[binfordw]

Hi,

I'd like to hear some thoughts on how to correctly set preload of a spindle. I figured it was a matter of testrunning and monitoring for heat issues. Too much heat= too much preload?

Is there more to it?

[RICHARD ZASTROW]

As usual, depends. Preload can be built in by the manufacturer. If that's what you have, the preload is set by clamping the bearings together.

Preload can be set by grinding a bit from either spacers or inner/outer bearing ring faces. NOT A JOB FOR THE INEXPERIENCED. You will be dealing with microns of material removal, absolute parallelism and perpendicularity of bores to faces of bearings/spacers.

Better to pay the piper and order your bearings with the proper preload built in.

Dick Z

[binfordw]

I'm speaking of a spindle setup that uses tapered bearings. Spanner nuts on end of shaft are used to set preload of bearings. Im curious what is tight enough- and what is too tight?

The way they are currently being rebuild inhouse- the spindle shafts are so tight you can barely turn them by hand. The bearings are failing in short fashion ( we have roughly 50 of these spindles in use) It was previously thought to be a non-greasing issue, but after rebuilding a small mill spindle at home and having issues with heat and correct preload, I was thinking the problem is with high heat from tight preload.

[packrat]

Timken has engineering literature, here - http://catalog.timken.com/WebProject...Code=trb12flx# Different settings for different uses.

[mactec54]

binfordw

So the part you are missing is a spacer between the 2 Bearings, you machine this spacer (tube) to give you the right preload on the bearings, you then can not over do the preload unless the spacer is to short

[dmalicky]

Agreed with Dick, packrat, and Mactec54. All bearings but especially tapered rollers are very sensitive to the preload distance / spacer length. Timken's double-row bearings (TDI, TDO, and older automotive SET series) or Spacer Assemblies make this easy, as the preload distance is built-in to the bearing/spacer. These also need a predefined press fit on the shaft. If the spindle is using 2 single row bearings, I'd look into ordering a spacer assembly.

Automotive wheel bearings are often torqued to around 150 ft-lb, but that simply compresses the inner rings together to their set distance. For a non-spaced assembly, a torqued nut isn't a good way to set preload -- thread friction is not consistent, so the preload is all over the place. Bearings can fail by heat at higher speed or excess contact loads at lower speed (and other ways).

Depending on the application, if a spacer assembly isn't feasible, another option might be a Belleville washer on 1 end of the spindle. As long as the washer isn't bottomed out, the load is reasonably correlated to the compression distance. But if the spindle has both push and pull loads, neither one can be more than the Belleville compression load or the spindle will have 'slop'.

[packrat]

Wow, 150ft/lb on wheel bearings! Think you mean 150In/lbs, then you back-off nut one flat. End play of .005 -.008 is what you shoot for.

[dmalicky]

Oh, there are a number of different designs and methods--light trucks and some cars often use a method similar to the one you describe (or with cotter pin). For most recent cars, common practice is indeed 150+ foot-pounds (no cotter pin or tabbed lock washer). I think the inner races are ground to give slight preload, as I've not detected any slop in the ones I've worked on.
http://www.vsm.skf.com/~/media/Files...ve/457377.ashx
Porsche 911 is 340 ft-lbs -- now that's a breaker bar!

[mactec54]

packrat
Wow, 150ft/lb on wheel bearings

This is a normal number, for auto front hubs,there are many that are much higher than this, older types & trucks, trailers Etc are similar to what you have described

But none of this has anything to do with what binfordw, wanted with his spindles
Post #1 I'd like to hear some thoughts on how to correctly set preload of a spindle. I figured it was a matter of testrunning and monitoring for heat issues. Too much heat= too much preload?

The only way to fix his problem is what I said in post #5

[packrat]

packrat
Wow, 150ft/lb on wheel bearings

This is a normal number, for auto front hubs,there are many that are much higher than this, older types & trucks, trailers Etc are similar to what you have described

But none of this has anything to do with what binfordw, wanted with his spindles
Post #1 I'd like to hear some thoughts on how to correctly set preload of a spindle. I figured it was a matter of testrunning and monitoring for heat issues. Too much heat= too much preload?

The only way to fix his problem is what I said in post #5

Or my post #4 going to Timken Engineering manuals for the type bearing arrangement the OP has. Which we don't know, other than tapered roller bearings at each end. Not all bearing set ups use spacers.'

The hub bearings you guys were talking about don't have an adjustment, wasn't even thinking of that type when I answered. They come from the factory with the .005-.008 end play built in. With out it they would overheat and die, fast.

[mactec54]

packrat
They come from the factory with the .005-.008 end play built in. With out it they would overheat and die, fast.

They may have .005 to .008 before assembly , but once locked up, assembled correctly, they have zero end play, so Not correct at all

Any machine spindles that are of any use at all, will have spacers to control in this case the timken bearing preload, no matter what type of bearing you use in a machine spindle, precision spacers are needed to have a stable spindle

[binfordw]

I appreciate the help guys.

The spindle setup (these are used on a chamfer-lathe machine) Uses a tapered, single row bearing, a radial bearing behind that, and another single row tapered bearing at the opposite end of shaft.

The drawings do not show a spacer/sleeve being used. However, there seems to be a couple different variations, I did find one that has a spacer. Why they do not all have spacers- Im not sure.

[dmalicky]

If it fits or can be made to fit, I'd suggest replacing the 2 nose bearings with a TDI or TDO pair (or possibly an automotive SET cartridge with built-in seals) -- then the preload is factory set, as Dick Z suggested. Or a pair of duplex ground ACs ($). The tail end usually gets a radial floated in the housing, unless there are heavy loads on that end, too, in which case a needle roller or another TDI/TDO (clamped to shaft but floated in the housing) is appropriate.

Slight preload gives the longest life for tapered rollers: http://www.pirate4x4.com/tech/billav...20pre-load.pdf
SKF shows optimal preload at around 0.05*C: http://www.skf.com/files/005031.pdf

Tapered rollers are very stiff and so very sensitive to preload distance, so it's very hard to grind a spacer just right, as Dick Z said. Usually, spacers have to be custom fit to each housing -- more cost and more to go wrong. If the spacer is slightly too short, the bearing overloads and fails. Too long, slop. Even if perfect, as the shaft heats up, the preload is thrown off. That's why most designs use TDI or TDO -- it keeps the highest precision requirements inside the bearing assembly, makes the preload ~insensitive to shaft temp, is extremely stiff in all directions, is easiest to make, easiest to service, most reliable, and least cost. 7 wins!

[RICHARD ZASTROW]

Precision tapered roller bearings for spindles are $$$$$$$$$$$ also. However, they can usually handle heavier loads than A/C bearings. A/c bearings can handle higher speeds.

I usually use sealed A/C spindle bearings. If I need more "heft", I use multiple pairs of bearings, sometimes with spacers. The spacers add to stiffness and spread the heat generated by the bearings to more area of the housing.

FWIW, latest one has two pairs of sealed A/C bearings separated by a roller bearing (matched bore and O.D.sizes). These are captured on the nose end and a floating radial ball bearing on the opposite end.

Dick Z