Ball Screw Noise - Normal??

Forgive the top posting. Netiquette or not it often saves people from unnecessary scrolling. Do you know who was used as the distributor of the Thompson ball screw? They should be advised of the problem, especially if this ball screw hasn't been in operation over a long period of time or under very adverse conditions (very fine abrasive dust or a lack of maintenance). At any rate, they should be willing to send a representative out to check the ball screw to see whether it is actually a problem. If another company put the linear translation stage together for you, they should still be able to tell you who they use as a distributor from archived Purchase Orders. If they can't then it's fair to ask them "why not?".

Mark 'Sporky' Staplet>

Thanks for the reply Mark. We purchased the Ball Screw directly from Thompson Industries, and we have made them aware of the problem. Our senior ME is convinced that the problem is due to a bad diameter/length ratio, meaning the ball screw was improperly specified for the application. This would result in a droop that may cause a compression load on the raceways. His inclination is to redesign and buy a larger diameter ball screw.

I just wanted to see if others had experienced/solved a similar problem. I'm not so convinced that droop is the cause, it doesn't explain the noise in only one direction of motion.

We will pursue getting a Thompson rep out to our facility, but we need to solve this problem in short order.

"strato" wrote in news:1107116172.387486.278910 @c13g2000cwb.googlegroups.com:

Here are some tips to diagnose the problem. In only one direction, leads to the initial conclusion that it is not the screw/nut, but the thrust bearing making the noise. When the screw moves toward the motor, the thrust bearing on that end takes the load. When the screw moves away from the motor, the bearing on the other end of the screw takes the load. (In a dual-bearing supported screw arraingement.) I strongly suspect either a bad, or mis-preloaded thrust bearing on the motor side. Take a magnetic base, and dial or last-word type indicator, and place the tip against the end of the screw. Grasp the mechanism being moved by the screw and try to move it linearly along the screw. If the screw moves, you have too much end play in the thrust bearings. If the screw does *not* move, place the indicator against the mechanism being moved, and attempt to move it again. If it moves, while the screw remains stationary, then you have lost-motion somewhere in the ball nut, or the load attachment to the nut. I still highly suspect a bad or mis-adjusted thrust bearing.

The long length of the screw, will lead to thermal growth. In this type of set-up, typically the screw is pre-tensioned by the expected amount of growth.

Anthony,

Thanks for your detailed response! No one on our team considered the cause might be a thrust bearing. I'll try and implement the tests you outline. We'll have another look at it tomorrow. I don't think thermal growth is a concern, we have a very low duty cycle and a relatively slow drive speed. In addition the environment is very well controlled. Thanks again,

Strato

Are you absolutely certain the ball screw is at fault, could the noise originate from a thrust bearing at one end ?

"strato" wrote in news:1107130751.515434.292750 @c13g2000cwb.googlegroups.com:

Please let use know how it turned out, / what you found.

Hi Anthony. After some extensive testing today, we are all convinced that the problem is in the nut with the circulating balls. We decoupled the nut from the stage and ran the ball screw with the motor while manually holding the nut. We noticed that any load placed on the nut, opposite to the direction of motion, caused the grinding noise and excessive vibration. The ball screw/nut is rated for over 1000 lbs, so I don't think we could apply more than 10s of pounds manually. Still not sure why the noise is more prevalent in one direction.

Regarding the thrust bearing idea, this design actually uses a radial contact bearing. There was no detectable axial slop in the ball screw stackup. There was some play in the ballscrew nut, but I think that's to be expected (it's NOT preloaded).

We were also able to rule out any ballscrew sagging as the cause as well.

Thanks again for all of you well thoughtout suggestions, they are most appreciated.

"strato" wrote in news:1107223052.238351.111650 @c13g2000cwb.googlegroups.com:

The ball nut should be preloaded. Usually the pitch (lead) of the ball track inside the nut is varied to a smaller dimension than the pitch of the screw as the track approches the center of the nut. Thus the slightly smaller pitch between the ball tracks in the center of the ball nut create load on the balls. Could be a bad screw/nut out of the box. Rare, but not unheard of.

Dan

"strato" wrote in news:1107223052.238351.111650 @c13g2000cwb.googlegroups.com:

Just out of curiosity, is the nut lubricated via central lubrication, or is it greased? or...no lube at all?

Hi Dan. Would you expect to have say .050 axial play in a typical ballscrew nut? Initially we had considered two opposing nuts, one on either side of the stage as a backlash takeup scheme. The idea was to create preload by adding some compression on one of the nuts.

Strato

Dear strato:

I find, using Google Advanced all the words: reverse the exact phrase: ball screw any of the words: noise vibration ... 737 hits, such as URL:

Your bearing-set may not be "symmetrically designed". I would recommend contacting the manufacturer of application guidance.

Preload seems like a good idea, as long as reverse acceleration is small.

David A. Smith

Thanks David. I did indeed do a google search prior, but not using the advanced features, they do provide better filtering. Indeed the article you point to describes what I think may be the actual problem, The ball tube outlet is misaligned relative to the ball screw grooves, similar to what you are saying above. The mfg's response has been to send the whole assy back, they agree that there is a problem, but we really don't have time to turn this into a big experiment. I think we're just going to request a replacement nut.

Regarding preload, our acceleration spans about 1 second and Vf is slow, ~1 rev/sec.

Strato

That sounds like wayyyyyy too much play. A lot of ball nuts have non-retained balls. In other words, if the nut runs off the end of the screw, the balls fall out. I wonder if you accidentally lost some of your marbles?

Don Kansas City

"Don A. Gilmore" wrote in news: snipped-for-privacy@individual.net:

In the back of my mind I was thinking the same thing (lost balls). .050? Are we talking fifty thousandths of an inch or fifty microns? Big difference. .050" is way too much backlash. Send it back. 50 microns is a lot for a new screw/nut, but could be in line with the set up and quality of the screw.

Dan

I have a device (a small shake table) that incorporates a motor driven ball screw arrangement to drive the table. I had a student setup the device & get it up and running.

He mentioned to me that the table seemed "a little noisey". I took a look at it . It was a lot more than a little noisey.

It sounded like s**t!

It had been damaged in shipment such that the keeper & wave washer allowed excessive axial play. Removing the axial play fixed the problem.; noise gone.

cheers Bob

I have a device (a small shake table) that incorporates a motor driven ball screw arrangement to drive the table. I had a student setup the device & get it up and running. awhile ago.

He mentioned to me that the table seemed "a little noisey". I took a look at it . It was a lot more than a little noisey.

It sounded like s**t!

Upon examination we determined it had been damaged in shipment such that the keeper & wave washer allowed excessive axial play. Removing the axial play fixed the problem.; noise gone.

cheers Bob

Bob, this sounds exactly like our situation. In fact we had concerns about the way these items were packed when we received the shipment. In any case we are going to replace the balscrew-nut. I don't have a lot of experience with these devices (mainly use power screws), so this NG has been of great help. Thanks to everyone!

Strato