FBinNY 

FWIW if this were true, spokes would never fail.

Analyzing what happens to individual spokes, whose tensions will change both up and down is a pointless endeavor.

Wheels are a tension structure and must be analyzed as a system. When loaded, spoke tensions will change such that the net vertical (vector) sum will increase equal to the load. The distribution of those changes depends on the radial rigidity of the rim.

In point of fact, tension changes due to vertical (radial) loads are relatively small. Changes due to side loads are of much greater size and significance.

jsallen

You restate the main result of the analysis made by John Forester in 1980, see https://cyclingsavvy.org//wp-content...ward-pull1.pdf, but you have neglected the slight increase everywhere but at the bottom of the wheel, as well as the much larger changes in spoke tension due to lateral and torque loads, described in Jobst Brandt's book The Bicycle Wheel, published a year later.

Kontact 

I neglected them because they are minor. Not minor loads in total, but minor because they are so distributed across so many many spokes.

Kontact 

Correctly built wheels generally don't have spoke failures. Spokes should outlast alloy rims and hub flanges.

When spokes fail, it isn't because their max tensions were exceeded, it is because some part of the spoke received too many work cycles of the tension dropping too low and then going way up. If the elbow isn't seated, that will cause the elbow the bend back and forth until failure. If the elbow is seated and the tension way too low, the work cycles could break some other part of the spoke, eventually, like at the stress riser going into the nipple. Butted spokes aren't actually stronger, they just distribute work cycles better across the spoke because they spring better than an elbow or a threaded section can.

I've seen more non-drive side spoke failures than anywhere else, and that is because they more easily go from full to zero tension than drive spokes.

FBinNY 

You and I agree on wheel theory 90% of time, but we need to agree to disagree on the low tension theory (because I doubt we'll ever agree on that point).

I consider the theory that spokes fail from low tension a myth unique to the bike world. It persists, despite contradicting known properties of the alloys used for spokes.

Steel simply does not fail from being understressed. It fails either from single overload events, or in the case of spokes, from work hardening and the resulting crack propagation (metal fatigue).

That process happens with repeated loading in a range close to yield, which is consistent with experience of more highly loaded wheels failing sooner. Spokes fail at elbows because elbow see bending loads that accelerate work hardening and fatigue, especially (as you point out so often) if not properly set and stress relieved. The rest of the spoke doesn’t fail because it isn't worked this way.

So while undertension is problematic in many ways, it is not a cause of spoke breakage.

Kontact 

Undertension alone doesn't cause spokes to fail. But changes to spoke shape do, and that can only happen if the tension drops to zero. Without dropping to zero, the spoke can't be physically worked between taught and play. With proper tension the spoke stays under tension and never changes shape.

That isn't an argument for high tension.

FBinNY 

As I said earlier, we mostly agree with this being a point where we don't.

Except in unique situations, slack spokes don't go into compression or change shape as you describe.

Two reasons.

1- if spokes are so slack (untense. If that's a word) to zero out, nipples would be free to turn and thr wheel would slacken and become useless long before spokes.failed.

2- even if nipples didn't unscrew, spokes still wouldn't deflect as you describe, since the nipple would be pushed back. For all practical purposes zero is the minimum possible spoke tension.

Despite disagreement on the details of the failure process, we DO AGREE on the fundamentals, including that once a wheel meets a minimum* tension requirement, additional tension offers no benefit, and may be detrimental.

*that minimum depends on spoke gauge and count, rim flex properties, and payload.

Kontact 

It doesn't need to be pushed into compression. The spoke is a linear spring. It will only be able to take a different shape if there is no tension, and you can't work harden something that never changes shape.

FBinNY 

It's a dead horse, and we'll never agree on it but I question your thought process.

here's what you said in post #56 and what I responded to:
"But changes to spoke shape do, and that can only happen if the tension drops to zero".

My point was that dropping tension to zero doesn't change shape other than let it relax to the zero point. Moreover it cannot unless the nipple is bolted to both sides of the rim. So, while we agree that shape change (aka flex) is the issue, what we disagree on is the mechanism of that flex. Specifically, HOW reducing the tension to zero would change shape any more or differently than working a spoke within the normal tension cycles.

In any case, this can go on forever, and we've both staked out our positions. Maybe one night we'll be the same place and can thrash it out over a few beers.

Kontact 

Just to make it clear what I'm saying, if you build a wheel with a spoke that has a curve to it, once you bring it up to tension the curve is gone. It is gone at low tension or very high tension. All the tension settings make the spoke straight. The spoke stays straight unless the tension goes to zero - then the curve comes back.


I'm not suggesting all spokes that break started out with a curve, but that all spokes are of an imperfect shape. And that imperfect shape is never seen unless the tension goes to zero. Then the spoke changes shape briefly and is flexed to a shape and back. Those are work cycles. And they are likely to happen where spoke has the most reshaping going on - the elbow.

Which is not to be confused with unseated elbows that break - because those are being flexed without ever taking a set shape, so all tension changes cause them to flex - breaking them quickly.

You don't have to agree, but you broken elbow poll has a fair number of NDS spokes for you to discount processes that explain why a low tension, well braced spoke should ever break.

FBinNY 

OK. But becomes a question of relative forces. The lateral force applied by a curved spoke straightening slightly (how slack would it gave to become to straighten any more than very slightly) is miniscule compared to what it takes to work the elbow. I very strongly it would flex the elbow at all.

As for the survey results, I have no preconceived notions. In fact, since you and I agree that failure to properly set and stress relieve elbows is a major cause of failure, and because I believe that lateral flex is the primary cause of fatigue failure, we both should expect a more even distribution.

Otherwise, I'm happy to wait for more data before drawing any inferences.

jsallen

[QUOTE=Kontact;23112440]
"I'm not suggesting all spokes that break started out with a curve, but that all spokes are of an imperfect shape. And that imperfect shape is never seen unless the tension goes to zero. Then the spoke changes shape briefly and is flexed to a shape and back. Those are work cycles. And they are likely to happen where spoke has the most reshaping going on - the elbow."

I had a real-world experience of this. There were loose spokes where the rim had been indented by an impact and trued by loosening the spokes. I could feel them lose tension when riding slowly -- the rear wheel noticeably lurched slightly sideways and back as these spokes were at the bottom. This went on for weeks, then one day, two loose left-side spokes snapped within seconds of one another -- an important lesson about wheel integrity, very early in my career working on bicycles.

But work cycles also occur when the spokes remain tensioned. The issue then is whether they are taken above the fatigue limit, see https://en.wikipedia.org/wiki/Fatigue_limit . Or maybe the rim gives way first around the spoke holes or spoke nipples break. Either way, bad deal!

mpetry912 

Interesting discussion.

No argument here that spokes fail because they are cycled in and out of the fatigue range because of too low tension.

I've had discussions with several people I consider to experts that the spokes see their peak tension as they rotate past the "top" position as the wheel rotates. Spoke cannot carry a compression load but the tension can be reduced which (I believe) occurs at the "bottom" of the wheel's rotation. I wonder if anybody has done a finite element analysis of the tension allocation across the spokes thru 360 degrees of wheel rotation.

Going back to the OP's question, it is darned near impossible to give recommendation for final spoke tension, except to say that for most wheels, it's in the 80-100 Kgf range and shoot for the minimum deviation across all spokes in the wheel. On highly dished rear wheels, increasing the drive side tension to the point that the NDS spokes pull into the lower end of that range seems to work well, at least for me.

But what do I know ?

/markp

FBinNY 

Sums it up perfectly.

sweeks

A few years ago at CABDA Mid-West, I saw a presentation by Bill Mould in which he discussed spoke tensions during loaded wheel rotation. I don't think he did a FEA, but he had an experimental set-up that could measure spoke tension at any point in the rotation. IIRC, his analysis was essentially what you have here ^^.

FBinNY 

There's absolutely no denying that (with axle loading) spokes lose tension at the bottom and gain tension elsewhere, though not necessarily ONLY at the top. The radial rigidity of the rim contributes to determining the distribution of the changes, but we can safely say that the net sum of vertical vectors changes equals the load. In any case, the details don't really matter because these tension changes are fairly evenly distributed and very small compared to the dead load.

It's my considered opinion that all this talk of tension and radial loading misses the point. IMO, Radial loads don't kill wheels, lateral flex is where we should focus our attention. Those interested should search their own minds and think about WHY spokes consistently fail at the elbow. Don't stop with the first and most obvious explanation. Keep going until you have at least 2 separate and distinct explanations. Doing this exercise will help you see the big picture rather than the minor details.

mpetry912 

OK that is interesting. It "seems" to me that the max spoke tensions are at the "top" of the wheel as it rotates, almost as if you are hanging from the ~4 spokes that are at the top of the wheel as it rotates, and the minimum tension is at the bottom.

Another interesting phenomenon is that the spoke tension is reduced when a (clincher) tire is mounted and inflated. When I took Ric Hjertberg's (Wheel Fanatyk) wheel building class, we experimentally confirmed this

Yes very good point ! A lot of these effects of dynamic loads on complex structures are hard to validate experimentally, so we'll keep thinking and observing.

anyway

/markp

sweeks

Yes! And meanwhile, mercifully, spoke breakage is a relatively infrequent occurrence. As well, one or two spokes breaking probably doesn't immediately take the wheel out of service (spoke count > or = to 32?).

FBinNY 

It may be interesting, but how is it meaningful?

Life is full of interesting observations, but we simply note and file them away, tension drop from tire pressure is just one of those.

sweeks

I don't see anything wrong with knowledge for its own sake. Those who don't care don't have to be bothered, but it's not always certain where and when a piece of factual knowledge will be of use.
I teach my dental students that while they are in school (and afterwards as well) they are constructing a "tool kit" comprising both physical tools (explorers, forceps, drills, etc.) and mental tools (concepts and facts). The more "tools" in the kit, the more types of problems the owner of the kit will be able to successfully solve. Some tools may never be needed; but if they're needed and not available... well, there's a saying that if the only tool you have is a hammer, every problem looks like a nail.
(FWIW, Francis, from what I've seen on this site, you have an extremely comprehensive toolbox. This digression was not meant for you!)

FBinNY 

Thanks for the compliment. Like you I have no problem with knowledge for its own sake. You never know what will come in handy down the road.
However, we need to be careful about dragging clutter into our thinking. The tension drop with tire pressure is a perfect example. Interesting to know, but not an excuse for raising spoke tension beyond already established goals or specs.

BTW I always preferred the Socratic method when teaching. I believe the learning process is like eating raw broccoli. If you don't chew it thoroughly, it goes in one end and out the other, leaving minimal nutrition.

sweeks

I thought the inverse relationship between spoke tension and inflation pressure was pretty interesting when I learned about it. Sort of like the shortening of a hollow axle when the QR is tightened. I see these phenomena as fitting into a larger (and nicely coherent) picture of how "solid" materials respond to stress.
Of course, you're right about having no influence on selection of final spoke tension.
OT: Happy New Year!