Hub Review

Welcome to the 4th edition of the Fair Wheel Hub review. In the previous review we brought Ron Ruff from White Mountain Wheels on board to give his thoughts as well. We figured having different points of view would be advantageous, so we’ve once again brought Ron back for the latest version. Ron, like us, really seems to enjoy the geekier side of wheel building and is one of the custom builders we most respect. We should also mention that while some of this review is taken directly from the original two we’ve changed much of it as well as added to it as hubs have changed quite a bit over the years that we’ve been doing this review. So don’t skip a paragraph thinking that you read it in the last one, just because part of it is the same doesn’t mean that all of it is.

Mistakes: The specs were compiled by us here at Fair Wheel and Ron at White Mountain Wheels, and while we did do our best to be careful there were an awful lot of numbers and calculations thrown around over those days. So with that in mind I’d like to put out the disclaimer that it isn’t impossible that we might have transcribed, written or recorded a number incorrectly. So please forgive any typos or mistakes. We’ve already corrected a ton and now like to think that most things should be correct, but with the scale of this thing it’s still possible that one will find a mistake.

Considerations when choosing a hubset: It’s important to understand that there is no one perfect hub. Each hub has its own unique strengths and weaknesses. That’s where a good custom wheel builder comes into play, helping you decide what best fits your needs. So a hub that is right for one person may not be the right hub for another. Anyone that tells you there is one ideal hub should be considered suspect. We’ve never found one hub that fits all needs ideally. With the combination of those of us writing this review, we estimate we’ve built 20,000 pairs of wheels, so it’s safe to say we’ve had a fair amount of experience. There are certainly hubs that could be used by any rider, but that doesn’t necessarily make them ideal for everyone. Just like being under-built, a wheel can also be over-built for a rider and their needs. We consider all the hubs in this review to be “good” hubs. That means they have a good design, quality control, reliability, service, etc… and each is capable of being a top choice in a particular application. But every design is a collection of optimizations and compromises, and every rider has different priorities. Things like price, weight, resistance, durability, ease of service, branding, and looks, are all considerations… along with the spokes and rim selection and the intended purpose of the wheelset. The idea is to balance the characteristics that are most important to one given rider and more importantly to balance the hub selection in harmony with the rim and spoke selection. Hopefully this review will help to point you in the direction of the hubs that will work best for you.

Durability is one important aspect, and also the most difficult to nail down. A hubset might be lighter than another due to better design and materials, but there can also be tradeoffs like smaller bearings, and simply thinner or weaker parts. Ultimately long term experience is the best indicator, but that isn’t a lot of help when a new or altered design comes on the market. To further complicate matters, the QC can be variable, so even an old design that was previously solid can have random issues. Even determining the kind of forces a rider might put on their wheels is complicated. Some 240 lb riders would have no issues with a hubset that some 120 lb riders would destroy. Rider weight is one important aspect, but so is riding style, so the rider’s previous experience with equipment needs to be considered. Also, a rider might be fine with durability compromises on their 12 lb climbing bike, but have a completely different set of requirements on a regular bike that is ridden in all sorts of weather.

Bearing material: this is what comes stock in the base model. Some hubs have upgrades available from the factory.

Bearing size: moving from left to right in the hub shell and then in the hub body.

Static load: Combined static load for the hub shell and again for the freehub body. Static load rating is the maximum amount of load a bearing can take without excessive deformation that would degrade the bearing performance.

Notes on Bearings and drag: Since ceramic bearings became the rage a several years ago, bearing drag has been a hot topic among cyclists. Unfortunately, there doesn’t seem to be much public information on just how much of a loss the wheel bearing drag contributes. According to Bicycle Science the drag of clean, lubricated, properly aligned and adjusted ball bearings is very small. The friction coefficient is ~.0015… which is the ratio of resistive force generated in the bearing divided by the load it is carrying. If you are familiar with tire rolling resistance coefficients, this functions in the same way… except that you need to multiply this force by the bearing/wheel radius to get a comparable factor. So lets say we have a hub with 15mm axle, and the bearings are on a radius of ~12mm. The wheel’s radius is ~335mm, so 12/335 *.0015 gives us an equivalent rolling resistance coefficient of .000054. To give you an idea of how small this is, typical tire rolling resistance coefficient is about .005… so the bearing resistance is ~100 times smaller. Another way to look at it is that a 200lb rider+bike traveling at 25mph will lose ~0.5W from the bearing rolling resistance. And these are not fancy bearings we are talking about… just decent steel ones.

There is another major component to bearing drag though, and that is the resistance of the seals. John Swanson did some interesting coast-down tests of wheels shown here: https://www.bikephysics.com/rails/wheel/list Aerodynamic drag was part of it, but his instrumentation was sophisticated enough to back out the bearing drag alone. Ron did the calculations on the bearing coefficients he obtained, and got an average power consumption of 0.25W for front hubs and 0.40W for the rear hubs at 25mph… or 0.65W for both wheels. Note that there was a lot of variation, but even the worst set of wheels was only ~1.3W. Since the only load in his tests was the weight of the wheel we’d consider these values additive to the 0.5W determined above… so typical losses are about 1.2W total. Though the losses in this test would capture any effects of misalignment or preload in the unloaded state, we should point out that under typical loads these factors can result in additional friction.

Does this mean that bearings don’t matter? We wouldn’t say that. Instead we’d emphasize that the most important factors are cleanliness, adequate lubrication, alignment, and adjustment. If any of these are off, then the drag can be much higher. Even though smaller bearings might have lower resistance in an ideal world, larger bearings (higher load rating) will be more tolerant of un-ideal situations, probably resulting in a lower practical resistance in addition to a longer life span. If your typical hub set in good condition is only consuming ~1W then be realistic about how much improvement is possible. The added expense of ceramic bearings and the added hassle of having light seals and grease (which probably will result in quicker bearing contamination and more frequent replacement) may not be worth it.

Axle diameter: Larger axles will typically produce stiffer wheels. It’s also important to note that a couple of axles are in different ways, butted or reinforced at the freehub body in the rear to help prevent cantilevering under acceleration.

Price: This is the msrp as it applies in the USA.

Flange diameters: Left / Right. As measured by us from center of spoke hole to center of spoke hole. A note or two on flange diameter. The biggest effect of flange diameter comes particularly from the drive side and in the form of torque transfer and a wheel’s ability to resist wind up during acceleration. Typically a larger flange will produce a better result in this category.

Center to flange: As measured by us. It’s been noticed that many of our numbers don’t match what is claimed by manufacturers. Our measurements are taken from center of flange to locknut. Some manufacturers provide outside of flange to center, while others provide numbers for both inside and outside but nothing center. Also some manufacturers may assume a 130 oln when their axle is not exactly 130. We use the actual oln measurement for our calculations. After the flange to center number is calculated it is rounded to 0.5mm.

Bracing angle: Based on a build using Kinlin XR300, 2x. Of course not all of these hubs would be recommended to be laced 2x, and with some it isn’t even possible. This was just a way to create an equalizer to show the differences in the hubs on a level playing field. Actual bracing angles and tension differences will vary based on the build.

Notes on Bracing angle: Bracing angle (or flange offset) is one of the most important factors affecting the lateral stiffness and stability of the wheel. 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.

As Dave Walker mentioned last year, rim stiffness has a great affect on the wheel stiffness as well, but since this is a hub review, we’ll focus on how the hub contributes to stiffness.

On a front wheel it isn’t difficult to get adequate offsets and stiffness. The limit is having clearance for the fork, and offsets of up to 40mm are usually fine… the wider the better the lateral stiffness will be. There has been some speculation that narrower spacings are more aerodynamic. It is also possible that a very flexible rim might experience a lateral wave if the combination of high tension and bracing angle and low spoke count were severe enough.

Bracing presents a conundrum on the rear wheel though, since the position of the DS flange is dictated by the 130mm dropout spacing, the wide cassette, and providing clearance for the derailleur. Because of this the spacing from the center of the wheel (and rim) is “stuck” being only ~16-19mm from the DS flange with a 130mm dropout width. Campy and Shimano/SRAM 11 spd hubs are generally in the 16-17mm range due to their wider cassettes, and 10 spd Shimano/SRAM specific hubs *can* be in the 18-19mm range. 11 spd hubs are inherently disadvantaged when it comes to making a stiff wheel.

The spacing on the NDS can be whatever the hub manufacturer wants. If it is same as the DS, then both sides will have the same tension… but lateral stiffness and overall stability will be very low. If it is twice as large… say 36mm… the NDS tension will be *half* as great as the DS, but lateral stiffness will be ok. The dilemma here is that a high bracing angle is good for lateral strength and stability, but lower tension on the NDS could cause these spokes to go slack when subjected to high radial loads. When spokes go slack the stiffness of the wheel goes way down and bad things can happen… from spokes coming loose due to nipples unwinding, to “taco”, wheel failure, etc.

So as you can see, the trick here is to find the best compromise. Now that nearly every hub offered is 11 spd compatible, we’ve seen some convergence of hub geometry. Most manufacturers are now “cheating” a little by making the axle 131mm long rather than 130mm long. I think Alchemy was the first to do this several years ago. This 1mm increase gives you an extra 0.5mm of DS offset capability. The maximum possible DS offset is now 18mm and most hubs are around 17mm.

Triplet lacing: Another approach to solving the issue of bracing angle and spoke tension on rear wheels is one that we think will become more prevalent as Shimano 11 speed with its Campag like dimensions takes hold in the market in coming years. The triplet or 2:1 lacing pattern on a rear wheel features 2 drive spokes for every 1 non-drive spoke. Because the non-drive side spoke head sits further out from the centerline of the hub it has lower spoke tension — typical non-drive side spokes may have only 45-50% of the tension of the drive side. On a 24h triplet rear wheel. You have 16 drive spokes and 8 non drive spokes. When you take away half of the non drive spokes the ones that are left have to pull twice as hard against the drive side spokes — effectively doubling the tension on the non-drive side. So if the non-drive was only 45% of the drive side and the triplet pattern doubles the tension on the non drive it is now only a 10% difference. Another benefit of the triplet pattern is that the drive side spokes are tangent which makes for the most efficient power transfer.

Like with all things compromises come with tradeoffs. One is that because you take away half of the spokes on one side of the wheel you lose lateral stiffness. To compensate for this you need a hub that has a wide flange spacing, ideally designed for triplet use. Another is that the rim needs to be stiff and center-drilled (holes not offset towards the flanges), which can limit rim choice. Also, if one of the NDS spokes happens to break, the rim could warp in an extreme way. None of the hubs in this review were specifically designed for triplet lacing.

The case for lighter spokes: There is an alternative to triplet lacing that can also address this issue; using heavier spokes on the DS and lighter (less stiff) spokes on the NDS, along with a higher NDS offset. Using lighter spokes increases their static “stretch” with a given amount of tension, and a small increase in the NDS offset can get back the lateral stiffness that would ordinarily be lost due to using lighter spokes.

There are now viable options in extra light stainless steel aero spokes (Sapim CX Super, and Pillar Mega Lite SS), plus a titanium spoke that is lighter still (Pillar Xtra Lite Ti). The stainless spokes are about 82% of the weight and stiffness of a CX-Ray while the Ti is about 60%.

Ordinarily we’d pick an NDS/DS bracing angle ratio of ~2.0 as being the best compromise for wheel strength and integrity. If we assume that the DS offset is 17mm for C11 and S11 hubs, then this is ~34mm. But if you used spokes that are 82% as stiff on the NDS, this would be increased to 2.2 (37.5mm) to retain the same lateral stiffness. For NDS spokes with 60% stiffness it would be 2.6 (44mm). In both of these cases you could increase the DS offset a bit more (and have improved lateral stiffness) and still have improved resistance to spokes going slack.

Hub shell material: Even though not every manufacturer will state the alloy they use, most manufacturers use a very high strength alloy (usually 7000 series), and at first glance this seems like a good idea. Stronger is better, right? In some applications though, we believe that 6061 might be a better choice. The reason is that 6061 has higher corrosion resistance, and more importantly resistance to something called “stress corrosion cracking”. The spokes exert concentrated and variable forces at the holes in the hub flanges, and high strength is a less important factor than ductility and corrosion resistance. Another advantage is that the softer alloy will deform more readily providing better support for the spoke in the flange. If you live and ride in a particularly corrosive area, anodized 6061 hubshells would likely last the longest. As far as we know, only White Industries and Alchemy use this alloy. Chris King won’t divulge the series of alloy they use for their hubshells stating only that it is proprietary.

Shimano 11 speed: The new Dura Ace 11 speed cassettes are wider by 1.85mm. This breaks with their tradition of keeping the cassette width about the same as they went from 8 speed, to 9, and then 10. The wider cassette will reduce the DS offset a similar amount, all else being equal. All the hubs in this review use the newer Shimano 11 dimensions.

Captured bearing vs free axle: These are two popular methods of hub design. Captured bearing means that the inner race of the bearings have a solid lateral support between them, either via shoulders on the axle or spacers that slide on the axle and join adjacent bearings together. The outer race is constrained in all cases by a press-fit inside the hubshell or freehub. In this design, the outer caps typically slide on and press directly on the inner race of the outer bearings, and no adjustment is necessary.

In the free axle design there are no lateral constraints on the inner races except for the external axle caps, one of which will be adjustable. If the adjustable end is removed, the axle can be slid out the other side. The adjustment is accomplished with either a threaded collar, a sliding collar with set screw, the cap itself threads onto the axle, or shims are used.

Either method can work well. With captured bearings the tolerances must be nearly perfect, else there will be a lateral preload on some or all of the bearings that will increase drag and wear. This is the biggest drawback. Some manufacturers have had more success with this than others.

With a free axle the lateral tolerances between bearings are not important, but the hub must be precisely adjusted, else there will be either a preload or excessive play. Also the outer bearings in this design are required to take all lateral loads (including preload if there is any). It is better to adjust these hubs with a little extra play rather than too tight. Either a threaded or sliding collar that allows for adjustment while the QR is installed, is a good feature to have with this type of hub… otherwise you must adjust with a little extra play to allow for QR compression.

Note that other aspects of hub design can also have substantial affects on wheel stiffness. Axle and shell stiffness, bearing size, tolerance, and arrangement, bearing to axle interface stiffness, and axle to dropout interface stiffness, are all important factors. Unfortunately, quantifying these is beyond the scope of this review

Note on Specs: All specs for this review are based on Shimano 11 speed versions of the hubs.

Note on brakes: All hubs in this review are for standard caliper brakes, we have not included any road disc brakes. Personally I’m not a fan of disc brakes on a road bike as I feel the disadvantages of them far outweigh any advantages they currently offer. However if they continue to gain popularity we will consider adding them to the next hub review.

A note on tools: When we talk about tools we will be talking about special tools. It will be assumed that a bearing puller and press is part of a standard tool kit. For the bearing press, we highly recommend the Wizard from Wheels Manufacturing, but designed by Jeremy from Alchemy. This is truly the most versatile press ever and with details such as an internally threaded shaft it’s uses go far beyond being a standard bearing press. We use it to pull axles, install axles, bearings, freehub bodies etc…

Thoughts On Each Hub

Now let’s get things started. Since many people mix front and rear hub brands we are going to look at them separately. So, on to the hubs…..

Front Hubs

REAR HUBS

We realize that this review was quite long, and unfortunately it only touches on a lot of the subjects. There is much, much more that goes into component selection and design for a custom wheel. However in the interest of keeping this review to a moderate length we decided to just cover the basics of hub selection. An alloy rim review can be found here. In the coming months we hope to write more on the subject including reviews on spokes and lacing patterns but for now we do hope that you found this an interesting read and have a better idea of what hubs may or may not be ideal for you.