bloke

I'm sure this has been asked a hundred times, but I have read just as many contradictory answers. Which is the correct way, radial lacing on drive side or non-drive side?

HillRider

Theoretically, the cross lacing should be on the drive side since radial spokes can't transmit torque. If I were using a small diameter hub shell that's what I'd do.

Realistically, the cross lacing is often done on the non-drive side because a stiff hub shell will transfer the needed torque adequately to that side with no significant distortion. I have a pair of Shimano WH-R560 wheels and the rear is laced radial on the drive side and 2X on the non-drive side and there is no problem because the hub shell is quite stiff.

carpediemracing 

A friend of mine worked for a wheel company for a long time. He is about the most inquisitive person I know, to a fault. He learned and relayed to me that the drive side should be radial. I don't remember his explanation.

Having heard that, I still build uneven wheels with the non drive side laced radially - easier to get spokes in proper lengths that way. Maybe next one I'll reverse it.

cdr

Dan The Man

The point is to even out the tension of the spokes on the drive/non drive side right? Because the hub is not symmetric with a cassette on it, the drive side spokes are inherently carrying more tension. Since I would imagine radial spokes carry less tension, you put them on the drive side to balance things out.

Metzinger

Since Dan just buttoned this up so well, I'll sidetrack if I may.
I've had a set of wheels built on Velocity OC rims. They appear to have almost no dish. I abuse the heck out of them, but have done almost no corrective truing. Why do I not see everyone building with some form of off center rim?

San Rensho 

What I heard was to radial lace the non drive spokes to avoid spoke breakage. The biggest cause of spoke breakage in modern high dish wheels is not spokes that are too tight, but those that are too loose. On the non drive side, when you apply torque tothe wheel, any lacing pattern will make half the spokes tighten and the other half looseen. Its this cycle of slack/tight on spokes that already have very little tension that eventually causes them to break. All the spoke breakages that I can remember on my wheels are the non drive, non pulling spokes, those that slacken when torque is applied to the hub.

With a radial pattern, all the spokes tighten when torque is applied to the hub, so you eliminate the problem of spokes going slack and breaking.

Dan The Man

Thanks, I didn't actually know if that was the case or not, but it made sense.

Sounds suspicious. The additional bending stress from deformations you would be getting by trying to torque radial spokes would far outweigh any cyclic loading. Not to mention that you would lose a lot of energy into a bendy wheel. If one side is radial and one side is leading/trailing, then that side just ends up taking all of the torque. The radial spokes are pretty much unable to transmit torque except under large deformations.

ultraman6970

It is funny but i have read this thing in both ways too. Another reason why some wheels are radial at the drive side is because if you do the crossing at the drive side with the dishing the cassette or the RD will touch some of the spokes.

Have a set of wheels with radial at the drive side and no problems so far anyways.

HillRider

Rear wheels have been built with crossed spokes on the drive side for over a century without interference problems with the cassette/freewheel or rear derailleur so that can't be the rational for radial drive side lacing.

AEO

all heads out on the drive side or all heads in on the non drive side would reduce dish.
all heads in and radial on the drive side, in theory, would reduce dish as well as increase lateral stiffness of the wheel in one direction.

you can't really get all the heads on the same side unless you use a 1x or radial pattern.
but you can't really use 1x unless the wheel is a 16h.

ultraman6970

Dunno dude just tell that to easton. The wheel came like that from factory. I have used wheels both ways and i have never noticed the difference with my 200 pounds. I rear long time ago in two different web sites about this and both websites came with the opposite solution. Until today have no idea what's the right way to do it because in both cases both authors make a lot of sense to their arguments. One was sheldon and the other web site was another guy who built wheels.

rschleicher

For what it's worth, the Mavic Aksium rear wheel on my bike has radial spokes on the non-drive side, and 2x laced spokes on the drive side.

Yet Mavic's more expensive Kyserium models use their so-called "Isopulse" lacing, which is the opposite - radial on the drive side and 2x laced on the non-drive side.

My conclusion from this is that there are arguments to be made for either approach, and that in the end it can be done either way.

conspiratemus1

Those who have experienced broken crossed spokes on the left side of high-dish wheels need to rebuild their wheels with radial spokes on that left side and see if they stop breaking spokes as the late Sheldon Brown predicts they will.

zacster

Easton wheels are a different animal, but I don't think they are radial on the drive side. They are radial non-drive and 2x drive side. The main difference with Easton is the twin threaded spokes, which have no heads as they thread into the hub and have none of the tension at the bend. I just built up a wheel on an Easton hub with these spokes and a Velocity OC rim and found it very easy to get right because of this. I've put about 500 miles on it so far on NYC and Seattle streets and not a wobble to be seen.

AEO

another method is to do a 2:1 triplet lacing.
brings the left side spoke tensions to about 90% of the right side.
downside is truing becomes harder, but the whole point of triplet lacing is to get a wheel with balanced tension.

Mr. Fly

As far as I can remember, this whole radial-on-drive-side thing was started by Mavic when the Ksyriums were introduced. Of course, according to Mavic's ad copy, this "Isopulse" thing was the cat's meow. But really, Mavic had to do that because their fat brick-in-the-wind aluminum spokes were too thick to be laced crossed on the drive side without running into cassette and/or derailleur clearance issues.

So essentially, radial lacing on the drive side is a workaround prettied up and sold as a "feature". I mean, why would anyone want to add weight to hubs so that it is strong enough to transfer torque to the non-drive side, when the drive side flange is perfectly capable of doing that already without any reinforcements? It makes no sense at all in terms of design optimization.

Sheldon Brown discussed the merits of half-radial (that's with radial on the non-drive side) here.

davidad

Radial spoking is a style thing and has no merit in the strength of a wheel and makes little difference in the weight.

HillRider

That's got to be because of Easton'e hub design and has nothing to do with the lacing of more conventional hubs. My Shimano WH-560 rear wheel also came laced radial on the drive side but their hub design is anything but conventional.

AEO

yet another method of reducing dish, is what can be found on ritchey, or AC hubs, where the NDS has the flange sitting very close to the center of the hub.

eddy m

That's the answer. Aluminum spokes are a solution that causes new problems without providing any benefit past the bling effect.
As far as Sheldon's opinion about radial spokes being more reliable on the non-drive side, I think he's probably wrong about that. The pedaling torque load is pretty small compared to the cyclic load from the weight of the rider. i think most of the problem of loose spokes developed because your typical bike shop wheelbuilder didn't realize how small the margin for error in tension became when 8s or 9s were developed.
There's basically 4 solutions that work:
You can equalize stress with thinner spokes on the non-drive side. That causes more of all the loads to be carried on the drive side. It's easy to do and it's both lighter and more reliable.
You can equalize tension with an offset rim, but there's a weight penalty for that.
You can equalize tension by using a hub with both flanges close together, but that gives you a laterally flexible wheel. I've never had a wheel that was too flexible, but some of my heavier friends complain about that.
You can use triplet lacing. That's the one unusual lacing pattern that actually makes sense, but you need to use a hub with widely spaced flanges to minimize lateral flexibilty. If you are worried about breaking a spoke you can carry a spare.

em

krazygl00

I'm thinking of building a rear wheel using a combination of solutions 1 and 2 with a Velocity Aerohead OC, 36h (I'm a clyde).

AEO

IMO, aerohead isn't a very good rim. aerohead OC is even worse. the sidewalls where the nipples are are very thin.
better to go with a beefier deep profile rim like sun venus, deep-V or similar and just allow the thinner spokes do their work.

superdex

I went from a pair of Shimano RH550 --radial DS, 2x NDS-- to a pair I built, Sun Venus rim (great rim), Sapim double-butted spokes, and laced 3xDS, 2xNDS. (Rob, if you're reading, I had these mostly built and sitting in the basement for two years, finally finished them about a month ago). The result? Noticeably stiffer --in hindsight I now know what 'windup' means, and the Shimanos had it. And after a few hundred miles I haven't had to true them. I'm 200lbs, and I hit potholes (unintentionally, but still).

Yes, totally anecdotal, but I thought worth mentioning.

conspiratemus1

I've been building with thinner double-butted spokes on the left side lately, ever since my LBS hired a mechanic who convinced his boss to start stocking (or obtaining for me) spokes in odd-lot numbers instead of by the box. I find I can put more tension in those non-drive-side spokes and the wheel stays true. But as a home wheelbuilder who weighs just 145 lb and never broke spokes before, I have neither the numbers nor the stresses to make any meaningful opinion about superiority. All I can say is that thinner spokes on the left side do work, and don't loosen over bumps. Might try that on the next tandem wheel (if it ever wears out....)

AEO

one way to check your build quality would be to reduce spoke count...
fewer spokes = more error prone.