Why do crankarms come loose? and a temporary solution
 

This has happened to me maybe three or four times in 40 years of riding: You're cheerfully pedaling along, and suddenly the non-drive side crank arm is rattling around loose on the spindle. It's not like it gets a little bit loose then rapidly gets worse--it gets very loose all at once. I seems to me that the crankarm fixing bolt gradually works its way loose until it's just finger-tight, but the arm itself stays pressed onto the spindle until suddenly it breaks loose, maybe as a result of standing up on the pedals to climb over a rise. Has this happened to others? Does anyone have any insight as to why it happens?

Anyway, it happened to me again last weekend about 20 miles into a 50-mile ride (returning from an overnight stop at a friend's house on a run from near Bellows Falls, Vermont to Amherst, NH and back). I was on my Gitane TdF, and it was a rainy Sunday with no obvious opportunities to borrow a 14mm socket wrench. I had just resigned myself to screwing the bolt in finger-tight and destroying the crank arm in the process of finishing the ride when I had a brainstorm while passing a place where the local utility had cut down a bunch of trees.

We pulled off the road and I lay the bike on its side with the drive-side spindle centered over the stump of a small oak tree. My friend Dave held it there while I found a two-foot length of oak, maybe four inches in diameter, that was sawn off square at both ends. I positioned the non-drive-side crank arm on the spindle, with the bolt removed and WHAM! brought the butt of the length of oak down on spindle end of the crank arm as if it were the business end of a wooden mallet.

To my surprise, that solved the problem. It pushed the crankarm far enough onto the tapered spindle that it stayed tight for the rest of the ride--no play at all, and no damage to the crankarm or spindle, as I determined by disassembling them later. The key, I think, was having the other end of the spindle resting solidly on the stump, so all of the force was transmitted into the spindle and arm, rather than the bearings and cups.

That said, I think I'm going to convert all of my bikes from 14mm hex bolts to those more modern 8mm bolts with the integral dust cover. They may be slightly non-original, but my folding allen wrench has an 8mm attachment. Even if this only happens once every ten years or so, it's a failure mode I can do without.

-----

One simple check you can make is to look at the end of the spindle vis a vis the arm after you have it torqued down.

Withdraw whatever fixing bolt you are using and have a look to see that there is some freeboard between the end of the spindle and the seat for the bolt head and washer in the crank arm.

Alloy cotterless arms can only be mounted so many times before they open up a bit to where the end of the spindle can come up to the seating surface for the fixing bolt.

One simple solution when this occurs is to grind a bit off the end of the spindle.

​​​​​​​-----

Loctite

Perhaps. But in my experience the "so many times" is a very large number. On my folding bike I pull the cranks every time I put it in its suitcase. So over the years they have been pulled and reinstalled well over a hundred times with no visible change in chain line or difficulty with them loosening or not seating properly.

When All Else Fails, Hit It With Hardwood
 

I cut four matching keystone shapes from an aluminum can and affix them to the spindle; maintains the chainline.
(Edit: Last instance I used roof flashing; the crank had suffered abuse before we met, and required more material.)

In my experience, the most common reason for crankarms loosening is a burr forming on the crankarm taper. Each time an the arm is installed, a small amount of aluminum is displaced, building up a small burr at the tip of the spindle. Eventually, the burr builds up to the point where the arm cannot be pressed far enough onto the spindle, regardless of the installation torque. Filing down the burr solves the issue in the vast majority of cases. This is far less of a problem with forged crankarms than cast crankarms, which typically use softer, lower strength alloys. It is fairly common on low grade, box store bicycles. In the OP's case, the impact method used to re-install the cranklarm may have been sufficient to overcome any burr, which a progressive torquing may not have achieved.

Lack of paying attention, to maintenance, occasionally..?




...

I'll second this recommendation. After having my left crankarm loosen at the start of the last lap at a CX race in Clarksburg, then fall free from the bike (and thankfully from my shoe) half way thru that last lap, the crankarm taper was left pretty chewed up from that half-lap at full-race effort. I lost several positions and had to go back to retrieve the arm and pedal after finishing the race pedaling one-legged! When I got home I reinstalled it using a generous amount of Loctite Blue in/on clean bolt threads. I then jumped on the pedals with the cranks horizontal and retorqued it. I was then able to finish out the season, several more races, on those old TopLine cranks without any further attention to the bolt. I'll dare say that the Loctite on the bolt threads can save a crankarm with a pretty badly damaged taper caused by loosening.

No one here has mentioned that, a few miles after riding, torque the spindle bolts again. In variably, I seem to get a bit more torque on the spindle bolts. Loctite on the spindle bolt threads is a good idea, however; I have never had to do this. Proper installation, maintenance and understanding will, usually, keep your cranks tight. While at it, take the time to check the ring bolts, ensuring that they are still tight.

For installation of the crank onto the spindle taper, I start by ensuring that the female fit is clean and burr free. I take fine emery cloth and attempt to round or bevel each of the four leading edges of the spindle. Doing this will dramatically reduce the chance of pushing aluminum out of place while driving the crank arm up the taper. A wee bit of light oil, on the tapers of the spindle will also be beneficial. I use 3 in 1 electrician's oil and only a very small amount.

I was going to but you beat me to it.

I was taught, and I don't remember where or when or how, to torque them then ride 30 miles or so then torque them again.

And as mentioned, it never hurts to check them every so often before starting a ride or when you are looking for something to do on a rainy Saturday afternoon.

I seem to recall Jobst Brandt or someone sternly (he was stern about almost everything) advising against that. The idea was that the proper amount of torque at the initial tightening would push the arm onto the taper as far as as safe, and that further tightening later risked expanding it too much and causing it to crack at the corners of the opening.

Does anyone else remember that? Maybe I dreamed it.

But I think that loctite is probably a good idea.

I remembered the same; here it is: https://www.sheldonbrown.com/brandt/...ng-cranks.html

Your recollection is correct. But if you have a torque wrench you can check the torque without driving the arm further onto the spindle. Just apply the recommended torque; if the bolt doesn't move, it wasn't loose. If it does move, you just restored the proper torque.

I believe this is where you saw the Brandt admonishment. Installing Cranks by Jobst Brandt

Back in the day, aluminum crank bolts used to be popular with the weight weenies. You would use the steel bolt to tighten the arm onto the spindle. Then you would remove the steel bolt and replace it with the lighter aluminum version, though torqued to a far lesser amount. I don't recall a deluge of loose or failed crankarms, despite the "loss" of pre-load caused by the lower installation torque of then aluminum bolt.

Thanks--that's the piece I was thinking of. Just re-read it, and it makes sense to me. Interesting point about dust caps serving as retainers to keep loose crank bolts from falling out and getting lost.

Not being an engineer myself (although I come from a family of engineers), a good part of what Brandt had to say about any given subject always goes over my head. But I love how definite he is about everything. Whether he was right or not (and I think he was generally right), he was never in doubt.

As far as 'best practice' from a theoretical point of view, I have no desire to address the subject at all.

As far as 'best practice' from a practical point of view, I consider it foolish to go for a long ride without the tool to tighten a crank arm bolt, whether that's a socket wrench of some kind, or an 8 mm allen wrench. Of course the crank arm shouldn't come loose! But... it can happen.

One time, on a bike I had ridden only a couple hundred miles since building it up, a crank arm came loose about 20 miles into a 100 mile ride. Searching through my tool kit I found I had not brought an 8 mm allen wrench (which was dumb) but I had brought three different 4 mm allen wrenches (which is arguably even dumber). But to my surprise, three 4 mm wrenches work pretty much as well as one 8 mm wrench. This does not explain why the crank arm came loose, or excuse the mechanic who allowed this to happen :innocent: but the fact is, this can happen. Be prepared.

Jon, I sent you an email, not about cranks. Please see if I sent it to the right address.

I will not touch theory either. I'm not entirely sure why this crank fixing subject became controversial, but it is. Maybe it's just as well it's being supplanted by pinch bolts and splines.

I will say that as a practical matter, it is a very good idea to use a torque wrench to fix cranks. Most of the time they come undone, it's because they weren't on sufficiently tight to begin with. If you're not wrenching on this stuff every day, 'feel' isn't necessarily good enough.

+1. I feel like beam torque wrenches are especially good at this, since you can watch the applied torque climb gradually toward the target value while checking. :thumb:

I never saw or heard this trick elsewhere before. I've been "saving" cranks this way for many years. (Second "I'm fhe first to do this" myth that has been shattered for me today. Read a documented case of another "triple chainline" fix gear; three chainrings, three cogs, three very different gears.

Ben

But don't keep torquing them. Once is OK. You can always torque them down tighter, but if you do, the time will come when the crank will simply break from a crack radiating out from a corner of the square taper. When you ride a properly tightened crank, as you apply pressure to the pedal, you leverage the crank slightly further onto the spindle, away from the bolt head. So you can always retighten the bolt back to the original torque. If you do this, next time you ride, the crank does the same small move inboard. If you tighten it again, eventually you will have forced the crank so far up the wider portion of the spindle taper that it has no choice but to break to relieve the stress.

This is a case where too much attention is a killer.

Ben

Tom, I'm not sure where that email went, but not into my inbox. It should get to me if you send it to jon@redclovercomponents.com

Same here although it wasn't a case of 'saving' a crank but rather one that came with a spindle that was too short on the left side. Tandem TA crank and the spindle that shipped with it may have been intended for a single. When tightened, the bolts on the inside of the cross-over chain ring rubbed against the bearing adjusting ring. Cut up a Coke can and put a shim around the square taper to hold the left arm out a bit farther. That was 42 years ago and I'm still using the Coke can shim whenever the bearings are regreased.

I'm thinking that even if you don't re-torque them, keeping them non-loose means the crank arm can't come off. So just don't let them get loose.

I would like to know where this information is documented. In all my years as an industrial mechanic, I never once saw a tapered fit push up the taper, during use. Down, yes, but not up. Please understand, I am not arguing, just asking where I can learn more about this.