bearing preload question

I will secondDavid.s concern. A tepered roller MUST have clearance when cold so it never gets tighter than zero clearance when hot. Check the procedure for setting front bearings on an '81-83 Tercel, as they used the hard spacer to set the clearance on their double tapered rollers.

As others have said, make sure the bearings won't get too tight when hot. On my american made car, 25 years ago, the bearings were tightened to a slight pre-load. When running, the heated spindle would expand lengthwise and the pre-load would lessen. Not paying attention, I pre-loaded the bearings on a VW bug front end. The book says not to do this, but I didn't read it. 15 miles later, 15 miles from home, down the freeway, they started to howl. Pulling over and getting out to check showed smoking hubs. The upshot was having to cut the races off the spindles. Check to see if heat will make your bearings loose or tighter. I went to a bearing web site, SKF I think, and was able to get pre-load data for rpm, axial, and radial load for any bearing I chose from their list. It takes a little time because of all the choices but you will be able to find out exactly how to run your bearings. ERS

Thanks for the input, guys. I am using the term pre-load, probably incorrectly, to mean the actual installed clearance of the bearing, positive or negative. What I am thinking, based on what others have done to other, similar, spindle setups, is a hard spacer to set the clearance. This will have the effect of creating a solid stack consisting of the inner cone, the spacer and the outer cone, fasted with about 50 Ft/lbs of torque on the fastening nut. This stack will have the effect of providing additional stiffness to the existing stub axle. Here's the deal - the triumph Spitfire OEM tire was a hard bias ply tire with a tread patch 4 inches wide. This is using soft slick tires 9.5 inches wide - big difference in loading, and the effect if pad knock-off from spindle flex. The car is from 1972 and it raced for 10 years like this, so this is a tweak, not a drastic fix.

Here's another question. The aluminium hub is going to heat up and expand, increasing the distance between the races of the two tapered roller bearings. The distance between the inner cones is going to be set by the combination of the expansion of the steel stub axle and the steel spacer. What are the exansion factors for steel and aluminium? Additionally, the hub has the brake disc fixed to it, so I can expect it to heat up more by some degree than the stub axle. Does any one think this can be calculated or is it a try it and see thing?

Thanks again, Brian

Thanks, Brian

manufacturer,

Brian,

I'm not trying to be an a**hole here, but I'm not sure I see the point. If I'm reading what you want to do correctly, you'll make more stable the relationship between the two outer races of the bearings in a hub, but the relationship between the bearing and the races will still need to be the same as it is currently. It's not going to change how much or little play you have in the hub/bearing/axle interaction once things get warmed up.

I guess I see the advantage of the setup you envision as perhaps making the installation of the hubs on the spindle a no-brainer; no futzing with the "feel", just torque it and go. And there's less chance of a nut backing off. MGBs used to have a similar arrangement and the rear end of stock cars often use this setup, but my understanding is that this is/was for convenience of installation and servicing, not feel or performance.

I'm assuming this is some sort of vintage FF or sports racer? I've driven plenty of these that felt tight, and I'm betting if the front end feels loose that the fault lies somewhere else in the suspension or steering. If this is you looking for the "unfair advantage", I completely understand. ;) I'm betting it won't make any difference, but let us know what happens if you do it!

Regards,

Peter

I think the idea is to make the inner race placement a bit more determined. If I understand it correctly, right now the outers are pressed into the hub, and the inners ride on the axle, with the cottered nut pressing them together. Because the inners are still a sliding fit on the axle, they are free to move a bit.

The setup he is envisioning is standard practice on many motorcycle wheels.

Outer races are shrink fits in the bore of the hub, with a fixed spacer that determines the separation between them. The spacer is a drop-in fit into the bore of the hub.

Then the inner races are spaced apart by another fixed spacer, and the entire stack is clamped up by a single nut on the end of the axle.

The preload (clearance, actually - some tiny amount) is determined by varying the relative lengths of the spacers. Typically a new bearing will seat in, and the clearance will increase somewhat, so the inner spacer has to be thinned a few thou to bring the inner races closer together and reduce the play.

This setup is used on motorbikes because one cannot have a spindle end hanging out in free air - both ends are captive either in the fork sliders or teh swingarms.

Jim

================================================== please reply to: JRR(zero) at yktvmv (dot) vnet (dot) ibm (dot) com ==================================================

Oddly enough, everything that matters in a setup like this is steel. The preload is determined by the relative lengths of the inner vs outer spacer - and becaue they are both steel they will move at the same rate. Properly designed it will be independent of temperature to first order.

Jim

================================================== please reply to: JRR(zero) at yktvmv (dot) vnet (dot) ibm (dot) com ==================================================

You need *two* spacers. An inner one that separates the inner races, and a larger one that sets the distance between the bearing outer races.

The inner races and spacer are constrained by a step in the axle (spindle) at one end, and by the nut at the other which provides the clamping force.

The outer races have to likewise be constrained in a bore in the hub, typically with a step at the bottom of one end to locate one race, then the spacer, then the other race, and then some kind of clamping setup to bring them all in contact.

Jim

================================================== please reply to: JRR(zero) at yktvmv (dot) vnet (dot) ibm (dot) com ==================================================

Jim,

Even though I called the inner races the "outer races" in my post (thinking in relation to the hub - not the bearings... brain fade), I do understand what he wants to do. I just don't think it's going to be of much benefit.

I haven't pulled a modern (read as made in the last fifteen years) motorcycle hub apart, but all the bikes I've owned had "regular" bearings in the hubs. Are tapered bearings being used now?

Peter

It is confusing, the 'inner races' are on the 'outside' of the stack of course. :^)

All of mine do, though they are all of the same brand (bmw). But they've been using the same setup since the early 50s, and possibly before that as well.

Tapered roller bearings, properly set up, will withstand tremendous loads and abuse.

Jim

================================================== please reply to: JRR(zero) at yktvmv (dot) vnet (dot) ibm (dot) com ==================================================

On Tue, 20 Apr 2004 15:50:59 -0400, "Brian" brought forth from the murky depths:

Like the others said, be careful. American cars (most Fords, my vehicle of choice) which ran with a slight preload had this method: Clean, dry, and regrease the bearings, reinstall the seal, mount the hub, torque the nut to 25ft/lbs (to seat the parts firmly) while spinning the hub the direction it will turn, back off to zero preload, retorque to 10 in/lbs. Install castellated nut, cotter pin, and cap. My '90 manual says 25ft/lbs, then back off 1/8 turn. Chebbies usually went with zero preload after the same pretorque and release.

If it were my vehicle, I'd try the Ford 10in/lb route, then check them when the vehicle has been run for awhile. I'd adjust as necessary for preload when they're within the temp range at which they will be run.

------------------------------------------- Stain and Poly are their own punishment

Comprehensive Website Design ======================================================

I understand. I think that the flex in the unit isn't going to happen in the stub axle, but in the upright, especially under braking. Just my take... I'll be interested to hear if it makes a difference in how the car feels.

Peter

Some manufacturers specify a 'breakway' torque setting for wheel bearings, a spring scale on a string taped to the tire surface. The wheel is supposed to start to turn at some particular force on the spring scale.

Jim

================================================== please reply to: JRR(zero) at yktvmv (dot) vnet (dot) ibm (dot) com ==================================================

Hi David, This is exactly what I am doing this to achieve. If there is no effect, then probably the upright is the next thing to lok at. Adding a spacer is a whole lot easier than fab-ing a new upright, and in the sourse of researching this, it appears that this has fixed a spindle few problem on this type of axle before, so I have hope...

Brian

Hi Jim. Your idea is pretty much exactly my thinking, but in this hub design there are two register shoulders that the races are pressed against in the bore of the hub, so the first distance is set by the hub machining not a spacer for the races. I can see where a spacer would be a help, particularly in a machining setup where getting repeatability to within a thou on the register location over many hubs would be difficult. But if the hub was aluminium and the spacer steel, the bearing race would be moved should the hub heat up - which it surely will. I would make the spacer match the material of the hub or the spindle, whichever it had to remain the same length as, to make sure that the bearings dont' lose their press-fit and spin in their bores.

Brian

Brian,

How has this problem manifested itself in the car's handling?

understand.