Steering head taper roller bearing

Tintin · Apr 12, 2012

DogT said:
The annular thrust bearings that came with my bike seem to be working fine for me. No sleeve, no tapered rollers to fall out.

Well, I don't know whether what I disassembled was the original stuff but my 1970 frame had the same cup'n'cone bearing stuff with 18 loose balls in as my Atlas did. That means 18 rollers to fall out in the first place.

One small advantage of TRB over ACRB actually is the fact that the roller/cage/inner ring assy can be separated from the outer race which makes installing them very simple. A bit more tricky on the ACRBs but still simple with correct drifts.

Tim

slimslowslider · Apr 12, 2012

Report this postReply with quoteRe: Steering head taper roller bearing
by Tintin » Wed Apr 11, 2012 8:27 am

comnoz wrote:
3. They don't require an accurate spacer and/or shims to set preload. Just a spacer of approximate length will allow the stem bolt to be torqued and keep the triple clamp assembly rigid.

That is exactly the point - the effort needed to do this correctly would make the costs explode. The ACRBs are a nice alternative but honestly they are obviously not necessary. However doing it right has a certain magix in itself ....

Tim

Seems like the spacer was deliberately designed longer than the length between the bearing recesses in the head stock. First you mount the bottom bearing, outer race fully home in the recess. Then spacer, then top bearing so far in that inner race just touches the spacer. Outer race is not home then in the top recess. Only the bottom bearing than takes the axial load (besides radial). Top one only locates radially, which it does easily and always. I guess, theoretically, if tolerances in head stock would be too big, when the wheel is off the ground, the weight of it all may pull bearings the extra length of the spacer.

An SKF 6205 has a Co of 7,8 kN, take max 50% for axial load.
Don't know what the actual axial load is that is to be considered, but looks to me the bearing is quite ok for the job.

Tintin · Apr 12, 2012

slimslowslider said:
Seems like the spacer was deliberately designed longer ...

Just to be precise: I was talking about a spacer in a paired axial bearing not the post-71 spacer. Off course this is longer, it is just a spacer which doesn't need to be anywhere near as precise as in a TRB/ACRB setup.

An SKF 6205 has a Co of 7,8 kN, take max 50% for axial load.
Don't know what the actual axial load is that is to be considered,...

Classis loadcase: Breaking under full load with high µ - basically a stoppie - plus a bumb in the road. Gives up to 2g easily. 360kg times 20m/s² is roughly 7200kN. Plus 50% axial load in a DGRB basically means it can't take any radial load anymore. 50% of 7800N is 3900N. Safety factor: 3900N load carrying capa divided by the actual load of 7200kN: 0.54. Zero poit five for.

.... but looks to me the bearing is quite ok for the job.

Sorry but no.

Tim

slimslowslider · Apr 12, 2012

Tim, that is 2G 100% axially? So diving with 2G?
I have not taken that many forks apart, the ones I did the bearings were ok. Still knocked them out for frame paint jobs etc., costs are little.
Where would this apparently being unsuitable manifest itself?
Maybe someone can chime in with experience on failed bearings as a result of excessive forces?

303 fmj · Apr 12, 2012

The only failed bearings I've seen were on BMW's. The tapered roller bearings. Owners who neglected to either have them adjusted periodically and greased had detents; divots; hammered into the races which you could feel as the steering head was rotated left-to-right. Looseness was the main culprit and also the use of too much grease and too stiff of grease which then causes the rollers to remain in one place on the race and make impressions--detents.
Tapered wheel bearings usually do not have this problem since they rotate all the time--not staying in one place such as the steering stem on a motorcycle going down the road for miles and miles without moving---loose bearings and road bumps take their toll eventually.

comnoz · Apr 12, 2012

I would have to agree with Tim, deep grove ball bearings are not designed for the load in the Norton steering head. That being said they have done a very good job considering they were a money saving bodge by the factory. Failure is rare unless the seal allows water into the bearing.

Several years ago at one of the INOA rallys back east I hit a large pothole with my bike. It was a hard enough hit to crack the from wheel bearing races and bend the rim. And yes I did stay on the bike [big "flying W'. I broke off the windshield with my head.]
After replacing the front wheel bearings I found there was a notch in the lower steering head bearing also. I went ahead and replaced both although the top was fine. Jim

slimslowslider · Apr 12, 2012

You certainly found out the hard way!
Then again, I guess it is fair to say any other type of (rolling element) bearing would develop a notch in such an incident.

Tintin · Apr 12, 2012

slimslowslider said:
Tim, that is 2G 100% axially?....

That is a very rough approximation obviously but if you see how a bike dives the axial force must be relatively close to the normal force. What I did to come up with this load case is a shortened version of what I do @work, for each structural part there is a defined load case (by our calc department) and the one I chose follows roughly the lines of what is done on suspension (although the g's are a little different on the cars I usually work on).

What I wanted to show is that if you follow the books and do a calculation according to common practise the inevitable result will be that theoretically they won't work. Why that is the case is the next question, either the load case is wrong (I never managed to lift the rear wheel on my Atlas) or the load ratings by the manufacturer are very conservative, e.g. . The thing is, the stuff works and is a proven and tested solution but still it is done against what bearing manufacturers recommend.

Failure on the TRBs is apparently not uncommon and they tend to bed in. Why this is the case I don't really know - overload and misallignment would also harm cup'n'cone bearings and ACRB. Maybe the higher stiffness of the TRB leads to overtightening in these cases? Which brings us back to the correct length spacer thingie BTW. One thing i do know is that the fork and the headlamp need a dedicated earth point to bypass the bearings unless you want to start tiny little random galvanisation processes. :wink:

Tim

ML · Apr 12, 2012

72Combat said:
I noticed front end juddering under braking on my '72, I discovered that the manual said it was sealed ball bearings. Is it a worth while mod to go to taper rollers or is that like starting an oil thread.....?quote]

Brake judder is not related to the steering head bearings. Its brake torque variation at the drum or disc due to uneven mating surfaces. The kinetic energy vibrates through a transfer path, i.e. the forks. There is probably zero movement at the steering head but the vibration path could still run through to the handlebars. I strongly recommend you remove the front brake components and take them to a specialist to measure and re-face as neccesary. So don't switch head stem bearings - just not going to help that issue.

Mick

lcrken · Apr 13, 2012

I'm with the folks who think the stock ball bearing system works just fine. I've raced Commandos with the stock bearings, Commandos with tapered rollers, and a featherbed with tapered rollers, and they all worked the same on the track. Converting a stock Commando to tapered rollers won't do any harm, if done properly, but I don't see any benefit from doing so, except for the need we all seem to have to keep "improving" our bikes. One problem with just substituting tapered rollers on a stock Commando is that the steering head nut is a fairly coarse thread, which makes it difficult to adjust to and keep the right preload unless you use an internal spacer, and getting that sized correctly is a real pain. I only used tapered rollers on a Commando along with a different front end, which had the usual fine thread nut between the top bearing and the upper triple clamp.

The best bragging rights will go to the first one who converts to ceramic rolling element bearings, with CNC machined magnesium triples (or fabricated from welded sheet titanium) and an aluminum steering stem. Keeping the stock forks and wheel for originality, of course.

Ken

Hobot · Apr 13, 2012

Alas books - equations don't cover all the real world proof of stock set up endurance.

Beware that normal spiral cut brake rotors loose significant grip compared to the cross hatched milling and or sand blast texture.

There is just not enough fiction loads in stem to benefit from lower friction of Expensive ceramics and too much wham bam in our thumbers to live on crank shaft, that leaves wheel and drive train and turbo's. If I ever need Bruce MacGregor gave me the p/n's for ceramic Drouin impeller bearings, but he's made due now for most a decade with just good steel turbo bearing w/o any cooling lube system.

If you do get restriction in the stem, how it shows up beside excessive effort is the bike fights your every correction to road texture twitching off the path. Its interesting to try as reveals the need of forks to ever so slightly oscillate to not fall down. A steering damper set on high for parking lot or flip backs in chicanes or correcting off road slip ups, can reveal what 1/2 way rusted up stem bearings feel like.

Steering head taper roller bearing

Similar threads