New SimScale user here - so far, I’m really impressed by the usability of the UI. I’ve just started my first project to evaluate the platform for use in future product design at my company.
Project basics: The design has a single heat source (a switching power supply IC), a PCB, and an aluminum heatsink. I’m modeling the heatsink as a chunk of Aluminum (solid) as a “heat spreader.” Eventually, I’ll replace it with a more realistic model including fins. The switcher acts as a heat source of 3.75W at the specific operating point I’m concerned about. I want to estimate the steady-state temperature field of the switcher, PCB, and heatsink.
I’ve imported a simple STEP model of the product and I believe I have the meshes correctly defined. I also followed the heatsink tutorial design step-by-step to get to the point of a converged simulation. However, when looking at the results, it seems like there is no heat conduction between the three regions of the simulation - the heatsink, PCB, or switcher.
I attempted to try and add Interfaces between the components specifically, but this caused Errors when executing the simulation.
Would anyone be willing to take a quick look at my project and see if there is anything obviously wrong?
The project link is here: SimScale
The appropriate CAD model is “Export B” and “Mesh 3.” I’ve executed the simulation successfully a few times, but the results don’t seem to be meaningful.
Thanks in advance and I appreciate your expertise!
Hi @koolscooby, I haven’t done many CHT simulations but I believe you need to define the interfaces between your domains. If you go to the interface section in your simulation the tool tips do a good job at explaining the options there. I think you need to define for each interface between solid-solid, solid-fluid and fluid-fluid.
Thanks for your review! The simulation indeed throws a warning and says that since not all surfaces have designated interfaces, that default interfaces of no-slip (for fluid), thermally coupled interfaces will be created by default.
I did try to create interfaces between the solid components at a minimum - but had errors in the simulation that were not obvious. (It seemed like it was segfaulting .)
The electronics heatsink example had no interfaces designed either - so it seemed like a safe bet to start. I’m hoping someone from SimScale that is experienced with CHT could guide on either the error or the methodology furhter.
sorry for my delayed answer and thanks for sharing your case and explaining the issue you are facing in detail. The warning that you get that tells you that not all faces and/or interfaces have been defined manually is in most cases fine. We are looking into an issue where the CHT solver converges very slowly. I will take a look at your project and check if this is the problem in your case as well.
By convergence I do not mean that the solution field residuals did not go down a certain level but rather the converged solution. You can easily test this by adding a surface plot item for the temperature on a test surface e.g. for a heat sink case on one of the fins. As long as the temperature is clearly rising and the rate is not decreasing significantly the solution field is not yet converged.
I will follow up on the topic!