Heat sink thermal simulation

[jbrownley5D42Y]

I am attempting to perform a thermal simulation on a heat sink assy. that i have created in inventor.  I get acceptable results when running the sim under conditions laid out in this you tube video. https://youtu.be/oVE5CQL6A_U?si=WPWb0m8UyWV0Zc7O. I am really wanting to account for thermal paste and gap pad effects by modifying contact settings but so far the heat sources do not seem to transfer heat to the heat sink. Does anyone have recommendations on contact settings?

Thanks,

[henderh]

Hi @jbrownley5D42Y,

Thank you for providing the youtube video as reference.
If there's a Bonded contact, and the Thermal Conductance default value is left empty, the solver treats that as having no thermal resistance (i.e. infinite conductance). Therefore, you should be seeing the effect of the heat transferring to the heat sink in the results.

I gave it a shot (model attached) and compared three different studies with the Thermal Conductance value as only difference. The symmetry technique was applied to split away 1/2 of the model (and halved the internal heat load) to cut the computation time in half. If we leave the faces created by the cut without any thermal convection or radiation loads, they'll behave as adiabatic (no heat transfer) to satisfy the symmetry rules.

The first study has no thermal conductance defined that reaches a 279 C max temp:

The second uses a high 38000 W / (m^2 K) as the value. The CPU body with the internal heat source will reach a bit hotter at a 285 C max temp:

Finally, if 2500 W / (m^2 K) is used the much lower conductance causes the CPU to be hotter still to 317 C max temp:  

You might find it interesting to see the mesh as well. After splitting the face of the CPU (using the imprint of the heat sink as the tool) I refined the mesh for the important zone, which allow a more accurate simulation at the contact interface. The temperature distribution appears reasonable to me:

Finding the correct value for thermal conductance isn't always easy, since it depends on many factors such as material types, material temperatures, surface roughness, contact pressure, gap distance, gap material, gap material temperature, etc. Almost all seem derived empirically through lab testing. You can find thermal conductance value charts on the web for different material combinations and conditions, and online custom calculators.

Here's an example of values I used (courtesy of the Autodesk Moldflow Insight product documentation):

Hope this helps!

ps: I'm cautious when watching some youtube creators, not all are providing accurate information or best practices.

[TimelesslyTiredYouth]

I don't think there are any legitimate youtubers that create properly explained simulations, not that I've seen anyway...

[jbrownley5D42Y]

Thank you! Appreciate the help!