CFD Online Discussion Forums

CFD Online Discussion Forums (
-   FLUENT (
-   -   extreme temperatures (

Andrew Garrard October 17, 2006 05:22

extreme temperatures
I am trying to solve an atypical tubomachines problem. The reason that it is atypical is that the stator baffles come extremely close to the rotor blades (i.e. < 1mm). The rotor diameter is 100 mm and so there is a very large change in length scale. The problem requires 3 dimensions and the solution to the energy equation. I am using a hex grid, which is tricky given the change in length scales, and I am about at the limit of what my machine can cope with (8 Million elements). When I set all wall B.Cs (there are no inlets and outlets in the initial case) to either no heat flux or 300 K, an initial fluid temperature of 300 K and no viscous heating, I get temperatures hitting the limits of 1 and 5000K in some cells. This situation is unphysical. I would like to be able to run the simulation as a time dependent, compressible problem, but at the moment the iterations are too long to consider running transient (and don't converge anyway) and the extreme temperatures destroy the ideal gas model.

If anyone has any suggestions I would be interested to hear them. Trying different ideas to solve the problem takes a long time due to grid size.

sam October 17, 2006 09:37

Re: extreme temperatures
it seems u have ill posed problem(Boundary condition unphysical)

tuw October 17, 2006 09:51

Re: extreme temperatures

I think this is because the wall boundary you set is incorrect. For wall boundary with no heat flux means the walls are adiabatic (no heat transfer occur in or out the boundary). or if you set 300K to the wall (constant temperature), means the temperature of the wall cant be changed (it is fixed!).

The heat generated couldnt dissipated through the wall, keep on accumulating, temperature increase continuously, residuel cant converged.

Try change the wall boundary to convective or else. Hope this works.

Andrew Garrard October 17, 2006 10:36

Re: extreme temperatures
Sam: please could you elucidate. If you mean that I am trying to solve an inherently transient system as a steady one, you are correct, but I am having the same errors using a transient sliding mesh calc.

Tuw: I have some boundaries that are fixed temperature, and some that are no-flux (adiabatic). The constant temperature wall will allow heat to move into or out of the system, Dependant on the local temperature of the adjscent cell. Is this incorrect? If all the walls were no flux, then there would be 1) no unique solution to the energy equation and 2) the ability for thermal energy to accumulate and diverge.

All times are GMT -4. The time now is 22:05.