5 sheared studs and the wheel falls off...

[GTTuner]

OK, well usually running with loose lugs you will have elongated stud holes and you'll see threads inside the holes also.

 

Does not look like the case. Ever have any aftermarket wheels or wheel spacers on that car?

 

Call your insurance company, comprehensive coverage will cover that.

[rob]

Yes I run aftermarkets for winter wheels for trips up to the ski hill and across the mountains. I would say maybe 15-25% total miles driven. Now that I think about it, there was one particularly hard impact from a pot hole 2 years ago that I *think* was on that corner. It was hard enough to dent and crack the winter wheel. I also should mentioned that I do have a utility trailer and tow up to 1000lb (trailer limit) once a year. I guess those factors could add up to 1-2 stud failure.

 

Yeah not so much worried about insurance coverage. More worried about defective parts. My wife doesn't want to drive the car anymore after this....

[GTTuner]

Are the aftermarket wheels "hub centric"? Subaru centers the wheel on the center hub, which also helps retain the wheel on the hub. If you using "lug centric" wheels you might have stressed out the studs.

[rob]

One set I think so, the other set not so sure, but I have only run the new set a few times this past winter. Is this a common failure seen with running lug centric rims for a prolonged period? I haven't heard of anything this severe, but I really haven't been looking until now. An off center wheel will have some amount of vibration that I should notice.

[GTTuner]

Well the lugs will center the wheel but the hub takes alot of the forces from the wheel away from the lugs. Just a theory,..... FWIW I would never trust a lug centric wheel on a hub centric car.

 

Replaced all the studs to ease your mind and get some centering rings for your aftermarket wheels.

[SubLo]

I had this same concern on my Ford Explorer when I bought "lug-centric" (no such thing really but that's going to incite an argument) aftermarket rims. I bought custom centering rings from here to hold the load properly and give me peace of mind.

 

http://www.motorsport-tech.com/hub_rings.html

[BAC5.2]

Is that bits of aluminum on the tips of the studs?

 

I have a hard time classifying this as a case of fatigue. Why? Because bulk materials are so consistent in terms of microstructure, that manufacturing errors are almost unheard of. Subaru almost definitely sources lug studs from another manufacturer (the same manufacturer that most other asian makes likely source from). Those manufacturers cut lug studs from wire spools thousands of feet long, knurl the ends, then die threads on. Then the lug studs are heat treated in batches of, likely, more than a thousand units at a time. For 5 studs to be manufactured incorrectly, an entire batch would have to be affected. The lugs might have been marked for batch, but there is no guarantee that all 5 lugs in your hub came from the same batch.

 

The statistics of this are mind boggling. It is clearly not a design flaw.

 

Properly torqued, there really isn't any danger of running lug-centric wheels. We are talking about shear, and to shear a single 12mm grade-8 bolt, you are looking at almost 16,000lbs per bolt... in shear! That is a LOT. So long as the wheels were installed using conical lug nuts (which center the wheel over the lugs, and increase the effective clamping area over flat bottom lug nuts), they should have been fine. Lots of cars are designed to use lug-centric wheels, remember. I think "lug-centric" is a misnomer though. Having the lugs carry the whole of the forces doesn't make them "lug-centric".

 

I don't see any evidence of any foreign contact that caused the breakage. The cause for failure MUST have been some introduced flaw to the lug stud itself. They were left loose at one point, overtorqued at one point, cracked, something. Maybe the lug nuts themselves were faulty or cracked? The lug nuts only need to be loose-ish for bad things to happen. If the lug nuts themselves were cracked, it's possible you had a false torque reading.

 

That BIG pothole MIGHT have done it, if it destroyed the wheel it could have damaged the lug.

 

I'd replace all the studs on the car. OE lugs are less than a dollar each, I think. Personally, I'd go to ARP studs. I think they are around $6 or $7 per stud, so $120 to $140 for a set. You'll need open ended lug nuts to go with them, but it's a small price to pay for piece of mind.

[rob]

Thanks for the insight BAC. How many lugs do you think it would take to fail before the rest gave out? I would think at least two would have to be compromised for three others good ones to fail? The car was loaded pretty good. And to the comment on hub centric wheels, I don't think the hub takes any of load. But I would like hear your opinion.

 

-Rob

[SubLo]

Well, this is the "lug-centric" argument I was trying to avoid. In referring to that term, it's in resepect to centering not load bearing. I agree with most of the other points. BTW, many heavy duty trucks do not use hub-centric wheels and hub loading but rather, yes, carry the entire load on the lugs.

 

Anyway, I believe there is indeed a danger of running non hub-centric wheels on a car that was not engineered for it as opposed to something like the trucks I mention above - properly torqued or not. There is a reason the engineers designed it this way.

 

I agree it's not a lug design flaw but discount throwing fatigue out the window as a possibility. The sheer strength of something like Grade-8 lugs is indeed astronomical particularly at a point-in-time which is why I brought up fatigue in the first place. Over time, substantial load (read: trailer, many passengers/gear, etc. + the use of said wheels) , extra angle/camber, I suspect fatigue does become a factor. That's true of any grade metal.

 

My 2 cents as a mechanical engineer...

 

<flame on>

[rob]

This isn't based on any empirical data (I am an EE after all and my ME friend says "shut up sparky" quite frequently ), but it seems to me that fatigue from above average, but what I would still consider acceptable usage should never result in this kind of failure. Am I off base?

 

Again back to my question if indeed I damaged one or more lugs somewhere along the way (giant ass pot hole being the most obvious ), how many good lugs could instantly handle the full load of the car? My gut feeling is four should be ok, three maybe not.

[SubLo]

This isn't based on any empirical data (I am an EE after all and my ME friend says "shut up sparky" quite frequently ), but it seems to me that fatigue from above average, but what I would still consider acceptable usage should never result in this kind of failure. Am I off base?

 

Again back to my question if indeed I damaged one or more lugs somewhere along the way (giant ass pot hole being the most obvious ), how many good lugs could instantly handle the full load of the car? My gut feeling is four should be ok, three maybe not.

 

You can dangle your whole car in the air via 1 lug in a sheer or strain setup and it would hold up. But that's a static case. You should have been fine with four for a long time excluding other factors. Something else was at play. I once drove an old Buick from Buffalo to NYC as a teen on 2 lugs (I would not recomend duplicating this effort ).

[LosAngelesLGT]

I once drove an old Buick from Buffalo to NYC as a teen on 2 lugs (I would not recomend duplicating this effort ).

 

Wow and I rolled the LGT out of the driveway with lugs on one wheel only hand tight and heard "whump whump whump".

 

I kind-of get the feeling that the lugs were removed/loosened and as soon as the wheel jammed up into the well. The still moving hub snapped the studs against the back wheel... Is that possible?

[torinalth]

based on where they all sheared... I think its metal fatigue. the guy, in 1 year, torques on the lugs more then I did in my entire time owning my WRX for 3.... like any bolt you torque too many times... eventually its going to stretch a bit... stretch too many times and it breaks...

 

personally I'd call it a win that nobody was hurt and go ahead and replace all 20 just to be safe (cheap insurance in my opinion)....

 

in the end its never going to be 100% on the cause, just hope it never happens again

 

T

[dr_sharp]

I think there's a bit of a misconception on fatigue.

 

Fatigue is not going to occur from removing and replacing the wheels a few times a year. Especially if the factory torque specs were obeyed. This is an axial force on the stud which is going to have a pretty hefty safety factor.

 

Fatigue is what would occur when the stud undergoes tens of thousands of cycles of loading. When a stud is at the 'bottom' nearest to the road, the shear force is acting away from the center of the wheel, the direction of gravity. When a stud is at the 'top' nearest to the car, the shear force is acting toward the center of the wheel. So as the wheel turns, the direction of the shear force that is acting on a given stud is constantly changing. The direction of this force rotates as the wheel rotates.

 

Now consider a car traveling at 80 mph on a highway and count how many times this cycle occurs in 1 minute. It's a lot more than 5...

 

Now take that 1 step further and add the potential problems of 1 or more loosened lug nuts or the lug-centric vs. hub-centric wheel and the shear forces acting on the studs could be much higher than what they were designed to see, even with a safety factor included.

[BAC5.2]

Yea, I don't think people throwing "fatigue" out really understand it.

 

Properly torquing a bolt keeps the metal in the elastic region of the stress-strain curve. That is the ONLY way to get maximum clamping force. You torque a bolt up to just below its yield strength (torque-to-yield is a common term for things like head bolts). Stay below the yield strength, and you can torque and re-torque the bolt as many times as you want without worry of failure. That's how bolts work.

 

Components like lug studs, drive shafts, almost anything that sees cyclic loading on a regular basis are designed with a similar fatigue limit. That is the idea of strain-life. The generally acceptable practice is to design for 10 million cycles. At 60 MPH, with a 215/50/17, that's around 10,500 miles. If it hasn't failed from fatigue in that time frame, there is a HUGELY small chance that it will fail due to fatigue. THAT is why I rule out fatigue, in a traditional sense. I don't believe fatigue is a factor for a properly torqued, good condition bolt.

 

The lugs must have been over-torqued at some point. I wonder if your torque wrench is out of calibration. I had a craftsman wrench that went out of calibration almost 10 ft-lbs once!

 

The lug-centric issue... isn't an issue. In a fully torqued scenario, the snout on the hub STILL doesn't carry primary loads. The lugs do. In a Subaru, where the snout of the hub does intrude into the wheel, still the lugs carry the primary load. If this weren't the case, then there would be massive scoring on the hub protrusion on Robs car. On centering, the lugs do that as well. The conical face of the lug nuts force the wheel into a specific location. The hub protrusion, literally, does nothing more than stick into the wheel a wee-bit. All wheels, in a variety of senses, are lug-centric. The hub doesn't center them, the lugs do. The only truly hub-centric wheels I can think of are mono-lug wheels like those found on the 911 Turbo, and the myriad of other supercars, where there is little else but the hub itself to center the wheel. The lugs always carry the load of the vehicle. The pictures in this thread prove that, case in point. There was no scoring or contact between the hub protrusion and the wheel here. If the hub were carrying the load, this wouldn't have happened.

 

As for what I think about the "how many until domino" effect, it's tough to say. I drove my old Legacy (my 95L) with only 4 lug studs on the front wheel. I put a few thousand miles on it like that, and didn't have any issue. I'd say as soon as two adjacent lugs failed, you'd be in jeopardy of having a problem.

 

Another possibility is dirty lugs. If the lugs had grime in them, and you went to torque them, that could give falsely high or low readings. That's how I broke the lug off of my 95 Legacy. I changed the tire at a tire yard, and a bunch of crap was all over the lug studs. I threaded on the nut, it bound up, and twisted the stud off well before the lug bottomed out on the wheel. Had I been using a torque wrench, it would have read peak torque, and I would have left it. It would have been loose, and the same problem would have eventually occured.

 

Yes. I think that the lugs have once, or multiple times, been over-torqued. That is the most logical explanation. Maybe a quick-lube place or someone at some point used an impact gun to tighten your lugs.

[proflatlander15]

I like this discussion. Reminds me of school.

[RochNY91TSI]

I like this discussion. Reminds me of school.

 

+1 Good discussion!

[Max Capacity]

someone at some point used an impact gun to tighten your lugs.

 

God no, I hope that wasn't done.

 

I hope we don't have to get into a training class on how to install lug nut's.

[SubLo]

Again, this lug-centric/hub-centric discussion gets ugly real quick and facts are quickly distorted and misinterpreted. It's like the synthetic vs dino oil, Small Turbo/Big Turbo, FMIC/TMIC, Al Sharpton vs Reality threads - avoid at all costs.

[rob]

Yes. I think that the lugs have once, or multiple times, been over-torqued. That is the most logical explanation. Maybe a quick-lube place or someone at some point used an impact gun to tighten your lugs.

Sure it's possible, every now and then I have had the tire shop put the tires on; as in the case with new tires (which wasn't that long ago). But you are saying that a one time over-torque could cause this to happen!? The number of dumbasses with impact guns working at the thousands upon thousands of shops, you would think if that was the case, tires would be falling off left and right.

 

The lugs must have been over-torqued at some point. I wonder if your torque wrench is out of calibration. I had a craftsman wrench that went out of calibration almost 10 ft-lbs once!

 

Again, I ask; you consider 10 ft-lbs over spec enough to cause this kind of catastrophic failure?

 

I have a pretty decent torque wrench that came with a calibration sheet. Not to say that it's not off, but at 70ft-lbs with a 250 ft-lb top end, it's in the low error range of operation.

 

Another update; I talked to the insurance company. First they were going to send an inspector out today, then changed their minds will be sending an "investigator" out tomorrow. From our description they are leaning toward the "vandalism" scenario. Which is alright with me, as that is covered no problem. If they deem it a manf defect, then they pay nothing and we have to go fight with Subaru. Ugh. Also by checking behind the wheel well cover I can see more damage to the body under where the corner of the bumper mounts.

 

-Rob

[TSiWRX]

someone at some point used an impact gun to tighten your lugs.

 

God no, I hope that wasn't done.

 

I hope we don't have to get into a training class on how to install lug nut's.

 

Sadly, I can't even begin to count the times that various tire shops - and even the dealership - over-torques. In the real-world, how realistic is such an expectation.

[GTTuner]

Again, this lug-centric/hub-centric discussion gets ugly real quick and facts are quickly distorted and misinterpreted. It's like the synthetic vs dino oil, Small Turbo/Big Turbo, FMIC/TMIC, Al Sharpton vs Reality threads - avoid at all costs.

 

 

 

+1. The hub DOES take some of the forces appiled to the wheel. To say it does not is inaccurate IMO but I ain't in the arguing mood today.

[robinlsb]

I need a better pic of each stud. Metal fatigue can be seen where the metal actually crystallizes. If two + studs fail then the others would tear off. Use a magnifying glass on each stud. You are lucky, I lost a front wheel on an RX and went end over end. Damn near killed me.

[robinlsb]

Yeah thanks guys. Indeed it could have been a whole lot worse. Not the best way to spend a sunday evening. No I don't anti-seize the family wagon. After changing several stripped studs, I do use the goo on my Impreza. With rallyx I change tires a lot more on that guy and yes I do use a lower torque value(which does get the lower torque value). On the Legacy the last torque was 70 ft-lbs.

 

And yeah, I am worried about the other rear corner now too. The fronts had wheel bearings changed under warranty at about 50k miles so at least those studs won't be from the same batch (hopefully)....

 

-Rob

 

That could be the answer!!!!!!!!!! Once over torqued, you can't go back. I am sure that the remove torque would not be noticed by you. Yet, if the studs were over torqued, they could have been weakened, setting up the ultimate failure.

 

That would explain the catastrophic failure of all the lugs, more or less at the same time. It would also tell you to replace the front lugs on the other side to be safe.

 

My guess is the dolt who replaced the front wheel bearings used an air wrench set for steel wheels.

 

However, since the bolt ends are so dinged up, I would stick with vandalism, unless the insurance company has an electron microscope.(

[BAC5.2]

Sure it's possible, every now and then I have had the tire shop put the tires on; as in the case with new tires (which wasn't that long ago). But you are saying that a one time over-torque could cause this to happen!? The number of dumbasses with impact guns working at the thousands upon thousands of shops, you would think if that was the case, tires would be falling off left and right.

 

I would peg this as a major possibility. MOST shops use torque-sticks when doing lug nuts with impact guns. It's usually the newbie who hasn't actually taken any auto-tech classes that makes the mistake. Impact guns can do up to 650 ft-lbs now-a-days. An uncalibrated finger can "butta-butta" a lug nut to over 100ft-lbs without even knowing it.

 

Again, I ask; you consider 10 ft-lbs over spec enough to cause this kind of catastrophic failure?

 

I have a pretty decent torque wrench that came with a calibration sheet. Not to say that it's not off, but at 70ft-lbs with a 250 ft-lb top end, it's in the low error range of operation.

 

Using a 250 ft-lb torque wrench on a 70 ft-lb bolt is less than ideal. You want your expected torque values to fall close to center on the wrench range. I'd check your wrench. It might only be calibrated from 100 to 250 ft-lbs.

 

10 ft-lbs might be enough. I haven't done bolt-stretch calculations in a while. It only takes one time exceeding the yield strength of a bolt to ruin it. Period. One dumb newbie zapping a lug to 100ft-lbs spells failure. Once you yield the bolt, all stability is lost.