what is the physical interpretation of the boundary conditions?
are some faces really always 20deg C?
you say "There is no problem when running the simulation with the alu part only" > so what is the difference between the study that works and the one that doesn't work?
are the parts touching? (does interference detection with treat coincidence as interference find the faces that are touching?)
to me your problem is somewhat underdefined, 40W = 2W and the 20deg C could rest on either side of the equation
i would recommend posting the models or sample files if you can't post those
The steel plate is kept at a constant temperature by active cooling, so yes some of the faces are 20deg C.
The difference between the simulations are, that this one is on an assembly of the alu and steel part and not just the alu part alone.
The parts are touching but not interfering (I just checked)
I am trying to model a laser which applies approx. 40 W into the alu part, while a peltier-element is cooling another spot on the alu part down by approx. 2 W. The whole alu part is sitting on a steel plate which is being cooled down to approx 20 deg C.
By under defined, do you mean that I am missing something to balance out the 40 W or that I should define the system clearer? So that the 20 deg C is on "the other side of 40 W" in the equation.
Unfortunately I cannot show the model, even though it would make it a lot easier.
I realise it would not be possible for anyone to help out without seeing the model, but any attempt will be appreciated.
does the problem occur if you just model 2 blocks? if the problem doesn't occur there, it is your models. if it does, post here.