As high as possible but .97 or better for emissivity
The load is very small, dealing with thin(.001") metal substrate sliding over another. Bearing load is small but has to survive over 1mil cycles of sliding back and forth about an inch each time.
What about debris, etc. from the abrasion of the substrate? Is that cleaned off somehow?
I don't see a simple solution here.
> anodize is probably the best for a simple surface, but you'd only get about 0.88 emissivity and it's unclear that anodize would survive 1M cycles
> etched pyramidal pits prior to anodization would increase the net emissivity, but I'm not certain that the exposed part would be able to survive 1M cycles.