I am trying to find experienced Flotran users because I am having some trouble getting decent results. I am dealing with supersonic external flow over a projectile (bullets and other ammunition). I have primarily dealt with subsonic flow in the past and do not have much experience with supersonic CFD problems. Any help is appreciated.

(1). You need to find out whether this code you are using is capable of handling shock waves accurately or not. (2). The supersonic flow itself is easier to compute because the region of influence is confined inside the local Mach cone. While the subsonic flow will affect the whole flow field. (3). So, if the method used is developed for the supersonic, compressible flow with accurate shock capturing, then normally, it is easier to get the solution. So, it is with the code, not with the problem you are trying to solve.

You might want to consider a code such as Floworks by NIKA. It uses a CAD program called SolidWorks as its pre and post environment. You can model a problem and get a solution started faster in this code than any other I have used. They have some examples of supersonic projectile modeling on there web site. The pretty pictures they have agree very closely with the schlerein (pardon my spelling) they have them compared to. I do not have much experience in doing SS problems but I agree with the other poster that they are easier to handle and converge due to the sharper gradients and the more localized areas of interest. the url is www.floworks.com

Hi i did so work as an under grad student, looking at supersonic flow over a flat plate using Ansys and Flotran. We were running on a pc and found several problems getting the code to produce any form of shock waves,let alone then compare the results to validation data. I used an example that came with the code, for Mach 2 flow over a 2D NACA 0012 aerofoil. This was back in 1995, and haven't used that code since, so I'm not sure that the example still exists. The code continually access the pre processor to modify the running parameters such as relaxation. You can modify the code and make these changes manually, reducing loading and saving times of the run. We also found that the pressure calculations did not work for our set up, and once these were modified then I could start getting some reasonable results out. Sorry for this being a little vague, but it was some time ago. If you are still interested, then I could have a look back at my thesis and identify the modifications we made. email me at funster7@hotmail.com if you're interested. good luck Stuart