crack in PSV spring 1

[nabhan81]

Hi,

we are suffering from PSV spring crack

Material: 17-7 PH stainless steel
operating temperature: 221 F
set pressure: 1760 psig
size: 2"
service: boiler feed water, good quality and well treated
There is some corrosion in the spring but it is not on the carbon steel piping!
See attached photo

Any one can help on why it fails???

 

[MikeHalloran]

It's mostly iron, and lives in a hot humid environment?

I'm guessing the spring itself, thanks to its location, is not as not as the piping, so isn't guaranteed to stay dry.

Dip the next replacement in Dykem, twice.

Or insulate the valve so the spring stays warm.

Or specify plastic-coated springs.



Mike Halloran
Pembroke Pines, FL, USA

 

[btrueblood]

Does this spring not live outside the water stream? If so, do you see deposits on the springs, e.g. salts from evaporating water?

 

[jimboux]

It's a bit hard to tell from the pictures but it appears to me that possibly the spring is cracking because of bending across the section, which would happen if the spring is in torsion about it's central axis.

This is often overlooked in spring design because the normal design practice is to only design it for torsion about it's section as a result of the compression.

Any chance the spring is being twisted about the longitudinal axis?

 

[TVP]

17-7 PH does not have outstanding corrosion resistance, so depending on your definition of "good quality" water, it is quite possible that the spring has pitting corrosion due to the combined effects of elevated temperature and small levels of Cl- ions.

 

[israelkk]

can you provide spring drawing, loads, deflections, heat treatment, surface treatment, cycles between loaded points, etc? The orientation of the cracks are as shear stresses should show. If the stresses are not too high then stress corrosion may be involved.

 

[Ron]

Looks like a possible lamellar defect in the material.

 

[metengr]

SCC, EAC, corrosion fatigue, material defect, ummmm..send it out for metallurgical analysis!

 

[nabhan81]

Spring is always live

There is no much salts in the treated water
Conductivity is 6400 micro simens/cm
Chloride < 2000 ppm
There is no indication of bending or permanent deformation
Oxygen < 10 ppb

I dont think 17-7 PH is not suitable for this service

What kind of metallurgical testing will help on this??

 

[metengr]

nabhan81;
If you can remove the cracked spring and send it to a metallurgical lab it can be destructively examined to determine the cause of failure. Otherwise, all you can do is guess.

 

[swall]

I was initially confused. Is the issue one of the spring tips breaking off rather than corrosion?

 

[metengr]

If you look at the attached photograph in the OP there are cracks in the coil spring itself. These failed springs need to be sent out for poper failure anlysis considering they are from a safety valve.

 

[btrueblood]

So the spring is always wet/submerged?

 

[nabhan81]

it's easy to say send to metallurgical lab, but what kind of test and what is the target of testing? what are the suspected mechanism?

I am waiting for experts !

 

[swall]

I will second what metengr said three times. Send the pieces to a metallurgical lab for failure analysis.

 

[MikeHalloran]

>>>it's easy to say send to metallurgical lab, but what kind of test and what is the target of testing? what are the suspected mechanism?
<<<

Don't tell the lab what tests to run.
Ask them to tell you why the parts failed.



Mike Halloran
Pembroke Pines, FL, USA

 

[metengr]

nabhan81;
You have been given sound advice by MikeHalloran and others. You have several possible damage mechanisms. If I could tell you exactly why it failed based solely on a public internet forum, I would quit my day job and become very wealthy.

 

[mrfailure]

The problem with specifying what type of testing is to be performed is that the nature of the work can change as a result of investigation findings. Many think they can get to a failure root cause just by testing for specific material properties. That is usually not the case. One of the most important steps in an investigation such as this will be visual examination of the suspect springs (optical and possibly in an SEM). Appropriate testing to identify the failure mechanism(s) has to be performed as part of the investigation, as well as identification of relevant material properties.

You should engage a competent materials lab. They will be able to propose an initial workscope to you based on the information you provide. Good luck!

 

[nabhan81]

Thank you all for your comments,
Mr. Failure, you are the most guy who could understand me..

anyway I showed it to a corrosion specialist and he commented it looks like hydrogen crack