Discussion Forum

Hi everybody,

I'm working with Comsol or any such programme for the first time for my thesis. So I hope I'm able to follow the answer if there will be any... =)

My model is a simple 2D one in stationary heat transfer mode consisting of several rectangles that stand for a stack of plates - electrical heated plates (lying in the centre of the stack) as well as insulation layers. The information for the boundary conditions I take from temperature measurements between the layers.
The aim is to know where the heat flows travel inside the stack and to quantify them via boundary integration.

My problem is that neither the sum of the boundary integration values of one rectangle nor the balance of the whole stack is zero. But isn't it supposed to?
I tried it with a one rectangle model and found out that the problem doesn't seem to be natural convection at the side of the stack nor the temperature gradients on the boundaries but lies in using temperature data on interior boundaries. Is that possible? How can I avoid it to get to my heat flows?

By the way, in one post I read that Comsol uses a special way to indicate the direction of the value of boundary integration. I noticed that, too, and tried to find an answer in the manuals, without success. Have I missed it? Or can somebody please solve the mystery?

In any case thank you beforehand for answering this!