The End Fell Off

[Ken I]

An M24 Cap Screw is obviously not from a model but I thought some members might be interested.

The end of this bolt simply popped off - it is due to hydrogen embrittlement and a lap crack underside the head (manufacturing defect). It was not very highly torqued and not subject to cyclic load.

Hydrogen embrittlement occurs when the metal is exposed to nacent Hydrogen H (as opposed to H2) - which then dissolves into the solid metal - termed a solid solution.

For reasons not fully understood the Hydrogen tends to migrate to high stress points - there it will find a void or grain boundary and it will meet and pair up with another H radical to become insoluable H2.

Believe it or not this process builds enough pressure to tear the material apart - grain by grain, until failure occurs.

One of the most common causes of hydrogen embrittlement is electroplating - where nacent Hydrogen is liberated at the Cathode.

The problem gets worse with higher tensile and hardness values - spring steel is particularly susceptible.

The lesson - when electroplating parts in high tensile materials be sure to de-embrittle them propperly by heating to 250°C for 2-3 hours immediately after plating. Immediately means right after removal from the plating solution.

This bolt probably missed that process. The stress raising lap crack prompted the Hydrogen to migrate and propogate the crack further and further - I came into work and the head had popped off during the night.

Ken

 

[Herbiev]

Thats what i love about this forum. I learn something every day

 

[Lakc]

Thanks Ken, any chance you have a decent microscope to get some pictures with?

 

[tel]

Interesting! I've seen that happen more than once, but never knew the cause!

 

[Brian Rupnow]

I have known about hydrogen embritlement for many years---Mainly because of my life long association with hotrods. There are many chrome plated front suspension parts being sold. Some are made in USA and some are made offshore. At best, its a gamble when running front axles, spindles, etcetera that have been chromed. Better a painted front end part that gets stone chipped than a catastrophic failure of the suspension part thats holding up the front of your car!!!

 

[Ken I]

Brian, Spot on - plated performance parts - especially springs and wheel spokes are very bad news - painted is safer.

I know one German spring manufacturer that point blank refuses to make plated springs - just simply not interested at any price.

Lakc - I'll see if I can rustle up some micrographs later.

Regards,
Ken

 

[don-tucker]

I bet this is why on undoing a 14mm wheel bolt on my motor home it sheared off,that was plated.
Don

 

[steamer]

That may explain the small ball joint breaking off on my son's RC car. They are suspension joints and they break where the stem meets the ball. They break before the plastic they are threaded into breaks!


I would bet they are chrome plated....hmmmm

Dave

 

[Ken I]

I posted this thread as this is a classic example.

Obviously not all breaks are Hydrogen induced.

An overload break normally evidences some plastic distortion and a "grainier" break. Alternatively a shearing kind of break like the torn end of a bolt we simply twisted off (we all know what that looks like).

A fatigue failure normally has a fairly smooth crack propogation zone as a series of smoothish half moons - with only the final break exhibiting granulation or shear. (If you've ever seen a broken crankshaft - they mostly look like this.)

A high impact brittle break looks pretty much the same although generally "grainier" - but in the absence of an obvious cause - then Hydrogen is the suspect.

Other causes of Hydrogen embrittlement - welding with damp electrodes / presence of water (that's why coded welders keep their rods in an oven).

Some forms of corrosion can cause it (especially where sacrificial anodes are employed) and can be created in HP steam & reactor environments.

Steamer - I have seen heads fall off the bolts "in the box" there is allways going to be some stress to get the Hydrogen all worked up.
You can destress parts long after plating but it takes much longer and some damage may already have been done. Try 250° for 24 hours next time and see if they last longer.

FYI

Ken

 

[steamer]

Thanks Ken...I'll do that! The break looks very velvety like the bolt that you show. It' is also happening at a stress riser where the ball meets the shank. It's about .156" diameter there and the nylon swing arm it's connected too can't possibly be that strong.

Dave

 

[MachineTom]

If you don't mind keeping with failure photos ( let me know otherwise) Here is a SS 3" propeller shaft from a fishing boat.

Starting at the 12:00 position a crack begins at the keyway, likely due to grounding, the beach marks show it progresses counterclockwise until the 4:00 position, then fails all at once. Guessing it took a couple month to get to the failure point.

Final failure

 

[Ken I]

Yup - the "beach" marks are classic progression of a fatigue crack.

Not much evidence of a starter crack from a grounding - that bump near the crack origin looks suspiciously like a defect inclusion of a nugget of harder material - but it aint necessarily so.
More likely as you suggested the grounding tried to tear the corner root of the keyway open somewhat - and the rest is history.

A badly vibrating prop wouldn't do it any good thereafter either.

Diagnosis from a couple of photos is dicey - but there a few things I would want to look at closely - the suspect nugget - the area just above it looks odd could be a defect or simply where the crack started - it looks like either a circumferential groove or scratch or work hardening of the skin on the OD - but what the heck its still broken.

3" wow - there's no upper limit to failure.

Ken

 

[Lakc]

Starting at the 12:00 position a crack begins at the keyway, likely due to grounding, the beach marks show it progresses counterclockwise until the 4:00 position, then fails all at once. Guessing it took a couple month to get to the failure point.

I origionally read that as clockwise, and looked at the pictures and said to myself the darker areas were cracked the longest, then I reread your comment correctly. Thanks for the pic, Im a rank amateur at this type of failure analysis, learned something new today. (beach marks)

 

[Ken I]

Lakc - Heres a micrograph - sorry it didn't come out too well but depth of field is limited at this magnification.

The black zone to the lower right is the lap crack (material folded over itself during hot heading process - a manufacturing defect) - what you can see thereafter is simply failure grain by grain - there is no shear or plastic deformation evident - a sure sign of Hydrogen embrittlement.

An impact brittle break would also tend to produce a 45° conical break, this is much flatter.

Ken

 

[Lakc]

Thanks Ken, that looks just like a rock that was cleved in twain.

 

[Swede]

Very interesting - it reminds me of hydrogen in molten aluminum, a common problem for home foundries, that comes from moisture in the air. Gave me fits until I made a simple homemade "lance" and bubbled argon through my melt. I know that's not the same as hydrogen embrittlement, but we're seeing the same deal... a gas "dissolved" in a metal.

Years ago, I electroplated a model gas turbine rotor shaft of 4340 steel with nickel to protect it and to build up the bearing journal areas. I never thought of hydrogen embrittlement. Fortunately, the shaft sees pretty high heat every run, so by now it's probably driven off.

 

[Lakc]

This failure should not be hydrogen embrittlement. My take its a simple fatigue failure, broke during cranking. Material is 4140.

And a few links I have been following lately, as this topic has been very relevant lately.

http://en.wikipedia.org/wiki/Fractography
http://www.tech.plym.ac.uk/sme/Inte...alysis/Fractography/Fractography_Resource.htm

 

[Ken I]

I'd agree - looks like a fatigue failure.

Probably starting from too sharp a corner rad or score line in the rad / corner.

Tip - roller burnish those corner rads - this imparts a compressive stress to the surface so that under tensile stress the greatest stress occurs below the surface.

Alternatively shot peen the hell out of it prior to grinding or the lightest of finishing cuts.

Ken