Best bearing type for reciprocating motion? 3

[btrueblood]

In general, is there a "best" bearing type to use for a shaft undergoing reciprocating motion (e.g. + and - 180 degrees of rotation, no continous rotation). My understanding is that most types of rolling element bearings will eventually develop "flat spots" on the rollers due to repetitive contact stresses. Would ball bearings be less susceptible, or more susceptible to wearing in this manner?

Thanks,

Ben T.

 

[GregLocock]

I would have thought that 180 degrees would be enough to avoid 'brinelling' as it is inaccurately called in the motor trade.

Basically if each roller gets to perform one or more complete revolutions, and runs over the conact patch of its neighbour on every revolution you've covered most of the bases.

However if you were to restrict your angular motion to say +-90 degrees, then only half of the rollers would ever see any significant load, so your fatigue life would halve.

You could try some preload to even things up, but since it is the stress range that is the killer I can't see that helping much.

Would you consider a high pressure oil lubricated plain bearing?

Cheers

Greg Locock

 

[MadMango]

What loads are we talking about here?

[green]"But what... is it good for?"[/green]
Engineer at the Advanced Computing Systems Division of IBM, 1968, commenting on the microchip.
Have you read faq731-376 to make the best use of Eng-Tips Forums?

 

[btrueblood]

Hi Mango, Greg.

Relatively high loads (~1000-2000 lbs), very slow speed (<1 rpm). The question is would cylindrical or "needle" bearings provide better performance, because (in general) the load ratings are higher than similar sized ball bearings. I could maybe make an arguement that balls have the "ability" or at least the possibility, of rotating around to put the flat spot away from the line of contact.

Greg: so, provided the roller elements complete a full revolution, and travel over neighbor's contact patch, would you say the risk of Brinelling/flat spotting is minimized? I could envision a very small rolling element bearing (relative to shaft diameter) which could presumably satisfy the requirement with less than 90 degrees shaft rotation, or am I missing something? Seems like a pretty good design rule to use though, so thanks.

BTW, right after posting this, I "stumbled" (connotation of a bumbling idiot appropriate) over the "bearing design" forum. Probably should have posted there, sorry.

 

[MintJulep]

I'd consider an oil-impregnated bronze journal, or maybe even one of the plastic journal bearings (often refered to a bushings).

At such low speed the grease may not get hot enough to flow properly in a roller or ball bearing.

 

[GregLocock]

btrueblood - yes, that'll be OK for 'brinelling', but you'll still get a reduction in fatigue life because only half the rollers ever get involved.

I agree with MintJulep that sort of load is usally taken in a plain bearing.



Cheers

Greg Locock

 

[btrueblood]

That's what I thought, but was wondering if there was any way to get a little better efficiency (less friction), i.e. using a roller bearing. We may be able to live with low fatigue life. Would an oil-immersed roller survive better? As usual, the right answer will probably be, "dunno, why don't you build it and test it." Which I'll probably end up doing. Thanks again for the help.

 

[GregLocock]

I think you need to see how much of your activation load is lost in bearing friction. The trick is to make the bearing small in diameter, then the torque it can exert for a given friction is reduced.

I don't think flooding a needle bearing will reduce its friction, in fact for optimum resistance we used to degrease ball bearings and just use CRC (we replaced them every day). If you want to use ball bearings you'll find self aligning bearings have the lowest friction usually, as they can accomodate shaft bending without generating side loads on the balls.







Cheers

Greg Locock

 

[sreid]

As a general rule

Ball bearings = lighter loads, higher speeds

Roller Bearings = heaver loads, lower speeds.

The edict against short stroke motion with ball bearings primarilly has to do with greased bearings. Oil does not get to the short stroke area causing rapid metal to metal wear.

+/- 180 degrees is not short stroke. And if you put the bearing in a oil bath, lack of lubrication will not be a problem.

 

[btrueblood]

Yeah, I know Greg. My problem is that I need to attach a crank, and carry a hellacious torque at very low speed... the shaft dia. needs to be pretty big, and the side loads are high. Result is a big bearing carrying heavy loads, and the friction torque penalty goes up along with it.

Yoiks - unlubed roller bearings? How many cycles could you get doing that?

Hmm, good point about shaft bending, will have to include it in the mix of "things to look at".

 

[GregLocock]

About 10^6 (it was all we needed)

Cheers

Greg Locock

 

[unclesyd]

I don’t know whether they will take your loads and confines or not, but I would take a look at Linear Rotary Bearings. We built a lot of our machinery and used any number of these bearings with very good success.

http://www.e-lrb.com/index.htm

 

[btrueblood]

Hmm. Looks interesting unclesyd. I don't need any linear motion in my design, but their idea of steps/jogs in the race -- is that to try and get lube "under" the balls? Wonder if that would work for a standard ball bearing too?

 

[btrueblood]

Greg - wow, 10^6 cycles. Were those at fairly light loads then?

 

[GregLocock]

Yes, it was the rear wheel on a solar car. We used a pair of 20x8 (from memory) self aligning bearings for a 90 kg load, ie 45 kg each. max shock loading was maybe 2g.



Cheers

Greg Locock