does "indexed steering" necessarily require complete headset replacement?
 

My campagnolo headset-equipped road bike is suffering from a severe case of "indexed steering". It just goes clunk into a straight ahead position.

does this necessarily mean the headset needs replacing ?
or could a simple replacement of the ball bearings be to cure?

I asked because I can repack the bearings in a headset myself (this afternoon) but for a HS replacement, I would have to hand it over to an LBS.
and I don't want to bother going to get the bearings, doing the bearing replacement only to find that the HS necessarily needed replacing.

if there is even a reasonable chance that new bearings would patch things up, I'll happily give it a try first before the time and expense of a HS change.

thanks

Peter

If you've caged bearings you might be able to fix the situation by putting loose bearings in there, as you'll need (typically) one more bearing than the caged set, and they'll not line up with the divots.

Yes, you can get by, we did it many times when working a shop in a student market where budgets were limited. Loose balls does seem to make the problem less pronounced, though Campy uses a full-complement cage, so you won't necessarily get more ball bearings in. Most of the damage/effect is at the lower cup/cone, as the upper does not suffer the same compression combined with the back and forth action that creates the problem. Campy generally uses 3/16" balls, but use whatever matches your current equipment. Fill the cup till the last one does not fit and remove it for the correct count.

I know we sometimes removed one of the lower headest parts and rotated it very slightly to try to get a mismatch of the damage, but I can't recall if that helped at all (over 30 years ago)

Im a campagnolo person to start, secondly i believe that you are talking about a threaded headset right? If the answer is yes probably the best thing for you to do is to change the headset for a shimano dura ace, an old 600 (if you can find one brand new from 1991 or so that is the 1st generation with water seals and stuff) or maybe a nice stronglight with needle bearings. As i said before, i love campagnolo but their threaded headsets and old BB sets always left a lot wish compared with the japanese ones. Those 3 sets i mentioned once are set you can even go underwater with them 10 times and they will continue working perfect w/o taking them appart even for a single inspection after 10 years or 20 years.

I know im talking about heresy, but it is the cruel reality.

cya

The shimano headsets solved the problem with the fretting wear on the cones if you can find them. The Kris King and Crane Creeks wil also do the job.

Even with the loose balls solution at best you're just buying time. The balls will thunk past the brinneled wear mark and damage themselves a lot sooner than normal. The eventual flat spots on the balls will wear away at the rest of the bearing race and soon the whole set will need replacing anyway.

But if you do go with fully packed loose balls it'll likely last the rest of this season unless you typically rack up thousands of miles a season.

Also the likely cause of the issue, assuming it's not a split headset race, is too LITTLE preload when it was set up. They should be set tight enough that you can feel a little drag in the headset. Not "cogging" tight but a little tighter than it was when the play first wasn't there.

Damage is from fretting. It's lubrication failure. There's fore-and-aft motion in the steerer (typically road vibration), and no rotation (except when you steer). this depletes the film of oil on the balls, which produces lots of friction, which produces heat, which welds the parts together microscopically, and then they're ripped apart. Getting the pre-load right certainly helps prevent the damage from fretting, but it's hard to get right if there's already damage.

Unless you go to a Chris King or other premium headset, replacement, even with an upgrade, isn't going to be a bank breaker. Get it fixed right. bk

Interesting to note that in the early days of rail transport, automotive wheel bearings suffered the same fate when transported over long distances. They couldnt figure out why bearings were trashed from the factory. When the root cause was found, The industry quickly resolved the issue by suspending the cars by axles and letting the wheels float. Didnt eliminate but greatly reduced it... Cartridge bearings are the best way around index shifting, just replace when steering goes bad... Im assuming that the factory preloaded auto wheel bearings properly and they still had the issue... Im not sure there is a way to avoid it.

I had not heard about fretting before now. In looking up this failure mode all the examples I'm seeing are more related to lubrication failure via moisture or other corrosive contamination that leads to rusting of the bearing parts and a resultant pattern left etched into the races. The other reference to fretting that I found involved the races spinning in the housings or on the shafts independent of the bearings themselves.

I've seen a couple of headset races with a pattern like this and could not for the life of me figure out what the problem had been. I assumed that it was due to some rust from sitting and that the pitting had been "polished" out by further use until the races formed a little hill and dale like pattern. Now I realize that these races had failed from fretting be it corrosion based or lack of grease based.

What you go on to describe here sounds more like the references I found to a failure mode called "spalling" but not quite. Spalling, occurs where the instant point load goes well beyond what is needed to cause a brinnelling mark and a part of the race actually fractures and comes loose from the parent metal.

The difference seems to be that spalling or brinelling is localized to the fore and aft points and is more likely to occur in a bearing with incorrect preload while fretting is more likely to occur in a bearing with a lack of lubrication and where there is significant vibration being absorbed. Along with the fretted bearings I've observed I've also seen more than a few headset races and wheel bearing cones where the high point loaded spots are really badly worn into a deep notch or short groove due to the bearings pinching past those points rather than rolling smoothly. This failure mode was clearly due to either brinelling or spalling.

I also found the reference to Jobst Brandt on this and it would appear that you're a fan since this is pretty much a direct quote.


Your post about fretting got me curious so I went searching. I turned up this write up about bearing failure that is quite interesting. There's more ways to mess up a bearing than I thought.... :D

http://www.brighthub.com/engineering...les/16068.aspx

Fretting in this article (in part 7) appears to be more related to the nature of the race mounted in the housing or onto a shaft. But there's other references to fretting on the actual bearing tracks to be found via google using "bearing fretting". They all seem to include references to corrosion. Fretting, from all the examples I can find, seems to be a more overall mode related more to bearing contamination due to corrosion or lubrication failure due to extreme vibration on standing bearings. The interesting thing is that this can occur, and in fact seems like it mostly occurs in bearings that are correctly preloaded but that are not in motion but where there is vibration applied to the bearing so that it still has to withstand a loading but without the aid of proper lubrication. But all the references I've seen for this sort of failure are all around the bearing instead of being limited to the fore and aft points. However there's no doubt that such damage would result in a notchiness or cogging.

From all this there appears to be four possible failure modes that are possible in a headset. Brinelling, spalling, fretting and finally just plain ol' splitting of the race that leaves a full depth crack that forms a gap.

So I guess that without seeing really good pictures of the damage in this case it is hard to say if it is a fore and aft only damage related to brinelling or spalling or if it's a full circumference damage that would indicate corrosion or poor lubrication and vibration based fretting.

But in the end one thing we can all agree on is that once the races are damaged to where you can see and feel a notchiness nothing is going to make the steering feel correctly smooth again. It's best to just bite the bullet and replace the headset once the races are not in ideal condition.

Threaded headsets, either get them cheap, and lots of them, or pay for a good one, which nowadays means Chris King. American Classic's Trilock was/is the best I've ever seen or used though.

I've been getting some Tange Passage headsets for dirt cheap. $10. Decent quality.

There are about a dozen ways to kill a bearing, short of a bomb. many real failures are a combination (bike gets ridden in the rain, gets water in the headset bearings. grease gets washed away, something rusts, rust causes mechanical damage (it's a decent grinding compound...)). Barden do a nifty pamphlet that explains some causes:
http://www.bardenbearings.com/scan pdf/Bearing Failure Brochure.pdf

Every headset that I've ever looked at with the indexing problem has had fretting damage. I've seen headsets that have failed for plain lubrication failure, from over tightening, from using lousy grease. But I've never seen a headset that wore the balls out.

Fretting is a pretty common failure in bearings that are exposed to vibration that they're not designed to cope with, as happens in storage, transport. It doesn't take much rotation of the ball to renew the film of oil, so it's rare to see it in bearings that rotate regularly.

Fretting is also a common cause of failure in parts that aren't supposed to move relative to each other -- like the left pedal to crank interface. If you have to change the threading to keep it from coming apart, you're doing something else wrong. Fretting used to kill car wheels, until spherical seat lugs were used. It's poorly understood, and widely ignored, because it's hard to understand.

Well, I took apart an old headset and found some balls with flat sides and distinct fore and aft spalling or brinelling damage. That was pretty extreme though. I've also seen headsets in both bicycles and motorcycles that again had obvious brinelling damage only in the fore and aft zones due to the preload loosening up and play developing. In those cases the balls were OK to the naked eye. On the cases that occured with motorcycles the "fix" was to go with tapered roller bearings. In the case of the couple of bicycles I just tossed and replaced the headsets.

But like you I've also run into a couple of fretted bicycle headsets as well. I guess I was "lucky"? to find the other ones where the damage was more clearly caused by improper or lost preload tension.