Spoke failures whilst descending

[Manotea]

I lost a spoke on my front wheel (SON Hub, Open Pro, plain gauge stainless spokes) today whilst coming down Bison Hill, Whipsnade. I finished the ride (65km to go) with a slightly wobbly wheel and avoiding use of the front brake from then on. The wheel is on it's second rim which is sufficiently worn that I have a new rim in the garage ready. Before the spoke failed the wheel was dead true.

Now this has happened before. I lost a spoke on the same wheel on The Dean in March, also on a steep descent. In both cases the rim was decidely warm when the spoke failed, the failures occurred at the same point, where the spoke entered the nipple. Prior to this I cannot recall the last front wheel spoke failure. I've broken a couple on the back but these failed at the hub whilst going over some bad road and not when braking.

So the question(s) is/are, is this type of failure associated with descending, and is it time to replace all of the spokes when I rebuild the wheel?

[Biggsy]

I would definitely replace all the spokes now since more than one from the same wheel has failed - with DT Competition or Sapim Race double butted - with Sapim Polyax nipples if there would otherwise be a bad angle from spoke to nipple.

I don't think the failure is often associated with descending, though anything can act as the "final straw" when a spoke is already about to break.

[mark]

Even if the braking surface of the rim was "decidedly warm", I find it hard to believe that enough heat would reach the spoke to affect the structural integrity. If the wheel was properly built and tensioned then two spoke failures would suggest that the spokes are starting to fatigue and, as Biggsy says, anything can act as the "final straw" and cause a spoke to break. Braking hard repeatedly on a long descent could be the "final straw" in this case. The fact that both spokes failed at the spoke/nipple interface does raise the possibility that the spoke to nipple angle is not optimum. The wrong length of spoke could also be a contributing factor, as too long or short a spoke will not allow all of the spoke threads to contact the nipple threads, placing too much stress on the spoke threads that do contact the nipple threads.

Double butted stainless steel spokes will definitely withstand more abuse than plain gauge, the thin center section will "give" a little to absorb shocks. Sapim Polyax nipples have a rounded seat that can sit in the rim at a wide range of angles, reducing or eliminating any bad angles between spoke and nipple. Reducing the number of spoke crossings can also eliminate bad angles between spoke and nipple.

SON hubs have a distinctly larger flange diameter than non-dynamo hubs, increasing the odds that the spoke and nipple will have a bad angle between them. Rims with a deep cross section can also exacerbate this problem.

[Ian H]

I have two SON hubs, both 28 hole built 2x (despite what Schmidt recommend). They 'look' right - no excessive angles anywhere. I didn't build them over-tight, so the flanges aren't under too much stress (which I presume is Schmidt's concern).

[border-rider]

I bought my first Schmidt wheels, which means I'm stuck with the 3x spoking that was originally used every time I rebuild them.  As Ian says, because of the flange diameter 2x might be better.  3x does give you a bit of an extreme angle at the nipple, and if that's not be properly taken account of in the build it will be a potential problem. Given the failure location, it sounds as if that's the issue.

[rogerzilla]

Yes, reduce the numbers of crossings (even x1 is adequate for a front wheel, where the objective is just to stop the spokes ripping through the flange), use the Polyax nipples or just stress-relieve the wheel VERY HARD when building (use an old crank as a lever to torture the spokes) so they take on a sharp kink where they exit the nipple, and won't flex afterwards.

I've built enough wheels with dynohubs, l/f track hubs and Sturmey-Archer hubs to have encountered this problem regularly.

Incidentally, replace the spokes with double-butted ones - they're much less likely to fail at either end, because the thin middle section can absorb a lot of overload.

[border-rider]

I'm not sure I'd change the crossing number on a rebuild of a SON though.  Mine show very distinct signs of the path of the original spokes, and I believe that the usual advice is to stick with that.  It's a matter of, as you say, managing the nipple exit angle.  I've never had spokes break in a dynamo hub wheel I've built, not even one with a hub brake*, so it must be a decent solution if you do it right.

I agree that for a new build, 2x is better

*though my first SA drum brake wheel  - which I didn't build - broke them like there was no tomorrow until it was rebuilt properly.

[rogerzilla]

Yes, I sold my SON to iddu and bought a new Shimano dynohub specifically so I could change from x3 to x2.

[Biggsy]

even x1 is adequate for a front wheel, where the objective is just to stop the spokes ripping through the flange

I would use x2 or x3 to prevent unscrewing without the need for threadlocker.

The flange might be strong enough to tolerate changing to 2x despite the grooves from x3, but sticking with the same pattern would be the safest option and should be OK with better spokes and Polyax nipples.

[Bianchi Boy]

You do not mention if the spoke broke at the rim or the hub?

I have never had a spoke break at the rim.

I started building wheels a few years ago. This was because my LBS wheels kept breaking spokes at the hub. Since this time I have not had a single broken spoke. Don't know what I did differently. I just took a lot of care and made sure the spokes were aligned correctly and where well tensioned.

I beleive you have boght good quality stuff so my guess would be poor workmanship - build yourself or change wheel builder.

[Biggsy]

You do not mention if the spoke broke at the rim or the hub?

That was mentioned and has been discussed.  Please read the original post again.

I beleive you have boght good quality stuff so my guess would be poor workmanship - build yourself or change wheel builder.

Probably it's a combination of a poor lacing pattern combined with poor quality spokes.

[andrew_s]

I have never had a spoke break at the rim.

It's a regular problem when the spoke and rim meet at a bigger angle than the nipple seat will accommodate (taking square on as a small angle). The end of the spoke gets bent where it enters the nipple, and eventually snaps off.
You get larger angles where there are more spoke crossings, and where the hub flange size is large relative to the rim size. Dynamo, gear and drum brake hubs are the usual culprits, though you'd probably also have problems with a standard hub built into a 4x 20" wheel.

solutions are
a) avoid the problem by using fewer spoke crossings
b) use polyax nipples, which will accommodate larger angles than regular nipples
c) put a permanent bend in the spoke where it enters the nipple

[rogerzilla]

Here's a moderately bad one from the back of my TT bike (Goldtec hub laced x3, which with hindsight was a mistake)



It's held up for 5 years racing/Dun Running/knocking about without the touch of a spoke key, but it is mega-tight.

[Greenbank]

Bit of a smudge mark on that sidewall there Roger (below the T and H of WITH). Tut tut.

[hubner]

c) put a permanent bend in the spoke where it enters the nipple

This should be done on all wheels anyway.

When building a wheel, with the the spokes laced but still loose, bending the spokes near where the nipples are will help to get better spoke line. Also bend the spokes at the hub end, both will also make sure the wheel stays true.

[Manotea]

A set of Sapim race spokes (with brass polyas nipples) has been ordered from Spa Cycles.

Looks like I'm gonna be building my first wheel from scratch!

Thanks all.

[Manotea]

I spoke to soon. My magic calculator suggests I need length 291mm. Spa only offer Sapim Race in 290mm or 292mm. However they will sell me the quantatity I need at half the price of other suppliers, so unless anybody stops me I'll get them in 290mm...

[David Martin]

I have never had a calculator give me the right length. I'd go ith 292 and file them down if neccessary - easier than building the wheel and discovering they are too short.

.d

[MSeries]

I have never had a calculator give me the right length. I'd go ith 292 and file them down if neccessary - easier than building the wheel and discovering they are too short.

.d

That was my thoughts too. I use three different calculators and see which agree (if any)

[Biggsy]

I woud get the ERD from the rim manufacturer, measure the hub dimensions, punch the numbers, err on the short side.

[Manotea]

The variability of my amateur wheel building efforts suggests that if the spokes are 1mm short of spec it will not be an issue.  The rationale for opting for 290 rather than 292mm reflects a concern that on the longer spoke I might run out of thread when applying tension ( there's plenty of room in the rim for longer spoke ends  to protude through the nipple without fouling the inner tube, so thats not the issue). Or have I got it wrong?

[Biggsy]

The variability of my amateur wheel building efforts suggests that if the spokes are 1mm short of spec it will not be an issue.  The rationale for opting for 290 rather than 292mm reflects a concern that on the longer spoke I might run out of thread when applying tension ( there's plenty of room in the rim for longer spoke ends  to protude through the nipple without fouling the inner tube, so thats not the issue).

Agreed, and also I vaguely suspect that spoke calculators don't take tension into account enough.

[Manotea]

Cross post with Biggsy.

Spocalc says 291 and thats the length of the spoke I took out to measure which, looking at it in situ in the wheel rim,  seemed to have the end of the spoke more or less flush with the nipple, i.e., about right. Having said that there were a bunch of spokes which were poking out about the end of the nipple. Whether they are different lengths or just under different tensions its hard to say. Possibly both ( I built the wheel using the original spokes about two years back, and then trued it about 2 months back when I replaced a spoke).

So looks like to go with 290mm

[Biggsy]

Knowing the length of an exsiting spoke helps a great deal.  I'm sure you'd be alright with 290mm.

[Ian H]

A mm either way doesn't matter, and spokes can poke out of the tops of the nipples without any untoward consequences in box-section rims.

Because spoke threads are rolled the threads stand proud of the spoke and there's very little resistance if the nipple goes beyond the thread.