Premature cracking of 8620 component

[UCengno1]

We are experiencing cracking from the top corner of a keyway in a load carrying component that mounts to an rotating output shaft. The problem is that the cracking is occuring at 1/10th of the actual operating load during a controlled assembly prove out process.

The part is fabricated from AISI 8620. It is carburized to 61 Rc to a depth of .015" and a 46 Rc core is resulting. The component does not have an especially thick cross-section, .25", and this is reduced by 1/3 in the region of the keyway.

This is a legacy material choice that has been in use for many years,so it has a decent track record. It appears that something has change in the processing that could create or promote cracking at the keyway corner but I am uncertain where to start. Any help would be greatly appreciated.

 

[TVP]

First thing should be to have a thorough failure analysis performed, including fractography, microstructure evaluation, etc. The core hardness looks a little high to me, but the size is somewhat small. The fracture analysis will be able to show any defects that may have been present, or whether the failure is by overload. Do you know the heat treating parameters, such as carburizing temperature, austentizing temperature, and tempering temperature?

 

[dbooker630]

It would be useful to note the furnace carbon potential, what temperature the part was quenched from, any evidence of carbides at the surface, interruptions in the heat treat recipe. The microstructure check will help determine if the heat treat process is suspect.

 

[UCengno1]

Will report on microstructure later today. From an off-shore vendor so control of HT is always suspect.

 

[UCengno1]

Turns out we were being provided 1117 steel in lieu of the expected 8620. Confusion with another similar part at the supplier.

Metallurgical indicates a 91 15N case, .011 deep, with a 23 Rc core. Is a lower core hardness more typical of 1117 versus 8620. We hear heat treaters telling us that the core follows the case but that is certainly not true of 1117. Is that more true of the 8620 as we try to achieve a 40 Rc core?

 

[dbooker630]

I would follow TVP's advice on the fracture analysis. 1117 is a resulfurized steel so it is inherently 'dirty.' Inclusion content should be evaluated as a possible cause based on the steel mix.

 

[UCengno1]

Is there truth to the statement that 8620, with its higher chromium and nickel content, will inherently have a higher core strength when carburized to the same case and hardness as a identical part fabricated from 1117?

 

[TVP]

UCengno1,

Yes, 8620 has higher hardenability than 1117, and therefore will have a higher core hardness when processed to a given case depth. 1117 will not be able to achieve a core hardness of 40 HRC.