FFaust

Dropped by the race shop today, the same that Boinkus mentioned above, for a courtesy visit, and the topic of broken studs came up. They again said that they had never had a stud failure, and that they have been using these same studs on many race cars for a long time. Also said that nuts were often over-torqued during races (or at least, not treated gently) with the impact drivers during expedited pit stops.

In light of the above, comment was: "This has to be a batch problem"...?

deputydog95

What about these nuts? This is what I just put on my car. They seem to fit fine.

http://www.race-studs.com/servlet/th...-Series/Detail

Boinkus44

That's what I've been using. No issues with fit and that's with 4 different sets of wheels.

FFaust

Same here, except that I only have 3 sets of wheels

NJ-GT

If I'm shown these pictures, my answer would be: this is what happens when you move your American manufacturing to China. Product looks the same, quality not so much.

Time to call TC Kline.

powdrhound

According to the Factory Porsche 996tt/GT2 maintenance manual, the wheel bolt threads MUST be lubed prior to installation and torquing to 96ft/lb. The verbiage is copied below from page 776 of the manual:

"Always apply a thin coat of Optimoly TA (aluminium paste) on the thread of the wheel bolts, on the shank and between the bolt head bearing surface and spherical cap ring (under the head) −arrows− . Do not grease the spherical cap bearing surface −X− of the wheel bolts that faces the wheel."

I know this applies to wheel bolts but I would imagine the same would apply to studs/nuts.

deputydog95

I've got 4 days worth of DE's on my stud conversion. So far, I'm really pleased. I used blue loc-tite (per the directions) and hand tightend till they snugged down. I had a friend hold down the brake pedal when doing the fronts. Made that part a little easier than try to hold it from spinning with two wrenches.

They don't look pretty, but they sure do make wheel changes easy...

Rob S

I thought I'd revive this thread, even though there have been several since this one. This one had some good and spirited debate about what may be causing wheel studs to fail on GT3s.

After previously pontificating from my keyboard in this thread, I, too, have now suffered the dreaded wheel stud failure. It happened late this summer on my 04 GT3 track-only car. After a day at the track, I went to remove the right rear wheel and there was less resistance to loosening torque than I expected. When I tightened it, the stud broke off -- right where others have reported it. It broke at the first thread into the wheel carrier. Pics are below.

Some facts (and unknowns). I *think* these are H&R studs. If anyone can confirm that from their experience, that would be appreciated. I did not buy them or install them -- that was done by the previous owner. A well known race shop did the work and (to my knowledge) maintained the car. That was four seasons ago. I've had the car three seasons and have run a total of about 12-15 track days myself since purchasing the car. I use OZ wheels on the track and torque them to 96 ft-lb religiously. I'm certain that during my ownership, I've never failed to torque them properly. I don't lubricate the threads or the nut surfaces (though there is likely some residual lube on the threads). I have been known to retorque them between sessions, though I've rarely found that they have loosened much or at all. The tracks I run are all counterclockwise, so the right rear wheel, where the stud failure occurred, would be most heavily loaded. I haven't removed any of the other studs yet to inspect them, and I haven't found any that don't hold torque so far. I ran one more track day after this failure by extracting the broken stump and using a factory wheel bolt in its place. I'm not going to run again until I come up with a solution -- which will involve removing and discarding the other 19 studs and either replacing with studs or going back to factory wheel bolts. Of course, I'm looking for the best solution.

Here's what I'm sure of: the failure is clearly caused by bending fatigue. It's a classic, handbook case. The crack began on one side and progressed over 90% of the way through the material before it failed. The lip on one side was the final fracture zone. It's all that was holding it. The failure took many thousands (perhaps hundreds of thousands) of cycles. It's remarkable how little of the stud was left before it broke. That is an indication that these wheel studs operate at a fairly low stress level. If the crack had only gone halfway through before failure, it would indicate a much higher operating stress.

Here's what I think is true: The failure was not due to overtorque -- at any point. There's no indication of a crack that was initiated in that way, and if it had, I would expect the failure to happen on the spot or very soon thereafter without such obvious evidence of very-long-term fatigue damage that could only have accrued from being run on the track. I also have no particular reason to think that the material used for the stud was inferior in some way. It would take some workup to confirm that, but I doubt that the choice of materials would be the biggest factor here.

Here's what I don' know: why did it fatigue? I *suspect* that the most likely reason was that the lug nuts were not adequately "torqued" (or more formally, the studs were not adequately "preloaded") on a regular basis. But how can this be? I used an accurate torque wrench and never operated the car without checking the torque. Could it be that someone before me (four years ago) failed on some occasion to torque the nuts? Could be, but I'd be surprised if a misstep that long ago would take this long to surface, especially given the long-term damage we see.

Could it be that 96 ft-lb is simply an inadequate amount of torque, when using studs, for the GT3? The factory wheel bolts have a two-piece head -- a spherical part that bears on the wheel (and likely stays stationary during torquing) and a flat washer-like bearing surface that allows the bolt to rotate, likely with less friction. So with factory bolts, maybe the percentage of torque that actually preloads the bolt is higher than with studs and nuts, and as such, the 96 ft-lb spec is adequate for the bolts, but is too low for the typical stud/nut system?

This could all be figured out formally, but it would be a lot of work to do. Any ideas on whether this might be the reason that people have been having wheel stud failures (but not bolt failures) on GT3s? If so, that would explain why different wheel stud brands have failed. If this theory is true, then it's possible that to keep studs alive, you should torque the nuts to a substantially higher level than 96 ft-lb. I don't know a number in mind yet, but I'm thinking something closer to 115 ft-lb might be a good starting point.

It makes no sense to me, especially after seeing how low the stress level is in these studs, to replace them every year or two if they're fatiguing. They should have an infinite life if their preload can be maintained.

Burger

Maybe the stock bolt are two piece because ball seat centering isn't a good way to do it, and the two piece setup allows for some variation in centering without stressing the bolt from not being perfectly in line with the hole?

rbahr

Stating the obvious, but torque them cold...

FFaust

This, or there is some kind of rotational force being applied on a stud v. a bolt with a "floating" ball washer?

My cousin Marc, in Montreal, has an 04 and uses stock-style bolts (a bit longer so OEM?): He broke 4 or 5 BOLTS this summer. Now, they were many years old, but it was revealing/shocking to learn that bolts too could break.

Failure mode has been the very same on every broken stud, or bolt, that I've seen. Broken parts all look the same: First thread, signs of longish-standing fracture, etc.

I wish someone could figure this out. Just saying that this is a maintenance item, which should be replaced regularly doesn't say much...

Rob S

All,

Thanks for your thoughts.

Francois, so your cousin broke bolts, eh? That's the first I've heard of such a thing. And 4 or 5 of them? Factory parts? On the same wheel? Do you know the circumstances? I'd be interested to know more. Perhaps he could post about what happened (or maybe one of you could send me a PM if so inclined). If failures were at the same location as we've all seen from studs, it sounds again like fatigue, which would typically be because of under-torque (or more properly, under-stretch). That's possible, of course, but per my theory it's "easier" to unwittingly under-torque nuts/studs than bolts because the factory bolts have that built in "bearing" that converts more of the applied torque to usable clamping force (instead of wasting it in friction).

Like you, I don't buy that studs should have to be considered frequent maintenance items. I'm not suggesting that it's not advisable to replace them periodically, but if they're regularly breaking from fatigue, then that's a problem that should be solvable. And a greater amount of torque may be the simple solution.

I heard a rumor that the newest Boxster has a torque spec of 115 ft-lb. Can anyone confirm? Of course, I don't know anything about the bolt/thread size, but if Porsche is upping their torque specs on stock wheels, it would be worth investigating why. Anyone know what the spec is for the 991?

Rob

FFaust

I will try to get Marc to chime in.

I am no metallurgist, so I find it odd that you mention under-torque as a possible cause. It certainly is counter intuitive.

I recently broke a rear axle, and the break looks very similar to what is happening to our wheel studs. Odd thing is that it broke within the wheel carrier, within the splined area, at a location that you would think is well supported and not as susceptible to breakage. I would be curious to hear your thoughts. PM would be fine in order not to derail the thread, unless the community was interested.

Here is what it looked like

powdrhound

Rob,
How did you manage to extract the broken stud from the hub? Seems like a major PITA especially since the studs are lock tighted in there.

Rob S

Francois,

To your thoughts about under-torque being counter intuitive:

I'm not a metallurgist either, but several of my colleagues are. I'm a mechanical engineer. I agree that at face value, you'd think that a broken bolt or stud would be the result of too much torque, if anything. And that could be the case, of course. But when the fracture surface looks as it does in my post (as well as the photos of the OP, though it's a bit harder to see), we see a progressive fatigue failure.

Fatigue is caused by the cycling of load (bending in this case) of the stud such that a crack starts at some location (a thread root in this case) and grows a small amount with every load cycle (each rotation of the wheel). The scary thing about fatigue is that it *doesn't* require some big traumatic event, and the stud fails at a normal load, over long (and undefined) period of time. It can be very hard to find or detect a growing crack of this type prior to catastrophic failure. So fatigue is generally very bad in an engineering design. The key to avoiding fatigue is to avoid cyclic loading of the damaging variety. In this case, that means cyclic bending. One way to avoid cyclic bending is to torque the wheel fasteners sufficiently so that they can't ever bend to a damaging degree. In most cases, a wheel stud that fails in fatigue did so because of insufficient "stretch" or preload in the bolt, and it suffered too much cyclic bending stress. Since we can't easily measure bolt stretch on wheels, we use torque as a substitute. We figure if we torque it to some published spec, it will not allow such damage to occur. But what if the spec isn't right (it's too low)? Or what if the clamping force from the torqued connection is lost because the clamping force diminishes due to something compressing or stretching in the joint? Or what if there are some other factors that prevent the published torque spec from achieving the goal: adequate bolt stretch to prevent cyclic bending?

Though material factors can also be important (some materials are more fatigue resistant than others), and manufacturing techniques can be important too (rolled threads tend to start cracks less readily than cut threads), I think the biggest factor is more basic: torque level. That's why I think that with all the fatigue failures we're seeing, on studs by all sorts of manufacturers, the most common thread (no pun intended) is that they are under the optimal torque. even if they are torqued to the Porsche factory spec of 96 ft-lb (which, incidentally, for a GT3 is for special bolts -- not studs with nuts).

I will PM you about the axle failure.

Powdrhound,

I removed the rear rotor (which required removing the rear caliper first), then I drilled out the center of the broken section of stud and used a stud extractor. I also used heat from an oxyacetylene torch to loosen any threadlocker. There are at least two kinds of stud extractors, I chose the kind that has a tapered rectangular end on it, not the kind that has reverse threads on it. The stud came out with little resistance, but I didn't know how hard it might be. The worst thing would be to do break off the stud extractor in the hole! What struck me was that if you suffer such a stud failure at the track, good luck fixing it there. I generally don't go to the track prepared with this much equipment. And I'd *never* run on just four studs. It made me nervous running even one track day on the remaining four studs and a bolt. So it's really not a good thing to do in the field unless you're very well equipped. And all it takes is one broken stud to shut down your driving weekend.

Rob