[xue sheng]

I am solving 1D shock tube SOD problem by using Harten modified flux TVD scheme.

For Harten's modified flux TVD scheme, I am using Right Eigen Vector Matrix (and its inverse) of Hoffmann book chapter 12 page no. 104, equation 12-29 (see attached file).
For this matrix my code is stable for CFL<1.


I also tried to use Right Eigen Vector Matrix (and its inverse) of "Lecture notes for 24th Computational Fluid Dynamics" (March 15-19, 1993; H. Deconinck) Section 4.1 page no. 55, equation 4.6.
For this matrix my code is stable till CFL <0.25 after that it becomes unstable.


1. Does selection of Right Eigen Vector matrix and it's inverse affect the scheme stability?

2. Have anyone tried both Right Eigen Vector Matrix?



Waiting for your prompt reply.
Thanks and Regards