Compact Thermal Model

[Anonymous]

When using a CTM, the device must contact a PWB and a fluid. I have a heat sink (copper) placed on the Theta Jc surface and a PWB on the Theta Jb surface. I run the analysis with a fixed airflow. The fluid contacts the edge of the CMT but is this correct for the analysis. Analysis runs and results generated but are they valid?

I have run the analysis with a custom material (solid) using  Theta Jc to calculate thermal conductivity . The result differs from the CTM to make me question how to solve? My geometry has conductive paths on both surfaces and intuitively this is not correct.

Any suggestions how to solve?

[apolo_vanderberg]

The use of CTMs are perfectly valid and generally accepted as being able to achieve respectable results.

The fidelity of the results (as with any analysis) can be impacted by the inputs and assumptions used with the setup, along with any mesh sensitivities within the domain.

As far as your second comment. Just to be sure I followed that correctly, you too the Theta(Jc) and computed an equivalent conductivity? Then used this as a Material to assign to the entire CTM volume as a custom Solid material?

With a generic solid material, the assumption will be that from its Junction there is a constant / uniform conductivity to lose heat to the board as well as to the case/heatsink.

The benefit of the CTM is that you can account for any difference in the conduction path between the Jb and Jc.without having to model the chip in detail.

Both paths (Solid or CTM) can provide useful and viable results, however, depending where you are in your design trend and what you are trying to achieve may push towards leveraging one method over the other.

Apolo

[Anonymous]

Apolo, you understood my post correctly.

Placing the heat sink on the surface is acceptable? The requirement for using a CTM is that a fluid is in contact with the surface. If I can use with the heat sink can I use a film coefficient on the fin surfaces or do I need to provide a fluid with appropriate constraints?

[apolo_vanderberg · Accepted solution]

Yes, the CTM just needs to touch the fluid domain so the fact that the lateral sides are exposed is fine. You can have a Heatsink mounted to the Case side of the CTM.

If you are looking to have the flow domain modeled to see the impact of the flow in (and around) the heatsink, yes we would need to have a fluid domain with appropriate constraints. You may want to step through some of our Example models as we go through many of these aspects with them.

Apolo

[Anonymous]

Just trying to understand this correctly.  If I'm using CTM's for my chips in my analysis, my understanding from this thread is that I can place a solid material aluminum heat sink directly on the chip, between the chip and the fluid (air), and this analysis will work properly.  Is that correct?