How to measure TIR of a hydrodynamic bearing spindle?

[jim rozen]

Good paper, thank you.

 

[luke8888]

The grinder I worked with was an old Levin-Tsugami OD instrument grinder. Quite rare. The wheel spindle looked very similar to your taper design, but had tapers on both ends, not rolling bearings. The oil was crazy thin, like water thin. There may have been a lower Velocite number back then, but I can't find any reference to it. Though it may have been Velocite #3, I remember something like #1 or #2. Thinnest stuff I've ever seen, but that's what worked correctly with the taper design.

Interesting. There is definitely Velocite 3. In fact I would get it for my spindle to mix with Velocite 6 in equal proportion if they sold it in small amounts (the least I can buy is a 20L drum). The manual for my machine specifies some ancient oils no one makes anymore. There is a preferred and an alternate oil. Velocite 6 is a little to viscous even as alternate. Velocite 3 is closer to the one they say is preferred. If I remember correctly it is supposed to have viscosity of 4 in 40C. So it would be ISO no 4, but no one makes one as far as I know. I could set less of a clearance if I had less viscous oil.

 

[mottrhed]

The pump is actually a spiral groove cut into the shaft. It pumps as it spins... But as another poster said the dynamic behaviour will be different. There are various effects at speed like the oil wedge etc

This is not really a "pump" this just allows oil to distribute across the bearing, the actual pressurization is done by the eccentricity of the bearing causing pressure to build as the shaft spins and the oil get "squished" through the pinch point-therefore creating float and load-stiffness

The brand is Jotes Spc20a, the spindle is made by a Polish bearing manufacturer FŁT (also known as PBF). BTW, it is a machine made very long time ago.

Here is a spindle drawing:
View attachment 394357
The zigzag shape on the shaft is the pumping part. It is supposed to use angular contact bearings 7206 class 5 in the back. It runs in Mobil Velocite 6.


I do give the details above, but they are really just a curiosity. I'm almost certain people that know anything about this specific spindle will not say a word here, because any published information affects their ability to charge pretty penny for rebuild services. One of the reasons why I started this rebuild myself is to put more information about this spindle on the net. It is an old model. Many people consider it's "plain" front bearing inferior and there is a market here in Poland for a rolling bearing replacement service for this one and similar spindles.

Dont expect to find anyone who knows this exact spindle, I agree with you there. But believe it or not, there are people that come here to give out data and help others understand what they are looking at. Sometimes its enough to get you over the hurdles and complete a successful rebuild, sometimes it shines a light on how complicated this really is and the tooling required to do it correctly, therefore avoiding expensive mistakes. Either way getting the info out there is good for the machine owner IMO.

Hydrostatic and Hydrodynamic bearings have their place, and when done correctly are still superior to a roller bearing in a few ways-depending on the application. The reason some people convert them is cost, part availability, and knowledge to service correctly. Almost never will the roller conversion be as rigid, repeatable or have the same life span as a properly built hydro spindle, so yes there are compromises.

This spindle uses a tapered tri-lobe design, shiguya okuma and a few others use the same style, usually bronze of some sort. Sometimes babbited, but the philosophy remains the same. Very finicky and sensitive to set correctly without deforming (destroying) the bearing. Usually when these are damaged, they need to be replaced, otherwise you can open them up a tad and make the shaft bigger to compensate.

Runout; theoretically a hydro spindle should produce virtually zero runout (although there is no such thing), or more accurately rotation concentricity should be held, within the limits of the roundness of the shaft, pressure consistent, and all other things being right. Now this doesnt mean at the tooling (grinding) surface you will have zero runout. What you should have is repetitive runout, virtually zero non-repetative runout. To dumb this down a tad, the shaft can be spinning perfectly true, but if the od taper that your hub sits on had 5 microns runout relative to the journals, you will still have the 5 microns.

In terms of dynamic runout, no not under load by grinding, but having the correct load at the bearing supplied by having the clearances all set right. When everything is set right hydrostatic bearings are super rigid, thats why they are used, they dont deflect much at all when working load is applied. My point is that runout, static or dynamic means nothing unless you absolutely know youre "loaded" (this term kid of confuses the topic-maybe stiffness is correct would be clearer?) -- essentially building the correct amount of pressure in the bearing to be held rigidly on center.

 

[john.k]

What about the old 'Filmatic ' bearings in Cincinnatti grinders ......rocking shoes with equalized oil film forces .......was claimed they would work with anything from soapy water to thick oil......Churchill grinders had 'air bearings' in the mid 1960s ......a wheel would rotate to heavy part down as soon as air pressure was applied .....no oil at all .....in fact the air had to be oil free.

 

[luke8888]

Thank you for a very comprehensive reply :-)

This is not really a "pump" this just allows oil to distribute across the bearing, the actual pressurization is done by the eccentricity of the bearing causing pressure to build as the shaft spins and the oil get "squished" through the pinch point-therefore creating float and load-stiffness

Interesting. I called it a "pump" in layman's terms as it certainly is a lot more effective in sending that oil to the front of the bearing than straight grooves and a whick supplied from above. What I failed to mention is that the shaft has a smaller opposing "pump" at the very end. I imagine it is there to prevent oil spilling out through the front clearance. As what looks like an oring groove, doesn't have any prints. It does work unless one over fills it with oil.

It can be seen here:

(BTW, the "band" seen on the shaft in this photo was one of the main reasons for the inability to adjust clearance. It has been fixed by cylindrical grinding since).

Dont expect to find anyone who knows this exact spindle, I agree with you there. But believe it or not, there are people that come here to give out data and help others understand what they are looking at. Sometimes its enough to get you over the hurdles and complete a successful rebuild, sometimes it shines a light on how complicated this really is and the tooling required to do it correctly, therefore avoiding expensive mistakes. Either way getting the info out there is good for the machine owner IMO.

I see. Yes, you're right. Regarding my machine, I think the photos I've uploaded are the only ones on the Internet showing that particular (very popular in my country) spindle.

Hydrostatic and Hydrodynamic bearings have their place, and when done correctly are still superior to a roller bearing in a few ways-depending on the application. The reason some people convert them is cost, part availability, and knowledge to service correctly. Almost never will the roller conversion be as rigid, repeatable or have the same life span as a properly built hydro spindle, so yes there are compromises.

This spindle uses a tapered tri-lobe design, shiguya okuma and a few others use the same style, usually bronze of some sort. Sometimes babbited, but the philosophy remains the same. Very finicky and sensitive to set correctly without deforming (destroying) the bearing. Usually when these are damaged, they need to be replaced, otherwise you can open them up a tad and make the shaft bigger to compensate.

Thank you for the proper name for this bearing design. At least now I have a term to search books etc with :-)

The spindle rebuild (touch up really) has been completed, but I would like to find anything useful about stuff like this.

Runout; theoretically a hydro spindle should produce virtually zero runout (although there is no such thing), or more accurately rotation concentricity should be held, within the limits of the roundness of the shaft, pressure consistent, and all other things being right. Now this doesnt mean at the tooling (grinding) surface you will have zero runout. What you should have is repetitive runout, virtually zero non-repetative runout. To dumb this down a tad, the shaft can be spinning perfectly true, but if the od taper that your hub sits on had 5 microns runout relative to the journals, you will still have the 5 microns.

In terms of dynamic runout, no not under load by grinding, but having the correct load at the bearing supplied by having the clearances all set right. When everything is set right hydrostatic bearings are super rigid, thats why they are used, they dont deflect much at all when working load is applied. My point is that runout, static or dynamic means nothing unless you absolutely know youre "loaded" (this term kid of confuses the topic-maybe stiffness is correct would be clearer?) -- essentially building the correct amount of pressure in the bearing to be held rigidly on center.

This is why it was so puzzling to me to have this instruction in machine acceptability document "measure the spindle runout at the front taper by gently rotating it by hand". I think what they are really asking to measure is the bearing clearance? Although I can't imagine it being set by any other means than a temperature reading near it's target setting.

Alternatively it might be a "beaurocratic thing". Here we have "machine tool acceptability criteria" set by the law (very long time ago). Things such as how much runout a "precision" machine spindle can have, table straightness and a hundred other things are defined there for various machines and their sizes as well as a way they are documented. It is a relict of a bygone era, but the advantage of those regulations is that every machine tool sold here has to come with this document. Including Chinese import machines, which filters the worst ones out the supply chain.

I sometimes wonder where this specific document format came from. I saw one using the same drawings in Russian, Czech, Polish, German and even Korean (for a North Korean lathe). If anyone knows, please let me know?

 

[TGTool]

Somewhere I was sure I had a picture of a bearing distinctly showing the lobed clearances but damned if I can find it now. I'd saved it more specifically because the poster was showing how to do the ID scraping. Still the 45 degrees from the axial line and alternating directions. I think he had gotten the whole thing to the right contact spots and had then followed with the slight relieving of areas that finally develop the hydrodynamic forces when rotating.

 

[luke8888]

Somewhere I was sure I had a picture of a bearing distinctly showing the lobed clearances but damned if I can find it now. I'd saved it more specifically because the poster was showing how to do the ID scraping. Still the 45 degrees from the axial line and alternating directions. I think he had gotten the whole thing to the right contact spots and had then followed with the slight relieving of areas that finally develop the hydrodynamic forces when rotating.

I sure would like to see that thread. Was it on this forum? Perhaps you remember some part of the title? Or specific wording used anywhere in the thread?

My spindle reconditioning was just a touch up of the shaft. I didn't do anything with the tri-lobed bushing. If extremely worn I think I would make a new one by turning it first then milling of the lobes. I had good results before with lapping a round bronze bushing with a purpose cast expandable lead lap. I'd probably do that with this one too to improve as turned surface finish and concentricity in its relaxed state. It would assume the tri-lobed shape internally once put under tension.

My old bushing has some wear that is the reverse of shaft's wear (it is about 10 microns / 4 tenths of an inch tighter in the middle). It did cross my mind to try and fix it, but I decided against touching it. No doubt I could achieve better stiffness by setting less clearance if it was straight, but that'll have to do for now. I need the grinder operational and I haven't got the time to make a new bushing at the moment. I'm hoping the tight section will wear itself true in time... (during start/stops).

 

[Robert R]

A answer to TGTool's question about the lobed journal bearing:
The machining process is not complicated. The lobed bearing is formed with a set of circular arcs. The bearing is first bored out on a jig bore to the diameter of the journal. The cutting radius of the boring bar is then increased by some small amount and held fixed. The bearing is then offset with, for example, .001" of table travel, and the boring bar is run through. The bearing is then indexed 120 degrees for a three lobed bearing and the boring bar is run through at each of the indexes. The profile of the lobed bearing will have its smallest radial journal clearance at the midpoint of the lobe. This allows the journal to be run in either direction without changing the oil pressure distribution in the bearing. My reference shows a oil groove machined between each lobe.
This description comes from a book " Design and Construction of Machine Tools by H.C. Town, 1971" It is describing a bearing used by a company called Gleitlager G.m.b.h. on page 81

 

[luke8888]

A answer to TGTool's question about the lobed journal bearing:
The machining process is not complicated. The lobed bearing is formed with a set of circular arcs. The bearing is first bored out on a jig bore to the diameter of the journal. The cutting radius of the boring bar is then increased by some small amount and held fixed. The bearing is then offset with, for example, .001" of table travel, and the boring bar is run through. The bearing is then indexed 120 degrees for a three lobed bearing and the boring bar is run through at each of the indexes. The profile of the lobed bearing will have its smallest radial journal clearance at the midpoint of the lobe. This allows the journal to be run in either direction without changing the oil pressure distribution in the bearing. My reference shows a oil groove machined between each lobe.
This description comes from a book " Design and Construction of Machine Tools by H.C. Town, 1971" It is describing a bearing used by a company called Gleitlager G.m.b.h. on page 81

Pity that book is quite rare. Is there a chance you can take a photo of the relevant page(s) and post a link (share via Google docs, Dropbox, imgbb etc)? There is so called "fair use" for copyrighted works and posting a couple of pages certainly falls into that (as long as its not hosted on the forum as an attachment, but posted as a link).

Regarding the bearing, it seems the book describes forming the tri-lobed shape in the inside of the bore during manufacturing. I may be wrong, but I suspect that's not how my bearing was done. I think the easiest would be to just make the inside circular and count on the deformation to form the lobes.

Here is how it looks from the back:

I haven't measured the taper, but it is substantial and the entire travel range is 20mm lengthwise. Also cranking on that nut requires ~160N/m (120 ft/lbs) of force despite putting a dusting of moly on the taper. I had to make a pin wrench like this to adjust it gently. (angle grinder pin wrench to scale)

No surprise previous owners were beating on that nut with the biggest hammer they had... (that's not how I treat my precision tools).

 

[mottrhed]

Thats a pretty basic/crude version. Likely not as rigid as the other styles as there isnt as much pressure built, but Im sure plenty rigid for the application. I agree the taper/crushing will deform the bearing and create some variance in the ID.

 

[johansen]

i researched the oil whirl problem a bit years ago, and found that the literature documents it as a problem when the diameter to oil clearance greatly exceeds the 1:1000 ratio commonly found in precision machine tools. and equally as much the literature talks about the fundamental frequency of the system and the journal bearing stiffness and dampening ratios. often systems are setup to cross over the resonances quickly and pass over them.

your 1-2 inch diameter lathe or grinder spindle with its .0001 to .002" clearance won't hit oil whirl numbers until the rpm exceeds practical limits. and i suspect many who think their machine had that oil whirl problem.. what really happened was it was out of balance and the rpm matched the fundamental resonance of the stiffness of their system. which is often lower than you think. some folks run 2-3" diameter main bearing crank shafts at 3 times the speed any engine lathe or surface grinder has ever run at, and they don't have a fundamental frequency oil whirl problem.

i am concerned it takes so much torque to affect the bearing shown in post #29.

if it takes 120 foot pounds to tighten it, how do you get that tapered bearing piece out of the tapered bore?

i once applied 800 foot pounds of torque to a 2.5" diameter maybe 12 tpi bearing retainer.. it turned out to be left handed and the casting blew up. oops....

i run low viscosity synthetic ATF which has a viscosity similar to velocite 6 in a southbend 9. it can hit 1500 rpm without warming up much, and has about a .001" clearance on a 1.875" spindle iirc.

if you can't get the clearance to close up as much as you need then run thicker oil and see if it works and doesn't warm up. the run out during operation will be as low as the spindle is round.. the oil film is self stabilizing and self centralizing, although under load the center point moves on you, but its a small predictable amount

any kind of dynamic movement from one turn to the next needs to be investigated if its the fundamental resonance of the shaft and the grinding wheel. its not oil whirl with surface grinder or lathe spindle oil clearances.

further more, regarding the oil whirl problem.. its not every turn. its a harmonic as i understand it. in order to make it work you have to have a swirling mass of oil follow the shaft round and round..
you can detect this by measuring the vibration and sweeping the rpm. you then have to account for the handful of different modes of vibration that could be happening at various frequencies. change the thickness or diameter of the grinding wheel and then test again.

thicker oil, running it slowly, will not damage anything. if problems go away.... well... problem solved.

 

[luke8888]

Thank you for an informative reply. Let me just clarify one thing. I have no oil whirl problem. Nor any other with that spindle at the moment. The only slight issue remaining, but I haven't noticed it affecting anything, is than perhaps I could set the clearance lower if the bushing didn't wear together with the shaft in a way that the bore constricts very slightly around mid point. If I did set that clearance lower, the resulting increase in stiffness could be beneficial when taking bigger cuts. However, this is quite theoretical, as I'm not noticing any problems with how it grinds. I expect when time comes to adjust that bearing again in perhaps a year or two that constriction should hopefully wear away.

i researched the oil whirl problem a bit years ago, and found that the literature documents it as a problem when the diameter to oil clearance greatly exceeds the 1:1000 ratio commonly found in precision machine tools. and equally as much the literature talks about the fundamental frequency of the system and the journal bearing stiffness and dampening ratios. often systems are setup to cross over the resonances quickly and pass over them.

your 1-2 inch diameter lathe or grinder spindle with its .0001 to .002" clearance won't hit oil whirl numbers until the rpm exceeds practical limits.

This is interesting. I read a text where they didn't say anything about the shaft diameter to clearance ratio, but they just stated "above 1800 rpm oil whirl starts to become a problem with circular bearings so other configurations such as crushed eclipse, tri-lobe etc are used instead" . The designers of this spindle must have thought similarly so they decided to go to the trouble of making that difficult to adjust tri lobe bearing instead of a circular one, like what I have in my lathe that reaches 1200rpm only.

The shaft of the grinder is 35mm in diameter, the lathe is indeed 2in (51mm) if I remember correctly.

and i suspect many who think their machine had that oil whirl problem.. what really happened was it was out of balance and the rpm matched the fundamental resonance of the stiffness of their system. which is often lower than you think. some folks run 2-3" diameter main bearing crank shafts at 3 times the speed any engine lathe or surface grinder has ever run at, and they don't have a fundamental frequency oil whirl problem.

I just found out what oil whirl is when researching that "weird" bearing before I decided how to refurbish it. If I had vibration or any other unexpected movement in the shaft I'd assume out of balance or a resonance condition most likely. There are really cool android aoos that let you use the accelerometer in your phone to measure vibration very precisely and desplay it as a frequency/power chart or in other ways. The one I use is VibSensor.

i am concerned it takes so much torque to affect the bearing shown in post #29.

Yes, I was too, I thought that's just silly, but I see no other explanation. At the end of the day when adjusting it we're squishing a solid tube of phosphor bronze that has walls thinned in a way that it squishes into a tri lobed star. But the "thin" wall thickness is still over 3mm. A lack of slots cut in that bushing to allow it to close easily is one of the "weird" features of this bearing.

However, with my pin wrench, and all 4 pins inserted applying that torque can be done very precisely.

if it takes 120 foot pounds to tighten it, how do you get that tapered bearing piece out of the tapered bore?

You can just leave it and run the spindle for a couple of hours or gently tap the front of the nut with a brass rod. It doesn't take much for it to spring back. It was like this even before I applied moly to both surfaces. It usually behaves like this
- immediately after the spindle is put together there is clearance between the brass bushing and the tapered bore, the nut turns freely.
- eventually the clearance is gone and it takes more and more force. It goes from "free turning" to that huge torque required within half a rotation.
- then you can continue turning with the high torque. I tend to measure the thread remaining in the nut when recording my settings. When all of the thread is gone, that is the end of travel. If I remember correctly there are around 5~6 rotations after it gets tight to reach my current setting. I still have about 5mm of adjustment left.
- if you want to back out you turn the nut the other way. The first tiny amount of backing it out requires some torque(nowhere near as much as to turn forward), then you can spin it by hand. Then you can leave it running or tap it gently on the nut varying the place you tap around it's circumference. It is very important to only loosen it a turn, no more, then tap, loosen more, then tap and so on. Never loosen by many turns then tap as this IMO risks the bushing going back unevenly and jamming.

The brass rod weights maybe around 150g (a third of a pound). The strength of the taps is comparable to what I use when I'm making a precise punch mark on mild steel. So not a lot of force.

i once applied 800 foot pounds of torque to a 2.5" diameter maybe 12 tpi bearing retainer.. it turned out to be left handed and the casting blew up. oops....

Wow, that's a lot of torque... I have my share of cracked castings too. Thankfully nothing important.

i run low viscosity synthetic ATF which has a viscosity similar to velocite 6 in a southbend 9. it can hit 1500 rpm without warming up much, and has about a .001" clearance on a 1.875" spindle iirc.

I run velocite no 10 in my lathe. 2in shaft at 1200rpm,similar clearance. The lathe does warm up if used all day long on top speed. Not excessively so. It feels warm to touch, that's all. Maybe 40C?

I tend to run the liquids specified in the manuals (or their equivalents).

if you can't get the clearance to close up as much as you need then run thicker oil and see if it works and doesn't warm up. the run out during operation will be as low as the spindle is round.. the oil film is self stabilizing and self centralizing, although under load the center point moves on you, but its a small predictable amount

This is an interesting idea. I don't think I'll do that as it is fine for me now, but it sure is something to think of in case it's needed.

any kind of dynamic movement from one turn to the next needs to be investigated if its the fundamental resonance of the shaft and the grinding wheel. its not oil whirl with surface grinder or lathe spindle oil clearances.

Yes, but detecting that movement is not that easy (especially when small). As others mentioned it should ideally be done under load (so not just spinning, but during grinding). Of course surface finish can tell you a lot, but having numbers would be neat.

further more, regarding the oil whirl problem.. its not every turn. its a harmonic as i understand it. in order to make it work you have to have a swirling mass of oil follow the shaft round and round..

The text I read said it is half the speed of rotation.

you can detect this by measuring the vibration and sweeping the rpm. you then have to account for the handful of different modes of vibration that could be happening at various frequencies. change the thickness or diameter of the grinding wheel and then test again.

thicker oil, running it slowly, will not damage anything. if problems go away.... well... problem solved.

Running it slowly? My grinder has only one speed... (no vfd). If I was bothered by that constriction I'd grind the ID of that bushing or lap it. But I still think this is a useful idea to have in ones toolbox.

 

[TaperPin]

The only slight issue remaining, but I haven't noticed it affecting anything, is than perhaps I could set the clearance lower if the bushing didn't wear together with the shaft in a way that the bore constricts very slightly around mid point. If I did set that clearance lower, the resulting increase in stiffness could be beneficial when taking bigger cuts.

As a novice bearing scraper only half way through my first lathe spindle bearings I can appreciate the situation you’re in. Trying to measure runout in worn out, mis-shaped bearings paired with a spindle of unknown condition is disappointing at best. You sound dedicated to the task, and in no rush for immediate results, so I’ll pass on the highpoints of my very limited experience - not for technical expertise, but for encouragement.

You are exactly where I was a few months ago. We all have to crawl before we walk and walk before we run. I had to get good enough measuring tools to admit to myself it was fruitless hoping for .0001” readings when the spindle was adjusted as close as I dared and deflected many .001”.

Looking at a grooved abused spindle, it was a punch in the stomach the first time I blued it up for bearing contact, but that was the moment I knew there was a lot to learn about scraping. Very little scraping experience and it becomes obvious that a grooved spindle has to be precision lapped. Precision lapping made it obvious my homemade scraper and technique needed improving. A good scraper made it clear that my sharpening stones weren’t cutting it. Something approaching a repeatable scraping pattern showed I wasn‘t good at moving the direction of the spindle nose. Lately I learned about multi lobed bearing contact to better hold and more gently distribute oil - from a video of huge rock crushing equipment in India, then another scraping video from 1942, then from half a dozen other sources that I had glanced over, but hadn’t registered. A tapered bearing doesn’t seem all that different from a cylindrical one as far as contact patterns and oil film thickness goes, but maybe I don’t know what I don’t know.

A year ago I wish someone had suggested I focus on getting the spindle as perfect as it could be before anything else, but I wasn’t ready for that on many levels. Whatever path you take I hope you look back and appreciate the journey.

Thanks to those who share their knowledge - there are many of us who never speak up, but are forever in your debt for the breadth and depth of it all.

 

[kamikazekato]

I'm in a process of reconditioning an old surface grinder that uses a hydrodynamic bearing as its main spindle bearing. The problem is that at speed the spindle relies on an oil film to provide rigidity and support. So slow turning by hand results in very weird measurements. I've been trying to run the spindle for a bit, hit the off button and try to measure as it comes to a stop, but I'm not entirely sure my test indicators can measure the full extent of a still fairly fast movement.

I'm supposed to set TIR to under 4 tenths (10 microns), but I can't if I'm not able to measure it properly... The machine manual doesn't specify any special method of measurement. There is a, drawing showing a meter touching the front taper and that's it. There are people that rebuild those spindles for a living and they are understandably tight lipped. So I decided to ask here.

For those interested in the details. The bearing uses a stationary cast iron taper into which a solid bronze bushing is pulled in by a clearance adjustment nut. This bearing is different from your typical plain bronze bearing in few ways. The most important are:
- no oil supply notches on the bronze bushing. Instead the shaft has a spiral groove which pumps oil under pressure into the gap.
- no slits cut in the bronze bushing to allow it to compress evenly etc. Instead some material is removed from the outside allowing it to deform when pulled in by the nut. As a result a lot of force is required to adjust it. Also the bushing is far from circular, but it appears to work well.

I have the usual assortment of dial and test indicators that measure down to a micron.

Edit: Alternatively, perhaps someone knows what the clearance for the oil film should be in such bearing? It runs as 2700rpm. The bore is 35mm (1.377in) and it uses iso vg 6 oil. Currently I have 30 microns of clearance and using my flawed measuring methods it seems the running TIR is ~8 microns.

Hello, Not being at the shop to talk to my 'Sleeve bearing' guy to verify clearances, I am a manual machinist at a electric motor shop.But just going by what you have sounds reasonable to run. Definitely run without a wheel on it first, and just do the touchy-feely around it. Without a wheel you may get some vibration. If it passes that, install wheel and dress it. Let it run for 10-15 minutes or so, and dress the wheel again.By that time all such temps and oil should be happy, and by dressing the wheel you are actually balancing it.Start by grinding the magnet to make parallel to head.If you can find some tool steel(O-1,D-2, or such) nothing soft, and start grinding. look at the pattern you are getting(you really shouldn't see one).

 

[luke8888]

@TaperPin and @kamikazekato thank you for the replies.

The spindle that is the subject of this thread has been running for last month a lot and I'm much happier about it than before. There is still room for improvement (better rear bearings), but for now I'm leaving it as is. However I'm monitoring what's going on with the oil frequently (checking for bronze particles).

Your post reminds me of my first non roller bearing spindle rebuild (this was my old lathe). Back then (~6 maybe 8 years ago) there were no online videos showing how to scrape bores and I wasn't very hopeful about figuring it out as I go so I didn't even consider scraping.

It used to be pretty common around here to find machine tools with spindles that are worn, but no one ever bothered to adjust spindle bearing clearances (gibs and ways are another matter).

My lathe was like this with ~4 thou of runout and the worst was it was pretty random. There was a lot of uneven wear on the bushing so it couldn't be tightened as is. The shaft was very surprisingly in very good condition. Either there was a lot less wear overall or the previous owner reground it. I bought a piece of cast bronze tube to make a new one, but I also tried to fix the old one as a "learning exercise". I first took a very slight cut with a boring bar (this was on my other very small lathe). Despite some chatter I was much happier to have "only" about a thou of taper and a bad surface finisg to fix. I cast 3 expandable lead laps (that's two of them still my "random stuff to reuse" box)
And I went to lap. I did have to recast and recut them to take that taper out, but in the end they did the job.

I used those German made abrasive pastes in various grits. They're made by Diamant and they are just called "an oil soluble paste". I think they contain silicon carbide particles.

For the first runs of the spindle I was very anxious. But I just set it loose first. I let it warm up (it took a very long time) going up in speeds one by one.

Then I started tightening it. Running it for 20 minutes and measuring the temperature. (with an IR thermometer - a very useful tool). Unfortunately I did manage to stall it once, but the clutch not being fully engaged saved it from having to be reworked. I'll never forget how this sounds. The machine sounds normally and suddenly you notice a very slight change in tone, by the time you have time to react and stop it it seizes. I hope to never hear that sound again. In the end I was happy with the result. About a thou of runout and very repeatable. I still have the bronze tube waiting for it's use...