I have this recurring problem on many of my fixed-gear builds where the non-drive-side crank arm starts to wobble and gets loose after a very short amount of time (sometimes after just a short ride where I skid stop a lot).

These are usually new crank arms, sometimes steel ones, they're not bottoming out and I'm liberally applying lock-tite to the threads of the bolt. The BB in the latest case is new as well, though I'm hearing a little grind with each rotation).

I don't have a torque wrench and if I did I wouldn't know how much force to apply anyway since I don't have specs for my parts, but I'm just wondering - what am I doing wrong (or NOT doing)? I don't want to overtighten (I've had seized cranks before) but other than just reefing on them, I'm not sure if there's anything else I should be considering (other than replacing the technician!!)

  • 4
    Do you grease the spindle? Worth reading - Torque should be around 45Nm, which is fairly tight (20 kg on a 200-250mm spanner). – mattnz Jun 26 '20 at 0:31
  • 4
    Don't use Locktite. Lightly grease the spindle and the bolt, and torque it in well. – Daniel R Hicks Jun 26 '20 at 0:37
  • Just confirming, this isn't the same set of wallowed-out crank arms each time? Next time you have that crank arm off, clean the bore and look for any sign of fretting, which is missing metal and a non-square hole. – Criggie Jun 26 '20 at 12:42
  • Small chance of this being the case, but since you mentioned it's on a fixed gear, what cranks are you using? Certain track-specific cranks still use an ISO-taper bottom bracket. it'll appear to work on the predominantly JIS-taper bottom brackets out there but isn't an exact mate and will likely cause problems. Specifically, the Sugino 75s and old Suntour Superbe cranks use the old ISO standard. Anything Shimano should be JIS. – Nathaniel Hoyt Jun 26 '20 at 20:33
  • 1
    @Criggie thanks for asking, no in this case brand-new crank arms on a fresh BB. I haven't checked the BB (it came with the frame) whether it's ISO or JIS, the crank arms would be JIS, but thanks Nathaniel for reminding me to check! – Tom Auger Jun 29 '20 at 13:47

You are undertightening the crank bolts on initial installation. 44Nm or 33 ft-lb work well as generic values for square taper cranks. A basic large beam-type torque wrench works well for crank bolts. If you aren't able to get one, it's safe for basic square taper cranks to reef like mad with something like a 12" ratchet or breaker bar. You'll be exceeding that torque value, but bad things won't happen. Fancier cranks with no extra material to give could eventually crack by doing so, but probably won't.

Always grease the threads and the bolt shoulder. Loctite is not optimal in this application, but it's also sufficient as an assembly lubricant to keep damage from occurring as long as there's appropriate torque.

Once the crank has come loose on the spindle, there's no fixing it. The press interface will never be right again. Throw them away. The spindle is hardened steel and is unaffected in almost all cases. If you're wrenching for others who have this problem, you must not re-install the crank. Doing so is irresponsible because the next time it comes loose, it might happen suddenly and cause a loss of balance and potentially a crash.

Square taper crank interfaces that are correctly installed don't have problems with seizure.

Some perceived loosening of the bolt is normal because as cranks are ridden, they can move around on the spindle enough to decrease thread preload. This is part of the reason why crank bolts are such high-torque fasteners; they must have some extra preload to give to avoid problems from this effect. Jobst Brandt wrote extensively about this on rec.bicycles.tech, material you can find archived by searching around. There are some dynamics happening that aren't readily apparent. Part of the conversation involves whether to grease the taper or not, which I won't get into here because it's outside the scope of the question, and in the end it's fine either way with some qualifiers.

  • 5
    +1. OP: don't worry about over tightening the crank bolts. Just make sure you grease the threads, then put as much torque on a 12" ratchet handle as you can muster. Overtightening is only a risk if you're using an impact driver or some monster breaker bar or some such. Channel your inner gorilla and crank on that thing. – whatsisname Jun 26 '20 at 2:36
  • In practice, if you torque it up while holding the crank arm (e.g. with your foot against it) you can use an arbitrarily long lever on the bolt. I also find it helps to tap the crank onto the arm with a wooden mallet when the bolt is well engaged but not tight, before finally torquing it – Chris H Jun 26 '20 at 7:57
  • My current procedure installing square tapper cranks is this: Put on the crank, tighten the screw. Take a piece of wood and a fairly heavy hammer, place the wood flush on top of the crank where it sits on the spindle, and give it a few smart blows with the hammer. This thoroughly seats the crank on the square tapper. Then tighten the screw for good. It's hard to actually overtighten these screws, but with a thoroughly seated crank, you don't need to seat it via torque. Using this method, I get my cranks installed well enough that retightening the screws after a few kilometers is pointless. – cmaster - reinstate monica Jun 26 '20 at 9:51
  • 1
    @Criggie Yeah, that's what I've been doing (more or less) before I started using the hammer. The hammer method is just quicker, and it ensures a good seat right from the off. No danger of working loose the crank on the first ride and thereby damaging it. – cmaster - reinstate monica Jun 26 '20 at 12:37
  • 1
    It's quite amazing that the article listed speaks specifically of the left-side (non-drive-side) crank arm being the main culprit due to the dual shearing forces. This has definitely been my experience over many years! – Tom Auger Jul 13 '20 at 21:00

It is likely you're under-tightening the crank arm bolt. There's a bodge way to estimate torque when securing a fastener.

45 Nm is very roughly the torque applied by ~4.5 kilograms at a lever length of 1 metre.

So 13.5 kilograms at the end of a 300 mm ratchet, 18 kg on 250mm, or 22.5 kg on a 200 mm lever.

  1. Measure your lever, and find a suitable weight in a form-factor that you can lift.
  2. Arrange the fastener and tool so that you are lifting against gravity. I find kneeling with the handle down near a hip works well.
  3. Lift the weight clear of the ground and get a mental idea of the effort required.
  4. Replicate that level of effort on the handle of your lever.

This isn't stunningly accurate, but is somewhat better than using "good-and-tight" Most people will under-torque the big bolts and over-torque the little ones.

This method should put you within 20% of "right" but you may need someone to hold the bike for you. The longer your lever the easier it is to get close to correct.

Of course nothing beats the proper tool, which is a torque wrench in the right range, but they're quite expensive.

Your passing comment about the BB getting a bit crunchy - this is a perfect opportunity to try tightening the bolts and see what happens, especially if the BB needs replacing anyway.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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