Effect of spacers 1

[JeanMicheling]

I think I have a good question for all the people in bearing design applications. I have two different opinions from experience bearing guys. If I add a spacer between two angular contact bearings, both agreed on the fact that it will increase the system rigidity. But one says that it will lower the load on each bearing, thus increase the bearing life, because the load is better distributed. The other one says that it's the opposite, because the system becomes more rigid, the bearings carry more moment load, thus shorten the bearing life. I find that really interesting and I'd like to know if someone has an answer or an idea.

 

[Tmoose]

Free body diagrams would be useful.

I think both answers are right, and wrong, without better definition. If the moment is due to a forced displacement (like 3 misaligned bores)individual bearing loads might be less with small bearing C-C distance.
If a radial load is hung from a shaft cantilevered by the 2 bearings the loads would be greater with closer spacing.

 

[carburized]

The system becomes more rigid. This is true. That is the only purpose to use spacer. Of coures, when system gets rigid and stiffer, bearing life shortens but on the other hand system is more accurate.

Ravi

http://www.gearsandbearings.com

 

[volpegrigia]

I believe the question is incomplete or phrased incorrectly.
Adding a spacer between two angular contact ball bearings will not increase the radial or the axial rigidity of the system unless the spacer increases the bearing(s) preload. As to the moment rigidity of the system, spacing the bearings farther apart will certainly increase the ability of the system to carry load and its rigidity.
However, icreased rigidity is a double-edged sword. Induced loads due to alignment errors etc. will subject the system to greater loads and thus lower life.

 

[JJPellin]

There is not enough information to comment. Are the bearings mounted back-to-back or face-to-face? Is the spacer between the inner races only or both races? When I read the original post, I picture a spacer shim between the inner races in a back-to-back configuration, which would decrease the preload. All of our angular contact bearings are ground for flush mount in pairs. But some users buy bearings made for differnt mounting and ground differently. How was the particular bearing designed to be mounted? I feel like we are all picturing different configurations and making different assumptions when we comment. If they are asking about a simple spacer plate (inner and outer races) which moves the two bearings farther apart, then the new configuration will be less able to accomodate any misalignment between this bearing pair and the radial bearing. If there is little or no misalignment, this may not be a problem. If there is misalignment, it could be disasterous. But I could not immagine any advantage

 

[JeanMicheling]

The spacers are between the inner and outer ring and they keep the same preload. It's a back-to-back arrangement. JJpellin, I think there is an advantage, except increasing rigidity, to use spacers. It helps to keep the right preload and lowers the bearing temperature in operation.

 

[JJPellin]

Now that I understand better what configuration we are talking about, I have a few other comments. I would expect the use of a spacer to shorten bearing life. With the bearings further apart, they are less able to withstand internal misalignment to the radial bearing. It starts to look like three bearings housings on one shaft, which I had always been told was a bad idea. There is always some internal misalgnment between bearing bores. With the extra ridgity, the loads on the bearings go up. This also introduces a bending moment that could result in shaft fracture. I don't see why the bearing would run cooler. The heat generated by each bearing would be the same (or more). The spacer plate (unless it has some holes or slots) would shield one side of each bearing which reduces the ability of oil to flow through freely. I also don't see how it keeps the pre-load constant. If the plate is perfectly ground flat, it is not likely any flatter than the ground face of the other bearing. Flat is flat. But perhaps I am missing something.

 

[Tmoose]

"I don't see why the bearing would run cooler. The heat generated by each bearing would be the same (or more)."

For angular contact precision spindle bearings adding a spacer improves the heat/temperature related speed rating significantly. The common explanation is a high speed bearing is walking on the edge of heat rejection, and the extra spacer (and shaft and housing) provides a larger opportunity for conduction and convection.

My wife likes to cuddle when she is cold. Eventually we both are warmer.

 

[JeanMicheling]

JJpellin,

You said using a three bearings housings is a bad Idea. Could you explain me why? Can someone tell me what would be the misalignment limit for back-to-back angular contact ball bearings? I calculated the misalignment created by the load by assumpting that the shaft was built-in (the spindle is overhung). So I calculated the shaft slope and than, the angle at the bearing.

 

[Tmoose]

3 bearing bores can not be absolutely concentric. Even if they were, the bearing races have some eccentricity. etc, etc. There always is a tolerance. The middle of a nice short fat shaft won't move off center even 0.0001 inch without a fight. The bearing books all talk about manufacturing limits with Lots of decimal places. Hard to make. Hard to gauge. Expensive to buy. Vulnerable to even slight temperature variations. One bearing (or a close spaced pair or group of bearings) at each end of a shaft is much more forgiving and practical.

If your calculated shaft deflection is more than a few 0.0001 inch then the middle bearing loads are likely to be hundreds of pounds. Making a spindle system stiffer by sprinkling bearings along the shaft can be pretty hard to do. With any overhang at all the ratio is likely to be something like 70% shaft, 30% bearings. Better to put a fatter shaft in there.

 

[JJPellin]

Tmoose states the problem with three bearing locations along a single shaft very well. However, as I read this chain again, I realize that there are substantial difference between high speed spindle bearings and the relatively low speed (1800 rpm, 3600 rpm) bearings that I typically work with in pumps. What works well in one might cause serious problems in the other. I need to be more carefull in my replies to be sure I am addressing an area where my experience is relevant. I would never consider adding a spacer between the angular contact thrust bearings in an API process pumps. But that does not mean it would not have substantial advantage in a high speed spindle application.

 

[JeanMicheling]

My application is 3600 RPM, the spindle is overhung. 7020 are the only bearings I can put in there. I want to increase my load capacity and stay retrofit. So the only way I have to do that is adding a 7020 (if someone has a better idea, please tell me!). I thought the spacer would be good to reduce thermal and preload issues. The stiffness really doesn't matter in this application.

 

[Tmoose]

What makes you think you need more capacity?

Can I get a reasonably accurate sketch of your bearing arrangement and loading somehow?

The bearing manufacturers are often technically very helpful, even with one-off applications.

Thanks,

Dan T (moose)

 

[JeanMicheling]

I suspect I have a misalignment (caused by the load) issue. I think increasing the load capacity could help. I can provide you a sketch. Do you know a good free upload website?