CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   Phoenics (https://www.cfd-online.com/Forums/phoenics/)
-   -   Residuals and .result files (https://www.cfd-online.com/Forums/phoenics/72359-residuals-result-files.html)

HPJE February 3, 2010 06:02

Residuals and .result files
 
Hi,

I'm using Phoenics as part of WindSim. I'm trying to figure out ways of checking convergence after a run has been completed by using the .result file I get from Phoenics.

Using the MIGAL solver, I get a list of residuals (Max, Min and RMS) for each iteration. Would it be appropriate to use these values to check convergence? I'm concerned whether the values provided are whole-field values or stems from a single location? Also, Which variable are the residuals showing; just the overall largest and smallest?

Further, can somebody help me understand the section shown below? I'm having trouble deciphering it.

************************************************** **********
Whole-field residuals before solution
with resref values determined by EARTH
& resfac= 1.000000E-01
variable resref (res sum)/resref (res sum)
P1 1.232E+05 1.383E-02 1.703E+03
U1 1.137E+06 2.078E-02 2.363E+04
V1 6.845E+03 7.748E-01 5.303E+03
W1 2.626E+04 5.407E-01 1.420E+04
KE 3.566E+03 6.060E-01 2.161E+03
EP 3.189E+01 5.820E-01 1.856E+01

************************************************** **********


Thanks for your help guys. Any comments are appreciated. And sorry if I'm asking stupid questions ;-)

Linh.Hoang February 5, 2010 05:07

Hello Hans Peter,
iīm not using MIGAL Solver but hope I can help you anyway. I check convergence after run by reading the hole field residuals in the result file. To be honest I donīt know how Phoenics calculate the resref so I take the res sum (hole field residual) and normalise it with the inflow quantities. You got this information from the result file right after the residuals. there you take the nett source (pos. sum) and nomalise yor res sum for each variable with the corresponding nett source. Your normalised residual should drop in a Order of E-3 to reach Convergence.
Hope this helps
Linh


All times are GMT -4. The time now is 11:36.