3.2L Steel Head Stud - Ready to Break

[Gordo2]

I was removing the dilavar head studs from the exhaust side of my 3.2L this evening. The intake side had steel studs that I planned to leave installed/reuse with the engine rebuild until I noticed this:



The left stud obviously twisted and stretched (probably as I was removing the head stud nuts).

In spite of my best effort with a Snap-On stud remover and oxy/acetylene torch it still broke as I tried to extract it. Left me with ~1.25" of stud to grab.


(working on a different stud in this pic - but same deal...).

Fought it with vice grips and torch with no luck. Decided to try the welded nut approach:



Reheated with the torch until it felt like it was breaking free only to find that I broke off another ~1/4" of the stud.

Decided I should be happy that I got the rest out without breaking - had a beer and decided to give it a week of PB before I go back at it (I also ran out of Oxy too...).

These things really can be painful. Can't wait to dig into the exhaust studs.

[Steve@Rennsport]

Don,

Two things:

1) PB is simply not very effective; its a consumer product and nothing like similar things sold to shops. Use Kroil or Zep Penetrant liberally and let it soak for 12 hours or so.

2) Patience is your friend: you really have to get things hot enough to melt the Loctite and that case acts as a huge heat sink. Take the time and heat the case at the base of each stud until it melts (325 deg F) and the stud comes out far easier.

[Lapkritis]

Gotta love aluminum jugs and steel fasteners. Good luck!

[Gordo2]

Originally Posted by Steve@Rennsport

Don,

Two things:

1) PB is simply not very effective; its a consumer product and nothing like similar things sold to shops. Use Kroil or Zep Penetrant liberally and let it soak for 12 hours or so.

2) Patience is your friend: you really have to get things hot enough to melt the Loctite and that case acts as a huge heat sink. Take the time and heat the case at the base of each stud until it melts (325 deg F) and the stud comes out far easier.

Thanks for the tips Steve,

Most of the studs came off with no heat at all (~16). I heated the stud-base/block for the studs that didn't turn right away but it only took ~ 20 sec of torch before the others broke loose.

I found I could heat one and get it to break free, take a moment to set the torch down, and return to find the stud stuck again. Sufficient and constant heat seems to be part of the trick.

I guess on this one, I'm going to have to get out of the comfort zone and heat it a bit longer/hotter. There's definitely a feel for it - something that comes with experience like yours.

Will pick up some Kroil before I start into the exhaust studs.

Thanks again, I always pay close attention when you contribute to threads (and the advice from the other pro's on the board) - greatly appreciated

Gordo

[safe]

Try using an impact wrench on the nut right after you have welded.
If it doesn't back out, try to drive it in a bit further and then go back.

[911pcars]

A last or next resort method. Remove the studs via EDM (electrical discharge machining). Luckily, you won't need that service, but try not to hurt the crankcase with excessive heat. Find a shop nearby.

Sherwood

[Lapkritis]

Ya know... I thought about this more tonight. If the bottom row is dilavar then roughly twice the pressure on the steel studs would be present when the stack and fasteners are heated over an all steel scenario. This is a terrible combination to mix fastener materials by Porsche on the same cylinder.

[Walt Fricke]

Your picture of the necked down stud strikes me as a classic representation of a fastener whose plastic limit has been exceeded from too much tension. Since the amount of material doesn't change, making the rod permanently longer has to make it narrower somewhere.

So I don't think twisting had anything to do with this, at least not twisting to remove the barrel nut. That's not going to elongate the stud. Sure can snap it, of course, and likely where it is narrower.

[chris_seven]

Originally Posted by Walt Fricke

Your picture of the necked down stud strikes me as a classic representation of a fastener whose plastic limit has been exceeded from too much tension. Since the amount of material doesn't change, making the rod permanently longer has to make it narrower somewhere.

So I don't think twisting had anything to do with this, at least not twisting to remove the barrel nut. That's not going to elongate the stud. Sure can snap it, of course, and likely where it is narrower.

A great description of tensile failure and only one small change is needed to make it perfect.

The Plastic Limit of a material only occurs at failure i.e. - when it ruptures.

When a reasonably ductile material exceeds its elastic limit it starts to plastically deform and a 'work hardening' material such as steel will initially stretch in a uniform manner.

This means that as the stud elongates the diameter will reduce but evenly along the entire length.

The load that the stud will support continues to increase as typically the rate of work hardening that takes place is greater than the reduction in load carrying capacity that occuse due to the reduction in diameter and a equilibrium condition exists. Work Hrdening typically obeys a parabolic law.

When the material uses up all of its work hardening capability it will start to develop a local neck - which is what we are seeing in this case.

The formation of a local neck typically occurs at the point of maximum load on an Engineering Stress/Strain Curve (Slightly different to a
True Stress/Strain Curve)

The load carrying capacity of the component will also be significantly reduced.

The local deformation that is occuring in the neck will be very high and at the point of maximum stress triaxiality voids will form and the material will fail.

The problem with a fastener that has been overtightened to this degree is that when it is used in a hot condition the increased load that is generated by expansion will cause a permanent deformation and the stud would have lost a significant amount of preload when the engine cooled.

The head may have been leaking quite badly during any warm up phase.

Whovever initially tightened this stud must have been completey ham-fisted not to have been able to feel the stretch and the flalling load needed to keep turning the nut.

[Walt Fricke]

Duh, not really a small change (hand slapping head) as the better side of my brain knows the difference between elastic and plastic. I can't blame my fingers, so I didn't think that through.

My vivid image is plastic failure of rod bolts - necked down where they broke in two.

I had no idea there was a work hardening effect strengthening (for a while) a fastener. But perhaps that is useful for design engineers to know, but not practical for us home brew types to count on helping us out?

[SCadaddle]

The vintage VW/Porsche guys that I call friends use Kroil with a passion on just about anything. This is the one exception where the usual results were just the opposite of what was expected in at least one case; the Kroil on the stud and using heat seemed to make things worse. That being said, when we replaced the dilavar head studs on my SC last summer we used the Snap-On tool, heat from a propane torch and the secret ingredient: Beeswax. It's a 2 man job; Bill would heat the case spigot with the torch to about 350F and then I would apply pressure with wrench making sure I wasn't leaning with it in a bending motion. Apply the beeswax to the base of the stud at the case. It would instantly melt and start flowing in around the stud. Bill would keep the heat on the case and with the other hand start rapping on the end of the Snap-On tool with a hammer. Not real hard, just a real good rap. Probably about 40 ft. lbs or so the stud would begin to turn---keep the heat on until it is all the way out or it wouldn't be long before things would tighten up again. The one broken stud we managed with a cam lock type of tool.

The real trick was trying to find beeswax. I found a 1 lb. block of it at a crafts store but didn't want that much of it (about $20 worth), then searched the local building supply house for a toilet gasket as supposedly those used to be just beeswax but nowadays who knows.....about $1.25 later I found myself walking out of the drug store with a chap-stik sized tube of Berts Bees Wax lip balm which was like 95 or 99% beeswax as per the label....and the minty green smell of success was the ticket!

[Gordo2]

Interesting thoughts regarding how the stud came to be stretched.

I assumed that it stretched during disassembly mainly based on the odds:

-The stud was so close to the breaking point I figured it should have broke when I tried to remove the head nut.

-I expected to break a stud when I was removing the head nuts and was very happy when it didn't happen.

-I really didn't need to rebuild this engine but decided to go all-in (to upgrade). My decision obviously averted a potential near-term stud breakage - what are the odds

Other thoughts:

I considered the idea that the stud was stretched prior to removing the barrel nut, but I didn't find any signs that the head was leaking (no black carbon residue outside of the head/cylinder mating surface).

I do think it is possible that it was over-torqued (either originally when rebuilt by the PO or in a follow-up re-torque). I had 1 stud that extended too deep into the barrel nut making removal quite difficult (the allen key didn't seat very deep). However, I didn't note which stud this was, but I tend to think it was the same one that I broke later.

I also researched and found the bee's wax trick (and oddly enough the first thing that came to mind was Bert's lip balm). But then I thought about cleaning the wax/gunk out of the threads to install the new studs and decided I would stick to normal penetrants and heat. I would think you would want to hit the threads with a propane torch to clean them out after using wax to assist removal.

Prior to finding this, I was considering just ordering 12 new steel studs - bumped the budget up a bit to pick up a set of 24 ARP studs (street motor otherwise would have considered Supertec or Raceware). Plenty of confidence in OEM steel studs based on BBS research, but the sight of this (and difficulty in removing) pushed me over the edge.

Thanks again for the thoughts & insights guys.
Gordo

[Walt Fricke]

Pretty much the only things which could stretch a stud would be thermal expansion of the cylinder/head combination at a higher rate than the stud expanded, and overtorquing. Oh, and the force of combustion. And these would go somewhat hand in hand - overtorquing would get the stud closer to its elastic limit, and the thermal expansion differential could push it over the edge I suppose, aided by combustion pressures.

There is a system for tensioning fasteners called torque to yield or something like that. A fastener which has exceeded its plastic limit is not necessarily going to fail, at least not right away. It will still clamp the parts together, and it will do so with a force compressing them by the amount of residual tension, even though it is permanently deformed and its ultimate strength has been reduced. It may be that stock Porsche rod bolts are of this sort, which would explain why Porsche stresses that they are a one time use fastener. A guy could check that if he used the bolt stretch method (or just measured stretch after torquing as usual). If the bolt, after use, returns to its original length when unfastened, it hasn't exceeded its elastic limit. Those of us who use stronger aftermarket rod bolts do so (at least I do) in part because they can often be reused if they end up only a thousandth of an inch or so longer after use. But this kind of permanent stretch doesn't show up as a significant necked down area like yours had.

Still, as Chris Seven noted, this is an odd thing. There is a limit on how much torque you can apply to a barrel nut, and that limit decreases as the tool insertion depth decreases, as you found out. If your studs are too tall, the stud will force the tool out of the barrel nut before you get the torque you want.

Personally, I prefer to use flange nuts. Avoids all those issuesj. You can get them with a reduced head size which still allows a regular socket (say 15mm) to develop plenty of torque. The side view cross section of a flange nut is rather smaller than the barrel nut's, which presents less of an obstruction to air flow. Flange nuts are especially useful with twin plugged heads because they give more clearance to the second plug. Though that is not your issue.

Maryland Metric has a torque chart for bolts. For 10mm bolts, an 8.8 has max torque listed as 42 lbs/ft. 10.9 is 60, and 12.9 is 72 lbs/ft. My experience (hardly good enough to generalize) is that these are somewhat conservative, and you can overtorque by 20% or so without expecting to twist off the bolt or stud. Their figure for 12mm fine thread 12.9s (like the flywheel bolts for the six bolt cranks) is 126 lbs/ft. I torque mine to 150 (spec is 110 or so), and have never had one fail.

So this is, indeed, a mystery. But not one which happened while you were trying to unscrew the nut.

[chris_seven]

Originally Posted by Walt Fricke

Pretty much the only things which could stretch a stud would be thermal expansion of the cylinder/head combination at a higher rate than the stud expanded, and overtorquing. Oh, and the force of combustion. And these would go somewhat hand in hand - overtorquing would get the stud closer to its elastic limit, and the thermal expansion differential could push it over the edge I suppose, aided by combustion pressures.

Very unlikely that thermal expansion and combustion forces have caused this type of deformation.

For Example a Grade 12.9 bolt is likley to have an elongation at failure of at least 10% when subjected to a monotonically applied load.

This means that a fastener with a length of 150mm will stretch by around 15mm when it ruptures.

The ratio of uniform to non-uniform elongation is typically 1:1 for these types of steels. This means that the stud would stretch about 7.5mm before it developed a local neck and then a few millimeters more before the neck became as pronounced as we have seen here.

Even if the material of the stud is less ductile we are still talking of a few millimetres of deformation.

The differential expansion of the two materials we are considering would only result in a displacement of around 0.2 millimetres which is nowhere near enough to cause the amount of plastic deformation being displayed.

The torque needed to yield this type of stud is probably around 40-50 lbs ft and this would increase due to work hardening until the local neck developed.

Even if this was only 5mm of displacement it is still more than 3 turns - hence my 'ham-fisted comment- After the local neck has developed the torque would be reducing and a further couple of turns has probably been added which should have been easy to sense.

I don't think the socket head makes much difference as all of the deformation is 'down stream' of the threads so the socket depth won't change.

The head will still be clamped and if the forces generated by expansion and combustion don't exceed the prevailing preload than the engine will run quite happily.

As I said earlier, however, the load carrying capacity of this stud has been badly compromised so it wouldn't be surprising if it did lose preload when running.

[Walt Fricke]

I agree about ham fisted, if that was the cause. Having "torqued off" a few more fasteners than I should have, I know the sickening feeling when a fastener starts to yield. You get more movement of the wrench than you should get for the force you are applying. Oh oh. Depending on how hard it is to replace the fasteneer, you start thinking maybe this is due to getting past a rusted spot on the threads or something. Wishful thinking.

The normal installed depth of the studs is 135mm (I remember this figure from having just screwed in studs until they bottomed, and then finding I didn't have enough thread sticking above the thick washer). The barrel nut has a threaded part, and then the hex part for the socket. Typically, when the prescribed torque is achieved the top of the stud is roughly at the top of the threads. As you keep turning, the stud starts pushing the tool out of the hex. The shallower the hex engagement, the less torque the nut can take without rounding the hex edges.

I measured a barrel nut (well used) at 18.5mm. Calling it 18mm means 9mm of it is threaded I think (I can't find my stash of "good" nuts, as I don't use them any more). At a pitch of 1.5mm, six turns uses up all the threads. If, say, 26 lbs/ft is the target torque, you could calculate how much stretch that would already produce out of the roughly 10% (say 12mm, discounting for the part of the stud now above the lowest two threads of the nut). And how much more stretch to yield. Based on the Maryland Metric figures, I'd have thought that is rather more than 50 lbs/ft (if these are 12.9 - I don't know that, as you don't see that figure for head studs - the aftermarket ones just give the PSI rating of the special alloys used). But either way, a guy who knew the stretch/strain relationship for the steels involved ought to be able to calculate how many more turns it would take to various yield figures. And how many turns to achieve that, and how much farther up into the hex area that would move the tip of the stud.

Thinking another way, a guy ought to be able to tell the difference between 25 or less lba/ft of torque, and double that. After all, you get a feel torquing 24 fasteners. So ham fisted it is for sure.

I am just wondering if that is even possible with a barrel nut.

But it is hard to think of a stud being defective out of the box in this way, which seems the only other explanation.

And so far, no one else has chimed in and said Oh, I've seen that neck down failure of a head stud. Lucky Don - now he can say that.

[304065]

Walt-- there's a real chestnut in there-- six turns uses up all the threads.

I often wonder why the Factory used a nut that when tightened, eventually will push the wrench out of the internal wrenching feature, probably stripping the nut as it is expelled.

Perhaps this is the answer-- perhaps they did this as a way of yelling STOP to the assembler who is attempting to torque the nut down with a vengeance?

[911pcars]

There are an infinite number of ways one can tighten a threaded fastener and only a couple of them are recommended, maybe three. That's why.

Sherwood

[Lapkritis]

Originally Posted by 911pcars

There are an infinite number of ways one can tighten a threaded fastener and only a couple of them are recommended, maybe three. That's why.

Sherwood

I've seen a mechanic use an impact gun on head studs to tighten in VW land years ago. Maybe he got his hand on this Porsche...

[Walt Fricke]

Aha - leaving aside internal wrench limitations, an impact gun could do it.

[Lapkritis]

The Porsche wrenches are generally gentle... lots of ham-brains out there at shops too.