|
[Sponsors] |
August 25, 2003, 23:41 |
Help!!!
|
#1 |
Guest
Posts: n/a
|
Hi all! If the momentum source added in subdomain is very sensible to the velocity, such as the momentum is equal to "c*velocity", and the coefficient "c" is quite large, what should do to set to make CFX solver to converge?
|
|
August 27, 2003, 15:23 |
Re: Help!!!
|
#2 |
Guest
Posts: n/a
|
Miko,
Add a momentum source coefficient equal to "c", the solver will use this to linearize the equations. Also, if 'c' is positive, the solver will blow up on you. When multiplying by velocity, you should only use negative (resistive) coefficients. Regards, Robin |
|
August 27, 2003, 20:44 |
Re: Help!!!
|
#3 |
Guest
Posts: n/a
|
Dear Robin,
Thank you very much! You are really a good people! I think my problem can't be solved by setting a linear resistibility, for the momemtum sources at different direction are different. For instance, x-component is 0, but y-component and z-component are both c*velocity. So how to deal with it? Best regards! Miko |
|
August 28, 2003, 15:42 |
Re: Help!!!
|
#4 |
Guest
Posts: n/a
|
There is only one entry for the source coefficient, use whichever c is largest.
Robin |
|
August 29, 2003, 03:00 |
Re: Help!!!
|
#5 |
Guest
Posts: n/a
|
Hi Robin,
I think I can try! Thank you! Best regards! Miko |
|
Thread Tools | Search this Thread |
Display Modes | |
|
|