[gdbb89]

Hey guys!

First time poster here. Have been using FLUENT for a little bit now, so I'm quite new (but determined to learn).

Here's my problem: I'm trying to model supersonic (M = 1.5) flow over a quarterpipe-shaped wedge in 2D. If I take a square geometry and cut out an arc on the bottom right corner, FLUENT is able to solve the problem quite readily. The solution converges in no time, and I'm even able to adapt a mesh refinement where the separated shock is.

Relevant information:

- Density-based
- Energy eqn on, Inviscid solver
- Ideal Gas material
- Pressure farfield set to M = 1.5, gauge pressure 101325 Pa, 300K thermal

- Implicit formulation, Roe-FDS flux type, second-order upwind flow
- Courant number = 5
- Residuals set to 1e-6
- Solution initialized from farfield

This is a simple solution for FLUENT. Here is the refined mesh:

http://i.imgur.com/K7dr9oW.jpg

And a Mach Number contour (from CFD-Post):

http://i.imgur.com/i8tKzUk.jpg

And a plot of my residuals for this case:

http://i.imgur.com/dFySURy.jpg



Now, I'm trying to take this same scenario, except open up my domain in the +Y direction. I want to see the how the shock trails off at the edge. To illustrate, here is my new grid:

http://i.imgur.com/Yz6beIT.jpg

However, when I try to run this, my residuals are completely unstable. Even at a Courant number of 0.5:

http://i.imgur.com/iTNfqse.jpg

I am completely lost as to why FLUENT can't handle this simple doubling of the Y-component domain!

Any and all help is greatly appreciated.

[gdbb89]

I've tried running this simulation and setting the residuals to 2e-2, the lowest they go before oscillating. This is the result of the Mach number:

http://i.imgur.com/0blxkoE.jpg

It appears as though FLUENT is having difficulting resolving this case for when the shock wave separates from the wedge. Any ideas why? I checked and the first case seems to resolve as the shock wave comes off the edge rather nicely.

[mmkkeshavarzi]

Dear all,

I did steady simulation over a body in Inviscid and density based solver,
I monitor the Drag coefficient (Cd) during my Run, it is stabilized after about 2000 iterations and I continue my iterations to 5000 and it is still after 2000.
The Cd value is 0.2 and doesn't reach 0, I assumed that my simulation is converged but know I have doubt about my convergence in inviscid flow.
Is it necessary to reach 0 in inviscid flow for Cd?
and also is it necessary to monitor Cl for inviscid flow?
what is the necessitation to monitor Cl?

Thanks so much in advance for your answers