Thermal contact conductivity - first tryout

DaveStupple

I like this feature - it offers a way to avoid modelling fiddly little thin layers. I think that TCC = CTC/thickness (where TCC = thermal contact conductivity, CTC= coefficient of thermal conduction), but I have a vague recollection of suggesting a similar idea that you could avoid modelling super-thin layers by modelling them thicker and changing the CTC proportionately. I think Victor may have tried a demo that showed this not to be the case. I need to check.

My first attempts to use this feature were in a solid lump , just selecting identical slave/master faces from opposite directions. This did not work. So I made two separate components with faces in the same plane and it worked fine. I haven't tried physically separating the faces to see if the feature still works. The image shows three similar trials with an identical wattage going in on the central area of the LHS face, and an identical convection at the RHS face. The TCC decreases by a factor of 10 from left to right. It's obvious where the contact is

Just to add, I LOVE working with the 3D Connexion mouse in Mecway. This software just gets better and better to use -great work Victor, and thank you.

.

Victor

Thanks for sharing.

To be honest, I'm never quite clear on the definitions of all those TCCs and CTCs and all that. I named the parameter using what I hoped was the more obvious term "thermal conductance" (=1/thermal resistance). But it turns out that's fairly uncommon and even ambiguous. So it's sadly a giant mess. Sounds the same as what you said though - thermal conductivity / thickness.

It uses contact in CCX, so the same behavior for finding which faces to connect applies, which allows some gap between the two meshes.

DaveStupple

I've got a few papers on this where a thermal resistance at an interface is caused by a vibrational mismatch between the two materials rather than by poor contact or poor thermal conductivity of the materials. They vary in their naming: interface thermal resistance; thermal barrier resistance; thermal boundary resistance; thermal boundary conductance. The last of these I have seen with both watts per metre sq per K, but also as the reciprocal units that you alluded to. When I saw the feature in Mecway it turned out to be exactly what I I thought and hoped it would, so your choice is pretty logical to me.

Victor

Oh, that's interesting about the cause of thermal resistance. Thanks for confirming the name is OK. It looks like "thermal conductance" would have been alright instead of "thermal conductance per unit area", but hopefully the extra words add more clarity than wrongness.

DaveStupple

Victor, yep, it is interesting. Diamond has a huge thermal conductivity, but - as an insulator - it is entirely through phonon (vibration) propagation, as opposed to metals, where free electrons are the main energy messengers. If the receiving medium doesn't like vibrating at the frequencies it is getting from the diamond, you have a thermal barrier. This is a very basic explanation, because that's the territory where I lurk. This type of barrier is intrinsic to the interface.

If you lump in crystal defects, contaminants etc. that sometimes accumulate at interfaces, you can describe a bigger thermal barrier that includes the non-intrinsic stuff. I reckon you could follow the modelling and maths (I can't), but whether you would want to go to the trouble is another matter. Shout if you want copies of the papers. Julian Anaya is one of the people who has done much of the recent work that I have seen. He's cleverer than me by some way, but he's also very helpful and (mercifully) tolerant of dumb questions, so look him up if you're interested, or I can ask for contact details.

DaveStupple

This, by the way is why this feature is so useful. Yeah, it might be a shortcut to modelling thin layers, but it's also useful for these thermal barriers that have, effectively, close to zero thickness.

Victor

Thanks. I don't want to study it too much but that adds a little clarity to something that I never quite understood - how materials with different mechanisms for storing thermal energy can manage to completely transfer it between each other and reach thermal equilibrium at equal temperatures. If I understand you correctly, the energy does fully transform between say phonons and free electrons but with a resistance instead of being locked out completely.

DaveStupple

Oooh, (speaking from logic, rather than comprehensive knowledge now...) there must be some transference from phonons to electrons, otherwise - as I think you are saying - the barrier would be absolute. Whether the phonon transfers first to phonons in the receiving material (a metal), and then to electron mobility in the metal, or... phonon to electrons directly, I do not know (nor care).