cyccommute 

You probably wouldn't find much consensus with this statement in the touring crowd. 36 spokes is barely enough for a traditional touring load (not ultralight) and a fairly lightweight rider. Heavier riders with heavier loads pop spokes with regularity on 36 spoke wheels. Going to a 2.3mm butt on a spoke increases the strength of the wheel significantly in these situations.

Ric Hjertberg says that it's like adding 10 spokes to the wheel. I wouldn't go that far but it is like adding at least 4. A 36 spoke wheel built with 2.3mm triple butted spokes (there are more manufacturers than DT) is at least as strong as a 40 spoke wheel built with 2.0mm spokes. A 32 spoke wheel has a strength similar to a 36 and a 28 spoke wheel's strength is similar to a 32, etc. For heavier riders, the 2.3mm spokes are a solution to a very persistent problem.

Kontact 

More spokes are always a good answer for more load - especially to better support the rim. I just don't agree that broken elbows are "normal". They are from bad technique, not a design flaw in bicycle wheels. There's no reason the elbows should be breaking any more than any other part of the spoke, IF they are seated properly in the first place.

Anybody can build a wheel. Not everybody does it correctly.

Ghrumpy

In my experience, NDS spokes are more likely to fail at the elbows, because they are at lower tension and thus can end up momentarily de-tensioning under rim deformation. This cyclic unloading makes the elbows squirm in the flange, work-harden, and become brittle, eventually either popping off the head or just breaking. The higher the dish, the heavier the load, the more this is a problem.

Thinner spokes stretch more under the same tension as thicker spokes, so using them NDS makes them less likely to completely de-tension. I've built several wheels with such a configuration, and rebuilt a few that had had NDS spoke breakage issues. I can't say I've tracked the 20-year long-term success of the configuration, but none of them came back within the eight or so years I was at the shop where I did the work.

It's not really more complicated than having two different length spokes for a rear wheel build, which high-dish wheels always need anyway.

cyccommute 

How ever the spokes break, it is "normal". I would put it as the 3rd most common questions that comes up on these forums. First is derailer/shifting problems, second is chain cleaning but third is spokes breakage. Yes, most of the time it is the result of poorly built OEM wheels but it is still a very common occurrence.

As for the elbows breaking, there is a very good reason for that to happen. It is the most stressed part of the wheel. The 90° bend makes for a huge stress riser and the act of bending the spoke rearranges the crystal structure of the wire such that forces are acting on that stress riser. Spoke seldom break anywhere else along the length because the forced of impact is running along the straight lines imparted by the forging process.

This makes sense when you look at just about any failure of a structure or metal item. Fracturing isn't going to happen except at where parts bend.

cyccommute 

People say this all the time but I seldom see broken spokes on the nondrive side of a wheel...and I see a lot of wheels with broken spokes at my co-op. Only 1 or 2 out of every hundred is broken on the nondrive side.

And, while I agree with your mechanism, I disagree with your conclusions. The drive side undergoes the same detensioning/tensioning cycle. Since they are under more tension, the detensioning/tensioining cycle has more energy release/gain on the drive side which puts more stress on the head of the spoke. This works the spoke bend more and harder so that they embrittle faster and fracture more easily.

TallRider

This is interesting data, and it sounds like you have a fairly significant sample size. Not controlled environment (in that you don't know all the variables relevant to these wheels with broken spokes) but what you relate is significantly against the received wisdom (although idk if that received wisdom is based on solid data). Your argument about the mechanism - that tension cycles stress the metal independent of whether the spoke actually goes slack or not - may be true, especially for wheels where NDS spokes are properly seated/stress relieved as Kontact emphasizes.


This thread drives home to me how little we really know about wheels.
* If thicker-elbowed spokes are less likely to break at the elbows, how have so many people had good long-term experiences with 15-gauge (1.8mm) spokes?
* If stress-relieving spokes at the hub interface is key to avoiding nipple breakage, why does Ric Hjertberg (who has a massive sample size of built wheels to draw on for claims/evidence about what works) recommend spokes with thicker elbows?
* if NDS spokes are at greatest risk of breakage, why does cyccocommute see less than 5% of broken spokes on the rear NDS?

The biggest conclusion I come away with for traditional spoked wheels is that well-built wheels (consistently tensioned but not over-tensioned, stress-relieved) probably hold up well for most riders, whether using butted spokes or straight-gauge. Obviously heavier loads warrant more spokes.

The best answers to these questions would come via controlled studies where the same wheelbuilder builds a host of wheels, with variations in spoke type, build technique, etc., and get large data outcomes where statistical analyses can provide strong evidence for what matters most.

Kontact 

Just because something is common doesn't mean that it is expected or "normal". Spokes shouldn't break in the normal course of things. Nor should handlebars, fork blades or seat posts. The fact that some do should draw our attention to the underlying cause, not cause us to shrug as if it were inevitable.

Build a wheel right and you will wear out the brake track before a single spoke breaks. If you are breaking spokes, the wheel is either built wrong or is being used woefully past its maximum limits.


Most wheels are machine built. They never get the elbows stress relieved. That's the reason handbuilt wheels are usually superior - because all the best practices were followed, not just whatever the machine was capable of.

Hellgate

14g straight with brass nipples. The primary areas of concern is the bend of the spoke, and lateral rigidity of the wheel, especially if you're loaded with gear. If that's the case consider a 40 hole hub/rim for hard corps touring.

cyccommute 

To be honest, I've only heard this claim that nondrive side spokes break more frequently here on the Bike Forums and only relatively recently. I'm always puzzled by the claim because I've seldom seen NDS spokes that are broken nor have I broken many...any?...NDS spokes.

I would suggest that the use of 1.8mm spokes is a rather small sample set. Most spokes are 2.0mm straight gauge. I see only the occasional old 1.8mm spoked wheel at my local co-op. I even took apart an old zinc plated double butted 1.8/1.5/1.8 recently and was rather surprised that those spokes even existed.

I have build a set of wheels with 1.8/1.5/1.8 double butted spokes for my wife but she is tiny and doesn't need the strength nor weight of a thicker spoke. This could be one of the reasons why they may last and why they are rare. They are rarely used for heavy riders. Those that do use them could possibly be using them on a limited basis as well, i.e. for race day riding, and/or treating them more delicately because they "know" they are more fragile.

Heavier riders don't choose them because most heavier riders are already popping heavier spokes so they avoid the lighter stuff. Relatively lightweight riders can ride equipment that is a lot lighter duty without issues.

Forming the spokes around the hub (to avoid confusion with stress relieving following tension) helps but it's hardly a science. As with most things about wheel building, there is a lot of art to it. The spokes undergo good quality control when they are made...and they are made a lot better than they used to be...but the way that spokes are formed around the hub can vary a lot even from wheel to wheel or spoke to spoke when the wheel is made by the same person.

The problem I see here is the supposition that the NDS spokes are at greatest risk of breakage. If that is true, why use a lighter spoke on the NDS side? Having a greater chance of spoke breakage on the drive side makes much more sense to me since the DS is under more tension...about 80% more on a conventional rim. Pulling harder on the crystal structure of the wire would result in more fatigue especially if there is a flaw in the metal.

I have also seen a lot of wheels where some of the NDS spokes are completely detensioned. The spoke is carrying no load whatsoever. The spoke around them don't fracture and break nor do the loose spokes fracture. It's really a nonissue. Additionally, when a spoke breaks on the drive side, the NDS spokes have to carry more load but, again, I seldom see broken NDS spokes. If, as many people assume, the NDS spokes are more likely to break, wouldn't these two scenarios result in more NDS spoke breakage?

For most people even the standard OEM crap wheels will last a long time. I'm often amazed at how well a glorified coat hanger can stand up to the abuse meted out to it without just collapsing into a pile of pick-up sticks.

But for people who have problems with spoke breakage...and they are legion...the triple butted spokes solve a lot of problems and make it so that heavier riders and/or riders with heavier loads can move down the road without having to deal with endless broken spokes.

I think this is the crux of the problem. Most people don't track their wheels all that much. We are depending on anecdotes for a lot of the data. Even if you were to track the wheels, breakage tends to be random and might depend on outside factors. Designing an experiment to collect this kind of data would be difficult, time consuming and expensive.

Just getting mileage on the wheels might be difficult. The "average" rider isn't going to put 10,000s of thousands of miles per year on a set of wheels. Just putting thousands of miles per year on a set of wheels would be difficult. A thousand, 2 or even 3 thousands is possible but more than that goes from enthusiast to professional (paid or unpaid). Among the people I know, I'm a "high mileage" cyclist and it's difficult to manage a 5000 mile year while still working. And not all of that mileage is on a single bike. My highest mileage bike has about 20,000 miles on it but it took 12 years to get there on that bike. I did 37,000 miles over that same period on 14 different bikes. It gets confusing to keep all that data straight.

That does bring up one claim that has always bothered me. There are claims out there that Jobst Brandt rode a set of wheels with the same spokes for 300,000 miles. How could he accomplish that?? If he were trying to break the record for yearly mileage (and failing) year after year, it would take 4 years at almost 75,000 miles per year.

But he had a job as an engineer so he probably wasn't spending all his time chasing mileage records. At the almost Herculean yearly mileage of 10,000 miles (while holding a job and having something of a home life), it would take 30 years to amass that mileage. That's a cycling lifetime. If you only build one set of wheels, and then use them for 30 years, how do you learn anything about wheel building? Personally, I've found that I don't learn that much from overwhelming successes while I learn a lot from even small failures.

cyccommute 

You are drawing the wrong conclusion from what I said. Spoke breakage is normal as in "common" occurrence. Who said anything about shrugging it off as inevitable? Your argument is more along the lines of shrugging off the problem than mine is.

Here is what I mean about "shrugging off the problem". Rather than say that "you just have to do a better job of building the wheel", I say you should build with better components. Most wheels are "good enough" for most people. I'm not saying that wheels fail with any kind of regularity. But there a lot of instances where the wheels are being used past the limits that they were built for. The answer is to either not push them past their limits...i.e. the old "Doctor, Doctor! It hurts when I do this" joke...or build for bigger limit. If spoke breakage is the problem, address the problem with spokes that don't break.

We've reached the same conclusion but I've taken a much different tack then you have.

Yes, handbuilt is better than machine built but that doesn't mean that handbuilt wheels will never break a spoke, especially if the spokes aren't up to the job. On the other hand, if you build with a stronger spoke, you have a stronger wheel and the hand built wheel is even better than it was before.

cyccommute 

That's one way to do it. But 40 hole hubs and rims (or 48) are difficult to find and replace. Read the Ric Hjertberg
article to see a better solution to the same problem with easier to source parts. I've used 36 hole wheels for about 10,000 miles of heavy loaded touring without issues. The thicker headed spokes make a very large difference.

pdlamb

This is my experience also.

ThermionicScott 

IIRC, his commute was a 20-mile loop on nice days, and he rode a century every weekend, along with his annual Alps trip, so he rode about 10-20,000 miles each year. I think he started riding seriously in the late 1950s, and only stopped in 2011, so that adds up. He had plenty of wheelbuilding failures in the beginning, but that set of Campy NR wheels held up once he determined what worked.

If my bicycling career lasts anywhere close to 50 years, I'll be a happy guy.

Kontact 

It doesn't sound like I made myself understood, so let me go back to something else you said:

The spoke elbow is NOT the most stressed part of the wheel, UNLESS it was never bent to properly conform to the hub flange. When that doesn't happen, the elbows are under tremendous stress - but that's bad mechanics, not the way the wheel was designed.


It is analogous to having bolts break because they weren't full tightened. The stresses that should be distributed to the neighboring components become concentrated at a point on the bolt not designed to carry a load that way - so it inevitably fails.

You can certainly substitute a bigger bolt or a bigger spoke head to make up for failing to assemble the parts correctly, but the best solution is to assemble the parts correctly.



This is exactly the same issue as the recent thread about the ovalized caliper mounting hole in the frame. The hole didn't ovalize because the frame isn't strong, it ovalized because the brake mounting nut wasn't tight.

Seat and stress relieve your spoke elbows. This is easier to get right with normal diameter spokes.

cbrstar

I've been thinking...What if frame flex has something to do with stress on the tension of spokes? Certain frames flex less and I'm wondering if that energy is transferred back into the wheel? That would explain why you can have the same wheels and one fails and the other doesn't, and why you never heard of these problems as often on older steel frames.

Kontact 

There certainly could be something to that. If you are essentially trying to flex your bike - like during a spring - and the frame won't, any other place that can flex is going to be doing double duty.


But that probably doesn't have much to do with touring bikes.

ThermionicScott 

What the hell are you talking about? Broken spokes happened all the time on older steel bikes. Maybe it didn't make the news back then, but we're in a golden era of over-analysis these days.

Drew Eckhardt 

With a heavier load, you may need spokes thinner than 2.0mm on the rear non drive side so your wheels don't go out of true as the nipples unscrew when they slacken passing the bottom of the wheel. Trigonometry sets the fixed ratio between tension in the halves, and 2.0mm at 90 kgf to avoid NDS problems could require drive side tensions of 180 kgf which isn't feasible when deep rims are built for 130 kgf and box section 110.

With proper lubrication 1.5mm spokes windup about 1/4 turn on the drive side which makes failure to compensate for that a bigger problem, although a tape flag on representative spokes (first after the valve stem hole on fronts, plus the second on rears) or sharpie dot on each makes this evident.

Otherwise there's no reason to use different gauges.

Some hubs won't properly support 1.8mm elbows on 15/17 gauge butted or straight 15 gauge spokes. If you don't correct spoke lines, 1.8mm threads are more likely to break at the rim. Using 15 gauge spokes lets you try then fail to use 14 gauge nipples.

You might want heavier spokes for lateral rear wheel stiffness if you're a big rider, although more numerous (like 24 or 28) thin spokes work well too.

2.0/1.5mm spokes have become ubiquitous on boutique wheels in the form of Sapim CX-Rays and DT Aerolites. The round Laser/Revolution spokes work as well.

You can use whatever suits your sensibilities and aesthetic senses. I prefer the uniform butting on DT spokes over the step in Sapim and Wheelsmith. I use 2.0/1.5mm Revolutions with alloy nipples because I can.
.

cyccommute 

You can go back to something else I said but you aren't making yourself any clearer. Nothing you say below has much to do with what you quoted above.

WHAT!!!??? Name one other part on a wheel that is more stressed.

Even when properly bent to conform to the hub, the spoke at the bend is the most stressed part of the wheel. Wheels seldom fail to the point of uselessness because of any other part breaking. With the exception of the hub flange, there is just about no other part on a bike that can fail that spells the end of a wheel but the spokes. Rims can be replaced if they crack and or fail. Break even a single spoke and the wheel is questionable. Break two and you are sliding towards complete failure.

And the weakness of the spoke bend is the way the wheel was designed or at least the way that wheels have evolved. J-bend spokes are a pretty crappy idea for something that is meant to be under the tension that spokes are meant to be under. The only problem is that other ways of attaching the spokes introduce their own problems which are worse.

This is not analogous to that at all.

Again with the wrong analogy or the wrong interpretation of the analogy. If you were fastening two plates together would you use fewer smaller diameter bolts to do the job or more larger diameter bolts to do the job. If the stresses are equal, the larger diameter bolts are going to do the job better with less possibility of failure than the thin ones.

And, since there is little weight penalty to using the larger diameter spokes but a huge strength increase, using the larger diameter spokes in a properly built wheel makes for a better built wheel.


Again, that has nothing to do with the topic at hand.

Who is saying that you shouldn't seat and stress relieve your spoke elbows? I certainly haven't. I've said that in a lot of cases that isn't enough. The parts simply aren't up to the job at hand. If the part breaks, get a better part.

As for being easier to get the seating and stress relieving right with "normal" diameter spokes, if that were true, then using spokes with heavier heads should result in more broken spokes, not less. The same number of spokes with heavier head section result in fewer broken spoke, not more.

Kontact 

The part of the wheel more stressed than the spoke elbows in a correctly build wheel are the spoke holes in the rim. Rims either wear out from braking or the spoke holes crack. Which makes sense given what we all know about aluminum, steel and work hardening:




I know you've had certain experiences, but I'm speaking as a professional, and all pro wheelbuilders are going to tell you the same thing - elbows aren't weak if seated, and rims are supposed to wear out first. Brandt said the same thing.


This really isn't a complex argument - either correctly seated spoke elbows break easily or they don't. And pro wheelbuilders everywhere tend to agree with me. Big spokes won't break more because they aren't stress relieved - I just said they are more difficult to stress relieve, not that the difficulty prevents it from happening.


Using increasingly heavier components isn't making a "better" wheel. It is making a heavier wheel to correct for poor assembly methods. A 30 pound bicycle should be quite durable - but it weighs 30 pounds. The point to all bicycle construction methods is to provide the necessary amount of strength, stiffness and longevity without just adding material.

Ghrumpy

First off, it's worth noting that bikes that show up at a co-op are IME rarely the kinds of bikes whose wheels got a fair start. Few if any will not have been machine-made from medium-quality components or worse, and fewer still would have been set and stress-relieved by a competent wheelbuilder. Many will have steel hubs, and I'd say nearly all would have been under-tensioned from the start. Any and all of those are a recipe for accelerated fatigue.

So even setting that aside, note that I didn't say broken spokes occur more often on the NDS, I said broken spoke elbows (and heads) happen more on the NDS. Of course spokes on the DS break more because of overshift damage, but that's an entirely different situation. Broken elbows are practically always from fatigue, which can largely be mitigated by proper build techniques and equipment choices. That's the point of a handbuilt wheel. I'm pretty well in total agreement with Herr Kontact on this.

It's not about energy loss/gain. It IS about tension/de-tension. Where you missed is that spokes that are at higher tension experience the same stress as an under-tensioned spoke, but will not undergo complete or even near-complete de-tensioning.

That's the point of wheelbuilding best practices like optimum spoke tension, crosses/spoke angle, stress-relieving, and seating of the heads/elbows. It allows the spokes to absorb and release the stresses of riding, while remaining below their fatigue limit, and all the while remaining under tension and so not moving relative to the hub flange, which is the primary cause of elbow fatigue and breakage.

Kontact 

Not to make a big thing about it, but a properly built wheel should never be going to zero tension. Even though the NDS has less tension on each spoke, the amount of tension applied to the rim is equal to the DS side spokes. Going to zero tension on the NDS spokes would require so much rim movement that it should also de-tension DS spokes. I would only expect to see NDS spoke failures on really low tension wheels - the kind that flex out of center with minor loads.

What is true is that the DS spokes are closer in tension to the breaking point than the NDS spokes, so if the tension varies 10% on a spoke at 60% of its maximum loading that is probably going to be worse than varying 10% when it is only at 30% of max.



However, spoke elbows aren't supposed to break. Spokes are strong through tension. A properly seated elbow is like a rope going through a pulley. The spoke's tension goes through a bend, but that tension remains evenly through the bend.

An unseated spoke carries the load both as tension and as a cantilevered load. The spoke is attempting to hold itself out away from flange, and tension pulls it toward the flange. As tension varies, the elbow attempts to pull away from the flange and work harden. Spokes are not made of very springy alloys, so getting them to bear side loads like that just damages the metal over time.


If all spokes were fundamentally stressed at the elbows, every spoke that has chain-rub damage at the elbow should have popped. Yet they do not. If the elbow is supported by the flange and isn't in a state of elastic side load it will hold up, even with minor damage.


Spokes break because they aren't seated, or because the tension goes to zero at times. And even when the tension is too low I would not expect the elbows to go first if they were seated.

Ghrumpy

Of course. That was my point about the co-op wheel data set that has a tiny percentage of its wheels in the "properly built" category.
But let's also clarify that we're not talking about the every spoke going to zero tension, since that's total collapse. We're talking about an individual spoke going to zero tension momentarily because of load and consequent rim deflection. Once that deflection passes as the wheel rolls on, the spoke returns to tension. This happens in a split second.

Or on high-dish wheels, where the NDS tension may be 60% or less of DS. In such a wheel, it would take proportionally that much more rim deflection to de-tension the DS spokes. Possible, but not likely.
These such wheels, by the way, are the only ones I have ever felt the need to do the different-gauge spoke patterns on. Or those that have already experienced chronic NDS fatigue failures.

Well, yes, but your second case is more the reality than your first.
Spokes that are in the general range of optimum tension on a fairly standard wheel (around 90-120kgf of tension) are at around your 30% of the tension that will cause them to fail directly from max loading. That's part of what makes that range optimal; the fatigue limit for stainless steel is around half the UTS, so a spoke at 30% has quite a lot of headroom to gain tension (say from torque windup) and still be safely under the material's fatigue limit.
Another part is that it's close enough to its elastic limit that you can give the spoke a set and have it take.
The third part of optimal is that it's stretched enough that it can take up the decrease in tension from loading without going near zero tension.
That's your sweet spot right there.

Your theoretical tension of 60% of max would make for spokes with a short life, since the spokes would be permanently past their fatigue limit.

cyccommute 

There are problems with your supposition. The rims spread the tension out over a larger area so there is less load over a given area. The spoke elbows are have the same force applied over a much smaller area and they have to deal with vector forces that have two different directions. This puts more stress on a weaker part then is experienced at the rim.


Yes, I know you are a "professional". But, as I've pointed out before, I've worked on as many bikes as most professionals have in a lifetime. I've also built a large number of wheels and as a wheel builder I can tell you that any wheel builder worth the title knows that the spokes are the weakest link in the system. Yes, rims can wear out but that's not the end of a wheel.

Spokes will fatigue eventually which is also something that Brandt said. And when spokes fatigue, that's the end of the wheel.

Go read the Ric Hjertberg article. He's a professional builder and makes a very good argument that the 2.3mm spokes are less prone to fatigue.

As for thicker spokes (so as to not confuse the issue with the term "heavier") not breaking because they aren't stress relieved that's 180° counter to your previous arguments. You say that spokes break because they aren't properly formed to the hub. But now you are saying that heavier spokes don't break because they aren't stress relieved.

I would agree that the spokes need to be formed to the hub. I do so with each and every wheel I build and have done so since I started building wheels about 30 years ago. 2.0mm spokes broke regardless. 2.3mm spokes stopped breaking. That's why I suggest them. They are harder to break and thus result in a more durable wheel. Ric Hjertberg agrees.

"Heavier" is this case doesn't mean more mass. It means "thicker". The mass differential is 7g over a set of 36 spokes. That's not a noticeable mass increase for a very noticeable strength increase. And the material is added to a point where wheels are known to be weak.

Kontact 

I was talking about instantaneous drops in tension to zero. They shouldn't happen at all. The amount of tension on the rim should make this pretty much impossible on DS and NDS spokes.


It should take the same deflection to de-tension either NDS or DS spokes. They are both effecting the rim equally, one set of spokes is "working harder" because of the bracing angle, but the rim is getting exactly the same compression from low or high braced spokes. The spokes are different vectors producing the same effect on the rim.

This is like a pulley problem - it doesn't matter how hard you're pulling on the rope, you still have to keep in mind that the weight you are lifting has a real value. The rim is under the same compression regardless of how the spokes are angled. NDS spokes aren't doing less work on the rim than DS spokes.

I was just throwing numbers out there to illustrate the point that DS spokes are under more load, and more is more in terms of approaching fatigue limits. I would always expect a structural member worked closer to its limits to fail before one worked by the same amplitude but around a lower center point.