Who makes the toughest/strongest chain ring bolts?

I've got Shimano Ultegra gruppa, but pretty continually break these bolts....I'm an engineer, and today had all the measuring equipment out trying to figure this out, and it seems the bolts are stretching, and then breaking.

Scenario goes like this: biking along, get a ticktick happening from the crankset. Re-torque bolts, sure enough there is one or two loose. Ride more, the tick comes back, go to re-torque, and the bolt breaks. Turns out the bolts are not actually loosening off, but are instead stretching. You can only re-torque them a couple times before the stretch affects the strength and then they break.

Shimano manual calls for 106-141inlb on these bolts, I've been doing them to 125 inlb. Less seems to give me a ticktick. I tried going to max and the bolts broke right away.

So ya, I need something stronger. Even if the bolts went from using a #30torx to #25torx, that would add a lot of strength too! I'm pretty hard on equipment overall...LBS loves me

Matt

Are you running OEM or replacement rings?

I ask because there are two scenarios. Quality cranksets support the rings on the shoulders, and the bolts only do two jobs, act as driving pins and holding them in place on the arms.

However, many replacement rings, and some cranks don't properly support the rings on the crank arms, causing them to depend entirely on the bolts. Because there's always some slop in the bolts (no matter how little) the rings undergo precession, since the load direction rotates with the crank. That causes them to work at the bolts, eventually loosening them.

One thing you might try is some lapping compound between the rings and crank arm faces. That will improve binding between the two, hopefully enough to combat the precession which is the core problem.

FWIW - stronger bolts WON'T help. You have to address the precession or you'll be dealing with this forever. Once you do address it, you'll eliminate movement, and the rings will stay tight forever.

In case it isn't as explained above but due to some other reason, and you are using aluminum bolts, switch to titanium. And don't overtighten them. Use thread locker instead (aka Loctite).

It's not the bolts, and if it were, then steel would be best.

The key is the precession which underlies the entire issue.

A thread locker will prevent loosening, but the OP will still hear the creaking as the rings shift with the change in load direction.

BITD the makers of top end equipment held such close tolerances between the rings' inner circle and the crank spider shoulders, that the rings almost had to snap into place. That's why bolt loosening on good stuff wasn't a problem back then.

I'm using OEM everything.

I had a bunch of bolts around the shop today (all shimano cranks) and was measuring them. The longer the bolts the less torque they could handle. The longest bolt broke before 80 inlb. The shortest bolts could handle the 125inlbs. That's what got me onto the bolts-are-stretching theory. We are talking ten-thousandths of inches here, not much. This could also just be from the manufacturing process, but it sure seemed to be a trend this afternoon.

I'll buy a new bolt set from LBS and Loctite them in, maybe physically marking them as well. That'll show if they're actually rotating. I'm not totally convinced, but it could be precession of the ring is putting a different load on the bolt as well, causing them to break when re-torqueing.

Matt

I will try the lapping suggestion as well!

Matt

The stretching is weakening the bolts. A bolt can stretch under tension only so far before it breaks (the elongation at break). But the stretch you are observing isn't reversed when the tension is released. It is creep. You have a stretched bolt even at zero tension. So it takes less tension to stretch the bolt the rest of the way to break. It already has a head start.

not sure of many chain ring bolts that are rated that high 125inch lbs =14.x nm and 141= 15.9nm sounds pretty high for chain rings. i put 12nm on my steel chain rings that came with my quarq but they are all steel not sure what the shimano ultegra are if they are part aluminum those torq numbers sound pretty high

Something is very wrong here.

Is it possible that the longer bolts are too long and somehow bottoming before they tighten?

Also, the precession I referred to earlier, related to chainring movement, not the bolts themselves. Using a threadlocker may help keep the bolts from loosening, but won't address the core issue. You need to increase the friction between the rings and tabs of the crank spider, so they can't shift under chain load.

Obviously tighter bolts can be a part of that, but you seem to be hitting the limits. So, start with an absolutely dry interface, (solvent clean, and dry). If that isn't enough, use my lapping compound suggestion, or apply an anaerobic adhesive, like a bearing retaining grade, between the rings and tabs. There are also anaerobics that swell as they dry, so applying one along the boundary between the chainring and crank shoulders may prevent the precession.

The bolt breaking before it's recommended torque would lead me to believe it's bottoming out.

I think some blame has to fall on Shimano's wacky asymmetric 4-arm cranks. 12-16N-m for chainring bolts? All of the bikes I'm riding/wrenching are 5-bolt 110BCD, and call for 7-8N-m. I've never broken a bolt.

fwiw, I think Campy's 4-arm cranks only spec 8Nm for chainring bolts.

The idea of the bolts bottoming out is a good one. Could maybe the bolts be for a thicker crank? Like 11-speed bolts on a 10-speed crank?

Interesting discussion. Two of my bikes need a chain ring bolt tightened every few weeks or few hundred miles. Hadn't occurred to me it might eventually snap. I'd assumed it was just loosening, not stretching. I can't reach the lock ring properly without disassembling the darned things and just haven't bothered to do it properly with Loctite.

never had a problem on any of the new 4 arm cranks with bolts or rings squeeking, and I've swapped around a bunch to run combos from 50/34-50/36-52-36. All OEM bolts and I always grease the outer heads and locktite the threads

I'd switch to steel chainring bolts. Sugino and Truvativ both sell kits of 5x 8mm long bolts and nuts. I kept cracking the heads off aluminum chainring bolts for a while before I switched back to steel ones -- one at a time. I think the 5th aluminum chainring bolt is still in service with 4 steel ones carrying the load.

Seems like a better question for the mechanic's corner than the road forums.

That does seem like a crazy-high torque spec, but there it is: https://si.shimano.com/pdfs/dm/DM-RAFC001-02-ENG.pdf

This sounds like a warranty issue to me, given the rarity of what you describe. Unless you're putting out 600w constant, or something....

So are the bolts failing from tension or shear? You talked about testing them, but which were you testing? Torsional force might also have been what you where describing for your test.

So which mode is the failure for your actual chain ring issues.

Powerful people and tandems use alloy bolts - sometimes.

While steel would be stronger (NOS Campy 2 speed and triple), there is something going on besides bolt strength.

Seems very strange to break bolts with a torque wrench, how confident are you in the accuracy of your wrench?

I checked SRM's torque chart (happen to have it handy) and they have Shimano 11spd chainrings at 10-12 Nm, roughly 90-106 in-lbs.

Yes, you're right if you are following a recommendation for the torque. But if you are continuing to crank up the setting to see how tight you can get it, that is a different story.

Just an update to this: My torque wrench is regularly sent out for calibration, so is accurate. I haven't been able to do much riding since replacing the last bolt I broke. However, I did mention this while I was at my local Trek Store for an unrelated repair. Their suggestion was to replace the bolts with OEM, and if any of them break they would warranty the crank as defective. Sounds like a good plan so I'm going that route.

Matt

Were the bolts that broke not OEM?

They are OEM, yes. But since I've had them out, and been replacing them, measuring, and generally dicking around, Trek store won't pursue warranty on the crank at this point. They want to replace them all, and then if they continue to break will pursue replacing the crank...they think the crank is faulty, and this is the only way they can/will warranty the crank.

Matt

I suppose that's reasonable, if a bit untrusting of the customer.

Well the bolts are definitely stretching, not turning loose. I torque them, and marked their position. Then I'd go riding, and then torque them again to the same value, adding approx 1/8-1/4 turn after each ride, but not always the same bolts, or all of them. After 3 rides, I went to torque them again today and the one felt like it was going to break so i pulled it out..a solid 3/32" longer than the rest...I pulled out all 4 and they were all different lengths...could just tell by eyeball.

I happened to look at my old Deore XT crank I replaced earlier this year. It had steel bolts so I put those into my road bike this morning. Torqued to max-141inlb, and marked position. We'll see how this afternoon's ride goes.

Matt