Discussion Forum

Hello,

for a given arrangement with a device and a desired temperature at its surface I'm trying to compute the power I need for the heating process.
The overall model includes the device (with the desired temperature as its boundary condition) and the direct surrounding, as well as some boundaries with a heat transfer coefficient regarding external convection. At first glance the solution looks fine, giving seemingly realistic results, yet I struggle getting a definite solution for the used power.

Since the Model is 2D-Axissymmetric my attempt was to use the Expression "ht.tfluxMag" along its boundaries, integrate along the lines and set the check for "Compute surface integral".
Now to the results that let me stumble:
First I see a difference when using up/down operators. This is completely OK, when my boundary is a heat source, since it will be able to heat both sides (if both are solved for), in my case the device itself is then heated up after a very short time and consumes no more energy after that.
But doing the same on boundaries that are just that - boundaries between to adjacent material layers - I also get differences between up/down-results, which should not be the case in my sense.
Plotting the Expression "ht.tfluxMag" along the arc length shows that the graphs are rather of bad resolution (showing very visible steps), can the mesh be a problem?
Also the plots for up and down diverge towards the end of the structure, can it be that heat flux is propagated along the boundary itself?

The other issue is, that there are seemingly a lot of possible choices for evaluating the heat flux through a boundary. I also tried "ht.ntflux", which gave me different results. Then there is the multitude of "internal boundary fluxes up/downside"...
Which option should I use when I want to evaluate the heating power that is emitted by an object through its boundaries?

Thanks and regards
Maximilian