We chatted a couple months ago on A/C bearings for an upper drive unit in a knee mill. We have since balanced the components and today finally installed the A/C bearings in the drive unit. They are making noise!! We lubed with 30% NBU15 Kluber, the configuration is face to face with 1in spacers between bearings.
Work was done in a meticulous fashion but I know that may not be enough. The bearings are quiet but have an intermitent rattle or ticking sound. I am wondering if this might be ball skidding or FOD? Are these bearings toast with only 1min run time? The sound is very confusing and we are stressed to say the least on this. Can this be a manufacturer defect? Thanks for all your help.
Were they precision bearings, was the lube contaminated, are they exactly parallel, was the preload on the bearings set properly, are they within the speed range of the bearings? So many possibilities with so few answers.
Just because they are new bearings does not necessarily mean that they are without defect. There is a possibility that with 1 minute of run time, they have suffered no damage.
Steve
Is the noise at any speed or one certain speed. I am guessing this is a low speed noise but what I have been reading that to run high RPM is has to run in stages until the rpm limit is reached, sort of a breakin kinda thing. I dont know much about these high quality bearings & setting them up. But this goes to show as many have warned about its not as easy as it seems to accomplish something like this. I am sure you went to extremes to install these correctly but it may be beyound that to do them right.
But the reason I ask about the speed & if they have been run high RPM is it may have damaged them to spin them up to high limilts before they where broke in & then the noise is a result of this. But like i said I dont know enough to evn begain to anwser your question. Just kinda thinking out lound, maybe it will help in some way.
Grease is brand new from Kluber so contamination would be very unlikely. Preload is light and determined universal grind on the bearings. The spacers are equal width so no additional preload is added with the spacers. Spindle was started at 1K RPM and noise was obvious. PUshed to 3K for a few seconds and noise is still there. Bearings are good for 10K in grease.
ONe thought I have is that the pinch nut being about 2.5in in diameter, is not being pulled down hard enough to set the preload properly causing skating. I know skating is just as bad as FOD so will not run them until I know what is going on.
A minor bit of damage to the bearings as a result of excessive force will EASILY brinnell a raceway. If you, by chance, "popped" the A/C's appart with force while handling them, you DEFINITELY brinnelled the shoulder and probably one or more of the balls in the process.
The ONLY way to disassemble an A/C is to heat the outer ring and let the inner ring drop out (into a pie pan of course so as not to lose stuff). To reassemble, you cold fit it and do essentially the opposite and support the outer and heat it and let the assembled inner and cage and balls "drop" into the outer ring once you heat it.
Again, if you used ANY force to disassemble the bearing to grease or clean it, you most likely brinnelled the balls and the raceway shoulder and the bearing is RUINED.
Some greases are notoriously noisey due to the thickening agent(s) use. I'm not sure of what Klube uses but I don't recall it being a "noisey" grease. Some greases that use sodium nitrate to thicken the grease actually feel/sound like marbles are being shaken in a coffe can.
More likely than anything, I'd suspect that an "OOPS" took place and you've gotten a flat spot generated that shouldn't have been someplace on the balls and/or raceway.
A new bearing might hum or perhaps whine - if it has a steel cage, perhaps you might hear a chinking/rattlying sound. But a repetitive rattle/noise is usually the sign of assembly damage or debris or debris dents or both....
These are new bearings and they were not separated per my last mistake. I used heating and cooling methods to install them and they went together with no effort. the rattle sound comes and goes but obvious at all RPMs. I am wondering if I actually somehow got debris in the bearings, if they are toast at this point or can we clean them and try again? I know it takes very little to damage them so I have been very careful.
When the assy is assembled, I froze the whole thing to install in the chassis but it still to a couple taps with a plastic hammer to get it to drop into place. Minimal effort and I would not think that would be enough to damage them but you better than I. thanks again.
Viper,
My work is vibration analysis and balancing of rotating equipment. Without hearing or looking at the spectrum of the noise it's difficult to say what is wrong. Different bearing defects have different noise signatures but do indicate what the problem is. There are defect frequency "sets" that are dependant on inner race dia, rolling element dia, outer race dia, etc. SKF has a calculator that will tell you what frequencies to look for in a given bearing. http://www.skf.com/skf/productcatalo...=&action=Calc6
You said there was an intermitant rattle or tick sound, The good news is that intermittant sounds are not generally bearing defects. Excuse my ignorance here, is this mill belt or gear driven ? Second question, is the noise actually coming from the bearings you replaced ? Try one of those automotive stethiscopes to locate the exact source.
The sound is definently intermitant and random and what is really funny is we heard a very similar noise before we tore the unit down. It is belt driven. I removed the A/C bearings (they are brand new), cleaned them, inspected for obvious defect, rolled them dry while listening with my stethescope, and reinstalled. They sounded great. I am at a huge loss for what is happening. I cannot get right to the side of the bearing when installed but I can guarantee it is this drive coupler because we can run just the motor with no noise and the spindle is not installed. Very simple setup here.
This is a noise I am not familiar with. Almost a chatter sound and it gets worse as the bearings warm up. One thing I did not check for but might actually sound like this is if one of the outer bearing races is somehow spinning. Seems extremely unlikely since is does not just drop into place but I have to look at everything. I think the goal today is to pull the assy and try to bench test it. Obviously, this may tell is alot in what is really going on.
put a small mark on bearing outer race and on housing when assembled, then check after run. if it's spinning you'll know. Check those pullys, differing belt tension through rotation can do funny things.
Steve
Timing belt drives sometimes have lead or "pull" in them. Thus, they can generate an axial load to themselves and/or the pulleys they're running against. They can also wander from side to side in the sprockets. When the belts hit the pulley flanges, funny noises can be evoked.
Chances are that the A/C's may not and probably are not damaged - I suspect you learned your lesson the first time on what NOT to do with/to them. The A/C's may NOT be emitting the noise but the bearing on the othe end (regular deep groove, probably with metal cage), may be rattling the cage (this can come and go depenind on the speed and acceleration patern of the spindle and/or any bearing misalignment.
When I say "misalignment", it doesn't take much (a few minutes) to cause the balls to travel from one side of the raceway to another. This causes a speeding up and slowing down of the ball and the cage dut to the instantaneous radius change that the ball paths are making. This begets a chucka-chucka or tink-tink sound as the balls and cages bounce against each other. This can come and go.
Like vger says, there are different signatures that are/can be emitted, and each depends on where the problem exists and how bad it is and how many problems there are lurking. You'll have to systematically "find" the offending member.
Don't be surprised if the problem is elsewheres. By fixing the major problem with the A/C's, you may simply have uncovered another problem that was below the noise threshold created by the prior A/C problem that is now fixed.
Welcome to world of unintended consequences and the never ending perils of progressively fixing used bearings in resurrected machine tools. It eventually ends when you get all the noisey stuff fixed and can only hear the 60 cycle hum of the electric motors running.
When/if you do get all the bearing issues fixed, you can get them this quiet. It is scarey when that happens because even the slightest noise emitted startst to make you think of nothing but spending more money on expen$ive bearings.
You are right on NC Cams, this thing is getting so quiet now that any small noise is heard. I will have to try and figure our how to put a picture on here so you guys can see what I am working with. In short, the way it is running right now, The only radial ball bearings that run are the two in the spindle motor. The drive coupler sits afront coupled through the ribbed belt.
Yes, I would say that the noise kind of sounds like ball bearings slapping each other or something similar. The reason the noise gets worse with heat is because of the change in grease viscosity I am thinking. OK, I will try to get a pick or two so you better understand this deal.
OK, I think I have some pics uploaded. Bad news, the noise IS coming from the A/Cs. Really sounds like they are not preloaded or the balls are dancing around in there. They feel a bit rough when you first change dirrection and then smooth up. I manualled spun this assy in the mill with a drill motor and upon coasting, it makes noise until about down to 50rpm and then smooth as silk. The phenolic cages are VERY free in the bearings but I am assuming that is normal.
As you look at the pic, the left side or hub side is the up side in a verticle mount. It couples to a pulley. The lower or right side has splines on the inside for the spindle to move in and out of. This assy has been run totally by itself and I know it is the A/Cs. I am wondering if FOD was an issue, would they not whine first? In my experience with bearings, it would seem the FOD would damage the balls and their rough tecture would cause noise eminating as a whine, not a tick. Tell me where to go. These bearings came from Germany so who knows how they were shipped. IA, maybe dropped, I don't know but getting replacements will be like pulling teeth I am sure.
Based on the foto, the bearings MUST be mounted DB in order to attain preload when you tighthen the nut into position.
As you tighten up the nut, you should feel some drag develop on the othewise free spinning bearings but this should remain fairly constant once the clearance is taken up.
Example: We have some 500 lb preloaded ball screw bearings and they feel snug yet spin glass/silky smooth when rotated under preload.
Make sure the raceways (inner and outer) are NOT dragging on the shoulders of the spacers. You might have to chamfer the spacers a bit. The spacers should fit the new bearings as-is but you never know. Make sure a cage isn't protruding and dragging someplace - it could happen.
If the bearings feel smooth when rotated and have a drag or noise ONLY when assembled, I'd look for interference. Recall that cages are usually ball guided and they can move around slightly. See if they'r dragging on the spacers.
I've seen cases where guys have inadvertantly loaded the bearings 'backwards' (I did it myself so I KNOW it is possible so don't even try to think that it isn't).
When you do this, it is super easy to cut (scored circumferentiall) a or the balls by shoving them against the wrong side shoulder of the raceway. The shoulder is there to keep the bearings assembled - not at all to accept load.
Result: cut ball or balls with brinnelled shoulder. Sounds somewhat like baseball card in the spokes of bicyle. The only fix is to replace the bearing.
Sadly, this could be turning into a very expen$ive learning experience in how NOT to replace bearings.
However, this will illustrate to the hard to convince just how damn easy it is to screw up a bearing when you install them.
FYI: on this spindle, the bearing against the shoulder could be lightly pressed up against the shoulder. However, if you freeze shaft and warm up the bearing on light bulb or hot plate, it could should literally DROP on with no need for ANY hammer work (plastic, feather or otherwise).
In light of the nut however, I'd look for both bearing inner rings to slip fit/no slop. Use scotcbrite to dust the shaft and use care/time and patience.
The spacers too should drop/slip on.
HOWEVER, The bearing retained by the nut had better slip on (as in slip fit/NO slop). Ultimately, both inner rings will be retained by the nut against rotating or axial motion. Besdies, the inner rings HAVE to to slip in order to allow preload to develop (think about it for a while).
Again NO hammer should EVER be needed!!!!.
The nut should be tightened JUSt enough to draw the rings so as no preload offset remains. It does not hurt to turn the bearing slowly while tightening the nut manually.
I've seen cases where guys run up the nut and brinnell the balls simply while running the nut home (with an impact of course, which actually should remain in the frigging drawer NEXT TO ANY HAMMERS you may be thinking about NOT using).
Trust me when I say it, when you use the thermal fit method, the bearings on a machine tool spindle willl LITERALLY FALL TOGETHER. For that matter, the thermal method also works the same way for monstrous railway locomotive bearings.
I literally saw a guy at a diesel locomotive shop heat a housing with a rosebud torch, then pull the outer ring of a bearing out of a pizza freezer they used to chill them (and their lunches).
He then damn near threw the ring across the table (like throwing a playing card into a hat) and it fell into the hole. They then gave it a quick twist to assure square seating before it cooled and it never moved again.....
The bearing suported the armature of a monstrous locomotive armature bearing for the next 250K miles. And they didn't have a hammer anywhere's near the operation, either.
It will be nice to know, however, that once you get thru this, you'll be ANOTHER spindle bearing replacement expert. All I ask is that you pass the knowledge on some time down the road. No need to let some future enthusiast have to learn how to reinvent the wheel.
Well I will say that the bearings and the shaft are NOT a slip fit without heating and cooling. The bearings on the spindle shaft are and it is a sweet install. I will admit that in order to tighted the large spanner nut, we used a spanner wrench and a hammer. Also, upon removal of the bearings from the shaft, there is really no way to remove them without a hammer. They are tight on the shaft. I am able to gain access to the inner race through the bolt holes and it is a time consuming process but if they are screwed anyway, it must be done. Now I am VERY careful in the removal and did not even scratch the race with my punch.
My questions are could we just not have enough preload set in yet by tightening the nut? Do you have any idea how tight this nut should even be? I am basing my tension from the original setup as I scribed the nut before removal.
I have done plenty of differentials with taper bearings and crush sleeves but a differential this is not. What I would like to know is if the torsional force can be checked to verify the proper preload is set in? Can you overload the bearings in this setep? My understanding is the grinding of the races makes overloading much harder.
Per your last reply, are you saying I should scotchbrite the shaft until the bearings are a slip fit?
By the way, there is a slight noise like slack when manually changing direction of rotation of the bearings. I am kinding of wondering if this is a lack of preload I am hearing. I think just how tight the nut should be is the big question you might be able to help with. thanks again.
Just looking at that assembly, I'm thinking you should be able to directly measure to ensure that the nut is pulling the inner races against the inner spacer ??
That is if you recorded the height of the inner races and the shoulders on the shaft ?? the thickness of the nut and the lock ring are easy because they can come back off to be measured.
Hum, this sounds like maybe the cage has been distorted. It being a high quality bearing it would only take a very,very little of distortion to cause this.
In your case, and assuming the beaings are mounted DB, The outer rings should touch each other first and then, as the inner rings are tightened and preload fully applied, only then will the inner rings fully and properly touch.
Thus, if you measure the width of the two inner rings and the spacer, when the assembled stack height of the inner rings is achieved, you've got the designed prleload. The trick is to somehow being able to measure the assembled stack dimesion.
Regarding the slip versus press fit, it is going to be pretty hard to achieve the desired preload when you're trying to overcome the press fit on the shaft - you'll never figure out how much is being applied to preload versus how much is overcoming friction. How the OEM got the preload is beyond me - probably by guess/golly and/or experimentation and a bit of luck.
This is why a slip, almost drag fit is what I'd be looking for between the inner rings and the shaft. THus, as/when you tighten the nut, you WILL apply preload properly when the the nut is tight.
If you want to figure out the amount of torque needed, that is going to take some figuring. Basically, you're going to need to do a torque-turn method. The trick is to figure out how many turns, once contact is made, to continue to tighten them so as to apply preload.
Essentially, the amount of preload in the bearings will determine how many turns of the nut is required. To figure this out, you need to know the pitch of the threads, the cross sectional areas of the spindle and the amount of preload in the bearings.
YOu then do some figuring via minipuplation of the following information:
First, figure out how much offset deflection is present in the offset rings.
Once you know the offset of each ring, you then find the set total. This is how much compression that has to be generated by the nut.
Let's say that there is 0.005" deflection per bearing (purely a number, you'll have to measure). Double this and you have to compress a total of 0.010".
At a pitch of say 40 threads per inch, you'll effectively move the nut 0.025". Thus, if you tighten the nut an additional 1/2 turn, you'll move it about 0.0125" (probably close enough to 0.010", especially since you'll have some nut friction to overcome as well)
Again, this assumes that the inner rings slip fit on the shaft. If the inner rings are pressed, you're guess is as good as mine as to how tight is tight enough.
Once the preload is applied, you should feel some drag increase. If you don't, you probably haven't preloaded properly if at all.
A rule of thumb is that you need roughly 0.002" to 0.004" of compression on the inner rings to adequately compress the preload and hold the parts tight. This is in ADDITION to the theortetical torque turn which is why I gave it a half turn instead of a tiny bit less (see above calc).
I'd argue that a press fit ID inner ring in this case could be preventing you from attaining the proper preload for reasons already cited. If it were my spindle, YES, would go to a very precise slip fit (h4 or h5 actually).
Yep, I am kind of leaning on lack of preload because I just think the sound I hear is unorganized chaos. The hope is that I have not damaged anything. I am thinking this might be why they were failing in the first place. Incorrect fit and thus a lack of preload.
NC, I absolutely agree with you about NO HAMMERS ! When I have a shaft that is just a touch to large to slip fit, I'll cut a strip of steel shim and put some 2000 grit under it wrapped halfway around the truning shaft. That way I get even pressure across the surface of the shaft. I also agree that the inner race does need to slip to get the preload right. A method I've used for press fit is to put the bearing in a hot oil bath and pack the shaft in dry ice, let em both soak for a while then QUICKLY slip the bearing over the shaft and get it in place just as fast as possible.
Viper, I'm not sure how you may be able to get this back apart without damage short of building some special tool to press/pull on the inner race. I was successfull once by drilling a hole that was just smaller than the outer diameter of the inner race, countersinking it on the underside to clear the shoulder, slicing the thing in half so I could get it in place setting on a steel sleeve, and then pressing it. Not sure you would have enough room to do that. Do you have the specs on this bearing set ? Or can you get them ? Even in german ?
NC Cams another bearing question
Posting Permissions
Reply with Quote