mihlbach

I've seen a few wheels built with a 2/1 drive side/non-drive side spoke ratio. Fulcrum wheels are like this and I've also seen it done with normal wheel components, such as a 32 hole WI hub with a 24 hole Niobium rim. This pattern is easy to spot because all the rim holes are used but half the nds hub holes are empty.

I understand the potential benefit of having equal spoke tension, but it seems to me, a traditional 1/1 spoke ratio rear wheel might be more stable, despite the uneven tension. Sounds like a 2/1 ratio would be a big PITA to true and if you happen to break a nds spoke, your wheel is going to be seriously out of alignment..so bad that there is no way you could ride it home. Is a wheel with n spokes and a 2/1 spoke ratio going to be stiffer stronger and/or more durable than a wheel than a wheel with n spokes and a 1/1 spoke ratio?

I'm curious because I'd love to build a wheel like this, but I'm just not convinced there are any real benefits.


Thoughts?
Waterrockets?

ElJamoquio

Tightening up the NDS spokes will make them less likely to break. Of course, as you surmised, if one does break, the effects will be worse, too.

Personally, I'm not worried about the hub; the DS holes don't pull through, so a similar amount of tension on the NDS is no big deal.

Bottom line... risk/reward is 'worse failure mode'/'failure mode is less likely'.

waterrockets 

True, the failure mode is worse, but I don't know that it's more likely. The NDS spokes will experience much less fatigue, so they will last longer. When traditional symmetric wheels fail, it's often the NDS spokes that are fatigued, and that wouldn't happen nearly as fast with the 2:1 spoke count.

The hub would be a concern if you are lacing the NDS radial, and the flange isn't designed for that. I know that Shimano (at least in my '06 hubs) didnt' warranty radial on the rear. Still, there's no reason you couldn't lace 1x or 2x on the NDS and avoid the problem.

I want to build a wheel like this so bad that it hurts. The 32h hub/24h rim combo is perfect for it. Unfortunately Saris pooh-poohs assymetric lacing on PowerTaps, so I didn't get to have any extra fun with my last build.

I haven't ever trued a wheel built this way, but I don't see how truing it would be any more painful. Symmetrical wheels have a huge tension difference that should complicate truing, but it works out fine. In an asymmetric wheel, the NDS spokes would have twice as much influence, so a 1/4 turn should have the same effect as 1/4 turn on two NDS spokes in a symmetric wheel.

I also haven't seen one post of a Fulcrum NDS spoke failure. Maybe there has been one?

ElJamoquio

Basically, a 16+8 spoked wheel will have less chance of fatigue when compared to a 12+12 wheel.

waterrockets 

Yeah, I think I misread your first post. Although, I actually think a 16+8 will have less chance of fatigue than a 16+16, and should be just as strong. I have no data though -- just going by feel.

ElJamoquio

I was a little confused, and figured you were responding to the OP, WR.

I'd have to run through some numbers, but in fatigue (as it seems you know), having a force that reverses itself is a huge no-no, which is why high tension is a good thing. 16+8 virtually guarantees this will not happen. The DS spokes, though, will have more stress on them when compared to a 16+16... but obviously the NDS is the problem with that setup. So at first glance, I'd agree with you, that a 16+8 would probably be more durable from a spoke-fatigue standpoint than 16+16.

waterrockets 

I'm missing why the DS spokes would have more stress on them in a 16+8 vs. 16+16. You're still going to tension them to 110 kgf, and the NDS would be pulling just as hard against DS, just with fewer spokes. I'm probably missing something, since I like this lacing pattern so much that my vision is cloudy

ElJamoquio

While you're building it, the stresses would be the same.

I was speaking about when you actually ride it. (Although you did made me think for a minute, WR.)

Let's simplify to a 2-D model of the wheel (circle, looking at the wheel from the side). Ignore the rim for a minute; but there's a load on hub - ~70% of your weight. You now have either 24 springs or 32 springs. The linear elastic model shows that pre-loading a spring does not affect the spring constant; we'll also assume that there is always *some* load on the NDS spokes, whether 8 or 16. So... basically you have more springs, and 'deflection' in each will be less if you have 32 spokes. If you build them to the same initial tension either way, then more deflection = more stress = more chance of a DS spoke breaking.

I think this is a moot point, though, because *any* failure represents failure, (i.e., weakest link). With a 16+16, the weakest link is the NDS spoke. With a 16 + 8, who knows... I'm guessing it's still the NDS spoke, but either way, the weaker of the two (16+8) will probably still be stronger than the NDS 16+16.

mihlbach

Thanks guys...this is all very useful.
I recently built up a 14+14 (2x) White Industries/Niobium30/Sapim CXRay rear wheel. While the wheel is fine, I regret not trying to do a 16+8 wheel instead. I imagine an 18+9 would be possible too, if you go radial on the nds.

mihlbach

Also should note that on the 16+8 builds I've seen, the nds spokes were 1x with the spoke heads all facing inward..makes sense to me, because you want the nds spoke tension to be high, but this probably prevents the nds tension from exceeding the ds tension.

waterrockets 

Hmm. I see where you're going with that spring model. But what about this?: For a given spring, the force exerted goes up with displacement. So, to compensate for the DS forces, half as many NDS spokes would have to stretch twice as far. This means that they'd be exerting twice as much force, and would resist loads twice as hard.

I guess we could look at it this way: If you and I are holding Pcad up in the air, and you're doing 60% of the work, what do you care if I use one arm or two?

waterrockets 

The problem with 18+9 is finding a 27h rim. I guess you're talking about using a 36h rim, and having spoke triples, similar to the Campy asymmetric wheels. This would leave half the NDS spoke holes empty on the rim. Hmm. Yeah, that should be good too.

mihlbach

Yeah, I wasn't being very clear, but thats what I meant.

ElJamoquio

OK, we're holding Pcad up in the air, with three arms. Then, botto comes from out of nowhere and drops an anvil on top of Pcad.



Wait, where was I going with this? Oh yeah.

All of our arms are going to buckle a little. If you were using both of your arms, our four arms would buckle a little less, compared to the three arms.

Hocam

Velocity will drill as many or as few holes as you want, if your LBS is a dealer they can get you a 27 hole rim.

I think a 32 hole wheel with an asymmetric rim will be stronger than a 2-1 lacing, but obviously 8 spokes and nipples heavier and limited to the 2 asymmetric rims available (that I know of). With the asymmetric rim you're getting 16 non drive side spokes somewhere around 80% of the drive side tension, which is pretty good. Compared to 8 spokes at what probably ends up close to or above 100% of the driveside spokes, I think the 32 spoke build is stronger in terms of load distribution on the spokes.

If I wanted a deep section rim and wanted to go weight weenie style then 16-8 sounds like the best choice, but I'm no weight weenie.

carpediemracing 

I've built such wheels in the past, maybe 10-12 of them, not for weight reasons but simply because I realized really quickly that even a 28H deep section rim was way overbuilt for road riding. Also it was easier buying a 32H hub than a 24H hub (or for that matter, a 28H hub at that time). Ultimately my goal was to reduce the number of spokes to increase the (perceived) aerodynamic efficiency of the wheel, nothing to do with weight.

Likewise I stopped buying 36H wheels since they seem overbuilt. My mountain bike has a 28H front and 32H rear rims, no eyelets, and 1.8mm straight spokes. 32H is even sort of overbuilt but it's just easier to find hubs and rims for it.

I prefer 28H since the spokes cross at the seam. 32H they sort of pull the seam apart. Not sure if it makes a difference with welded seams but for the plugged seams I felt like they helped.

Even with a 2:1 or 16:8 set up, the NDS spokes were very loose. They were tighter than if I had built an even wheel, but they were definitely looser than the DS spokes.

I don't get why the NDS spokes would be more likely to fail so I'll have to re-read the argument. However, in actual experience, I found that the DS spokes fail more often (and not due to the chain dropping into the spokes). This seems to be due to the higher tensions the DS spokes see and the resulting lower amount of elasticity available for additional loads. If you have a rubber band that's already stretched really far and you stretch it more, it'll break easier than a rubber band which isn't stretched as much to begin with (given same age of rubber etc).

cdr

waterrockets 

Still, say pcad weighs like, I dunno 300 lbs, for the sake of argument. If you're responsible for 200 lbs of that load, why would you care if I use one arm or two for my 100 lbs of that load?

Botto's anvil will get banned before it hits, and there will be a botto-shaped cut-out in the white garage door.

jamiewilson3

Great discussion so far. Now is the time for a mechanics engineer with some free time to jump in with a good FEA analysis.

CrimsonKarter21

That would be 11.4 from the Track Forum. The guy is brilliance.

Psimet2001 

FEA is for wussy. FBD is simple enough to solve the problem. Thank God I am too drunk to attempt a solution at this point though. Didn't read it all, but any loading condition that reduces the tension differential between both sides, and prevents unloading of the NDS spokes is will last longer.

Discussing failure modes at that point is like discussing what exit route you will take when the building you are in explodes.

ElJamoquio

True.


But it's statically indeterminate, so of course you do have to use elastic theory as well. Perhaps I will solve over X-mas.

Window, baby. Window.