Civilian version of a military heat sink system

Discussion in 'General Electronics Chat' started by hp1729, Jan 25, 2016.

1. hp1729 Thread Starter Well-Known Member

Nov 23, 2015
2,087
232
No, I'm not building this at this time. Just a question in the theory of the design.

Consider, if you will, a heat sink system using liquid silicon in an enclosed tank, a Peltier junction with a heat sink cooling the liquid and a heat sink for power transistors with fins down into the tank. A small pump circulates the liquid. How does one calculate the cooling capacity per square inch of heat sink?

What are the intelligent questions that need to be asked?

The military system is an electronic warfare pod that straps to the bottom of a plane. It replaced a previous Freon (or some such gas) cooling system that kept running out of gas at the wrong time.

2. SLK001 Well-Known Member

Nov 29, 2011
849
242
Well, silicon has a liquid point of over 2600°F, so not much cooling capability there. Perhaps you mean "silicone", or more specifically, "silicone oil"? To be able to calculate the cooling capacity, you would have to know the rate of flow of the liquid, the heat capacity of the liquid, the transfer efficiency of the liquid to the heatsink and the transfer efficiency of the heatsink to the environment (plus probably other things that I forgot).

If I had to do this, I would go crack open my old college thermodynamics textbook and start reviewing.

3. hp1729 Thread Starter Well-Known Member

Nov 23, 2015
2,087
232

Is this important? For liquid silicon ...
Liquid Heat Cap., cal/g mol ºC 6.755 (at m.p.)
Liquid Density 2.533 (at m.p.)

Heat capacity. Does this imply that liquid silicon has about 6 times more cooling capacity than air?

Liquid silicon is non-corrosive and non-conductive? So can I put the pump inside the tank?

Dang, I might build this after all. How would I test it? Do I measure heat rise (is heat drop correct?). Do I evaluate it by seeing how much power I dissipate before the temperature gets back to normal room temperature?

My Peltier junction runs on 12 V and draws 5 Amps. I power it by an old computer power supply when it is not in the car.

4. wayneh Expert

Sep 9, 2010
12,374
3,226
It should be easier than resorting to that [responding to #2]. The maker of the heat sink should be able to tell you the specifications, the heat transfer coefficient. Likewise the supplier of the silicone oil should be able to tell you how it performs at various temperatures (viscosities) and such.

One of the most intelligent questions is how much heat needs to be moved, in how much time. Don't forget that the TEC will consume as much as 10 units of heat for every 1 unit it moves from the hot side to cold side, and so the hot side needs to shed 11 units.

Last edited: Jan 25, 2016
5. hp1729 Thread Starter Well-Known Member

Nov 23, 2015
2,087
232
Re: "Well, silicon has a liquid point of over 2600°F"
Yes, thank you for the correction. Silicone oil. Solid silicon, yes. Liquid silicone is used as a lubricant and anti-foaming agent. Also used to help give floatation to electronics in silicon tubes in military SURTASS (Submersible Towed Array Surveillance System, or something like that) and the oil industry.

Re: "college thermodynamics text books"
That would be high school for me and that was a long time ago. Like almost 40 years ago?

I did find heat capacity, but I don't know what to do with the info. Empirically, I can always judge it by the result. How cool does it get with no heat load applied and how much heat load do I have to put into it to get temperature back to room temperature.

Last edited: Jan 25, 2016
6. hp1729 Thread Starter Well-Known Member

Nov 23, 2015
2,087
232
Thanks. I am still trying to dig up info on silicone oils. Viscosities? A higher viscosity cools better? Do I compare it to water?

7. wayneh Expert

Sep 9, 2010
12,374
3,226
High school level thermodynamics isn't going to cut it! Let us know what you're doing and folks here can step you through it.

Feb 24, 2006
10,340
1,850
9. hp1729 Thread Starter Well-Known Member

Nov 23, 2015
2,087
232
Re: "the hot side needs to shed 11 units"
Sounds familiar. Back when I was on the road a lot I build a small ice chest for the car. No, efficiency was not the objective.

10. hp1729 Thread Starter Well-Known Member

Nov 23, 2015
2,087
232
Thanks for the offer. I suspect I will need it.

11. wayneh Expert

Sep 9, 2010
12,374
3,226
A lower viscosity is better for convective heat transfer, because the thin layer where the temperature gradient goes from high (the heat sink metal surface) to low (the bulk temp of the circulating fluid), is thinner.

Nov 23, 2015
2,087
232