CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   Modeling Hypersonic Flow (http://www.cfd-online.com/Forums/fluent/74691-modeling-hypersonic-flow.html)

pawan1989 April 6, 2010 14:26

Modeling Hypersonic Flow
 
Hey guys. Been browsing these forums for more than a year and a half now but decided to post.
I have used fluent for 2+ years and I have been able to figure things out along the way but I am a little stuck here.
I am having trouble achieving a converged solution for a 3d model I have been working on.

The inviscid case was ok, but the residual level of convergence was just not good enough (most residuals converged to 1e-2 instead of my criteria: 1e-6)... on this note, how do we choose what's our criteria to consider a solution successfully converged? How do I determine the accuracy of my flow parameters which I am measuring?

The laminar case fails one way or another. sometimes, there are Divergences detected in the AMG solver, most times it complains about my absolute pressure and temperature limits are set too low.. I don't know what's going on. Sometimes it says there are reversed flows at the outlet..

Is there anyone who's willing to read a little more into this and offer me some advice? I can post or (preferably) send some of the pictures and the steps for some of my runs.

I am running at M=6. I have pretty much a sharp edged nose cone (like from a missile or such) and a conical domain sorrounding it - the outlet is flush with the end of the nose-cone. Far field conditions at the inlet and conical boundaries and a pressure-outlet at the outlet.

Thank you.

DoHander April 6, 2010 14:38

Try with far - filed conditions on all boundaries (except of course the walls).

Paul

pawan1989 April 6, 2010 14:43

Hey, thanks for your interest!

I tried pressure-far field on all boundaries at first but I felt like the flow was "forced" near the outlet because the flow right around the body, at the outlet isn't exactly far field. Pressure-outlet does an extrapolation of flow data if the flow is supersonic - which mine clearly is.

Here's a figure of the pressure-contours of one of my (million) failed attempts at a converged solutions

http://www.freewebs.com/pkh89/pressure_contours.png

Just to give you an idea of the geometry - the real data on there is.. well I have been trying to tell from the figure if there's anything wrong but I really can't tell.

pawan1989 April 8, 2010 15:54

Thanks for the help guys. I appreciate it. I shall keep anyone interested updated on my progress.

pawan1989 April 27, 2010 17:23

So, I have exhausted all my knowledge.

I have tried obtaining a converged solution for the laminar case of a 3d mesh at M=6 and it fails to converge. I have tried pressure-based coupled methods, density based explicit and implicit methods. I don't know what else to do. Any suggestions?

Chris D April 29, 2010 09:30

Quote:

Originally Posted by pawan1989 (Post 256552)
So, I have exhausted all my knowledge.

I have tried obtaining a converged solution for the laminar case of a 3d mesh at M=6 and it fails to converge. I have tried pressure-based coupled methods, density based explicit and implicit methods. I don't know what else to do. Any suggestions?

You're solving without a turbulence model, but is the flow turbulent in reality? A turbulent flow is unsteady, so without using a turbulence model, you might not get a steady solution. I really don't think there is anything to gain by solving it as a laminar problem.

Also, you have your outlet boundary at the end of your body. If you do that, then you should make sure the flow is supersonic everywhere at the boundary. If the flow is subsonic at any point, you'll be setting a pressure, which you probably don't want to do there. I would move the outlet boundary back a few body lengths, just to be sure.


All times are GMT -4. The time now is 14:57.