Decrease hardness/dimensional change of quenched/tempered 100Cr6 steel

[kapot]

We use 100Cr6 quenched and tempered steel from two different suppliers. The steel is quenched in oil (60°C) and tempered for 1.5h at 165°C. The working temperature of these components can be up to 170°C continuously. The component needs to have a certain hardness and stable dimensions.
We did a test: we put the components in a furnace at 170°C for 100hours. After that we measured the hardness and dimensions.
For the first supplier we see that the dimensions have increased (appr. 0.03mm), but with hardly any decrease of hardness.
The second supplier shows a decrease in hardness (appr. 4HRC), but with a less increase in dimensions (appr. 0.01mm).
How could the difference between the two suppliers be explained? Is there an inverse relation between changes in dimensions and hardness?
Does the content of the retained austenite may be of influence? How can these changes be eliminated?

 

[redpicker]

A small drop in hardness would be expected with service temperature slightly above the tempering temperature. One possible explanation is that the first sample (which experianced the greatest growth) contained a signifigant amount of retained austenite. During the 100 hour 170C treatment, carbides percipitated from the austenite so that it formed fresh martensite on cooling (which results in a volume expansion). This fresh martensite offset the loss of hardness of the tempered martensite. The other sample contained little or no retained austenite so it saw little or no volume change and the hardness loss was observable.

This can be eliminated by a double temper treatment, perhaps with a low temperature treatement between tempers(which the second supplier may have done as normal processing for this material).

rp