137 page PhD Thesis on the Bicycle Wheel by Matthew Ford, Dec 2018

[MikeD]

- I've had more spokes fail than rims.
- Stiffess feels good, and brake rub sucks. Also, a stiffer wheel will last longer.

Spokes breaking is indicative of a wheel that was not properly built; i.e., the spokes were not bent to conform to the hub or were not properly stress relieved. Spokes are only stressed about a third of their yield strength and should last hundreds of thousands of miles, according to Brandt. He explains in "The Bicycle Wheel" that he wears out rims and just tapes a new rim to the wheel and moves everything over, reusing the hub and spokes.

A stiff rim can cause brake pad rub when pedaling out of the saddle. In this case, stiffness works against you.

[bm0p700f]

Actually you side tracking. Stiffer wheels (it has been shown) minimise tension and therefore length changes in the spokes for a given load. pedalling loads will cause length changes in the spokes and increasing flange PCD reduces the size of the those lengths changes (that why some hub have over sized flanges). So wheel that are stiffer can last longer but they need a balance of high radial (pretty much guarnateed unless you have really messed up compnonent selection), lateral and torsional stiffness.

I can build a wheel very well but there are some wheels combination lets take a Mavic GEL 280 32 spoke with sapim race spokes on a rear hub with 57mm flange spacing. There is no way I can make that wheel relaible for all. Why it not stiff laterally - radial stiffness is fine it the side loading that will kill this wheel. Hense wheels that exibit flex can lead to spoke failures. It not just about how well a wheel is built. I refuse some wheel jobs because I know I cant guarantee a spoke wont fail during the life of the rim. I expect a spoke failure on such builds.

If your getting brake rub then it means the spoke cannot keep the rim between the pads. the solution increase the lateral stiffness the spoke provide by the varioous available.

Marin to your thought experiment yes build a wheel symmetric with thin and think spokes and the lateral stiffness should be the same both both side because both sides are coupled. In fact Mr Ford deals with this page 62 of his thesis.

[bm0p700f]

Jugi, when you pass the tension where you get a wave in the wheel the rim has passed it tension limit and a stable wheel cannt be built until the tension has dropped. What matt ford implies though is building the brandt way i.e tensioning to the limit then backing off half a turn is it it not ideal. your better off dropping the tension quite a bit from the limit to maximise lateral stiffness. Incidently Campagnolo build there wheels the brandt way tension to the limit minus a bit. So it cant be too bad. As with all thing there is a trade off.

Mr Ford has also given me new ideas about why 2:1 lacing works. We know it does but explaining why has been diifcult. It also answers some other questions.

What page 63 shows is the difference in spoke tensions for a normal wheel decreases with decreasing DS flange to centre spacing. Hense putting the NDS flange closer to the centre (DT swiss road hubs) increase tension balance and radial stiffness or limtis the drop compared to symmetric hub. This however decreases lateral stiffness compared to an 11 speed hub with the NDS flange further out board. So a DT swiss hub may be more useful on rear wheels where radial stiffness is low or needs to be maximised, as these hubs would provide an useful increase.

It is also important to note and my own far more basic sums agree that asymmetric rims do not affect lateral stiffness much. again page 63 illustrates this. However increasing tension balance is of benefit to radial stiffness.

So 2:1 lacing address this by increasing tension balance and therefore maximises the radial stiffness for that flange seperation. Some may think that therefore equal tension either side is desirable and for a 17mm DS to centre spacing that is achieved by with a 34mm NDS flange to centre spacing. However if the 2:1 hub had a NDS flange 46 to 50mm from centre then the tension balance is lower but the lateral stiffness provided by the increase in bracing angle is much larger than the drop in radial stiffness. Mr ford shows for a normal hub with offset 10mm (ie 17/37mm spacing - DA hub) the drop in radial stiffness compared to a symmetric hub of the same total spacing is not large maybe 5%. o it questionable trying to get 100% tension balance I think.

A 17/49mm spaced 2:1 hub with a rim with a 3mm offset would have ~87% tension balance which is good enough. there will barely be a reduction in radial stiffness there.

I have had difficult in the past trying to explain this about 2:1 lacing. finally I think I might have a better explanation of the ideal hub goemtery Incidently Campagnolo/fulcrum are pretty much there. the shamal from in wheel measurement has hub flange to centre spacing of 17/46mm with a offset rim. Thus a near 100% tension blance (~95%).

Every time I read parts of this thesis I pull more out of it.

[alcatraz]

If campagnolo increased the flange even more they would have even more torsional and lateral stiffness. So by saying "they nailed it" is giving them a bit too much credit, no?

I think they are pleased to cover 90% of their customers requirements with what they've got. If 50 years from now people are putting out 30% more power on average the bora ultra flange would probably grow even more. That's my prediction.

Seems to me they are going out on a limb with their rim stiffness by having those large spaces in their spoke patterns. They are kind of telling us that they prioritize torsional stiffness over rim stiffness. They rely on the rim to be stiff enough (or not stiff enough) to warrant the spaces.

The rim is unsupported in many places. Are the gaps a way to deal with brake rub maybe?

They could have chosen any spoke hole pattern but they chose this.

[bm0p700f]

I was talking about hub geometry not there lacing pattern or there rims. I made no mention of that.

[ergott]

The two considerations with triplet wheels optimized for lateral stiffness (17/46 or thereabouts) is if and when a NDS spoke breaks the wheel will be grossly out of true to the point of being unridealbe. It's not the ultimate deciding factor for most, but worth considering if rinding away in unsupported areas and pushing the limits of the wheel's intended use.

Another consideration is the rim design. With a NDS flange spacing that wide the spoke holes should ideally be drilled accordingly and there should be some testing to be sure the rim design is up to the task of the increased leteral load on it. It's not difficult to address this, but not all rims are ideally suited for this.

I do like triplet rear wheels and would go even further to 17.8mm/52mm if designing from scratch. Thing is, then you have a proprietary rim and hub. Then you are right back to a wheel that isn't easily supported worldwide unless you are Shimano or other major company. Even then, you shouldn't keep changing it around and making it impossible to get spare rims 3 years after a wheel is made. That's not fair to the customer and my biggest gripe with Mavic, Shimano, Campagnolo etc.

[bm0p700f]

ergott, exactly so when my triplet sample hubs arrive in march I will be ordering a few carbon rims drilled specially for this hub. In fact I need to get that order off next month really to see if the bend at the nipple can be avoided. Most 24 spoke wheels are unrideable if a spoke goes anyway. Alloy rims need to have a production run to large too test samples.

If your getting rims made then it easy to keep the ERd the same if the rim design changes slightly. However there is also no need to change rims design so often. The carbon wheels I build have rims that are interchangeable with the first ones I built 5 years ago. also since rim brake wheels seem to be on the decline for me there is little need to continually update the rim. One a rim is picked, thats probably it for a long time. If rim brake builds continue to slide the whole thing will never get off the ground.