Lawrence08648

I'm a firm believer that the factories do not put enough grease in wheels bearings or they don't put any grease in wheel bearings. How do you feel about removing headset bearings and greasing them? The wheel bearings are turning continuously but the headset bearings are turning a minimal amount and get very little use.

FBinNY 

Perversely the lubricant demands of stationary (and nearly stationary) ball bearings (the entire bearing, not the balls) can be more severe than when there's constant rotation.

When bearings rotate the balls spread the grease and continually keep the bearing races lubricated. But when stationary, or almost so, the vibration causes the balls to work their way through the grease film and beat up the races through a process called fretting or false brinelling.

It's this process which explains why the headset is often the first bearing to fail on a high quality bike.

So it's not whether I ever grease headsets. I always grease them, using the best grease I have on hand.

Burton

Higher end headsets with sealed bearing inserts are a lot better than the exposed ball bearing type that was the norm 20 years ago and is still common on lower end machines. For those ball bearing headset I completely agree with you and FB. The lower race seems to fail more ofter - maybe because when left in the sun the grease has a tendency to leak out more. The top is usually sealed against water but not the bottom. I guess someone figured gravity would keep water out. Unfortunately it also helps grease escape!

JohnDThompson 

Roller bearing headsets are also superior to parabolic cup and cone headsets.

BCRider

I agree. And this is why you find tapered roller bearing headsets on the better motorcycles or one can easily buy tapered roller bearing as replacements for most motorcycles.

On a bicycle? It might be nice but it's overkill. Cup and cone ball bearings have generally proven more than sufficient for bicycle applications. And now we have the new sealed "cartridge" angular contact ball bearings in some headsets.

Or are SOME of these cartridge headsets being delivered with tapered roller bearings?

mechBgon

I made it part of our assembly routine to add grease to traditional headsets, and to liberally grease the contact areas on cartridge headsets. For traditional headsets, it definitely prolongs the life (and on threaded ones, makes it more likely someone can get them apart 15 years from now). For cartridge headsets, it helps prevent creaks and ticks, as well as cartridges corroding to the cups and/or the upper wedge ring not wanting to release for fork removal.

At some bike shops, that level of proactive thinking would get a guy fired but I have quite a bit of lattitude. Tangentially...

LesterOfPuppets

The first aftermarket headset I installed got all grumbly after just a month or so of rain and snow rides. Top was dry as a bone and rusty. Uggh.

So from then on, I just assume that factories don't pump enough grease into any bearing.

gyozadude

Ditto the extra grease on headset bearings. I tend to use a thicker grease as well, or one with micro-filaments or fibres/additives mixed in to keep the grease in place (like Phil Wood grease or a thick automotive bearing grease). Helps keep the grease trapped in the crevices.

HillRider

The lower cup race and crown race fail in cup-and-cone headsets due to impact, not from water incursion or from the grease leaking out. The upper bearing just keeps the steerer centered and allows it to turn but the lower bearing supports the bike and rider weight and absorbs every road shock so it obviously will wear out first.

JohnDThompson 

Not sure about current designs, but the Stronglight A9 and B10 units we used by the thousands at Trek hold up quite well. The OEM price on the A9 was something like US$6.00/each.

davidad

No they are not. They fail just as often for the same reason as angular bearings. On a road bike the bearings do no move much so that they don't maintain grease and are exposed to fretting wear. shimano solved this problem with the combination of a taper seat and bearings.

LesterOfPuppets

Dunno much about cartridge bearing headsets, but my olde 2001-ish non-cartridge FSA headset has roller bearings in a plastic cage. They aren't tapered rollers though, rather cylindrical.

mechBgon

Unfortunately they didn't work very well... headset play or smooth rotation, pick one. The Onza Mongo II was pretty good, though.

BCRider

That's a very odd design of bearing in that case. Headsets pretty much need to use tapered rollers since they need to both rotate AND support. A cylindrical roller bearing in a headset would need additional thrust bearings to support the bike. A cylindrical roller bearing has no ability at all to withstand axial forces. And axial forces there's much of on a bicycle.

LesterOfPuppets

I guess needle bearing is how FSA describes them. Almost jibes with the needle bearing definition found in the Wiki, but I don't think the rollers are 4x longer than their diameter though.

On my FSA the rollers are arranged in a conical formation. I reckon that handles the forces from different axes well enough.

A-ha, found a good pic of replacement innards for a Stronglight needle bearing headset:

LesterOfPuppets

Thinking about it a bit I can't think of how tapered rollers would work for anything. There are tapered roller bearings but the rollers on those are still cylindrical.

JohnDThompson 

FWIW, I did an experiment where I installed a Stronglight A9 roller bearing headset on my commuter bike in 1987 and did *ABSOLUTELY NO MAINTENANCE* on it until 2007, whereupon I opened it up and found this:



As you can see, there is trivial fretting damage on the lower races and virtually nothing on the upper races.

I swapped upper and lower races, repacked, and will report back in 2027 with another update.

JohnDThompson 

Umm, if they're cylindrical, they're not tapered.

Tapered roller bearing in principle have lower rolling resistance than simple cylindrical rollers, but in practice are quite fussy about alignment.

Cylindrical roller bearings are less fussy about alignment, but do have sliding contact with the races due to the difference in radius between the inner radius of the roller and the outer radius of the roller. Since headsets, unlike hubs and bottom brackets, are not in constant rotation, this is not a significant concern in real-life use.

LesterOfPuppets

Ahh, yes now I see clearly in this sketch. It was difficult to see the conical nature of the rollers in photos of tapered bearings. Looking forward to the 2027 report