ballscrew support idea / advice please

[DrStein99]

I'm upgrading my homemade XYZ cnc machine to use ballscrews - 3 lengths,30", 24", and 12". Using the 5/8 rolled ballscrew with thompson ballnut; simular to the ones I see commonly used offered by NOOK.

Unforetunately, I dont have a lathe - I have to get someone at machine shop to turn down the ends of my shaft. (Please dont reply and advise me to "simply" do what comes so easily for those who own lathes)

My original idea was to press-fit one end into 1/2 bore support bearing like this:

http://cgi.ebay.com/NEW-1-2-PILLOW-B...ayphotohosting

and on the other end, have a slightly loose or snug fit the same bearing; to allow heat expansion.

-------------------------------------------------
Does anyone have any advice if this would/wouldnt work ?

[vacpress]

people here will tell you all about $700 angular contact bearings... they will insist... however, i would imagine that spending many times on bearing blocks what you spent on ballscrew and nuts does not appeal to you.

this said, i think the pillow blocks would work. I am not sure about 'press fitting' though. if the screw has a max OD of 5/8, you still are trying to force 1/8" material into a 1/2" bore bearing. this might work... but you dont want to stress out your bearing. it will be taking an axial load it was not meant to take, and damaging it mounting the screw would be the first in what would turn into a series of blunders.

If you happen to be in illinois you can 'borrow' my shop for an afternoon and use the lathe.... otherwise you may want to see if someone else on the zone will extend similar invitation.

[mxtras]

Oh - NCcams is gonna have a field day with this one. Be gentle, Cams......

Scott

[DrStein99]

Please allow me to reiterate;

I'm doing design artwork for the guy at the pro-machine shop, who has years of experience producing military specification assemblies working with steel. In return for my design work - I asked him to help with the shaft (which doesnt mean 40 hous of labor).

Its up to me to do my reseach first; and know what to do.

My idea is to create a shoulder on the drive-side, and match the turned-down o.d. shaft to the i.d. of the bearing, and then create threads for a flange locking nut to pull the shaft through- and solidify tight on the motor side, much like a collet assembly locks into place by using two wrenches.

He has all the other jigs, clamps, vises, arbor presses, and bearing pushers to press any size shaft into any sized bearing - in addition to a 100% accurate set of measuring tools.

Those bearings are used in lots of heavy-duty machines; I see. And it has plenty of rigidity. Using a larger bearing; such as 5/8 bore, with a wider length-grab is also an option, as I can turn down the shaft and have a sleeve made to match a larger bearing.

This upgrade would mean alot to me, however - I do not expect it to work more then 6 months.

[JRoque]

Hello. Your idea works. This is more or less what I did in my build:



I thought that any old, cheap bearing would work if you *pulled* on the ballscrew outward instead of using precision bearing that push inward. If you push inward there's a more likely chance that the screw will bend under torque. This is minimized when you pull the ends out. Hope I'm making sense this late in the evening.

I invested $7.50 on a flange bearing with a collar. I put the screw through the collar and tighten it down while the base of the bearing was leaning forward, toward the screw, a bit. When I mount the bearing on it's base, it pulls the screw outward and prevents any slack/backslash.

JR

[DrStein99]

Originally Posted by JRoque I invested $7.50 on a flange bearing with a collar. I put the screw through the collar and tighten it down while the base of the bearing was leaning forward, toward the screw, a bit. When I mount the bearing on it's base, it pulls the screw outward and prevents any slack/backslash. JR

Wow - the exact info I was looking for! Nice to see the pics!

Can you help me out a little, I am unclear - is it a bearing with a collar as one PART? Or do you mean you bought a bearing AND a collar ? The bearings I have just have two set screws in them.

I am assuming you turned the shaft so it can slide in, otherwise - I dont understand how you pulled the shaft through if it was press-fit.

Do you have two collars on both sides?

[JRoque]

Hey Dr.

Ok let me give it another try. There are two parts to the assembly: a flange bearing and a collar. If you have the same function in one piece, that works too I would think. I have one collar and one bearing per axis end.

Yes, I turned the ends down from 5/8" to 1/2" so it would slide through the bearing and collar. I did not press fit; it slides in and out relatively smoothly. In my opinion, press fitting is obviated by the use of the collar and this setup where the screw is kept taught outward. A precise screw shoulder is also not needed for the same reason. However both of these can be added if you wish.

Below's a rough diagram of what I did. Note how the base, bearing and collar are tilted forward - exaggerated here for illustration. Once in this position, I set the screws on the collar against the ballscrew. As I tighten the frame screws, the collar pulls the ballscrew outward. In reality, the amount of tilt is minimal and just enough to take any slack off the bearing. Note also that this particular bearing I'm using swivels the bore a few degrees to accommodate for imperfect alignment. This is perfect for this application.

BTW, I'm also using the exact same ballscrew you describe. These run about 6' on X and Y. If I were to build a machine of this size again, I would go for a larger diameter ballscrew. For the money, however, it's a great value. I added a second ballnut as well to take up the backslash there.

JR

[DrStein99]

Noticing that there is a TINY little play inside the bearing itself; I was having 2nd thoughts about using this idea. Seeing the visual - now, I understand how you've worked it. Thank you.

With the offset bore; does the speed have any effect on orbit of the shaft?

With LESS parts which cost the least amount of money; I found this is one amazing feat for you! I am very impressed that this method has proven to work for you!

I really appreciate the advice. Before hearing from you - I was getting depressed over the idea of having to make an elaborate setup with metric angle bearings. Its clear to me that no matter how hard I try, my skills ATM will never be able to line up a perfect shaft anyway; and I was looking to use the most efficient means of fabrication w/ acceptable accuracy.
-----------------

Since my bearigns have two set-screws on them, I think I am going to take it a step even more, and try turning my shafts down to fit snug as possible, and try the built-in set screws as a collar/clamp.

[JRoque]

Hey no problem at all. Folks here help me all the time too.

"With the offset bore; does the speed have any effect on orbit of the shaft?"

I'm not sure I follow you. If you refer to the ballscrew end cut, that's perfectly centered in relation to the rest of the screw. If you mean that the bearing collar is offset, yes, it is so it holds better to be bearing. If both the bearing and collar are perfectly round, the collar would easily spin on the bearing. I made a reference to wanting a larger diameter ballscrew. The reason for that is that the screw tend to "whip" when spun at high speeds. This only happens, in my case, on one axis which seems to indicate the screw is not perfectly straight there. Pulling outward even helps minimize this effect, too.

I know, I know, I need to shut up about pulling the screw outward and whatnot. But it seems so logical to me I'm not sure why everyone else doesn't do it this way too. Sure beats paying >$100 for each precision end support while having to maintain dead on accuracy everywhere.

JR