Wheel Dishing Dilemma

[531colin]

Its got to be tighter than that!
Much tighter! 100kgf average of both sides

[axel_knutt]

I'm dishing a wheel again (with Park Tool WAG-5) but there's the usual problem that seems to happen every time...

Spokes that aren't tight enough as it is (on the non-drive side) need to be loosened to send the rim the right way.
- and/or
Spokes that are getting too tight as it is (on the drive side) need to be tightened to send the rim the right way.



Doing either of the above will fix the dish, but doing either is wrong. The only "solution" could be to deliberately dish the wheel slightly off center? I don't know how else it can be fixed. Help! Isn't it more important to have spokes tensioned properly than it is to have your wheel exactly central?

Either your gauge is incorrect, or your definition of too tight or too loose is incorrect. If the dishing is correct, the ratio of left to right spoke tensions must be correct, because it depends only on the wheel geometry. It is not possible to change the tension at one side only without changing the dish as well.

I wrote this about 18 years ago for someone who claimed a wheel builder had taught him how to dish a wheel without using different spoke tensions:

The length (L) of the spokes is comprised of three orthogonal dimensions, one radial (R), one tangential due to the cross pattern (X), and the lateral one (W) due to the width from the centre to the flange.
For the drive side, by Pythagoras:

Ld^2 = Rd^2+Xd^2+Wd^2 [1]

Similarly for the non-drive side:

Ln^2 = Rn^2+Xn^2+Wn^2 [2]

Since the tension acts along the axis of the spoke, the three orthogonal components of the tension are proportional to the dimensions above, so multiplying eqs. [1] & [2] by a constants of proportionality, Kd & Kn:

Td^2 = (KdLd)^2 = (KdRd)^2+(KdXd)^2+(KdWd)^2 [3]

and

Tn^2 = (KnLn)^2 = (KnRn)^2+(KnXn)^2+(KnWn)^2 [4]

For the rim to be in equilibrium, the lateral force to one side has to equal the lateral force to the other, so:

KdWd = KnWn [5]

therefore:

Wn/Wd = Kd/Kn [6]

Dividing [3] by [4]:

Td/Tn = KdLd/(KnLn) [7]

Finally, substituting [6] into [7]:

Td/Tn = (Wn/Wd)(Ld/Ln) [8]

So you can now see that the only factors affecting the tension ratio are the ratio of the spoke lengths, and the inverse ratio of the width of the flanges from the centre line of the wheel.
With the same spoke pattern both sides, the lengths are likely to match to within 0.5%, in which case the expression will approximate to:

Td/Tn ≈ Wn/Wd [9]

Taking my wheel as an example, the tension ratio is 1.75:1 calculated approximately from flange spacing only, or 1.742:1 calculated exactly from spoke lengths as well.

Changing the spoking pattern from side to side will affect the tension ratio in as far as it affects the length of the spokes, but again, since this difference is only small compared with the difference in the flange spacing, the benefit is pretty negligible considering the loss of torsional stiffness with radial spokes. On my wheel, the tension ratio would be 1.67:1 if I were to use radial spoking on the drive side.

[Manc33]

After more tinkering with the tensiometer and listening to the pitch to get them evened out a bit more, I tightened the whole lot and ended up with NDS average of 107 kgf and DS average of 83 kgf. When averaged that's 95 kgf across both sides.

I rounded off a nipple, so don't want to mess with it anymore.

Since a solid tyre is going on this (that isn't going to reduce the tension like a pumped up pneumatic tyre) I'm just going to leave it as it is.

Don't forget this isn't a rear wheel, it's a front disc wheel hence the NDS is slightly tighter