[Willbird]

OK now I have read a lot posted by NC Cams on bearings, LEARNED a lot. ABEC 7 AC bearings are the stuff I get it now.

However, lets talk about tapered timken bearings, we touched briefly on them before and the problem with them was that when you get enough preload for them to be stiff they get hard to turn.

As to precision however......if you borrowed two of the tapered timken bearings used in a 1 ton truck rear end (new ones not used) and you set them up with the OEM preload.....for an application like an OD/ID grinder work head that runs at typically fairly low rpm, would they work ?? how "stiff" would they be compared to AC bearings ?? I know they would COST a lot less....but they seem to offer the benefit of handing radial and thrust loads both at he same time.

For the stated application the HP required to rotate them is not going to be an issue...most motors we would find secondhand will be BIGGER than what the job requires.

This is just a mental exerscise and should be treated as such :-)

Bill

[NinerSevenTango]

I use them in a lot of machine designs where thrust is going to be an issue. The drawbacks to them are; yes, they get harder to turn when they are preloaded, and that means heat gets generated, and when heat gets generated, the size of everything changes. Another thing to watch out for is 'run - in'. Basically what that means is you have to run it for awhile, then set the preload again. So, your 'stiffness' changes over time. And, if you are designing the mount for them, try to copy the idea from the front wheel bearing off an old truck. Use a fine thread on the spindle and a castle nut with a cotter pin to retain it. Then bury the whole thing in grease or oil.

The drawbacks don't matter in my applications (or I make it so they don't). The nice thing about them is that they last forever if you keep your assembly sealed. For fairly low RPM, tough machines, I love them.

They don't get used on ball screws much because you can't just buy them and put them in -- everything in the design has to revolve around them, sorta like certain kinds of women.

[NC Cams]

97T's reply is pretty much spot on.

This "run in" issue can be addressed somewhat at build by continuously rotating the shaft while you turn it so as to take out the cumulative end play as you get the rollers to fully seat on the raceways.

This way you can set the endplay at 0.000" for max stiffness with min torque. If you want max stiffness, however, you can now more precisely find preload and set it properly without overloading the bearings with negative clearance - rotating torque WILL go up as you go into negative clearance/preload with a taper.

The rise in rotating torque comes from the fact that the ends of the rollers are eventually absorbing the axial thrust as they ride against the flanges - this is where the friction comes from, namely SLIDING friction. The ends of the rollers are "moon walking" against the flanges as they roll along the raceway.

Whereas an A/C has pure rolling friction under preload, the taper is a combination of line contact of the rollers (tremendous radial capacity) combined with lots of flat surfaces sliding against one another (sliding friction and high thrust capacity) at roller to flange interface.

The issue of heat can be addressed by a pressurized jet of oil or oil air mist. This requires a recirculating lube system. Grease will work but thermal stability and resultant size stability will also be compromised over time.

97T's comments about low RPM machines is again, spot on. Low speed grunt lathes run superbly on tapers. High speed pretty much necesssitates balls - the inability to take massive cuts with high speed ball fitted spindle is made up for by many more light cuts - speed vs grunt.

Normal class "axle bearings" are quite affordable. However when you start looking for PRECISION tapers, price goes UP and availability goes DOWN.

ANY bearing application depends on design integrity for the machine to function well. Misapplied parts are usually quite inexpensive and highly affordable - they may or may not work or they will work but for a short time. If you're willing to accept the compromised limitations of tapers, they are perfect for your needs and/or budget.

HOWEVER, don't buy a plow horse and expect to win the Preakness or Kentucky Derby with it. Ain't gonna happen, dude.

[Willbird]

This is all just food for thought :-). I'm not understanding the ends of the rollers running against the flanges, I'm going have to look at some and get that part in my head, I thought that properly done only the OD of the rollers would touch anything...

Bill

[Bear]

Hi all !
muchado about nouthing !
have rebuilt many grinders, and tapered roller bearings
are way better than the bushings they replaced.
Bear

(however, the ball bearing assm. out of early model
automotive front end wheels are more than adaque)

[NC Cams]

"muchado about nouthing !" - I dunno about that.

"front wheel bearings" are quite adequate for many applications, especially applications that require "wheel bearing" runout performance.

However, for SOME grinder applications, wheel bearing bearings are woefully inadequate - too much runout potential and not enough stiffness. Wheel control (finish and size control) will suffer from deficiencies in these areas.

Taper rollers are great for some grinders but NOT as a hasty generalization/universal fits-all for anything that has/uses a grinding wheel.

Some very fine precision grinders use precision bored sleeve bearings - some oil and some air lubricated. Why? Sleeve bearings don't allow the shafts to "orbit" as ball and roller bearings do/can/will allow the shaft to do. Think not, check the runout potential for the various ABEC levels, including ABEC 9.

Keep in mind that a lot of front wheel bearings were actually fairly decent angular contact bearings suitable to be "taken appart" (essentially they're a magneto bearing). They had axial thrust capacity (critical for cornering stability) and sufficient radial capacity for long life. They also had controlled radial runout for smooth, vibration free rides.

"Wheel bearings", even though they may not be absolutely optimum for a machine tool spindle, are quite charming for use by the DIY'er for three reasons: their cost, their capacity and their generally suitable accuracy for semi-precision operation.

It takes a lot of care and precision to make a seemingly generic "wheel bearing" perform as well as they do. Anyone who has been involved with a wheel bearing development program knows that they are anything but lame a$$ generic ABEC 1's that merely hold up the corner of a car.

[Willbird]

"Time in service" is another factor to think about. I think about automotive bearing examples because I know the conditions and such that they are designed to work in..and the horsepower, torque, and some idea of how they are loaded in their original application, so by looking at common bearings I think we can "steal" the engineering already done.

"Time in service"...if we look at 200,000 miles on a vehicle, at 50 mph that is only 4000 hours of service, in machine shop time that is 2 shifts a day, 40 hours a week, for 1 year...200,000 miles sounds like a LOT for a car, but in machine shop time it is not very long, if we had to replace our spindle bearings every year we would think we had a poor machine, if we get 200,000 miles out of a set of wheel bearings we are quite happy with that :-).

I do agree NC that a sleeve bearing takes 3 parts out of the runout equation, the inner and outer races of a ball bearing and the balls themselves. The spindle bearings on a centerless grinder are one example of a plain bearing that is anything BUT simple :-).

Bill

[NinerSevenTango]

The nice thing about automotive technology is that scaling for mass production brings the cost down -- you can get an awful lot of engineering trial and error and an awful lot of precision for literally pennies. I cheat like that wherever I can. Many times if you just pick a bearing size and rating out of a book the cost is hundreds, but if you look for ones that are mass produced, you can get something close for 1/10th the price.

Water pump bearings are another modern miracle. Where else could you find a bearing that takes loads like that and stays sealed for so long while doing it? I use them for part locators in machines that spray them hard about once per minute. You can imagine the cost for a from-scratch design to live under those conditions, which my competitors have always done. I searched for water pump bearings, designed them in, and they outlast everything else hands down.

Anyway, the things I work with spin hundreds of RPM, not 20,000 RPM. And accuracy of around a thousandth of an inch is usually good enough. But the environment is tough. So I go for maximum ruggedness by trying to find parts that are designed for much harsher conditions, and made in huge quantities. When I have the luxury of being able to design my machine to take maximum advantage of their qualities while making their drawbacks inconsequential, then the payoff is nice.

CNC mill spindle cartridges and grinder bearings are another whole field altogether, though.

Oh, and on the friction in roller bearings -- the flanges trap the rollers to more or less keep them from 'squishing out'. The outer race and the inner race are not parallel, a line from each intersects at some point ahead of the bearing. So the rollers are conical in shape. To keep them exactly where they belong, a flange is machined in at the wide end, where some scrubbing occurs. It's not much, but it's a lot more than zero on a ball bearing. It's the price you pay for having so much extra contact area.

--97T--

[Willbird]

Those water pump bearings are so great that Wheel Horse used them for mower deck spindle bearings :-).

So I'm told anyway by a guy that worked on them, they were a cartridge type bearing that wheel horse ground thru the outer shell in one place to let grease from the zerk in the housing enter the sealed unit, the grind tru looked like it was done by hand not on a machine.

Bill

[NC Cams]

The so-called "time in service" factor has a lot of additional considertions associated therewith that aren't always documented.

In other, easier to understand terms, these undocumented factors are the same as "factors of safety" that one uses when designing a part.

A wheel bearing WILL see shock load - almost thermonuclear shock load in some instance. Sometimes it is a puddle in the road and it could be a rail road tie. For the bearing to survive, the instantaneous shock load that it has to endure will make it pretty much bullet proof in simple roll down the highway service. How many times have wheels failed yet the bearings are still fine after hitting a chuck hole at 60mph???

The friction differences of tapers versus ball bearings are well documented and duly noted. For many applications, car or truck wheel bearings make good candidates for adaptation into other uses - the economies of scale and development they undergo for abusive use make them a keen value for DIY projects.

Keep in mind, however, that the runout requirements of wheel bearings is nowhere near what you may want/need in a machine tool spindle. Although they deserve consideration, the last sentence of posting #3 still applies.

[S_J_H]

As to precision however......if you borrowed two of the tapered timken bearings used in a 1 ton truck rear end (new ones not used) and you set them up with the OEM preload.....for an application like an OD/ID grinder work head that runs at typically fairly low rpm, would they work ?? how "stiff" would they be compared to AC bearings ?? I know they would COST a lot less....but they seem to offer the benefit of handing radial and thrust loads both at he same time.

Hmm, well I have installed many roller bearings into automotive rear ends and the typical rotating drag measurement used to set preload on a brand new bearing in the typical automotive size is about 25-30 inch/pounds.
They loosen up after run in by quite a bit though.
I am not sure if that is considered a lot of drag or not for a machine tool.

Steve

[Bear]

Well,
maybe I should not have called them wheelbearings!
Should have called them Hi-precision circular rotary
rolling devices,or sum P.C.hootofruty name!
The point is they work fine. Have for several years and
and still are.This is not rocket surgery.
Bear
P.S. I dont think anyone could the difference!
(tell)thank you.

[NC Cams]

Bear: It is not what you call the bearings that is the issue at hand here, it is the drastic oversimplification of the application/selection process that I"m taking exception to.

Bearings are robust devices that many people oversimplify when they try to apply/use them. Automotive wheel bearings are merely better developed, high volume, low price variants of generic bearings - and very often contain special, undocumented internal features which drastically differentiate them from their generic cousins.

Few if ANY people read the applications manual when they "select" bearings. Moreover, many people, including and especially engineers who should know better in light of their training, pick stuff on size or SWAG - fewer yet do the load/life calculations.

Tis a pity they don't. But why should they? "This ain't rocket science" and yet the user quite often doesn't pick the right stuff or install it properly and then wonders why it doesn't work or "loosens up when it beds in".

You have no idea how many times some ground based engineers SWAG'd a bearing appication, had problems with it and gave me the "rocket science" line.

Oh yeah? If it ain't rocket science, how come so many guys screw up bearings? How come they don't know how to preload them so that they don't "loosen up when they run in" (it is possible to minimize or eliminate this, IF YOU KNOW HOW).

BTW, the reason why "wheel bearings" are so robust is usually due to the beyond aircraft grade, undocumented special features that are incuded in them by the bearing manufacturers/developers.

The kick-ass, scienced out undocumented special stuff that is taken for granted (this ain't rocket science afterall) by the users of auto grade bearings is the unheralded reason why wheel bearings are as robust a performer in "off label" use, especially at the cost and size of the bearing.

Don't belittle the efforts of the engineers who develop wheel or other automotive bearing applications by saying "it ain't rocket science".

Outside of aerospace use where lives depend on it, automotive grade bearings are some of the most highly developed and extensively tested bearings that are ever made - especially wheel bearings, afterall, they hold up the corner of the car.

Moreoever, wheels may break but the bearings will survive some bone crushing impact loads - can a rocket can survive similar treatment? Not hardly.

BTW: although they still inspect quality into aircraft bearings but they build quality into automotive bearings with quality levels (PPM defects) that the aircraft industry lusts over.

I know for a fact that bearings made on an automotive producti line were merely inspection documented and sold otherwise "as-is" into aircraft applications - the auto stuff is/can be THAT good.

The auto bearings may not have the accuracy of machine tool bearings BUT they will take way more abuse via the factors of safety that are, out of necessity, a part of the design and/or construction.

"...P.S. I dont think anyone could (tell) the difference!..."

Perhaps you can't tell the difference, but a person who knows what to look for in bearings can tell the difference between generic, automotive, machine tool and/or aircraft bearings. And it don't take a college degree or a rocket scientist to find/uncover the differences if you know how and where to look. Anyone who's ever PROPERLY set up a Ford 9" or Dana 60 or GM 12 bolt can tell just by the sound if the bearings are right or wrong.

If the average user CAN'T tell the difference between a machine tool bearing and an automotive one used in a machine tool environment, the automotive bearing engineer and bearing maker did their jobs admirably well - the bearings don't need to be that good - afterall, "it ain't rocket science" nor was it intended to be machine tool grade (my corollary to the too oft repeated rocket science line).

[Willbird]

I was thinking more about the pinion bearings BECAUSE, in their original application they were designed expressly to be run with a preload, where the tapered timpken wheel bearings are designed to be run at "0" play or with just a bit of slack.

Also several OEM pinion bearings, and even the ring and pinions themselves are used in Halibrand sprint car rear ends where they hande a lot more horsepower and torque than in the OEM application, albiet with somewhat larger dia tires than most OEM use so the pinion rpm is maybe a touch lower. Lighter vehicle too, but more applications of high horsepower than OEM.

I know there is nothing free in life, and bearings are interesting to study on.

Bill

[NC Cams]

Yes, pinion bearings are substantially preloaded (negative axial clearance)whereas wheel bearings are run at 0.000 to 0.005" positive end play/clearance.

By the way, the preload is considered as part of the loading that the bearings see as a constant force, the reactive force from applied torque is cacl'd as part of the variable force histogram when/if a proper load life calc is performed by the system designer.

The preload is critical in establishing stiffness in order to keep the pinion depth constant - besides the applied torque can't move the pinion until/unless it overcomes the properly established/qualified preload.

Another problem you have in Halibrand and/or other Q/C's is that the case is aluminum. As it heats, if forks up the preload and/or fit calcs of EVERYTHING in the housing. That's why you MUST warm up a Q/C before you run it - and it takes some real experience to properly set one up due to the thermal growth and compliace issues.

It ain't the rocket science that some guys like to talk about BUT it involves high falutin math and/or experience. You have to rely on these tools to properly compensate for the thermal and mechanical growth/deflections that are endemic to aluminum or almag housings. BTW, these thermal and mechanical deflection issues are NOT experienced to the same degree in iron housings - God said so when he created the elements that we learned to use in the rear end housings.

If you don't do it (set preloads & clearances) properly, the beast will go primitive and eat itself. This usually generates a "DUH what???" reaction the first time you encounter it. Beginners use "bad bearings" as the allibi but veterans know better and just smile.

[Willbird]

Some Halibrands were magnesium :-).

Yes the case is a floppy noodle compared to the OEM rear end the ring and pinion were borrowed from. I worked as pit crew and machinest for an outlaw sprint team, The car owner went thru several ring and pinions until my Dad set one up, that one went 3-4 years without a hitch...he didnt use any rocket science beyond normal ring and pinion setup. HOwever the moron that owned the car was putting it back together with the front pinion support bearing bore in the case cracked, dipstick figuired it had two OTHER bearings so that one wasnt all that important :-).

So dad made him have it tigged and we re-machined it proper...and Dad put it together, and as I said it worked just fine after that no problems.

I know the one case he had was Mag because the guy that tig repaired the alimunum one remarked on it, and said he had tig welded mag stuff and you had to be damn careful because if you got the whole thing too hot it would start to burn and then all you could do was drag it outside and let it burn itself out.

It is real hard for us self taught guys to dig into the literature and wrnagle out the answer we want and need, kind of similar to learning how to do quadratic equations, a lot of hard mental work and study, however I feel it is worthwhile to do it, and having hands on working examples in your own shop and home to look at for a gut check that you have at least a rudimentary grasp on the concept is always good.

Bill

[Bear]

Well said,

Beginners use "bad bearings" as the allibi but veterans know better and just smile.

Bear

[NC Cams]

With appologies to Bear & Willbird:

I take strong exception and/or aversion to the "not rocket science" comment that gets thrown around. The reason why stuff is so robust (ergo "wheel bearings") is that someone, at sometime went thru the hassle of working out the numbers and came up with some real robust bearings that will:

a. take a 4" chuckhole hit at 60mph and survive
b. run 100K miles with no service
c. be cheap as all get out
d. be made in multi million piece quantities (do a parts count and you'll be amaxed at how many parts are invovled in the front wheels alone for 250k's worth of cars).

The guy who tig'd the Q/C rear end may not know how to do a quadratic equation (I hardly recall mystelf anymore) but then again, I can't run a tig or mig very well - at least not well enough to properly repair a Q/C rear end or weld up a roll cage that I'd feel safe in.

It is very wise indeed to used something that has been well and highly over engineered in a much more benign applcation - no need to reinvent the wheel. The problem comes into play however when someone SWAG's something improperly and uses something that is barely adequate for the off label use or unsuited for some unseen reason (the part simpy wasn't developed for that application).

Here's an example: the ball screw bearings in a particularly famous mill are encased in seals that say 6204. HOWEVER, the cores may have been 6204's but the special assembly and matching practices used to make the ball screw variants make them anything BUT 6204's.

Imagine the grief when Joe Tech opens up the machine, sees "6204 bearings" and orders some to fix his mill. Afterall, "it ain't rocket science, anybody can plainly see that the p/n's are written right on the bearing!!!!". Trouble is that the "6204's" aren't what they appear to be (different internal geometry) and the generic 6204's simply will suck as ball screw bearings.

I'd be inclined to ask the owner of the car with the cracked Q/C case as outlined in post #16 if he'd use a race car with a cracked main roll bar upright - afterall there are usually 3 more there to suppor the car if/when it flips.

Yes, sometimes you can and have to simplify things - other times, the use of oversimplified SWAG's can create more harm than the $$$'s or effort that they might save.

[Willbird]

The sprint car owner with the cracked housing, he was an auto body shop owner, we often thought the paint fumes had done some perm damage to him :-).

Tig welding a mild steel roll cage is probably one of the easiest welding jobs out there to do safely :-)...the Chrome moly structure of a sprint car ?? A bit tougher :-).

I can vividly recall some of the "roll cages" I have seen in stock body "bomber" division dirt track cars, black pipe with the threaded ends still on it, welded with AC stick welders, one comically with the welding rods the guy kept getting stuck in the weld still in place under the still wet coat of gold spray paint :-).

Inbred tech was to make a "fuel cell" of of schedule 80 steel pipe (big stuff) and half arsed fasten it in place in the left rear of the car somewhere...often welded to the black pipe "roll cage" with the AC stick welder every inbred dirt track racers has :-).

To be clear not all dirt track racers are inbred, not even all the ones racing stock body $500 calimer engine cars........the reverse is true enough that the examples I speak of stood out from the rest of the guys.......many spent far far too much on engine and not enough on chassis......stock DOT tires limited HP that could be put on the ground anyway. It sure was soemthign to look at and giggle about tho :-). And nobody got hurt as a result of it.


Running 13,000 rom with a cat50 spindle for hours and hours takes some rocket science I'm sure.

making a workhead for an OD/ID grinder that starts it's life using a Dumore toolpost grinder as it's grinding head as I am thinking on doing (it will have room for a red head grinding spindle however) is not quite as high tech.



Bill

[NC Cams]

If you saw the tech specs involved in the bearings that were fit into the Dumore's, you woudldn't say that.

We almost got our asses into a sling when a sales tech "interchanged" a genric electric motor bearing (because it fit the generic size) into the Dumore grinder extended nose spindle. The bearing would have either shattered or shook the grinder out of the person's hand had the substitution been made. The bearing had all sorts of trick specs spelled out in a "secret handshake spec" that the bearing developer came up with and wouldn't publish. Only by knowing the application did I prevent a fiasco from occurring.

The sales guy couldn't fathom why a generic bearing would cost nearly 4 times the market price. When you starting adding the options needed to get the bearing to run at the rpm and low runouts needed for a Dumore grinder, the costs added up AND FAST. In this case, the bearings were damn near rocket ship accurate - they had ot be!!!!

Make sure you check out the published tolerances for a high speed grinding spindle housings and/or shafts if you try to make one. There is a reason why OEM spindles work and do so for long periods with minimal service. Failure to get adequate shoulder squareness and/or shaft runout will result in an affinity for the spindle to eat bearings (usually break cages) or vibrate badly.

Guys who try to copy grinding spindles don't copy ALL the tolerances never seem to understand nor figure out why the bearings don't hold up or the grinders don't hold size or finish. After all, "it isn't rocket science".

I too have seen some lame roll cages. Fortunately guys haven't gotten hurt or killed. Then again, one guy I know contends that God created accidents to "thin out the herd and clean out the gene pool every so often". In that regard, the aircraft industry has a ruthless efficient way of eliminating BAD pilots. Fortunately, racing organizations do a decent job of keeping totally bogus roll cages out of competition anymore - they have to or should for insurance reasons.

The guy with the cracked Q/C probably barely makes any money in his business. Why would you tear up 2-3 sets of ring and pinions and/or Q/C gears and NOT fix/replace the housing? Would he use contaminated thinner to paint with? Talk about poor decision making process!!!! And people wonder why the same guys always win and the same losers always don't/can't win......