem with piecewise constant data separated by a discontinuity. Its solution for the Euler equations (fluid dynamics) involves waves such as shocks, rarefactions, and contact discontinuities. Solvers developed by researchers like Philip Roe (Roe solver) and Bram van Leer provide efficient approximations. The work of Stanislav Osher and Sukumar Chakravarthy on ENO methods also relates to high-resolution treatments of these problems.

Higher-order extensions

The original method is limited by Godunov's theorem, which states that linear monotone schemes for scalar conservation laws are at most first-order accurate. To overcome this, extensions like the MUSCL scheme by Bram van Leer introduce TVD limiters. Other approaches include the Piecewise parabolic method by Paul R. Woodward and Phillip Colella, and the WENO methods pioneered by Chi-Wang Shu and Stanley Osher. These methods reconstruct higher-order polynomial states while controlling spurious oscillations near discontinuities.

Applications and limitations

The scheme is extensively used in Computational fluid dynamics for simulating compressible flows in aerodynamics, astrophysics, and shock tube experiments. It underpins many codes in fields like combustion modeling and plasma physics. Limitations include computational expense from solving many Riemann problems and the first-order accuracy of the base scheme. Despite this, its robustness for problems with strong discontinuities ensures its continued use in major software like the NASA OVERFLOW code and in research at institutions like the Courant Institute of Mathematical Sciences.

Category:Numerical analysis Category:Computational fluid dynamics Category:Finite volume methods