PlunnnggggK! Tak, tak, tak, tak… Not the sounds you want to hear from a back wheel. But better at the bottom of a fast descent, as in this case, rather than the top, or in a curve on the way down. Broken spoke, rear wheel, non-drive side (the overwhelming majority will occur on this side).
It was inconvenient for this particular rider (he had to get transport home having had to abandon the ride) after a few hours of climbing and descending. It also signaled the failure of one of the wheels in a wheelset production I wrote about in the previous post.
So, what’s the problem? Well, it is not going to be simply with the rim, hub, or spokes in isolation. There’s an interaction going on here amongst the three elements that needs to be identified. Or as was suggested to me: a pathology of interaction. Say what? Something is not quite right, that’s for sure. Compatibility across different manufacturing regimes (ok i can accept this technospeak) is the main issue.
I talked to my wheel assembler who suggested that it may need a higher spoke count. The original design was Front 20, Rear 24 2 cross design. So I upped the ratio to 24/28 3 cross design, which was a shame. Minimalism, elegance, simplicity are my guiding principles. Yes, as long as the product is technically sound. That this particular spoke ratio was the problem was not proved. It just seemed the most likely explanation. So, I made the change. Hopefully this would restore the hub-spoke-rim balance.
The next production proceeded and went to market. Everything was fine. And then, the same thing. A broken spoke on one back wheel. Same problem as before. Both breakages represented a small proportion of the overall productions. However there is still an unresolved problem with the particular combination of elements in this design.
Luckily the next production was just about to go through, so I had a chance to backup and rethink. On the day that the hubmaker delivered the hubs for the latest production to the wheelbuilder, I got a call. Problemo. Hmm.
The ‘master’ builder, the boss himself, checked the spokes and hubs. The outside-to-inside spoke threads through the hole ok. Not so for the inside-to-outside. The hole is the tiniest bit too small. This is a problem because of the shape of the hub, with the driveside flange having a larger diameter than the non-drive side.
Aesthetically, I had just about decided to change this, anyway, to equal sizes–the hubmakers will issuing the same model with that as an option, increasing choice in the range. Great, since this is a tried and tested hub, and proven durability is the most important consideration. (I guess good design builds great aesthetics into, or on top of, great engineering–safe and sound is the place to start).
Building a wheel requires exact specs to be supplied to the hubmaker. They have to machine the spoke holes to accommodate the particular spokes. When they are given the spoke specs, they automatically apply a tolerance. Suggested is 0.2mm. The specified spoke hole was 2.5mm meaning an upper limit of 2.7mm and a lower limit of 2.3mm.
This is what they did in the previous productions, and the holes were able to accommodate the spokes during assembly. But it seems that slight movement wears away the paint (powder coat), and go on to enlarge the diameter, after wearing the paint powder-coat off. The measurements are tiny. But even the smallest excessive movement can lead to breakage. It is possible that this was a significant factor in the incidents of spoke failure.
Thus I decided to specify a smaller diameter, 2.3mm, assuming that the hubmaker will err on the side of a +0.2mm tolerance meaning the maximum diameter would be 2.5mm. This would go much further to ensuring an exact fit to minimize the possibility of movement and thus breakage.
The issue was that the hubmaker machined the spoke holes in this latest production of hubs at the lower tolerance limit of 2.3 mm. The upshot was that the inside-to-outside spoke would not quite fit.
Anyhow, the solution was for the hub-maker to re-machine the hubs (photos of this process in the next post) to 2.5 mm maximum. The problem basically came down to one of communication and procedure. I needed to include an actual example of the spoke to be used so the hubmaker could doublecheck the fit. They did not think to remind me; I did not think to go that far. Both sides needed to be aware of what each exactly meant or expected in the application of the tolerances. By clarifying, they could adjust their processing exactly to the requirements of the production.
To date, so far, so good. This particular design would appear to be near the optimum. Time will tell, but I am quite confident the right balance has been struck. And all parties to this wheelbuilding experience have got a bit better at what we do.
Next up, hub manufacturing: a look inside the production procedures and processes of a high-end hubmaker.