Cycling UK Forum

Here's the rear wheel on one of my bikes. Shimano 7-speed MTB hub, 135mm OLN, Velocity Synergy rim with offset spoke holes. With this setup, there is virtually no dishing required to have the rim correctly centred. The 7-speed freehub takes a 9-speed cassette with the smallest cog removed - I discarded the 11t cog off an 11-32 cassette, giving a 12-32. I just needed to scratch up a lockring intended for a 12t small cog, as the 11t lockring provided with the cassette is too small an outer diameter to engage the 12t small cog. Nine cogs off a 10-speed cassette will also fit if desired.

slowster wrote: ↑13 Sep 2023, 6:02pm That has prompted me to wonder what my expectations should be of 700C wheels on a bike that is mostly ridden off-road on stone tracks and paths, all with between 4kg to 10kg on the rear rack. The wheel was built by Spa with an Exal DC 19 rim, 36 spokes and XT disc hub, so is suited for that sort of use, and I would guess it has done no more than a few thousand miles. I have yet to have any problems with my MTB wheels, but they have much wider tyres and no dish.

Is it the experience of others that rough-stuff with several kg of luggage on a rear rack, i.e. deadweight, is so hard on touring wheels that slight loss of trueness after a few thousand miles is not uncommon, and perhaps even to be expected?

Not at all - I honestly can't recall the last time I knocked a wheel out of true. Possibly my commuter wheels (SS road bike) which broke a spoke on a huge pothole hit and the rear went out of true enough that it jammed on the brakes. However I was only a couple of miles from home so just disconnected the brake and rode it in. Not had wheels "just go out of true" since the very early days of my riding when the only rims available were box section aluminium and in trying to buy lightweight, you accepted that they'd be more of a pain.

All the focus on dish as well I find baffling. Modern rims and hubs are designed to mitigate this as much as possible - offset spoke holes, different size hub flanges. I rebuilt my old CX hubs into new rims a while ago and didn't even consider it* - had the hubs, bought some rims I liked (DT Swiss), ran it all through a spoke calculator and built them 3x with double butted spokes. That's on an 11sp hub for a CX/gravel bike which routinely sees substantial off-road use and (less routinely) a bit of bike packing and it's been flawless.

*OK I "considered it" in as much as I used a dishing tool during the build but at no point did I feel the need to try getting clever with moving spacers or cutting cassettes down or minimising it - it is what it is and it's designed for that.

You can't have it both ways......it can't be unnecessary to re-space hubs to reduce dish, and at the same time be a good idea to offset the spoke holes in the rim in order to reduce the effects of dish; particularly when the rim allows such a small offset.
Playing with spoke flange diameter changes the bracing angle of the spokes.....but not by much.
The big thing is developments in rim extrusion, giving the modern stiff, light rims.
(If only that development wasn't squandered by fitting ever more gears, wheels would last for ever)

531colin wrote: ↑15 Sep 2023, 10:13am In round numbers, on an 8,9,10 speed 135mm hub;
Driveside spoke flange is roughly 45mm from the dropout
Non-driveside is roughly 35mm
7&8 speed sprockets are roughly at 5mm spacing
.....so if you fit a 7 speed hub body to such a hub and transfer about 5mm of axle spacers to the non-driveside, you have as near as makes no difference a dishless (ie symmetrical) wheel.....certainly less dish than a front disc wheel!

(I have never done this....I have never had any trouble with conventional 8,9,10 speed 135mm wheels)

No, I have not made up seven speed on an 8,9,10 hub either. I have only made them up on seven speed hubs. Not sure I would do the same now, as 8 speed is more flexible. For me, 7-8 speed works well, also because I like to use friction down tube shifters (which I don't like with, say, 20 speed.)

Build better wheels?
My commuter does 100 miles a week, 86 off road (roots, big gravel, stoney paths) and the rest on shocking tarmac. 35 weeks a year minimum. 5 years so far and haven't touched a spoke. I build awful wheels, no idea about tension and as long as they straight and nearly round they do. 700c with 35mm slicks.
My play wheels haven't done quite as much but close and they get what most call MTB treatment. Big roots and stones, drainage grooves, jumps, drop offs etc. Usuallly as fast as I can trying to keep up with younger mates. Not touched a spoke. 700c with 35/38mm semi knobblies.

Build better wheels?

Maybe a bit glib but there’s truth in it, though, to be fair, the OP did buy good wheels and them shifting a little is basically bad luck.

Building better wheels isn’t that simple though (one has to understand what’s not working for you) and what a (say) bantam weight rider can use without damage the rest of us probably couldn’t. As a younger man wheels would regularly go out of true and eventually I learnt how to true them myself and then how to build my own, I rarely need to true wheels now because I follow good / better practices - and I spend more time on them than a bike shop could afford to. What changed? All of the spokes are tight and at a similar tension, the rims are (more) ridged, I try to avoid highly dished wheels (YMMV), I don’t use narrow tyres *, I do release spoke wind-up (surprisingly important, IMHO), and I do look at seating the spokes well and using brass washers under spoke heads, etc.

Are my wheels ‘bullet proof’? No, not quite that good but these days my expectation - and experience - is that they will stay true. The wheels on hub geared bikes have proved to be remarkably trouble free, once sorted its rare to need to do anything to them - less time on repairs and more time riding them . I accept that the rear wheel on a derailleur geared bike might benefit from a tweak every six months or so and ‘cause you never know how long a wheel is going to be good for I carry a spoke key in my tool kit - perhaps that answers the original post .

Are my wheels actually used on rough stuff? Yes, they certainly are used on rough stuff but I’m also a sympathetic user too - ask too much of anything and it will either break or need repair.

* Narrow tyres. Over time I’ve realised that tyres need to be wide enough - plus a fraction - to absorb (sufficient) shock from uneven hard tracks (35mm works for me and 1&1/4” has for others) and importantly rims need to be wide enough to be suitably ridged. Much more than that goes unused so is a waste and potentially unhelpful. Fat tyres are more about both not sinking into / rolling over soft surfaces and the greater challenges found in serious all terrain riding. All terrain riding is something beyond rough stuff journeying (were you ride where you can & walk the difficult stuff), it isn’t obvious to all but the two ‘styles’ of riding are different.

Lot of talk here about stiffer rims or offset rims and juggling hub sprockets and spacers to reduce dish, but I very little talk about the most important part of a wheel: SPOKES. The OP has 36 of them and that's a good start, but are they the same gauge of spoke in both sides? They probably are and that's very sad; because in a dished wheel, given the same number* of spokes each side it can't be helped that those on the flatter side must be tighter than those on the more sloping side.

Every wheelbuilder stresses the importance of getting all of the spokes to precisely the optimum tension for the wheel's intended use and the components it's made from. Then most of them make a complete mockery of that by putting exactly the same kind of spoke in each side of a dished rear wheel, which guarantees that half of them will be much tighter, and the other half slacker! The truth of the matter is there's a range of tension either side of optimum that'll do. So when it comes to a rear wheel, the wheel builder's highly-esteemed craftsmanship is all used up in making the best of a bad job, ensuring that none of the right side spokes are too much tighter and none of the left too much slacker. But sometimes the dish is so much or the usage so hard, that the necessary difference in tension exceeds the range of near-enough and no amount of skill can make the wheel last.

The primary aim of "optimum tension" is to minimise the effects of metal fatigue, due to the repeated slackening and restoration of tension every time the area of rim a spoke's attached to is pushed towards the hub, as the wheel rolls against the road. Being too slack - so a spoke momentarily goes completely loose and bows sideways as it passes bottom-dead-centre - is just as bad as being too tight. Either can cause a plague of snapping spokes, either on the left (too slack), or more commonly on the right (too tight), or even both sides of the wheel.

Optimum tension also helps the wheel resist buckling. Too tight and the wheel is vulnerable to spontaneous collapse into a 'Pringle' shape under its own tension. Too slack and the spokes provide insufficient support against sideways loads, allowing the affected area of rim momentarily to be pushed far enough to take on a permanent sideways bend.

But (and here we get to the root of the OP's problem) the rim of a dished wheel moves sideways even in response to purely vertical loads! If the spokes are at a different tension and angle either side of vertical, the rim will deflect sideways as it deflects upwards towards the hub. This is how dishing (together with narrower flange spacing and heavier loading) makes a rear wheel more likely to go out of true, just from riding along. And no amount of tensioning finesse can prevent that. As I said, it's a fundamentally bad design, nevertheless good enough for most purposes, except long, hard use, i.e. commuting or touring.

I know three ways to transform a rear derailleur wheel from a bad job into a good job, immune from the destabilising effects of dish.

1) Reduce the dish by reducing the number of gears and offsetting the spoke holes in the rim. Both these have already been discussed quite exhaustively and both help, but non-standard cassettes are a hassle (plus who wants fewer gears?) and road rims are seldom wide enough to help very much.

2) Shift the chainline further right, either by dishing the frame instead, that is to offset the rear triangle (tricky, but has been done), or making the rear end much wider (typical on tandems). Neither of these are an option unfortunately, unless you're a tandemist or in the market for a custom-built frame. And a wider chainline comes with a wider pedal track (so-called Q-factor), which may not suit some people's knees.

*3) Instead of same number of spokes at a higher tension, simply have more spokes on the right at the same tension! Go look at spoked wheels on old cars, they're often dished too, and that's exactly how they built them. Because those wheels were designed by engineers!

The undeniable engineering logic of dish-matched spoking has fortunately begun to penetrate the minds of the people who design bicycles, but so far this feature is available only on some readymade 'factory wheelsets' for sporting use (a few of which may nevertheless be strong enough for touring, thanks to straight-pull spokes). If you want a rear hub for building up with the overall number of spokes and rim of your choice however, you are stuck with the same number both sides (although I did once design and have made, a flange with twice as many holes that attached by small countersunk screws and the existing holes in the right side of a Campagnolo small-flange hub).

4) Dish-matched spoking is nevertheless also possible with the same number of spokes, but different thickness. It's obvious when you think about it. Thicker spokes are designed for more tension, so put them on the side where tension is inevitably higher. Thinner spokes vice versa, so put them on the left. For optimum dish-matching, the elongations of each bank of spokes, right and left, should end up the same when the wheel is tensioned. Essentially you are matching spoke stiffness to the dish. This way every spoke plays an equal part in maintaining the shape of the wheel. As the weighted wheel rolls against the road, its rim at the point of contact will now deflect purely upwards towards the hub, as spokes on each side slacken by equal percentages of their initial built tension.

But dish-matching does not have to be perfect to help a lot. Simply specifying 13/14 gauge single-butted on the right and 14/16/14 gauge double butted on the left, will be a whole lot better than same both sides.

^^ great stuff, thanks very much CJ.

plus who wants fewer gears?

Probably very few people realise that having a gear or two less on the rear cog (for less dish) is a pragmatic and low cost (£) exchange for having a wheel that doesn’t break and does stay true - well that’s my inexpert experience. I really do like it when my brakes don’t start rubbing and I really don’t like riding home with a broken spoke and wondering whether (before I get there) I’m going to get an unridable wheel.

But dish-matching does not have to be perfect to help a lot. Simply specifying 13/14 gauge single-butted on the right and 14/16/14 gauge double butted on the left, will be a whole lot better than same both sides.

A bit pricey for an old skinflint like me but I hear you and you’re undoubtedly correct. Thank you.

CJ wrote: ↑20 Sep 2023, 2:32pm ..........

4) Dish-matched spoking is nevertheless also possible with the same number of spokes, but different thickness. It's obvious when you think about it. Thicker spokes are designed for more tension, so put them on the side where tension is inevitably higher. Thinner spokes vice versa, so put them on the left. For optimum dish-matching, the elongations of each bank of spokes, right and left, should end up the same when the wheel is tensioned. Essentially you are matching spoke stiffness to the dish. This way every spoke plays an equal part in maintaining the shape of the wheel. As the weighted wheel rolls against the road, its rim at the point of contact will now deflect purely upwards towards the hub, as spokes on each side slacken by equal percentages of their initial built tension.

But dish-matching does not have to be perfect to help a lot. Simply specifying 13/14 gauge single-butted on the right and 14/16/14 gauge double butted on the left, will be a whole lot better than same both sides.

Jobst Brandt's book on wheels was published in 1981. His diagrams show rims made from a single channel; my memory isn't good enough to be certain, but I guess thats what was available then; sometimes with clumsy spoke eyelets, like shoelace eyelets. You certainly didn't get modern deep section rims with thicker extrusion where the spoke holes go, and I was certainly riding 27 x 1 1/4 inch tyres, with the associated narrow rims.
Brandt calculated that a 50kg load at the axle gave a 0.15mm vertical displacement of the rim at the bottom spoke.
When the rim is displaced vertically, if the left and right spokes are not perfectly matched so that their stiffness perfectly matches their tension, there will be an associated lateral displacement. How much lateral displacement will result from a vertical displacement of 0.15mm?
On an 8,9,10 speed 135mm wheel, the driveside spokes are one and a half times as tight as the left side, so the lateral displacement is maybe measured in hundredths of a millimetre for a 50 Kg load at the axle?
I came to the conclusion that it wasn't worth bothering using different thickness spokes for my own wheels, and I have used double butted throughout for quite a few years on my own wheels. I still haven't broken a spoke probably since about the time Brandt was writing....I think my last broken spoke was on a machine built 126mm 6 speed wheel, i know it had a Maillard Helicomatic freewheel, if you remember them!

I'm 11 years retired from Spa, but Spa touring wheels were always built with 13/14g spokes driveside.

CJ wrote: ↑20 Sep 2023, 2:32pm are they the same gauge of spoke in both sides? They probably are and that's very sad

Thank you for that very detailed post. In my particular case (and in Spa's defence), the wheel has plain gauge spokes on the drive side and double butted on the non-drive side.

Drive side plain or 13/14g single butted?

I was very impressed with the Rolling Existence world tour YouTube video where the guy had to replace spokes on his girlfriends / wifes wheel at the side of the road -- trueing it before refitting and carrying on.

Some boy that.

I posted a link on a thread I started about it but there wasn't much interest --- truly remarkable journey around the world on bicycles. Well made video too.

This is the place and explanation of why the spoke broke.

https://www.rollingexistence.com/routes ... -santiago/

viewtopic.php?t=158005

531colin wrote: ↑20 Sep 2023, 4:59pm Drive side plain or 13/14g single butted?

I was not sure about that, but having checked again, I can feel that they are single butted.

In other words, the wheel is like the standard Spa offering below with a DC19 rim, but with ACI spokes instead of Sapim.

https://www.spacycles.co.uk/m10b0s176p1 ... ce-of-rims

slowster wrote: ↑20 Sep 2023, 4:32pm

CJ wrote: ↑20 Sep 2023, 2:32pm are they the same gauge of spoke in both sides? They probably are and that's very sad

Thank you for that very detailed post. In my particular case (and in Spa's defence), the wheel has plain gauge spokes on the drive side and double butted on the non-drive side.

Later checked and clarified as single butted on the drive side.

To be fair to CJ he wasn’t taking aim at anyone but rather making a general comment. In my experience the overwhelming majority of wheels have plain 14 gauge spokes and I have but rarely knowingly seen anything else. What’s become clear is that your wheels are pretty much as good as they come … even the best derailleur wheels need to be trued occasionally.

As an unassociated and general comment it occurred to me earlier today that, when combined, any and all best practises move us towards, and hopefully past, that point at which wheels stay true and don’t break. Things don’t need to be perfect, they don’t even have to be ‘the best’, they just need to be good enough for the purpose that they’re used for .

531colin wrote: ↑20 Sep 2023, 4:16pm Jobst Brandt's book on wheels was published in 1981. His diagrams show rims made from a single channel; my memory isn't good enough to be certain, but I guess thats what was available then; sometimes with clumsy spoke eyelets, like shoelace eyelets. You certainly didn't get modern deep section rims with thicker extrusion where the spoke holes go, and I was certainly riding 27 x 1 1/4 inch tyres, with the associated narrow rims.
Brandt calculated that a 50kg load at the axle gave a 0.15mm vertical displacement of the rim at the bottom spoke.
When the rim is displaced vertically, if the left and right spokes are not perfectly matched so that their stiffness perfectly matches their tension, there will be an associated lateral displacement. How much lateral displacement will result from a vertical displacement of 0.15mm?
On an 8,9,10 speed 135mm wheel, the driveside spokes are one and a half times as tight as the left side, so the lateral displacement is maybe measured in hundredths of a millimetre for a 50 Kg load at the axle?

I also have a copy of Jobst Brandt's book and think it's a pity he says so little about dished wheels, just one diagram (Fig. 18 on p43) to which I can't find any further reference in the text (lack of an index doesn't help). It looks like his finite-element analysis was probably up to the task of calculating how far sideways the rim of a dished wheel deflects when simply pushed upwards, but he hasn't done that. His radial load deflection calculation is for the simple case of a symmetrical front wheel. So we are all left guessing.

I studied and practised engineering before computerised finite-element analysis was reaily available and don't have it now, but I can remember how to use a displacement diagram to predict how far sideways the rim of a dished wheel wants to deflect, solely under the influence of different changes in spoke length on each side. The answer may surprise you. Were it not for the lateral stiffness of the rim itself, it would deflect further sideways than vertically!

First I calculated for the dished wheel in Brandt's Fig. 18, that has left & right spoke inclinations of 7.3° & 3.5°. This results in right side spokes 2.1 times tighter than on the left and a tendency to shift 3.3mm to the right for every mm it's pushed upwards. The 135mm Shimano HG hubs referenced by Colin have inclinations of 6.5° & 4.0° according to the data in SpoCalc, resulting in a less extreme 1.6 times difference in spoke tension and a rim that wants to shift 2.4mm rightwards per mm upwards.

I've used italics and the vague expressions 'wants to' and 'tendency' in the above paragraph because the rim obviously does have inherent lateral stiffness, that enables it to resist this rightward excursion somewhat. But only so long as the resulting sideways deflection remains very small and only if the overall spoke tension is not very high.

In his book (p30 under Wheel Collapse), Jobst Brandt gives a very clear and graphic description of how a wheel can collapse sideways into the classic 'Pringle' shape as the result of an almost purely vertical load. He says:

The spokes in the load-affected zone become slack, leaving the rim with no lateral support. In this unstable condition, the rim can be deflected easily to either side and lateral collapse results.

What he did not say (and perhaps didn't realise since he hasn't mentioned this effect anywhere in his book), is that for similar spokes on each side of a dished rear wheel both to be slack, the rim must already have deflected some way to the right.

Brandt goes on to describe how hoop compression induced in the rim by all the other spokes still pulling inwards, pushes on the ends of the sideways bending section so as to make it bend further. It's like when you push on the ends of a ruler: at a certain load and especially if you give the middle a sideways poke, it suddenly bends sideways. The loaded section of the rim does that, then adjacent sections bend the other way and finally the diametrically opposite section bends the same way.

Wheel builders who've experimented with ever higher tensions will have discovered that above a certain level, a wheel will buckle and collapse spontaneously into the Pringle shape with no external force at all, purely due to hoop compression overcoming the lateral stiffness of the spoking as well as the rim itself.

Surely the last thing you want in a wheel is for some of the spokes that are supposed to hold the rim straight, to be actually kicking it sideways, whenever you simply run over a bump in the road. When you do that of course, the vertical load will momentarily be several times the dead weight supported. Maybe sometimes it's enough to slacken spokes and in that case remember what Jobst wrote:

In this unstable condition the rim can be deflected easily to either side.