tgot 

With all due respect, I believe that you are mistaken. Were your argument to be correct, one could ride a wheel with no more spoke tension than is required to keep the spokes in place.

The tensioned wheel is not a tensegrity structure, but rather a prestressed structure: https://en.m.wikipedia.org/wiki/Pres...idge%20systems.

There may be rims strong enough to bear you weight with just the top-half of spokes populated, but I do not believe that it is true for the many simple box section rims widely used. They are only strong in compression, and only kept in compression by the strong tension in the full 360deg of spokes.

I believe Jobst's underlying results from finite analysis are valid and interesting: That the largest *change* is stress between and unloaded and loaded wheel is in the spokes at the bottom. And that rim deflection inwards in that region causes compressive force, which the spoke can handle as long as that force is less than the pre-tensioning, so that each spoke remains overall in tension.

RiddleOfSteel

Glad my post could be the catalyst for your comment. The more implied tack I took with it was ride feel and not strength, strength assessment, or testing. Rims have their place in the wheel strength equation, as do spokes of course--I think we can agree on that. I concentrate on the rim as a definite contributor to the ride quality component, at least in the case of the aforementioned TB14 build because I've done an A vs B comparison using the same hubs and spokes, just MA2 vs TB14. The highly butted spoke + TB14 build yielded an undesirable ride quality, and re-rimming that wheelset with MA2s made the ride quality (using the same spoke tension and same tube/tire/pressure setup) good.

I, too, am a Clydesdale rider, mostly by virtue of height. Still, I can lean on a set of wheels and frame, though I am not a gorilla nor do I try to be. I have broken one spoke in all of my riding, and it was a wheel that had been built for some time. I don't know the spoke tension on it as I had bought it as built (7400 hubs, 2.0-1.7-2.0 spokes, MA2 rims), but I knew I didn't want to start chasing the rabbit of spoke replacement, especially as a rear wheel, so I rebuilt it. The rim was also experiencing cracks at the eyelet in a few spots. Took me several years to get to those points of failure, but I didn't mess around.

Anyway, I am glad that there are people that test component strength and get to the bottom of what really drives success or failure and where stresses are. I'm not that guy, nor will I be that guy. My comment culled remarks from the OP and several others and used in service of observable ride quality/feel differences, especially in my current in-progress wheel build.

*****

Perhaps more relatedly to spokes' strength in the wheel build equation, a year ago I had a Trek FX drop bar conversion build where I initially used DT Swiss R23 wheels @ 1600g or so. R460 rims (469g each, 23mm wide external, parabolic profile), their slightly heavier aero/bladed spokes, and R350 hubs. Ran that wheelset with 35s and 42s. In the name of cutting more weight and seeing what ride quality difference it made, I changed the R23 wheels to lighter (~1500g) Shimano RS81s--basically a Dura-Ace 9000 C24 wheelset with a heavier freehub and likely spokes--the carbon-over-aluminum rim was the same, as was the spoke count. Ran it with the same 35mm tires, may have even done the 42s. The differences were quite stark. On 35s, the RS81s absorbed the road imperfections better, but on anything involving medium to high speed turns--whether sweeping or more quickly executed--the RS81s gave troubling handling characteristics. Unstable and unpredictable throughout the turn. R23s didn't flinch, and neither did any other vintage wheel build with those tires previously or since.

This is probably an obvious horses-for-courses situation, with the RS81s built for road/race tire sizes, light weight, and great manners exclusively, to the detriment of 'flexibility' (or adaptability) when someone like me decides to introduce large tires to it. The R23s are also a road wheel, but they are heavier overall, with slightly wider rims and a higher spoke count. Spoke strength, rim strength (or at least rim profile), and spoke count, IMO, certainly are in play here (in this comparison). RS81s deliver a much nicer ride at 25-28mm tire sizes vs the R23, as do most vintage wheelsets, which is plenty good, because the R23s look cool and handle a 48mm tire (on my '09 Tricross) just fine.

cyccommute 

This dissertation essentially argues exactly that. “Contrary to both popular belief and expert consensus, increasing spoke tension reduces the lateral stiffness of the wheel…” The dissertations is a very deep dive and I haven’t finished reading all of it and digestion will take much longer but it is a well know phenomena that too much tension on spokes can result in a spontaneous “tacoing” of the wheel. There’s a balance needed and Matthew Ford measured the balance which he discovered is much lower than many people think.

From the same source

This description better describes a bicycle wheel than the definition given for a prestressed structure

This article states straight up that the wire bicycle wheel is the most common tensegrity structure in use. It goes further to say that

It would be more correct to say that a tensegrity structure is a special prestressed structure.

To be fair, the article as says

Or, in a better parlance, neither idea is completely right or wrong. But he also

This is more in line with what I’ve said above than that the wheel is “standing” on the lower spokes. There is nothing to “stand” on.

I may not have stated that exactly but have certainly implied it. Perhaps I’ll test it (I still have the half spoked wheel) but I have little doubt that I could mount a bike with the spokes on the upper half of the wheel but certainly not roll it. In static load, I’m reasonable certain that it would hold my weight without collapsing. Any lateral movement would, of course, cause it to bend.

On the other hand, putting the spokes on the lower half of the wheel would result in instantaneous collapse when any load over that of the bicycle (and perhaps not even that) were placed on it. I won’t be testing that!

I fully agree that the largest change in the tension on the loaded vs unloaded spokes but Jobst (and others) have interpreted the decrease in tension as an increase in compression. Only the rim is compressed which results in the reduction of tension but there is no compression on the spoke. Tension and compression are opposite acting forces but they are not force that act together. Decreasing tension does not result in increasing compression. Tension has to reduce to zero before compression can be applied. That’s what the quote about the weight of the rider not being translated through compressive force to the rim is all about.

cyccommute 

Yes, I understand that the J-bend is undergoing shear but the shear is the result of the tension on the spoke imparted by the way the wheel is built. Without the tension on the spoke, there would be no shear on the spoke.

And, I’ll point out again, that there are very few measurements of any kind when it comes to spokes and spoked wheels.

noglider 

I have proven this to be possible by building wheels with minimum tension and riding them. The question isn't whether you can ride them. The question is, how long will they last? At low tensions, I believe the change in tension when riding it is greater, and that greater change leads to quicker fatigue.

And speaking of the rare case of a spoke breaking in the middle, it happened to two or three spokes in a wheel I had. I mentioned it to an expert, and he told me that they must have been caused by faulty metal. The wheel was out of warranty, so I rebuilt it with a new complete set of spokes.

tgot 

The failure mode proposed by Jobst for low tension spokes was *wear*. That undertensioned wheels have the spokes go slack at the bottom of rotation as tension goes to zero, and that "forces negative to tension in the direction parallel to the spoke" cause it to move slightly at the J-bend interface with the hub.

Photos of various failure surfaces under a microscope were posted with descriptions of the grain-level evidence for different failure modes. I didn't have the metallurgy knowledge to judge.
judge.

seedsbelize2

I've also always heard that we're not supposed to mix straight gage and db. If this is truly an option, then I likely have the sizes I need, in hand.

ThermionicScott 

It seems intuitive that one would want to use the strongest spokes, but that doesn't necessarily lead to making good choices for a durable wheel. I am sure that motorcycle spokes would perform far better than anything we use, in one of those "strain to failure" tests. But it should be obvious that they would be the wrong choice for bicycle wheels -- you'd weaken the hub flanges by drilling out the holes bigger, and you'd crack all but the heaviest rims by bringing the spokes up to the tension necessary for them to stay put.

So, like with frames and handlebars, we're not choosing the strongest parts available, but ones that are strong enough for the job, but not unnecessarily heavy. And it turns out spokes can be pretty skinny and still accomplish that. I have over 8,000 miles on a wheelset built with 1.8-1.6-1.8 spokes, and I'm starting to feel that even those are overbuilt!