CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   ERROR #004100018 (http://www.cfd-online.com/Forums/cfx/21191-error-004100018-a.html)

falling_stone April 23, 2005 07:21

ERROR #004100018
 
Hi, I'm very new in CFX simulations. I'm trying to simulate electronic chip heatsink inside the enclosure. I make a setup similar to the heat exchanger example from CFX help. It generates error #004100018 if the air ideal gas used inside the enclosure domain and converges with acceptable results if air just replaced by water. What may be a reason for that error? Thanks, F.S.

Ted April 23, 2005 08:05

Re: ERROR #004100018
 
maybe the turbulance model is wrong so that the iteration overflowed.I met the same problem with another example too. try to use another turbulance model. -_-

Glenn Horrocks April 25, 2005 18:35

Re: ERROR #004100018
 
Hi,

What is the text description of the error? Where in the simulation did it occur?

Also note changing the fluid from air to water will result in a completely different Rayleigh number, so of course the behaviour of the simulation will be different.

If you have managed to get it to converge using water but not air, the most likely error is using too large a timestep for the initial timestep (if transient) or iteration (if steady state). Use a smaller timestep for the first few iterations.

Glenn Horrocks

falling_stone April 26, 2005 02:34

Re: ERROR #004100018
 
Thank you Glenn, The simulation is steady state. The error happens in the first iteration, description follows: "ERROR #004100018 has occurred in subroutine FINMES. Message: Fatal overflow in linear solver." The timestep was 2s. Best Regards, F.S.

Glenn Horrocks April 26, 2005 18:45

Re: ERROR #004100018
 
Hi,

In that case your initial timestep is probably too big. Try a much smaller timestep to start off, and once it has got going increase the timestep size.

Glenn Horrocks

falling_stone April 27, 2005 02:49

Re: ERROR #004100018
 
Glenn, Decreasing timescale doesn't seem to work. (If I understood you correct) I tried to change timescale control in solver control to both auto timescale and physical timescale varying from 2s to 1e-20s, the result was the same error. What else but timescale may cause it? Thank you in advance, F.S.

Glenn Horrocks April 27, 2005 18:57

Re: ERROR #004100018
 
Hi,

In that case your simulation is most likely not physically possible. Check that your simulation physics is correctly set up.

It may also be caused by a poor mesh but usually you can get poor meshes to start, they just don't converge very well.

Glenn Horrocks


All times are GMT -4. The time now is 21:45.