One thing I don't understand is, surely a bearings durability is a function of its quality rather than its diameter? So I don't understand why a 28/22 setup would be inherently better than four 17s. Assuming the build quality of the bearings were equivalent. Anyone know the answer? Bb
Durability of a bearing is determined by quality and the correct size bearing for the job.
Cartridge ball bearings are so-called deep-groove ball bearings. This means that the balls run between a (shallow) groove on the outside of the inner race and inside of the outer race. The contact surface area of the balls determine how much stress the balls place on the race. Obviously a larger ball will create a larger contact area and vice versa. The contact area is important because the steel deforms when the bearing is loaded. The ball if flattened somewhat at the contact area and the race is dented. These cyclical flats and dents (they grow and disappear with each revolution) places stress on the steel that deforms/relaxes deforms/relaxes etc etc as the wheel turns. Eventually the steel fatigues and small flakes peel off, leaving the bearing rough and contaminated with bits of steel. This process is inevitable but if the bearing is too small for the job, the deformations are too large to remain under the steel's fatigue limit.
Further, the durability of the bearing is determined by the longevity of the grease. Each bearing is packed with grease which is kept inside by two seals. These seals are visible as coloured plastic rings or as two metal rings, depending on the requirement for the bearing. As I previously explained in either this thread or a another currently active one, the plastic seal cannot run on a dry surface. It will simply abrade away like a pencil rubber rubbed on a smooth table top. Therefore the seal is designed so that some of the grease inside the bearing escapes from under its lip and migrates outwards. This increases the seal life since it now runs on a layer of grease instead of dry. It also means that the grease is a consumable with a finite life. The obvious solution to make the grease last longer is to pack more of it in. It works, but this increases drag. In some applications, including bicycle wheels (bar one exception) this drag is absolutely meaningless in the bigger scheme of things. We create more air drag on our bicycles at 1 mile per hour than the bearing grease creates. However, the slightest bit of drag multiplied by two - there are two bearings in a freewheel - causes the freewheel to stick and throw the chain forward. Thus, freewheel bearings have to be very loose. The ideal bicycle bearing then would be one with say 30% grease in for the freewheels and another one for the hub with say 70% grease in. This would make sense because a freewheel bearing only works when you freewheel. When you pedal, both the inner and outer race rotates so the bearing doesn't actually work. But, they are not manufactured that specifically for bicycles.
I've previously explained how a single seal cannot separate two liquids. However, I'm yet to see a bicycle wheel company other than Suntour in its day that understood this principle and put in a second labyrinth seal to protect the catridge bearing from water. On its own a cartridge bearing cannot tolerate water. It needs extra protection.
Deep groove cartridge bearings manufactured in their millions for general purpose use in machinery, appliances, cars, hoovers, trolleys...you name it. They're come in zillions of sizes, two seal varieties (contactless and contact) and in a range of grades which relates to surface smoothness. But the one thing in common with all these bearings is that they are not designed for lateral loading, as in bicycle wheels. They are designed for straight-ahead running. A good example would be a bench grinder. The wheel spins radially and receives no axial loading at all. A car wheel receives huge axial loading when the car corners and its wheel bearings are therefore Angular contact bearings, not deep groove bearings. A deep groove bearing performs very poorly in lateral loading. If you just cup your hands in two Cs facing each other and imagine the ball running that groove and then offset the two grooves, you can see what happens to the ball. This is what happens in bicycle wheels and hence, the bearing doesn't live long.
Bicycle wheel manufacturers choose cartridge bearings nevertheless because this puts the bearing manufacturing in the hands of someone else. They use off-the-shelf products to save money. Good wheels have ACBs (angular contact bearings) which can be adjusted for play. The leader in that game is Campagnolo, followed by Shimano. Campag has the best system because it allows the user to buy the individual bearing races and replace them when they are damaged. Shimano doesn't have a replaceable race and if the race (not cone or ball) is damaged, it is time for a new hub. The reason Shimano doesn't do replaceable races is because of a Campagnolo patent that prevents them from doing so. Nevertheless a Shimano hub is still better than any cartridge bearing hub.
Some cartridge bearing hub manufacturers do use larger bearings to try and compensate for the problem but in reality, even a huge bearing and stacking up to five of hem on the axle (Hope does this), still doesn't improve the situation. Hope bearings live margninally longer than AC bearings.
Mavic attempted to get past the problem by offering an adjustable deep groove bearing system but this poses other problems, not the least of which is its poor understading of the second seal, mechanic ignorance about adjusting and in some cases, a "cone" that can only be used once.
In spite of all the evidence against cartridge bearing wheels, consumers will keep on buying them and keep dissing cup and cone bearings. The best wheel is the one that weighs the least, as we can see from the wheel review article in a post above. That is, until the bearings fail or spokes break.
