Tapered bearings..end float or pre load

[mgcjag]

Hi All......So as per the title whats the current thinking......for our E types the service manual specifies.. front/rear wheel/hub should be set with end float......lower fulcrum in rear hub carrier pre load......diff bearings pre load........Personaly from what i have read re tapered bearings in general they should be set up with pre load....So should we steer away from the service manual for wheel hubs and set them preloaded...Just to give my opinion iv set my wheel bearings at 0....no end float no preload.......will see how they go.........Steve

[Joes66]

Basically taper bearings are very strong and designed to be under some pressure
Rule of thumb after being in the trade for 35 plus years (new to e types) is that there should be no rocking or slack in the bearings once set up UNLESS ITS SPECIFIED in a workshop manual for some reason
If a manual says a way to set up a bearing using end float then this should be adheared to
Generally once a taper bearing is set up it should have no play but slightly stiff to turn by hand
Interestingly I recently watched a video rebuilding an independent jaguar rear suspension which was produced by jaguar cars
When he is re fitting the outer fulcrum bearings he mentions that if you have set it up properly with the shims the rod should be slightly stiff to turn but you should be able to turn it by hand
This is a feel thing that comes with doing many bearings on many different applications
Years ago I was head mechanic at the maidenhead MG centre and we would never send a car out with rock in the bearings
Having said that I have never in 35 years had a new wheel bearing I have fitted fail
One other point is that if you have fitted a new bearing it should be given 100 miles or so to bed in and then re checked and re adjusted if need be
Many workshop manuals dont give end float figures and an MGB front hub bearing is just the same as an e type one
Joe

[Joes66]

[angelw]

Hello Steve,
If the bearings are due for renewal, almost invariably, you will fins a groove the width of the Inner Bearing at 6 o'clock on the Shaft. This is due to the Bearing spinning on the shaft. Most stub axles being sold for E Types and others, are made so that the Bearings are a close running fit with the shaft. They are made this way for easier maintenance (the bearing and seal won't be left on the shaft requiring a puller to remove them). Adjusting the Bearings to be in Pre-load for this type of arrangement will only entice the bearing to spin and exacerbate the groove being worn in the shaft at 6 o'clock. Accordingly, its not advisable to set the bearings up in Pre-load in this assembly arrangement.

If a spacer is used between the Inner and Outer Bearings, the bearings can be locked in position by tightening the nut. Pre-load of the bearings is achieved by modifying the length of the spacer by adding shims to lessen the Pre-load, and subtracting shims, or shortening the spacer to increase the Pre-load.

An alternative to the spacer, is to use a stub axle where the Bearing Journals are an Interference fit with the bearings. With this arrangement, a puller will be required to remove the bearings when required.

Regards,

Bill

[mgcjag]

How about rear hub bearings....end float as per the manual or pre load...thanks...Steve

[angelw]

Steve Wrote:

How about rear hub bearings....end float as per the manual or pre load...thanks...Steve

Hello Steve,
The Rear Bearings have a reasonable Interference fit with their Journal; accordingly, I set these up with Pre-load.

Regards,

Bill

[christopher storey]

End float as in the manual. It fascinates me that people always know better than the manufacturer. There will be a reason for the end float specified. What it is I do not know , but conceivably it stems from the fact that as everything gets hot under running conditions, clearances alter

[angelw]

I put my trousers on in a similar manner to other mechanical engineers. You think that the Front Wheel Bearing system of an E Type, or others with a similar system is great, just because the manufacturer made it that way; a running fit of the bearing with the journal is not that clever. The working end of the Stub Axle and the Bearings are the same as used by a box trailer. Clearance for the front bearings is specified because anything tighter would guarantee the bearing to spin on the shaft. There are many cars that use Back To Back, Angular Contact Bearing that are under considerable preloaded.

Its a well known fact that life expectancy of Taper Roller bearings is in the Zero to Light Preload area. I'm fortunate in that, apart from my own engineering knowledge, I have the ear of a past National Design and Production Engineer of Timken Bearings; he spent his whole working life with the company.

[mgcjag]

Hi All.....I posted this topic becaus there were different opinions when searching the subject in our forum......Not becaus i wanted to know how to set my own bearings.......also i could not find just one post covering tapered bearings but spread out on dozens of posts........As we all know setting rear hub bearing is not a straight forward job i have done it quite a few times on various Jags..however its not unknown to have to reset these due to exessesive end float even though they were set correctly in the first instance.......not just by me but by others in the trade that im aware of.....this is why i wanted to discuss setting with pre load as opposed to the service manual end float.....Not that we know better than Jaguar or their reason for doing certain things...but the service manual is not 100% acurate with many revisions and service bulletin changes.....Steve

[christopher storey]

Steve : as far as I am aware, the workshop manuals specified 2 to 6 thou end float for the rear hub bearings throughout the entire 25 year life of Bob Knight's rear suspension design , and Jaguar are hardly likely to have gone to the trouble of providing 15 different sizes of spacer if this were merely an academic or optional setting. Incidentally, I do not think it is possible for excessive end float to develop in service , ( except perhaps in the rare instance of the bearing collapsing ) for the simple reason that to increase the end float a thicker and not a thinner spacer is required. Thus wear on the spacer or the bearing faces results in pre-load developing , which I think is what gives people the idea that pre-load is acceptable

In relation to Bill's point , rather different considerations apply to front wheel bearings, where end float of 3 to 5 thou is prescribed using a dial gauge , but ( and this in my view is very significant ) if a dial gauge is not available then the age old method of tightening until " sticky" and then going back 1 to 2 flats of the hub nut is prescribed. Thus unlike some later designs, the provision of end float is expressly required no matter how the bearing is set up, and as Bill states, pre-load will probably result in damage to the stub axle

P.S. having watched the Timken video, I see that they stipulate the "sweet spot" as 1 to 5 thou of end float even today. Although I have some vague recollection of 1 or 2 designs in the 1980s requiring pre-load ( IIRC Volvo 244 was one ) it seems that with Timken at any rate, end float is still needed

[angelw]

Christopher Wrote:

Steve : as far as I am aware, the workshop manuals specified 2 to 6 thou end float for the rear hub bearings throughout the entire 25 year life of Bob Knight's rear suspension design , and Jaguar are hardly likely to have gone to the trouble of providing 15 different sizes of spacer if this were merely an academic or optional setting.

Hello Christopher
By the time the S3 was introduced the End Float specified in the Workshop Manual for the Rear Hub Bearings was 0.025 to 0.076mm (0.001" to 0.003") a mean of 0.002, half that of what it used to be. Exactly the same construction in the area of the bearings.

The number of spacers provided is rather irrelevant . The increment between sizes is 0.003" with a spread of 0.042" from thinnest to thickest was to accommodate, I suspect, the accumulation of tolerances in the various parts of the assembly. If the measured End Play is 0.004" (0.001" outside the specified setting tolerance), using a 0.003" thinner spacer will land you 0.001" inside at the other side of the tolerance range.

Preaload of the Differential Pinion Bearing, measured by determining the torque required to rotate the Pinion Shaft Assembly alone (without the Crown Wheel in place) has climbed from circa 8 lb/in early cars to circa 26 lb/in later cars. No change in the Bearing arrangement, but evolution in setting up bearings exists.

With the regards to the age old method of adjusting your 0.002" to 0.006" End Float of a Tapered Roller Bearing setup, by "backing off" the Hub Nut from snug by one to two flats, is totally flawed. The Lead of the Thread of an E Type Front Stub Axle is 0.056" (18 TPI), therefore, one to two flats of the Hub Nut equates to 0.0093" to 0.0186" Axial movement. Sticky, would equate to circa 0.001" preload at most. Accordingly, you would end up with between 0.0083" to 0.0176" End Float at best; well outside the specified max End Float. If you want to see the life expectancy of Tapered Roller Bearings tumble, run them with that type of clearance. Running Tapered Roller Bearings slightly loose compared to slightly tight, results in a greater life expectancy drop off. However, running this bearing setup, where the fit of the bearing with it's journal is a running fit, with too little End Float, or with Preload, will almost certainly result in the bearing spinning on the shaft. This doesn't mean that running Tapered Roller Bearing in Preload is wrong, its a case of the design fit of the bearing to the shaft is technically wrong.

The Bearing set up of Machine Tools, CNC machines in particular, have used Tapered Roller Bearing and Angular Contact Bearing in quite heavy Preload, for their Main Spindles since day dot. The revs of car Wheel bearings is pedestrian in comparison. The revs of a 205 x 70 tire on a 15" rim, at 100kph, is circa 15 times slower than the revs that can be continuously maintained by a reasonable CNC Machine Main Spindle. In their life, they will cover more territory than an E Type in many lifetimes. And its not a case of the Machine Tool Spindle Bearings having elaborate lubrication systems for their bearings. Some of the very quick spindles do, but the majority rely on lubrication that equates to 10% of the available air space in the bearing, being applied at assembly of the spindle.

Regards,

Bill

[christopher storey]

Bill : your arithmetic is fine. All I am saying is that the Jaguar specifications are there for all to see, ( together with their recommendations for setting up front bearings without a dial gauge ) and it seems to me to be folly to ignore them. This is particularly so when in effect they are reiterated in the Timken video which appears a few posts back . However, I suppose the counter argument is " what did Bob Knight know about mechanical engineering ".........

[angelw]

christopher storey Wrote:

Bill : your arithmetic is fine. All I am saying is that the Jaguar specifications are there for all to see, ( together with their recommendations for setting up front bearings without a dial gauge )

Well, based on the specified procedure of setting the End Float without a dial indicator, it would seem that the Engineer that thought that up, had no regard for a design setting.

In early manuals, it states the correct End Float of 0.004" +-0.001" (0.003 - 0.005). Without the aid of a dial indicator, fully tighten the axle nut (that's subjective right there), then slacken by one, or two flats. Depending on just how tight the nut is, you might get lucky and fall within the tolerance range with one flat, but for a second incremental, axial move that equates to 4.5 times the total tolerance to be acceptable, is a bit of a nonsense. Just because its there to see, doesn't make it right and clearly that instruction is flawed, no matter what spin you like to put on it.

[PeterCrespin]

“..throughout the entire 25 year life of Bob Knight's rear suspension design...”

Thirty five years Christopher. You’ll recall that apart from switching to outboard discs near the end, , the system remained in use until the end of XJS production in 1996.

I realise that wasn’t the main point of your post, of course.

The front hubs of later cars used a pressed 12 point nut cover that extended to form castellations. Because the slip-over cover had twelve points not six, and because the castellations were slightly offset, it was possible to adjust the nut by much less than one flat and hold it there with a split pin. I’m not sure about XJ40s but X300s definitely used this system and I expect the part is very cheap and could be used on E-types. After ‘96 they switched to deep twin-groove cartridge ball bearings and easy DIY adjustment became unnecessary during the service life of the bearing. Which is just as well because it also became impossible. The cartridge bearing and stiffer press-fit ball joints were amongst the changes made to tighten up the front end and avoid ‘tramlining’ on wide low profile tyres when Jaguar introduced their 18” and upwards wheel choices. Being able to tighten the bearing solidly onto the stub is a boon but an amateur can’t undo with without a fancy special tool.

[angelw]

Peter Wrote:

The front hubs of later cars used a pressed 12 point nut cover that extended to form castellations.

Hello Pete,
I had forgotten about that change. Also through the production of the XJS, different Stub Axles were introduced that were a lot heavier, with lager Bearing Journals and Thread. The OD of the Bearings are the same as that used on earlier cars, so the internal dimensions of the Hubs remained the same, but smaller diameter rollers were employed to facilitate the larger Bearing Journals of the Stub Axle. I know that the 12 point cover was used with the larger nut of the larger Stub Axle, but do you know if that system was used on any earlier cars equipped with the earlier Axles?

The 12 point index makes sense, where an instruction to move the nut one or two flats makes none.

A method of turning a feature to a size within a tolerance (on a lathe), say a bore, when only a plug of the minor size is available (no internal Micrometers, or Telescopic Gauges), is to rough close to size by comparison with the plug, then all subsequent Depth of Cuts (in terms of diameter) are made using the tolerance value. Eventually you will get to the point where the plug will not quite entering the bore, then after taking the next cut using the tolerance value, it does. Lets say that the last time the plug couldn't enter the hole it was, say 0.001" under the plug size (Minimum Bore Size). By taking another cut using Tolerance, the hole size will be 0.001" inside the maximum hole size and therefore, in tolerance.

The same method applied when adjusting wheel bearings when a dial indicator is not available. The specified size range of End Float for later cars is 0.002" to 0.006", a tolerance of 0.004". By moving the axle nut axially by the tolerance (0.004") from Zero (no End Float), will result in an End Float within the specified Tolerance Range.

Regards,

Bill