Building up 9 x 4 spoke pattern rims

[reohn2]

What are the advantages of cluster spoke wheels?

There aren't any beyond fashion

[NATURAL ANKLING]

Hi,
Acquired several on skip bikes, if they have aluminium hubs I normally break them up and keep the hubs.
Or just skip the whole wheel, they came on a cheaper range of bikes some years back, Decades even?
Just had a quick gander I have a set of 2+2 cluster? Whether this is W cluster I do not know.
Two spokes a gap two spokes.

[Brucey]

clusters of two spokes are different to either pattern I showed upthread. If the spokes are close together enough at the rim and the hubs are not drilled in a weird fashion (some are) then the spoke length normally boils down to a crossing value of ~ x2.5 or x3.5 instead of x3, depending on how it is done. Again to be sure you can compare with a conventional wheel.

cheers

[SimonCelsa]

Maybe I'll try this instead, far less complicated:

[Brucey]

... far less complicated:

and far less good....

cheers

[AndyA]

Unfortunately my fashion sense is usually 20 years behind the times!

What are the advantages of cluster spoke wheels?

The only advantage I have heard of is that the staggered gap makes it easier to get a D-lock or thick cable lock through......hardly a major benefit but for a hack bike I think the price is right and if I get a couple of years out of 2 rims I'll be happy.

Maybe that's why they're often seen on modern dutch roadsters - to make it easier to use the nurses lock

[Brucey]

I've certainly seen a lot of wheels from such bikes with heavily 'reshaped' spokes. I'm not sure of this but I think a lot of the damage comes from attempts to move the bike when the lock is already on. If this is the case then it might be a touch counterproductive; a bigger gap means a longer 'run up' and more damage to the spokes...?

cheers

[SimonCelsa]

Well all the components turned up so I spent a quiet couple of hours putting them all together and hey presto;

OK not the best of pictures but the finished wheel actually built up pretty straightforward.
Theoretically the spoke lengths required (by Brucey's custom calc) were 285.2/286.5/286.9 & 288.2mm but that rounded into 27 pcs at 286mm and 9 pcs at 288mm.

The rims started off perfectly flat and didn't really need much tweaking throughout. The braking surface minimum thickness is around 1.8mm.
Quite pleased, will fit them when the Spring finally arrives and see how they last.

They also sell these rims in job lots of 10 for £40 (plus a tenner for postage).

[Brucey]

nice job!

FWIW I was looking at a wheel with a rim like that the other day and it was indeed laced up differently; More or less it had 'W' cluster spoking in an X-cluster rim. Possibly this is what you saw when you mentioned that the outside spokes looked longer in one photo?

Anyway the spoke lengths for an X-cluster rim, W-laced, would be x3.125 (inside spokes) and x3.375 (outside spokes).[edit; where inside and outside are in relation to the cluster, rather than the usual wheelbuilding meaning]

I suppose there is a fourth permutation which is W-cluster rim, X-laced. But this would make the gap in the spoking for valve access as small as possible, so it isn't a very good idea. Anyway the spoke lengths (for a W-cluster rim as drilled in the picture upthread) would be x3.125 (inside) and x2.875 (outside).

Anyway of the possible ways I think the X-cluster rims laced the X-cluster way (as you have done it) look the best of the available options.

cheers

[SimonCelsa]

All good info for the future.

I am trying to understand why these fractional crossings are factors of 1/8 (one eighth). If it were 1/9 (i.e 9 groups of 4) it would make a bit of sense but the theory is a little beyond comprehension (mine anyhow). Perhaps I'll study it more during a dark day next winter.

Cheers

[Brucey]

1/8 crossing values arise because the drillings are (to a reasonable approximation) about half as far apart as normal.

This is almost exact in the case of the X-cluster rims (both yours and the one in my photo) but in fact it is a little less exact in the case of the W-cluster rim I photographed; if this has caused confusion I apologise. It would be exact for the W-cluster rim in the photo if each pair (where in this context there is a pair each side of the valve hole) was exactly half the normal distance apart and still centred on the usual position for that pair, but this isn't quite the case. In fact for this particular W-cluster rim there is a small error because x2.875 is a slight underestimate and x3.125 is also a slight underestimate.



The true values for this particular rim are probably closer to x2.90 and x3.15. (If the spoke is one hole out of position (1/36 of a turn) then this is an error of 0.5 in the crossing value).

Using the same values as previously as an example; (60dia 28offset hub 600 rim)

36x3 = 286.3mm, x3.125 = 287.5mm, x2.875 = 285.2mm, x3.15 = 287.8mm, x2.9 = 285.4mm

so in each case there is an error of about 0.2 or 0.3mm.

This error would be smaller if the hubs were small flange ones (instead of the large flange 60mm dia value used) but equally might be larger with a larger hub and/or a different starting value for the crossings.

so if the hub is 90mm dia then x3 = 280.4mm, x3.125 = 282.2mm, x2.875 = 278.6mm, x3.15 = 282.6mm, x2.9 = 279mm

and we are up to 0.4mm error in length. This might be enough to tip you over into using a different spoke length, but probably not; you are still far better off than if you didn't make any correction at all.

In point of fact it isn't exact but it is normally OK to use a linear interpolation between crossing values. So even with a 90mm hub flange diameter

36x2 = 267.4mm
36x3 = 280.4mm
36x4 = 295.7mm

i.e. from x2 to x3 is 'worth' 13.0mm and from x3 to x4 is 'worth' 15.3mm. If you worked in increments of a tenth of a crossing then near x3 each crossing error of x0.1 (which is about 10.5mm position error around the rim circumference of a 600ERD rim) would be 'worth' ~1.4mm in spoke length.

These errors go pro-rata with hub diameter, so with a small flange hub they are straight away at least half as much. If your estimate is ~5mm out of position (a crossing error of ~x0.05) around the rim with a SF hub, then you are incurring an error of ~0.35mm.

I hope that makes it a bit clearer (rather than the other thing). It is simply a question of estimating how far out of the 'normal' position the spokes will be and translating that into a 'crossing value' that is acceptably precise.

BTW 'acceptably precise' is in relation to the other errors you will encounter in spoke length. It is easy to be blasé about these things but errors in spoke length calculation can just as easily turn cumulative as cancel one another out.

If you are careful about the usual measurements, one of the biggest errors you can encounter is if there is a variation in the J-bend length/head shape and the shape/thickness of the flange. If you always use the same brand of spoke and hub (and they are made consistently) you probably won't even notice this but if you don't, it is very easy to have a head that settles differently into the hub flange and also to have a small variation in the J-bend length. If there is (net) an extra 1mm of J-bend, by the time the wheel has been stress-relieved properly the spoke lengths can be 1mm out. I recently fell foul of this; just using a different make of spoke (even in a rim/hub combination I'd built before) was enough to tip it from 'perfect' to ~1mm too short. Why did I not just go +1 mm in the first place? -well +2 was the closest length I could obtain, it was a single-walled rim and I didn't want to dress the spokes if I could avoid it, and of course if the spokes are ever adjusted they can start out flush with the nipple top and then go through the rim tape. I'd previously built this rim with all the spoke ends finishing mid-slot ( a near-perfect result) and I was hoping to do the same again. Some hope.

Because I was re-using a hub I followed the original spoking pattern which was 'skew' (inside spokes leading on one flange and trailing on the other) which means that the spoke lengths are all a little bit out anyway; the inside spokes should all be ~0.5mm shorter than calculated and the outside spokes are ~0.5mm longer. In a 'symmetric built' wheel (inside leading (or trailing) on both flanges) this isn't seen (because the hub can twist very slightly and this can make the error disappear) but in a skew-built wheel it just adds more errors to those present already. I've (somewhat depressingly) come to the conclusion that whilst the wheel in question would probably work OK, and most folk would be happy if they had built it, I shall never be 100% happy with it and I shall end up rebuilding it with +2mm spokes (the next nearest length available in these spokes, I normally work to 1mm if I can.... ), even though some of them will (unavoidably because of the skew spoking) be poking out of the nipples slightly. So several small things (the J-bend/settling variation, the single walled rim, the coarse increment in available spokes, and the skew spoke pattern) can add up to one big thing and you end up with some of the spoke ends being 1mm below the slot bottom.... The net result of this is (for the first time in -literally- decades) that I shall spend another £20 on spokes, and I shall have to waste time dismantling and rebuilding the wheel... .
Ho hum....

cheers

[SimonCelsa]

I think I get the gist of it, the bit in the middle made some sense. I'll take some time and digest it better later.

Good luck with your re-build, the ACI Alpina DB spokes from cyclebasket haven't let me down yet over several builds, 25p each with nipple.

Cheers

[Brucey]

just today I saw another bike with cluster spoked wheels; a Batavus with rim brake front and coaster 3s rear, it was built x2 front and rear with X-cluster rims/spoking. This may allow closer/fewer spoke lengths to be used (than x3). I'll probably do some sums later.

edit; the two spoke lengths in a 36x2 SFF X-cluster build (x1.625 and x1.875) are only 1mm different from one another and the mean value is only 1mm shorter than a standard x2 build, i.e. it is much easier to build these rims this way, you only need one spoke length. However not every hub will take radial spoking and this is getting close to it (in terms of the loading in the hub flange).

edit2; the rear 36x2 in a Nexus SG-3C41 uses spoke lengths that are 0.8 and 2.5mm shorter than normal (for x2) respectively. Not as forgiving as the front build (mainly because the flanges are larger diameter) but not too bad.

cheers

[NATURAL ANKLING]

Hi,
I have been using Alpina for some time now and happy with them.
They are quite generous on the J bend, so I have washered a few.
Also slightly smaller diameter than most at middle 1.7mm.
I think if you are getting all spoke ends within 1mm of each other at the nipple slot you are doing well, not all wheels even when perfect on the same side same length are that close, dont beat yourself up.

[Dr pepper]

Dont you like me using 2 nipple lengths?

My rims came, I thought they had eyelets, they do not, ah well.
The ebayer ERD spec of 530mm was correct surprisingly.

For the front spokes I calc'd:
244.4mm & 247.2mm, I'm going to try 246mm spokes.

The rear:
257.2mm & 261mm one side, I'm going to use 260mm.
and
258.9mm & 262.6mm the other side, I'm going to use 262mm.

I have printed out the photo, I'm going to colour in the diffo spoke lengths, & put the same colour insulation tape on the actual spokes, then build the wheel looking at the photo, hope thats not too hard to do.
8 of my spokes are a little naff, the galv has come off and theres some rust, might try them & replace with new once I know they are correct.