Plain bearing spindle design

[Borealis]

We need to design and build a couple small spindles for some grinding operations. They need to be reasonably but not exceptionally accurate: 1 tenth TIR would be perfectly adequate. I have one small Blockhead air bearing that's the right size, but it's overkill for the task and I don't want to risk putting it in harm's way when we usually use it for special measurements. I thought about using a Harig grind-all spindle, but the center height won't work out with the mounting arrangement.

I can design and build a spindle with a couple pairs of preloaded angular contact bearings for this. That wouldn't be a problem. Before commiting to that, I wanted to consider a plain-bearing spindle. Years ago, I had an ancient Hendey conehead lathe with plain bearings that turned exceptionally round. Thousands of SB lathe spindles had plain bearings, and the old Crystal Lake Grinders claimed to achieve sub-0.00001 TIR with their conical spindles riding in scraped bronze.

Can anyone point to a reference that discusses the design of plain-bearing machine-tool spindles in some detail? I've never disassembled a precision plain-bearing spindle and I'm unclear on several details...for instance, how are thrust loads resisted in both directions, and how is the oil-gap maintained as the spindle warms up? It seems like the spindle will get hotter than the housing, which would cause the spindle to grow faster and for the oil-gap to increase.

 

[eKretz]

You might take a look at Mattison grinder spindle info if you can find it. They used a conical plain bearing also. Very good spindles.

 

[John Garner]

Adding to eKretz's suggestion, take a look at Studer's plain-bearing spindle design . . . also an "if you can find it". Might be easier to locate, because Studer grinders are still being made.

 

[TGTool]

A mechanical engineers handbook from the 20's or 30's might be a good place to look for a discussion of plain bearings and calculations thereof. I think they had a fairly good grasp of the hydrodynamic characteristics at that time, and plain bearings were still common.

In the simplest case you just have a round shaft in a round bushing and you're making sure it gets lubricant. For higher speed and precision you would have the shape of the bearing such that it builds pressure in the lubricant that then supports the spindle without contact between the two.

This isn't exactly relevant to your problem, but I learned some interesting things rebuilding a turbocharger some years ago. It's a plain bearing, and the surface speed between spindle and housing are way, way above usual expectations. So what they have is a ring between the spindle and the housing that looks something like a bearing cage without balls. That is a bunch of radial holes. The design theory is that this cage spins at some intermediate speed.dividing the surface speed between spindle to cage, and cage to housing. It's fed with pressurized oil and seems to work.

 

[michiganbuck]

As much as I love plane bearing spindles for their ability to often give a better finish than a beringed spindle, the warm up time is a problem factor for much work. So many grinding jobs may be a 15 minuet or hour task, the 20 minuets warm up can be a cost factor.

Building one I would likely go with a catalog bronze bearing, and build a tube for it. Seems like schematics for Brown & sharp plane bearings should not be hard to find.
I have a B&S 13 plane bearing spindle if any is interested, should be a deal for $500 plus postage, it fits all B&S 13s..

Some catalog bearings.

Amazon.com : bronze sleeve bearings

 

[Cole2534]

What about taking a page from engine design and using pressure lubed plain bearings?

 

[technocrat]

What about taking a page from engine design and using pressure lubed plain bearings?

I believe that you add too much drag at high speed with pressure lube.

 

[Cole2534]

I believe that you add too much drag at high speed with pressure lube.

He doesn't give a shaft speed, but I took the post to mean relatively low RPM.

 

[john.k]

In fact ,the random runout of the wheel spindle influences finish and wheel wear ......its the work head bearings that define accuracy of the work.......rigidity of both wheel spindle,work spindle and the whole machine has a greater influence on accuracy.

 

[Strostkovy]

I believe that you add too much drag at high speed with pressure lube.

That doesn't sound right to me. Turbos use pressure lubrication and spin extremely fast.

 

[technocrat]

That doesn't sound right to me. Turbos use pressure lubrication and spin extremely fast.

that is a pretty specialised case, light load plus very high energy source plus the need to remove heat. Not many other bearings operate in that situation. My general understanding was that with spindles, once you moved from plain bearings you went to ball bearings with light oil as the speed and load increases.
Anyway, you don't get edjumacated unless you say what you think.

 

[Borealis]

I lost track of my own thread for a few days--thanks to all of you for the excellent suggestions.

I've decided to split my efforts: I designed and started building of the spindles using 7009 P4 angular contacts--very conventional and simple. The other one will be plain bearing. If it works, great. If not, I'll learn something.

It's hard to find specifics about a lot of spindles...I couldn't find diagrams of a Mattison or Studer spindle...at least searching any obvious phrases on google images. There are quite a few x-sectionals of Hendey and SB lathe spindles, however. Those diagrams answered a few questions.

Machinery's Handbook has some guidance, as does Mark's Standard Handbook for MechEs. Neither of those seems to address the analysis of tapered spindles, though. For small angles, it probably doesn't matter much, but for spindles with significant taper, it may be a different story.

 

[Doozer]

Cincinnati had some plain bearing grinding spindles called Filmatic. This was 4 pinches instead of the normal 2 pinches. This means 4 eddys circulating forcing the spindle to self-center rather than 2. More rigid when running is what that says to me.

-Doozer

 

[Robert R]

Hydrodynamic bearings are still being developed for the magnetic media disk drive industry.
For example:

Hydrodynamic bearing assembly and spindle motor having the same

Complete Patent Searching Database and Patent Data Analytics Services.

www.freepatentsonline.com

The descriptions of the older machine tool hydrodynamic bearings can be found in the above patent search website by typing the manufacture's name and product description.

For example: "Cincinnati machine tool bearing " or " Toyoda hydrodynamic bearing" , or "Landis machine spindle bearing"

You can narrow down the search to the older technology by restricting your search to patent numbers less than 5,000,000. This corresponds to the early 1980's.

 

[Conrad Hoffman]

IMO, a plain bearing grinding spindle isn't trivial to build and you need a decent cylindrical grinder to do it right. I'd stick with the angular contact bearings or even deep groove, just the way DuMore did it. Google "Dumore grinding spindle diagram" or look at the diagram near the end of this- https://www.homemodelenginemachinist.com/threads/grinding-spindle.33281/
Lube it with a high dm grease like Chevron SRI or some expensive Kluber goop like NBU-15 or similar.

 

[TGTool]

Cincinnati had some plain bearing grinding spindles called Filmatic. This was 4 pinches instead of the normal 2 pinches. This means 4 eddys circulating forcing the spindle to self-center rather than 2. More rigid when running is what that says to me.

-Doozer

Two? I would have expected at least three for stability. Four would be workable but overkill. In operation it has to develop a centering force if the spindle moves out of position and two seems like it could not develop the same force for displacement in any direction.

 

[Doozer]

http://vintagemachinery.org/pubs/10764/19843.pdf

 

[Borealis]

IMO, a plain bearing grinding spindle isn't trivial to build and you need a decent cylindrical grinder to do it right. I'd stick with the angular contact bearings or even deep groove, just the way DuMore did it. Google "Dumore grinding spindle diagram" or look at the diagram near the end of this- https://www.homemodelenginemachinist.com/threads/grinding-spindle.33281/
Lube it with a high dm grease like Chevron SRI or some expensive Kluber goop like NBU-15 or similar.

Conrad, what does "high dm" mean?

 

[Peter from Holland]

I would look for the additional spindles for ID grinding on a ID/OD grinder
Plenty available New or used
There is more into making a plain bearing for a grinder as meets the eye
If it is for fun?? Please go ahead

Peter

 

[just Dave]

Is this occasionally used or a production set up?