Fretting Fatigue 15-5 2

[suddenDEBT]

Hey all,
Been lurking for a few years and can usually find what I'm looking for by searching. But I'm stumped on this one and would really appreciate some input from the experts.

I've got a valve body (15-5 Rc= 36-38) and valve stem (17-4 Rc= 36-38)used for pumping fairly clean high pressure process water (40k). During the valve travel, the stem seems to be fretting the pressurized body bore. The main bore of the body is ballised post machining. We have recently experienced several failures with the fatigue failure with the crack origins in the fretted main bore. Cavitation is not likely as the system would have indications elsewhere. I've attached a picture of the bore in the area of interest. The material is Chinese (unfortunately) and the MTR's seem ok, and of course the 3rd party micrographs don't tell much of a story either.

We have made quite a few of these units over the years and have few failures. Suddenly we are having a rash.
I've looked into the ferrite stringers and banding, seems that what I've read 15-5 is less prone to this, but I could be wrong. The application and use of this valve has not been a problem with moderate fretting. I seem to be stuck on the fact that the fretting takes off the Cr2 03 and since its in water the oxide layer can't reform. There is some discoloring in the fretted area (right side of image), and I am not sure if it is material transfer from the stem to the bore, or corrosion taking place due to the absence of the protective layer.

So my question(s) are:
Is there a significant mechanical effect of high pressure water on this newly unprotected surface? Or could this type of failure linked more closely to the residual stress present due to the ballising and acting like a SCC problem?

Thanks for reading.

 

[israelkk]

Did you considered galling between the 15-5PH and 17-4PH?

 

[metengr]

Based on a review of your post and seeing many valve stem guide failures, I would concur with israelkk this is probably galling damage. If lack of clearance is not a cause, I would look into considering a Stelllite or other hardfacing inlay for this bore against the valve stem material.

 

[mrfailure]

You may want to consider a mechanism of fretting fatigue. Galling usually only induces wear damage, not fatigue. The picture is a little unclear to me, but I have the impression the crack is longitudinally oriented initiating on the inside of the valve body. If this is correct, then loading is occurring in the circumferential direction (think hoop stress). If this is fretting fatigue, then the crack direction indicates tiny vibrational movement in the circumferential direction, as opposed to longitudinal movement associated with the valve stem. You should look closely at the contact surfaces to determine if fretting is occurring rather than galling (some clues: red-brown "rouge" or oxide in and around the contact region, patches of material that look smooth, a lack of parallel galling wear lines).

Aaron Tanzer

http://www.lehightesting.com

 

[metengr]

Ballising, if done correctly, introduces a compressive stress along the bore that should not result in fretting fatigue. If done incorrectly, this could be the cause of your observed problem. I think you need to carefully examine the surface for wear between the stem and valve body, and look at several new valve bores to check the condition of the bore hole after ballising.

 

[suddenDEBT]

israel/metengr,
I questioned calling it galling or fretting it but it does seem to me like either way its mechanical damage with slip stick behavior lowering the fatigue strength. The indications are mostly parallel with the bore but some do run circumfrencially (bottom of image right side). There is plenty of radial clearance between the parts and when the unit is operating it is possible for this valve to experience thousands of vibratory cycles without actuation.

I apologize for the picture quality the flash/focus weren't cooperating with me. I've attached some more images I captured with a USB microscope of another valve of a similar failure mode. It looks like mechanical damage initiated a spot for washout (jetting) to occur. Direction of flow is top to bottom in the picture of the sectioned valve.

The valve in my first post did not experience the jetting and consequential material removal. It did however share the same contact appearance and longitudinal running crack that appears to originate in the transition between the damaged and undamaged portion of the bore. Transition zone is out of the frame towards the bottom of the image. (more pictures to follow).

Mr. Failure is correct there is some red/copper/gold transfer coloring in the severly abraided regions. Do you see the material removal/smearing in the linear direction as well as some indications in the circumferential direction?

metengr,
The inlay idea sounds like its worth looking into, have you had any success with it?

Thanks again.

 

[metengr]

suddenDEBT
Before jumping off the cliff to re-engineer the current problem I re-reviewed your OP and found this statement to be significant…

We have made quite a few of these units over the years and have few failures. Suddenly we are having a rash.

This statement above indicates a process change has occurred resulting in the rash of failures. A root cause analysis approach needs to be implemented to determine what process change(s) have occurred.
I would go back and review each manufacturing step to determine what has changed from past manufacturing practice or supply vendors. Review tolerances, clearances, heat treatment and most certainly the ballising process. I believe you will find a process change has taken place and resulted in the sudden increase in failures.
Longer term you could investigate the use of a Nitronic 60 split sleeve as an insert or Stellite weld inlay for the valve body. However, this would add cost but possibly could further improve your valve performance. You would need to develop a mock-up and test performance in service.

 

[mrfailure]

Thanks for the new picture; that helps. I see what you mean about there being circumferential and longitudinal markings consistent with galling. I not that the crack shown does have the kind of contact patches one would expect in a fretting environment. The lines of discoloration on the edges away from the crack are consistent with rouge oxide pushed out from continuous contact between the valve plunger and the body. And fretting does induce fatigue cracks by its nature - galling does not. I right now would lean towards fretting fatigue as the mechanism.

Aaron Tanzer

http://www.lehightesting.com

 

[EdStainless]

I am with Metengr on this, something in the production of the bodies has changed.
I suspect that the result is an inhomogeneity of the material in the bore.
This could be chemistry, or heat treatment, or finishing.
My fist thought is that the boring and ballising are not not being done correctly.

= = = = = = = = = = = = = = = = = = = =
Plymouth Tube

 

[suddenDEBT]

Thanks metengr and mrfailure.
Completely agree on the process verification, I'm leaning towards that as a root cause with fretting fatigue being the resulting failure mechanism. The ballising/cold work hasn't been well documented, so here's the opportunity to do it.
Cheers!

 

[suddenDEBT]

Ed,
After talking to the AM operator, he tells me he verifies the process "by feel" (feel = .0007). Since we've recently expanded the night shift with new guys, I'm guessing the night operator might not have "the feel".