7 replies to this topic

[mini-mad-mark]

Hi All,

Apologies firstly for starting a new topic - reason is that my original post seems to have been removed - maybe not a bad thing as it could mislead so apologies for that also. 

 

My comment was that wheel bearings should be run with endplay not preload.

- this was based on my general experience with manually adjusted bearings where they are run with very slight endplay as it is not possible to gauge/measure the amount of preload you might impart with the hub nut; here we are NOT talking mini, we are talking bearings with no spacer between races where the adjustment IS meant to be made using the hub nut.

 

You can measure/perceive a small amount of endplay in a bearing - and it is safer to run a bearing with a little (known) endplay rather than too much (unknown) preload as excessive preload can cause bearing failure very quickly , which is why this endplay setting  is recommended for manual adjustment

 

N.B. As mentioned correctly many times you cannot "adjust" the bearing in a mini hub by playing with the CV nut torque as this just tightens across several flat surface interfaces but as also mentioned many times it should be tightened correctly (flat washer first and correct torque for CV joint type) to prevent other grief with CV and/or bearings

 

For the mini, and now I suppose many bearings/cars, the adjustment is made by the bearing manufacturer by selection/adjustment of components - I assumed (incorrectly) that the setting they applied would be the "safe" endplay setting. This is not correct for the reasons below.

 

Apologies to MokeSpider - he pointed out correctly (which I didn't disagree with!) that a bearing will give optimum performance and life when used in a slight preload situation (see attached file if it works) - which is why manufacturers DO run bearings slightly preloaded - then, as MokeSpider  pointed out, the mini does run preloaded bearings, which I was incorrect to question as I now find this to be correct after looking at the various tolerances.

 

Based on what I found out the bearing is given a bench setting of 0.003-0.005" endplay (or clearance) 

(from the picture in the previous topic you will see it is marked 0.004")

 

The bearing is 2.4404-2.4414" outside diameter and the hub should be  be sized at 2.4424-2.4434" based on the normal interference fit range of 0.003" to 0.001" tight  (tightest fit = 2.4434-2.4404 = 0.003 and loosest fit 2.4424-2.4414 = 0.001)

 

When the outers are tight fitted in the hub this reduces the bench end play and (by calculation) will give a mounted setting of 0.006" preload (or tight) to 0.002" endplay (clearance)

 

The exact figure in that range is dependent on where in the tolerance ranges the exact sizes of the hub bores, the outside diameters of the outer races and the actual adjusted bench end play are.

Note - this does not include any variation in hub shoulder width (which I have not been able to find out definitely but is likely to add 0.001" extra endplay assuming a 0.001" width tolerance )

 

So this does mean that bearings will predominantly be in a preload situation BUT if you have the maximum bench end play with the smallest interference fit (so smallest bearing and largest hub bore) and smallest hub shoulder it is possible that you will end up in the small mounted endplay range; or more endplay if the hub bore/shoulder is worn (reduces the interference fit so reducing the change from bench end play to mounted setting 

 

The cheap bearing manufacturers have less control over all these tolerances so as people have found some work and some don't and even if these tolerances are correct, then the actual rolling contact surfaces have to be half decent as well, and again some are and some aren't. 

 

It does pay to buy a reputable brand as 99.73% or more (!) will work perfectly - the cheap ones will work maybe, but you may or may not be lucky......

 

I should say here that I rebuilt some hubs recently and re-fitted the (reputable brand) bearings after inspecting them carefully - neither side had any discernible play (in preload I guess) so I kind of ignored my own (incorrect) advice but assumed the play was so small that I couldn't feel it .......

 

The attached file (if it works!) shows the graph of endplay versus preload and you can see that if you go too far into preload you will lose bearing life very quickly whereas a little bit too much endplay will reduce the life but in a "safer" way - it also talks about manual setting of non-spacered bearings and that they should be "primarily end play"

 

 Bearing setting graph.pdf   492.97K   19 downloads

 

 

Regards, Mark 

 

[nicklouse]

It is still there http://www.theminifo...wheel-bearings/

[Spider]

Hey Mini Mad Mark!

 

Brilliant posting and additional info.

 

I agree with what you have said re: Pre-load vs End Play. If one is not sure which way to go, I tend to think you're better off with a small amount of End Play than too much Pre-load. As the Timken Graph shows, while the correct Pre-load is helpful, too much is a disaster! And there can be a fine line on that side, while there's a much broader safe area in the End Play area.

 

Supposedly one of the 'beauties' of the bearings we are supplied with for the Mini is they take little skill to install. In the case of Drum Brakes, I would say this is true. In the case of Disc Brakes, some more skill is needed I feel.

 

Also, and I did skip over it in my other post, but what you've bought up in regards to the interference fit between the Bearing Cup and the Hub is very important and bang on. That is the root cause of 90% of Wheel Bearing problems. I didn't actually go there in my post as it requires some odd and expensive measuring gear to be able to measure properly, which I doubt most garages would have and even less likely, the guy at home. I must say, I don't even have it, but I have ways around it!

 

What I have found is that once that interference has been lost, the Bearing Cup spins in the Hub and the hub being quite soft, wears very quick, even with half a turn, given the loading on them, they wear by a measurable amount.

 

The post I put up was one fairly way I could think of for the guy at home could do to check if the Hub is usable. Sure and I did neglect to add, that if the Bearing Cup when driven or pressed in suddenly goes easy or even 'drops in' then it's all over right there.

 

I'll go back and update it.

 

Cheers

Chris

 

PS, I did a Rear Wheel Bearing assembly following this front one. As an 'experiment', before Greasing the Bearings, I was able to sit them in the hub and easily get 0,004" feelers between them, in fact, I'd say if I could bend without damage, i would have gotten a 0,010" strip in there.

 

I did take some photos, I'll get them up later.

[mini-mad-mark]

Cheers Chris - my mistake was assuming that all bearings are run in the "safe" endplay zone when they aren't.

 

I broke the cardinal rule of the forum - don't comment on what you don't know - I should have done my (home)work before commenting!

 

Agree with the hub wear issue - if you haven't got interference fit of the bearing outers, bin the hubs as you will never get them to work satisfactorily. (Assuming the bearings are not defective which might be the case although that should be an easy dimension/tolerance to achieve, even for the cheapos)

 

Also agree most people will not have skills/access to suitably accurate measuring equipment - fortunately I do, as mentioned I work for one of the major manufacturers (which is why I felt qualified to discuss the issue, but it showed up my lack of knowledge/research)

 

by the way I also looked at the tapered rear wheel bearings (so I didn't screw up again!) - the original spacered assembly is designed to have 0.0045" preload to 0.004 endplay mounted setting, but the newer version with extended inner races is a SetRight assembly

 

- if you are familiar with the Timken website you will see the principle of how SetRight works on there ( type "preload vs endplay" into the search bar and it will bring up the Timken publication for setting techniques - for those that are not familiar)

 

SetRight will be aiming for the same sort of range of mounted setting but based on statistical principles rather than an individual adjustment of each bearing.

 

 

Regards, Mark

[Spider]

Cheers Mark

 

I should also lay some of my cards on the table.

 

I know next to nothing about cars - True. I know a little about Minis and Mokes, but not much about cars.

 

Also, the Front Wheel Bearings on our Minis are actually a Metric Assembly. the D of them is 62.00 mm and the d is 31.75 mm (yes, I know, but that does come up in the Metric Listings).

 

I did look in a few workshop manuals I have here on other cars (not that I understand them);-

 

1980 Honda Civic, rear Wheel bearings is says to torque the nut to 10 metrics, loosen, then retorque to 3 metrics, tighten further to next split pin hole.

1976 Isuzi Gemini - Front Wheels - Torque to 30 metrics, loosen off, then hand tighten only, then to next split pin hole.

VN Holden - Front Wheels - Tighten to 18 Nms, rotate 3 times, loosen off, retorque to 13 Nms, back off to next split pin hole

Triumph MkI and II - tighten to 5 ft/lbs, while turning the hub, then back off to next split pin hole

 

All these manuals, except for the Triumph one, are factory manuals by the way. But, these few seem to point to Preload, though I don't doubt there are some out there that have end float. I'll have a look in a couple of truck ones I also have.

 

Reading what Timken said in their manual they do say sometimes end float is set, however reading on they say that it's usually to compensate for temperatures and when at operating temp, there's usually no end float.

 

Going back to the Talk from the Timken Tech guy that we had, I do recall him saying that Timken's advice to all manufacturers for wheel hub bearings is to set with Pre-load. He did go on to say, that what the car manufacturer actually did with the bearing assembly after receiving their advice was sometimes a different matter as sometimes the Manufacturer's Engineer(s) coff coff knew better coff.

 

<EDIT: Mention 'End Float' to the Brake Manufacturers, then run as they'll be chucking big hammers at ya! >

Edited by Moke Spider, 23 July 2016 - 12:56 AM.

[nicklouse]

Here is the service info for my trailer brakes.

[sonikk4]

Love all the info here, just to put things into perspective with other bearings, when we change a wheel on a B747 its initially torqued to a max of 525 lbs ft whilst spinning the wheel, then backed off then retorqued to 150-250 lbs ft. Now within this range we have two anti spin bolts to fit just for good measure.

Next time my lads do a wheel change I will post up a picture of the bearings we have fitted. And considering the loads these have to deal with they cope extremely well. In 36 years not seen one go yet.

[tiger99]

This is all very good information, and about time too! The bearing problems have been occurring for quite some time now. Well done, all!

 

As for the Triumph bearings, they almost always end up with small but discernible end float when nipped up and backed off to the next split pin hole. Some people try go get as close to zero end float as possible by gently abrading the back of the nut on a carpenters oil stone, or sheet of abrasive paper on a very flat surface, but it was never necessary to be as exact as that on any of my various cars with adjustable bearings.

 

What seems to be well established now is that a very small preload or very small running clearance are both safe, and that is what matters. Many people will not be able to measure, but you can feel just a hint of clearance or preload as you turn the wheel. The past threads where people were having their rear wheels practically seize up as they torqued the nut were of course an unsafe condition, due to something being wrong with the assembly. I recall that there may have been a problem with some cowboy who was packaging bearings and spacers as sets without doing accurate measurements.

 

Timken are saying that very light preload is better and I am going to believe them. Other cars may need bearings from a different manufacturer, but the same considerations ought to apply.

 

We still need a good way of assessing hub castings, CV shafts and drive flanges (drum brake) etc for dimensional accuracy. Most of us can't measure these things at home. A professional hub reconditioning service (as is done for radius arms) might be the way forward.