Hydrogen Embrittlement ASTM A574

[TugboatEng]

Today I have inherited several vessels with obsolete drive units and no factory support. The drives have submerged.structural flanges and are notorious for breaking bolts. 30 years later, I think it's time to look into a cause. The fastener are socket head caps screws of ASTM A574 standard and some shoulder bolts of no specific standard. My understanding is that the shoulder bolts have never been found broken. It appears the shoulder bolts are 7/8" and 32 HRC. The bolts that are breaking are 1-1/8" and 37 HRC. Another manufacturer of similar equipment uses exclusively class 8.8 fasteners, I assume to reduce the risk of hydrogen embrittlement.

Here is a picture of one failed fastener. Are their any visual clues of HE? Can HE be determined if the fracture faces are corroded?

I'm thinking of specifying ASTM A193 B7. The problem is that lower strength socket head cap screws are not commercially available in imperial dimensions.

 

[3DDave]

Off the top: https://www.boltcouncil.org/files/HydrogenEmbrittlementInSteelFasteners-Brahimi.pdf

Not sure it's hydrogen. It could be, but what I have seen in hydrogen embrittlement has been like the part is made of glass; just shatters right away.

However I cannot say I have seen every possible case and usually the fracture, such as from fatigue, goes perpendicular to the axis rather than this weird cone.

The usual source of hydrogen is some part of the manufacturing process, such as pickling or electroplating or both. This report describes what the metallurgists I worked with had to say on the subject; mainly that plating or passivation should be followed by baking to cause the hydrogen to exit.

It also mentions the failure of fasteners that are subject to high temps and the problem that cadmium plate poses as a stress risk. There was some engine the supplier had used cad plating on the exhaust headers and they had a 100% fail rate. The spec for the fastener specified "NOT FOR USE ABOVE X00ºF" of some similar and suspiciously low limit. ( I forget the exact temperature. It's been 20 years.)

So, if unsure, you could simply require baking the fasteners at a few hundred ºF for several hours before installation as mentioned in the report.
---
This site has pictures of various failures; the one that matches is stress corrosion fracture: https://smrp.org/News/Solutions-Arc...y-Learn-to-Recognize-Mechanical-Failure-Modes

Stress Corrosion Cracking

Bolts being under constant tension leave them susceptible to stress corrosion cracking (SCC) if they are exposed to water-based corrosion agents. The effects of any corrosive environment are made worse at warmer temperatures. Over time, bolts exposed to the wrong compounds can begin to suffer cracking and eventually fail. Each material type is susceptible to different corrosion agents. Hardened carbon steels are at risk of SCC when exposed to a variety of compounds including ammonia, caustics and sulphur-based compounds. Stainless steels are most susceptible to chloride-based cracking.

---
Other failures of this type produce a cup-and-cone surface pair when the fastener is not strong enough, though those seem to have a lot more tensile section reduction. Maybe the stress corrosion cracking causes the tensile failure without allowing much reduction in section.
---
Recommendation - install the bolts with polysulfone rubber sealant. Apply liberally to threads and ensure all small cracks that could let water get into the threads are filled, like under the head. It cannot corrode if it cannot get wet.

 

[MintJulep]

ASTM A574-17: 5.5 Standard Finishes—Unless otherwise specified, the screws shall be furnished with one of the following “standard surfaces as manufactured” at the option of the manufacturer: (1) bright uncoated, (2) thermal black oxide, or (3) chemical black oxide. Hydrogen embrittlement tests shall not be required for screws furnished in these conditions.

So, the folks that wrote the standard weren't worried about hydrogen embrittlement with these three finish options.

Looks like your screws have one of the two black oxide finishes.

The shape of the fracture looks most like stress corrosion cracking from the pictures at the link provided by @3DDave.

Probably not hydrogen embrittlement.

Black oxide is mostly useless at preventing corrosion, as evidenced by the all the corrosion on your screw.

These days you should be able to get A547 SHCS with Aluminum-Zinc Flake coatings (i.e. Geomet or Dacromet). Far better corrosion protection than black oxide, and the process does not cause hydrogen embrittlement.

The Unbrako catalog sort of suggests that it's possible. https://unbrako.com/images/downloads/Unbrako_US_Product_Guide.pdf

Where are these installed? What's the loading?

 

[LittleInch]

Do you have a sectional diagram of the bolt installation?

That looks more to me like crevice corrosion / fatigue of the bolt on the left caused by stress concentration at the point it enters the hole. but its not really possible to see is it he bolt that's the issue or the design?

 

[TugboatEng]

Stress corrosion cracking, ISO 898-1 calls this out as a problem for these fasteners. The zinc-aluminum flake coating sounds like a very good solution and should be commercially available. I will pursue this route.

 

[TugboatEng]

 

[mfgenggear]

Is there sea water contamination?
Are there any chemicals stored in the same area?
Possibility of electrolysis and there are no zinc anodes to mitigate?

 

[Andrew ONeill]

I once had to deal with a significant stress corrosion cracking failure of B7 studs. These looks similar, but in reality, to know anything is to send them away for testing.

In saying that I learnt 2 super important factors in relation to fastener manufacturing, that are not obvious, and are not mandated by a lot of standards, but are usually followed (i.e. most manufactures do the following, because it give a better product, but the standards don't mandate it, occassionaly a manufacturer doesn't follow these requirements, because it's cheaper)

1) Rolled threads are vastly superior the machined threads, in particular for fatigue and stress corrosion cracking prevention
2) It is critical that heat treatment is done prior to rolling threads, not after thread rolling.

The above may or may not relate to your failure, but if you get stuck on why some fastener is failing vs another that seems identical, it's worth looking at the above two items.

 

[TugboatEng]

Yes there is sea water exposure. The flanges and counterbores are corroded so sealing is no longer ideal. There is surprisingly little corrosion on the bolts considering the conditions. I am going to heed MintJulep's advice but cannot find zinc-aluminum flake fasteners on short notice larger than 1" diameter. Instead, I plan to coat all surfaces of the fastener with Krytox 227 or Tef-Gel. There is water present but no flow so a grease should provide durable protection. The Krytox 227 has the benefit of a film forming corrosion inhibitor.

 

[MintJulep]

cannot find zinc-aluminum flake fasteners on short notice larger than 1" diameter

Too bad, as with so many things these days, they are "available", on paper, until you ask for them.

Some possible alternatives to Tef-Gel:

LOCTITE® LB 8023

LOCTITE® LB 8023 is a black, metal-free, brush top lubricant and anti-seize made from graphite, calcium, boron nitride and rust inhibitors. ABS approved, it protects assemblies from fresh and salt water. It works especially well in high humidity conditions. It has excellent lubricity , superior...

next.henkel-adhesives.com

Vibra-Tite 906 Marine Grade Anti-Seize

www.vibra-tite.com

 

[TugboatEng]

I am starting to divest myself from anti-seize. It should be called anti-gall. I find greases provide better seize prevention as they contain more oils that protect surfaces from corrosion. Anti-seize is mostly powdered EP lubricants with a little oil to make them sticky. Avoid the soap thickened grease. We have polymer thickened grease now.

 

[MintJulep]

Try these guys.

Fastener Supplier for OEM, Distributors, and Industrial Applications

Big Bolt is a fastener supplier providing next-day manufacturing, even on custom orders. Learn more about our OEM and distributor fastener products and services.

www.gobigbolt.com

I have no affiliation, but they've supplied to a lot of train manufacturers that I've worked with over the years.

 

[TugboatEng]

I'm working with a local supplier. They can source custom bolts but I've only given them one week. We're going to install what we can get with some method of sealing for the threads and under the heads and then perimeter weld the flanges. They were welded previously and both the bolts and welds broke. We'll do some back gouging, I don't think that was done previously. Apparently this problem has persisted for nearly 30 years. It's just new to me and I don't like doing the same thing and expecting different results.

Autocorrect did me dirty in the earlier post. I'm going to "heed" your advice.

 

[MintJulep]

How many of those arms are there between the pod and the ring? What is is made of?

I can imagine that setup hitting a resonance at some conditions, leading to fatigue failures.

 

[TugboatEng]

The nozzle has 3 attachment points, two struts and a flange at the top. All 3 points have a history of fastener failures. It's worse on some vessels than others. We have 3 identical vessels now totalling 6 units.

With the flanges and counterbores corroded I cannot guarantee no perpendicularity of any surfaces and I know that this causes bolt failures on its own. But, these failures started when the units were new and these surfaces were in good condition.