Lathe spindle bearing selection

I have decided to replace the spindle bearings on my Chester DB10 due to surface texture issues. Not before time either, one of the inner races is cracked.

The bearings are NBGE 32007 P6.

A supplier local to me has a selection to choose from, the link below is to their page.

Is any one of these more suitable than the others for use on a lathe?

Also what grease should I use? No oil feed to the bearings on this lathe.

https://www.engineersmate.com/search.aspx?s=32007

Thank you in advance

I bought SKF and Timken bearings for the Tom Senior R8 spindle and they are excellent, and I would go for a big name brand such as Toyo, FAG, SKF, Timken, NSK, NTN and similar. As for grease, a quality lithium grease for car wheel bearings from Castrol, Mobil, Carlube, Shell, from any car parts shop.

When packing the headstock, be sure to leave 50% airspace, overpacking will lead to churning and overheating. Not too much preload when adjusting, the headstock should not get over about luke warm after 15 minutes running at speed. Back off the preload slightly and wait for things to cool before retesting if it gets too hot.

Edited By old mart on 12/12/2022 18:46:31

None on your website mention P6 bearing tolerance, but seems plenty elswhere. You could also obtain correct grease from the bearing supplier chosen.

Old mart,

Thank you for the Information very useful, I can work with that.

What is P6 referring to?

While the budget ones are half the price, it's still only a tenner difference between those and any of the branded(SKF, Timken, FAG, etc) ones. I'd buy whichever of those they actually have in stock.

Ketan of ArcEuroTrade will advise you from a font of great knowledge. Just ask.

I've had FAGs in my Warco 280 for at least 10years

He also has a good reduction on 32007 bearings

According to SKF P6 is closer tolerance than their standard bearings and P5 is closer than P6. **LINK**

Brian G

You cracked inner race suggests that the inner race may be too tight on the spindle, and that can be a leading cause of adjustment difficulty leading to surface finish issues too. If the inner race is too tight it wil not slide suitably enough to let you adjust the bearing clearances, inevitably resulting in either too loose or too tight a clearance.

See here for description of P5 and P6 tolerances.

https://www.skf.com/uk/products/rolling-bearings/principles-of-rolling-bearing-selection/general-bearing-knowledge/tolerances

Using the wrong tolerance might contribute to your race cracking if it was at small end of normal tolerance.

Robert G8RPI.

+1 on go for one of the better brand names for something like this. IE SKF, FAG, Timken, etc

Cheap no-name or obscure brands are a false economy.

Definitely track down a set of the P6 close tolerance bearings if that is what it used originally. There would be a reason for using them in the first place.

Use big brand names as stated, they are a known quality, not only dimensional but also material specification. Use at least the dimensional class to suit the purpose, plenty of information available on manufacturers sites. Follow advised fitting method and lubricate accordingly, if grease packing follow recommendations given for type and volume, too much is as bad as too little, overpacking will cause heat build up. If you decide to make design changes for sealing or mounting again follow design requirements suggested.

I replaced the spindle bearings in my HM50 mill head. The new bearings were the correct sizes, but the shaft luckily was oversized and needed a little over 0.03mm to get to the nominal diameter, and then a little more off to allow it to slide with the pressure of the wavy bearing rings. Keep in mind that if the spindle is not round, the inner race will be distorted to that non round shape, same for the housing the bearing sits in, also needs to be round. Otherwise the better quality bearing becomes waisted money and can fail earlier than a standard bearing of less concentricity and roundness. There is a white looking lithium grease for spindle bearings, and is worth getting some.

Neil

SKF don't seem to list a "P6" grade for metric taper roller bearings - they have "Normal" "CL7C" and "P5". Brian's link shows this - references to P6 are in the "except taper roller bearings" heading.

More important than boundary dimension tolerance tightening between grades is the running accuracy - for a 35mm bore bearing (32007) the radial runout limit for a Normal class is 20 microns, CL7N class is 10 microns & P5 is 6 microns. Bore tolerance is the same on Normal & CL7C classes (0 to -12 microns) and only 2 microns tighter on a P6 ( 0 to -10 microns). Bearing runout directly affects the roundness of the parts you make on the lathe, so a better class of bearing should result in rounder parts. In reality, with today's manufacturing techniques, you would be pretty unlucky to get a bearing from a reputable manufacturer that was close to the top limit - a "Normal" grade SKF bearing may well run truer than a "no-name" maker's P6.

Tightness of fit is a bit more difficult to call, but pretty well all of the industrial sized lathe spindle bearings I was involved in changing were an interference fit, requiring the use of bearing heaters to fit the replacement inner races to the spindles (and a fight to get the old ones off ! ). Bearings fitted loosely run the risk of spinning on the shaft / in the housing in service. While rebuilding machine tools at my last-but-one employment spindles regularly had to be sent out for worn bearing seats to be rectified (ground true, hard chromed oversize, ground back to size) due to wear from a bearing that had spun.

Normal lithium wheel bearing grease will be fine - not recommended to use grease with additives like molybdenum disulphide or other additives in rolling element bearings. For a light hobby lathe I doubt there would be any advantage to using an expensive "special" grease like the Klueber long-life, high speed products. Overpacking is less of an issue with taper roller bearings than ball beariings, as any excessive grease can easily escape. I have a recollection of reading in some bearing literature that TRBs actively pump lubricant through the bearing due to the difference in diameter along the length of the rollers ( but no longer have access to that data library to be able to quote the actual source) whereas ball bearings churn the lubricant within the bearing & generate heat..

Nigel B.

If your DB10 has been designed and sold to you for general purpose use in a hobby workshop however long ago it was purchased, then it would be fine to use a general purpose 32007, branded or otherwise, from reputed U.K. distributors.

If it was sold for general purpose use, then the assembly process for the spindle would have followed general factory assembly procedures. Here the factory would have used whatever TRBs were available - Standard or higher precision. They could have been general purpose 32007, or bearings marked with higher precision which may have failed the higher precision requirements - but fine for general purpose use, or surplus.

If your spindle/lathe was specifically specified for tool-room use, the spindle would have been assembled in a more precise way to ensure that the higher precision of the bearings being used retained their precision attributes during the assembly process. If assembled in the general assembly way, then the precision attributes are more likely to be lost.

Based on the above factors, I would guess that it is more probable that this was a general assembly produced spindle/machine. If so, in my opinion general precision branded or unbranded 32007 - either of which procured from a known U.K. supplier would be fine.

1. Possible causes of the inner race being cracked could be assembly issue eg:

- the shaft might have been too tight

- the installation force/process may have put too much stress/shock on the bearing

- if there was some dirt/shaft between the inner ring and the shoulder on the shaft against which the bearing may sit, then the inner ring may not be sitting correctly on the shaft. This could be ascertained by looking at the wear on the inner ring.

2. Another cause could be abnormal shock on the bearing during use of the lathe, or too much pre-load combination, but this would have had to been really bad... and extremely rare in such an application.

3. It is possible to crack inner rings during dismantling.

If you found a higher precision bearing, and if you wanted to have such bearings installed, you would also need the skill, equipment and ability to instal them correctly onto the spindle, or you would need to ask a company which would have such equipment, ability and skill to carry out such installation for you - correctly. Otherwise, the higher precision attributes of the said bearings would be lost if you were to do it yourself in a hobby workshop environment.

As for the grease, I would suggest general purpose NLGI2 LM bearing grease - approximately 20~25% internal volume (guess). Over packing will only result in excess heat generation unless the grease has extra area for it to escape. Moly grease sticks to the rolling elements like s**t to a shovel, and less recommended by most bearing manufacturers and dealers of old. However, there are people who swear by it. If you decided to use it, use far lower quantities than normal grease, as because of its properties it may not allow the whole assembly to bed in properly (although makers of Moly would disagree). Compromise could be 20% normal grease with 5% Moly (well mixed before introducing), but then again, that is open to debate. Whichever you decide, you should continue with the same idea at every maintenance. Moly extends time between maintenance.

If you had 'finish' issues with your work piece, did you consider increasing the pre-load?. After your installation of new bearings and applying the pre-load to achieve the finish you expect, take a sample piece of the finish and store it with date showing when you did the sample. It takes 'x' period of time for the bearings to bed in. At that time, you may see 'finish' issues with your workpiece, so you will need to check, adjust, add pre-load to the TRBs.

Ketan at ARC

Edited By Ketan Swali on 13/12/2022 13:33:38

I was told that molybdenum disulphide chemically changes the surface of steel, lowering its coefficient of friction. Fine with plain bearings, but not so good with rolling elements as the surface can be made brittle by the changes.

Thank you all for your responses masses of information to go on there.

This is the position I am now in, with a request for some additional assistance please.

I have now cleaned down the whole assembly and it appears to me that the bearing that has failed (it is the one nearest the chuck) has been running hot as it has a brown tinge to it like heat has been applied and little to no lubricant compared to the bearing at the other end.

Before attempting the next procedure I checked and adjusted the slides on the saddle, cross slide, compound slide, tool post until I was happy that no play existed and the machine was still useable (not stiff to turn the handles)

I did attempt to adjust the preload of the bearings before dismantling what I did was fix in the chuck a Dia 20mm solid carbide bar aprox 300mm long (internal boring bar) with a DTI gauge measure the amount I was able to deflect (I am believing the carbide bar will not flex under finger pressure).

The amount I was able deflect the bar measured aprox 240mm from the chuck jaws was + 0.125mm - 0.125mm with not to much force from me. The gauge always returned to zero as I removed the pressure from the carbide bar. I also tried to push and pull the spindle in the Z axis with a total movement in this direction of 0.03mm. Slackening the locking grub screw on the spindle locking ring and using a C spanner I made incremental tightening of the spindle. Checking the end float and the ability to deflect the carbide bar. The end float reduced to zero but I was still able to deflect the carbide bar + 0.08mm - 0.08mm. At this point whilst rotating the chuck I could feel it was notchy so decide4d there was an issue with the bearings and went no further, the plan was to replace the bearings. If I am honest I have never been happy with the machine always chasing size and surface texture on the more precise workpieces.

The bearings on the machine as purchased are NBGE 32007 . P6. I do not know the quality of these bearings but were unable to find a supplier locally.

I have purchased a FAG32007-XDY-XL bearing and the lubricant multi purpose lithium high temperature grease. I am confident in reassembling the spindle and adjusting the preload.

With the lathe spindle removed and all cleaned down the front cover of the headstock removed I was able to see two of the bolts securing the headstock to the lathe bed. With an allen key in hand I decided to check the tightness of these two M8 bolts. The result was they were not at all tight I would describe them a nipped up very little pressure required to tighten them further. At this point I removed one and found it came out with only a few turns. placing the bolt back in the hole without any screwing in the gap between the head of the cap screw and the casting was 5mm. So only 5mm of thread would be engaged when fully tightened, I checked the depth of the threaded part in the bed and found that the M8x 25mm set screw could be replaced with a bolt of up to 40mm in length.

So now my question is should I change the bolts for longer ones and I am thinking M8X35mm one at a time so the head is never loose. I am unable to get a torque wrench in so its by feel tight of very tight.

This was probably the main reason for the poor performance of the lathe but a cracked bearing needed to be rectified.

If you are sure the M8 SHCS are intended to secure the headstock to the bed, then it would be a good idea to use longer ones, I would not be happy with less than 3 diameters of thread engaged in cast iron. You will likely have to realign the head, but that can come later. Do you have the manual and parts list for the lathe? The FAG bearings will be fine for a hobby machine, all the quality makes exceed their specifications.