Hi, I am modelling a fluid to solid heat transfer problem. The fluid is at a very high temperature (2100 K), but when I get the solution the solid wall in which it comes in contact with shows a very small increase in temperature (initial 300 K, final ~350 K). Also the material of the solid is copper. Does anybody have an idea what values I should play with to get some reasonable temperatures (exp. anlaysis gives me a wall temp. of around 650 K) I am using a Spart-allmaras turbulent solver for this 3D problem.

Thanks.

Try different viscous models - im using k-e-RNG for some heat sink analyses. It works OK but I'm not going up to the temperatures you are.

I also had to go for the enhanced wall treatment which meant adjusting my grid to get the correct y+ value.

Also, just a thought but is your fluid density correct at that temp?

Sorry, I can't offer more help, but I got past the problem I faced with alot of trial and error and even more patience, good luck

Thanks, I am also playing around with all the vicous models. Till now the spart allmaras has given me the best convergence and results. also my fluid is ideal gas, so fluent takes care of all the density matters. Also could you tell me, how much of a rise in temperature are you getting and what y+ values are you trying for?

I'm only dealing with the heat sinks you might find in your computer, so only a 40degC temp rise. The y+ values i go for on the enhanced wall treatments are around 1 - easier said than done if you have a large 3D model.

Another problem I've come across are extremely high heat transfer coefficients at the joints between the pin and pin endwall (in the region of a few thousand W/m^2.K). These seem to arise when the cells in the boundary layers of these surfaces have high skew. So another tip I'd suggest is make sure the boundary layer grid has a low equi angle skew as possible.

Good luck with it!

Dear jack smith,

The problem lies in the near wall treatment. U should provide more layers near the wall to capture the boundary layer and give the correct values at the solid interface.

SA model is only for the beginners. it will give you approx. results compared to RNG k-e. So better try with RNG k-e and more refined mesh near the solid-fluid interface.

mail ur email id.

all the best.

MKP

What kind of bc have u applied to the waal? a convective bc? what do u mean by initial temperature? are u studying an unsteady case? If u expect so different temperatures between the flow and the wall u should get more attention in the radiative model that u're using than in the turbulent model. What kind of model are u using and why?

thomas