August 14, 2012 at 9:24 pm #93483Dave K
Excellent review. Any chance of Dash hubs being included in a future update?August 15, 2012 at 4:36 pm #93484Brad
C-4 suffers quality issues, IMO. I’ve had nothing but grief with C-4 hubs. Fairwheelbikes dropped C-4, and I don’t blame them for that decision.
I don’t think Fairwheelbikes has reviewed any hub it doesn’t sell. So, that would explain why the low cost Tawainese hubs like Circus Monkey, etc, largely OEM hubs, have not been reviewed. Someone mentioned Hope hubs. Im a fan of the Evo Pro for mountain bikes, and the weights and price point of their road hubs are close to white Industies. Worth looking at carrying Hope, IMO.
Fairwheelbikes added Dash products recently but no their hubs.August 16, 2012 at 8:17 pm #93485
The Shimano and Campy hubs do not come in low holecounts, which limits their application.
Jeremy’s (Alchemy’s) lack of marketing can be seen as a virtue… though I do agree that logos on the hubs would be nice.
The low-cost hubs from Taiwan and China are a good value, but since everyone contributing to the review builds customs with higher priced hubs, we just didn’t have any experience with them.August 30, 2012 at 10:30 pm #93486
Very comprehensive review and wonderful archive of the current state-of-the-art in hub design and construction; thanks to all of you who put in the long hours to assemble this. I have a few comments:
1) In the the section headed, “Notes on bracing angle,” I staunchly disagree that “Bracing angle (or flange offset) is the most important factor effecting the lateral stiffness and stability of the wheel,” unless you meant to include the word “hub” between “important” and “factor.” Yes, it matters, but it’s no more than the second most important factor for any reasonable hub design. The most important factor is the rim! The rim bending stiffness (second moment of inertia) is proportional to the cube of the section depth for rectangular or triangular rim sections in either the x- or y- (radial or lateral, let’s say) axes, so a small difference in section depth makes a big difference in bending stiffness. Just going from a 19 mm wide rim to a 22 mm wide rim, all else being equal, increases the stiffness by (22/19)^3=1.55, or 55%, while the difference in bracing angle between the narrowest front hub (Soul-Kozak Aero) and the widest (Alchemy ELF) is 7.9/5.7=1.38, or 38%. See my point?
2) You go on to say, “The lateral stiffness imparted by the spokes goes up with the *square* of the bracing angles, while using more or heavier spokes only results in a linear increase in stiffness… and an increase in weight.” How is it proportional to the square of the bracing angle, unless you mean to imply that it’s squared because it’s linear on both sides so the combined effect is quadratic? The lateral spoke force is proportional to sin(bracing angle), and sin(x)/sin(y)=x/y for small angles, but Hooke’s law says the force is directly proportional to the displacement. That’s a simplistic view, though, as a stiffer rim also spreads the load out over a larger region, so perhaps the square law is appropriate for linear spoke and cubic rim proportionality on each side of the wheel? The full answer, of course, is FEA analysis, and Henri Gavin, Prof. of ME at Duke University, did that years ago.
3) I agree with those who say that narrow hub flanges afford no aerodynamic advantage vs. wider flange spacing. Jobst Brandt addressed this generic issue years ago in _The Bicycle Wheel_, where he points out that radial lacing affords no aero’ advantage over crossed lacing, since the spokes in either case still see the wind, only at a slightly different apparent angle relative to vertical: the spoke cross section into the wind is the same in either case. Same goes for hub flange spacing. Thus, I would never recommend a Soul-Kozak hub Aero front hub for a wheel build since there’s no aero’ advantage and lateral stiffness is compromised compared to other high-quality hub offerings.
Keep up the good work, and ride on!
PAKETA CYCLESSeptember 5, 2012 at 7:54 pm #93487
Thanks for your comments, Dave.
1)We know that rim stiffness is important. Rim stiffness does indeed increase greatly as the section width and height goes up, but when evaluating the stiffness of the wheel system, it is a matter of degree… how much force does it take to displace the spokes vs displacing the rim? The old very shallow section rims were quite flexible, but the stiffness of the spokes made the wheels stiff. More typical modern designs have much stiffer rims but fewer spokes. I did some FEMs a few years back and even for typical “stiff” rims, the spoke bracing angle still dominated, IMO. At any rate, this is a hub review so we assume that rim selection is another topic.
2)You seem to understand structural calculations, so you could analyze a simple model where you have one spoke of the appropriate length and bracing angle, and calculate how much force it would take to displace the rim end say 2mm laterally. You will see that the force required is proportional to the square of the angle for the small angles we are interested in. Both the spoke length change and the lateral component increase proportionally.
3)Though it would make a small difference, radial lacing should provide an aero benefit because it avoids the interference drag and turbulence near the hub. I’m skeptical that a narrow flange would help at all. But for really deep rims, a wide flange spacing isn’t necessary to get a good bracing angles, and really high bracing angles could cause problems (maybe) with the high nipple angle in the rim. For the deep rims there is often more lateral flex taking place within the rim itself than elsewhere.September 6, 2012 at 2:24 am #93488Participant
with narrow flange spacing, while i dont think there are any aero benefits to the spokes, i think there may be some gains around the axle. there can be an airfoil shape from the dropout to the flange that is free from the disruption of the churning spokes. sure, it’s small, but if you are looking at possibly too great a bracing angle with a deep rim, it could be a win/win.
i definitely feel that narrowing a front hub flange to “hide” the spokes behind the rim is completely pointless for a number of reasons.October 4, 2012 at 10:46 pm #93489
It took me a while to uncover some of my older notes; I apologize for not replying sooner. Thanks for your reply to my first post; comments regarding your three points:
1) I don’t understand the distinction you make between, “…how much force does it take to displace the spokes vs. displacing the rim?” as when one uses the term “wheel stiffness” it’s the whole system—as you mentioned, but then what’s the difference? Without knowing the parameters you used in your FEM analysis it’s difficult for me to comment on your modeling, but I still assert that the rim can—and usually does—make a bigger difference to overall wheel stiffness than the spoke bracing angle for the typical ranges encountered with (a) rim section depths vs. (b) hub bracing angles. Sure, I understand it’s a hub review (!) but this is an important issue to keep in context.
2) I must be missing something here. As I said previously, the lateral force at the rim is directly proportional to the displacement for a single spoke; that force being equal to sin(bracing angle)—call it simply “sin(a).” Let’s assume a different hub with bracing angle b. Then the (lateral) force at the rim is proportional to sin(b) and the ratio between the two cases –what we’re interested in for wheel design purposes–is (sin(a))/(sin(b))=a/b for small angles, where the “=” sign is understood to be approximate. That appears to be linear to me. For a symmetrical (e.g., non-disc front) hub with the same bracing angle on each side, you could call the ratio proportional to 2a/b since there’s an equal contribution (with a pre-stressed structure, that is) due to the spokes on the other side, but it’s still not a quadratic relationship. This simple analysis also agrees with the graphs produced from FEM analysis done by Mark Rodamaker, “Design and Analysis of an Optimum Bicycle Racing Wheel,” Finite Element News, Feb. 1989, pp. 33-37. Figures 4 and 8 show that the lateral displacement is linear with force up to the point where some of the spokes go slack.
I also found the paper I referred to in my first post. The citation is “Bicycle-Wheel Spoke Patterns and Spoke Fatigue,” Henri P. Gavin, Journal of Engineering Mechanics, August 1996, pp. 736-742. Gavin’s work is analytical, not FEM—my error—and verified experimentally. It was the discovery of Gavin’s paper that led me to work on deep-section wheels in 1998, where I performed a series of wheel stiffness tests on various wheels (15 in all; 5 rear and 10 front wheels) including several intended for tandems as part of the development process for an early entrant in the “deep section rim with low spoke count” sweepstakes. The most fascinating revelation (to me, at least) was Figs. 5 and 6, which show curves for radial and lateral wheel stiffness vs. rim bending moment of inertia about the axis parallel to the axle—in other words, these are graphs for radial and lateral stiffness vs. in-plane (or radial, if you like) rim stiffness. The key takeaway is that the wheel’s lateral stiffness increases just as fast as the radial stiffness for an increase in radial rim stiffness. Hmmm…
This is (or was) the result: http://www.sandsmachine.com/a_bil_t8.htm. The lateral stiffness of these 24 spoke wheels was nearly the same (within 10%) as a pair of 48 spoke wheels I tested that came off of a Santana Sovereign. Interestingly, the Phil rear hub I used for my wheels was a 145 mm version (flange spacing of 55 mm), whereas the Santana rear had a 160 mm hub with 60 mm flange spacing and the same Wheelsmith DH-13 2.3/2.0 mm single-butted spokes as I used. The difference, clearly, is the rim: 40 mm deep FIR Rialto for my wheels vs. 19.5 mm deep Mavic T217 (a popular box-section tandem rim at the time) on the Santana wheels. The rim widths are close, with the Mavic being slightly wider (22 mm) than the FIR rim (20 mm) I used. Even I was amazed at how much of a difference the radial rim stiffness makes in lateral wheel stiffness. To reiterate, the difference in radial rim stiffness is approximately (40/19.5)^3=8.6 times. The fact that there are only half as many spokes along with a lower spoke bracing angle yet the lateral stiffness is comparable indicates that the rim is the dominant effect, not the bracing angle. A wider rim and/or more spokes and/or a wider hub would’ve made these wheels even stiffer laterally, but the FIR rims only came in 16 and 24 holes (I also have a 16 spoke front we’ve used on our tandem and even occasionally on the triplet), that’s how I designed the Bilenky (i.e., 145 mm rear), and they’ve served us well for 15,000 miles of riding to-date, so they’re obviously stiff enough as well as strong and durable enough: we’ve never broken a spoke.
3) I would agree with this assessment.
Feel free to contact me offline if you’d like copies of the articles mentioned.
Thanks for listening,
Dave WalkerOctober 8, 2012 at 4:17 pm #93490
All good stuff, Dave… hope I can explain things as well.
You are missing part of the equation for the spokes. The lateral component of the spoke’s force (tension) will be proportional to the bracing angle. But in addition the change in spoke tension for a given lateral displacement will also increase proportional to the bracing angle… resulting in stiffness being proportional to the square of bracing angle.
It might help to show an example… calculating the lateral component of force if the rim end of the spoke is laterally displaced 1mm for two different initial offsets of 20mm and 40mm:
Radial hub to rim= 280mm
Offset 1= 20mm
Offset 2= 40mm
Spoke stiffness= 100kg/mm
Lateral displacement= +1mm
Spoke length initial L1a= (280^2+20^2)^.5= 280.713mm
Spoke length final L1b= (280^2+21^2)^.5= 280.786mm
Change in spoke tension= (280.786-280.713)*100kg/mm= 7.30kg
Lateral component~ 7.3*(21/280.786)= .55kg
Change in spoke tension= 14.30kg
Lateral component~ 2.07kg
Doubling of the offset results in a ~4x increase in spoke lateral stiffness.
In either case the force vs displacement will be nearly linear. For instance, you could calculate the lateral force for 3mm displacement and it will be about 3x as high… and still 4x difference between 20mm and 40mm.
Many years ago Damon Rinard did some tests of wheel stiffness (http://sheldonbrown.com/rinard/wheel/data.htm), and several of them were front/rear sets with the same type and number of spokes. I counted 10 road sets that meet this criteria. As a ballpark value I’d expect the spokes’ contribution to lateral stiffness to be ~60% higher for a front wheel vs a rear wheel, if the number and type of spoke are identical. This is based on a typical 36mm x2 offset on the front wheel vs 36 and 18mm offsets on the rear. In Rinard’s tests the ratio of deflection between the front and rear ranged from 1.35 to 1.61 with an average of 1.48. 48% is high enough compared to 60%, that I feel confident in saying that the spokes were the greatest contributor to lateral stiffness for these wheels.
The Gavin paper is online if anyone is interested: http://www.duke.edu/~hpgavin/papers/HPGavin-Wheel-Paper.pdf
As you mentioned, figure 6 shows the modeled increase in lateral stiffness vs rim radial stiffness. In his example increasing the rim radial stiffness 3x resulted in a ~40% increase in lateral stiffness. IIRC, this is with 36 2mm spokes. This is in the same ballpark with your experience, where you increased rim radial stiffness by 8x and removed half the spokes to end up about where you started. An 8x increase in rim radial stiffness is a very large amount… and a 700g 40mm deep rim is certainly on the uber high end for stiffness. I’m also not surprised that you experienced good reliability with the 24 DH13- 700g 40mm deep set. Even that one is a beefy build. I’ve built tandem sets with 28f 32r CX-Rays that weigh about half as much (and are half as stiff) as the DH13s, and rims that are only 445g… and even those hold up under modestly sized riders.
You stated that rim depth is a more important factor than hub bracing angle… for the typical ranges encountered. But that last part is key. As I showed from Rinard’s tests, the difference in bracing angle between a front and rear wheel makes a large difference…48% on average. Based on Gavin you’d need a >3x increase in rim stiffness to equal that… which is certainly exceeded if you compare the least and greatest stiffness rims available. But typically riders limit their choices to a much smaller range, because they are also interested in other criteria… like weight, aerodynamics, and ride quality.
If you compare all 100mm front hubs to each other, and all 130mm rear hubs to each other, you won’t see a big range in bracing angles. The reason is *because* it is so significant. The hub manufacturers use about all the space they can… and the good ones use every small fraction of a mm on the rear hub. If it didn’t matter so much, there would be good reasons to have hubs and axle spacings narrower than they currently are. In fact bracing angle is the driving force behind the steady increase in rear dropout spacing as cassettes get wider, and why tandems have 145 and 160mm wide rear axles.
I agree that the rim stiffness is a substantial contributor to lateral wheel stiffness. That is one reason why I tend to favor rims with some depth to them even if the wheel ends up a little heavier, since you can also reduce the number of spokes and improve aerodynamics. But the bracing angle of the spokes is important as well… and there is everything to gain and nothing to lose by optimizing those angles within the constraints of the frame and drivetrain. Not just for lateral stiffness but for keeping tension on the NDS spokes, which I believe is more important.October 28, 2012 at 9:29 pm #93491
I appreciate your comments and couldn’t agree more: all good, and the intent is to increase the knowledge base and disspell myths and misinformation when it comes to good wheel design and construction. Let the information overload continue
Thanks for the clarification on spoke contribution to lateral wheel stiffness being proportional to the square of bracing angle (and the semantics are important: it’s spoke contribution at issue). One small note I’d make is that if you go through the math, using Gavin’s spoke modulus of 206 GPa (appropriate for stainless steel), 100 kg-f/mm stiffness corresponds to a 1.3 mm diameter spoke. More realistically, for the 1.83 mm diameter Gavin uses (odd, but slightly larger than 15G or 1.8 mm) the spoke stiffness is very close to 200 kg-f/mm. This only helps your argument, though: the lateral component per spoke is effectively double what you calculated in your example.
Where I’m having an issue is the sentence, “You stated that rim depth is a more important factor than hub bracing angle…for the typical ranges encountered. But that last part is key…” I’d say that’s just the point, but turn it around: if the purpose of discussing and listing bracing angles in the hub review and hub specs is so readers can make informed decisions on hub selection, then isn’t it more appropriate to compare front hubs and rear hubs separately?
As an example, neglecting the outlier Soul-Kozak front hub, the range of hub bracing angles is 7.1 (C-K R45) to 7.9 (Alchemy ELF) degrees. Squaring the ratio (7.9/7.1)^2 gives a 24% increase in stiffness due (only) to the spoke bracing angle. A 24% increase in lateral wheel stiffness due to the rim can be achieved by increasing the rim depth by (1.24)^3 or 91% increase in rim depth. The thing is, one has a wide range of choices for rims; not so for hubs: a front hub cannot be substituted for a rear to get a substantial increase in spoke bracing angle, but one can easily choose a stiffer rim to more than make up for it. Fairwheel sells KinLin rims in section depths of 19, 22, 27, 30, and 38 mm, for example. If I felt a customer asking for, say, a XR-200 22 mm deep rim (most likely for the weight savings) but weighed 200 lbs., I’d probably steer them towards the XR-300 30 mm deep rim (and 65 grams heavier) for overall strength as well as lateral stiffness. The XR-300 is 254% stiffer radially than the XR-200, meaning building a front wheel with the widest flange spacing in the review on the shallower rim would be less stiff laterally than the narrowest hub in the review (outside of the Soul-Kozak, as stated previously) using the XR-300 rim, all else being equal.
The conclusion is that, for a given wheel–front vs. rear, MTB vs. road, etc., etc.–the most important practical considerations for wheel stiffness is, in order of importance: (1) rim; (2) number of spokes; (3) hub bracing angle, neglecting the odd hub outliers; (4) spoke gauge (Rinard says it’s a small effect; 11% increase for 2.0 straight gauge vs. 2.0/1.45/2.0 butted with 32 spokes in his test). Interesting.
Your comment on the FIR Rialto rim at “700 grams” (it’s more like 650 g–still substantial, but with many Al rims at 400-500 grams it’s not a huge stretch, particularly for a tandem build) only matters to a few people, and I agree not many would choose that rim…although I’d readily point out that Fairwheel offers the KinLin XR-380 38 mm deep rim, which at 550 g might indicate there’s some market for such things. A more relevant response is one word: carbon. There’s plenty of light weight carbon offerings in a wide variety of rim depths, so for those wheel customers who are concerned about lateral wheel stiffness, an easy solution is to steer them towards a deeper-section carbon (assuming they’re in the market for one in the first place) rim. Note that ENVE, as a good example, has refined their Smart System 6.7 wheels by using a 60 mm section depth for the front rim and a 70 mm section depth for the rear rim–an interesting and relevant point overlooked by the reviewer in the Fairwheel product introduction athttp://fairwheelbikes.com/cycling-blog/products/enve-smart-system-6-7-rims.html. Doesn’t sound like much, but that 10 mm difference front to rear is worth (70/60)^3=60% increase in radial stiffness on the rear, which helps make up for the inherently lower lateral rear stiffness due to the hub bracing angle.
We haven’t even talked about rim width, which obviously also makes a big difference to those who are concerned about lateral wheel stiffness. I’ve always thought the crazy super-wide DH rims with shallow section depth (e.g., Sun Rhino and the like) are a poor design as the wide rim might be OK laterally, but not without enough section depth to keep the wheel stiff and strong radially…and add to the lateral rigidity.
The last issue I’d like to address is wheel strength, not stiffness. As Rinard, Jobst Brandt, and others have said, the strength of the wheel is only maintained until certain spokes lose tension. In a severely biased rear wheel, that happens sooner, but that must be balanced against the engineering advantages of a wider flange spacing. As the original review stated, one way to achieve that is by using lighter-gauge spokes on the NDS vs. DS–something I’ve done in all my wheel builds since 1991. To the extent that all of the rear hubs in the review end up with a NDS:DS spoke tension ratio of around 50%, that argues strongly in favor of the 2:1 triplet rear lacing pattern, since the spoke strain can be equalized using the same-gauge spokes on both sides, thus making the wheel as strong as possible but also as stiff as possible (vs. using differentially-guaged spokes NDS:DS, assuming conventional materials where you’re limited in the choices of available spoke gauges). That also indicates an optimal left-side flange distance/diameter so that the tension ratio is exactly 50% if used with triplet lacing.
It’s too bad Gavin’s analyses didn’t include similar graphs for lateral stiffness vs. J (rim torsional moment) and I_rr (rim bending moment about the radial axis), as that’s what you really need to look at if you’re obsessed with lateral wheel stiffness. Maybe someone out there can add to the knowledge base with this info? There’s always another layer on the onion to be peeled.
All the best,
Dave WalkerJanuary 25, 2013 at 3:31 am #93492Brendan Montgomerie
Hey, great review! I started riding with
shimano hubs, but I bent several rear axles,
tried a Nukeproof hub that immediately cracked a flange.
I heard that White Industries was the answer back
in ’95, and ever since then I have been riding their
Speed racer with the double 55 mm flanges.
It started with about 10 MTB rims, all Mavic,
then another 8 road rims, mostly velocity. The same
hub has been in continuous use for 18 years!
Currently built 3x drive radial ND, Sapim race
drive, laser ND to a Kinlin XR-200 (this is a 32 spoke wheel),
brass on drive al. nipples ND. Amazing wheel
amazing hub! When I wanted to convert from 135mm
to 130 mm I just walked into White Industries
and they gave me a free ND flange to convert.
The first 3 bearing changes were just ride to their
factory in Petaluma, pay for new bearings and
they replaced them while I waited for free. The bearings
Typically last 3 years on the road, half that on dirt.
I recently bought a ten pack of 6902′s and a ten pack
of 6802′s for $25, so I guess I’ll never run out.
Best rear hub for durability and stiffness!
TMI- extreme product loyalty. If you use this hub,
get a bag of 2 mm set screws from White,
replace these every bearing change or so,
and never try to tighten them without cleaning the
inside of the Allen head, don’t go overboard tightening them,
and you will have a no trouble. That’s it. I should try to sell them!February 19, 2013 at 9:55 pm #93493Geraint
Great article, really enjoyed reading this and the rim review.
I would be very interested in your view on the CK classic & classic cross hubs, in particular in comparison to the R45s. Judging from the comments you made on the R45s, I expect you’ll say that they’re doubly overbuilt! Were these hubs covered in your original review? I can’t seem to find a link to that.
Thanks again :0)March 14, 2013 at 11:13 pm #93494Participant
Pity you guys didn’t get your hands on some British Goldtec hubs, 48t engagement, super super smooth bearings, developed in conjunction with Britiish Cycling, they have a great reputation in the UK.
Maybe next time?April 15, 2013 at 2:07 am #93495Tim Voegele-Downing
Or even have a look at the fabled Gokiso Hubs…April 17, 2013 at 7:05 pm #93496Dave
Gokiso? I had never heard of them before. In looking at their spec’s, it would appear the the bracing angle of the spoke wouldn’t approach that of someone like Alchemy (if I read their spec’s correctly).
Interesting none the less. ~$3k for a set of hubs? 29 year warranty? Wow…April 17, 2013 at 7:58 pm #93497Brad
Gokiso “super climber” at a combined weight of 675g is hardly a contender for a weight-weenie climber’s needs. The ‘suspension’ doesn’t seem practical to me, gimic, and would actually increase loss of tension on the non-pulling spokes compromising durability. The spokes are elastic enough to handle stress of most uses. The freehub body design is interesting and the Gokiso materials seem high grade with precise manufacturing. The rear hub geometry seems less than ideal. And someone suggested $3k for a set of hubs?? ouch.