Working on a blog post and podcast episode about torque which is something that keeps coming up in the questions we get for the MarginalGains Podcast… I was wondering what entertaining stories you all might have regarding broken components, stranded riders, etc related to torque? I’ve worked some 100 Ironman events and seen hundreds of last minute broken parts, panic attacks and general freak-outs related to the topic, but I’m sure collectively we all have some awesome stories to share!!
My personal story is from back in 1999-2000 when the Campagnolo Record Carbon post first came out, we were making the Record crank for them at Zipp and one of our employees bought the new seat post with a gruppo from Campy. He set it up in a brand new Pinarello and on while adjusting it a week later heard it crack! Nobody had torque tools back then and there really weren’t torque specs per se, so we sent an email to our contact at Campagnolo who sent us back a scan of the little paper pamphlet that ships with every part… the pamphlet had a little line drawing of a seat post with axial crack in it just like what we were seeing, and it said something to the effect of ‘if your carbon seat post develops a crack like this, you have overtightened it’
I since think about this every time we develop a new product and have to do the instruction sheet/warranty info, etc… I think about a little guy in Italy deciding to write something to the effect of ‘if it’s broken, then you’ve done something to break it…’ HA!
I stripped the threads on a cartridge cassette before I knew what I was doing & fortunately was able to get a local framebuilder to essentially drill out the cassette. RIP Tom Teesdale
the problem with trying to idiot proof a product is that we can always design a better idiot.
I once told a friend (new home mechanic) that “hand tight” was good when tightening his stem faceplate. He took that to mean as tight as he could turn the bolt with his hand.(s?)
While I do agree with the sentiment that you don’t “need” a torque wrench and they often aren’t terribly accurate depending on lube/etc… I’ve found most people’s tendency is to go with much more than much less torque when first getting into wrenching and going by feel. It’s a good tool to keep the ham fist under control.
I’ve had the opposite problem - components not staying tight enough when torquing to ‘recommended’ specs. Biggest offender is the seatpost on my Cervelo P2C; the screw holder piece says 5Nm-m but even with ample carbon paste, the seatpost invariably slips downwards if I only torque it to less than 6. Even 7 is dodgy! (And I’m not big - only 70kg).
Same with cassette recommended torque settings - too loose, resulting in cutting into the hub by the cassette splines when I pedal very hard.
Haven’t broken anything yet, but I did take care to NOT overtorque my road bike’s carbon stem and and plates. Those I do to the 5Nm spec with carbon paste - it’s a LOT less torque than I would have ever applied without the torque wrench. (I was wayyyy overtorquing my stem+faceplate when I was using alloy - lucky I didn’t break anything.)
Hey Josh, I’ve really been enjoying these podcasts. Keep up the great work!
One thing I’d like to see you discuss as part of the torque discussion that seems to be missed on many home mechanics is the difference between dry torque and wet (lubricated) torque and how it related to the bike industry; as well as where and when to use thread locking compounds.
My good mate had one recently. Press fit BB86 that he was convinced was squeaking. He went through 5 bottom brackets in a few month span and the creak never went away. Went bat shit crazy. Finally brought it over to me because I was convinced the creaking was coming from the seat post or something else. Nothing we did helped the situation until I observed him remove his cranks and reinstall. He barely was turning the Shimano crankset preload cap and it was resulting in the crank moving back and forth ever so slightly. I grabbed my torque wrench and tightened it to 1 nm and the bike is dead silent.
It was one year in Kona (don’t recall which one but it was pre-2010 as that was when I got my Cervelo), I had my QR TiPhoon & the headset all jacked up, it was loose. Took it to the bike shop on the "Dave & Mark Hill from Iron War on the ocean side of the Queen K…the guy took his allen wrench and worked on it, and did not use a torque wrench. I asked about it & he said he had worked on headsets for “x” number of years and done thousands so his hand WAS his torque wrench.
Needless to say, he cracked my steerer tube. I raced not knowing it was cracked. I know it wasn’t cracked when I put it together & when he took it apart and installed a new expander bolt, because I watched right in front of me when he took the assembly apart entirely–I think it was a Stronglight or Cane Creek. Then when I took it apart to pack up after the Ironman into the bike case, it was cracked. I did not jump off any curbs at all so the fact he jammed down on the stem bolts without a torque wrench may have cost me more than a few dollars had it failed during the race.
Love this one and also love Tom Teesdale, he was a total pro, expert in the field and so early to the thinned lug and dropped seat-stay game, legendary work!
How to know if your torque wrench is still good, or is torquing to a higher value than it’s set for? And can it ne fixed?
I’ve recently found plenty of bolts that seem as tight as I think they should be but the wrench hasn’t “given way” yet. Best 2 examples, just 12 days ago I broke the seat post retaining wedge bolt on my Slice and was nowhere near the torque setting specified in the instruction manual. #2, in the Alpha one thread I was advised the stem cover would hold shape when torqued past 6-7 Nm however mine deforms at 2 Nm.
There are millions of stories , not bike related , where torque was not a good indicator of clamping force. To many variables.
The best one was we had , an operator cross threaded a nut and bolt. Loads of torque, zero clamping force. They were correct , the torque wrench said X but it wasn’t what we really wanted as a result.
Most engineers use a very crude formula to calculate clamping force from torque. It’s basically a big swag hidden behind a formula.
I got a pro wrench. Used it for a few months, no issues. Then i suspect i got lazy, and did not know any better at the time, and began to use it to loosen bolts too.
On day i was tightening a stem bolt and i popped it. A day later i was tightening a seatpost clamp, and pretty much every orher bolt on My pretty nice road bike. a small crack developed in the post.
I stopped, suspected my torque wrench was out of spec, researched that i should not have been using it to loosen bolts, and took my bike apart. I had marked, but not damaged the carbon steerer. Everything else was ok.
I never used the torque wrench again until a replacement of a different make arrived. The pro was torquing 2nm tighter than the new wrench!!! So 4nm was 6 on the pro.
The pro went in the garbage.
Fyi, a good torque wrench is something i would pay a lot of $$$ for. Why risk ruining parts worth hundreds and thousands of dollars, to save 50$ or so?
Lots of great stories here so far!! It has been interesting talking to the industry about this for the piece I’m writing. At Zipp, we would develop these numbers by looking at performance over the range of torques… so for a stem, does the bar slip at 3, 3.5, 4 etc until no slip, then keep going until it breaks something. We would do this for high limit and low limit parts as this makes a big difference, something like a handlebar has 0.2mm diameter tolerance and the stem interface is similar, so we then match up samples as low/low, low/high, high/low, high/high and see where each one stops slipping and then fails… in the end you find a band of allowable torque that won’t slip or break.
Now the challenge becomes making sure torque wrenches are accurate, that the band is sufficiently wide to allow for this variation, etc… yet I’ve been surprised at how many people I’ve talked with (in the industry) who have said things like ‘we always spec 5Nm for that fastener…’ or who didn’t understand where in the range slippage was possible and where fracture was possible.
Despite all of this, I know when I was still at Zipp we would warranty hundreds of Ti stem bolts per year, these bolts had an onset failure torque of 12.5Nm MIN, meaning that in testing no bolt was allowed to begin plastic deformation, much less fail below 12.5Nm and the average bolt would fail around 14-15Nm… and the torque spec on the stem was 5Nm!
I had numerous cases where people would call in (including shop mechanics) having broken 2 or more of their stem bolts at the same time!!
One day way back I noticed the torque spec on my Felt stem bolt was 2Nm. Asked my coworker about it. His response? “That can’t be right”.
Fast forward to IM Cozumel 2014, first time travelling w/ my bike. Broke it down, flew with it, got it there and assembled it. As I’m tightening the stem bolt I feel it strip slightly. Ruh roh. Turns out they used aluminum stem bolts. Thus the 2Nm warning (which was in fact right). It took a little effort getting it out after the race but luckily I was able to. I got a replacement shipped to me when I got home.
First day when I brought my brand new bike home - I start installing my own saddle, use my cheap ass Nashbar torque wrench to tighten the seat post - set it to 5Nm and start tightening and tightening and tightening - the damn thing doesn’t click. At one point I think - something’s wrong here - I stop and remove the bolt, put it back in, wrench clicks and I’m all good to go. Next day I go to ride my bike on the trainer and notice the clamp seats crooked and one of the nuts has been ripped from the frame. I almost had a heart attack. I took the bike to the shop and luckily the bike shop guy agreed to send it to warranty and they got me a replacement frame three days later.
Needless to say, my torque wrench went straight into the thrash bin when I got home.
Woah… I didn’t know anybody was spec’ing anything at 2Nm… very interesting, particularly as that is a really hard torque to accurately read/measure for any standard torque tool. In our testing, putting a click-type torque wrench in the refrigerator will on average cause it to click more than 0.5-1.0Nm higher than it does at room temperature, so not a lot of margin for error when the target is 2.0!
You can send your torque wrench to be calibrated. We calibrate hundreds of tools per year – some we do in house, some we send out. This thread reminds me to bring my torque wrenches in and get an informal calibration from our shop.
Crazy thing is- all of my stories is when using a torque wrench.
Now, I start to torque it and if it feels tight and torque wrench doesn’t click, I call it tight! Not worth breaking stuff.
Replaced the ignition coils in my Porsche Cayenne. One bolt appeared different than the others. Others looked like steel and this one was aluminum. Whatever. I used the torque wrench. Snap. I was so pissed. Thought I would have to drill it out of the engine. Luckily removing the ignition coil left enough of the broken bolt exposed that I could get it out with pliers. Ordered a new bolt from Porsche and waited a couple days til I could finish the car.
Swapped stems from the integrated to the standard style on a Trek speed concept. Bolts were getting tight, but hey- trust the torque spec. Snap. Bolt broken in fork. Fork was now junk. Turns out that the standard stem uses 20mm length bolts and the integrated used 25mm bolts. I bottomed out the longer bolt in the fork and broke it before it tightened to the stem. No clue why Trek doesn’t include shorter bolts and info regarding this.
