Different spoke count front-rear, still OK with hub brakes?

Quite recently Sheldon Brown wrote something along the lines of "If you have the same number of spokes front and rear either your front is heavier than it needs to be or your rear is weaker than it should" somewhere on this site.

For a "traditionally" configured bike, rim braked and with a dished rear this makes good sense, but what about bikes with hub-mounted brakes, either drum or disc?

Rear still has the disadvantage of dish, carries the heavier load during most riding, and has to deal with the propelling torque. But the front can become heavily loaded during braking, particularly on pavement/tarmac, so is it still recommendable to use a lower spoke count front?

(I think I'm clear on the forces involved, but I'm uncertain about order of influence/magnitude..)

I assume the original recommendation is mainly a spoke fatigue issue, so (how) would the addition of hub-mounted brakes change the recommendation? Is there a possibility of causing front wheel failure through brake generated torque? Or through dynamic weight transfer during braking?

(this would be with a drum brake hub(high flange), so front would be symmetrically dished)

I'm not thinking of anything particularly extreme here, maybe a 36(3X, half radial)-28(2X) combo on a lightly loaded commuter.

Functionally I could very probably get away with a much more generic build, but where's the fun in that?

Cheers,

Are there really NO opinions available regarding this matter?

It's not like this forum to miss a chance to debate the various merits and drawbacks to different spoke counts and configurations!

(ah well, there's always the fallback of trial-and-error...)

You could write an email to Sheldon and see what he says.
I think that using a front hub brake, even one that is symetrically spoked, increases the dynamic loading significantly.... especially due to the fact that front wheels carry most of the braking load.
If you don't already have the front hub and the rim for it, why not go for 36 spokes front as well?

__________________
David Ornee, Western Springs, IL USA

FWIW. I treat the brake side on a front disc hub as if it were the drive side on a rear hub -- consequently, I chose to lace a 32h 700c CXP14 rim onto an XT hub with a x4 on the BRAKE side & x1 on the non-brake side -- straight 14g spokes, of course [just the facts ... not necessarily trying to keep THAT controversy alive].

The rear was x4 on both sides with 14g spokes.

A front disc hub is NOT symmetrically dished, BTW.

Attached Thumbnails

 

Thanks for replying.

I've been thinking along those lines too, but been a bit uncertain about how far to carry the comparison. Undoubtedly a rear wheel would see more "drive impulses" than a front would see "brake impulses", but OTOH I could probably brake harder than I can accelerate....

Can't say that it's a combo I would have thought of anytime soon, but I can see what you're aiming for. Isn't x4 considered a bit awkward to lace, denpending on the hub?

But the drum brake front hub I'm thinking of using would be.

Thank you for replying.
I believe I did that. But either it got lost when my email provider was having some difficulties or he'd been too busy to reply.

That seems highly probable, but to what degree? Enough to have negative consequences? Wheel failure is usually fatigue related(and not sudden overloads) and you don't brake half as much as you pedal.

I have the front hub already. It's actually a 36H but I planned to leave some holes empty. If Rinard can get away with it so can I...

Leaving holes empty in the rim though isn't tempting even for me...

Well, mainly because the statement from Sheldon and the final line from my original post:Functionally I could very probably get away with a much more generic build, but where's the fun in that?
Unless it's too weak to function it'd still make more sense than a crow foot lacing for instance.

It's not like I really need this bike anyhow so I thought I'd indulge myself with some light experimenting - unless you guys think this is a certain recipe for wheel collapse and road rash.

Cheers,

Yes, x4 is problematic on a low flange hub, but not as much as most people would like you to think ... and, it really isn't a problem on a high flange hub, IMO ... but, that's just my opinion, I suppose -- i.e., IMO, having a spoke in contact with (or, lay over in the case of a low flange hub) another spoke in the flange area is no more a real problem than the J-bend OR the contact between interlaced spokes ... so, it really becomes more of a cosmetic issue (at least, IMO!).

Lacing the non-brake side x1 came down to a cosmetic issue (radial lacing is a generally BAD IDEA, IMO). A x2 lacing & x3 lacing didn't look the way I wanted (at the time) AND I deduced/decided that the x1 lacing would provide adequate lateral "support" with the stiff CXP14 rims.

You know, the only thing that you need is a good spoke calculator (I use Dan Halem's http://www.geocities.com/d_halem/wheel/wheel.html because I find it exceedingly easy to use, FWIW) & an inclination of how you want the wheel to function & look ...

If your drum brake hub is truly symmetrical, then I would probably lace it x4 on the drum side & x3 (or, x2) on the non-drum side.

NB. Having a mis-match in the number of holes in the hub & rim introduces a REAL, potential aggravation that you need to be aware of -- that is, you will probably need at least THREE DIFFERENT SPOKE LENGTHS (using more lengths is probably slightly better) ...

I added FOUR holes to a 28h 3-speed hub (steel) so that I could lace it to a 32h 700c Open Reflex rim, and THREE different spoke lengths is what I ended up using ...

And this hub would qualify as high flange... Dang! I reall didn't need another option on how to build these wheels!

So I've heard, yet I have a radial front wheel built around an old LX hub that seems to be doing fine so far, several seasons down the line.


Makes sense to aim for something that's more than "only" functionally pleasing while you're still at it, doesn't it?


It has an illustration feature? That WOULD make a change from the DT one that I normally use.

Cheers,

Duly noted. I've seen (and toyed with) some spoke calculators / formulas that claim to deal with this problem. but I don't feel quite confident with the answers given. Some LBS will get a repeat customer for some days when it comes to final assembly. The "normal" length should be enough to get the wheel to hold its shape though, which would make it easier to determine the length of the other ones.

Now THAT sounds tricky, it must have been an exceedingly helpful flange design to leave you with enough material for such an operation.

Radially lacing places more stress on the flange than crossed lacing UNLESS the hub has been designed for radial lacing -- e.g., DT redesigned/re-spec'd their flanges to have a larger diameter.

Since a MTB generally has a front suspension AND fat, low-pressure-by-comparison tires, the negative ride characteristics of radial lacing aren't noticed ...

Radially laced (front) wheels on a ROAD bike are for those who are slaves-to-fashion OR who are gluttons for punishment, IMO.

The flange size on a 3-speed hub is equivalent to the size of a high flange hub, so there WAS enough space (certainly, no more tightly spaced than the spoke holes on a 36h low flange hub) -- two extra holes per side -- I would NOT have done it with an alloy hub, BTW.

Now I remember. I found a link some time ago that offered a modified spoke calculator that dealt with mismatched no. of holes in hub/rim, but it only generated ONE new recommended spoke length. This seemed odd to me since the spoke length deviation shouldn't be static.
If you have a group of spokes with an empty hole at each side at the hub the middle spokes in this group will be pretty darn close to where they would have been in a matched configuration, while the ones closest to the hole will show the most deviation, particlularly the ones reaching "across" the hole.

BTW. Dan Halem's spoke calculator does not have "an illustration feature." It does have an extensive (though dated & incomplete -- e.g., the ubiquitous MAVIC Open Pro is not included, so I use the Open Reflex data & generally adjust the ERD, accordingly) database whose variables you can modify quite readily -- i.e., start with a something that is close (if not included), modify the one-or-two variables ... et voila!

So, you MAY actually have to measure the ERD (effective rim diameter) and locate/select a similar (in dimension) rim in the database ... OR, hold a "new" rim next to all your other rims to see if anything appears to be close enough to use as a starting point. For example, I have some UKAI 700c (622-17) rims which are not in the database, but which I ascertained to have an ERD comparable to a "standard" sew-up rim; and consequently, I simply used the Campagnolo Strada (tubular) rim for the spoke calculation.

When I laced up the CXP14 rims to the XT disc hubs, I selected the Campagnolo Hi-Flange front hub & changed the value for the left flange offset.

When I've looked at OTHER calculators, they were not any easier/faster to use -- at least, the way I use it -- so, I continue to use Dan Halem's.

BTW. With Halem's calculator, you can/should ROUND DOWN to a slightly shorter spoke.

Yes. I (would) wonder how that calculator worked out the ONE, compromise spoke length ... in theory, half of the spokes on each side would actually have a different length, but the centimeter-of-threading (some spokes do have MORE) does leave wiggle-room in the calculation.

FWIW. I essentially used trial-and-error to get the THREE different spoke lengths that I used (thank heavens I have my own spoke threader!) ...

I think I began with the calculated length, laced up the rim, and then inspected for (looked at ... measured/reckoned/guessed-at) the variance ... and, THREE different lengths was the minimum I was able to use on a 32h rim (I suspect that FOUR different lengths might have been slightly better).

If I had been lacing up a "dished" wheel whose rim & hub had a mismatched count (BTW. there is a miniscule offset on a 3-speed hub which I chose to ignore as being inconsequential), then then I would have needed (?) SIX different lengths.