Page 2 of 2 FirstFirst 12
Results 11 to 19 of 19

Thread: Italian mystery mills, and spindle bearings therein

  1. #11
    Join Date
    Mar 2019
    Location
    W. Nowhere, CT
    Posts
    10

    Default

    Quote Originally Posted by J Tiers View Post
    Perhaps I am looking at a different part, but the bearing balls I see in the picture do not look so bad.

    They have a little more scratching than usual on the very best bearingss, but they are "used".
    I'm actually not particularly bothered by the way they look - hell, they look nicer than most of the other parts of the machine. The way they feel, though, is another matter. Unfortunately the forum software here is a little on the old fashioned side, and I couldn't find the standard "Link Tactile Sample Here" button in the toolbar, so I had to compromise and post a picture of the balls instead, vaguely hoping people might be able to extrapolate from that. A video didn't do a very good job capturing it either, but I haven't completely ruled out interpretive dance.

    The bearing itself appears to be a double row, and may likely be a "zero clearance" or "internally preloaded" bearing, similar to what was used on Logan lathes.
    Yep, they're double row angular contact bearings (at least the lower one is - I suppose the upper one could be a a double row deep groove and I wouldn't know the difference) which, if I understand properly (and I probably don't), inherently have the rows preloaded against each other.

    They do appear to have the remnants of ancient grease still stuck to them in places (the dark stuff). That can be removed. I tend to use lye-based cleaners, because they have no tendency to rust anything or otherwise discolor steel.
    Unfortunately that's not grease - I really wish it were. Or rather, if those dark bits on the balls are grease, it's grease that's trapped by the rough surface down in cracks or grooves, not sitting on the surface. The bearings will usually turn freely (at least unloaded), except when they catch on one of the rough bits and stick or crunch. I kept trying to convince myself that it's not that bad, and a little crunching is perfectly acceptable in mill spindles, but I have yet to find anyone else willing to reinforce my delusion.

    It's a tricky thing to capture in a picture, since the balls are small, round, and reflective (and I'm impatient, shooting on a phone without macro or manual focus, and a lousy photographer), but these three pictures are closeups of a few of the more visibly worn balls from the lower bearing that, zoomed to full scale, might make it a little easier to see the problem.

    I have heard of people using Pine-Sol, and swearing it cleaned everything up perfectly, but cannot confirm that. It discolored the only bearing I ever tried it on, and I will not use it ever again unless someone can explain the discoloration in a satisfactory way.
    I've mostly been using mineral spirits with the occasional WD-40, PB blaster, and Evapo-rust, all of which have thoroughly discolored the natural deep browns, blacks, and reds of old grease and rust - but I haven't yet caught any of them changing colors in ways I didn't want. I'll have to remember the warning about Pine-Sol and bearings, although I don't usually have any lying around. Pine-Sol, that is - obviously I've got a few bearings lying around.


    -Bats
    ( a few balls short of a bearing )

  2. #12
    Join Date
    Jan 2004
    Location
    Missouri
    Posts
    30,713

    Default

    OK, those last two pics really show it clearly. Very good. (very bad, actually).

    It initially appeared that those were stuff stuck "on" the balls. a 2D picture does not give a great deal of depth perception, the pic from an earlier post in the thread seems to shift back and forth in appearance, to me. The clearer pics show that they are cavities, that the bearing has spalled, and it is "done". Needs replaced.

    That is the failure mechanism of ball bearings and roller bearings. Metal fatigue that leads to spalling off of the surface like that. It depends on load, and happens after a certain number of rotations of the shaft or spindle at that load. The makers have graphs of the life vs load and number of rotations. Could have started as corrosion, or may be legitimate fatigue spalling. Once there is one pit, the edges start to crack off also, and in a short time it is serious toast.
    Last edited by J Tiers; 03-16-2019 at 12:03 AM.
    1601

    Keep eye on ball.
    Hashim Khan

  3. #13
    Join Date
    Jan 2016
    Posts
    156

    Default

    if the spindle was welded and then reground too much oversize, then the entire clearance of the bearing would be taken up and the bearing will destruct.

    so that cheap C3 clearance bearing may actually be what you need.

  4. #14
    Join Date
    Mar 2019
    Location
    W. Nowhere, CT
    Posts
    10

    Default

    It initially appeared that those were stuff stuck "on" the balls. a 2D picture does not give a great deal of depth perception, the pic from an earlier post in the thread seems to shift back and forth in appearance, to me.
    Yeah, I had trouble finding any single shot that gave a decent sense of depth, but didn't want to overload the first post with, essentially, a bunch of pictures of exactly the same thing. The second collection definitely shows it better, but takes several shots to do it.

    The clearer pics show that they are cavities, that the bearing has spalled, and it is "done". Needs replaced.
    Yep... that's pretty much the conclusion I started out with - I just needed some help to completely crush the nagging hope that maybe I was wrong. I think at this point the only questions left are how, with what, and whether that 'what' should be shielded or open.

    That is the failure mechanism of ball bearings and roller bearings. Metal fatigue that leads to spalling off of the surface like that. It depends on load, and happens after a certain number of rotations of the shaft or spindle at that load. The makers have graphs of the life vs load and number of rotations. Could have started as corrosion, or may be legitimate fatigue spalling. Once there is one pit, the edges start to crack off also, and in a short time it is serious toast.
    So clearly the problem is that the manufacturers didn't use a large enough piece of paper for their graphs, and the bearings fell off the edge. Damn penny-pinching bastards couldn't even be bothered to tape on an extra sheet to extend it.


    -Bats

  5. #15
    Join Date
    Mar 2019
    Location
    W. Nowhere, CT
    Posts
    10

    Default

    Quote Originally Posted by johansen View Post
    if the spindle was welded and then reground too much oversize,
    You mentioned this once before. While it would explain a few things, I'm still curious what brought you to that conclusion.

    There does seem to be some fairly obvious welding that was done, but it doesn't look to have been in the service of a complete remachining. On closer inspection of that weldy-looking squiggle with the grinding around it, it looks like what happened was a slit was cut (rather sloppily) through the wall of the spindle and a little piece of keystock was welded (or maybe brazed?) through the hole to the inside of the taper as a collet index pin. I'd be more worried about the strength of the weld or braze, except for the number of recommendations I've seen to forcibly remove the damn things anyhow.


    This is another angle on the squiggle, which again makes that particular work (damage) look localized:



    then the entire clearance of the bearing would be taken up and the bearing will destruct.

    so that cheap C3 clearance bearing may actually be what you need.
    It looks like the situation might actually be a little more complicated than that - or possibly backwards.

    You can see when the spindle was first pulled that the bearing wasn't actually flush against the bottom collar:


    ...it turns out may be a reason for that. When it is pushed down flush, it does this:

    The bearing must be installed backwards - the wrong part is turning.

    So it looks like the spindle is distinctly undersized at the end, tightens up over the ugly bit (if you for the red reflection in the third picture, that seems to be a high spot that was polished smooth by the bearing on its way off), and then very gradually tapers down over the rest of the length. Once it's past the grinding, it's just a hand fit - although the inner race doesn't twirl again until about 2/3 of the way to the other end. I still need to make some measurements of the housing - I'm not sure yet whether this means the spindle was spinning in the bearing and only shoved up to the tighter position in the course of removal, or that it had been deliberately installed that way from the beginning (with "beginning" meant in the loosest possible sense), but my suspicion is that going with even higher clearance is likely to be a bad thing.

    There's also a discussion to be had about installing the new bearing without a press (or the fancy thousand dollar bearing fitting kit that SKF recommends), since I took this one off by working my way around the inner race with a trashed brass punch (honest, officer, it was trashed when it got here! I never served it, and had absolutely no idea it was underage!), which is probably not recommended and definitely not very efficient. That can probably wait for another thread, though.


    -Bats

  6. #16
    Join Date
    Jan 2016
    Posts
    156

    Default

    So you're saying if you press the bearing all the way on the spindle its a loose fit? but about 1/4" away from the end, its a tight fit?

    my speculation on what happened is that they had to turn it to clean up the weld for the keyway. Chucking it up in say, a 3Jaw chuck, when they turned the weld off.. they probably kept going all the way to the end.. because the weld distorted the spindle.

    so that spindle is probably more like this () than round, by however much the run-out was when they chucked it up to turn off the weld. (couple thousandths probably.) and the lathe wasn't turning straight so there is a taper.

    if you put the spindle on V blocks you can use a test indicator to measure the run out on the taper. the V blocks do not need to be precise. 4 nails in a block of wood is actually good enough. you need a 5th point of contact on the "V block" to contact the flange of the bearing seat so the axial location of the spindle is defined to measure the run-out on the taper.
    Last edited by johansen; 03-17-2019 at 03:05 PM.

  7. #17
    Join Date
    Mar 2019
    Location
    W. Nowhere, CT
    Posts
    10

    Default

    Quote Originally Posted by johansen View Post
    So you're saying if you press the bearing all the way on the spindle its a loose fit? but about 1/4" away from the end, its a tight fit?
    Exactly. And then it loosens up again afterwards.

    All the way down it's toward the looser extreme of clearance fits, by the time it's 1/4" away it's getting into tight press fit territory, which extends for about the next 1.5". After that it still doesn'tturn freely, but it's no more than hand tight.

    my speculation on what happened is that they had to turn it to clean up the weld for the keyway. Chucking it up in say, a 3Jaw chuck, when they turned the weld off.. they probably kept going all the way to the end.. because the weld distorted the spindle.

    so that spindle is probably more like this () than round, by however much the run-out was when they chucked it up to turn off the weld. (couple thousandths probably.) and the lathe wasn't turning straight so there is a taper.
    Possible, except that it looks like what they primarily used for cleaning up the weld was an angle grinder (or maybe a die grinder?). The fact that there was a high patch left in the middle of the grinding to be polished smooth by the bearing's removal makes me think that, while it was obviously turned at some point in time, it must have happened before the welding.

    This shows the ground area a little more thoroughly, along with the almost mirrored-looking polished patch that crosses the camera just before the squiggle (it's most visible just before it rolls under the bottom):


    Based on where that polished patch is and the fact that it coincides with where the bearing was tightest, I think the issue was a single high point (well, patch) rather than a (necessarily) out of round and/or tapered spindle. That portion of the shaft does seem to taper slightly from one end to the other (probably by less than 10 thou over the whole 8" length) but I don't think that had much bearing (no pun intended) on anything else. My initial assumption was that it was designed that way so that bearings would only have to be press fit at the end, rather than forced along the entire length.

    if you put the spindle on V blocks you can use a test indicator to measure the run out on the taper. the V blocks do not need to be precise. 4 nails in a block of wood is actually good enough. you need a 5th point of contact on the "V block" to contact the flange of the bearing seat so the axial location of the spindle is defined to measure the run-out on the taper.
    Thanks. I meant to ask last time you brought this up about just how accurate the v-blocks really needed to be for the operation and whether I could get by with cutting something approximately the right shape out of aluminum on a light CNC router or hardwood on a bandsaw (nails hadn't even occurred to me, but I've probably got a few of them around here somewhere too). Intuitively it seemed like I should be able to use just about anything, but my intuition tells me lots of crazy **** that'd cause dreadful injuries, produce not entirely reliable measurements, or get me arrested if I started listening to it.

    I'll see if I can get some numbers tonight or tomorrow & see what the situation looks like (both inside & outside)... assuming, at least, that I can still operate an indicator with my fingers crossed.


    -Bats

  8. #18
    Join Date
    Jan 2016
    Posts
    156

    Default

    the surface finish from top to bottom seems to be the same. perhaps the weld shrunk the spindle enough that the bearing is no longer a press fit.

    anyhow i think the idea of buying 100$ bearings is a waste of money. find some cheap ones and get it running again. you can knurl the spindle to make the bearing a press fit if you really wanted to, or just epoxy or loctite the new bearing on.

  9. #19
    Join Date
    Mar 2019
    Location
    W. Nowhere, CT
    Posts
    10

    Default

    Quote Originally Posted by johansen View Post
    the surface finish from top to bottom seems to be the same.
    More or less. The feeds & speeds may have differed (the "thread pitch" of the turned surface differs slightly on the different thicknesses), but aside from the outside of the the collar at the bottom, the inside of the taper, and that messy patch around the "weld", it's all essentially the same.

    perhaps the weld shrunk the spindle enough that the bearing is no longer a press fit.
    Except that the "weld" (which I'm thinking is actually a braze - and not a very solid one at that - seeing as there doesn't appear to be any sign of melting on either the spindle or the keystock, just the material filing the cutout, half of which came out in the course of cleaning & bearing removal) doesn't continue down to where the bearing sits. In fact, passing over that area is the only place it is a press fit. Above that point it can be shoved on/off by hand, and below there it can spin freely.

    anyhow i think the idea of buying 100$ bearings is a waste of money. find some cheap ones and get it running again.
    Shhh! Don't let the Bridgeport owners hear you say that!

    (earlier today I found some nice shiny ceramic bearings... at the price they were asking, I could buy a used Bridgeport and new bearings for it)

    Right now I'm waffling between ABEC-3 (the recommendation of a local machinist friend, which I've found at $75 for the pair after some rather excessive $12 shipping) and ABEC-wecan'tbebotheredtospecify (which includes NOS - although I don't know whether I should worry about the already aged grease in sealed ones) for maybe two thirds of that.

    you can knurl the spindle to make the bearing a press fit if you really wanted to, or just epoxy or loctite the new bearing on.
    I did think about knurling, but was turned off by the fact that it would presumably require regrinding the taper afterwards. Of course, at this point it's looking like that's going to be required regardless. Aside from the fact that pretty much every way of making the bearing fit looks like it's going to throw off (or at least fail to preserve) concentricity, I finally got around to tossing the spindle up on v-blocks last night. The good news is, I'm not seeing any sign of being bent, and it's no more than half a thou out of round where the bearings sit (possibly less, but between the rough surface and bargain bin .001" indicator, that's about as close as I can say). The bad news is inside the taper, which shows about 4 thou runout.

    I've got a local friend (who conveniently is much better equipped than I, and, even more conveniently, is much less clueless) who'll hopefully be taking a look at it later this week to consider options. In addition to the weld-and-remachine route (whether or not something similar already happened at some point in the past) he was tossing around the idea of either plating or some sort of flame spraying procedure that I'm not entirely clear on.


    *sigh* I was really hoping I could get through this with just a simple teardown, cleaning, and reassembly.

    I suppose I should've known better. "Just cleaning" my South Bend required a whole new gearbox before it was done.


    -Bats

Tags for this Thread

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •