In a recent thread regarding bearings, someone mentioned the old Timken Engineering manual, indicating it to be a useful and interesting reference. For 4 bucks, plus 2 postage, I confirmed that the version was all that. By the way, the Timken website currently has a lot of the same info.
A recent article in one of the home shop mags showed plans for building a high speed spindle for a milling machine, using a dremel tool motor. The design has the dremel offset from the mill spindle and fastened to a 1/2 inch plate which is mounted on a holder that fastens into the mill spindle. The holder in this design is bored out and carries a high-speed spindle which rotates on two separate ball bearings. One of the design limitations mentioned is the play between the two ball bearings is taken up by some Belleville washers (I think the author actually used a wave washer). Anyway, the load capability of the unit was specified as "limited" because of the flex in the washers.
So I thought, "That would be a cool application for a tapered roller bearing". With preload, I thought, play could be eliminated, and the tapered roller would handle a heck of a lot of thrust, removing the limitation of the current design. My little Atlas lathe has Timken tapered roller bearings, so why not? Heck, taking this a step further, if I wanted a project I could retrofit the double-row ball bearings on my drill press with tapered roller bearings! The world was my oyster: I would improve a couple of machines, fix the economy, and repair all that was wrong in the world with tapered roller bearings.
Then I read that mills and such typically use angular contact bearings. These seem much more expensive, and seem to me to be less robust that the tapered roller. So, aside from my rather ambitious plans to install tapered rollers in everything from my desk chair to the hard drive on my laptop, would someone be kind enough to explain the benefits of angular contact versus tapered roller bearings in applications like DPs and the aforementioned high-speed spindle?
I did look around on the web for such information, and did not find anything. If there is a a site I missed, a URL would be appreciated.
Thanks,
Jim
Here's the notions I've been carrying around forever.
Angular contact Vs taper roller bearingss other things being equal (envelope, lube preload etc)
Angular contact are characterized as capable of much higher RPM
Taper roller have mote thrust radial load cap and are stiffer that anguarr contact
Taper roller are a bit more expensive than anguar contact
Prices triple or more as you go up per precision grade
Machine tool spindles absolutely positively have to hac precision grade bearing grade 5 and above for angular contact and comparable for taper roller.
High speed spindles require high precision bearings grade 7 and above.
Angular contact and taper roller bearings require housings/spindles reflection their grades of precision.
Designs having the bearings separated by some distance should arrange the bearing direction so that spindle expansion from heating reduces the preload.
Designs featuring bearings rigidly paired near the business require a "float" bearing at the other.
Lubrication should reflect the speed and service of the spindle. Some modern greases are suitable for high spindle speeds 20,000 RPM for 60 mm bore for example. Open cycle oil mist has long been the standard for very high spindle spees (100,000+ RPM)
I can't imagine using a Dremel motor for anything thats actually precision. They are crudely made and unless their bearings are swapped out for presision grade (usually requires a complete redesign of the tool) the shaft tracking is erratic. They are poorly balanced, and the collet is laughably crude. Compare the Dremel to the Precise line of electrichigh speed grinders in the same size class. This is not a sneer at those who love ther Dremel tooling. I have two or three and I use them frequently. I'm merely putting things in perspective. A Dremel is not suited for application to most anything where the word "precision" is used as a descriptor.
Good questions and a great read Jim, I'm waiting for further illumination on this, it has come to me as well.
I'm suspicious that some of the preference is tradition, making the assumption that angular contact bearings are inherently more precise, though I don't quite grasp the reasoning, willing to learn though.
Clearly, the tapered roller bearing has a greater ability to generate heat at max loading, (though much above ball bearing capacity) more contact surface, though ironically, the first general use for them was to eliminate "hot boxes" on rail cars.
It seems to me that a well adjusted pair of tapered roller bearings could be sized for high speed application and not produce objectionable heat, lube tailoring to aid, they most certainly would carry a much greater load, size per size and as you suggest, readily take adjustment.
I hope that some of the answers rise above classical reiterations of the qualities of ABEC7 and ABEC9 spindle bearings in a vacuum and delves deeper into a rational comparison of what tapered rollers can and could do, with similar massaging that drives the ABEC7 and ABEC9 bearings into pound per pound platinum pricing.
Are tapered rollers just klunky and no way around it? Seems improbable to me, got really cool modern CNC grinding machines and even better modern methods of measurement/sorting/matching.
I am able to grasp that very high precision angular contact ball bearings should excel at very high speed/light load but your drill press?
Bob
Edit;,
"Taper roller are a bit more expensive than anguar contact
Prices triple or more as you go up per precision grade", didn't know that, thanks Forrest.
What Forrest said, except I thought tapered rollers were cheaper.
I used to work for a company that manufactured turbines with tapered roller bearings and the biggest limitation we ran into was RPM's. A competitor making a higher RPM machine used angular contact ball bearings. The angular contact bearings on the competitors machines seemed to last longer than ours, but there were probably lots of other contributing factors other than just bearing type.
One of the issues I recall was that the tapered rollers are pretty good oil pumps and at high RPM's (over the design rating) would pump the lube right outa the bearings. In a spray oil lube system it was important to spray the oil in at the small end or else it would pump right back out.
I like tapered rollers and as far as I know the only major limitation for machine tools would be the specific bearing RPM limit. Another issue was getting the preload correct. On the turbines it had to be set cold so that it was at the right preload when the shaft and housing were at operating temp. They had different temps and it took some trial and error (read burned up bearings) to get it right.
"On the turbines it had to be set cold so that it was at the right preload when the shaft and housing were at operating temp. They had different temps and it took some trial and error (read burned up bearings) to get it right. "
Yes, if find the applications of tapered roller bearings somewhat odd.
The old wisconsin vf-4 v-4 air cooled engine used timkens on the
crankshaft, setting the end play (.005 IIRC) cold, had end play, not preload,
using shims under the pto end bearing cap.
Electric motors (large ones I deal with) use a standard ball bearing
on the pto end, and as the motor heats up, the shaft grows in
length, out the opposite end, using straight roller bearings.
Thereby not affecting the pto end of the shaft.
Thanks, guys. Some points:
Forrest, thanks once again for sharing your knowledge and experience - I see a lot of folks benefited. FWIW, the Dremel in this tool drove a small splined pulley which drove a larger splined pulley on the spindle. The coupling was a Gates brand splined rubber belt. So the new spindle rotated on it's own bearings. The author of the article didn't spec the class of ball bearings he was using. The step down was in speed and so there was an uptick in torque. No doubt a jig grinder is better!
BobM3, yeah, the Timken manual mentioned that there must be some thrust load in operation. A lot of times (e.g. in my lathe) that's preload. Sometimes the unit warming up generates preload (adjustment of preload is recommended after a warm-up period).
Robert, What, you haven't retrofitted all the water valves in your house to tapered rollers? Your dog's water dish hasn't been swapped completely to ball bearings? Your doorknobs haven't all been dynamically balanced? Why would you question me using TR bearings in my DP? Actually, the only reason I'd do that is because the original bearings are shot and I have to replace them with something, and if rollers would give me better thrust performance and would last longer, I'd be willing to give it a try. Call me
I'll take a look at the RPM and precision class limits of smaller roller bearings and see if I find something that might match up.
For now, let me thank everyone again for sharing their expertise.
Jim
The "Precision" in a precision bearing derives from the quality of the axis of rotation they constrain the shaft to. Ideally an axis of rotation should be a line. Practically speaking, the errors in the bearing races, the balls, the housing the bearing seats etc produces an error trace than when hugely magnified looks something like a potato. The more pecise the bearing the smaller the potato shape. Off the shelf bearings can have as much as 0." roundness error. When you condsider the roundness error is but a component of the tolerance for a machined bore 0." error can make precision to 0.003" tolerancess almost unacheivable.
The spparable errors sources in a round bors are roundness, taper, rainboe, finish, tool deflecction, tool wear, material inconsistancy, temperature, obervational error, measuring error etc. Not all of these will be apparent but5 you can bet all will bre present in the small decimal places.
The bored hole or the truned diameter is but a reflection of the spindle's axis of rotation.I would hazard the permissible proportion of roundness error in general machine work is 30% of the tolerance. About half of the roundness error is an inseparable artifact of the rolling element bearing that support the spindle generating the bor. 1/2 of 20% of a 0." tolerance is 0.". That's 50 millionths: definitey precision bearing territory.
You may never have to hit a 0. tolerance size unless you have to bore or turn some ball bearing fits. So whenever "spindle bearings" comes up insantly revert to precision thinking.
The ends of the cylindrical rollers rub against the flanges when the bearing is loaded axially. This added friction prevents Timkens from replacing angular contact bearings in high speed applications.
Are you sure about that? I don't have a tapered roller in front of me but I've replaced a LOT of heavily loaded ones and as I recall the ends of the rollers are not even finished machined and don't contact the flanges. Maybe once the bearing has worn out and failed, yes. I thought the high speed limitation is due to the greater surface contact area between the rollers and the races and resultant heat generated by the lubricant as it gets "squished" under the rollers at high speeds. But if I knew everything I'd be rich.
BTW, the PhasePerfect is working great.
Bob
Are you sure about that?
According to the little Timken book, Glacern is correct about the contact. There is a little lip on the cone. Under no load, the rollers just touch that lip. Under thrust load, the rollers are squeezed out and are forced against the lip and the point contact becomes a little oval.
I thought about this and reckon it has to be that way: think of the direction of force perpenticular to the cone and cup faces. Those faces aren't parallel because of the roller taper. If you didn't have the little lip, thrust would shoot your rollers out of the bearing radially. The saving grace of the design is that difference in cup and cone angles is small, and so the difference between forces perpendicular to the cup and cone faces is very small, too. The contact area is also small. Small area, low force, good lubrication. And it obviously works.
But Timken tapered rollers ARE used on lathes, horizontal milling machines, and the like. I think that for high velocity, even the small area and low force rubbing might create enough heat to be problematic and that may be the reason for angular contact bearing use.
My conclusion is that for a high-speed (11,000 rpm) spindle or a drill press that maxes out at rpm, I may do better to just use ball bearings for now, due to the higher speed.
Jim
Robert, What, you haven't retrofitted all the water valves in your house to tapered rollers? Your dog's water dish hasn't been swapped completely to ball bearings? Your doorknobs haven't all been dynamically balanced? Why would you question me using TR bearings in my DP? Actually, the only reason I'd do that is because the original bearings are shot and I have to replace them with something, and if rollers would give me better thrust performance and would last longer, I'd be willing to give it a try. Call me
Jim
Jim, no, no. When I said,
"I am able to grasp that very high precision angular contact ball bearings should excel at very high speed/light load but your drill press?" I left out, "why not?" I was acting as an "enabler", tapered roller bearings in a drill press make eminent sense to me. I should have been clearer.
Oh and yes,
I have retrofitted all the water valves in my house to tapered rollers. My dog's water dish has been swapped completely to ball bearings and my doorknobs have all been dynamically balanced, spin 'em very fast.
Bob
Robert, What, you haven't retrofitted all the water valves in your house to tapered rollers? Your dog's water dish hasn't been swapped completely to ball bearings? Your doorknobs haven't all been dynamically balanced? Why would you question me using TR bearings in my DP? Actually, the only reason I'd do that is because the original bearings are shot and I have to replace them with something, and if rollers would give me better thrust performance and would last longer, I'd be willing to give it a try. Call me
Jim
Jim, no, no. When I said,
"I am able to grasp that very high precision angular contact ball bearings should excel at very high speed/light load but your drill press?" I left out, "why not?" I was acting as an "enabler", tapered roller bearings in a drill press make eminent sense to me. I should have been clearer.
Oh and yes,
I have retrofitted all the water valves in my house to tapered rollers. My dog's water dish has been swapped completely to ball bearings and my doorknobs have all been dynamically balanced, really spin 'em up!
Then there's;
"According to the little Timken book, Glacern is correct about the contact. There is a little lip on the cone. Under no load, the rollers just touch that lip. Under thrust load, the rollers are squeezed out and are forced against the lip and the point contact becomes a little oval. If you didn't have the little lip, thrust would shoot your rollers out of the bearing radially." Now see, that's why I can't stay away from this place, so little known, so much to learn. That one
should have been obvious though. Not always do I ask "why"? when something works as well as a tapered roller bearing, remiss.
Forrest!!
"I hand scraped and half moon flaked my driveway." The very idea, the first exaggerator hasn't a chance.
Bob
you can download the Timken engineering catalog here
http://www.timken.com/en-us/products/Documents/Engineering Section.pdf
Tapered roller bearings are well suited for lathe or mill applications provided they are used within their limits. Bearings have a thermal reference speed that is based on how much heat the bearing generates in operation and for the same envelope dimensions, the angular contact bearing will have a higher speed. The TRB has an advantage that using rollers it has a much higher load capacity over a deep groove ball bearing or angular contact bearing. This results in a much longer life for the same operating conditions, or allows for a smaller bearing to be used if the application permits it. I would not have any concerns with buying a lathe or mill that used tapered roller bearings provided it was not a high speed application.
For your high speeds of the spindles in question, a ball type bearing, either deep groove or angular contact would be a far better choice then the TRB. Keep in mind the DGBB when used as an angular contact bearing (axial and radial loads) has a lower capacity then the same sized ACB due to the difference in contact angle.