Does suface area affect thermal coupling design?

Question

I am building my own thermal couplings for a experiment I am doing at home. I want to build a very small thermal generator with them. At each contact point does the amount of surface area involved in making contact affect the amount of electricity being generated?

Answer 1

Thermopiles are not well suited to generating large amounts of current, typically only a few tens of milliamps. They also do not generate much voltage per thermocouple hence you have to put them in series to boost the voltage. even with this you need to generate the highest possible thermal difference from couple to couple to get anything out of them. Still it would be possible to generate enough to power low wattage loads. You could buy a thermoelectric cooler module and clamp it between a cold plate and a hot plate and it would generate some power at its leads. Those devices would be optimum for many applications. Don't expect to get much out of a string of thermocouple wires all twisted together though. The short answer is to improve the thermal coupling to the junction itself.

Answer 2

At each contact point, the area will affect the resistance. This will not affect the open-circuit voltage available, but it will affect the maximum current output.

So for high current you need large contact area and large diameter wires to minimise resistance and maximise the current and therefore power available. Do you need to flatten the wires? Well, as long as the contact resistance is suitably far below the wire resistance, there is little to be gained from reducing it further.

250 ma sounds feasible : but at what voltage and what temperature differential? Gas boilers in the UK use the current from a single thermocouple at dull red heat to hold a solenoid gas valve open. This cuts off the gas if the pilot flame dies out. I don't know the actual operating current but it must be quite significant to power the solenoid.

The problem with thermoelectric generators is that increasing the wire diameter also minimises the thermal resistance between hot and cold sides - as you increase the useful power output you also increase the wasted power (heat directly conducted from hot to cold) - the efficiency remains constant at about 1% to 2%, dictated by the thermoelectric potential, and the ratio of thermal to electrical resistance.

(Unless you have access to exotic metals like tellurium - I forget exactly, but suitably expensive materials can give 5% or even 8%)