2007-06-11

Rotating Poles

So we saw these really nifty wind sculptures by Lyman Whitaker while visiting the Missouri Botanical Garden. Unfortunately, I did not get a very close look at them because of how they were installed, but I've been thinking about how to make some sort of less attractive approximation.

The only critical engineering element is the bearing. The Whitaker sculptures apparently use a sealed ball bearing "at the top of a vertical rod", which I suppose is to reduce the thrust load on the bearing. I'm unclear of their definition of top, however... It does make sense to place the bearing as high as you can, reducing the amount of the sculpture that will spin and produce drag.

There are lots of types of bearings. Generally I would expect a spinning pole to need to support a moderate thrust load with very little radial load, but at fairly high speeds. The speed probably mean that a ball thrust bearing would not work all that well, so I'm guessing a roller thrust bearing might be the way to go if a ball bearing doesn't cut it.

Wouldn't you think that there would be readily available plans on the Internet for making some kind of rotating pole thing? As far as I can tell, two kinds of rotating pole assemblies exist electronically: rotating flag poles, and trick stripper poles. Both of these use bottom-mounted bearings, which spin the full weight of the pole and anything hanging off of it -- which I guess means it is silly for me to worry about whether I need roller thrust bearings instead.

I guess what I need is a nice 12' galvanized steel pole to fit a sealed ball bearing's inner dimension, and a shorter length that fits the bearing's outer dimension which can depend from it and will have my windsocks fastened to it. Or I guess I could even use a separate bearing for each propeller element.
Something I don't fully understand -- remember, I'm a software guy -- is how to fasten the bearings. Do you ever have any choices besides pressure fits and straight welds? Larger bearings often come with cast mounting accessories on the outer piece, but the inner piece is often featureless. Flanged ball bearings provide another option, but don't really relieve my consternation about how best to connect one piece to another.

Maybe I'll email Lyman.

3 comments:

KRM said...

Reproduced without my father's permission:

"I guess I didn't realize that the whole sculpture rotated together. From the picture you sent, I thought each 'arm' had to rotate independently. For these 'pole perchers' all you need is one hardened steel ball at the top. You don't even need to seal it...it just won't matter. You do need a long (strong) internal rod, and if you want it to go in the slightest breeze, it would be a good idea to keep it from touching the rod by using a 'regular' ball bearing at
the bottom to maintain clearance...it would have essentially zero load on it. You still have to make all those perpendicular welds. They probably used two outer tubes, one continuous and the other concentric in short pieces, one for each 'wing'."

KRM said...

I think, actually, that at least for the sculptures with mutually opposed arms, the arms are single rods with loops bent in the middle that are crimped around the pole -- no welding.

Anonymous said...

I'd attach the "arms" to a hollow tube that rotates around a central pole (extending down into several feet of concrete, if this thing is going to be tall enough to do damage if it falls on somebody), with tapered roller bearings at both the top and the bottom. Weld or otherwise put a tapered race at each end of the outer post (just flaring the pipe might work); then use tapered collars on the inner post to hold the bearings in place which can be removed for service or repair. The bearings themselves don't fasten to anything, they're just wedged between the races. If you need an example, take the wheels off your car :)

Tapered Roller Bearing