Hi,

I implemented Steger Warming scheme (first order in space) for hypersonic Euler flow over cylinder. Results are little inaccurate on the surface of the cylinder.

It seems the problem is because of dissipative nature of this scheme.

Will making this scheme second order in space help reducing the dissipation problem ? Or is it better to change the scheme itself ?

Thanks in Advance, Sunny

If the solution appears "instable" you have to change the scheme. You can use a class of FVS (flux Vector spitting) scheme which can be easily implemented such as Van Leer, Hanel or AUSM; comparing different schemes you can find "your way"...

All these schemes can be coupled with a MUSCL approach to obtain second order schemes.

Making your scheme second order should definitely reduce the numerical dissipation in the solution. However you should be careful since less dissipation means solution can be unstable. To counter that problem I would go with mar's suggestion.

Also making the method second order can make your solution grid sensitive. You may need a good quality grid to have stable solution.