Uneven Thrust Loads on Needle Roller Bearings

[Madison]

or "How I Need to Make A Turntable That Will Not Bind."

So, I am in the midst of designing a mechanism that rotates about a vertical axis and I'm worried that the enormous attached moment arm will cause binding in this vertical pivot joint.

McMaster-Carr shows needle roller bearings that are rated for 1900+ lbs. of thrust, but I assume that is assuming even thrust on the washers at each end of the bearing. How does an uneven load affect the performance of such bearings? Are they a viable solution for this problem, or is there some other method that's commonly used for supporting vertical axles under torque load?

Attached is a graphic and, by circumstance, a vague look at the arm design. I hope it makes my concerns clearer.

[kiracofe8]

Originally Posted by M. Krass

or "How I Need to Make A Turntable That Will Not Bind."

So, I am in the midst of designing a mechanism that rotates about a vertical axis and I'm worried that the enormous attached moment arm will cause binding in this vertical pivot joint.

McMaster-Carr shows needle roller bearings that are rated for 1900+ lbs. of thrust, but I assume that is assuming even thrust on the washers at each end of the bearing. How does an uneven load affect the performance of such bearings? Are they a viable solution for this problem, or is there some other method that's commonly used for supporting vertical axles under torque load?

Attached is a graphic and, by circumstance, a vague look at the arm design. I hope it makes my concerns clearer.

Well, the easiest thing to do is to support the shaft by two bearings that are spaced a good distance apart. Then, instead of one bearing that has to take a large moment, you have two bearings that each have to take a small force. This is the principle behind why doors always have (at least) two hinges. If you only had one hinge, it would have to take a tremendous moment. But if you have two hinges about 6 feet apart, each only has to support a few pounds of radial load.

[Madison]

Originally Posted by kiracofe8

Well, the easiest thing to do is to support the shaft by two bearings that are spaced a good distance apart. Then, instead of one bearing that has to take a large moment, you have two bearings that each have to take a small force. This is the principle behind why doors always have (at least) two hinges. If you only had one hinge, it would have to take a tremendous moment. But if you have two hinges about 6 feet apart, each only has to support a few pounds of radial load.

Well, I should I have mentioned that I plan to support the shaft with two bearings. Despite that, because of its design and application, the load that the arm exerts (~200 ft.*lbf.) must be cantilevered off the shaft. I can't support the shaft above and below the arm, only in two places below the arm. The separation of the bearing was planned to be only about 4", but I can investigate increasing that distance to something larger; maybe as high as 12".

Kevin -- I'll look into angular contact bearings. I'd never heard of them until now, so I don't know what they're typically used for, nor do I understand how they work Thanks for pointing out their existence.

[Kevin Sevcik]

As long as you have two bearings supporting the shaft, it won't be that much of a problem even if they aren't on opposite sides of the shaft. Of course, the more seperation the better, but 4" can't be THAT bad. wil a ~200ft-lb moment, it'd put less than 1000 lbs of radial force on a bearing. MSC actually has some angular contact bearings that look more than capable of dealing with that kind of load.
Also, think about way to counterweight the thing so it doesn't apply that kinda moment in the first place.

[Kevin Sevcik]

I would think a decent solution would be to use two angular contact ball bearings. Using two of them helps turn the pure moment into two radial loads, and angular contact bearings are meant to handle thrusts, so it'd make sense to me.

Here's a link to a page about angular contact bearings:
Torrington Angular contact bearings

[Jay TenBrink]

[quote=M. Krass]or "How I Need to Make A Turntable That Will Not Bind."

mechanism that rotates about a vertical axis and I'm worried that the enormous attached moment arm will cause binding in this vertical pivot joint.

We used a bearing from Kaydon for this very same application on last year's team 494 Martian robot (it had a turret and elevating arm). It was a turntable bearing that we mount in an aluminum disc that gives a a sort of "lazy susan" platform that will take well in excess of 1000 in-lbs of torque. You can mount whatever you want on this platform .

We cut teeth in the outside of the disc and drive it with a window regulator motor to get our azimuth control.

[Madison]

I know you're all perched at the edge of your seats awaiting the solution that was found for this problem.

In the end, we lucked out and found a widely available surplus mechanism that was used to rotate satellite dishes into alignment. At ~6 lbs., they're heavier than one might like, but they're substantial enough to take this load without a problem and include gearing to slow down the rotation to a manageable speed.

While I typically like to design from scratch, this was a clean solution to a sticky problem that freed up our resources to tackle some other pressing issues.

Thanks for everyone's suggestions and help.

[Frank(Aflak)]

Originally Posted by M. Krass

or "How I Need to Make A Turntable That Will Not Bind."

So, I am in the midst of designing a mechanism that rotates about a vertical axis and I'm worried that the enormous attached moment arm will cause binding in this vertical pivot joint.

McMaster-Carr shows needle roller bearings that are rated for 1900+ lbs. of thrust, but I assume that is assuming even thrust on the washers at each end of the bearing. How does an uneven load affect the performance of such bearings? Are they a viable solution for this problem, or is there some other method that's commonly used for supporting vertical axles under torque load?

Attached is a graphic and, by circumstance, a vague look at the arm design. I hope it makes my concerns clearer.

I think what you are saying is you are trying to prevent the shaft that has the arm on the end from twisting in its bearing.

We also have a long arm on a turret, we pulled it off using a long (4 feet or so) pipe. It has an oil-impregnated bronze plain bearing at both the top and bottom of it, so there will be no radial loads, only side loads, which plain bearings stand very, very well. The drawback is you have to design around it because it gets in the way, and its a little late to be working one in, but it is the best way we could think to do it.

I'm so nice, I cropped a pic of ours to make you a special teaser. Maybe I'll talk to my team about posting full images tommorow or something. For now, be content with this sliver of a picture. Both the top plat that has a bearing and the bottom plate are secured and very stable . . the pole will be driven and our arm will be mounted to the top of it. Thus, no radial loads on the bearings, only a side load.

The pole and bearings weigh maybe 3 pounds. There 'may' have been other stuff used that has mass and whatnot, but the entire turntable system (not including motors and arm) is probably aroun 5 pounds or so.