heat transfer simulation - contour

Dear FEAT team,

While performing 3D heat transfer in a simple shape-design, the contour temperature stays (or even goes below) ambient. Equations and boundary conditions are set as shown, with a heat source of 10W.




rho,Cp,k,Q,T0=1,1,5,10,25 - other variables 0


h,Tinf=20,25 - other variables 0

How is the temperature in the contour can be set to change accordingly?

Thanks for your time and reply.

If I understand your question correctly you wonder why the temperature does not reach a uniform 25 degrees?

You have set up a heat transfer model with a constant heat source and heat flux boundary conditions (proportional to the difference between surface and exterior temperatures). In this case double or triple times heat loss will occur for the edges and corners, respectively, since they share two or three surface. You therefore will have lower temperature at the edges and corners. This is to be expected and consistent with how an object in the real world behaves (consider a block heated to a very high temperature so it is glowing, when it cools the edges first will show dark indicating regions of more heat loss). If you switch of the heat source and let the simulation run to steady state then the whole body will eventually assume ambient temperature.

Yes, the model is the one you describe, with a constant heat source (10W) and heat flux boundary conditions (h=20 and Tinf=T0=25degC) at all surfaces. You are right that temperature should be lower in the edges, but i was expecting 2 or 3degC, not that much (also the temperature at edges now, based on temperature contour, is same with ambient which i think is not correct).

The heat source should stay on until the simulation finishes, is it possible to adjust the solution for stationary application, with this heat source model?

 I m afraid that i neglect something and the result is not correct.

Thanks for you time and help.

Dear FEATool team

Any ideas/suggestions on this temperature behavior on the contour?



Kind regards

kth wrote

Yes, the model is the one you describe, with a constant heat source (10W) and heat flux boundary conditions (h=20 and Tinf=T0=25degC) at all surfaces. You are right that temperature should be lower in the edges, but i was expecting 2 or 3degC, not that much (also the temperature at edges now, based on temperature contour, is same with ambient which i think is not correct).

No, the temperature at the edges and corners should approach 25 degrees for your problem setup. If you solve the problem with the FEniCS solver (or for example Ansys) instead of the built-in you will get the exact same solution. You will of course get a better solution with a finer grid/mesh.

kth wrote

The heat source should stay on until the simulation finishes, is it possible to adjust the solution for stationary application, with this heat source model?

A stationary problem can of course not use a time dependent/variable heat source (it must be constant in time).