[raydiator]

Hi

I am very new to OpenFoam and have been trying on and off for a few weeks to get a simulation to run. I will outline what I am trying to do, what I have done and what I where I have gotten.

1. What I am trying to achive: Optimise the heatsink and copper block size (in x and y) for a PCB. Ideally I would like to know if I reduced the copper block size from 50x50mm to something smaller what the temperature rise would be.
The heatsource is 10W and is 1x1x0.1mm. This sits on top of a 2mm 50x50mm copper block. Below this block is an aluminium heatsink 60x60mm suitable to take a 60mm fan blowing roughly 20L/sec of air in positive z direction up to the heatsink. I have the base of the copper block at z=0 and an air boundary of 100x100x100mm (air is then +/- 50mm in each direction around 0,0,0). For initial conditions, I expect everything to be at 300K, the heatsource is 10W and the air is 300K blowing at 20L/sec (I realise presently the 20L/sec is over 100x100mm inlet rather than more realistic 60x60mm fan but that is a bridge too far for me presently)

2. What I have done? I started with OpenFoam V12 and looked at the HeatedDuct example. I created stl files for the heatsource and copper block. I have ignored the heatsink for now trying to simplify the solution to get things to run. I changed the stl for the heatsource to be 3x3x3mm as I couldn't seem to get meshing (snappyhexmesh) to converge (lack of a better word) with a smaller stl file. I did increase refinement surfaces to 10 and still didn't get it to work as I expected.
My next mistake was since I couldn't get it to run due to numerous errors, I enlisted the assistance of AI. This eventually got the mesh as I expected (note I have never done this before) and managed to get foamMultiRun to run to completion. The results however look nothing like I expected. Paraview is also new to me so that might also be part of the problem.

3. Where am I now? I have initial conditions that I don't trust, a dataset that is equally questionable and wondering whether someone can advise me whether I am on the right track and give some advice? I have shared my project below. The controlDict has time running to 25sec. It took two hours to run this morning on my computer which seems excessive compared to the heatedDuct example so in my efforts to eliminate errors to run foamMultiRun I have introduced excess computational tasks.

https://drive.google.com/file/d/1N5w...w?usp=drivesdk

[raydiator]

On further playing with paraview, it is showing the air velocity correctly (at least matching the initial conditions). The heating however seems to not be running. At T=1 there is a temperature gradient but by T=2 it is pretty much all gone to 300K (very very small gradient). Looking at the heat source, it isn't actually running. I increased it to 100W and still there is no transfer. Any suggestions?

[raydiator]

I think I sorted it out :-)

[rpach]

hi Ryan,

I am trying to run a similar case. do you happen to know any methods to speed up the simulation? I am a newbie to OpenFaom.