17-7 PH disc springs failure!!! 13

[zztop]

I had disc-springs made of "equivalent" material to 17-7 PH called 1.4568 (european standard).

They are now completely broken and I don't know why. The machine has been out of use for two years near the sea... till they were going to use it.

The springs were loaded, but no vibration and no shocks at all.

Does anybody know if this material shows any type of "stress cracking" with H2S or marine environments?

ZZtop

 

[NickelMet]

To answer your question zztop, yes, a disc spring working under those conditions. Very low chloride levels normally though.

I too am enjoying this thread and would also, like you it seems, to see a graduated list of materials for chloride environments.

~NiM

 

[btrueblood]

Now I'm getting curious too, and agree that this is an interesting thread.

I discussed this briefly with a colleague, who mentioned that he generally ignores aqueous chloride ions (in contact with stainless steels) unless the temperature is above about 150 deg. F, because the chloride is not "liberated" below that temperature. Is he right?

Thanks,

Ben T

 

[unclesyd]

btrueblood,

No one can ever ignore Chlorides and Austenitic SS no matter what the variety or temperature though you might get by in the arctic.
Industry wise we have experience SCC of SS at all temperatures but we usually talked 100?F which is a good ambient in Florida. I've seen all sorts of data on Cl levels, temperature, alloy, condition, enviroment, and time, to name a few, saying that SCC will or want happen. It works out that Chlorides + SS mean trouble and they need to be kept apart if at all possible regardless of the temperature.

Anecdotal:
I used to do a little commercial Red Snapper fishing and we used 316 SS cable on or reels. The cable was 3/64" 7x19 McWhite, aircraft control cable. Several times I've witnessed that after putting a brand new cable on a reel and after less that an hour use, the cable would part underwater. Metallurgical examination of the cable revealed SCC of the finest kind. The cable could fail anytime and place.
The muffler on this boat was made from scrap 316 SS pipe (10" and 4") and sat exposed to the weather. In heavy seas you see the salt spray frying on the surface. It was 30 years old with no cracking. I fought off every attempt to clean it up.

 

[zztop]

unclesyd, I have also a 18-8 stainless steel spoon in my salt shaker, and it's been working fine for years, without signs of corrosion... Oh God!. Corrosion is a too much big mistery for me!

I just want something which helps me to find the right disc spring.

Anyway,

thanks to all of you by your interest

 

[israelkk]

zztop

NASA MSFC-SPEC-522B, DESIGN CRITERIA FOR CONTROLLING STRESS CORROSION CRACKING is a good reference guide for selecting material for stress corrosion resistance. It can be found on the web as a pdf file.

A loaded disc (belleville) springs has both compressive and tensile stresses zones at the same time. Ususlly, bellevile spring are designed for compressive stresses which can go up to 300% of ultimate compressive stresses with no problem especially if they are used for static applications. However, care should be taken regarding the tensile stresses zones which are the responsible and suseptible (I believe) for stress corrosion cracking.

Therefore, maybe instead of replacing the 17-7 PH CH900 which is categorized as high resistance material to stress corrosion according to NASA MSFC-SPEC-522B you can try to lower the tensile stresses in the belleville spring. For example by preseting the springs to induce comressive stresses in the tensile stress zones. Or use shot peening on the tensile stress zones on the bellevile springs.

Could you give:

1. The dimensions of the bellevile spring (or the catalog data)
2. The load and deflection
3. The space available

http://www.webspawner.com/users/israelkk/

 

[btrueblood]

Israelkk, Unclesyd, a star for youse, thanks for the reply.

I am in the midst of a re-redesign of a regulating valve, part of which incorporates a big honking spring, which will have a loading right up near the fatigue limit. Normally, the valve would see fresh water and maybe some glycol (chilled HVAC water), but since some areas (Florida is one!) may have cooling towers exposed to salt air, the possibility of trace chloride in the water is not so remote.

Since the concentration/time history of exposure will be beyond my control, I think I will ask my supplier if we can use 15-5PH wire for the spring. We were originally looking at 17-7PH, but it is listed in the MSFC spec as "ok" for SCC only in condition CH900 -- which it looks like I can't specify, since it requires a special rolling mill treatment. 15-5PH has tensile properties similar to 17-7PH that he proposed, in the attainable post-winding condition of H1050. Am I way off base with this logic?

Thanks,

Ben T

 

[israelkk]

btrueblood

17-7PH spring wire is a standard spring wire and it comes in condition C. After coiling it is treated (heated to) 900F to become CH900. Every respected spring manufacturer can make the spring to your design.

To my best knowledge 15-5PH does not come in "spring wire condition" and 15-5 PH H1050 strength is much much lower than 17-7PH CH900.

 

[btrueblood]

Thanks yet again, israelkk. I just got the same info. from my spring manufacturer.

Ben T.

 

[zztop]

israelkk. Thank you very much. The report from NASA looks very good.

zztop

 

[NickelMet]

Here's another suggestion, did you check the springs to make sure they were the correct length? I just reported on a failure here where the springs were a bit too long, were put into service, and failed within 10% of the normal cycling. The only explanation we came up with was that they were of the wrong length, thus over stressed due to over compression. (At least that is what 2 labs agreed on ... I'm still not 100% convinced ... another story for another time though.)

~NiM

 

[israelkk]

NickelMet

Disc springs are just coned washers made of a strip or sheet they are not coiled from a wire as a helical compression spring.

Concerning your example, this is simply a matter of bad design where either the designer did not design the spring correctly or the one who chosed the spring just took the first spring he could find and used it without calculating deflection and stresses (you will be suprised how many of such cases I have encounters durig my 25 years as an engineer in the aerospace area).

http://www.webspawner.com/users/israelkk/

 

[NickelMet]

Thanks for the clarification Israelkk... I got turned around in the list of commentary. And, that is interesting about the aerospace incidents, but I guess not really not too surprising. Things happen in every industry that makes us scratch our head.

~NiM

 

[EdStainless]

Going back to a question earlier by zztop, about the ranking of materials.
There is one crevat on this list. Any material that is improperly heat treated, especialy senitized, will fail rapidly. This list assumes correct heat treatment and clean surfaces.
1. At the bottom of the list are the 7-10% Ni austenitics. Give me a few Cl ions and an atom or two of S and I can crack them.
The work by Copson is still valid for austenitic alloys. The CSCC is governed by the Ni content.
2. Alloys with 25%-35% Ni have some limited useful resistance to CSCC. Alloys like 6%Mo superaustenitics and 310 and 800 are in this group.
3. The duplex stainless grades also have some reasonable CSCC resistance. This is also a function of chemistry, but not Ni content. The higher Cr alloys (27%) have the highest resistance, the 25% Cr are in the middle and the lean alloys (19-22% Cr) are lower. Though the lean alloys are still much better than low Ni austenitics.
4. The ferritic stainless alloys fall into two groups. The alloys with no Ni (<0.5% such as 430) are totally resistant to CSCC. Ferritic alloys with some Ni (1-2%) are more resistant than even high alloy duplexes, but they can be cracked. All of the field failure data that I have see with these has been in sensitized material.
5. PH grades and Ni based alloys are tougher to catagorize. A lot depends on the specific strengthening mechanism. Alloys like 13-8, 15-5, 17-4 and Ni alloys like 718 and 188 are not very CSCC resistant. While alloys like 17-7 and A-286 are fairly resistant to CSCC (perhaps similar to 25% Cr duplex).
6. The top of the heap are Ni-Cr-Mo 'C' type alloys (276, 22, 686, 59). A couple of odd alloys also are in this class, alloy 33 and Allcorr. These alloys will not even crack when they are highly cold worked. For high strength the multi-phase alloys like MP-35N and Elgiloy are the answer. Some aero alloys are in this group also such as X-750. The other option would be Ti 3-8-5-4-4.

I need to reinterate, if any of these alloys are not properly heat treated, or are exposed to service temperatures that are too high (for microstructure stability) they will crack easily.

= = = = = = = = = = = = = = = = = = = =
Corrosion never sleeps, but it can be managed.

http://www.trenttube.com/Trent/tech_form.htm

 

[NickelMet]

Good post Ed. (You get a star for that one!) This little discussion is being hard-copied to my materials files. Good stuff.

~NiM

 

[Maui]

I'll add this discussion to my reference list as well.


Maui

 

[DiscSpringGuy]

A very informative and interesting, two month, dialog about disc spring materials. Who knew?

As an actual manufacturer of disc springs, I absolutely concur with these posts. The only caveat is that one should only take into consideration materials that are actually available and used in disc spring making for the particular material thickness of spring required to do the job. This alone will chop the innumerable types of corrosion resistant steels down to just a very few.

17-7PH and X7 (1.4568) are essentially the same materials. It is not a material we recommend because it is inherently sensitive to stress corrosion and other commonly available materials display better corrosion resistance and are much simpler to process.
I.e. The basic, very unsexy, SAE 301 (AMS 5519L) or what we call X10 (1.4310 – DIN EN 10151) has proven quite adaptable to corrosive environments: sea water, sour-gas, etc. Since this is a cold worked material it is not available in thickness above 2.5mm. Above that X39CrMo17-1 ( 1.4122 – DIN EN 1008-2) is that standard material used. This is generally a temperature resistant material, but is OK for corrosion resistance. This is a material susceptible to stress corrosion so the possibility of delayed brittle fracture can not be totally ruled out.

Nickel alloys of course are the other option. The ones actually used are Nimonic 90 ( 2.44632 / 2.4969) or the Inconel X750 or X718. The downside to these are their cost and availability, the complexity of processing and some functional issues such relatively high setting due to their toughness (high strength / low elastic ratio).

The inherent design of the disc spring and its specific performance criteria plays no small role in all this as well. I’d be more than happy to look at zztop’s basic application criteria and try to recommend a design, including material that would be both appropriate to the application and commercially available at reasonable cost and delivery time.

A.S.


PS : I’m new at this sort of format, so please let me know if any of this has been helpful

 

[schrijen]

the most possible reason is hydrogen embrittlement because this type of steel is martensitic with high hardness.
and Chloride SCC can not occur in this type of steels only pitting.
The cathodic hydrogen caused by the corrosion reaction can occur the embrittlement. Cl SCC in 17/7 or 17/4 does not occur.

Hydrogen embrittlement in combination with H2S for martensitic chromium steels is well known.