Bronze Spindle Bearing RPM Question

[jspivey]

I have a pretty long commute every day and like to use that time to think about interesting things. Recently I’ve been thinking about plain spindle bearings.

Say a person wants to build a relatively small, fast spindle but doesn’t want to pay the premium for actual Timken (or other) spindle bearings. Obviously a plain bearing is much cheaper.

Here are some assumptions:
1: Bearing bore diameter is nominally 2”ID x3” long. Bearing is of appropriate bronze alloy.
2: You have a hardened and polished spindle with appropriate clearance running in the bore.
3: You have a pressurized lube system for the bearing (ie not an oil cup)
4: There is a pressurized coolant system built into the spindle/spindle housing to prevent over heating.
5: Everything is properly aligned.
6: Axial and radial loads will be relatively light (think the kind of force that a small benchtop lathe could handle comfortably)

Everyone knows that older plain bearing lathes have relatively low spindle speeds, but I understand this is because the were expected to have relatively high cutting forces. A “force on spindle vs spindle speed” relationship. Assuming cutting forces are very low, what is the determining factor for spindle speed? Thermal expansion that can be compensated for with a cooling system?

How can one roughly figure the max safe rpm you could run such a spindle at without serious issues?

 

[Milland]

To some degree you'll be limited by molecular stress within the oil itself. Even with cooling there will be thermal effects of oil shear vs. breakdown of hydrocarbon bonds and temporary and permanent viscosity changes.

My WAG for a well designed hydrostatic spindle with decent separation of bearing journals and no external influences (shaft flexing and whirl, off-balance loads, long overhangs, etc.), would be somewhere in the 7-10K RPM range. Fancier lubricants, optimized surface features for lubrication uniformity, best filtration and temp control might take you up to mid-teens. If we take "fluids" to include gasses and the loadings are light, then mechanical strength will be the limit.

But, as always, my guess and $10 will buy you a fancy coffee.

 

[john.k]

Pressure lubrication with thin filtered clean oil will cope with 10,000rpm.....heat buildup is the limiting factor,remove the heat and plain bearings are more rigid and more reliable than any roller .

 

[Cole2534]

Pressure lubrication with thin filtered clean oil will cope with 10,000rpm.....heat buildup is the limiting factor,remove the heat and plain bearings are more rigid and more reliable than any roller .

So a large sump with heat exchangers?

Sent from my SM-G973U using Tapatalk

 

[ferretlegger]

Perhaps you might think about a different kind of bronze spindle bearing. An AIR bearing. As I am sure you know, air bearings can be extremely high speed (hundreds of thousands of RPMs), low friction, very strong with respect to axial and radial forces. I once designed high speed laboratory spinners for magnetic recording studies using NSK air bearings. We also had much larger ones made by Professional Instruments. The biggest single difference between these super high performance bearings and your Thought experiment spindle is that they used pumped air instead of pumped oil or an oil film to float the rotating spindle in a "porous" bearing. In the air bearing, this was a sintered (I think) bronze like material which was porous on a small scale. The air was pumped at 100 PSI or so into a cavity surrounding the porous end and side bearings, and the flow into the inner space where the spindle was caused the spindle to float. If the spindle got off center, the local pressure between the side (or end) of the spindle and the walls went up drastically, pushing the spindle back into the middle. Just like an oil bearing. The clearances between the spindle and the porous bearing material were very, very tight, and the cylindrical and end plate tolerances were also very tight. I am sure that the NSK or Professional instruments spindles were cylindrically and flat ground to very exacting specifications, as were the porous bearing material parts.

I found many air bearings on EBay, for very reasonable prices. One a Professional Instruments one for $1400. Many of them had motors attached. For high speed operation, the motors need to be very well balanced. In the spindles I built, I used 3 phase brushless pancake motors and a custom controller I designed with 1 ppm stability from 300 to over 10000 RPM. This was MANY years ago. I am sure better performance is easily achieved today.

Anyway, while enduring your long drive, you might enjoy pondering the air bearing, and possibly scheming how you might build one in your own shop. I have not looked, but I would be astonished if there was not a mountain of design information available, and even more so if there were not people on this site who were experts at designing and/or building them. But please, don't get lost in your head contemplating all the possibilities and drive off the road!!
All the best,
Michael

 

[john.k]

Just think about all the pressure lubricated plain bearings in your car that turn at up to 6/7000rpm.,and the plain bearing in the turbo that turns at 60,000rpm.

 

[4GSR]

Rather than bronze bearings, look at cast iron. Far much less thermo expansion when it starts to heat up. Look at the spindle on a 10" L South Bend lathe, most of them are running on cast iron, with oil that is recirculated. That spindle can run up to around 2000 RPM with the correct lubrication. Another example, is to look at the large thrust bearings used on heavy turbo machinery, they are cast iron. Ken

 

[tdmidget]

Rather than bronze bearings, look at cast iron. Far much less thermo expansion when it starts to heat up. Look at the spindle on a 10" L South Bend lathe, most of them are running on cast iron, with oil that is recirculated. That spindle can run up to around 2000 RPM with the correct lubrication. Another example, is to look at the large thrust bearings used on heavy turbo machinery, they are cast iron. Ken

I can't say that there are no cast iron turbine bearings, not having seen them all, but all that I have worked on and seen are Babbitt.

 

[johnoder]

The 2" bronze bearings in my B&S Universal grinder run at 1500 all day with very mild heating, drip fed Mobil Velocite 8, with NEGLIGIBLE clearance, 14" grinding wheel.

That's a bit shy of 800 SFM at the journal

Entirely satisfactory results for a 72 year old machine

 

[4GSR]

I can't say that there are no cast iron turbine bearings, not having seen them all, but all that I have worked on and seen are Babbitt.

You are correct. Way too early for me to comment on.

 

[jspivey]

To some degree you'll be limited by molecular stress within the oil itself. Even with cooling there will be thermal effects of oil shear vs. breakdown of hydrocarbon bonds and temporary and permanent viscosity changes.

My WAG for a well designed hydrostatic spindle with decent separation of bearing journals and no external influences (shaft flexing and whirl, off-balance loads, long overhangs, etc.), would be somewhere in the 7-10K RPM range. Fancier lubricants, optimized surface features for lubrication uniformity, best filtration and temp control might take you up to mid-teens. If we take "fluids" to include gasses and the loadings are light, then mechanical strength will be the limit.

But, as always, my guess and $10 will buy you a fancy coffee.

Thank you for the guess, but I’m not much of a fan of fancy coffee haha. 7-10k is more than I would have thought, but then again that’s why I asked. I appreciate the input, gives me plenty more food for thought.

 

[jspivey]

Perhaps you might think about a different kind of bronze spindle bearing. An AIR bearing. As I am sure you know, air bearings can be extremely high speed (hundreds of thousands of RPMs), low friction, very strong with respect to axial and radial forces. I once designed high speed laboratory spinners for magnetic recording studies using NSK air bearings. We also had much larger ones made by Professional Instruments. The biggest single difference between these super high performance bearings and your Thought experiment spindle is that they used pumped air instead of pumped oil or an oil film to float the rotating spindle in a "porous" bearing. In the air bearing, this was a sintered (I think) bronze like material which was porous on a small scale. The air was pumped at 100 PSI or so into a cavity surrounding the porous end and side bearings, and the flow into the inner space where the spindle was caused the spindle to float. If the spindle got off center, the local pressure between the side (or end) of the spindle and the walls went up drastically, pushing the spindle back into the middle. Just like an oil bearing. The clearances between the spindle and the porous bearing material were very, very tight, and the cylindrical and end plate tolerances were also very tight. I am sure that the NSK or Professional instruments spindles were cylindrically and flat ground to very exacting specifications, as were the porous bearing material parts.

I found many air bearings on EBay, for very reasonable prices. One a Professional Instruments one for $1400. Many of them had motors attached. For high speed operation, the motors need to be very well balanced. In the spindles I built, I used 3 phase brushless pancake motors and a custom controller I designed with 1 ppm stability from 300 to over 10000 RPM. This was MANY years ago. I am sure better performance is easily achieved today.

Anyway, while enduring your long drive, you might enjoy pondering the air bearing, and possibly scheming how you might build one in your own shop. I have not looked, but I would be astonished if there was not a mountain of design information available, and even more so if there were not people on this site who were experts at designing and/or building them. But please, don't get lost in your head contemplating all the possibilities and drive off the road!!
All the best,
Michael

Michael,
Thank you very much for the info. I hadn’t really considered air bearings, mostly due to a lack of knowledge. I know what they are and their working principles, but that’s about it.

However, for what I am imagining, it sounds like I could be confident in achieving the desired speeds with the hydrodynamic bearing I described above. It also sounds like building an air bearing spindle might be a little beyond what I am capable of in terms of both skill and machinery. Also, one could build a plain bearing spindle for significantly less money, even if $1400 for a good spindle is cheap (not counting lost free time of course). It will be something else to think about while I try and stay between the lines!

 

[ferretlegger]

jspivey,
I am sure that there are beter solutions to your problem than air bearings. But since you seem to be a curious, thoughtful bloke, here is a video about an air bearing based CNC lathe. Pretty amazing!
YouTube

Regards,
Michael

 

[jspivey]

jspivey,
I am sure that there are beter solutions to your problem than air bearings. But since you seem to be a curious, thoughtful bloke, here is a video about an air bearing based CNC lathe. Pretty amazing!
YouTube

Regards,
Michael

Michael,

The link takes me to eBay but I believe I know the video you are talking about. It’s quite an impressive machine he built. Perhaps a little more sophisticated than what I would consider tackling!

 

[DDoug]

How is having a pressurized lube system going to be cheaper than some ball/roller bearings purchased off the shelf ?

 

[Milland]

How is having a pressurized lube system going to be cheaper than some ball/roller bearings purchased off the shelf ?

Heh - I wonder if this is an actual application, or just farting around. But if a real need, perhaps the fastest way to go is repurpose an engine block, anything from a lawnmower engine (with pressure lube) up to a motorcycle or car block. With any engine that uses an oil pump, you could adapt it to electric motor drive, and use a shaft the right diameter with a couple of thrust collars where the crank's would have been.

You'd have to plug some oil passages, but otherwise it's almost ready to go. Plenty of sizes out there, accurately machined, easy bearing replacements (and usually undersized available), sumps, etc.

 

[DDoug]

Heh - I wonder if this is an actual application, or just farting around. But if a real need, perhaps the fastest way to go is repurpose an engine block, anything from a lawnmower engine (with pressure lube) up to a motorcycle or car block. With any engine that uses an oil pump, you could adapt it to electric motor drive, and use a shaft the right diameter with a couple of thrust collars where the crank's would have been.

You'd have to plug some oil passages, but otherwise it's almost ready to go. Plenty of sizes out there, accurately machined, easy bearing replacements (and usually undersized available), sumps, etc.

...and then it needs to go over to the Hobbymachinist.com web forum.

Not practical at all.

 

[jspivey]

How is having a pressurized lube system going to be cheaper than some ball/roller bearings purchased off the shelf ?

Well, I am assuming that the parts (pump, fittings, etc) one would have to buy for a pressurized system are, on a small scale, going to be cheaper than multiple spindle bearings at $500 or more per bearing.