HVAC modeling problem
I followed the tutorial 17 for modeling of an HVAC system in a room, and made some changes to the geometry of the room. Now when I run the solver I get the error # 001100279 has occurred in the subroutine ErrAction - Floating point exception: Overflow.
The solver will make it through the first iteration, but always gets hung up on the second iteration at the point when it is calculating turbulence kinetic energy and eddy dissipation. I have tried bumping my allocated memory up and I have also changed the eddy dissipation length from the .25m value given in tutorial 17 to automatic, and automatic with value. I still get the same error during the same iteration. |
Re: HVAC modeling problem
Are there other ways to dumb down the model? Maybe my mesh is too complex?
|
Re: HVAC modeling problem
Hi,
Floating point exception means the linear solver has diverged big-time. You need to numerically stabilise the simulation. You can do that by a combination of: improving mesh quality, smaller timesteps, double precision, simpler turbulence model, upwinding for advection (just to start things off), time stepping based on local time scale, or a few other things (but they are the main ones). Glenn Horrocks |
Re: HVAC modeling problem
Thanks alot. From your help and others on this board I was able to make this simulation work.
What I did was practically double the mesh size, cut the time step in half, and I chose the SST turbulence model. I think I will try increasing these attributes in small increments until I get the result I want. This board is very informative. I have learned more CFD analysis in the past week, than my entire college career. |
Quote:
|
All times are GMT -4. The time now is 13:09. |