Advice needed - ballbearing solution for rotating table 1

[makerjack]

Hi all,

Brand new to the forum, lovely to meet you all.

I am currently undertaking a project to build a large round table composed of a series of concentric rings that all spin independently.

The basic idea is, there is a single stepper motor attached the small circle in the center of the table. The center can be spun, which locks into the other rings around them and forces them all to spin either clockwise or anti-clockwise.

Each of the rings of the table will be mounted with progressively larger arms down to a central metal 'trunk', which will allow them to all spin independently. Like those mechanical models of the solar system.

The thing I am struggling with currently is the cylindrical ballbearing solution that will allow the 'arms' to attach down the central trunk, be supported, and also spin freely. I have attached a quick exploded 3D example of the design (except the tabletop is interlocking rings, not circles). The solution I am struggling with is for the 'red' pieces in this drawing.

Any advice much appreciated!!

Cheers,

Jack Colley

 

[EdStainless]

How much load is on them? What speed?
There are some great sleeve bearings made from high performance plastics (PEEK with Teflon) that might work well.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[Jboggs]

Google thrust bearings. A combination of thrust bearings and radial bearings will accomplish what you need.

 

[makerjack]

Thanks so much for the response guys - so I could weld supports onto the thrust bearings and have that sort of vertical weight and they would be ok? It would only be a few kilos... probably the heaviest would be a 300mm wide ring of 30mm ply at 1300mm diameter...

 

[SwinnyGG]

Definitely don't weld to any bearings, ever.

Find the size appropriate here:

http://www.mcmaster.com/#standard-rolling-element-thrust-bearings/=132bhqt

Note that the smallest bearing available, for a 10mm shaft, has a rated thrust load over 2,000 lbs. Including the washers (which you will need, top and bottom) you're looking at $6 per part. We don't know your budget, but I have to imagine that's cheap enough.

Each ring would be supported by a machined ring around the central shaft, which you would weld to. Inside this shaft ring you would use a plain or roller bearing, with a thrust bearing between each one to handle the axial load. Simple.

 

[BUGGAR]

Stack a bunch of automotive hub unit bearings, each with its own extension arms.

Ps.: I've started to use these bearings for everything.

 

[makerjack]

Wow - thanks everyone. This is so useful

"Each ring would be supported by a machined ring around the central shaft, which you would weld to. Inside this shaft ring you would use a plain or roller bearing, with a thrust bearing between each one to handle the axial load. Simple."

This sounds really great but I'm having some trouble visualizing it... specifically the part about 'inside the shaft ring'. So the machined ring is what I weld the arms onto, and inside of that we use a plain or roller bearing which sits around the central shaft? Where does the linked rolling element bearing come into the picture? Sorry!


@Buggar - that definitely sounds more simple. You mean something like this ?

And they are ok to sustain the axial load if I attach the arms onto them?

 

[BUGGAR]

No, it's more of a double flanged bearing on a splined shaft. One flange rotates and you can hang things from this. You can use a plain shaft instead of splined. I found pictures but couldn't download. I like Moog Part No. 513282; they're for Camaros and maybe too strong for your use. You can get smaller and cheaper. I have used them for children's merry go rounds and to adapt hydraulic motors to drive wheels for construction equipment.

 

[SwinnyGG]

"This sounds really great but I'm having some trouble visualizing it... specifically the part about 'inside the shaft ring'. So the machined ring is what I weld the arms onto, and inside of that we use a plain or roller bearing which sits around the central shaft? Where does the linked rolling element bearing come into the picture? Sorry!"

You have several ply rings with arms

you have a central shaft

You weld the arms to a machined ring. Inside this machined ring and riding on the central shaft goes a plain/roller/ball bearing, depending on your cost and load requirements.

So if you have 5 ply rings, you will have 5 machined rings, stacked vertically on the central shaft. The thrust bearings go in between, to allow the machined rings (and thus the ply rings attached to them) to rotate independently.

I would create a drawing but I don't have the time right this minute- hopefully this explanation is clear.

Buggar's solution is similar to mine- my idea will be cheaper, his idea will have higher load capacity.

 

[makerjack]

Ok that's starting to make total sense now, except one piece.

Inside this machined ring and riding on the central shaft goes a plain/roller/ball bearing, depending on your cost and load requirements.

How do you mount the mount the machined ring to the bearing?

Think I may just buy a hub bearing for ease of use, but now I'm just desperate to make sense of your awesome solution!

Thanks so much guys

Jack

 

[SwinnyGG]

there's a lot of ways to skin that cat:

-design the machined ring such that the bearing is a press fit
-design the machined ring such that the bearing is a slip or light interference fit, with grooves for snap rings
-design the machined ring with a lip on the bottom of the internal bore to prevent the bearing from sliding down the shaft

Which method I would choose would depend on the load, cost target, and whether or not you expect the assembly to need to be serviced at some point. A press-fit bearing will handle more load, but will be more difficult to service if bearings ever need to be replaced. Slip fit with snap rings will positively locate the bearing and allow for very easy disassembly/service, but will be more expensive to machine because of the snap ring grooves.

Based on your description, I'm guessing that the loads this assembly needs to reliably support are small. If that's the case, I would use a small lip on the bottom of the internal bore in the machined ring, with the rest of the internal bore sized for a slip fit on your bearing of choice. Gravity will keep the bearing in place.

The hub bearing route is a perfectly serviceable way to go. Using BUGGAR's idea will ensure that your assembly is bombproof. The only advantage my concept has over his would be cost.

 

[Jboggs]

makerjack, I'm glad you've discovered thrust bearings because you will definitely need them. But I'm still afraid of another style of loading: moment load. That's what happens when the direction of your load and your support are not in line with each other, and it can lock up an otherwise good set of bearings in a heartbeat. Thrust bearings are good, but unless they are designed to, they cannot absorb moment loads. That means you will still require good radial bearings in between the thrust bearings. And the best arrangement is always the same - separation of support points. What that means is that you want to make the distance between each of your thrust bearings as large as possible so you can separate the upper and lower contact points of your radial bearings.

 

[makerjack]

You guys are amazing, really. I can't thank you enough.

In regards to this most recent point about moment load and thrust bearings - does using the wheel hub bearing assembly take care of all of this? I imagine it is totally self contained and all I would have to do is mount the inner shaft onto my metal pole and then it can deal with the various stresses that will apply to it following that.

Thanks!!

Jack

 

[Jboggs]

Depending on the exact design, it might very well work as desired.

 

[BUGGAR]

Hub unit bearings from a Camaro are good for about close to one g lateral on a 3000 lb car with 26" diameter tires. I'll let others do the math.

ps: Jimmie Johnson lets me drive his car if I sell his parts.

 

[SwinnyGG]

In support of your point BUGGAR, they are good for MUCH more than that. Think about how much moment load goes through a wheel bearing when, for example, a Camaro cornering at full speed brushes a curb, or hits a small bump in the road.

 

[makerjack]

That should be more than enough for my relatively weightless spinning table then!

One final question on the actual mechanics of the wheel hub assembly... so if I mount the inner shaft onto my 'trunk', the outer assembly which is normally what you anchor onto the wheel assembly will then spin and I can attach extension arms onto it? So all I have to do is find a decent wheel hub bearing, and then mount the inner shaft firmly and make sure it is supported and everything is golden... right?

 

[BUGGAR]

Basically right. Find a friend at an auto supply store who will show you several in different sizes and get a hands-on feel for the things. For experimental purposes, I use free worn-outs (not legal for re-sale) from, of course, Jimmie Johnson Chevrolet. You may need an internal adapter to mount to the diameter of your trunk or this could be built into your longitudinal spacer.

 

[SwinnyGG]

Yes.

My suggestion would be to use a machined part, which is pressed on or slips on around the wheel bearing, and not to weld anything to the outer bearing race. If you weld to the race, A) it will be hard to weld because the bearing race is likely a very hard alloy not selected for weldability and B) you WILL warp the bearing race, possibly bad enough that the bearing won't spin.

How big of a stepper are you using? Wheel bearings do have a fair bit of preload designed in, so they aren't super free spinning like a small ball or roller bearing is. Just something to consider.

 

[MikeHalloran]

The idea of a post with a bunch of stacked bearings and bent rods seems to have closed your eyes to all other opportunities.

Is there some reason you couldn't just put ball supports upside down on the bottom of all the concentric rings and let them all skate around on the same surface plate, separated by simple radial bearings all in the same plane?




Mike Halloran
Pembroke Pines, FL, USA