[nitewatchman]

The original spindle in my 1100 has developed a definite bearing noise and knock in the spindle. That's okay because it has made a pile of chips about as big as it is.

Some time back I aquired a used spindle to keep as a spare to use when this event occured. The future "spare" had a VERY noisy bearing that seemed to be at the top near the pulley. I assumed it was the result of the drive belt side load. When my original spindle started knocking I grabbed a set of bearings and pushed the "spare" apart.

Encountered a bit of a surprise. The bad bearing was actually on the taper end and the outside bearing of the pair. It appears that coolant had completely washed the grease out of the bearing leading to the failure. Both the bearing and the lower part of the spindle showed corrosion. While there is a steel bearing guard flange pressed below the lower bearing, there is no seal of any type and coolant can freely enter the bottom end of the spindle especially if it is being splashed or directed upward. You might want to consider this when placing coolant hoses. Even without corrosion or staining coolant is not a very good high pressure lubricant.

The spare is setup and ready to go in this weekend when we change over parts in the fixture. There is also a set of bearings waiting to go in the original spindle so it now becomes the spare. It will be interesting to see what caused the failure when this one is pushed apart.

nitewatchman

[BAMCNC.COM]

Good info, do you have bearing #s so we could buy spares?

[tbaker2500]

In their document SB0024, they quote:
"Upper bearings are PN 30316 and are provided as a matched set. Lower bearings are PN 30303 and are also provided as a matched set. You need to order both part numbers to get a full set of 4 bearings."

[nitewatchman]

In their document SB0024, they quote:
"Upper bearings are PN 30316 and are provided as a matched set. Lower bearings are PN 30303 and are also provided as a matched set. You need to order both part numbers to get a full set of 4 bearings."

Upper Bearing is a 7007C/DT P5, Lower Bearing is a 7008C/DT P5. Best deal going is from Tormach but they are Chinese Bearings.

nitewatchman

[GJeff]

For people using high pressure coolant what do you think the best way to prevent this would be? Is it possible to put some sort of guard up?

[nitewatchman]

Mine pressed fairly easily, used a 25 ton HF press. I don't believe that my fits were that tight. Just make very sure they are started straight.


nitewatchman

[anglin]

Mine pressed fairly easily, used a 25 ton HF press.

Out of curiousity, how did you ensure you were pressing only on the outer ring of the bearing and not on the inner ring?

[nitewatchman]

Out of curiousity, how did you ensure you were pressing only on the outer ring of the bearing and not on the inner ring?

Depending on what was being pressed a short tube to contact the surface being pushed. The press fits were really light, like pulling the short jack handle with one finger type light. If you don't have a press, I think it would be easy to lightly tap the bearing into place.

As a matter of a fact, the old bearings fell apart when pressing the spindle out of the housing and the outer races were tapped out with a piece of 1/2" key steel and a copper hammer. Same with reassembly, the lower bearing and spindle could be pressed home but the upper could only be pressed to within 1/8" or so without having a longer tube than I had. The final 1/8" was lightly tapped into place until seated.

Spindle seems to be doing okay. After taking up the last preload adjustment, surface finished improved and the spindle seems to run very well and quiet. This weekend the spindle ran esentially continous for about 30 hours at 5140 rpms (yeah I know I need a High Speed Spindle, looking at the 2.2KW water cooled 24,000rpm units). Spindle temperature rose to 90F to 95F and stayed there.


nitewatchman

[SCzEngrgGroup]

Is the bearing really directly exposed to coolant? Spindle bearings would normally be protected by a labyrinth seal integrated into the retaining nut at the lower end of the spindle. Hard to believe Tormach would miss something like that, and allow coolant to get into the bearings.

Regards,
Ray L.

[nitewatchman]

Is the bearing really directly exposed to coolant? Spindle bearings would normally be protected by a labyrinth seal integrated into the retaining nut at the lower end of the spindle. Hard to believe Tormach would miss something like that, and allow coolant to get into the bearings.

Regards,
Ray L.

Slinger Ring below the bearing is all.

Replaced the original spindle with the spare tonight and put new bearing in the removed spindle. The front bearing was also the culprit in this spindle. Grease was fine and not washed out but the outer race was spaulded and the balls were scored. New spare is back on the shelve in the bottom of the Tormach Cabinet until the next time.

nitewatchman

[SomeWhatLost]

anyone ever consider upgrading to ceramic bearings? seems like less friction may be a good thing, but more important is that they would probably be ok with having the grease washed away, and I am guessing they don't rust?
probably not cost effective, but still, may be cool

[LeeWay]

Ceramic bearings are not as strong a steel. They cannot take the pounding induced to spindle bearings. Ceramics are built for high speed and fairly static loads.

[nitewatchman]

Update on spindle after 15 hours of use.

I have now increased the pre-load on the replacement spindle twice and still get no heat buildup during operations including one ID engraving cut where the machine is running WAO. My 1100 actually seems to be cutting better now than when it was new with less tendency for high frequency howl or chatter than before on heavy cuts. The spindle and Fenner drive belt is also very quiet with the VFD howl being the loudest sound at high speeds. There is still a little sheet metal rattle somewhere at the highest speeds which I think indicates a slight imbalance in the rotating assembly. I think that I will take up the preload one more time during the next few days unless the spindle starts building heat during operation.

Looked at the original spindle bearings a little closer and question the choice of grease used to factory pack the bearings. The grease is a semiclear material that is fairly thin, almost like a gelled oil. I used a waterproof EP Wheel Bearing Grease on the new bearings. I was a little concerned about heat buildup from using the slightly heavier grease but this has not been a problem so far. I think that this type grease would have definitely helped to prevent the washout of the grease from the "spare" spindle due to coolant.


nitewatchman

[tbaker2500]

Interesting!
I'm glad you were able to salvage it.

Was there a difference in the grease used between the two spindles?

[nitewatchman]

Interesting!
I'm glad you were able to salvage it.

Was there a difference in the grease used between the two spindles?

Yes, the new spindle had the grease that almost looked like thinned vaseline, the "spare" was a more conventional looking grease but I don't think that it was waterproof.

nitewatchman

[anglin]

I pulled out my bearings over the weekend when one of them started screeching intermittently during my belt testing ( http://www.cnczone.com/forums/tormac..._than_oem.html ) when I loosened the belt almost all the way and ran it. I found the grease as you describe it in both the top and bottom bearing assemblies (Vaseline-like). The lower bearing appeared to have a little bit of coolant mixed in (the grease was a milky color and it looked wet). I'm going to be ordering replacements. Anyway, three of the four bearings fell apart when they were pressed out of the spindle housing or off the shaft. Only one remained intact and couldn't be disassembled. I took the bearings to my work to have our chief bearing engineer look them over and the only thing he notices was the track indication that showed up on the balls. It's common for an angular contact bearing to do that. The fact that there was a single track (on most bearings) indicated lack of procession of the rolling element which indicates that they were under some preloaded. Some balls did indicate that they shifted orientation from one rolling path to another. However, he did say that the orientation of the bearings (each bearing set with the orientation of the thrust face of the outer ring pointed the same way) was unusual (in his experience, but maybe normal on a machine tool spindle) and required careful preloading of the bearing to unsure correct load sharing of the bearing sets at each end.

With that mumbo jumbo out of the way. Did your bearings separate when you got them free of the assembly? How did you torque when you reassembled? By feel? Does Tormach provide any guidance? I couldn't find a torque requirement in the manual.

Thanks.

[nitewatchman]

Interesting, the spare which was quite rough when rotated by hand, did exactly as you describe and three of the four bearings separated when the spindle was pressed apart leaving the outer races in the cartridge, the inner races on the spindle and balls all over the floor. The second original spindle was not in as bad a shape and had just developed a rumble at all speeds. It pressed apart and all four bearing stayed intact. Inspecting the bearings, it appear that the rumble was in the bottom bearing closest to the taper and the top bearings look fine.

I pressed the bearing on in the following sequence. First the lower bearings were packed with a lithuim moly EP waterproof grease then matched up and pressed onto the spindle being sure to position the slinger ring below the bearings and the arrows pointed up. The the top bearing were packed, matched and pressed into the top of the cartridge flush with the top of the cartridge housing withthe arrows pointing down. The spindle was then inserted into the cartridge body and the lower bearings pressed home while supporting the top bearing on the inside and outside on the press bed. The spindle was firmly seated to make sure the lower bearings were tight to the shoulder in the cartridge housing (don't ask how I knew to do this and neither of us will be embarrassed). The spindle was then removed from the press taken to the bench and the top bearings tapped down into place using a piece of 1/2" key stock turned on a diagonal so that the corners contact the inner and outer races. I use steel because I have had brass drifts leave flakes behind before. Tap lightly and you will hear when it is seated by the change in sound.

Add the top slinger ring, install the nut and tighten as tightly as possible by have using a spanner. Loosen the nut and bring it back to snug. Tormach suggests tightening another 5 degrees but I think it requires more, this is however a good place to start. Reinstall the spindle. pulley, nut, etc. and power it up. Let it run for a while at low speed to work the grease through the bearing and expel excess grease. Step the RPM's up and down in similar manner to your vibration test. Pay attention to the spindle nose AND the top of the spindle housing for heat buildup. Both the spindles (spare and original) that I reassembled made noise at this stage in the 800 to 2000 RPM range. This later indicated that the bearing were too loose.

I put the machine back into production cutting Ductile Iron Castings and immediately noticed the the finished generated by cutting with the end of a 1/2" end mill was horrible while cutting with the side was actually better than I had remembered. After a couple of cuts all I could figure was that the spindle had end play and was floating up and down in the cut. Applying an indicator it would "float" by 0.001" or so when pryrd up and down. So again, off comes the PDB, the pulley while dropping the cartridge down to get the pulley out and I took up the top nut by perhaps 10 degrees using a hammer and punch. Started up again and the first thing noticeable was that the spindle noise was gone. Entered a cut and the finish was vastly improved and as good as I remember except for a few random swirls. Ran the part which is a 17 minute cycle with 8 tool changes and checked the temperature and it was at ambient. After several more parts, I repeated the adjustment and took up a little more lash in the spindle. Spindle now runs dead quite with great finishes and I believe that it cuts better than when it was new. After running continously on the castings for 8 hours including a 5/8" End Mill Plunging two holes at 100% of the green on the meter and an engraving cutter at 5100 RPM the spindle temperature was about 95F. I am pleased.

The only issue that I have now is some vibration at high speed (4800 up) that causes a rattle. The machine, yes with a Fenner belt is silky smooth and in my opinion quieter than the OEM belt. My OEM belt seemed to have a bump or stiff place that knocked and roared at high speed. Granted I didn't try a preminum belt from Gate or equal to compare. I had 3/8" and 1/2" Fenner stock on hand and just slapped her right on.

As to the bearing arrangement, it is also a little strange to me. I have used Angular Contact bearing on stuff I have designed before and normally I would have turned the lower bearings to face each other with a retainer nut on the spindle body to clamp the inner races. There would then be a retainer ring bolted to the spindle cartridge face to clamp the outer races. This would preload and set the lash of the main bearing as a unit. The rear bearing could then be a simple roller bearing or another pair a AC Bearings. Clearly a more expensive approach. I would also have used some type of seal on the lower spindle bearing. But these changes would increase the cost more than putting in an ocassional set of Chinese Bearings. I would have whinned about that also.

BTW - If you want a shock look at what a set of SKF matched bearing sell for even on eBay.

nitewatchman

[anglin]

Thanks for the thorough reply. I just reread my own post and I'm pretty unimpressed with how incoherent it is. I think my 5 month old must have been sitting on my lap at the time demanding too much attention.

Anyway, the real reason why I came back into this thread is because I explored installing the spindle bearings without pressing them into the cartridge housing (heating the housing) and thought I'd share the results here for anybody who might be considering the same thing. The bearing fits are tighter than I expected. Here are the fits, as I measured them:

Upper spindle bearing set: .0035" tight (interference) fit
Lower spindle bearing set: .0025" tight (interference) fit

I was using a caliper and not a micrometer, so they might actually be .0030" tight each, but I'm confident the fits are within the range listed.

I made a few assumptions with my calculation. First, the bearing assemblies would be installed at room temperature (70°F) since cooling bearings can result in rust-inducing condensation. Second, I assumed the spindle cartridge housing is made of steel and not iron (iron makes these numbers worse because it has a smaller coefficient of thermal expansion). The crossover temperature, where the fit goes from tight to loose, is as follows for each set.

Upper spindle bearing set: between 260 and 270°F (310-320°F if the cartridge housing is iron)
Lower spindle bearing set: between 200 and 210°F (230-240°F for iron)

I'd prefer to have a little bit of clearance in the assembly to address heating of the bearing outer ring when installing (don't want it to stop moving when it's half-installed), so to get .0010" loose fit, the temperatures need to be up near 325°F (upper) and 250°F (lower) (assuming steel).

I'm not willing to let the bearings sit in a 250°F heat sink with their phenolic cages on the rolling elements, so that rules out heating. Not much value to this post other than to show that the work has been performed already.

FWIW, CO2 cooling of the bearings (assuming the steel of the outer ring follows the 7.30E-6 in/in °F coefficient of thermal expansion) makes the upper bearing a loose fit (.0001") at 90°F cartridge housing temperature and the lower bearing is a loose fit (.001") at 70°F. Using LN2 gives something like .004" loose and .007" loose, upper and lower respectively, if the housing is at 70°F. Perhaps with the bearings fully greased up no condensation and resultant rust is a problem.

[cncoperator]

I was using a caliper and not a micrometer, so they might actually be .0030" tight each, but I'm confident the fits are within the range listed.

The accuracy of my 0-6" Mitituyo is stated as +/.001 and I expect most other calipers have similar accuracy specs. So if the bearing bore and housing seats were both measured the numbers given, .0035 and .0025 tight should include an uncertainty spread of +/- .002. If just the housing was measured and the bearing bore tolerance were looked up then it's +/- .001 plus the bore tolerance which for this size bearing in a high accuracy class is to the order of +/-.00015. In fact I find it hard to use a caliper for measurements that need .001 accuracy let alone a bearing seat that needs tenths measurements since just changing how hard I squeeze the caliper causes the measurement to change a couple .001. Use a micrometer.

Acceptable interferences is not an absolute number but dependent on the size of the bearing and it's application, but invariable a larger bearing will have more interference than a smaller one in the same type of application.

[anglin]

After running continously on the castings for 8 hours including a 5/8" End Mill Plunging two holes at 100% of the green on the meter and an engraving cutter at 5100 RPM the spindle temperature was about 95F.

Where did you measure the temperature and how? I recognize that somebody will criticize your method if you publish it, but I'm looking for something to use as a comparison when I test my spindle temperature, since we can't measure the temperature of the bearing outer ring directly.

Thanks.