Why do my spokes keep breaking? - Bike wheel science.

[silva]

New stainless steel spoke. New brass nipple.
[Take care when googling.]
See this thread for more: https://www.cyclechat.net/threads/rusty-nipples-ok-corroded-but-wheres-the-drama-in-that.198848/
I'd be rather surprised if the eyelets on your rim were steel. So the 'rust' you report is not from those.
A magnet is your friend.
Consider replacing all the nipples with brass ones, while you have the tyre off.

https://www.customcutspokes.co.uk/product-category/bicycle-spoke-nipples/14g-spoke-nipples/
https://sheldonbrown.com/wheelbuild.html
"Many good quality rims have "eyelets" or "ferrules" to reinforce the spoke holes. The best double-wall rims have sockets which spread the load to both layers, allowing these rims to be lighter and/or stronger."

Rim without eyelets	Rim with eyelets	Rim with recessed spoke holes and no eyelets	Rim with recessed spoke holes and eyelets	Socketed rim with eyelets
Rim cross-sections
View attachment 730327	View attachment 730328	View attachment 730329	View attachment 730330	View attachment 730331

That magnet is an idea, chosed a small one, less than 1 cm diameter.
What I named "inserts" and you name "eyelets" are confirmed as non stainless steel: the magnet sticks, even when upside down = hanging and shaking. It is indeed rust, no dust or so, and it also showed that typical expansive deformation of rust hence I took it for non stainless without a magnet test.
You say that that would surprise you, it is the case nevertheless, what was your reason to not expect it?

What I also discovered with the magnet, to my surprise then, that the spokes differ in material. Some are stainless (A2 grade, magnet hangs very slightly, easy to shake off) and others steel, magnet sticks firm.
About a quarter are stainless, the others not.
And that also confirms their different looks, when I cleaned the spokes, some started blinking, others not. I assumed it was due to different ages, but it's due to different materials.
The one that sat in the broken nipple is a stainless one.

Apparently the dealer originally sold the wheel with steel spokes, to afterwards replace broken or suspected ones with stainless ones. The distribution stainless/non stainless appears random.
Stainless steel has quite different mechanical properties than steel, and getting a wheel straight is all about balancing tensions. How does that work out when the materials differ?

 

[silva]

About the broken nipple, I have its two pieces in front of me, externally I don't see any damage or sign of wear. It's broken a couple mm inwards from the "flange" that acts as support against the tensioning.
It's not magnetic, not the slight of A2 / 304 grade stainless. A4 / 316 is totally not magnetic. Certain treatments of A2 / 304 can cause it to loose its diamagnetism. So I don't know. On sight there is like a yellowish but hard to be sure, which maybe indicates brass. Maybe a magnifying glass on the break surfaces tell something.

The spokes and nipples on my travelbike rear wheel I do know since I asked and got a parts list / price offer for my spare rear wheel.
The spokes are named as "Sapim DB Race", which duckduckgoes to "double butted stainless steel"
The nipples as "Nippel Sapim Secure" with an addition that is translated to "locking agent ; sealing agent ; lock compound ; locking compound ; locking device"
Duckduckgo shows

Sapim Polyax Aluminium Nipple 2mm - Secure Lock. Sapim's anodized Polyax nipples are treated with a special coating to increase durability and allow for easier wheel building. The nipples have a 2 mm thread for rims with Ø 3.5 mm holes. Aluminium saves weight, compared to other materials like brass.

During the 7 years I used the travelbike, I have had not a single spoke break/losening/whatever, despite the first Ryde Yura rims outer wall broke in the middle all around, and the inner and sidewall also started cracking, rim replaced by an EXAL like this older/current bike now, didn't even know upto this story it was an EXAL.
Based on this experience, I would say the Sapim spokes and nipples did a quite good job.
The comparison is not fair though, since a 62 mm wide tyre has much more air in it, so absorbs shocks more thereby saving the spokes that. That was one of the reasons to want 62 mm on my travelbike, whose frame and setup allows two different tyre widths so I could have chosen in a range 42-55 mm instead.

The currently mounted rear wheel in the previous/older bike also has a mix of stainless and non stainless spokes, tested 8 sequential and found 2 stainless so it looks like a same story.

 

[rydabent]

3 in 3 months! Ridden for 3 years prior with no broken spokes. I get them fixed by lbs within a few days after each one has gone.

Shizuoka hoy 1 bike, Commuting down canal towpath 8 miles each way with paneers with clothes and lunch in them. i weigh 87kgs and haven't put on weight recently.

Anything I should be looking for before just replacing another spoke? is it easy to do for a fairly hands off cyclist who usually uses lbs for much more than puncture or minor tweaks? Should I look at new wheel?

I have never broken a spoke. In the last 16 years sicne I have retired, I have ridden over 25,000 miles on a recumbent or a trike that you cant "unload" going over a bump or pot hole. Im sure the 5 wheels I was riding on were computer built. In all those miles, I have never once had to put a spoke wrench on one, and they spin absolutely true.

 

[silva]

I neither on my travelbike, 7 years.
The rear rim, and now the frame, turned out to be the weakest parts on the path that the load travels along.
On this older bike I'm now riding on, it is the opposite, I never had a rim or frame broken. Not on earlier bikes either.
Apparently spokes or spoke nipples were there the weakest parts in the path.
One would expect the latter is the better turnout, since replacing a spoke or nipple is alot less costly than replacing a rim let alone a frame.
And it's not that I put on bigger loads on the travel bike, I did and now do the very same with the current older bike.

What i wonder, and asked before here, why is it that the average tension of the spokes of the wheels of the older bikes is quite lower than the one of the wheels of the travel bike? What determines which average tension a wheel builder choses?

 

[Yellow Saddle]

What i wonder, and asked before here, why is it that the average tension of the spokes of the wheels of the older bikes is quite lower than the one of the wheels of the travel bike? What determines which average tension a wheel builder choses?

Average tension on a wheel is not determined by a wheel's age. It isn't as if the old days they did this, and now they do that.
We experimentally determine that when building wheels. The optimum tension for any wire-spoked wheel is the maximum tension the rim can handle before it collapses. This used to be determined by overtensioning the wheel until it collapses, then undooing the spokes again and settling on a lower maximum.

No rim manufacturer that I've come across has ever stated what the maximum tension should be but engineering principles, such as those I've explained in this thread, are the determinining factors.

Obviously wheels are different and will have different optimum tensions. Here's some examples.

1) Rims see spoke tension accumulatively. At 1000N of spoke tension, a rim with 32 spokes sees 32 000N in total and a rim with 28 spokes sees only 28 000N.

2) A wheel with a low spoke count - 28 and lower, cannot be tensioned optimally because fewer spokes contribute to the overall accumulative tension. therefore each spoke has to be tensioned more on a low-spoke wheel than on a high-spoke wheel. In these situations the limiting factor is galling in the nipples. The nipples seize at high tension. Therefore, wheelbuilders keep the tension below optimum.

3) Shallow, box-section rims buckle quicker than deep-V rims. Therefore, they can't accept very high tension. An exaple of where this presents a problem is a 36-spoke Mavic MA2 rim that needs high tension because of the application. It cannot be done. In that situation you'll choose a heavier rim.

4) Rim weight is in direct proportion to rim strength. Reason being that thehy're all made from the same material and thus, more weight is more material.

Now that you have the background, I'll answer the question.

A wheelbuilder's experience from building successful wheels with various, recognisable rims.

 

[Ian H]

About the broken nipple, I have its two pieces in front of me, externally I don't see any damage or sign of wear. It's broken a couple mm inwards from the "flange" that acts as support against the tensioning.

Sounds like (some of) the spokes are a little too short.

 

[presta]

No rim manufacturer that I've come across has ever stated what the maximum tension should be

Mavic do.

Mavic Technical Manual, 2002.

By the time I discovered that, I'd been using 1kN (mean on the rear) for years, and had had no trouble, so I saw no need to change.

 

[Yellow Saddle]

Mavic do.

View attachment 732063
Mavic Technical Manual, 2002.

By the time I discovered that, I'd been using 1kN (mean on the rear) for years, and had had no trouble, so I saw no need to change.

No they don't. That information there is meaningless because it does not take into account spoke count. Like I said, spoke tension is a function of spoke count when attempting to know the upper average tension limit. I can tell you now that a 28-spoke wheel with only 700N of spoke tension will go out of true routinely.

Further, the divisions on a Hozan tensiometer mean nothing on their own. You have to plot the divisions for various spoke diameters in a curve that then gives you a reading in N. That bit of Mavic information is utter crap.

 

[presta]

it does not take into account spoke count

Yes it does, different rim, different number of spokes, different tension recommendation:

the divisions on a Hozan tensiometer mean nothing on their own

Which is precisely why it's quoted with a tension as well, and not on it's own.

 

[silva]

I don't see rim spoke hole counts at all there, that 1st yellowish number 10 is a spoke package quantity, not a total rim quantity. The given tensioning forces are "per spoke", regardless their number.

 

[Yellow Saddle]

Yes it does, different rim, different number of spokes, different tension recommendation:

View attachment 732430


Which is precisely why it's quoted with a tension as well, and not on it's own.

I like a good sleuth. You are right....in this instance. This is a proprietary wheel and yes, Mavic does give all the stats in this instance. I was referring (mostly in my mind because I didn't specifically specify) to rims intended for home and shop builds. In other words, standard 28/32/36 hole rims. I have come across Mavic rims (probably Open Pros) with a spec that refers to all three spoke-count variants, leading to the problem I cited upstream in this thread.

That's why wheelbuilders over-tension a rim and have it harmlessly prezel and then re-doing it with just-just less tension that what would make it collapse. Once you have determined the max tension, you can then calculater the total force on the rim and from that, what the tension should be for other spoke counts.

I wish rim manufacturers will supply us with that information but they don't (unless you are going to prove me wrong again).

 

[Yellow Saddle]

I don't see rim spoke hole counts at all there, that 1st yellowish number 10 is a spoke package quantity, not a total rim quantity. The given tensioning forces are "per spoke", regardless their number.

Yes, but that information is useful to the builder for that specific wheeld. As I said, that's a proprietary wheel that cannot have more or fewer spokes and therefore the information supplied is useful.

 

[raleighnut]

I like a good sleuth. You are right....in this instance. This is a proprietary wheel and yes, Mavic does give all the stats in this instance. I was referring (mostly in my mind because I didn't specifically specify) to rims intended for home and shop builds. In other words, standard 28/32/36 hole rims. I have come across Mavic rims (probably Open Pros) with a spec that refers to all three spoke-count variants, leading to the problem I cited upstream in this thread.

That's why wheelbuilders over-tension a rim and have it harmlessly prezel and then re-doing it with just-just less tension that what would make it collapse. Once you have determined the max tension, you can then calculater the total force on the rim and from that, what the tension should be for other spoke counts.

I wish rim manufacturers will supply us with that information but they don't (unless you are going to prove me wrong again).

The thumbs of a decent wheelbuilder are generally a good judge of tension.