I bought a new frame (online) intending to build up the bike. It will be used for recreational rides on gravel/dirt paths mostly.

The frame has an English-threaded bottom bracket shell. I happened to have a new I.R.D. BB manufactured by Tange Seiki in Japan, 68x127, which I'd bought for another bike and never used. The BB shell's threads were clean, I applied a thin layer of Park anti-seize to them with a toothbrush, and then gently threaded the BB in by hand, first the drive side and then the non-drive side. Everything was smooth as butter, no cross-threading, so I torqued them to 40Nm.

On the non-drive side, between 1 and 4 o'clock, the outermost threads of the cup are proud of the BB shell face by 0.56mm (at 2 o'clock) whereas on the 7-10 o'clock side of the shell, the cup sits pretty much flush with the face.

There's bare metal on the face of the BB shell, so I assume the BB shell has already been faced.

Is 0.56mm sufficient imperfection to try to rectify?

Is it possible the axis of the threading is itself askew?

I don't have the bike built yet, so I can't say if this is affecting the pedaling.


A lot here depends on the "shape" of the imperfection.

I'd be looking for contact around something like three fourths of the circumference.

If you have contact most of the way around the shell, with a little "divot" gone so there isn't contact in one small area, it's probably not worth facing .5 mm off the shell to get contact on that last little bit.

On the other hand, if you basically have a "point" sticking out on one side, and essentially the entire rest of the shell is .5mm away from contact with the face of the shell, it's almost certainly worth cutting that spot down to get contact around most of the shell.

If it's a matter of the shell having been cut off slightly crooked, so the face is flat, but not square with the threads, I'd face it off so there was contact at least half (and as I said above, preferably more like three quarters) of the way around.

  • Good point on the difference - if it were a single nubbin I'd very-carefully hand file it down with a flat file (not a rotary tool) – Criggie Jan 30 at 21:53
  • Imagine a 1mm thick circular clock face with an axis running from 11 o'clock to 5 o'clock. Now press in at 8 o'clock ever so slightly, so that the clock face at 8 o'clock moves away from you and the clock face at 2 o'clock moves towards you 0.5mm. That is how the cup will sit -- the perfect circle has become an oval. I would have to try to remember my trigonometry to calculate the extent to which the 9 o'clock - 3 o'clock diameter has been reduced. – Tim Jan 31 at 21:05
  • "The perfect circle has become an oval ..." if you're looking into the bottom bracket shell from the drive side. The non-drive-side cup is not presenting a circle to the spindle, but an oval. – Tim Jan 31 at 21:12
  • But if the threads on drive-side are on the same axis as the threads on the left side -- then the inverse would be true on the drive side, the right-side cup would be proud of the face on the bottom right quadrant. – Tim Jan 31 at 21:19

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.