Spindle bearing preload issue?
I finally finished up the servo driven bt30 project, mostly anyway. The new head assembly is built but i wanted to do some testing before rebuilding the z axis with new head. I bolted the new head down to my machime table horizontally and bolted some aluminum stock to the existing head for doing some test cuts. I had to do some dynamic balancing on the big 80t servo pulley, but the spindle runs smooth as glass at 12k, no balancing needed there. I proceeded to do some test cuts to see if the 1kw servo was performing as expected. It seems to top out at about 1hp at its peak torque rpm before it gets a lost phase error. Should be capable of much more than that for at least short durations, but thats a separate issue that i will talk with dmm about to hopefully solve.
Theres 2 other issues that im trying to figure out, most lilely related to eachother. After testing peak power i did a couple finish cuts to see what kind of surfave finish im getting. It was not so great. Not terrible but definitely a chattery look to it. Finish cuts were done with an 8mm 3 flute that i use pretty often. In my current 2.2kw chinese spindle, this endmill gives a beautiful finish and usually has 3 or 4 tenths runout on the flutes. In the new bt30 spindle, run out was even less for these test cuts, about 2 tenths. Also keep in mind that vibration levels are very good, at least what i would consider good. With my hand on the nose of the spindle, i dont think i could feel the difference between running and not running if i had ear plugs in. So why the bad finish if there's no vibration or runout when the same tool gave perfect finish in the 2.2kw spindle? The other issue that i believe is related is an odd occasional noise. To me it sounds like a bearing race letting go now and then but i can be sure. its an intermittent noise that happens for a couple seconds at a time with breaks between and seems to mostly go away after warming up . I Mentioned in my other thread that the top bearing of the spindle had a not so tight fit of the outer race. It would slowly spin in the housing when turning the spindle by hand. I addressed this issue with a wave spring and ring above it to apply pressure to the spring. The spring pushes down on the outer race of that top bearing with about 25 pounds so i doubt that race is slipping. When assembling i put a mark on the outer race and the housing so if i pull it back apart, i will know if it slipped but i very much doubt it. My concern is the 2 preloaded bearings in the nose. Maybe the nut was not torqued properly? Im not sure how i would check this without completely tearing down the whole spindle. The top bearings need to be pulled to expose the nut, but theres no way to pull them. As far as i know, the spindle shaft needs to be pushed out first, i guess with the preloaded nose bearings coming along with it? There is a bolted on nose cap that clamps the outer races of the nose bearings. Im not sure if maybe removing that could reveal anything. I suppose theres 2 possibilities for a slipping race in the nose. Could be an inner race slipping from the nut not being tightened, or could be an outer race slipping from the nose cap not applying pressure to the outer races properly.
Does the noise and bad finish sound like a preload issue? With no runout and no vibration, i dont know what else it could be. Any ideas on what else the problem could be or how i can pin down what that problem is?
Re: Spindle bearing preload issue?
Could the Z axis be a bit loose? If the material was not rigid, the effect would be the same as a loose spindle. Maybe run the test again, but try increasing the feed or stepover to load the tool a bit more.
Re: Spindle bearing preload issue?
I made sure everything was solid when fixing the material. As for the z axis being loose, i would think that would have caused problems before when running the old spindle which was attached to the same z axis. I did try a few different feeds, rpm, stepover. Same result. Keep in mind this is the same tool that i have perfect results with in the cheap chinese spindle, so i dont think it's a feed/speed issue.
The only difference in the test other than the new spindle is the backwards setup. Material on z, spindle on table, but everything is just as rigid. I suppose its possible that could introduce some variable thats affecting finish, but theres still the strange bearing noise. I could move forward with swapping out the z axis, but its a big project. I did a lot more than build a new head. Z travel is increased with new linear bearings, ball screw is getting swapped out. All the parts are ready for the swap, but once i do it, theres no switching back to old spindle without completely undoing the swap. If i do the swap and then determine theres a problem that needs some machining to fix, im stuck. Also the possibility that the spindle goes back to get replaced. Im either waiting a couple months with my machine down, or undoing the swap to go back to old spindle. Really would be best to do the swap and test spindle in its final resting place, but to me it seems pretty unlikely that this issue isn't spindle related and i would prefer to solve it before the swap. If there is no convenient way to rule out a bearing preload issue, then i guess i have to move forward with the swap and see what happens.
Re: Spindle bearing preload issue?
Quote:
Originally Posted by
QuinnSjoblom
I finally finished up the servo driven bt30 project, mostly anyway. The new head assembly is built but i wanted to do some testing before rebuilding the z axis with new head. I bolted the new head down to my machime table horizontally and bolted some aluminum stock to the existing head for doing some test cuts. I had to do some dynamic balancing on the big 80t servo pulley, but the spindle runs smooth as glass at 12k, no balancing needed there. I proceeded to do some test cuts to see if the 1kw servo was performing as expected. It seems to top out at about 1hp at its peak torque rpm before it gets a lost phase error. Should be capable of much more than that for at least short durations, but thats a separate issue that i will talk with dmm about to hopefully solve.
Theres 2 other issues that im trying to figure out, most lilely related to eachother. After testing peak power i did a couple finish cuts to see what kind of surfave finish im getting. It was not so great. Not terrible but definitely a chattery look to it. Finish cuts were done with an 8mm 3 flute that i use pretty often. In my current 2.2kw chinese spindle, this endmill gives a beautiful finish and usually has 3 or 4 tenths runout on the flutes. In the new bt30 spindle, run out was even less for these test cuts, about 2 tenths. Also keep in mind that vibration levels are very good, at least what i would consider good. With my hand on the nose of the spindle, i dont think i could feel the difference between running and not running if i had ear plugs in. So why the bad finish if there's no vibration or runout when the same tool gave perfect finish in the 2.2kw spindle? The other issue that i believe is related is an odd occasional noise. To me it sounds like a bearing race letting go now and then but i can be sure. its an intermittent noise that happens for a couple seconds at a time with breaks between and seems to mostly go away after warming up . I Mentioned in my other thread that the top bearing of the spindle had a not so tight fit of the outer race. It would slowly spin in the housing when turning the spindle by hand. I addressed this issue with a wave spring and ring above it to apply pressure to the spring. The spring pushes down on the outer race of that top bearing with about 25 pounds so i doubt that race is slipping. When assembling i put a mark on the outer race and the housing so if i pull it back apart, i will know if it slipped but i very much doubt it. My concern is the 2 preloaded bearings in the nose. Maybe the nut was not torqued properly? Im not sure how i would check this without completely tearing down the whole spindle. The top bearings need to be pulled to expose the nut, but theres no way to pull them. As far as i know, the spindle shaft needs to be pushed out first, i guess with the preloaded nose bearings coming along with it? There is a bolted on nose cap that clamps the outer races of the nose bearings. Im not sure if maybe removing that could reveal anything. I suppose theres 2 possibilities for a slipping race in the nose. Could be an inner race slipping from the nut not being tightened, or could be an outer race slipping from the nose cap not applying pressure to the outer races properly.
Does the noise and bad finish sound like a preload issue? With no runout and no vibration, i dont know what else it could be. Any ideas on what else the problem could be or how i can pin down what that problem is?
The Bearing noise is most likely the cage is skidding which is common on a start up of a new spindle, this can be a problem until the grease has moved around, to much grease is usually what causes this problem when new like this it pays to run at lower RPM and increase up to the max in steps, this help to disperse the grease
You can check if you have preload or not by putting an indicator on the bottom of the spindle and see if you any movement in the spindle, you apply some pressure at the top / bottom of the spindle to check this
Re: Spindle bearing preload issue?
Ah, that does make sense that the cage could make noise, especially since my shop is only about 60F right now so the grease is probably pretty thick until warming up.
As for end play, i forgot to mention that i did put an indicator on the nose and watched it while activating cylinder (about 1 ton). There was about 4 thou of movement, but i imagine the bearings must have at least some amount of give and maybe 4 thou of movememt with 2000lb of force is ok? Either way i think that reading was probably not very informative since i only applied force in one direction which isn't going to tell me anything about backlash in the bearing stack. When i get home ill take a reading again but apply force in both directions. As long as the indicator returns to zero in both directions, would you consider 4 thou of give with 2000lb ok? Or should it not move that much? When i took that reading i also tried placing the indicator on the fixed part of the spindle housing while actuating the cylinder and there was no movement in that case, so the 4 thou when checking on the nose of the shaft was definitely movememt in the bearings.
Re: Spindle bearing preload issue?
Not sure what bearings you are using, but based on my best guess of the size the load rating should be at least 8000 lbs static. 0.004'' seems like a lot of movement.
Re: Spindle bearing preload issue?
Quote:
Originally Posted by
Jim Dawson
Not sure what bearings you are using, but based on my best guess of the size the load rating should be at least 8000 lbs static. 0.004'' seems like a lot of movement.
Ive been doing some reading and im having a really hard time finding specifics about checking preload. All recommendations are similar, put an indicator on nose and apply pressure back and forth. No mention of how much pressure to apply when checking, or if its to check for actual backlash in the stack, or just give that returns to zero when pressure is removed. In more than one video they were checking for actual backlash in the stack and not using any more than hand pressure to push it back and forth. I will check again tonight and make sure to apply force in both directions to see if there is literally slop, but beyond that im not really sure how exactly this should be done. What i didn't see in any videos is the use of the drawbar actuator as the source of pressure for the test. There would be no way to apply that much force in both directions anyway. Could be that 4 thou of give at 2000lb is irrelevant. If i can get that same movememt at 50 pounds, then im sure its a problem. Ill keep searching and see if there is some more detailed info on this.
What thing that is on my mind is the fact that i asked them for a 14k capable spindle and i originally wondered what they really did differently for this one vs the 10k max version. I can imagine one way to allow a spindle to go faster without burning up is to build it with little to no preload. I hope thats not the case. One thing i did observe was a temp increase while running. Went from 60F to 80F in about 15 minutes so that does give me hope that theres some preload. If there was none, i wouldn't expect much temp increase. Ill see what i can learn tonight.
Re: Spindle bearing preload issue?
Quote:
Originally Posted by
QuinnSjoblom
Ah, that does make sense that the cage could make noise, especially since my shop is only about 60F right now so the grease is probably pretty thick until warming up.
As for end play, i forgot to mention that i did put an indicator on the nose and watched it while activating cylinder (about 1 ton). There was about 4 thou of movement, but i imagine the bearings must have at least some amount of give and maybe 4 thou of movememt with 2000lb of force is ok? Either way i think that reading was probably not very informative since i only applied force in one direction which isn't going to tell me anything about backlash in the bearing stack. When i get home ill take a reading again but apply force in both directions. As long as the indicator returns to zero in both directions, would you consider 4 thou of give with 2000lb ok? Or should it not move that much? When i took that reading i also tried placing the indicator on the fixed part of the spindle housing while actuating the cylinder and there was no movement in that case, so the 4 thou when checking on the nose of the shaft was definitely movememt in the bearings.
Even though the bearings can take a load like this, there should be zero pressure applied to the spindle when doing a tool change, this is easy to have in your design when doing new builds like this
With the pressure of 2000lbs other parts could be moving, put the indicator on to the front bearing retainer plate, and see if that is moving
The Bearings with preload have no clearance, they may have added clearance with your spindle for high speed but if it has .004" that would be way to much
Re: Spindle bearing preload issue?
Quote:
Originally Posted by
mactec54
Even though the bearings can take a load like this, there should be zero pressure applied to the spindle when doing a tool change, this is easy to have in your design when doing new builds like this
With the pressure of 2000lbs other parts could be moving, put the indicator on to the front bearing retainer plate, and see if that is moving
The Bearings with preload have no clearance, they may have added clearance with your spindle for high speed but if it has .004" that would be way to much
Yep, thats what i meant by this:
"When i took that reading i also tried placing the indicator on the fixed part of the spindle housing while actuating the cylinder and there was no movement in that case, so the 4 thou when checking on the nose of the shaft was definitely movememt in the bearings."
As for designing it to not apply the 2000lb to the bearings when tool changing, that was definitely on my mind, but when asking the seller about it, he said its not a problem for the bearings. I can still modify the design if you think its an issue for the bearings, things just get a bit more complex since the cylinder needs to float. This is a gantry setup so i felt much better about rigid mounting the cylinder, but it can still change if necessary.
Re: Spindle bearing preload issue?
So i went ahead and set up a clean area to attempt pulling the spindle apart. Didnt take too much effort. Removed the front end bearing retaining cap and was able to push the shaft out along with all the bearings. Everything seems to be in order as it should be. The 2 nose bearings seemed to have a fair amount of preload. I was not able to spin the outer races separate from eachother by hand. I was also able to take some measurements of the depth of the front bearing bore and compare it to the stack. Stack plus end cap was about 5 thou longer than bore depth so the outer races are definitely being clamped in the housing properly. Safe to say there are definitely no races anywhere free to spin and the noise must be a bearing cage as mactec suggested. I did check end play again and there is 4 thou of give when the cylinder actuates, but i tried also loading it in other direction with a clamp and the indicator always returns to zero. When applying force back and forth by hand as hard as i can, there is absolutely no movement at all. I guess i will just move forward with completing the swap and hope the finish improves
