Joining VeloClub not only supports the work we do, there are some fantastic benefits:
by Matt Wikstrom
September 18, 2018
Photography by Matt Wikstrom
The evolution of road bikes may have provided cyclists with frames that are lighter, stiffer and more aerodynamic, but there has also been a substantial increase in the number of bottom bracket and crank axle designs. The once ubiquitous threaded bottom bracket shell has given way to larger threadless designs while the diameter of crank axles has also grown.
Both may have helped elevate the performance of contemporary road bikes but consumers have been left to contend with myriad options and some frustrating incompatibilities.
In this post, Matt Wikstrom updates his original article from 2014 with a look at the range of bottom bracket and crank axle designs that are found on the market today and discusses the details that are important for matching one to the other.
When I started working as a mechanic in the mid-90s, there were essentially two types of bottom brackets for road bikes, English- or Italian-threaded. The only other thing that varied was the length and offset of the crank axle.
Two things stand out from that time: first, servicing a bottom bracket was labour-intensive; and second, there were rarely any complications when fitting a new crankset to a frame. Few riders ever seemed to complain about clicking or creaking from the bottom bracket, however the cranks were troublesome, sometimes creaking, or more often, coming loose on the square-taper axles that had dominated the industry for a couple of decades.
Things were changing, though. In 1992, Shimano started bringing sealed bottom bracket bearings to the masses with the introduction of its innovative cartridge system, and then came Octalink in 1996, a larger diameter crank axle with a new interface for the crank. Meanwhile, Magic Motorcycle developed external bottom bracket bearings so that an even larger axle could be used for its ground-breaking cranks.
Shimano’s Hollowtech II cranks, unveiled in 2003, essentially ushered in the current era of crank design by integrating the axle into the design of the crank. From that point on, the weight, stiffness and reliability of cranks improved, plus, they were easier to install and service.
Italian-made frames often feature Italian-threaded bottom bracket shells, such as this classic Concorde Squadra from the ’90s with Cinelli’s iconic spoiler shell.
BSA-threaded shells are a common sight for frames made from metal, be it steel, alloy, or titanium.
T47 is the newest threaded bottom bracket design that is well suited to 30mm crank axles, yet it can be easily adapted to smaller diameters like Campagnolo’s 25mm Ultra-Torque axle.
Cannondale has remained true to BB30 since it introduced the design in 2000.
BB86 was the first threadless shell that was specifically developed to suit carbon fibre frames.
At face value, Trek’s BB90 may look a lot like BB86 but the internal diameter is a little smaller.
PF30 is a direct descendant of BB30 with a slightly larger shell diameter so that cups could be used with the same bearings.
BB386 was developed by expanding the width of PF30 to match that of BB86.
The first of the new wave of bottom bracket designs broke in 2000 when Cannondale unveiled BB30 at the Tour de France. The oversized threadless shell was designed around a 30mm crank axle with bearings that were pressed directly into the frame. Rather than patent and protect the design, the company offered it openly to all frame manufacturers to encourage uptake by the industry.
The rise of composite frames encouraged further development of threadless bottom brackets, and the designs — PF30, BB86, BB90, OSBB, BBright, BB386EVO — that emerged over the next ten years were a mixture of open and proprietary inventions. Importantly, none of these designs would go on to become a standard for the industry and to this day manufacturers remain free to adopt (or invent) any bottom bracket design that satisfies their needs.
While consumers may have grown weary of innovation in this realm, there is no indication that it is slowing down. The recent introduction of the new T47 oversized threaded bottom shell as well as SRAM’s 29mm DUB axle prove that manufacturers — big and small — are still prepared to experiment with new designs.
When manufacturers started developing new bottom bracket designs, two things happened: first, the diameter of the shell increased; and second, the width of the shell tended to increase, too. Both allowed engineers to increase the lateral stiffness of the crank and the frame, which has been embraced with enthusiasm by racers and enthusiasts alike.
Compared to the once ubiquitous BSA-threaded bottom bracket, the diameters of the new designs are up to ~30% larger (Figure 1A) and the width of some shells has increased by about the same amount. While most bottom bracket shells are symmetrical, there are a couple of asymmetrical designs, such as Cannondale’s BB30A and Cervélo’s BBright (Figure 1B), where the non-drive-side of the shell has been extended beyond the traditional 68mm width.
Figure 1: a comparison of some common bracket bracket shell designs for road bikes. (A) Shell diameters have increased with the introduction of new designs. (B) The width of the shells has also tended to increase, and while most a symmetrical, there are some asymmetrical designs.
One thing that should be obvious from comparing the various shell diameters is that they will all require different-sized cups or bearings. Indeed, there is little interchangeability of parts for the bottom bracket shells found on the market today, so in the event that a frame is replaced or upgraded, a change in the bottom bracket shell will probably demand new cups/bearings.
Unfortunately, frames are not often marked with the specifications for the bottom bracket shell, and it can be difficult to determine what they are on the basis of sight alone (especially when the cranks are installed). Changes in terminology have added to this confusion, so the only reliable way to identify some bottom bracket shells is to measure the width and diameter after the cranks and cups/bearings have been removed.
Table 1 details the key specifications for many, but not all, bottom bracket shells that can be found in road bike frames today. The number of threadless shells clearly outnumbers threaded shells because they are better suited to composite materials. Nevertheless, some steel, alloy and titanium frames feature threadless bottom bracket shells, just as some composite frames have threaded alloy inserts for fitting threaded bottom bracket cups.
Table 1: Overview of the specifications for common bottom bracket shell designs found in road bikes.
In general terms, most cranksets can be fitted to a variety of bottom bracket shells, though much of this compatibility depends upon the availability of suitable hardware (see next section). Any incompatibilities that do arise are often related to the length and/or diameter of the crank axle.
In the past, crank axles were classified according to length and could be exchanged to suit different frames. That no longer applies to most contemporary cranksets since the axle is integrated with one of the crank arms, so if the axle is not long enough for the bottom bracket shell, then the crank will have to be replaced.
Table 2: overview of specifications for the common crank axles for road bikes.
The distinction can be a difficult one to make since manufacturers often don’t spell out the spacing for their cranks or the length of the axle. Thus, any incompatibilities may be not be obvious until an attempt is made to install the crank. In practise, though, the matter is often decided by the availability of hardware for any given crank and shell combination.
The diameter of the axle can have an impact on the compatibility of a crankset, though this is much less common. One clear example concerns Trek’s BB90, which cannot accommodate a 30mm crank axle. The narrow diameter of BSA- and Italian-threaded shells can also be problematic for 30mm axles, however oversized external bearings can address at least some of these difficulties.
Table 2 details the diameter of the axles for many, but not all, cranksets on the market along with known incompatibilities. For those cranks with 24-25mm axles, steel is typically used, while 30mm axles are normally made from alloy. The difference in material has some impact on the final weight of the crankset, and indeed, the lightest cranksets favour 30mm axles.
While modern road cranks are compatible with many different types of bottom bracket shells, some combinations are better than others.
Shimano (left) and GXP (right) crank axles share the same diameter — 24mm — however the end of each axle is quite distinct. The diameter of the Shimano axle remains unchanged while the GXP axle steps down to 22m to provide a shoulder for the non-drive-side bearing.
Campagnolo’s Ultra-Torque (left) and Power-Torque (right) crank axles are both 25mm in diameter, however the former is split in the centre, while the latter is not. The difference affects the placement of bearings on each axle and therefore requires different hardware for installation.
A Praxis M30 (top) and BB30 (bottom) crank axle share the same diameter, however like GXP, M30 steps down to a smaller diameter (28mm) at the end of the axle to provide a shoulder for the non-drive-side bearing.
If it isn’t clear from the discussion above, the bike industry is highly inventive, and manufacturers have always been willing to experiment with new bottom bracket designs. Most come to life as a proprietary feature for a new frame design. Cervélo’s BBright and Trek’s BB90 are two good examples of this, but there are other, lesser-known designs that occupy the same realm.
Wilier designed its BB94 (later re-named BB93) around Campagolo’s Ultra-Torque cranks. Seats for the bearings were moulded into the frame à la BB90 so there was no longer a need to install cups. Wilier was able to provide two bearings sets, one to suit Shimano’s Hollowtech II cranks, and another for SRAM’s GXP axle, that could be pressed into the seats so that those cranks could be installed. This design was subsequently replaced by BB386EVO.
Look created its massive BB65 shell in order to accommodate its one-piece Zed cranks. It’s a design that is very similar to one-piece BMX cranks that require a bottom bracket shell with an internal diameter of 51mm. In the case of BB65, the shell has a diameter of 65mm and measures 90mm wide while the diameter of the crank axle is a massive 50mm. BB65 is still used by Look for some of its frames (eg. 795).
The original OSBB shell created by Specialized for the first Venge (c. 2011) was 61mm wide with a 46mm diameter, so it was essentially a narrower version of PF30. A few years later, OSBB was replaced by OSBB Carbon and OSBB Alloy, however there was nothing novel about these designs, they were just re-labelled versions of PF30 and BB30, respectively. That distinction has disappeared for the company’s newest road bikes and now OSBB refers to a shell that is 68mm wide with a 42mm diameter, aka BB30.
When Colnago created the C60, the company implemented a new bottom bracket design dubbed ThreadFit82.5. Comprising a pair of threaded rings that secure an aluminium shell within the carbon fibre bottom bracket lug, ThreadFit82.5 has an internal diameter of 41mm that is directly compatible with all BB86 hardware. Threadfit82.5 remains a feature for some of Colnago’s current framesets, including the new C64.
Finally, Cannondale has created a few variations of BB30 to suit specific models in its catalogue. BB30A, discussed above, is just one of those: there is also PF30A, BB30A-83, and PF30A-83. PF30A has the same asymmetrical 73mm shell as BB30A but with a larger 46mm diameter suit PF30 cups, while the -83 variants have a 10mm wider shell to provide more tyre clearance. All are in current use, and curiously, Cannondale applies the same BB30 logo to the frame regardless of whether the shell is BB30, BB30A, or BB30A-83.
The bottom bracket of any frame is designed to fulfil one simple, yet crucial task: housing a set of bearings for the rotation of the crank axle. Given the size of the loads involved, reasonably large bearings with generous races are best suited to this task, though it is not strictly necessary that they be housed within the bottom bracket shell.
Some bottom bracket designs (e.g. BB30) provide a seat for the bearings within the frame while the rest depend upon some kind of cup (alloy or plastic) that is either threaded (e.g. BSA) or pressed (e.g. BB86) into the shell. A sure and accurate fit is necessary for resisting creaking under load, yet it is important that the hardware is reasonably easy to install and remove. In this regard, threaded bottom bracket shells have proven more reliable, however the industry has yet to create a foolproof system.
For those assembling a new bike or replacing the cranks, bottom bracket hardware is rarely supplied with either. Instead, it must be purchased separately, and this is where an appreciation for the specifications for each is indispensable. Even then, it can be a confounding experience due to the sheer number of potential combinations.
Shimano’s Hollowtech II crank design is easy to install and service and the system has proven very robust. The company offers just three options for bottom bracket hardware: BSA (shown), Italian, and BB86.
SRAM’s GXP cranks are well suited to BSA and BB86 bottom brackets, hence the company has hardware to suit. These cranks can be fitted in larger shells however owners will have to look elsewhere for suitable hardware.
BB30 cranks are best suited to large diameter shells such as BB30, PF30, and T47. SRAM has hardware to suit BB30 and PF30, while other brands offer hardware to suit other shells, including BSA.
Campagnolo’s Ultra-Torque cranks have bearings fitted on each half of the axle, so all that is need is a suitable cup to match the size of the bottom bracket shell. Campagnolo accommodates a wide range of possibilities with its range of Ultra-Torque cups and the 25mm axle will fit most shells.
Campagnolo’s Power-Torque cranks ultilise the same right-hand bearing and cup arrangement as Ultra-Torque coupled with a left-hand cup fitted with a suitably sized bearing. Campagnolo’s range of Power-Torque hardware options is not quite as extensive as Ultra-Torque, which creates a few incompatibilities.
Praxis M30 axle design now permeates its range of cranks with a diverse range of bottom bracket hardware to suit a wide range of shells.
To its credit, the industry has done a good job addressing the number of potential crank and bottom bracket combinations with an enormous range of products. Now, almost every crank on the market can be installed in any given frame.
Most of this hardware takes the form of suitably-sized cups and bearings to fill in the gap between the axle and the shell, though spacers, shims, and/or reducers are sometimes needed. Crank manufacturers have shouldered some of this work by supporting their preferred combinations, while enterprising aftermarket brands (e.g. C-Bear and Kogel) have been founded on a commitment to tackling those that have been overlooked.
Be that as it may, incompatible combinations still remain (see Table 2 for some examples), and for those that are affected by them, there is very little that can be done to address the issue. As a result, it is worth considering potential incompatibilities before purchasing a new frame, crankset, or bike.
C-Bear is one company that has created an extensive catalogue of bottom bracket hardware that solves previous incompatibilities, such as fitting Shimano cranks in a BB30 shell.
When SRAM set out to design its new crank axle and bearing system — DUB (durable unified bottom bracket) — there were two important goals. The first was to address the bearing durability issue that has afflicted oversized crank axles; and the second was to ensure compatibility with all of the bottom bracket shells that prevail in the current market.
At the heart of the system is a 29mm axle, a seemingly trivial distinction, but according to SRAM, it makes for a better selection of bearing sizes than those available for 30mm axles. The result is a crankset that enjoys the benefits of a lighter and stiffer axle with bearings that are as durable as those matched to SRAM’s GXP axle.
SRAM created a suite of bearing cups for the new cranks so they can be fitted to the more common bottom bracket designs (BSA, BB92, BB30, PF30). At this stage, DUB is limited to SRAM’s MTB cranksets, however it will be surprising if it doesn’t become part of the company’s road catalogue in the next year or two.
MTB have long employed wider bottom bracket shells and crank axles so as to provide more clearance for fatter tyres and shedding mud. In the beginning, an extra 5mm was added to the width of BSA to yield a 73mm shell, so there was no need to change the cups and bearings. All that was required was a wider crank axle, which was a relatively simple demand for the industry to meet
In the time since then, the industry has continued to apply this strategy, and thus, many of the crank and bottom bracket designs found in the road market have been carried over to MTB with one simple modification: an extra 5mm (or more) for the width of the shell and axle.
BSA, BB30, PF30, and wider versions of BB86 are the most bottom bracket designs for MTB, with shell widths starting at 73mm. Downhill and fat bikes have even wider shells, however the formats remain unchanged. As for crank axles, Shimano’s 24mm axle, GXP, and 30mm axles predominate the current market, however SRAM’s new DUB design (see sidebar) has added a 29mm axle to this short list of options.
SRAM’s new DUB axle design is compatible with both small threaded shells (BSA) and large unthreaded shells (PF30). Photo: James Huang.
These wider bottom bracket shells and crank axles have yet to appear on cyclocross and gravel bikes, though that may be changing. Cannondale recently opted for an 83mm bottom bracket shell (BB30A-83) when re-designing the SuperX and while the MTB-inspired width works well to provide extra tyre clearance, it does make for a couple of complications.
To start with, this wider shell is incompatible with the majority of road cranks, so MTB cranks must be fitted to the bike instead, and with that comes a wider q-factor (i.e. the distance between each pedal). For those riders with a wide stance, this will be a welcome change for a cyclocross bike, however road riders that prefer a narrow stance will have no way to compensate for the wider q-factor.
Given the way that bottom bracket shells have been growing for MTB in response to enthusiasm for wider tyres, it won’t be surprising to see the same thing happen for at least a sub-section of gravel and bike-packing offerings. These bikes already blur the distinction between traditional categories, plus, a wider q-factor may also help the general appeal of these bikes.
Road cyclists, especially racers, have been obsessed with the stiffness of frames and various components for decades. And it’s a preoccupation that has continued despite the lack of evidence that it has anything to offer other than an edge in the closing moments of a race when riders are sprinting out of the saddle.
The cranks and bottom brackets of road bikes have received more attention than most parts of the bike, presumably because of the proximity to that race-winning effort. In this regard, bigger bottom shells and larger diameter axles have been important, though it is the relocation of the bottom bracket bearings so that they are much closer to the cranks that has probably had the biggest impact.
As a result, contemporary cranks now flex less than earlier designs. When Fairwheel Bikes measured deflection of various cranks under a load of 200lb (~91kg), a square-taper crankset (2006 Campagnolo Record) exhibited up to 50% more deflection than contemporary cranks. In contrast, there was generally much less variation in deflection for modern cranks, and interestingly, the diameter of the axle had no impact until stiffness was compared to the weight of the crankset.
According to Fairwheel’s calculations, a reduction in crank flex has the potential to save a rider some energy, though the magnitude of these savings is just a handful of watts. However, that is based on the assumption that a flexing crank cannot return energy to the rider, which may not be the case, since Jan Heine has data that suggests that this kind of flex can reward the rider with free speed.
Thus, there is little point in concentrating on crank stiffness in the current market, if only because the amount of variation between products is relatively small. Weight, cost, and appearance do far more to distinguish the current range of cranks, so shoppers should feel free to let any and all guide their purchase decision.
The bottom bracket is perhaps the most susceptible part of any bike. Water, grime, sand and mud are always quick to collect around the shell, so contamination of the bearings is inevitable. And while regular washing will do a lot to remove the obvious muck, water will make its way past the seals to attack the bearings, carrying particulates that will speed up erosion of the balls and races.
Before the widespread adoption of cartridge bearings in the ‘90s, a traditional cup and cone assembly supported the crank axle in the bottom bracket. It was a design that was well suited for contending with the combined loads that were placed on the axle. All of the parts could be serviced and the system was highly adjustable to the point where the chain-line could be varied when needed.
Cartridge bearings simplified crank installation and reduced servicing times but a lot of adjustability was lost along the way. The chain-line of contemporary cranks is no longer adjustable, but once a bearing starts to deteriorate, there is no need for labour-intensive servicing: the whole cartridge is simply replaced with a new one.
The range of options for replacement bearings is large. The balls and races within the cartridges can be fashioned from steel, stainless steel, or ceramics: low cost bearings are made from steel; stainless steel adds durability for a bump in price; and ceramics offer a bit of free speed due to a reduction in friction, but come at a significant cost.
Cartridge bearings with ceramic ball bearings that come from brands like CeramicSpeed have become a halo product for the bottom bracket.
The latter is a tempting proposition for racers and performance-oriented riders, but ceramics alone do not account for those savings. It’s a matter of highly accurate balls carefully matched to hardened races, both of which are expensive to produce. Cheaper ceramic offerings typically compromise on the quality of the balls and/or races and as a result, fail to provide the same level of performance.
At least some high-end cartridge bearings (typically ceramic offerings) are available with a choice of seals: a light road-oriented seal that offers a minimum of friction, or, a more robust seal that is better suited for keeping the grime at bay. The extra sealing adds friction but it’s a necessary compromise if the bearings are going to be used in a grubby environment.
The majority of cartridge bearings employ a radial design where the balls are sandwiched between two parallel surfaces. The alternative is an angular contact bearing with a cup-and-cone arrangement for the races (just like a traditional bottom bracket). At face value, there is no way to distinguish between the two, but on paper, an angular contact bearing promises to be more durable and perhaps better suited for the cranks because of the potential for combined loads. With that said, radial bearings generally do not suffer in this setting, and accelerated wear may be the result of poor seating and/or misalignment of the cartridges.
A small sample of the range of bottom bracket hardware that Kogel has developed to improve the compatibility of different cranksets for various bottom bracket shell designs.
In an era when an enormous amount of time and effort has been devoted to developing new designs for bottom bracket shells and cranks axles, it is perhaps a little surprising that this area of the bike continues to be plagued with difficulties. Indeed, there is no such thing as a trouble-free bottom bracket, but that can simply be a product of the amount of use and neglect that it has to endure.
In this regard, regular servicing has a lot to offer. Most cranks are easy to remove and without any small parts to spill out from within, it doesn’t take long to clean the axle and bearing cartridges. A smear of fresh grease on the cartridges is often all that is needed to restore the bottom bracket and eliminate creaking. With that said, there are many other sources for this kind of noise (e.g. chain, skewers, pedal axles, saddle clamp, seatpost, stem) yet the bottom bracket is often deemed guilty until proven innocent by a methodical process of elimination.
A system of reducers and spacers allows hardware manufacturers like Token to adapt one bearing set to a wide variety of cranks, however the extra pieces adds to the risk of creaking.
Praxis was an early innovator in bottom bracket hardware with a cartridge-like unit with an expanding collet to fill the bottom bracket shell. Now the company makes use of an elastomer ring (in orange) to brace the bearing unit within the shell for a better (creak-free) fit. Photo: James Huang.
Parlee Cycles relies on a deep shoulder to provide a more secure fit for its PF30 cups. Photo: James Huang.
Press-fit cups that thread together, like Enduro Bearings TorqTite, also promise a better fit for threadless shells. Photo: James Huang.
Instead of using two separate cups, BBinfinite houses both bottom bracket bearings in a one-piece aluminium unit that is pressed into place, and as long as the shell is up to spec, it shouldn’t creak under load. Photo: James Huang.
An ongoing problem with noise or bearing wear can often be traced back to some kind of poor fit, be it the crank, axle, cups, or bearings. In this regard, complementary combinations — large crank axles for large shells and small axles for small shells — tend to be the most robust.
Squeezing a 30mm axle into a small diameter shell (35-41mm) often makes for a compromise on bearing size, which can affect durability. Steel axles (24-25mm) that require shims or reducers to fill a large diameter shell (42-46mm) or bearings can start creaking because of an inaccurate fit. And a crank axle that needs a stack of spacers to compensate for a narrow shell can suffer the same fate. An accurate, fuss-free fit will bypass these issues, plus, the system will generally be easier to install and service.
These measures will only go so far when it comes to compensating for a poorly manufactured bottom bracket shell. Shell tolerances are normally defined in terms of fractions of a millimetre, which can be difficult for consumers and mechanics to measure. Similarly, there is no easy way to measure the roundness of the shell yet small deviations can have a profound effect on the fit of the bearings. Bonding agents or hardware that offers an adjustable fit can be used with some success, however a flawed shell really demands to be replaced.
Imperfect spacing of the cranks can also produce creaking under load. Some systems, like Shimano’s Hollowtech II, offer enough adjustment to bypass this issue; the remainder that depend upon careful selection of shims and/or wave washers can have trouble eliminating lateral play. Like bearing fit, this can be a matter of a fraction of a millimetre, so a lot of trial-and-error may be needed to address the issue.
All of these issues means that tackling a troubled bottom bracket can be a convoluted and time-consuming process. And in some cases, it may not always be possible to find a robust solution for the root cause, even when it is left in the hands of an experienced mechanic.
Preparation is often important for successfully fitting bottom bracket hardware. In the case of threaded shells, the threads often need chasing for a clean fit.
Threaded cups need a suitable tool for driving them into place. Photo: David Rome.
The range of patterns and shapes for the cups has been growing in recent years, which means more tools are required to contend with them all. Photo: James Huang.
The number of cup removal tools is smaller and not nearly as elegant. Photo: David Rome.
Wheels Manufacturing has come up with a set of fittings for its bottom bracket bearing press that allow common bearing sizes to be removed for the cups in the frame. While this makes for a welcome addition, it can’t be used to remove the cups from the frame, so mechanics still have to resort to some kind of hammering to get this job done. Photo: David Rome.
The last 20 years have ushered in a lot of innovation and refinement for bottom brackets and crank axles. The former has contributed to advances in frame design while the latter has come with improvements in crank design. Aside from added stiffness and less weight, cranksets are now easier to install and service, however the range of possible combinations and potential incompatibilities is more confounding than ever.
While some kind of standardisation could make a huge difference for the consumer, it would stifle innovation and slow progress for the industry. And at this point in time, there is no sense that bottom bracket and axle design has been perfected, or that a smaller number of designs could satisfy all the needs of the industry and the consumer.
Indeed, there is room for improvement, particularly where threadless bottom bracket shells are concerned. At present, various cups and bearings can be gently pressed into these shells, however their removal is another, more brutal, matter. Mechanics must inevitably resort to hammering them out of the shell with the risk that repeated replacement may be detrimental for the frame.
This is where threaded shells have a distinct advantage, but unfortunately, they are poorly suited to composite frames. Some aftermarket hardware employs cups that are threaded together, however this strategy is at odds with a threadless fit (at least one cup must be able to turn freely in the shell, which compromises its fit). Installing a threaded sleeve in a threadless shell poses its own set of complications, so what is really needed is a fresh strategy. Until then, consumers (and mechanics) will have to live with the current compromise.