maddmaxx 

Does anyone here have any idea of how much of the load bearing forces are carried from the bottom of the wheel to the top by the rim. It would appear that the bicycle wheel is a classic monocoque construction with large parts of the forces being supported by the "skin" or in this case the rim. The spokes are there to keep the rim from crushing.

For the rim to collapse, deformation inward by part of the rim will cause (prior to the point of failure) an attempted expansion outward of other parts of the rim. The spokes prevent this expansion and thus the inward compression also.

Spokes work under tension.........not compression.

bubbagrannygear

Agreed and agreed and agree with everything you'e said........ however - not to dismiss the standing on the spokes crowd too lightly - the crane is holding a platform attached to a lift line. The steady line is affixed to a giant concrete slab on the ground and held in place by a turnbuckle tightened to produce 150 lbs of tension on it. A 150 pound man now steps onto the platform. The crane will not experience an additional load of the entire 150 lbs. Do you agree ? Is this signigicant with respect to the construction and behavior of a bicycle wheel ? and if so, how should it be described ? ( I think thats what Brandt and the "standing on the spokes" crowd are concerned with)

BCRider

Quite right. If the upper cable can stretch to some degree then yes the upper cable may only see a fraction of the man's weight or not seen ANY load increas if it can stretch to where the lower cable goes slack. The lower tension being replaced by the man's weight.

But which of the cables can you remove and not have the man and platform fall to the ground? That would seem to be central to this whole concept of which provides the support.

Let's take our wheel to the same extreme but useable for case of demonstration zero load example. The rider's weight matches perfectly the preload in the spokes along the center of the contact patch. As they reach zero tension you can remove them just as you could remove the lower steady line. So if the spokes don't need to be there to hold the rider up then which are doing the work? Obviously things get messy if you try to roll the wheel with that many missing spokes but that's part of the fun, right?

I agree that the study and concept is an interesting and unique way to look at the loading involved but it just went over the top when it started attributing lifting powers to spokes losing tension. That very loss of tension is what lets the other spokes hold up the load. The interesting part being that it can do it with increases in tensions that don't seem to add up because some of the riders weight is carried by the preload thanks to the de-tensioning of the lower spokes.

While I know that some tempers have maybe come to a simmer over this (mine included ) it's been a great debate and made me think a lot about all this. And I'll take stuff like this over a crossword puzzle any day of the week. Well done to one and all and I trust we can still all get together for that E-beer at the E-pub regardless of our deeply held final beliefs on the matter.

bubbagrannygear

Sounds good I'll E-buy the first round.

The concept of something being suported by spokes does not work for me either,but the point of showing how the crane does not experience an appreciable increase in load, despite the application of an external force was to get everyone thinking about the effect of preloading the rim in compression. If we stick to the simple hanging from the spokes theory, the starting tension of the spokes shouldn't make any difference in the wheel behavior. But we all know from experience that the higher, the initial tension (short of buckeling the rim or failure at the spoke holes) the more durable the wheel. I haven't thought it through, about how to explain it - and in my opinion neither has Brandt - but I think thats what they are trying to do. Just something to chew on besides a crossword puzzel

dabac

That's mainly down to semantics, and everyday use don't quite match strict engineering use. The reason for this seemingly illogical statement is that if you were to stick a tensiometer on each spoke and roll a loaded wheel around you'd see that the the spoke that sees the biggest change is the one between the hub and the ground.
So if that's where the biggest change is then naturally that must be where most of the "work" is being done.

Equally true is that a spoke can't take much load in compression, so of course the wheel isn't really standing on the spoke. But the load for the other spokes is so well distributed that the proportional change is much smaller than for the bottom spoke. And with no change there's no work being done, hence no "hanging" from the top spokes.

DannoXYZ 

The main problem with the analysis on that website you posted is that the model is not set up correctly to begin with, and the resultant data is incorrect. A second mistake is in not using experimental measured data to confirm the model. If he had used a tensiometer on the spokes and actually measured the tension-changes, he'll find that it doesn't match his analysis.

The first problem of the incorrect model comes in assuming that the forces on the upper-spokes comes from the hub, it actually doesn't, it comes from the upper part of the RIM. He assumes the rim is some super-elastic structure that doesn't deform all around and that's incorrect. The rim is actually super-ridid structure that refuses to deform easily and transmits forces all around. The sequence of events is kinda like Cooker's god/devil analogy, but it's not static.

The first event that happens is load pushes down on the hub. This then pushes the bottom of the rim and flattens it. HOWEVER, the next step is where the guy went wrong, he assumes that the rim ONLY deforms at the contact patch. In reality however, the RIM must maintain a fixed circumference. The shorter distance of the bottom flattened section has to go somewhere; it goes to expanding the entire top of the rim that's not in contact with the road. This expansion of the upper-section then goes to increasing the tension on ALL of the upper-spokes. The picture should really look like the right side where the rim maintains the same circumference by expanding the section not in contact with the ground to compensate for the shorter flattened section:



Another way to see this is with the difference in distance between the hub and the spoke-nipples. Obviously the spoke-tension will change with deformation in the shape of the rim and the closer the rim is to the hub, the lower the spoke-tension. The change in the rim's shape changes the spoke-tension and vice-versa. If the spoke-tension really changed like the way he modeled, the rim would take on a flower-shaped pattern with some sections expanding more than others:



The main problem is he's modeling movement of the hub as if it doesn't affect the shape of the rim at all. And he models it as if each spoke acts independently without affecting the rim's shape or its neighboring spokes. When in reality, it's actually the rim re-acting to load and deforming which then results in changes in tension of ALL the spokes. Since the upper-part of the rim expands evenly, the increase in tension on the upper-spokes increase evenly as well; and by a miniscule amount due to having so many more of them than the few at the bottom.

You can test this yourself with a balloon. Hold it up in front of your face with one hand in front and one hand in back of the balloon. This way, you can see pretty much all around the outside circumference. Then push it down onto a table and you'll see that the upper part not in contact with the table will expand evenly.

eddy m

I call BS on this. Have you run your finite element analysis prpgram to come up with those pictures? That was what Ian did. He didn't assume any particular deformation in the rim, nor did he assume an extremely flexible rim. He just wrote a mathematical model of a wheel, applied a reasonable load, and let his FE program determine how the wheel works. The assumptions he made about rim flexiblity look pretty good, and finite element analysis is a generally accepted method for analyzing flexible structures.
How did you develop your model of how a wheel works? Did you actually calculate anything, or is it just based on your opinion? By your model, it appears that the circumference of the wheel INCREASES as load is added. How does that work?
BTW your statement to the effect that the rim expands outward was confirmed by the FE analysis. The only difference is that the outward expansion is uneven and very small compared to the inward bending at the bottom.


em

maddmaxx 

I think that Dano has done a pretty good job in pictographically representing what is happening to the rim during the proposed deflection. Given that the picture has to be an exageration for us to actualy visualize what is happening, there may appear to be an increase in diameter that really isn't there.

Given that the rim doesn't actually get much longer or shorter then the area between the round rim and the deflected rim at the point visualized as deflected is counterbalanced by the area between the round and deflected rim over the rest of the view.

What is not clear to me in the FE examination of the wheel in the article we are looking at are the compressive forces in the rim (I think, neither the bending or angular forces specified in the model) In other words, all of those forces represented as the sum of spoke forces must be conuterbalanced by compressive forces in the rim.

In simplified form, the spoke tension is attempting to collapse the rim inward and the compressive forces acting in a tangent around the rim + some relatively small forces opposed to the bending inward of the rim at any given spoke are keeping the rim round.

I do not see these tangential forces represented in the FE model but I do see them in Danno's pictographic model of the inward deflection creating an outward pressure on the rest of the rim. This is what I was implying in the earlier post to this thread when I spoke of monocoque construction. I believe that there are relatively large forces traveling around the outside (relatively) of the rim that are not accounted for in the spoke and rim bending model.

eddy m

The difference is that the finite element analysis calculates changes in tension in each spoke, while Danno assumed that each spoke changes equally. There is no reason to make that assumption. The FE analysis shows the the big changes in tension, and therefore the greatest movement of the rim from its original position, occur near the bottom. The FE analysis takes into account the bending forces in the rim (you know that because Ian estimated the bending strength of the rim) and there is also another analysis on Ian 's website that models the internal bending and shear forces in the rim. Compression in the rim is straight forward up to the point the strength of the rim is exceeded, but that doesn't seem to be much of an issue with bicycle wheels.
I wish I had an FE program. I'm getting ready to build new wheels and I'd really like to be able to estimate the loss in strength if I build with less than 32 spokes.

em

DannoXYZ 

You can't run an FE analysis on an incorrect model. Take a look at his model here: https://www.astounding.org.uk/ian/wheel/build_wheel. He slides the wheel into wedges and duplicates it. The problem is each wedge is independent and doesn't push on the next. He assumes that the spoke attaches to a fixed point that doesn't move. Then he moves the HUB and measures the change in spoke tension. But that's backwards. He needs to model the rim first and transmit forces circumferentially around the rim and THEN measure the change in spoke-tension based upon changes in the shape of the rim.

If you want to set up a more accurate model, I've got Solidworks w/FEA option that you can try out. The main catch here is that the model MUST reflect empirical measured data. And it's so much simpler to pull out a tensiometer and measure actual spoke-tension changes. That data is real-world and it doesn't match what his model predicts.

eddy m

Danno
You are confusing different things. First, the program you referred to is only a program to work out the geometry of the wheel for different spoke patterns. It's nit a finite element analysis. Second, whether the rim moves when you apply a load at the hub, or the hub moves and the rim stays put, is a frame of reference problem, and has no effect on the results. I haven't seen his FE program, but he has estimated the bending strength of the rim section, so i assume that is included. Finally, with the relatively low changes in tension among most of the spokes, you will not be able to prove or disprove much about the model by measuring with commonly available instruments, beyond the fact that most of th eload is carried by reduction in tension in a few spokes.

em

Randallissimo

Damn! There's some smart fuggers in this forum!

BCRider

Let's say he did include a lot of what has been discussed already. That's fine, let's accept the data from the FE program. But it's pretty easy to see from all the posts here and factors brought up that he hasn't analyzed the outcome and presented it in an integrated way. All his conclusions are based around isolated factors or small groups of factors instead of bringing in the obviously interrelated aspects and looking at how the tensions all around the wheel all work together and what forces induce the changes. That's where we lose a lot of the value of the FE results because we don't then bring it out as a full wheel analysis like it should be where various spoke groups aid or counter each other and why.

Please don't take all this to mean that the FE results are garbage. They are far from that. Even if not quite all the factors are in there it provides some fascinating numbers to look at and ponder. It's the lack of a really meaningful follow through that is lacking.

eddy m

i don't get it. Ian has the change in tension in every spoke, and the change in shear stress and bend stress in the rim. If you add back the pre-load, you essentially have all the stresses in the spokes and in the rim, except for the local stresses related to the spoke attachments and the stress related to tire pressure (which should be a constant as well).
what else do you believe he is missing?

em

nwcdr200

This may constitute a new thread entirely, but from what I've heard the stiffness of newer low-spoke rims largely makes up for the low spoke count, but no FE analyses to back that up here, sorry. My own personal, underexperienced, probably underinformed advice would be to stay above 24/28 spoke count front/back, but that's just what I would be comfortable with building myself.

BCRider

He did fine in the first part. Where he failed was in not fully and open mindedly analyzing the FE program results and offering a more overall and integrated interpretation of those spoke tensions and how it all adds up from a full wheel persective. Instead he chose to use isolated parts of the results to further his viewpoint of hubs standing on reduced tension spokes while misinterpreting or ignoring the rest of the results in whatever way suited his goal. For example he points out that spokes adjacent to the reduced tension spokes are showing higher than normal tension and that tension is pointing down. Well.. yeah... the rim is distorting so the spokes at the point where the rim is trying to push outwards the hardest are responding. This has nothing to do with holding up the load and is only a local force that is locked into the internal web of balanced forces. However this isn't brought out in the conclusions at all. Instead he stops at the observations of tension changes in the lower spokes and a few linked observations in the spokes around the affected area. There's nothing else about how all this affects the rest of the wheel.

eddy m

He identified all the changes in tension in all the spokes, confirmed that the only large changes in tension are decreases at the bottom. He can vary the number of spokes, the rim section and load. That's enough to determine the optimum size and number of spokes, and the preload, for any particular rim.
What he didn't do was apply torque loads or side loads, which are the cause of important failure modes. I don't see any reason why he couldn't do that.
The weakness is that he assumed how the load was applied to the rim. He could confirm the model by measuring the tension change in a few spokes. Most of the tension changes are below the threshold that can be measured, even with the best instruments, but the large changes could be measured.
What else do you need to know to design a wheel? What's misleading? What did he misinterpret? If he didn't discuss every element you want to discuss, you could ask him. Just remember that you normally need to pay someone to do that kind of work
What's your better method of determining the loads, or selecting components to match the load?

em

maddmaxx 

What are the compression changes in the rim that correspond to the changes in spoke tension? These would not be the bending forces on the rim but rather, those that are following a vector around the rim. Several of us have mentioned these forces in posts above.

eddy m

That's a good point. Over loading the rim in compression will cause the wheel to taco. Compression is easy to calculate. You just resolve all the force vectors acting on the rim, add all the parallel components in any direction, and divided by 2. It's easier to visualize than to describe, but I have no graphics capability here.
From that FEA, it looks like the compression on the rim increases by half the load, or 500 Nt for a 1000 Nt load. The compression due to preload is equal to one half the number of spokes, times the tension, divided by pi.

em

eddy m

I'm a heavy guy, so it makes me nervous to ride a 24 spoke wheel. Using a stiffer, deep section rim definitely makes fewer spokes possible, but I'm concerned that the deep section is not much stiffer laterally, and that could lead to a folding failure. I also don't understand if it's reasonable to use fewer spokes on the rear wheel. The total possible load under braking on the front wheel is much greater then possible load on the rear wheel, and the lateral load on the front wheel is probably greater as well. Maybe the rear wheel needs more spokes because half of them are undertensioned, but I'd like to see some real analysis of that.

em

miamijim

So....if your sitting there on your bike and I run up and cut all the spokes above the hubs center line you wont fall over?

Squeazel

OK, gents,

Say you have a 2-spoke wheel, with a spoke above the hub and one below, each tensioned with 100 Kg pulling on the hub, and a rigid rim. Then you apply a 50 Kg weight to the hub. I submit that the spoke above now has 150 Kg of tension and the spoke below has 50 Kg of tension. So is the weight "standing" on the lower spoke or "hanging" from the upper spoke? The answer is both- the sum of the spoke tension is constant over the whole wheel. If you assume that the rim is less rigid but the circumference is fixed, the sum is distributed less symmetrically, but the sum is still fixed and the top spokes take some extra tension and the bottom ones are relieved of some tension. If you allow that the rim can change circumference, well, then you used a rubber rim strip instead of a rim, and I wouldn't ride on that wheel if I were you...

facial

Here's my two cents, a priori:

There must be a specific load such that the bottom spoke has zero stress, but still does not buckle. Thus the bottom spokes, under human-like loads, take little or no load at all; it's like they become dead weight.

In general, the majority of the load (and reaction force) is carried by the spokes roughly situated in the top semicircle, with the greatest tension increase in the top spokes. They should roughly correspond to the tension relief in the bottom spokes, but Danny's geometrical construction shows that to be exactly precise, the top spokes do not compensate for all of the stress loss and the stress distribution is rather much more distributed.

This assumes, with a convenient angular position, that the left and right wheel halves are symmetric, which is not the case under braking and acceleration.

Thirty-six spokes should have a sufficient number of elements/nodes to represent a mostly solid wheel with the interior under tension (not unlike tempered glass) for FE analysis under SAP or Solidworks. If at a later time, I can provide some input with graphics.

cooker

No, the upper spokes don't have to increase tension to correspond to the loss of tension in the lower spokes. They have to increase tension by an amount equal to (the loss of tension in the lower spokes)-(the weight of the rider on the hub).

So if the weight added is 50 kg, and the combined loss of tension in the lower spokes is (guesstimating) 45 kg, the combined increase in tension in the upper spokes is 5 kg.


I can't give you exact numbers but you're certainly wrong. The up and down tension on the hub have to be equal. The force is more likely something like along the lines I guesstimated above: 105 kg on the upper spoke and 55 on the lower spoke.

Yes I will. This is why Brandt's quote as reported is so misleading. He allegedly states the weighted hub "stands" on the lower spokes. That's either engineer talk or (dare I suggest?) troll talk.

In plain language, his point is that when you add a rider's weight to the hub, you don't see much increase in tension in the upper spokes, but you do see a marked reduction in tension in the lower spokes.

VenturaCyclist

Folks,
Read the entire thread. FEA is an engineering modeling tool. You can use it to say that you're standing on a spoke but if you examine the argument you wouldn't build a wheel that way (with the upper spokes tight and the bottom spokes loose). He used semantics, FEA and an arbitrary arrangement of spoke tensioning to present an interesting idea. We have all proved that we've been entertained.

2-3 engineers here have made some very good efforts at presenting the true picture but as you can see with all their mathematics, modeling and measurements they don't know 100%. But they have a pretty good idea. This is proven by the excellent bicycles that these engineering concepts have given us today.