What's new
What's new

1964 FP2 horizontal spindle adjustment

thanvg

Hot Rolled
Joined
Mar 3, 2015
Location
Greece
Dear all hi,

allow me to present you with my problem and ask for the forum's assistance. I am sure the problem I am facing is, most probably, trivial, but bear with me and provide your feedback if possible.

A couple of weeks ago I noticed an enormous axial play at my FP2's (500 mm X version), around 0.2 mm. I can't tell if this was here since I got the machine (1.5 years ago), or whether I induced it when I opened the spindle for the Kluber treatment, or whether it showed up at some point. I hadn't used the horizontal spindle much, and only for boring/drilling, that's why I hadn't noticed the play before.

I checked my manual and I was happy to find that axial play is adjusted via a threaded ring, no need to grind spacers etc (there was a relevant mention in a similar thread here, but I can't relate to that I think). My problem is that I can't seem to be able to ajust it, no matter what I do. Here is a magnified picture of my manual:

attachment.php


To my understanding, we have a spacer C held captive on the spindle, sandwiched between the spindle shoulder (at one side) and spacer F, inner race of the top roller bearing and the nut 9 (at the other side of the spacer-collar C). So, since collar C is fixed on the spindle, we have quill shoulder A and thrust bearing B that hold the spindle from exiting the quill, taking axial play in the outwards direction. And then there is the set consisting of thurst bearing D, roller bearing outer race E and threaded ring 6 that push the collar towards the spindle nose and takes up all axial play.

My problem is that, no matter how tight I get the threaded ring 6, I can't get the axial play to get any smaller. Resting the spindle on it's nose I can lift the quill up and it does rise up. So, it's like the threaded ring and the parts it is pushing cannot reach the collar C to push it towards the spindle nose. But I can't see anything interfering there to cause that. Has anyone faced something similar or has any idea what might be the problem here?

Here is also a scan of the full page of the manual:

attachment.php


Last, I would like to ask whether my understanding is correct: in order to set the thurst bearing properly, spindle nose should have a bit of a clearance from the quill, right? Something like this:

attachment.php


Thanks in advance for any assistance guys. I know it's no rocket science, but I am at a mental block here currently.

BR,
Thanos
 
Some thoughts:
First off the ball thrust bearings, B and D are directional. I am mentioning this because you have had his assembly apart i assume to grease the spindle....

If the ball thrust bearings are not in the right direction it will affect your ability to adjust the end play i think....
The ball thrust plates for each baring are different sizes....One plate is made to be a precision fit on its OD while having clearance on its ID...The other plate is the opposite.
It will be smaller than the housing bore , while its ID will be a fit on the spindle.....

So bearing "A" should fit tight to the housing (quill) and have clearance to the spindle on its most forward plate....Then come the balls and finally the rear most plate of bearing "A"
should be opposite....Having a fit on the spindle (snug) while providing clearance on its OD to the quill....

This setup allows the bearing to have one stationary plate (forward) and one that is rotating with the washer and the spindle....

The washer should be s snug slip fit on the spindle (rotates with the spindle) and have clearance to the housing (quill)
Finally there is bearing "D" it should be stacked opposite to the forward thrust bearing....So the forward plate of bearing "D" should fit the spindle and clear the housing (quill)
Then the balls, and finally the rear plate, fitting snug to the quill and having clearance to the spindle....

Also, locking ring #9 needs to be tight as well (not an adjustment) , but if memory serves it locks up or captures the collar "C" in the thrust pack....

Cheers Ross
 
Hi Ross, thanks for chiming in.

Regarding positioning of the thrust bearings, I learned that they are directional back from the ALG100 days when I had the spindle apart quite a few times. So, yes, I paid attention to the following (assuming that the spindle is vertical, standing on it's nose, so that we have up and down to orient ourselves):

- bottom plate of bearing B fits snugly in the quill and rests on shoulder A.
- then come balls
- top plate of bearing B fits snugly on the spindle
- collar/spacer C fits snugly on the spindle and has generous clearance from the quill
- bottom plate of bearing D is loose on the spindle ! (no tight snug fit here, but I guess it does not matter, does it?)
- then come balls for D
- top plate of bearing D fits snugly in the quill (actually it fits snugly in a sleeve that fits snugly in the quill)
- finally comes the outer race of the roller bearing and the threaded ring.

So, only thing that's 'strange' here is that none of the plates for the upper (D) thrust bearing fit snugly on the spindle, but I can't see how this might cause an issue here.

Also, I did the following experiment: I removed the spindle but kept the rest of the parts in the quill. So I was left with the thrust bearings sandwiching the collar C and the whole system being pressed onto the quill's shoulder A by the threaded ring. So, as I tightened the threaded ring, then indeed collar C was starting to get compressed and was difficult to move radially (clearance to the quill is very large so it's easy to move the collar radially when it's free). So, indeed, threaded ring can apply preload on the bearings/collar, at least when the spindle is not there. Given this, since collar is captive on the spindle when everything is put together, spindle should have no axial play...but it does....

Also, at the first phase of assembly I compared the height of the spindle's shoulder to the top plate of the bearing B, having everything in the quill of course. Indeed, spindle's shoulder is a bit higher (.6 mm) than the top of bearing B, as expected. So, when the collar comes in and rests on the spindle's shoulder it has somewhere to travel to when it is compressed from the top by the bearings/threaded ring combination. Collar should be pushed downwards dragging the spindle along by the shoulder. So, sanity check passed here as well.

What I think I will try next is to measure the distance from the top of the threaded ring to the top of the quill after snugging the ring. I will take the measurement with the spindle out, checking that the collar is indeed compressed and then with the spindle in. If I see that the threaded ring ends up higher for the latter case, then something stops it from threading further when the spindle in installed.

Any extra ideas would be appreciated. I wish I could discuss it with someone in real life, with the parts at hand and not asking you to burn your head trying to imagine what's going on there through text and a couple of pics...I am sure it's something simple, I just can't see it currently...

Many thanks,
Thanos
 
Thanos:
Sounds like you have things assembled properly...
Additional question: Is the thrust washer that sits between the two thrust bearings made as a plain washer or is it made with a shoulder on one side...Have seen different versions of
the thrust setup on these machines...
So if the washer is just that, a plain washer, is it possible that it has a relief on one side (chamfer) of its center hole where the spindle passes through....
If so perhaps it needs to have that chamfer facing toward the spindle nose....allows the washer to go full to a shoulder step on the spindle.....(clearance for the shoulder fillet)
Just a thought.....
Cheers Ross
 
Hi Ross,

thanks for the reply.

I think I found the issue, I'll post later, I just came home.

BR,
Thanos
 
Hi to all,

so, after a bit of extra investigation, I figured out that what I am seeing could only be caused by the collar (C) not being captive to the spindle, as it should. So, with the spindle standalone I inserted the collar C, then the T-cross-sectioned spacer, then the roller bearing inner race and then lock-ring and nut. Snugged the nut and confirmed my suspection, collar and spacer could move around 0.3 mm axially....Here is the configuration, arrow shows the axial play:

attachment.php


The thing is that I cannot take up this play by tightening the nut because the roller-bearing inner race bore is tapered! It matches a tapered section of the spindle, that can be seen here:

attachment.php


Now there are two issues concerning me:

1. I think that the collar is possible to have been ground thinner. Grind marks on the two faces are not similar and it also measures uneven all around, from 4.76 to 4.80mm. Here is one face:

attachment.php


and here is the opposite face:

attachment.php


Also, chamfer is smaller in the former face than the latter.

So, initially, I assumed that, for some strange reason, someone ground the collar down in the past and I should correct it. But as I am typing right now, there is something I don't get, and this is the second issue.

2. If the stacked height of the collar plus T-sectioned spacer is not exactly spot on, there are two options: if the spacers' height is smaller then I get the axial play I am seeing. If it is larger, then the inner race of the roller bearing cannot be pushed all the way down and in this case the tapered bore will not mate with the tapered section of the spindle!!! So, we'll have removed axial play but introduced radial play on the top (rear) roller bearing. Well, how about this now?

I guess I am still missing something, since, as far as I can tell, in similar cases things are adjustable by shims, spacers or nuts; I would not expect that design called for extreme accuracy on the stacked height of two spacers or else we get radial play!

I think I might have to contact F. Singer and ask about that collar, whether it's availabe and what's its width....

continuing in next post...
 
Finishing with some notes, for completion:

1. earlier I mentioned that none of the two plates of the higher thrust bearing are a snug fit to the spindle. This of course happens since there is clearance for the T-sectioned spacer that has to fit in there as well...

2. No matter how many times I came across this bearing, I always ready 'china' here :)

attachment.php


Please share any ideas, if you have any...

BR,
Thanos
 
First thing i would do is check the clearance of the roller bearing to its outer race when assembled.....
Bearings like that (roller that is) are designed so that tightening the bearing on the shaft pushes it further down the taper and makes the inner race "swell a bit" and tightens the running clearance....

Should be assembled with some way oil on the taper to allow it to move...thinking that you are simply not tightening the
retaining nut enough.
In a perfect world that spacer is ground to set the position to the bearing and thereby set the running clearance....
If the bearing has been changed some time and not adjusted for proper running clearance the spacer might be too short....

To set the bearing clearance i would think that it needs to be tightened till the side play on the spindle as assembled runs somewhere around .0002-.0004" of side play measured close to the bearing ,with no grease on the bearings (light oil for test assembly.

Need a proper spanner for tightening that retaining nut...might need to be tighter than you think to get everything correct.

So as summary the bearing fit sets the bearing position, and that determines the exact length of the spacer....

Cheers Ross
 
Thanks Ross, all this makes sense.

Intuitively I don't feel that I can tighten the nut to take up the 0.3 mm slack, at least without splitting the bearing in two, but that's the path I am going to follow.

So, I'll put back the vertical head, without horizontal spindle in the machine (:( ). I'll turn plugs for the quill side and the rear side to keep things from entering the headstock. I need the vertical spindle on to make the spanner, that's what happens when you only have one milling machine :)
I'll make a proper spanner, with a sleeve to keep the locking rign centered as per manual's suggestions. I'll have to file down the nut a bit first, neither the previous owner nor my used a proper spanner till now....

Then I'll do a feasibility test with the spindle out: I am going to put the collar, spacer, inner race (with a bit of oil as you suggested), lock-ring and nut on and start tightening to see if I can, indeed, reduce the gap. Success or not, I will not down either torque value or distance from top of spindle.
Then, repeat the test with everything reassembled. The goal is to reach the axial distance of no axial play and measure proper radial clearance at the same time.
(I just don't want to think about the senario where I will have tightened beyond reason and still have too much of radial and axial play....bearing does not seem worn so I hope I won't end up there)
on edit: CORRECTING MYSELF: I 'don't really care' about the axial play, I'll compensate for this with a thicker collar, it'll only require some extra waiting. What I really need to see is that by reasonably tightening the nut I am getting proper radial clearance. That's the important thing, or else it's going to take a lot of time and money to fix....

Many thanks!
BR,
Thanos

P.S. I checked the manual for the second gen (400 mm X) FP2. Apart for the extra roller bearing (3 in total!) it's much simpler. No tapers there.....
 
The tapered inner race for the bearings is a feature that Deckel used it seems for a short time. And not sure it was used on the vertical.

Before doing anything do a test. Setup your spindle on the machine, set to be flat and parallel with your "X" axis movement....
Use a good test indicator and record the rate of taper on the bearing seat of the spindle (read rise over run, calculate taper) this way you will have a good idea of how much the bearing will expand for a given
movement , so that when you test assemble and measure your radial clearance you can get a good idea of how much longer or shorter the sleeve needs to be to give a reasonable radial play....

Might hope that the manual would tell you this information along with a side play specification, the values i suggested above are just an educated guess.....

Make sure that you oil the thread and faces of the washer on the retaining nut (use way oil) as well as oil on the seating surfaces of the bearing and spindle.

If you make the retaining nut tight, and the radial clearance seems to be in range and still have axial play, then guess you will need to make a new spacer, of add some shims. I would vote for new. Not a big fan
of stacking thin shims in an assembly like this.

Keep us informed as to how this works out....
Cheers Ross
 
Hi Ross,

thanks for the extra advice on measuring the taper and oiling everything.

I din't have enought time to make it to all these today, but I managed to turn some steel in order to make a proper spanner plus a centering tool for the lock washer as the manual suggests. I also got a friend to help me reinstall the vertical head on. I really need to do something on the lifting issue, that long-reach is very heavy and I am not tall enough...

Regarding the vertical head, I can't tell about the standard one but the long-reach is all straight shanks, no tapers. I have a feeling though, not quite sure, that the long-reach was introduced late in the 500 mm X life or together with the 400 mm X machine, so it's newer design compared to the horizontal spindle.

Manual does not mention specs for the horizontal, I'll use the values for the vertical spindle.

As I said earlier, I am prepared for the possible need to replace the collar with a fatter one. I realy hope that's all that's going to be needed, I hope/think that the bearings will be fine, radial-play-wise. Acquiring the spacer will not be trivial, I don't know someone that can make it for me since I don't have a grinder, I may even ask Mr. Singer for one. In any case, certainly easier than a new bearing....
Also agree on this, new collar is better than shims for me.

Till next update.

BR,
Thanos
 
Hi to all,

no actual progress to report here since yesterday that I had some workshop time there were other tasks running that had some dust flying around so I wasn't very keen on messing with the spindle bearings under those conditions.

I did manage to make a proper spanner for the castle nut though. Prehardened chromoly steel for the spanner itself, to give the teeth some fighting chance, good fit to the spindle in order not to rock. Handle off mild steel with a 1/2" square hole for a torque wrench, it might come handy if I want to replicate settings. Quick and dirty:

attachment.php


attachment.php


That nut had the hammer/punch treatment one time too many, at least in my hands.

I also made a sleeve to fit both the nut and the spanner that would center the lock ring.

Tomorrow with real update, hopefully.

BR,
Thanos
 
Hi to all,

reporting back my progress as promised, with kind of mixed feelings...

Summarizing the results before describing the procedure:

1. radial play at the back bearing (tapered inner race) did not seem to change no matter how tight I got the nut.
2. radial play at the back bearing did, strangely, get a bit smaller when tightening the threaded ring
3. though it didn't affect radial clearance, tapered inner race DID indeed move a bit towards the front and took up axial play!

And here are the steps I followed:

- I put the spindle in the quill, thrust bearing installed but no threaded ring. I held the back of the spindle in the vice and started tightening the nut after having oiled the tapered section of the inner race and the nut threads. At the same time I measured radial play at the rear bearing. I got 0.008 mm. I tried to reduce it by tightening the nut. There, I went a bit too far, I hope I din't damage anything, I tightened up to 100 Nm, which, now that I think about it is crazy. Radial play did not get any better. Here is the setup:

attachment.php


And here is the measurement:


- Next, I measured radial play at the front side of the spindle, getting great results, around 0.001 mm!

attachment.php



- Then, I measured axial play and got around 0.18 mm

attachment.php



That was the point I was fearing before starting, not to be able to adjust radial clearance employing the taper. Also, I had no progress regarding axial play, not nice.
For completeness and to address possible questions by the forum, I installed the threaded ring and tightened it by hand, as far as I could go. And what do you know....axial clearance was gone! It seems that inner race moved axially to take up play but didn't swallow enough to affect radial play. Strangely though, installing treaded ring did reduce radial play by 0.002, so I ended up with 0.006 mm!

- Next, I released the crazy torque off the nut and repeated torquing it back again checking axial and radial play as I went on. It seems that 50 Nm, much more reasonable value, are enough to bring axial play down to 0.01 mm. Radial play stayed there, at 0.006mm at the back and almost zero at the front bearing.

- Throughout the procedure spindle was free to rotate nicely.

So, the way I see it, though I would have liked rear bearing radial play to be a little better, I think I'll be ok with out any modifications. I will retorque to 50 Nm with the spindle off the quill, and confirm that axial play of the collar/spacer is, indeed, taken up. If this proves ok, I think I'll clean everything, Kluber them and put them back together.

It is a bit annoying that I don't understand the use of that taper in the spindle and the bearing, it could have easily been parallel there since it didn't do anything regarding axial play. But, I can live with it.

I will report back with measurement of the spindle alone. Seems I might, indeed, be able to put it back together till the end of the week.

Br,
Thanos

P.S. While I was at it, I checked SK40 taper runout. I was amazed to find it to be less than 0.002 mm! Much better than my vertical spindle....
 
It seems that 50 Nm, much more reasonable value, are enough to bring axial play down to 0.01 mm. Radial play stayed there, at 0.006mm at the back and almost zero at the front bearing. P.S. While I was at it, I checked SK40 taper runout. I was amazed to find it to be less than 0.002 mm!

Thanos, in your shoes I would call this good and not mess with it further. The axial play is (just) within the Deckel spec and I suspect that the radial play also is (you need to check it with a test bar at two locations).

Note: it's possible that when the spindle is heated up from use that the play changes (and it might decrease).
 
Thanos, in your shoes I would call this good and not mess with it further. The axial play is (just) within the Deckel spec and I suspect that the radial play also is (you need to check it with a test bar at two locations).

Note: it's possible that when the spindle is heated up from use that the play changes (and it might decrease).

Hi Bruce,

well, that's what I think I'll do. Hopefully today I'll make sure that, indeed, 50 Nm eliminate play in the collar/spacers stack (so what I saw yesterday with the spindled assembled is actually this and not some other interference somewhere).

Then, I'll take it to work, we have a dust free room in the lab, I'll clean everything, grease and put it back together. And then back to my workshop for adjustment.

I'll feel better when it's back on the machine, instead of the aluminum plug I have now to cover the void :)

BR,
Thanos
 
Some thoughts:
First off Torque should not be the determining factor in setting your axial play....
If things are correct the bearing should come down hard on the spacer ring, and the torque (if measured) should go up on a pretty much straight line once contact has been made.(No compression, little stretch)

Also some care should be taken here to verify that everything including the spacer and ring as well as the nut are absolutely flat. Any out out of flat on any of these parts can deform the spindle and draw it up into
a banana shape, which will affect overall run out...
Nut face should be checked relative to its thread. Turn a thread on the lathe, install nut (Snug fit) and test contact face for truth to that thread....


Once you have verified that the spacer is captured and crushed tightly, that is the value you need to use when tightening the retaining nut.
Axial clearance (and you must have some) totally is made by the outer threaded ring at the end of the quill....

Not sure about about the radial play...Might be fine. Should be noted here that later spindle designs do not even have that bearing.....
Higher clearance might be in the design, as its possible that the designers were trying to have the bearing at the end not fight with the action of the forward two spindle bearings...

Personally i would be tempted to grind some off the spacer and see if the radial play improved, that bearing may not original, would be interesting to do a bit of research to see if the class of that bearing matched the original factory specifications.
A call to Singer might answer some reliant questions here...anyhow that's a difficult call, that only you can make biased on feeling how everything is taking up and such when the retaining nut is tightened
Bruce's opinion is prudent, and it might be best to leave it be, but i would still wish to see what Singer had to say here....

Cheers Ross
 
Hi to all,

Ross thanks for the comments, they appear in line to my latest observations. Details in the description below.

Here is the setup:

attachment.php


I set the spindle up on its own and installed collar, spacer, bearing inner race, lock ring and nut. Plus the spanner and the torque wrench at 90deg.
I went ahead and tightened the nut to 50 Nm, as yesterday only to find that, indeed, collar was not tight, there was still some play.

So, point one you were right: I made the mistake last time to have the thrust bearings tight, no play there, so required end clearance I measured was all from the play in the collar! Totally wrong way to go.

Next step, I tightened the nut a bit more and saw that, indeed, collar was getting more difficult to rotate, but still had some axial play (around 0.01 mm). But then I had to leave the workshop and was not able to go on with this experiment, I will continue tonight hopefully.

Also, second point you rose, collar is not dead flat, I can't remember now but I think it was off by more than 0.01 mm. I'll have to check and correct.

Regarding radial play, you are right in that I did not give your approach a fighting chance: I have to install a thinner spacer to allow for the bearing to move a bit further down the taper and see if this makes any difference in radial play. I will make a test ring on the lathe (mag chuck) as flat as I can go, check for flatness on the surface plate and correct to reason and try this as a proof of concept. Depending on my results I'll source a proper ground collar on not. Regardless of what I'll do in the end, I owe it to the forum to see this through in case it helps someone in the future.
(now that I think about this, I might just try tightening the nut without the collar altogether and measure radial play. Now I have measurements (torque) so I can avoid going too far and breaking something!)

Regarding spindle bearing, this is the older design with the single front bearing, so both bearings should, I guess, have similar radial play specs.

Let me also note here that my initial observations were so off due to the fact, of course, that I didn't have a proper spanner to tighten the nut, so the collar was never going to be compressed and I was getting that ridiculous play (0.3 mm). Not having the proper tool not only risks damage to the parts (nut) but also leads to totally wrong conclusions.

I'll also ask Mr. Singer, it has been in my plans, I hope I can explain this clearly enough.

Many thanks for the contribution once again.

BR,
Thanos
 
Thanos:
I want to congratulate you on working through all this. Looks like you are following a sound and logical progression at solving this and getting your spindle back in service with sound
and correct bearing setup...keep moving forward, think you are heading in the proper direction.

I needed to correct some of my previous statements. I spoke about later spindles not having "that bearing" referring of course to the bearing at the end of the quill....
I was confused. That end bearing is one of the spindle bearings and in my mind it should be running with clearance similar to the forward bearing.
In the above description, believe you stated that the radial play at the rear was .008mm.....To me this is just fine...would say that is about what i would wish to see.

Also stated that you believe that the spacer is not flat...be careful here...any truing of the sleeve will of course reduce its overall height and change the point where the bearing fully takes up on
the ring.....and also as a side result that change will also affect the clearance of the bearing...
Before making any correction to the spacer carefully measure to get the ID, OD and of course the length...be sure to record the length as a function of its high and low dimension....

Make a note of the required torque on the retaining nut to make the sleeve and ring fully tight with the old spacer...
If any work is needed on the spacer my suggestion is that you make new....starting with a part that is just at the high length of the old spacer assuming that is of course that you can tighten the retaining nut enough to
actually compress the sleeve and hold the ring tightly....
If you can't get the sleeve (original) to tighten up with the radial clearance as stated above (.008mm) then you need to make a longer sleeve.

Small changes here make a big difference, work carefully, double check every step and be sure you get the expected result form any change.
The good news is that you have recognized and isolated a problem and its cause. There is a clear workable solution , it just needs to be implemented with care and precision....the end result will be a better
machine....Looking forward to Herr Singers response, and your final evaluation numbers post repair...

Cheers Ross

ps. The forward bearing on your spindle is also on a taper and like the rear bearing can be moved to tighten the radial (running) clearance....Personally, i sort of like this arrangement...not sure why Deckel abandoned this
design in favor of having sized needle rollers to adjust the running clearance....Guessing that the later setup is more rigid, having a larger spindle OD , and having the bearings closer to the taper on the front end...
 
Hi to all,

Ross thanks for the kind words, I highly value them, along with your advice of course! I hope that someone might find this documentation useful in the future, though a bit misleading in the first posts since I had started with wrong initial conclusions...

Please have a look at my latest observations and advise on my conclusions:

- Initially I continued from where I had left it yesterday, that is tighten the nut until the spacer stack is compressed and no allowed to move anymore (spindle out of the quill, as in yesterday's picture). What I found out was that at around 60-70 Nm, spacers where tightly pushed against the spindle's shoulder without any play, measuring at the T-sectioned spacer right after the collar. However, if I measured at the collar, be it edge or closer to the spindle, I was getting around 0.001-0.002 mm needle's movement.

Conclusion: Collar flexes and I measure this flexure at its edge. But closer to the center? I guess, what I am seing there is the result of the collar not being paraller, so there are tiny gaps that allow for some breathing? (in hindsight, I should have measured all around the periphery to verify this assumption)

- Next, I removed the T-sectioned spacer but left the collar and everything else and put the assembly back in the quill. The aim was to confirm that indeed, if given room to move down the taper, the inner race would shallow and radial play would be reduced. Indeed, as I torqued the nut down I reached 0.005 mm radial play at around 75 Nm! Did not push anymore, I had my proof!

Note1: I had made provision (with shims thinner than the spacer I removed) that when pulling the spindle back out of the quill I would not press against the rollers but the inner race, not to damage anything.

Note2: I found that keeping the mag base close to the point of measurement, error introduced from flexing is reduced. I had more consistent results having this setup here (compared to a couple of posts ago where the mag base was further up the spindle)

attachment.php


Again in hindsight I should have put an arbor in the spindle and secure it's business end with a screw jack and clamp. This way spindle would not be allowed to flex (not much at least) and movement would only be due to quill's radial play! I am not used to sub-0.01 mm measurements so all this does not come natural to me, it requires thought!

Conclusion 1: Apparently, radial play IS influenced by the axial position of the inner race on the spindle's taper! (as Ross suggested...)

Conclusion 2: We have proven that at 75 Nm spacer stack is already compressed, so bearing will never be able to travel that far and allow for this reduced radial play. This is why, even at the 100 Nm of my initial measurements, radial play was fixed at 0.008 mm.

So, my plan I think should be to check everything for flatness, rectify if needed, clean, lube and assemble. If I find something out of parallel (I now collar is) and modify it to be (more) parallel then I eliminate distortion sources plus I end up with a wee less radial play, since the modified part will be thinner, which is welcome. Of course, I'll measure carefully everything before I do anything to it. Lacking a surface grinder I think I will be able to remove the high spots with emery cloth and selective pressure on a flat surface. In any case, I'll work in small steps and measure, relatively safe approach I guess.

So, if no other suggestions change my plans, I'll come back tomorrow to report work on the collar (or/and the nut).

Thanks for the help.

BR,
Thanos

PS. Regarding the front bearing Ross, that you say is also on a taper, there are no spacers for it to rest against! So, we might conclude that spacers are not required to keep the bearing in place, taper is enough for this. In our case (rear part of the spindle) the extra parts are only to work with the thrust bearings in setting the axial clearance. So, I guess, the official way to adjust this thing would be to use a thin spacer/collar that would allow for the tapered inner race to travel as far down required to set the radial play and THEN measure axial play. Disassemble, replace thin spacer or collar with a fatter one as calculated by the axial play measurement and you're good to go!
 
NSK has a superb spindle bearing reference document available here. There is a whole chapter on mounting tapered-bore roller bearings. My Acieras use that type of arrangement for the bearings at the spindle working end. The Aciera manuals say to leave a tiny amount of radial clearance, whereas NSK suggests having a small amount of preload for bearings at the working end.
 








 
Back
Top