Kimmo

Some peeps talk about the replaceable outer race aspect like it should matter, but damn, you really have to be thick to let a bearing get that bad. The cone is always thoroughly toasted before the balls get affected, and then the cup. If it's a decent hub, it should be pretty well sealed, so if it's adjusted properly (which is hardly rocket science; there's usually a pretty huge window of acceptable preload), it should require next to no maintenance anyway.

Does my head in how so many folks can look at an Ultegra hub or similar, and look at hubs by pretty much anyone except Campy or Shimano (ignoring all the low-end loose-ball stuff like Joytech), and not see that one is fully designed for purpose using all bespoke elements, and the other is a result of the engineers just throwing up their hands and saying 'generic part goes here' when it comes to one of the most important aspects.

Unca_Sam

@Kimmo: the engineers involved here were concerned with how to 'cleverly' kill two birds with one stone. I'm sure they saved gobs of money simply because they didn't need to shut down a production line to retool for a longer axle.

WizardOfBoz

I agree, Unca. But it kind of sucks that a fairly mid-premium wheelset ($~1000) has design weaknesses that I could fix using the drawing tools in PowerPoint in about a half hour. I mean, the excessively long bore for the bearing mount (the holes in the hub that have an ID that matches the bearing OD) is so obvious. And frankly, fixing this might have saved money. Reprogramming a CNC lathe is pretty easy for those experienced in it. An improved design would use a hub that would be about the same, but the depth of the bearing bore would be less (so that's less machining.) The axle would be longer, but this is minimally machined tube. You'd change about 2 parameters for the hub and 3 for the axle in their CNC programs. The plugs (which I assume are machined from larger diameter solid bar) would be shorter and again, reprogramming would be about 5 parameters. As to material costs, you'd buy a bit longer cheaper lower diameter tube and a bit shorter more expensive bar stock. So material would be cheaper. Machining the tube would be about the same. Machining shorter plugs would be less machining. I actually think that they might have saved money in manufacturing had they implemented a proper design. If the Bontrager diagram (not mine, theirs, in post #1 ) reflects current design, then they DID improve the design by making the axle longer, and the plugs shorter (with less unnecessary distance between inner and outer hubs). I'm disappointed in Bontrager on this one.

Mad Honk 

Wiz, Kimmo, and Unca,
At what point in the design process, did anyone think about down the line problems that might occur? I'm thinking (just like the last girl I asked out) NEVER! These guys are taught to do the quick and dirty production process, and save the company as many dollars as possible. So post production problems are up to the QA people who will learn how to explain the problem away. Stoopid easy and stoopid simple is the new way of engineering, and don't try to think about the future in this,,, more to come in the next generation... Smiles, MH

Andrew R Stewart 

MH- I was just talking about this aspect of current bike design/manufacturing with a co worker that I will likely end up teaching how to build a frame. The life cycle of a modern bike is less then the industry wants you to believe. Why? Because the marketing side (and one can lump into this the world of magazines, blogs and such that parrot ad spec/hype) are working harder to create a reason to buy the latest/newest with the promise that your Strava status will go up.

We're never going back to the abundance of small replacement parts availability being important or frame resiliency is what makes a frame better Andy