Rear chain line

Has anyone measured the rear chainline (i.e. the distance from the wheel center/hub center to the cassette center) with a Shimano 11 speed MTB cassette on a 142mm OLD hub?

Additional details (but I think they don’t matter):

  • Cassette: 40-11
  • Hub: DT Swiss 350, straightpull, Centerlock.

Maybe there is a way to determine this without measuring? It should be possible if I knew where on the freehub the cassette exactly sits (i.e. so and so many mm outwards from the end of the freehub). Then you can calculate it because the cassette width and enough hub dimensions are known…

There are certainly dimensioned drawings available of both cassettes and of freehubs. I have Shimano versions of these but if you wanted to get an exact dimension you’d need the equivalent for the DT hub. I’m sure their freehub body will be close to the same position as a Shimano on the hub but that’s just an assumption. I’ll ask why you need that dimension though?

Beides these drawings wouldn’t you also need to know where on the freehub the cassette sits? Because the cassette overhangs the left end (towards the hub center) of the freehub by a few millimeters onto the drive side flange and it’s a guess what the extent of this overhang is.

I need it because I would like to achieve a “correct chainline”, i.e. I would like the midpoint between the two chainrings to have the same distance to the bike centerline as the center of the cassette. Now if the rear chainline is rather big then I would install a longer crank axle to adjust the front chainline to match the rear chainline. If I know that I will do this in advance I have more room between chainrings and tires to play with which has implications for chainstay forming and / or tire clearance. If the crank axle increase is not enough to give me a roughly correct chainline I might also have cancel my plans to offset the dropouts by 6mm.

ok, so you’re using a square taper crank axle? I think if you have the parts in hand then you’d be better off physically measuring. I don’t want to say you’re over thinking things but maybe you’re over thinking things?

Chainline for MTB 11speed on 142 OLD is 48-50mm. I’d be building to that spec. I’m not sure why you’d need to narrow in on a more accurate number if you’re running a multispeed cassette?

1 Like

You might be onto something there :crazy_face:
The complication is that I want a 6mm offset frame (think Cannondale AI). That increases the chainline 6mm. If the chainline front and rear are too far off I might have to change to a 3mm offset instead.

Are you specifically referring to the rear chainline? Do you have a Shimano source for that? I couldn’t find anything in the Shimano documentation. The front chain line is pretty well documented but not the rear chainline (maybe because it also depends on the hub).

I use standard Rotor cranks which allow to swap crank axles to adjust chain line.

Well that’s the thing. There’s not really any such thing as rear chainline. Chainline is just chainline. Shimano will publish recommended chainline for specific transmissions (10,11,12 spd, Boost, Non-Boost etc). There’s so much variation of where the chain sits at the back, that to publish an actual number doesn’t really make sense. They do publish a chainline for single cog hubs (Nexus, Alfine etc) as do Rohloff and other single sprocket hub manufacturers. Chainline numbers makes sense for things like chainsets, bottom brackets and front mechs. Not so much cassettes. Back in the day, you could theoretically adjust the position of the multi-sprocket freewheel by moving spacers around on the axle but with cassettes and freehubs that’s not really an option. Again, I suspect you’re overthinking things and for your specific use case, I think if you assume the ‘rear chainline’ to be 48-50mm you won’t go wrong.

1 Like

I don’t know about the terminology but I think it makes sense to know what the distance between hub center and the middle of the cassette is. See below why.

Shimano and co publish (front) chain lines, i.e. the distance between bb center and the center of the chainring (1x) / center between the two chainrings (2x).
I have measured 3 bikes in the past few days and the “rear chainlines” were 43mm (11-36 12-speed 105 cassette on a 142mm hub), 45mm (11-42 11-speed GRX cassette on a 142mm Shimano Deore hub) and 46mm (11-51 12-speed Shimano MTB cassette on a 148mm hub). The method I used to measure rear chain
line is described here: Bicycle chainline explained | BikeGremlin US

This means that front and rear chainlines are only the same for 2x Road setups. By “front and rear chainlines being the same” I mean: you have a straight chain when the chain is at the center of the cassette (e.g. cog 6 of 12) and where the manufacturer defines it’s chainline in the front. Everything with a wider front chainline than a Road Double has mismatched front and rear chainline. The reason drive train manufacturers do this is:
(a) With 1x, chains would rub the next bigger sprocket when the chain is on the small sprockets (i.e. right side of the cassette) if the chain is angled too much. If front chainline is bigger than rear chainline then this isn’t a problem anymore.
(b) The trend towards more tire clearance and shorter chainstays requires wider front chainlines and it is easier to make the front chainline wider than it would be to introduce a new hub spacing standard to compensate for this increase in front chain line.

This Park Tool video https://m.youtube.com/watch?v=0lN3Zf9gpp8 measures rear chainline to be between 45mm and 47mm for boost MTB setups which is similar to my measurement of a similar cassette and hub. It also confirms that front and rear chainlines for MTB drivetrains differ significantly.

My conclusions:

1x is no good if you like efficiency (nicely illustrated here https://m.youtube.com/watch?v=X1jdTDPwFVc&pp=ygUuSW5jcmVhc2UgZGlzdGFuY2UgYmV0d2VlbiBjaGFpbnJpbmdzIGNoYWlubGluZQ%3D%3D)

An offset frame is a great idea SPECIFICALLY FOR BIKES WITH DOUBLE CHAINRINGS (potentially problematic for 1x, see above) not only because it makes the rear wheel stiffer (due to more even bracing angles) but also because it can, in some specific circumstances, give you a better chainline than if you would assume that your rear chainline equals the front chainline as specified by the manufacturer. For example: I assume that my rear chainline with 142mm hub spacing and an 11-40 11-speed cassette will be somewhere between 43mm and 45mm (probably more towards 45mm). With a 6mm offset frame my rear chainline would be somewhere between 49mm and 51mm. This would be 0.5 to 2.5mm bigger than my 48.5mm front chainline. And a little bigger front chainline is a good thing anyway (if you don’t have 1x) because you will be slightly more often in the big chainring on average which comes with a little less chain friction because the chain articulates less over the bigger chainring (this is the reason time trial bikes combine relatively big chainrings with big sprockets). The other two effects cancel each other out: the easiest gears (small chainring and big sprockets) are little more efficient while the hardest gears (big chainring and small sprockets) are a little less efficient.

You are definitely overthinking this. If you have a known cassette, hub and dropout spacing, the only thing you can control is the dropout position (lateral offset). Anything that isn’t symmetrical will need to be similarly offset at the BB - so if your BB is offset by 6mm, your dropouts should be too. Or more precisely, if your front chainline is 48.5mm but the spec for that groupset is 44mm, then you may find offsetting the dropouts by 4.5mm to the right works well. OTOH, if the spec chainline is 48-50mm, just put the dropouts in the normal place (symmetrical).

There is so much lateral flex in geared chains, and so much normal variation in the actual route of the chain across a cassette, that 2-3mm of “rear chainline” (or front chainline TBH) makes very little difference. Say if the front was 2mm too narrow and the rear 2mm too wide, you may have problems with rubbing on larger cogs as you said. But if they’re both off by 2mm the same way it cancels out.

I think this assumption that an offset rear triangle requires that you also offset the bb laterally by the same amount and in the same direction (hope I understood you correctly there) is crucial to what are are saying. I’m curious why you think that.
A 6mm offset for example would mean to move the following elements all 6mm to the right: right dropout, left dropout, hub, cassette, hub flanges, rotor, caliper. But the rim and tire stay where they they normally are otherwise the front and rear wheel would not be in line and the bracing angles also wouldn’t adjust favorably so there wouldn’t be a reason to do it in the first place.

This is the part I don’t get about your question - why would you even want to consider changing the rear chainline, when it means changing all that as well?!

1 Like

It is simple:
(1) The front chain line is a given (I’m not changing anything about it): 48.5mm.
(2) The rear chainline would be ~44mm without the offset.
(3) A 6mm offset provides two advantages:
(a) The rear chain line (~44+6=50mm) and the front chainline (48.5mm) become much more equal than without the offset: the difference is now only 1.5mm.
(b) Bracing angles (i.e. the angle at which the spokes leave to rim towards the drive side and non drive side hub flanges) are more similar on the drive side and non drive side than without 6mm offset. I hope the following pictures make this clearer:

Illustration of “bracing angle”:

Illustration of offset concept by Cannondale (Cannondale calls their offset frame concept “Cannondale AI”):


p4pb12618650

Maybe there are flaws in my thinking. This is partly why I talk about this here :slight_smile: So don’t hold back any critique or questions!

1 Like

I get that the wheel is theoretically stronger the closer it is to having zero dish but as with many things I see in bicycle design, it’s easy to overlook the fact that for almost every purpose, a dished wheel is ‘strong enough’. I can’t think of a single issue I’ve come across in many, many years of doing this where I’ve thought ‘this could have been avoided if the wheel was zero dish’. I’m not trying to downplay the issue (I build many bikes with Rohloff hubs and I love the fact that they are not offset) but I’m not convinced that you aren’t actually just building in a whole load of complexity to fix an issue that was never an issue in the first place (wheel strength due to dishing).

And you said it yourself, you’re using rotor cranks with options on spindle length so if you just want to build something with a front chainline closer to the rear, then you could leave the frame alone and just adjust the front by changing the axle length.

I want my bikes to work and to be easy to fix/repair when they don’t. Being able to slot in an off-the-shelf wheel if I need to, certainly beats the fact that my wheel is now stronger because it has no dish.

And don’t get me started on heal rub with a stay shifted an extra 6mm or trying to fit a rack or mudguards. :stuck_out_tongue_winking_eye:

3 Likes

I agree with everything you said except maybe for a few small details.

Could be true but it is also hard to know what the cause is of a broken spoke or an out of true wheel is. Would the wheel have held up better if that wheel would have been more symmetrical? Also imagine if we would have a blank slate and could choose to make wheels symmetric or asymmetric. It would seem insane to make them asymmetric.

True. But that would only have disadvantages: less space for tire clearance.

Can’t argue against that :grinning: But given that the industry has moved to wider front chainlines a change to a 6mm offset being the standard would make sense.

That is a good point especially with short chainstays but with my 450mm chainstays that is not a problem.

My head hurts. Ive tried to asorb all the points and data but I honestly am not seeing the problem here. 55mm chainline is great because it gives more room for frame design. In the the rear the chain moves so much and we generally spend more time in the middle sprockets, that making it all more complicated by having an asymmetric rear end doesnt make sense. Specialized tried it and the industry basically told them to go jump. Use an asymmetric rim if you want to improve the bracing angle of the spokes. A properly built wheel with a sufficient number of spokes works really well and has few problems.

4 Likes

I think pushing the design envelope is cool. I’m not trying to dissuade you of anything but posting ideas for feedback on a forum is going to illicit feedback. I guess you must be more concerned about the problem(s) you’re trying to solve than me. I see two issues you’re trying to improve, less chainwear due to a chainline more parallel to the frame centreline than normal and you want a stronger wheel? Is that correct? If that’s the goal and you’re design considerations address that then go for it. My feedback would be that if to get those two problems solved I’d have to run a non-standard/proprietary wheel, not be able to use 99% of wheelbuilding jigs/truing stands, not be able to use most frame jigs without modification, have problems fitting mudguards, have problems fitting racks, have potential heel rub issues, then I’m out I’m afraid.

2 Likes

Lots of builders with way more experience than me have shared some good insight, but I will offer my opinion - For some use cases, I think offsetting the rear a few mm to the drive side can be beneficial. On my hardtail, I spend a whole lot of time in the lowest gear (more than I should admit), and I have chosen to offset the rear end 4 mm to the DS to straighten out the chain a bit in that gear. Does it actually improve chain life or efficiency? Probably not measurably, but I think it runs a little quieter in the granny gear. Any gains in wheel strength are probably even more marginal, although I did also offset the rear end of my touring bike with that in mind.

I haven’t had any heel rub issues, but it is a pain to not be able to swap wheels between bikes.

If you are building the frame for yourself, you spend a lot of time in the lowest gear, and you build your own wheels anyway, I think an offset rear end is worth considering.

1 Like

My front chainline is only 48.5mm. But if it were 55mm the benefit from a wider rear chainline (for example through an offset frame) that comes close to that 55mm front chainline would be even greater.

It’s exactly what I want!

A full list of benefits as I see them:

  • Less chain wear.
  • Less chain noise in the biggest sprockets.
  • Less chain friction in the biggest sprockets.
  • You will be more often on the big chainring plus a bigger sprocket than you would be if the front chainline was bigger than the rear chainline for a given amount of cross chaining. This way the chain articulates less (because it runs around bigger circles) leading to less chain chain friction.
  • An offset frame allows me to run a wider (48.5mm) front chain line (if I compare an offset vs a non offset option and both of these have even chain lines front and rear) which increases tire clearance or reduces the need to ovalize/dimple chainstays.
  • Stronger/more durable wheel. I haven’t measured it but I would be surprised if the benefits are marginal. The bracing angles change substantially with a 6mm offset from 4.5°/8° to 5.8°/6.8°.
  • Particular only to my rear rack and seatstay design: the rear rack tubes that attach to the dropouts have a slightly wider base because of the wider flanged dropouts which increases the lateral stiffness of the rear rack.

The costs are:

  • The combination of a 6mm offset and a seatstay design that guides the NDS seatstay laterally around the brake caliper requires a dropout with a wider flange width so I can’t use off the shelf dropouts.
  • Slightly more bending of the DS chainstay near the rear axle to avoid heel rub but with the 450+ long chainstay I plan with the bending here is modest anyway.

Not being able to swap wheels between other non offset bikes is a down side. But this will be my only bike so I won’t be swapping wheels between bikes.
Good point about the truing stands. Hadn’t thought of that. It seems that only having one of the truing stand’s indicator arms at the rim is possible (see https://m.youtube.com/watch?v=JF8p1koGmoM 15m30s). So that isn’t a problem.

Another point I hadn’t thought of. Thanks! With standard mudguards you would need the DS to have longer struts. In my particular case it isn’t an issue because my mudguards are attached directly to the WELDED rack.

Minor solvable issue. See above.

I came across this article today by Wolftooth about chainline and I remembered this thread.

As a bike shop owner, I truly appreciate this :slight_smile:

3 Likes