The problem is it must last for 3 years and be intact. It will be attached to a gas turbine rotor, under the heat shields. We tried temperature indicating paints (like tempil) and they just oxidize away in that time.
I contacted Tempil directly. Here is their response:
Hello,
Thank you for the inquiry. We do not offer temperature labels with ratings above 554F. The reason is the paper used to produce the labels will not withstand the higher temperatures.
I was kind of thinking that there may be a calibrated metal strip that has an oxide thickness that correlates to time@temperature. Anyone know of such an animal?
Is the application show temperature especially if it exceeds 1000F then oxide layer thickness isn't going to show you that? A thermocouple would do that but it would require some power.
I have seen closed metal foil packages that had low melting point elements inside them.
There are alloys that melt very near 1000F making this practical.
McMaster-Carr is showing a "Tight-Joint Silver" brazing rod with a melting temp of 1125-1145 degrees F.
Maybe you could fix a cup to the turbine and put the brazing rod in it.
Best regards, David
Ed's and gibson's suggestions are good, but only give an answer to the question "did the temperature ever exceed x", and not the added question " and for how long". Along the same idea as theirs, imagine a cup/vial with a very small vent hole (perhaps piped to a location deemed safe), and the cup is filled with a precise volume of, for example, tin. Based on the known vapor pressure of tin vs. temperature, and the size of the vent hole, one could (by measuring the mass left in the cup after some temperature-time exposure), estimate the temperature vs. time profile. Would likely work better with an array of these devices, with different temperature vs. mass loss rate characteristics. Patent pending