Black anodizing of parts to stand up at 125 °C

A. I believe most dyes will start to break down around 225-250 °F. It's not the anodizing that is failing, but the dye.

Not too sure if changing alloys will make a difference or not, you could always have your anodizer run a sample coupon for testing.

Your anodizer should be well aware of the limitations of dyed coatings, and should contact his supplier to see if there is a product that will meet your needs.

A. Is this hard anodize, or is it black dye? Either way, 125 °C should not be a problem. Have your anodizer seal the product in boiling water and you should not have a problem.

A. The thermal coefficient of expansion of the oxide is 0.2 that of the underlying aluminum metal, therefore the coating starts to crack above 80 °C. 2024 is worse than 6061. I can send you a xerox of that page the Sheasby/Pinner Book. If the dye is fading, that is probably a sealing inadequacy.

A. Could be a couple of issues here. First, 2024 and high copper 2xxx series of aluminum do anodize poorer than other alloys like 6061 as you've suggested. 2024 comes out of the black with a smut which can be wiped off but detracts from otherwise a good black anodize finish. Secondly, the "coming off" you're alluding too is likely due to poor or no sealing of the anodize following the dye step. If it's fading at 125 °C, that's pretty low temp as many connector manufacturers require upwards of 550F with no visible degradation. Also, your anodizer could be using a cheap dye with poor heat fastness - ask your anodizer what the black dye's heat fastness rating is? Lastly, the dye could just be spent - low on effective dye concentration, high in aluminum, sulphate, phosphate, etc. Last comment, you're getting a pretty bad anodize job and you might be best off finding a more reliable source!