Cycling UK Forum

Deflecting more laterally than radially chimes with what happens when you true a wheel.....for a lateral correction you only need to turn the nipple fractions of a turn, for a radial correction its whole turns.......I didn't think of that!

So in Brandt's example, the vertical shift is 0.15mm, the lateral shift (for a 135mm wheel) would be 2.4 times that, which is 0.36mm.

So for a vertical load of 50kg with a 1980's single channel rim, you might expect a lateral shift measured in tenths of a millimetre, not hundredths.... my guess was wrong.

But I won't be worrying about a wheel pretzel as a result of a lateral displacement of a few tenths of a millimetre.
In the real world, when did anybody last see a wheel pretzel? Even in the 1980's, with single channel rims, it wasn't common.

The first person I heard of using thicker spokes on the driveside was Hewitt, that was probably last century?

Slight thread bump but I remembered the mentions of Jobst Brandt on here and if anyone is interested the latest issue of Cyclist Magazine (December 2023) has a good profile article of JB.

Mine is a subscriber copy, I guess the December issue will be in Smiths etc in the next week or so.

Cowsham wrote: ↑20 Sep 2023, 5:10pm......This is the place and explanation of why the spoke broke......

had he just straightened the spoke and stress-relieved the wheel, the chances are that there would have been no breakage IMHO

Brucey wrote: ↑30 Oct 2023, 3:06pm

Cowsham wrote: ↑20 Sep 2023, 5:10pm......This is the place and explanation of why the spoke broke......

had he just straightened the spoke and stress-relieved the wheel, the chances are that there would have been no breakage IMHO

Maybe -- some tour all the same.

I've said it before but I'll say it again anyway; heavily dished wheels are basically a crap idea....

531colin wrote: ↑21 Sep 2023, 8:46amSo in Brandt's example, the vertical shift is 0.15mm, the lateral shift (for a 135mm wheel) would be 2.4 times that, which is 0.36mm.

But I won't be worrying about a wheel pretzel as a result of a lateral displacement of a few tenths of a millimetre.

With respect, the original question is about nothing as extreme as a wheel 'pretzel' (I prefer 'Pringle' myself, describes the shape better!), it's about wheels going out of true for no apparent reason other than you've loaded them for touring and ridden rough roads and tracks. That does still happen, quite a lot, even when the wheel was built by a pro.

And with respect to Brandt, his example was for a load of 50kg: a typical amount of weight on the rear wheel of an unladen bicycle when standing still, or being ridden on an incredibly smooth road. Real world roads have bumps however, even potholes, which means exit steps, that produce impact loads many times greater than the static weight. The magnitude of an impact load depends on many things, which obviously include the size and steepness of the step, the speed at which it's approached, the dimensions and pressure of the tyre and the amount of weight on the bike. But not all weights are equal. The elastic properties of the weight itself also matter and the human body has quite a lot of give, especially when one takes one's weight off the saddle and uses one's legs as shock-absorbers. Of that 50kg rear load, the overwhelming majority is rider's 'live weight', plus perhaps only 10kg of 'dead weight' from the bicycle, so much bigger bumps can be ridden than a simple analysis might suggest. Luggage however, adds dead weight, typically as much as 20kg directly over the rear wheel, added to the bicycle's 10kg. This tripling of the dead weight is probably why touring is so much harder on a rear wheel than just riding around unladen. Even if the rider were 20kg heavier, it wouldn't be anything like as hard on the wheel as 20kg of luggage.

So you've got momentary impact loads several times greater than 50kg. But how much greater? Let's work it out. Take a typical touring tyre 32-622 size, inflated to 80psi, as it might well be if you're carrying a touring load and wanting to guard against pinch flats. Pinch flats still happen, don't we know it, so how much force does it take to flatten that tyre against the rim?

A 32mm tyre stands about 32mm above the 622mm diameter bead seat and 26mm above the rim edge. The geometry of intersecting chords says the length of the contact patch is 2xSQRT((622+32)x26) = 261, call that 26cm. A 32mm circle flattened out would be 32xPI/2 = 51mm but the part inside the rim is still curved so lets say the tyre flattens out to 4.5cm wide. So the flattened area is 26x4.5xPI/4 = 92cm2. The force resisting that flattening is the internal pressure (80psi = 5.6kg/cm2) acting over that area which is 5.6x92 = 515kg.

So already, without even getting a pinch flat yet, we've got ten times Brandt's 50kg trying to kick the rim 3.6mm sideways. And now we're riding roughstuff, so these nearly-flat incidents are coming thick and fast. Okay, so a lot of the objects we're impacting are not flat, but rounded, so the impact area and force required to press the tyre against the rim is less. Or else sometimes the tyre does get pressed all the way to the rim. It'll still take a lot more force on top of that to pinch the inner tube hard enough to puncture it, cushioned as it is by two thicknesses of tyre cover.

So I reckon we could be getting half-ton impacts quite often riding roughstuff with a touring load and not necessarily suffer that many punctures. Plus there must be a whole lot of quarter-ton impacts, that kick the rim sideways 'only' a couple of mm. I reckon that adds up to quite a hammering. Are we all a bit less surprised now, that dished rear wheels built with same diameter spokes both sides go out of true?

Brucey wrote: ↑30 Oct 2023, 4:33pm I've said it before but I'll say it again anyway; heavily dished wheels are basically a crap idea....

Not when they're really designed to be dished, by a real engineer.

CJ wrote: ↑30 Oct 2023, 6:01pm

Brucey wrote: ↑30 Oct 2023, 4:33pm I've said it before but I'll say it again anyway; heavily dished wheels are basically a crap idea....

Not when they're really designed to be dished, by a real engineer.

-which is why spoked wheels are so popular on cars these days.
I happen to think that tension-spoked wheels are 'just right' for bicycles unless they are heavily dished.

Brucey wrote: ↑30 Oct 2023, 6:22pm unless they are heavily dished.

I find heavily dished wheels are horrible to build. The differences in tensions on each side are so extreme that it is hard to believe the wheel can have any integrity.

CJ puts in a good word for the various methods you can use to improve things and there are indeed gains to be had. But none of this escapes the fact that heavily dished bicycle wheels will always be weaker and most folk are riding wheelsets where the front is much stronger than the rear. As it happens I think that the OPs wheels probably went out of true because the NDS nipples backed out; [not uncommon with recently built dished wheels] and I'd advise threadlock -
or just wait until corrosion helps you out.

Brucey wrote: ↑30 Oct 2023, 7:51pm I'd advise threadlock -
or just wait until corrosion helps you out.

I do use threadlock.
I doubt I will build another extreme dish wheel.

Brucey wrote: ↑30 Oct 2023, 7:51pm I'd advise threadlock

Do you mean undo each nipple and apply a medium strength threadlocker, or leave the nipples as they are and instead use a wicking threadlocker to pentrate the thread? If the latter, any suggestions for a suitable product? (Loctite's wicking grade threadlocker, 294, appears to be unavailable in the UK.)

CJ wrote: ↑30 Oct 2023, 6:01pm

Brucey wrote: ↑30 Oct 2023, 4:33pm I've said it before but I'll say it again anyway; heavily dished wheels are basically a crap idea....

Not when they're really designed to be dished, by a real engineer.
20230903f008.jpg

It’s a CAR wheel, it’s built like that to withstand lateral loads, irrelevant on a bicycle

slowster wrote: ↑30 Oct 2023, 8:53pm Do you mean undo each nipple and apply a medium strength threadlocker, or leave the nipples as they are and instead use a wicking threadlocker to pentrate the thread?

I have done the former myself, and I've seen/heard of the latter [which I don't think is 100% reliable].

I’m away from home so reliant on my 76 year old memory, which is never a good thing.
However, I’m pretty sure Brandt doesn’t tell us whether his finite element analysis refers to fifties style single channel rims, or more recent hollow box section rims.
As I recall he doesn’t mention hollow box section rims much; there is a diagram , but I don’t remember him mentioning how much stiffer the box section rims are.
I have assumed the finite element analysis refers to single channel rims, because of the date of publication and because that’s what he illustrates.
If that assumption is correct, then Brandt’s analysis doesn’t have much relevance to modern rims