Some things are easy to screw up if you’re not an experienced wrencher. Nobody would advise a novice to attempt a full engine rebuild, or to recondition their own high-pressure diesel injectors. But changing your own wheels, that’s easy, right? Oh, but it can go so wrong.
Let’s jump back to 2007, in the nascent days of video on the Internet. We see an FD Mazda RX-7 tackling an autocross course in anger. It’s the best rotary road car ever built, but it’s about to be in trouble. Fair warning, this is possibly the worst quality video you’ll see all year, but the events within are fantastical.
The RX-7 jinks right, slides left, then jinks right again. And all four wheels spontaneously eject themselves from the vehicle. Wait, what?
The video was shared far and wide over the years, usually with a minimum of context. Mockery ensued, with the braying masses insinuating the owner forgot to tighten down their lug nuts like a gosh-darned fool. But that’s not what happened here.
How do we know? Quite simply, the video tells the story. If loose lug nuts were the cause, we’d most likely see individual nuts bouncing free and more gradual wheel shake before the wheels came off. But that’s not the case. Instead, we see a near-instantaneous ejection of multiple wheels, suggesting the failure of multiple lug nuts at once.
The best insight we have comes from a NAISOC forum post dated December 12, 2007. User go_speed_go indicates they were a witness on the scene.
Hello all,
I was at the event and know the guy who owns the car. (No it wasn’t me)
I was in the paddock area next to him changing wheels at the same time as him. I watched him torque all of them. I even borrowed his torque wrench to tighten mine (I always forget something after I get on the road).He went to a wheel/tire business and asked for spline drive lug nuts for his particular make/model/year car for the Enkeis, and that’s what they gave him. He did not purchase the tires or wheels from them, just the lug nuts.
The lug nuts where sheared off the studs.
This theory lines up with what we see in the video. The shearing of multiple lug nuts at once would allow a wheel, or four wheels, to suddenly pop off like that.
But why did the lug nuts fail? That’s not entirely clear, but the prevailing theory is that the wrong nuts were used for the thread on the RX7’s hubs. As standard, an FD RX-7 has M12 metric lug nuts with a 1.5mm thread pitch. Perhaps the most likely explanation is that a set of 1/2-inch threaded nuts may have been supplied instead. These would potentially have been close enough to thread up, though their lack of proper thread engagement would allow them to shear off quite easily.
Let’s examine the spec of US-style 1/2-20 UNF threads. 1/2 refers to the 1/2-inch nominal diameter. We’re talking about lug nuts with an internal thread, so the basic minor diameter is o.446 inches, or 11.3 mm. That’s the smallest inner diameter of the thread. The maximum allowable outer diameter of the thread on the M12 stud is 11.97 mm as per international specifications. Between a 1/2-inch UNF lug nut and an M12 x 1.5 stud, at best 0.02 inches (0.67 mm) of thread would engage. No wonder the lugs stripped so easily! The proper combination of metric lug nut and stud should have a full thread engagement of over 0.04 inches (1 mm)!
As for pitch, the -20 part of the UNF spec refers to the fact there are 20 threads per inch. This equates to a metric thread pitch of 1.270 mm. 1/2-20 UNF lug nuts are close enough to M12 x 1.25 mm threads that they’ll readily thread on, even though they don’t quite mate properly due to the slight difference in diameter. However, as mentioned above, the FD RX-7 uses M12 x 1.5 mm threads. The difference in pitch of 0.009 inches (0.23 mm) might not sound like much, but it would normally lead to the lug nuts binding on the studs in short order.
Basically, the pitch is different enough that the owner should have realized something was wrong while fitting the wheels. As per the forum post from a witness on the scene, though, the owner apparently didn’t notice anything was wrong when torquing them up.
I’m not a mechanic, but the mechanics theories there was that the lug nuts where the incorrect size, but they were just small enough to grip the outer edge of the threads (as I mentioned, they torqued to 75-80 ft/lbs without complaint).
We’ve all made mistakes, though, and the M12 vs. 1/2-inch confusion does seem the most likely theory for why the lug nuts suddenly stripped off. It’s the most plausible explanation currently on hand. If we had higher-resolution video of the aftermath, we might be able to see marring on the wheel studs which would shed more light on the matter. The fact that the failure happened to all four wheels at almost the same time is a coincidence to some small degree. Regardless, when the wheels were loaded past some basic critical point, it’s no surprise the lugs popped off.
Thankfully, the story has a moderately happy ending. According to go_speed_go, the car wasn’t a total loss.
The lug nuts where sheared off the studs. Several members lifted a corner of the car to slide a jack underneath it. With a few more jacks, blocks of wood, etc, they were able to mount the stock wheels and stock lugs nuts (at least 3 per wheel) back on, torque them and drive the car off course. The design of the car, possilbly for aerodynamics, tucks all mechanicals on the underside a little higher than the body line, or lowest part of the brake rotors. Not even the exhaust was really scratched.
Hopefully, the only damage is to the body and brakes.
Ultimately, this old video is a sage lesson to us all. Pay close attention to what you’re doing when you’re torquing down your wheel nuts. They should thread on smoothly, and you should never have to force the threads. A cross-threaded wheel nut, or one of the wrong size, can lead to ugly surprises like these.
Image credits: topwriter2 via YouTube screenshot
I had a similar failure on my r53. The tire shop that installed the wheels used too short of bolts so when my brother’s and my fat hawaiian behinds tried to take a corner in it the driver’s side front wheel came off and bent the hell out of my hood.
High-res video is available to use today, so I say the Autopian should test this Mythbuster-style!
Replicate the conditions, affix a camera pointed at each wheel, and let ‘er rip!
That’s a job for Garage 54!
Somebody ugga dugga’d the wrong ones on really well
I once went to a Jeep dealership in Rochester Hills, MI for some service and a tire rotation. On the drive home, something felt off. Couldn’t quite pinpoint it, but the steering felt loose. The car was wandering. After a mile or so I pulled over to investigate and there it was: all four wheels were about ready to fall off. All 20 lug nuts were only hand tightened. Bonus: there was there was metal shavings all over the studs as the wheels had been sliding up and down the lugs and filing the holes into an oval shape. Took it back to the dealer and they nonchalantly blamed the new guy without so much as a care in the world. What a joke. Could’ve been a disaster had I not pulled over before it forced me to.
Classic vid. I came close to losing a wheel once (heard the groaning and grumbling of a loose steelie) and immediately knew the culprit was the tire shop I JUST LEFT FOR THE SECOND TIME after I bought snow tires. (first time they installed 195/70/14 instead of 195/60/14 and I rubbed over every bump on the way home with a lowered Jetta)
Hey its not rocket science.
Oh wait –
https://www.simscale.com/blog/nasa-mars-climate-orbiter-metric/
There could be more explanations to dig through. My company experienced bolt thread failure on about 250k bolts recently. We sent some bolts off for analysis and found the surface of the threads were decarburized. That’s a process where the carbon burns off when the threads are overheated during the cutting process, and you end up with a soft, low-carbon steel. It only occurs hundredths or thousandths of an inch into the thread surfaces, but makes the bolt useless. Our specific bolts were only able to hold 20-50% of the rated torque before failure.
Something similar could have happened to those lug nuts resulting in threads that couldn’t take full torque without being damaged. Then, when loaded in a turn, threads begin to give up until one nut pops off and begins a snowball effect around each wheel and from wheel to wheel. Also, this can happen to any manufacturer if they do not have a specific program to inspect batches for decarburization. This would be a manufacturing defect.
I’ve also seen bolts that are not rated to cinch down components get damaged by 4 or 5 extra loaded turns during the torquing process. The bolt threads are damaged by the act of pulling components together, and they fail before they reach full torque, or are heavily damaged when they reach full torque. If the lug nuts were not capable of pulling the wheel to the hub, then that could have caused the failure too. I’d classify this failure as a manufacturing defect.
♫ You picked a fine time to leave me loose wheels ♫
With four hungry children and a crop in the field
If that’s the case, he might not even have noticed a difference in required socket as 19mm is effectively the same as 3/4″ and odds are good that’s the size it normally used. He definitely should have felt it when tightening, but there went the wheels, so apparently not.
I had a 2000 Nissan Frontier with 6 lug wheels. I was taking the wheels off, and 2 or 3 studs snapped on every wheel. Some dipshit at the tire shop gave the lugs one too many ugga duggas. I had to drive back to that shop with like 3/5 of my lug nuts to have them fix it.
Root issue: The owner was an idiot for using incorrect lug nuts.
True, but if US customary units and car parts measured in these units didn’t exist, it wouldn’t be quite as easy to make that mistake.
If rat poison looked, felt and smelled almost exactly like candy, it would still be your fault to eat rat poison instead of candy, but it would still be safer for rat poison that could be mistaken as candy to not exist.
And thus, US customary units are similar to rat poison and is the main reason why the US should switch to metric.
:-p
🙂 Well, you could say that I guess.
I actually meant that lug nuts and wheel bolts made to US customary measurements that are so similar to the correct ones to metric measurements are like rat poison to this Mazda.
The US and 2 tiny nations elsewhere are the only countries in the world using inches and pounds. A great deal of the trade goods the US produces and ships elsewhere are made in America using metric in their manufacture. Significant portions of US manufacturing and trade good have been metric for a long while
My mom also had a wheel come off. A “professional” shop had done something to damage the lugs/studs and ended up welding on the wheel(s)!!!!! (She was on a slower gravel road, thankfully.) No wonder she was paranoid of garages. Same with oil changes; claimed one charged her for oil never changed, though I wonder if the engine was just dirty enough to darken the oil quickly (that happened to me once.)
Welding
WELDING
What???
Had near miss about 15 years ago, felt a wheel vibration the day after having tires replaced. Stopped and found ALL 5 lugs on left front finger tight. Tightened with the 4 way and retorqued all 4 wheels as soon as I got home. I now loosen and retorque all lugs as soon I get home from a repair center.
This is standard procedure with us now.
Just good common sense.
If anyone is in the market for a torque wrench, I love my Icon split-beam and it can be had relatively cheaply with an HF coupon.
Easy to set, read, you can leave it at its torque setting indefinitely, and it has a nice solid click to let you know you’re done.
And the main reason I’ll take all my wheels off the car, take them to the tire store for new tires, and then bring them back and put them on myself. If we’re in a jam and can’t do it like that I will always unbolt and reinstall afterward. I’ve seen these jokers just send lugs home at 400 ft-lbs with an impact and then come around and “torque” them after that. Click, click, click. All set! That’s not how any of that works. Ugh.
It’s a Mazda. They fell off because something rusted 😛
All this could have been avoided if US Americans didn’t insist on using customary measurements for car parts and used plain metric like everyone else.
Better alternative: the rest of the world stops using the metric system and starts using US customary units.
That is not possible because without metric, the US customary units would not exist, as they are only funny multiples of metric (or SI) units. They are literally defined as such.
Your statement makes no sense. The Metric system was invented in the Eighteenth Century, but the “English” system goes back so far that it’s a mix of Saxon and Roman standards.
See, in the US, we reject Johnny-come-latelys. Or, I should say, “Jean” come latelys.
It makes perfect sense if you know the facts.
The customary units that the US use today are not related to the old English system at all. They only share the names, the sizes relative to each other (mostly — cough pint cough), and the approximate dimensions, with them.
An inch today is legally defined as 25.4 mm. And all other US customary units are also legally defined as multiples of SI units.
US customary units have been defined as multiples of SI units since the 19th century. (Even if an inch was slightly different from what it is today, it was still a multiple of a metric unit.)
That is the reason why the US units are “customary” and not “imperial”. The US aren’t using the “imperial” units (which were defined by law in 1824) anymore; haven’t for more than a hundred years.
I can choose to define an inch any way I like. I can say it’s 1/12 of a foot, or 1/36 of a yard. I can also say it’s 1/63,360 of a mile.
And speaking of miles, the majority of the United States was surveyed using this system. A rod is 16.5 feet, a chain is 4 rods (66 feet, the standard width of a residential public right-of-way in the United States).
80 chains is a mile, and of course every standard section in the survey system is one square mile, and every standard township is 36 square miles.
All of this was worked out by GW and TJ and BF and their wig-wearing buddies while they passed the Northwest Ordinance of 1798, one of the most important pieces of legislation in American history.
So back in the 1990’s, the federal government got “metric fever” and decreed that all public infrastructure projects with federal funding be designed and built using metric measurements exclusively. This caused massive problems both on the design end and in the field for construction. It only took about a year for the feds to get wise and take it all back.
In sum, the Metric system was the solution to a problem that no one- excepting a few very uptight French scientists- had.
Didn’t those same French people try to make a 10 day week and a fixed 30 day month and other weirdness?
I don’t think we’d use inches and pounds if we were designing a system from scratch today. But they work fine, and there’s nothing inherently more rational about the metric system. It’s all humans vainly trying to create order and a shared “language” out of the cosmos.
Yes, it’s very important that I measure the picture layout on my living room wall in relation to the wavelength of certain chemical compounds when they are incinerated in a vacuum.
That is only partly true. Most human languages (English and my native language German among them) count in powers of ten. It makes sense to align the measuring units with this. The customary system with its many different multiples between units is just bonkers.
Also, basing units of measurement on natural phenomena and the dimensions of the earth is inherently more sensible than on parts of the human body. Yes, I know that is not the current way SI units are defined, but that’s the idea behind them. Water thaws and boils at (almost) the same temperature for everyone.
The idea of a ten-day week and a hundred-minute hour are not inherently bad. They just didn’t stick.
I agree that the relationship between units is more logical under SI, but I don’t think there’s anything more rational about the length of a meter or the weight of a gram vs a foot or a pound. They are all ultimately arbitrary.
Oops correction, the NW Ordinance was 1789. I hate dyslexia.
Excuse me, but no. The metric system was the solution to a very real problem that more or less everyone had. At least everyone who wanted to trade goods.
The traditional units were not properly defined anywhere. A foot, one of the oldest “units of estimation” (you can’t really call it “measurement”), was … well, the length of a foot. My foot? Your foot? That’s fine if you want a rough estimate, but for a precise measurement, that is no good.
Being a bit radical, they decided to start from scratch and create a logical system that also aligns with language (hence the powers of ten … that’s how European languages count). That was good because the traditional units varied so wildly, also in relation to each other. How many feet to a mile? 5000, if you ask the Romans or the Persians. But then suddenly in the time of Lizzy One, the English added 280 extra feet to a mile. WTF?
The English tried to solve the same problem in a different way: by properly defining units with traditional names, in 1824 (the “imperial” units). They defined them in the “Weights and Measures Act”. The idea is the same, but they usurped the traditional names and gave them a new, exact, meaning. But they tried to solve the same problem.
You’re not wrong, and in the unconnected Regency Era, it made sense. Standardization was part of the evolution of business and commerce.
But making up a system out of whole cloth- and tying it to obscure and arcane scientific constants that everyday people never interact with- seems like just the French doing French things, i.e. hating on the Limeys.
There was a perfectly good way to solve the standardization problem- standardize the traditional English units.
The Metric System is the most obvious case ever of “reinventing the wheel”.
The US system is currently defined using metric units for convenience, but there is no reason it couldn’t be defined other ways. The idea that the US system cannot exist without the metric system is not true.
Admit it: you don’t like the US system because you hate Freedom!!!
The current US system is dependent on (and younger than) the metric/SI system. That is a fact. Yes, you could devise another system that defines units of the same name and similar dimensions in another way. But the US system as it is used today is SI units with funny multiples.
And of course “the US of A” and “Freedom” are two concepts that have very little to do with each other. But that is another discussion for another day.
All I know is that China, Russia, and North Korea have two things in common:
1 – They hate freedom
2 – They use the metric system
Guys, guys. The answer is clearly a third standard. I propose one that uses metric units, divided into base 12, base 2, base 5280, base 1000 (don’t forget mils!), base ideal vs. actual where you get to cheat off 10-15% of a measurement, and base whatever the hell a yard is (36? 1760?), of course all depending on what you’re measuring.
I say we just toss it all away and use Whitworth.
Even Mater, the dumbest truck in the movie, knows Whitworth is junk.
Do you work in purchasing at my company?
How many stones does your car weigh?
I don’t even know how many kilogrammes it weighs.
Laden or unladen?
You can take my Freedom-Units from my cold, dead, hands!
P.S. I’m an engineer/scientist and I actually use SI units all the time. The thermostat in my house is even in Celsius. Working with mechanical engineers is the bane of my existence. Around here they are the only legitimate engineers (or are they?) that insist on using US units.
I recently encountered a unit of area called a kilo-circular-mil (kcmil); my head about exploded.
Surely none of this beats acre-feet per fortnight, which my water engineer wife tells me (here in the UK) is a US measure of rainfall.
Gentlemen…
https://www.youtube.com/watch?v=JYqfVE-fykk
No. Not buying it. Just feels like the typical You Tube “look at me shit, even for how old the video is.
Seriously.
That or someone was to stoned to be in charge of the lug nuts perhaps?
Nah, disagree. If this was a stunt they would have actually captured it properly.
On top of which, it’s pretty obvious this is a sanctioned event. They wouldn’t allow such bullshittery.
You are 100% correct my esteemed writer friend. . No offense ever intended.
I have a brother who is a car dealer of some 35 years.In a crappy part of Florida.
Self made millionare but also a total lunatic. He has had over 100 cars stolen over the years and recovered by the cops. Most were so badly damaged that they were worthless when the cops found them. If they still moved under their own power he would pull the lugs off 2 cars at a time and have a demo contest with either his part, or son… Last car able to move wins. Then off to the crusher they would go.
This was also a common thing with his SCCA Miata racing buds. If another competitor was a real turd, it was not uncommon for a set or two of lug nuts to disappear before practice on track. Really. My bro experienced that at least once.
At Daytona. After a broken shoulder from hitting turn two wall, he settled down, eventually.
So I did appreciate your piece a lot. Really all your stuff is great reading.
I tend to be pretty cynical of a lot of stuff these days. Mea Culpa. Thanks again, good work Lewin.
Bought a set of tires and wheels for my truck. 1990? or so.
Driving home at 70 mph on the interstate, a car pulls up next to me and the guy says, “Hey you just lost a wheel dude!”
Pull off, and right left rear is gone…Another guy stops and says wow. I saw where it went. So he helps me find it, a half mile through snake and gator infested grass. 110 fucking degrees out. Fun.
Call the tire shop and they send a wrecker, I had just left the place 10 minutes ago.
Of course the idiot installer did not install lug nuts on that one. Still sitting on the rack. WTF?
Tires and wheels are not brain surgery. Seriously. Soon after this shop started to require techs to video tape when installing wheels.
To paraphrase Mr. Gump, “shit happens.”
Even documenting the installed lugs with a photo and attaching it to the invoice is a good idea.
Every lug on each wheel was finger tight. I still can’t believe nothing worse happened. I was about 5 miles away from an interstate raised bridge that is eight miles long. Was already planning to test the tire balance by hitting 100 mph.
Sometimes we just get lucky?
I’m a pretty experienced wrencher, and still narrowly avoided f-ing up something as simple as lug nuts. I did rear brakes on my C30 over the weekend, and put the wheels back on while up on jack stands. I usually spin them on with a 4 way after hand starting them, but just till they make contact with the wheel. Then I lower the car down and torque them down.
Well , while I had it up in the air, I decided to also fix a loose heat shield, and that wound up taking longer than I thought. it was at least an hour before I lowered the car, and backed out of the garage. I was starting off down my driveway to go bed in the brake pads when I thought “Wait, did I tighten those lug nuts?”” Of course, that was the oh sh*t moment, I had not. And they were just finger tight. The rather violent stops of bedding in the pads would have almost certainly loosened them to the point of failure.
2 minutes with the torque wrench in the driveway and I was all good. Whew.
I’ve done something similar… Now, before I pull off wheels, I put the torque wrench in the driver’s seat. Good reminder to torque the wheels before setting off.
Great idea! I’m always looking for ways to help avoid my ADHD from costing me too much money.
After reversing and yanking out my blocker heater wires once, now I wrap the cord around driver side mirror.
I’ve done it as well. Got called in to work towards the end of swapping out an axle. Forgot to torque the nuts once wheel was on the ground. Thankfully, after 5 miles on the interstate, I felt the vibration just before the narrow, no shoulder, winding part of the road to my store. Disaster averted—but I sure felt the fool
I just did this last night!
I was replacing rear brakes, and I just used a 3/8 socket wrench to “tighten” the lugs while it was in the air, probably only 20-30 ft-lbs. (Of course going to torque them on the ground.) I lowered the car, but then got pulled away for a couple of hours. I came back and needed to drive to bed in the brakes, which requires hard stops.
Imagine my horror after the second hard stop when I suddenly, and thankfully(!), remembered that I had never torqued the lugs!!! I gently stopped and got out, all lugs were on and at least hand tight (no tools), and the wheel was seated against the hub, so I drove very gently home (1/2 mile) and torqued them down.
I like the idea (below) from Rippstik of putting the wrench in the seat!
I really thought that the conclusion was going to be that this was a staged stunt and doctored video. But this is crazy.
I always start my lugs by hand and tighten with a normal ratchet before giving final torque with the torque wrench. Helps catch small problems or damaged threads.
Same. I turn them by hand until I can’t; I end up scraping the hell out of my fingertips sometimes, but you just can’t replace that feel.
On my Focus’ spark plugs, I can almost get my hands in there to do the same, but have to settle for holding the socket in my hand instead.
Start with fingers, then spinning a socket on an extension without a wrench. Yes!
Yup, I would only recommend gloves if you’re already doing dirty car work. Saved me a few times when I hit a burr somewhere or a flaking paint chip/rust tries to slice open my finger.
I thought maybe it was like how a skink sheds its tail to distract a predator when frightened. Admittedly, that would be a bit self-defeating in a car race.
In my experience this is indeed self-defeating, although in the grand scheme of things not as high on the list as several other factors:
https://www.murileemartin.com/UG/LWA13/402-UG-Pacific_Northworst_24_Hours_of_LeMons.jpg