ad banner
  Reliability Magazine« Message Boards
  Alignment/Balancing/Vibration Forum
  Measuring bearing clearence

Post New Topic  Post A Reply
profile | register | preferences | faq | search

UBBFriend: Email This Page to Someone! next newest topic | next oldest topic
Author Topic:   Measuring bearing clearence
posted 12-17-2003 07:46 PM     Click Here to See the Profile for david_g   Click Here to Email david_g     Edit/Delete Message   Reply w/Quote
I have heard from some (may not be reliable) sources that clearence in a mounted antifriction or sleeve bearing could be checked by lifting the shaft (if possible)and measuring displacement with a dial indicator.

Does this method have any merits? How accurate it is, will it detect normal wear, etc.?

Thanks for the feedback.

[This message has been edited by david_g (edited 12-17-2003).]

IP: Logged

posted 12-17-2003 08:23 PM     Click Here to See the Profile for RHW2ND   Click Here to Email RHW2ND     Edit/Delete Message   Reply w/Quote
You probably know this, but the way we check a new antifriction bearing for installation is to hold the bearing outer, inner raceway in your hand and pass a feeler gage between the outer raceway and roller element until we find the feeler gage that is snug. That is the new bearing clearance we then slide the bearing unto a tapered sleeve (without any load of the rotor or supported with a sling or wood wedges) and tighten the locking nut thereby reducing the clearance to the manufacturers recommended displacement usually 0.001 to 0.002 inches.

I would think that a displacement reading would be just as good as a dial indicator except would you be sure that the actual displacement is the roller clearance or inner raceway to shaft, outer raceway to pillow block machined surface or just hold down hard ware?

IP: Logged

posted 12-17-2003 09:33 PM     Click Here to See the Profile for Ralph2   Click Here to Email Ralph2     Edit/Delete Message   Reply w/Quote
It's a quick and dirty method that works. Looking for "clearance" you are asking how much the shaft can "freely" move. A dial indicator can give an accurate measurement of this.
But, there are caveats.. Often you can not measure the shaft at the bearing and you likely will measure the "lift" where you are lifting. This can cause some errors, the geometry, and if the shaft is bending need to be considered.
Also, consider the possibility that a "vertical" motion is not necessarily in line with the wear. And, by not visually inspecting the bearing you may be missing a lot. A bearing that is in the process of failing may actually have less clearance than one that is good
Good luck

[This message has been edited by Ralph2 (edited 12-17-2003).]

IP: Logged

posted 12-18-2003 06:14 AM     Click Here to See the Profile for Eben   Click Here to Email Eben     Edit/Delete Message   Reply w/Quote
The common "lift test" also does not show you if the clearnace is in the bearing or housing.

I quick way to check is to measure the clearance and then pop the bearing cover and insert a piece of paper or thim shim between the outerrace and housing. Then replace cover and redo the lift test. If the lift test shows the same number the clearance is internal to the bearing if the indiactor reads lower then the clearance is in the housing.

This method of poping the cover and installing the shim also gives you chance to inspect the bearing and locking system.

I hope this helps,


IP: Logged

posted 12-18-2003 11:32 PM     Click Here to See the Profile for Rusty   Click Here to Email Rusty     Edit/Delete Message   Reply w/Quote
A "lift check" works extremely well and is very accurate, if done properly. The secret is to use a magnetic base and mount the base on the bearing housing, with the indicator point on top of the shaft. Then when you lift the shaft, even if the base flexs and the housing moves, you will still be measuring the shaft movement relative to the housing, not the baseplate, the ground, etc. (I suggest an Anyform magnetic base that will adjust to curved housings.)

What this tells you is the "total clearance" which will include shaft-to-inner race, outer race-to-housing, bearing wear clearance.

On really large machines you can -- very carefully -- used a porta-power to lift the shaft. Be aware you can damage the bearing this way since it's rather easy to put 10 tons of pressure on the shaft.

I prefer to just use a long prybar and a suitable stack of metal to position the end of the bar under the shaft. This will give you more than enough leverage to lift all but the largest shafts, and you have good "tactile" feedback. You can tell when you've taken up all the clearance, and you can tell if it's "bumping" within the bearing (not good).

Just remember: the indicator base should always be mounted on the bearing housing!

re: Jmarc's comments below..... He's right.....A "lift check" is definitely not adequate to determine if a tapered adapter bearing is set properly but I don't think david_g was suggesting that.

A lift check is only a tool to determine if there is a "gross" clearance problem -- I only use this when the vibration spectra shows gross looseness (multiple harmonics, 1/2x harmonics). Sometimes in double-row roller bearings you'll see gross looseness with only an extra thousandths or two clearance. If that's the case, then I am not going to change that bearings because it may run for years that way. I use a lift check to determine the severity of the problem only.

And he's right about the inadequacy of setting the bearings with a feeler gauge. You can do your very best, but it seems the difference between "just right" and "too tight" (usually "too loose") is often 0.001" or less. I don't think it's the measurement tool that's the problem though, but rather all the other variables involved in the installation. Banging on the lock nut with a drift and hammer certainly doesn't help. Instead, use a spanner wrench, but add at least a 30 inch cheater bar to it. The extra leverage makes all the difference in the world, and you can then actually set the clearance using the spanner wrench (not going to happen unless you use the cheater). Of course, the instructions that come with the bearing don't tell you this, because they'd sell fewer bearings if they were set up properly, and they also want you to buy their expensive hydraulic nut setter. (These companies are not run by the knowledgeable sales engineers and service guys that most of you know and trust -- they are run by marketing people who are just "used car salesmen" with degrees)

Any bearing guys out there? The biggest problem is the "square" notchs in the locknut -- the spanner won't hold. Why don't you use a dovetail notch instead - a hook spanner would then hold very well.

[This message has been edited by Rusty (edited 12-19-2003).]

IP: Logged

Junior Member
posted 12-19-2003 08:15 AM     Click Here to See the Profile for jmarc   Click Here to Email jmarc     Edit/Delete Message   Reply w/Quote
It is an area where we are still in the stone age. Everytime I adjust a bearing with a feeler gage make me remember many years ago when using the feeler gage was the way to perform coupling alignment... Some bearings companies suggest to adjust bearings by measuring the sleeve displacement. That method would be accurate but not often possible to perform. I never tried with the dial gage but will test it the next time. I think there is a need to develop something better to permit precision and repeatable adjustment.

IP: Logged

posted 12-19-2003 03:17 PM     Click Here to See the Profile for DavidM   Click Here to Email DavidM     Edit/Delete Message   Reply w/Quote
There are a couple of problems inherent with shaft lift and feeler gauges. On a double row bearing such as a spherical roller bearing, it is possible (and probable) that as the bearing is mounted it will force the rollers to one side or the other. Most mechanics measure the clearance on one side only. This is especially the case with large machines like dryer bearings. Also, as the bearing is mounted, the rollers will get cocked in the cage pocket. So there is no guarantee that the rollers are positioned properly so that as the load is applied, either up or down, the rollers seat optimally in the race. For these reasons I recommend that the shaft be turned before making clearance measurements. The geometry of the rollers is self-guiding. CAUTION- Do not turn the loaded shaft with out lubrication. On a large bearing, I saw rolling the shaft 1/4 turn make a .010 inch difference in measured clearance. If possible, measure clearance on both rows of rollers and average.

The most accurate method of mounting is the axial drive-up method. Think about it-most tapers are 1:12 slope. A .012 inch axial movement results in .002 diameter reduction in clearance. If you overshoot a bit axially, there is actually a very small error.

[This message has been edited by DavidM (edited 12-19-2003).]

IP: Logged

posted 12-20-2003 10:42 AM     Click Here to See the Profile for Arne.Lindholm   Click Here to Email Arne.Lindholm     Edit/Delete Message   Reply w/Quote
Just adding to the tactile feel of Rusty:
Using a spring scale - fishermans scale, you can add a force and note the deflection in steps. This way, it is easy to detect when rolls really get in touch and if you note the lever lengths used, it is not possible to overload.
I use the scale also for meauring the static stiffness of various items having a resonance. Tensioning in small steps yields a series of points (force vs deflection) and the stiffness is the slope of a good line adjusted to pass the most linear part of the points. A good feedback to calculations. /Arne

IP: Logged

All times are ET (US)

next newest topic | next oldest topic

Administrative Options: Close Topic | Archive/Move | Delete Topic
Post New Topic  Post A Reply
Hop to:

Contact Us | Reliability Magazine« Home

Except for message content, this material is Copyrightę 2000 by Industrial Communications, Inc. All rights reserved

Ultimate Bulletin Board 5.45b