I am trying to incorporate a TEC in a portable system. The heat from the interior of the system needs to be transfered to the exterior where it can be dissipated freely into the environment. Now, I have developed a thermal management design where heat is conducted outside via sheets of copper and will be ultimately deposited into a fabricated "water heatsink" where temperature differential is mitigated by water's high specific heat capacity and additional surface area. The water heatsink is consisted of coppersheets forming a thin container filled with water. The system needs to be compact and lightweight and this system seems to provide heat mitigation capabilities and an increased surface area.

water heatsink Attached is the visual illustration of the design. Is this viable?

  • \$\begingroup\$ how will the heat be removed from the water? \$\endgroup\$
    – jsotola
    Jun 15, 2019 at 4:12
  • \$\begingroup\$ I guess I should probably put a fan underneath. Anyway, what is the best method in constructing a very thin copper container filled with water? And how to achieve watersealing tightness? \$\endgroup\$ Jun 15, 2019 at 5:42
  • \$\begingroup\$ When using the heat capacity of water to absorb heat a useful factoid is: Water "absorbs" 850 litre.degrees_C per kWh. \$\endgroup\$
    – Russell McMahon
    Jun 15, 2019 at 9:13

1 Answer 1


Water has some great properties for cooling, but it's messy, and awkward to control. So if you're going to incorporate it, make it work.

In your present diagram, its high heat capacity means it can absorb a pulse of heat better than a metal heatsink.

However, water is a poor conductor of heat, so it's actually insulating the bottom of the heatsink from the top, it's worse than just having a metal heatsink there, if the heat is going to be generated for any length of time. Heat is flowing top to bottom, so natural gravity driven convection is not going to occur

What it does do well is convection cooling, as it has high heat capacity and low viscosity. If you're going to the bother of using water, then pump it, to flow from a small hot emitter under the TEC hot side, to a large heatsink elsewhere.

Heatsink options?

A large, finned, radiator, cooled by either natural air convection, or a fan.

A large mass of water, large enough that it doesn't heat up sensibly in your operation time (a bucket, a stream).

An evaporative (bong) cooler (fan assisted evaporation of the cooling water).

If you can't flow water to one of these heatsinks, then I wouldn't bother trying to incorporate it into what looks like a basic conduction-path metal heatsink.

  • \$\begingroup\$ Thank you so much! Just wondering if I go the water-route, do you have any idea how to properly achieve the copper sandwich in my prototype? I have tried it before using epoxy on two sheets of copper but water is leaking nonetheless. \$\endgroup\$ Jun 15, 2019 at 6:22
  • \$\begingroup\$ Water "absorbs" 850 litre.degrees_C per kWh. \$\endgroup\$
    – Russell McMahon
    Jun 15, 2019 at 9:12
  • 1
    \$\begingroup\$ @LyleKennethGeraldez I think you might have misunderstood Neil_UK's answer. Actually a high thermal capacity does not help to improve thermal dissipation in your setup. Instead it you need a high thermal conductivity, which actually gets worse when using your proposed water-copper-sandwich. I suggest forgetting that idea and simply using bare metal or a heatpipe instead. \$\endgroup\$
    – Sim Son
    Jun 15, 2019 at 11:27
  • 2
    \$\begingroup\$ Just to reinforce what everyone is saying: Liquid is only useful for cooling if it is moving. A pool of water will only delay the inevitable, you still need to dissipate the heat. \$\endgroup\$
    – Mattman944
    Jun 15, 2019 at 13:12

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.