Grinding spindle.

[stackerjack]

I've built a grinding spindle using bronze bushes. I haven't used it yet. Do you think there will be any problems?
Jack

 

[BaronJ]

Hi Guys,

I used dual bearings in my grinding spindle at both ends.

The shaft is secured in the pair of bearings at the wheel end and can slide through them at the drive end. Total cartridge length about seven inches. I actually did away with using an "O" ring to provide preload and replaced it with a shim, relying on the threaded end cap to prevent the bearing pack from floating in the housing.

 

[mu38&Bg#]

I've built a grinding spindle using bronze bushes. I haven't used it yet. Do you think there will be any problems?
Jack

If you've accounted for the heat generated at the RPM you intend to use it at, it should be fine. Is oil pressure fed?

 

[stackerjack]

If you've accounted for the heat generated at the RPM you intend to use it at, it should be fine. Is oil pressure fed?

No, I'm afraid it isn't. It's just got an oil reservoir mounted above each bearing housing, similar to a Myford Lathe. I cut a spiral in the shaft to persuade the oil where to go.
Jack

 

[TSutrina]

Hi Guys,

I used dual bearings in my grinding spindle at both ends.
View attachment 127101

The shaft is secured in the pair of bearings at the wheel end and can slide through them at the drive end. Total cartridge length about seven inches. I actually did away with using an "O" ring to provide preload and replaced it with a shim, relying on the threaded end cap to prevent the bearing pack from floating in the housing.

Worked for an aerospace company for 20 years. Built aircraft generators usually 2 pole at 400 Hz 24,000 rpm and rocket driven turbines that spin at a much higher speed. Rule for bearing is always to preload them. This insures the balls are always rolling. Do not want what happens with the tires of a plane touch the ground happening to the balls. Obviously the higher the speed the better the bearings needed and also the lubricant. Also preloading helps to insure the axis of the bore and the shaft are the same.
Journal bearings have to have clearance so the axis will not match up. Further wear exist. At high speed insuring lubricant is always present. Self lubricated is not going to cut it. Also need to remove heat from the oil. which is another reason to have a oil source. The load on the shaft angle changes will cause problem with lubricating the bearing. So the weight of the grinding wheel result in the force at straight down. But grinding the top of a surface causes the to be close to straight up. The bearing on the drive end will not be the same as the bearing near the grinding wheel. Never used journal bearing in our aerospace products.

 

[stackerjack]

You say that when bearings are pre-loaded, the balls will always be rolling. I cannot see how the balls will roll on two surfaces of different diameters, being the inner and outer races.

 

[TSutrina]

All bearing are angular contact when preloaded. as you can see that the balls contact at a circular line of contact on the OD and ID race that are different diameters. The contact is a very small area for each ball. due to the metal under load deforming elastically like for example press a rubber blall or even a basket ball against a glass you well see the contact is as shown.

 

[BaronJ]

You say that when bearings are pre-loaded, the balls will always be rolling. I cannot see how the balls will roll on two surfaces of different diameters, being the inner and outer races.

Its is the same situation as having three gears of different diameters running together.

 

[petertha]

James (Clough42) did an interesting series on his tool post grinding. There are quite a few episodes but this is the main video explaining his design rationale & some good illustrative CAD sections/commentary. He mentions one of the challenges was sourcing AC bearings (within his own imposed spindle OD constraint). Eventually, many videos later, you can see it run but it was kind of preliminary, lacking good wheel dress at that particular test moment.

I haven't tuned in to hear how the deep groove bearings are holding up. I might be wrong but I don't think this assembly lends itself to tailstock support while grinding despite being quite compact & toolpost dovetail mount. He also swivels his compound at the magic angle so 0.001" compound infeed = 0.0001" cross bed DOC. I didn't see him lock his compound so maybe that lathe doesn't have one. That's an unfortunate drift source from my experience. I didn't see an independent dial indicator setup either which may have helped, its strictly reading the grads. Anyways, this discussion is about spindles so providing another informational data point.

 

[stackerjack]

Its is the same situation as having three gears of different diameters running together.

Sorry, but I disagree with this.

 

[stackerjack]

All bearing are angular contact when preloaded. as you can see that the balls contact at a circular line of contact on the OD and ID race that are different diameters. The contact is a very small area for each ball. due to the metal under load deforming elastically like for example press a rubber blall or even a basket ball against a glass you well see the contact is as shown.

Thank you for your very explicit response, I'm afraid I'm not conversant with the formulae. So, what you are saying is that: the balls just deform to allow them to "skid" along the tracks. You are correct in stating that the contact area is extremely small, this is what causes very high pressures between the balls and the tracks, even when the load is not very high, which is presumably the cause for the deformation.

 

[Charles Lamont]

So, what you are saying is that: the balls just deform to allow them to "skid" along the tracks.

Eh? No, no, no - the balls roll on both races, in spite of the problem you seem to be having with the idea. I think BaronJ may have been refering to epicyclic gears, in which case the comparison is most helpful. I am expecting soon to hear the tinkle of a penny dropping.

 

[willray]

Sorry, but I disagree with this.

I disagree with a great many things, but this does not necessarily make me right.

For simplicity of illustration, ignore deformation/non-point contact patches. These actually do complicate things somewhat.

Take two horizontal, parallel infinite planes, and place a ball between them in contact with both. Do you agree that one plane can translate left/right while the ball rolls without sliding, in contact with both? If you do, then you disagree with your disagreement.

Consider that the contact between the ball and the upper plane is a point.

Because the contact between the ball and the upper plane is a point, is there any conceivable way that the ball knows that the upper plane is a plane, rather than a cylinder?

Since the answer is no, convert the upper plane to a cylinder. Balance it on the ball and roll the whole thing along left or right - does anything need to slide?

The exercise is left to the reader to extend the illustration to an outer cylinder and an inner cylinder, instead of a plane and a cylinder.

 

[SmithDoor]

The deep grove ball bears has made many changes. They made lot of older bearings old obsolete but still used by some manufacturer's.
Tapper bearings are for low speed and high loads.

The center and centerless grinders use plane bearings for higher quality work.

I found in replacing small grinders use seal bearings to keep dust out.

Dave

I disagree with a great many things, but this does not necessarily make me right.

For simplicity of illustration, ignore deformation/non-point contact patches. These actually do complicate things somewhat.

Take two horizontal, parallel infinite planes, and place a ball between them in contact with both. Do you agree that one plane can translate left/right while the ball rolls without sliding, in contact with both? If you do, then you disagree with your disagreement.

Consider that the contact between the ball and the upper plane is a point.

Because the contact between the ball and the upper plane is a point, is there any conceivable way that the ball knows that the upper plane is a plane, rather than a cylinder?

Since the answer is no, convert the upper plane to a cylinder. Balance it on the ball and roll the whole thing along left or right - does anything need to slide?

The exercise is left to the reader to extend the illustration to an outer cylinder and an inner cylinder, instead of a plane and a cylinder.

 

[BaronJ]

I disagree with a great many things, but this does not necessarily make me right.

For simplicity of illustration, ignore deformation/non-point contact patches. These actually do complicate things somewhat.

Take two horizontal, parallel infinite planes, and place a ball between them in contact with both. Do you agree that one plane can translate left/right while the ball rolls without sliding, in contact with both? If you do, then you disagree with your disagreement.

Consider that the contact between the ball and the upper plane is a point.

Because the contact between the ball and the upper plane is a point, is there any conceivable way that the ball knows that the upper plane is a plane, rather than a cylinder?

Since the answer is no, convert the upper plane to a cylinder. Balance it on the ball and roll the whole thing along left or right - does anything need to slide?

The exercise is left to the reader to extend the illustration to an outer cylinder and an inner cylinder, instead of a plane and a cylinder.

This is a much better example than the one that I used, but the principle is the same.

 

[cds4byu]

The thing that stackerjack has likely missed is that the balls are free to move at a different velocity from either the inner or outer races. The velocity of the center of mass of the ball will be just the right velocity needed to provide rolling contact.

For the flat plane case, with the lower plane fixed, the upper plane will move at twice the velocity of the ball.

For the cylindrical case, with the inner race fixed, the outer race will move at (radius of outer race + 3*radius of inner race)/( 2*radius of outer race) times the speed of the ball. Note that for the flat plane case (outer radius = inner radius), this formula goes to 2.

It's this self-adjusting speed of motion of the balls that allows you to have pure rolling, even though the inner race and outer race have different radii.

Carl

 

[TSutrina]

Thank you for your very explicit response, I'm afraid I'm not conversant with the formulae. So, what you are saying is that: the balls just deform to allow them to "skid" along the tracks. You are correct in stating that the contact area is extremely small, this is what causes very high pressures between the balls and the tracks, even when the load is not very high, which is presumably the cause for the deformation.

the balls are like your car tires. they roll instead of slide. The tires roll on flat, concave, and convex roads and stay or track in a worn cupped out track. Balls act the same way. The lubricant is like having a wet road. So long as the acceleration of the ball is low it rolls instead of slides. So a ball that in not in contact with the two races like a non preloaded bearing during some portion of the circle then it slows up. When it makes contact and loaded it must accelerate in a very short time period, skids. The formulas are not important except for those that design bearing. In reality the life of a bearing is determined by testing. The testing information is provided on each bearing. This use is within the range of the data. Have question give the company a call. Purchase the bearing from suppliers that provide access to the manufacture.

 

[Alec]

Hullo, I looking for ideas really. I retired from precision engineering, college workshop technician and 30 years in a shipyard and now have a small Colchester student and similar size vertical mill. I have been making the Quorn t/c grinding machine, which has been a challenge but enjoyable.Being me I am looking to remake the grinding spindle with a different design, perhaps angular contact bearings , just because I can, nothing wrong with the original design by the way. I'd welcome ideas on this please.

Further to my asking for your ideas please take this as a large thank you to everyone.
In the notes for the Mk3 Quorn it is suggested for the split joints ( as opposed to double cotter type) that the joint is first split,an undersized shim inserted -.025mm/0.010” and snugged down then lastly bore to size. I went with the double cotter way to clamp my machine. It’s integrity is unaffected but it’s only my decision on my machine.
I also kept all sizes metric . But that is my preference.

 

[Ken I]

Just a further few cents worth - having had the experience of rebuilding Cincinnatti #2 & #3 Filmatic bearings (tilting shoes bearings based on Mitchell thrust bearings) as well as Jones & Shipman grinding spindles.
As strange as it sounds - grinding spindles need to run with zero clearance otherwise the spindle orbits on its clearance which is bad for accuracy and surface finish.

Cassette bearings typically use an opposing pair of angular contact or tapered roller bearings which are usually spring loaded. These are in close proximity at the working end and an outrigger bearing supports the opposite end allowing thermal contraction and expansion to "slide" through the "ourtigger" bearing.

Plain bearings need to employ tricks vis :- (Highly exaggerated to show principal).

The J&S arrangement uses a white metal lined collet arrangement which is sprung and bored oversize. On being closed down to its calibration shim it allows the oil to form three very powerful wedges that allow the spindle to run clearance free dynamically (static it will have some clearance).

The Cincinnatti Filmatic uses three adjustable tilting shells on the #2's and five on the #3's. (Essentially a variation on the Mitchell tilting pads used for thrust bearings in marine propshafts.)

As the shaft rotate the oil pressure lifts the leading edge thus closing the trailing edge - kind of an irresistible force against an immovable object - so the spindle rotates with zero clearance dynamically.

Typically working at pressures approaching the film shear strength of the oil. I was trained at Cincinnatti to set them such that they run at 65°C - cooler is too loose and hotter too tight.

No matter what the arrangement is, grinding spindles run with no clearance and therefore at the limits of the shear strength of the lubricating oil - so expect grinding spindles to get quite warm.

And don't use anything except the recommended lubricant.

Regards,
Ken

 

[BaronJ]

Hi Ken,

Thank you for your post ! It answers some questions that I had when designing and making my grinding spindle ! I had wondered why the Cincinnati had those funny screws on the spindle. I just didn't appreciate what they did.