CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CD-adapco (http://www.cfd-online.com/Forums/cd-adapco/)
-   -   Negative densities at the start of a run (http://www.cfd-online.com/Forums/cd-adapco/53212-negative-densities-start-run.html)

nandu February 21, 2003 10:09

Negative densities at the start of a run
 
Hello,

I am trying to model the supersonic flow between two finite width wedges. The model has approx 250000 elements. I am attempting to run it as an inviscid calculation with an upwind scheme. The solver fails within the first three steps citing 'negative densities found in more than 100 cells as the error'. how can this be rectified, any help in this matter is greatly appreciated.

cheers nandu

cjtune February 21, 2003 12:17

Re: Negative densities at the start of a run
 
Try pseudo-transient runs. Adjust time step to taste. Also, try using double-precision. Worked for me.

azmir April 20, 2003 22:48

Re: Negative densities at the start of a run
 
Same problem... for my transient moving mesh case. I've tried: 1) changing the time step to be one order to order of magnitude smaller but didn't work. 2) using double precision but didn't work 3) changing all relaxation factors (everything not just pressure) but didn't work 4) all above with 2-3 different values as trial and error and still didn't work. 5) looking into mesh quality possibilities now... don't know yet.

Appreciate if someone with experience could give some tips here. Thanks.

Dave May 13, 2003 16:18

Re: Negative densities at the start of a run
 
I've had this problem quite often. From my expirence (I'm a fairly new user) mesh quality is the cause. I've only come across this when using PROAM to mesh.



All times are GMT -4. The time now is 13:05.