Corner pitting after pressure test

[Lyrl]

I work at a commercial heat treater that mainly does carburizing. We're working with a potential customer on carburizing some ~3" ID / 5" OD parts made of X4CrNiMo 16-5-1.

They tried nitriding these parts (at another company - the place I work doesn't have any nitriding equipment) to a depth of .004": 68 HRC at .002" and .004", dropping to 36 HRC at .008" and to the core hardness of 33 HRC at .012". The nitrided parts were pressure tested: pipes filled with water and pressurized to 25,000 pounds (per square inch?). The result was pitting at the corners of the bearing races (picture attached). They think the cause of the pitting is the case depth being too shallow (their reasoning for sending samples to us for carburizing). Does that sound plausible? Or could it be some other factor (the extreme hardness of the nitrided case causing brittleness, maybe?)

 

[metengr]

Hard to tell from the photograph, but this does look like some type of spalling at each of the corner locations on the ID surface. Did you perform the hydrostatic test? If this condition happened after the test, I would perform surface NDT, and see if you have any surface cracks at the corner regions.

 

[tbuelna]

Definitely looks like spalling. Were the races given any NDI prior to the hyrdostatic test?

If the parts were not properly cleaned prior to nitriding then there may have been surface fractures from the contamination. Also, I assume you performed some operation to remove the white layer after nitriding?

 

[Tmoose]

Gears that have case hardening ( like nitriding generally is) that is too shallow for the stress levels involved can fail along case/core interface, kind of like the candy shell coming off an M&M..
See "case crushing" on page 6 here -

http://www.xtek.com/pdf/wp-gear-failures.pdf

 

[tbuelna]

Tmoose-

The type of spalling failure described on p.6 of that document is due to a sub-surface initiated shear fracture from rolling/sliding hertzian contact. That type of sub-surface shear fracture can occur within the case, at the case/core transition or within the core, depending on the contact conditions. If I understand the OP correctly, the spalling shown on the bearing race inner edges was not the result of hertzian contact, but was caused by exposure to high fluid pressure (25ksi?) during a single cycle hydrostatic test.

I would speculate the spalls were due to fractures created during nitriding that propagated when exposed to the high fluid pressures of the hydrostatic test. If the structure of the nitrided surfaces exposed to the 25ksi hydrostatic test pressure were sound and free from defects, then they should not have spalled.

 

[Lyrl]

Thank you for the replies everyone, it gives me some helpful perspective going into our customer doing this same test on samples we have carburized for them.

 

[mrfailure]

Lyrl,

I may be misreading your post, but it sounds like the races were never in service and seeing mechanical wear - the red-brown hematite observed in the picture was found after the corners was found after water pressure testing. If this is true, how could you have galling damage - there would be no wear contact. I probably am missing something - clarification would be great




.

 

[Dougt115]

I don't see the spalling. I see cracks. Was it pitting? Did material come off?

Do you know which nitriding method was used? Gas, Salt bath or plasma. The location of the pitting (also looks like there are cracks between each pit) is in a transition point from one diameter to the next. This would make me look in the direction of thermal expansion and cooling as the root cause.

The pitting appears to shrink at each level and finally there are none. Each level has a smaller diameter and therefore a larger thermal mass than the previous level. Metal rings expand and contract according to the temperature so stress would build at the transition from one ring to the next. The top level would heat the fastest and expand the fastest causing stresses to build at the transition, pitting area. Cooling would do the same. At each level the mass increases so that the heating becomes more uniform and the stresses are more uniform until by the fourth level they do not result in pitting and cracks.

You might check on two things. Is the part heat treated to set the core hardness first and then nitrided? I expect so. The heat treating could cause the cracks to start to form, have them inspect for cracks at this point. Second, the method of nitriding will determine the amount of time the part is heated for during nitriding. Salt being the longest, then gas and finally plasma being the fastest. If thermal expansion is the problem then keeping the time at temperature short would be best to prevent thermal expansion, so plasma.

 

[Lyrl]

The poster who asked about service is correct, these parts never saw any service. The parts have been sitting in the customer's shop or our shop for a few months and I'm not sure if the rust appear at the time of the water pressure test or if it rubbed off from the steel tubs the parts are in. The pits are missing material and developed during the water pressure test.

The perspective in the picture is difficult, and I apologize for that. The top row of pits is at the bottom of a flat (well, all the same diameter) inner diameter surface. Just below that top row of pits is a curved race surface with a larger diameter. Then another flat inner diameter surface - the same diameter as the top flat surface. Again, at the bottom of this flat surface is a row of pits. Underneath this row of pits is another (identical) curved race surface. And again for the third row - another flat inner diameter (still the same diameter) with a row of pitting and a third race surface below that. Followed by a fourth flat inner diameter surface, below which the diameter shrinks. There are three step changes to shrink the diameter, none of which have any pitting.

I am not sure what kind of nitriding was done. Since our company doesn't have any nitriding equipment we are not pursuing any attempts to improve the nitriding, rather focusing on our area of carburizing.

 

[mrfailure]

Given the lack of wear, I would start thinking about corrosion pits forming in the corners forming preferentially because of the crevice geometry. One possible scenario is moisture from the pressure test becoming entrapped afterwards, allowing wetting-drying cycles.

 

[AndrewFinAustralia]

Hi Lyrl,

I read original post as items not being in service and the spalling post pressure test immediately.

Whilst I haven't read the link on case crushing above, in the past, I had issues in the past with a maximum stress concentration during pressure application being at the depth where case hardening finished, resulting in spalling due to the nature of massive stress and the interface. Some detail here ->

http://image.slidesharecdn.com/roll...5/rolling-mill-rolls-39-638.jpg?cb=1415579159

Case hardening produces a surface volume expansion - counterintuitively, it looks as though the case depth may be too thick and the material can not support the stresses produced at the interface as the expansion induced by the volume hardening occurs, with forces directed radially inwards - the uniform spacing between the pits tends to suggest that this might be the case. A thinner case depth would stop this from happening but would not really help with component life. The pressure test might be producing a localised compression/shear at the interface which exceeds maximum stress.

As metengr said, can your potential client do NDT post case hardening and pre case hardening? There might also be an issue as if the substrate is (cracked/spalled/galled/heavily work hardened) before heat treatment, you'd be on a hiding to nothing to try and fix it as the surface expansion occurs

Might end up being a substrate problem or a process change their end pre-hardening.

Former plant process engineer/materials/metallurgist - been out of the game for a while now.

Just another data point - hope it helps


edit - the failure mechanism I refer to is discussed in detail here:

http://www.slideshare.net/nakulrtm/rolling-mill-rolls

<- no longer available for download from esw company site that I could see