I was browsing specifications of internal gear hubs and noticed that many of them have flange center moved to left (relative to hub center). I'm wondering why, I think IGH should have symmetrical flange distances for same spoke tension on both sides.

Some examples:




Shimano SG-C7000, H is the offset I'm talking about

Shimano SG-8R20, E is offset

  • 1
    Could it be to provide better access to cooling air to the disc?
    – Carel
    May 21 at 10:20
  • 2
    " I think IGH should have symmetrical flange distances for same spoke tension on both sides." why? torque is still applied to the wheel/hub on one side, by the chain, so the rear wheel is intrinsically asymmetric and I feel it needs to be balanced.
    – EarlGrey
    May 21 at 11:15
  • "Could it be to provide better access to cooling air to the disc?" disc is on the side to which flanges are offseted, so vice versa - it lowers air cooling. correct me if I am wrong or don't understand you correctly.
    – AKmatiAK
    May 22 at 7:41
  • @EarlGrey: The hub shell provides a rigid connection. There are some rear wheels with radial spoke pattern on the drive side where only the spokes on the non-drive side transmit torque.
    – Michael
    May 22 at 9:52
  • I can’t think of any logical reason for this design, except maybe something in the internal construction makes it necessary? It could also be necessary to provide enough space for chain guards.
    – Michael
    May 22 at 9:57

It's a good observation.

Symmetrical flange spacing and the maximized total spoke tension it provides is exactly what one wants in some applications, but not all. If you're designing a hub you have to think about what is enough total tension and then parlay the benefits of getting "extra" strength above that amount versus the drawbacks of reducing the bracing angle, albeit only in one direction. When wheels fail in practice, a pretty meaningful percent of them fail from side loads. So when you see this asymmetry on hubs that could have been made symmetrical, it's the designer making decisions on what will actually produce better reliability on average.

An exaggerated version of the same thing is seen on track hubs. They're the most could-be-symmetrical hubs of all, but proper track racing hubs usually aren't. The reason is that radial overloading is just not that much of a concern, but crashes and heavy sprinting riders rocking the bike side to side are.

  • 3
    But how asymmetrical tension helps to improve wheel strenght?
    – AKmatiAK
    May 22 at 8:57
  • @AKmatiAK As an extreme example, picture a modern road wheel, with its nearly vertical drive side spokes and the left side spokes at a much wider bracing angle and about 60% of the tension. Why not make that wheel have equal tension too, i.e. make the flange spacing much narrower? If you did that, the wheel feels flexier and is vulnerable to failing from relatively minor side loads, even though it will have more total tension and be stronger in the straight up and down direction. The term 'column buckling' is sometimes used for this effect. May 22 at 16:44

So far my findings are that this offset may be to equalize bearings load, as it is in case of car wheels offset.

  • But the rim is still above the center line of the hub and the hub shell is rigid. So from a bearing point of perspective everything is symmetric and the asymmetric spoke flanges don’t achieve anything.
    – Michael
    May 22 at 9:55
  • 1
    shell rigidity is not the case. imagine two bearings - one on left end and second on right end of theoretical hub. let's assume you have flange on only one side, at the end of hub. torque will be put on that side and wear bearing on loaded side faster.
    – AKmatiAK
    May 22 at 10:14
  • 1
    For Japanese designers, symmetry is less important than with Western design.
    – Carel
    May 22 at 10:14
  • 1
    @AKmatiAK: Consider how the weight force (applied to the center line of the bike) is transferred to (symmetrical) drop-outs. The left and right drop out receive exactly half the weight. It stands to reason that bearings in a similar (symmetric) location would also receive exactly half the weight each.
    – Michael
    May 22 at 10:23
  • I don't know what's the case then. Maybe it's hub construction reasons. I doubt it's because chainguard - I checked it on my hub and chainguard (I didn't assembled my bike yet) and there is still some free space left. Also, there are different offsets in similar O.L.D. hubs - as in first image I posted - coaster brake has bigger offset than other variants.
    – AKmatiAK
    May 22 at 10:31

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.