That sounds like it would work. I'd probably end up with a tinned
thermocouple before too long - I'm not sure if, or how, that might
affect the accuracy of subsequent readings.

That is why I recommended an isolating pad between the metal and thermocouple.
Something with good thermal conductivity.

Ideally, you would measure the temperature with a non-contact optical
thermometer.

I'm not very familiar with them. Don't infra-red thermometers have
issues with shiny surfaces? It seems like 300 degree metal with a
rough black surface would emit more IR than the same 300 degree metal
with a polished, silver surface.

The ones I've used from Omega, Ircon and Micron are pretty good, especially at
the low temperatures we are talking about. They are really useful when we got
up to 2200C. Thermocouples really don't last long at that temperature. We were
holding better than 1 degree for long periods of time.

Anyway, ideally, you would have a cavity in the sample and aim the non-contact
thermocouple inside that hole to simulate a black-body target. I never saw any
great difference in temperature readings because of differences in the color or
polish of the metal.

One of the differences in cost of these devices is the optics which allow you
to read to a narrow angle. The tighter angle means a smaller area is read.
Sometimes you want to average over a large area and other times you need spot
readings. Check out the Omega website for details on various options.

Usual disclaimers.

Earle Rich
Mont Vernon, NH