I keep an old BSO near work, generally outside. The last few days, the left crank bolt (square taper) has loosened every few km. Admittedly the socket spanner I put in my bag isnt very long but I put a fair bit of weight on it and probably got 40Nm or not far short.

When tightening it just now, I noticed that the pedal bearings on that side are rubbish, really stiff and rough, and probably the cause of at least some of the dodgy noises.

But could they really cause the crank bolt to loosen? The point of a square taper is that the bolt only holds it together and doesn't see any torque.

The plan is to take a spare pedal to it tomorrow, as well as a longer bar for the socket.

  • 1
    At the moment I'm starting to think it's coincidence, that the pedal bearings failed and the crank worked loose at the same time. With a good pedal but not further tightening the crank, it still loosened this morning, but less. I may have overestimated the effective lever arm of the spanner I used before; anyway since this morning's ride I've really leant on it - probably 60Nm as most of my bodyweight was on the (30-cm long) spanner and pedal.
    – Chris H
    Jun 22, 2023 at 9:50

2 Answers 2


I'd think so.

It seems to me the only really significant force that's normally applied to the crank's attachment to the bottom bracket is from the rider's downstroke pushing on the pedal. There may be some reverse forces lifting the rider's foot and leg on the opposite side, but I think most riders would also be using muscles to some degree to assist the motion, with some riders actively pulling up on the pedals.

But with a rough, stiff pedal bearing, those "reverse" forces are going to increase, maybe even by a large amount.

That means the crank/spindle joint isn't going to be subjected to just a push-push-push series of force applications with maybe some minor pulls along the way, it's quite likely to be subjected to push-pull-push-pull series of forces, actively rocking the joint back and forth.

That sure seems to me to make for a situation almost designed specifically to loosen the joint.

And if it's a square spindle, once it loosened under load it'll never stay tight again if the motion distorted the surfaces of the crank arm. So the fact that it won't stay tight now doesn't necessarily mean anything.

  • The creak of a loose crank certainly demonstrates the cyclical nature of the forces. It being a BSO may have an advantage for once - the cranks are steel so won't have got chewed up like aluminium ones. It's amazing how much it managed to loosen in a 3 km trip.
    – Chris H
    Jun 22, 2023 at 14:32
  • 1
    How about when you coast over rocks on an MTB with the pedals horizontal to the ground? One of the arms would experience torque in the other direction than when pedalling. So coasting-pedalling-coasting-pedalling over time should loosen the arm, but that doesn't happen, right?
    – Robert
    Jun 25, 2023 at 0:35
  • 1
    I've since scrapped that bike for unrelated reasons, but before replacing the pedals it kept loosening. With decent pedals it didn't
    – Chris H
    Nov 21, 2023 at 15:55

The arm could have a hidden crack around the area that holds onto the square spindle. I would inspect it carefully. As you already know, it can cause a crash if it fails at the wrong time during a ride.

Given it's the left crank arm, it makes it easier for a quick swap if you have a spare from another crankset laying around. It would be curious to see if it still becomes loose, even with a new crank.

If the arm is not cracked, then maybe thread locker on the bolt should help a little bit. I've seen that new cranksets come with the threadlocker pre-applied on the bolts when bought new. Even if the arm becomes loose anyway, the threadlocker should at least prevent (or delay) the bolt from undoing completely and the arm falling off.

The spindle could also have a crack. Might be worth inspecting it too.

Being left outside, the elements might have accelerated corrosion of the interface between arm and spindle. I would inspect that as well.

Personally I would just replace the BB and at least the left crank arm, just to be safe from crashes.

Could the stiff pedal have anything to do with the crank arm becoming loose? I doubt. Let's imagine the pedal is completely locked, bearings as stiff as if they were being welded in place. Then you push down on it. What torque is produced, and where? Firstly, at the pedal thread, in the direction that undoes the pedal, so the pedal would become loose. And secondly, at the square taper interface, in the normal direction the interface was designed to withstand torque, right? My intuition might be wrong, but I don't see how the stiff pedal could cause the arm to loosen.

Another posibility could be that the bolt can no longer push the arm on the spindle far enough. Why? Because the surface on the arm that contacts the bolt becomes flush with the side of the axle (where the bolt screws into) too soon. Why? The bolt could have chewed away too much material from the surface of the arm. But this is very unlikely with a steel crank. Even more unlikely if there's a washer under the bolt.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.