Thermal compound conductivity choice...rule of thumb? Need to do the maths?

[bostonman]

An interesting experiment you can do is to get two glass plates like microscope slides, put a tiny blob of thermal grease between them, and then press them together. It is amazing how far a little dab of grease expands.

I've seen or have done similar experiments, and you're correct, it spreads out quite a bit more than assumed.

Using more grease may go back to what I was trying to explain on how I interpreted what exactly it does. Let's assume two pieces of metal are mounted together (the tab of a component and a heat sink) and a pit located exactly at the same points on both pieces of metal. Let's say they are half spheres on both, and, when pressed together, create a perfect sphere.

Obviously this sphere is an air gap. My thought on grease was that the grease fills the sphere and transfers the heat by the heat going through (I know, poor word to use) it and onto the heat sink. Better quality grease transfers the heat better/faster.

From what I'm understanding is that the grease heats and transfers its heat to the heat sink.

Technically both my thoughts are the same because the heat sink would be warming slower than the grease and the grease slower than the component, but to be a bit ridiculous and completely incorrect: I thought the grease doesn't heat (that's the completely incorrect statement that I'm using for purposes of discussion) and is just a compound that transfers the heat from one object to another.

Therefore, more grease, the more cooling.

Yes, I understand (now) that more grease is more wrong.

[T3sl4co1l]

Yeah, you've got it

A sphere obviously won't carry much heat across it, as having a spherical bubble in the middle of a metal sandwich; but a very flattened spheroid sort of shape might well describe a lot of the bumps and warpage that real mating surfaces have, and it is that gap that grease helps the most at.

Tim