I have a system which a couple PCB's, for which I have to evaluation possible concepts for forced air cooling. To implement the heat loss, we have done measurements in a free convection setup to determine the critical components and their heat generation in watts and gathered thermal imaging data for each PCB to validate the simulation.
As I do not want to use the detailed model of the PCB's, I created a block model of the components for my system simulation. Of course, this leads to some discrepancy, when I compare the heat distribution from measurements using a thermal imaging camera and my simulation.
Now my idea was to "reverse engineer" the heat loss using my thermal images instead of using the measured heat generation. My approach was to use the measured temperature of the critical components on the PCB as boundary conditions for my block models of the PCB in a free convection simulation. The resulting heat generation should then be used as input for my system simulation. Unfortunately, I only get a surface heat flux and no heat generation for a complete block. Is this approach even feasible and what should I use as BC for my heat sources in my system simulation?