Lateral strength of bicycle wheels

[Eichers]

Originally Posted by CAMPYBOB
115-120 would be better. I have not seen any cracking yet, but I suspect neither the rim or hub would live long at 130 Kgf. I cringe when I look at the 'new' wheel designs.

Don't get me wrong, I like the Aksiums and I can't seem to kill them, but I finally got fed up with their lack of stiffness. Also, the rim alloy does feel really soft, anywhere away from the spoke head seats. I've easily straightened the clincher bead seats and dressed of some burrs.

Campy Sigma or Strada solution aged, heat treated rims these are NOT!

Hi CAMPYBOB, i gave this a little more thought and I think that if you get a wider flanged front hub like one of these (better BA), http://www.bikehubstore.com/SL85W-p/slf85w.htm, and some Sapim CXray aero spokes, http://www.bikehubstore.com/category-s/131.htm, then 100-110kgf should be sufficient.

What do you think

thanks KL

 

[CAMPYBOB]

The extra 10 MM of width would help some with the build angle. They would need to be slotted for use with bladed spokes.

 

[WillemJM]

Originally Posted by CAMPYBOB
"Thank you CAMPYBOB for your reply, very much appreciated"

I just finished 42 years of road training and racing this year. I've posted pics of most of the bikes I've used and abused over the many years I've spent on the road. I built the wheels for almost all of them. I built a bunch of wheels for my team mates. I built and rebuilt trashed wheels for family, friends and complete strangers that were in need, over the years. I really do need to get a pic up of all the wheels on my bench and all the wheel boxes full of wheels I've built and rims waiting for the call.

I may just be an engineer, but I know what makes a bicycle wheel stiff.

LOL, sounds similar except about 8 years ahead this end counting 9 years old riding as a juvenile. Nowadays my friends have so much money, they are part of the throw away society, so I don't build as much as I use to.

Ever got caught up in the DT debacle, around 2000 when they lengthened their spoke bend just after the spoke head to fit their hub flanges better? We all had to figure out what went wrong, with the spoke head breaking off after 1000 miles on a back wheel. DT corrected the issue, took them a few months, a lot of builders switched to Wheelsmith due to that.

 

[CAMPYBOB]

No. I've used them almost exclusively thru the years.

I did, however, get a bad batch in the 1980's that popped heads at the bend. It was only one pair of wheels (Campy Record low flange 32H / 3X to Campy Omega rims IIRC), but I spent a summer constantly replacing spokes in both wheels. The team training ride that I popped two spokes on was the day I retired them as jinxed. Bad heat of steel? Improper wire draw? Incorrect bend? I never figured it out. Easier to get the lineman's nines out and cut them apart and build another pair.

I have to comment that both machine built, OEM wheels and production hand-built wheels are much better quality than 40, 30 or even 20 years ago.

Way back when the Pharaoh and I were getting into racing, building your own or seeking out one of the few wheel gurus were the only options. Even the so called 'high end' wheels that came on bikes were often not well built. Even the best 32 and 36 hole wheels did not have long lives sometimes. Some of those old sew-up rims were pretty soft and even back then the weight weenie instinct had us using 280 gram flyweights on roads more suited to 40-spoke tandem wheels.

There's still some clunkers out there...Mavic's Tracomp failures...Ksyrium failures...Oval's self-unscrewing nipples...carbon failures...

 

[WillemJM]

Originally Posted by CAMPYBOB


Way back when the Pharaoh and I were getting into racing, building your own or seeking out one of the few wheel gurus were the only options. Even the so called 'high end' wheels that came on bikes were often not well built. Even the best 32 and 36 hole wheels did not have long lives sometimes.

Yep, back in the 70's we use to build with galvanized, you either had to be rich, or learn to maintain and build your own.

 

[CAMPYBOB]

I was an impoverished (but still voted Republican!) college student. In 1974 I ordered a custom Schwinn Paramount and the bike came with chrome plated Robergel Trois Étoiles Spokes.

Can you say, "hydrogen embrittlement"?

They broke at the bend in Campy high-flange 36-hole hubs (4X) with such regularity that I usually carried 3 or 4 of each length in my Silca frame pump!

If I had not already learned how to wrench on wheels by then, those wheels taught me how to keep rolling.



The package states in French, "Practically Unbreakable"!!! Hah! I beg to differ!

 

[Eichers]

Quote:Originally Posted by CAMPYBOB .The extra 10 MM of width would help some with the build angle. They would need to be slotted for use with bladed spokes.


Hi CAMPYBOB, no slotted needed, the CXrays are only 2.0mm wide... thanks KL :smile:

 

[CAMPYBOB]

Yeah, after reading the dimensioned pic that hub has a φ2.6 MM (φ.102") drilling. A 14 gauge spoke is φ.064". Plenty of clearance for a 2 MM (.079") blade to pass.

My aero spokes are 3.05 MM (.120") and would require slotting.

Anyone remember hand cutting countersinks and polishing hubs to a mirror finish?

 

[Eichers]

Hi CAMPYBOB, no worries and good luck with your hand cutting :smile: It is a shame that a Metre/Meter and a Yard aren't the same length then both measuring systems could used inter-changeably :smile: Millimetres are way superior for small measurements (3.05 MM (0.120") and decimal usage, but Feet is nice to use for larger measurements like peoples height, etc ... not to worry It would also be better if a 1 Kg = 2 lbs (rather than 2.2 lbs), then both measuring systems, again, could used more inter-changeably :smile: thanks KL :smile:

 

[lukeevans]

Hello all, seems I opened a can of worms here!

Just to update you, I have run a two types of experiments, none of which is going to be conclusive as I could only afford 4 wheels on my budget. I trued two wheels (to the best of my ability - I have trued wheels for repair and built one wheel before but would not say that they were up to a high quality) to a tension of 1638.27 Newtons, and two to 568.98 Newtons.

I then hung weights on them and with the weights available couldn't get the wheels to collapse (had 130KG) Unfortunately I couldn't afford strain gauges and the Park Tensiometer I had borrowed from the local bike shop was awkward to use on 20 inch 48 spoke wheels so couldn't get accurate readings of what was actually going on with the spokes. I then loaded two wheels in an Instron machine and loaded as a point load at the rim. The wheels showed a similar curve, but the failure modes were different. The spokes with the higher tension deformed more with a lower pressure applied than the lower tensioned wheel!

I then went back to hanging weights, this time measuring deflection. The results were similar for both (high and low tensioned wheels) for low loads (up to about 58N) the deflection of the wheel was similar. The lower tension wheel continues to deflect at a steady rate up to about 345N, whereas the rate of deflection of the higher tension wheel is reduced. The lower tension wheel continues to deform at a similar steady rate (slightly increased but not much), but the deformation rate of the higher tension wheel begins to accelerate at 4000N. At 5000N the deflection is the same and beyond this point the lower tension wheel continues to deform at a steady state but the higher tension wheel rapidly deforms. Saying that, we are talking 0.1's and less of a millimeter in difference between the high and low tensioned wheels, but these wheels were not loaded as it would be if somebody was sitting on them, so I think the deflections would be exagerated. The loads were also static loading, so again I think the deflections would be increased slightly, but not by much.

How I am understanding it (which I am not claiming to be right, I guess I'll wait and hear what the lecturers say about my findings) I think the tension in the spokes holds the shape of the wheel, but also brings the rim closer to the Euler's stress. I think this would explain why the tension in the spokes gives initial strength, but once the deflection goes beyond a certain point the wheel buckles due to Eulers buckling formula.

This would explain why the angle of the spokes makes a wheel laterally stronger, as it is being tied down like a tent (think of trying to hold a stick up right using guy ropes), so a wide flange hub would indeed increase the angle and make a stronger wheel.

I would have liked to try with more wheels and with different spoke gauges, but I cant really see why they would make much difference, other than greater ability to absorb energy, so I don't think in my static loading scenario it would make much difference.

Using Eulers buckling theory would also explain why deep rims are stiffer, having a greater second moment of area than a standard rim. So for laterally strong wheels a rim with a high second moment of area would be good.


I'm in the process of writing it up but thought I would share with you the raw results as soon as I had them

Cheers

Luke

 

[dhk2]

Luke: Good stuff! If I understand correctly, you used hanging weights to put a lateral load on the rims. But I'm confused when you refer to 4000N and 5000N, after stating you were limited to 130 kg of weights. If those are lateral rim loads, my first observation is that your 20 inch 48 spoke wheels are much stronger and stiffer than a typical 700C road wheel. What gauge spokes did you use?

You mention that above 5000N the higher tension wheel began to deform rapidly. Was that due to the spokes yielding or rim deformation? When the load was taken off, could you tell that the overload resulted in a loss of spoke tension? One of your results then is that the higher-tension wheel was actually weaker, ie, less able to withstand lateral loads.

Perhaps your report will answer all my questions. When you complete the report, would be great if you could post a link here.

 

[WillemJM]

Thanks for posting.

In the Instrom machine, did you have the point load on the rim, and the weel supported at the hub?

When you talk about failure, are we talking about just elastic deflection, or did you actually permanently bend the the wheels?

 

[lukeevans]

Hi to answer all questions

I was both hanging weights from the rim and then loaded in the Instrom to destruction.

When hanging the weights, the axle was clamped in a chuck (think it was from an old milling machine, I borrowed it from the workshop). at first attempt the axle bent (they were cheap wheels but it was still a 14mm axle). I then knocked the cups out and put a substantially stronger bolt through. The wheels were loaded up to 130kg which deflected more than I could measure with the dial gauges. The strange thing was taking the weights of for both the high tensioned and low tensioned wheels the spokes returned pretty much to their original tension (can't be exact as the park tensiometer only just fit in to measure the spokes so couldn't measure at the same point for each spoke) and when put back into the truing stand the wheels were still true.

The spokes were 2mm gauge.

I then point loaded the wheels in the Instrom to destruction, where the machine was run until the rims were permanently deformed (plastically). In hindsight I would not have point loaded them, and would have spread the load over a section of the rim in a way more akin to that in real life where the tyre is in contact with the ground. the wheels where held using a high tensile bolt through the hub. I would also have liked more time to test them , but the machine is fully booked for a while now and the work is due in soon.

My results, and I wouldn't suggest in anyway that they are right as I only managed to test two wheels to destruction, even if I had managed to load all four it would not be very conclusive, but it does suggest that from the two tests that a higher tension wheel is ever so slightly stiffer (has a higher capacity to resist deflection), but a lower tensioned wheel is slightly stronger.

The only conclusion I would make is that it is worth a more thorough investigation into the lateral strength of wheels for load carrying three/four wheeled human powered vehicles. This has neglected a lot of things, and is an extremely simplified model of what is happening. Would love to get the option of carrying this on for a phd.

I'll happily post my report up once its finished and been marked (the uni uses some online searching tool for checking against plagiarism, although it would be hilarious to get done for plagiarizing myself, my supervisor has suggested not testing the system)

 

[dhk2]

Thanks for answering my questions so quickly. Agree that the point load isn't realistic, but at least it serves as a worst-case in exploring the lateral load limits for the test wheels. Most realistic of course would be to have a tire mounted and generate the lateral load via the tire contact patch.

Certainly don't risk anything by putting out your paper on the internet before it's all graded and approved for publication. Good luck with your future studies!