Viceroy T.D.S 1/1 G.B - Spindle refitting

[dazz]

Hello

After spending several days soaking and flushing out the old grease from the spindle bearings, they were finally ready to go back in. Installation is the reverse of removal.

If the headstock has been fully disassembled as I have done, it is easiest to start with installing the spindle and working down.

The photo shows the spindle being inserted. The shaft sticking out is for the back-gear selector. The selector is fitted immediately after installing the spindle.

The photo in the headstock shows the spindle partly installed. Note the spacer on the right is not fully in position. Everything is a lot cleaner now.

[dazz]

Hi

It was now time to fit the bearing and adjust backlash on the spindle.

To install the bearing race on the gear end of the spindle, I simply used the lock nuts (left hand threads) to drive the bearing onto the shaft.

I turned the spindle lock around the wrong way to allow it to effectively stop the spindle rotating while tightening the nuts. I didn't have a C spanner, so I made one from scrap.

I asked around on the internet for a good procedure to adjust the bearings. The best advice I could find recommended no pre-load. I also found advice recommending that the setting could be checked by measuring the bearing temperatures after a high-speed no-load test.

To achieve these goals, I came up with my own adjustment method. I set up a dial gauge to measure axial backlash. If the taper on the bearings is (say) 1 to 5, then 1 thou radial backlash would allow 5 thou axial backlash. For these example figures, measuring axial backlash is 5x more accurate than measuring radial backlash. I have not seen anyone else use this method to adjust spindle backlash but I can't imagine that someone else hasn't already thought of it.

I found that when the bearings are too loose, alternatively pushing and pulling on the spindle clearly shows backlash. The spindle settles at different locations after being released from a push and then pull. I tightened the locknuts until the spindle settled back to the same position after either pushing or pulling. This gave zero backlash and no, or minimal, pre-load.

I did a high speed run for around 20 minutes which gave a modest and safe approx 10 degree celcius temperature rise. If the temperature is excessive, the pressure will need to be reduced on the bearings. Too much grease will also cause bearings to overheat.

Once the spindle was back in, I then started to reassemble the gear train. It's all looking shiny and new.

[dazz]

Hi

To check that the spindle bearings were correctly setup, I ran the lathe at full speed for 30minutes, no load. I measured the bearing temperature every 5 minutes.

To measure the bearing temperature, I used a digital thermometer which measures ambient temp at the display. A second sensor on the end of the wire was clamped right next to the bearing outer race.

[dazz]

Hi

After resetting the bearing adjustment to give zero axial backlash, and therefore zero preload, I got the lathe doing some work cutting mild steel bar. The results were disappointing. Even light cuts resulted in significant chatter that would make any type of precision work impossible. Free cutting steel gave the same poor results.

It's possible that this could be a result of using lighter grade, modern grease compared to the Firmax 3 specified. The grease visibly flows as the bearing warms up and the reduced viscosity may not supress chatter like the old crud that was in there.

It's just as likely that these bearings require some preload to stop things shaking around.

I decided to leave the grease alone and tighten up the bearings until I could make a 1mm cut without chatter. It's been 25 years since I have used a lathe of this size so I don't know if 1mm depth of cut is a reasonable target, but seemed like a modest starting point. It's much easier to tighten the bearings than to loosen them off.

Tonight, I did a high-speed, no load test to make sure the bearing wasn't too tight. I ran the lathe until the temperature stabilized. The results are below.

Temperatures are in degrees celcius, time intervals were 5 minutes.
Delta is the difference between 2 consecutive temperatures
Ambient air temperature = 23.4 C

Time___Temp___Delta
0~~~~~23.0
~~~~~~~~~~~6.7
5~~~~~29.7
~~~~~~~~~~~1.2
10~~~~30.9
~~~~~~~~~~~2.0
15~~~~32.9
~~~~~~~~~~~0.9
20~~~~33.8
~~~~~~~~~~~0.7
25~~~~34.5
~~~~~~~~~~~0.0
30~~~~34.5

It's very important to monitor the deltas because increasing values indicate the temperatures are accelerating toward bearing destruction. Decreasing values as seen above indicate a path to stable running conditions. The overall temperature increase of 11 degrees C looks "safe" to me. I'll try the current setup for a while and see how it works out.

I suspect that most of the heat is produced by moving the grease around because this temp rise was only slightly more than the revious no-preload test run. If so, I should be able to add some more pre-load without wrecking the bearings if necessary.

Unfortunately, this whole process was trial and error. I don't know of a quick way of getting to this point. There is no specific torque setting or otherwise to aim for that I know of. I would be interesting to know how these were set in the factory. They must have had a quicker way.

One possibility might be to measure the axial force required to move the spindle a set distance (say a few thou). More preload should equal more force to achieve a specific movement.

[Denford Admin]

I would be interesting to know how these were set in the factory

Unfortunately, as with most engineers in the UK, those who built these machines have moved on or retired.

But I'll ask around anyway.

[dazz]

Hello

According to Timken, it isn't practical to measure preload. See here:

https://www.timken.com/products/bearings ... ol6No3.pdf

for details, noting that this is for wheel bearings, but the principle should be the same for any bearing.

[dazz]

Hi

I checked out price and availability of the spindle bearings today. They are readily available locally. The supplier had 500 of one type in stock. Both are used in aircraft applications (Cessna wheel bearings). Both are relatively cheap.

Given that the only two variables I can play with are lubrication and pre-load, I have decided to incrementally increase pre-load even if this means shortening the life of the bearings. They might only last another 10 years instead of 20. While thicker grease might reduce chatter, I don't think it would be the primary cause of my current problems.

Worst case scenario is that I would have to replace one or both bearings.

[Triac whizz]

Just had a thought... What are you putting this bar in? a Chuck? if so it could well be that the jaws are sh*gged, try using something firmer, like a collet chuck.

Could be the chuck mounting?

Did you actually take the bearings out/off? maybe still not seated right?

Just my 2 centimes worth

[dazz]

Hi

That's a good point. I am using the 3 jaw chuck that looks OK. I don't have a collet chuck but I'll try the 4-jaw chuck. It's heavier and has probably had little use.

No, I didn't take the bearings out, but I didn't do anything that would have unseated them either. When turned by hand, and with the backgear lever positioned at a point that the drive belt is disengaged, the spindle turns by hand smoothly and easily.

Taking the bearings out would be the last resort, because I then might as well replace them with new ones.

I'll let you know the results.

[Andy B]

Hi Dazz,

The manual for the later 280VS model lathes quotes a figure for bearing running temperature - I think it is 65 degrees from memory, but I'll check and get back to you.

Another point to remember is that as the spindle warms up it will expand, so decreasing the pre-load. I found some advice for the Atlas lathe here - https://www.atlas-press.com/tb_12bg.htm.

From previous experience in (automotive) gearbox assembly where a thread is used, it is normally calculated by the designer and the instruction to the fitter is usually given as, for example, 1/16 of a turn on the nut.

Andy

[Andy B]

Just checked the manual for my 280VS - the bearing arrangement is slightly different as the 280 has a pair of taper roller bearings at the nose end which are pre-loaded using a locking ring inside the headstock.

But it has this to say about temperature:
...Check that the preload is not excessive. The machine should run at top speed and the bearings should not exceed 65 deg C (150 deg F).
NOTE A certain amount of temperature rise must be expected when running the lathe at high speed but it should be possible to place the hand on the spindle nose after an hours running without discomfort.

Andy

[dazz]

Hi

Wow, 65 degrees celcius is quite hot. The roller/race interface would be a lot higher than that. I wouldn't be comfortable with using that much preload on the TDS.

I did a google search on the 280VS and couldn't find an image so I don't know how it compares to the TDS I have. The TDS has a pair of lock nuts at the gear end of the spindle.

[Andy B]

Dazz,

Have a look here - https://www.lathes.co.uk/viceroy/page5.html

The second picture shows the inside of the headstock (nose bearings on the right).
There is a lock nut next to the toothed speed readout disc (both are threaded), and these are used to set the pre-load.
The manual vari-speed is shown, but the arrangement is also the same on the electrical vari-speed version (which I have), and I suspect on the 'Synchro' versions which again have a variable pulley arrangement but a detent on the selector lever to give 6 fixed speeds (x2 with the backgear).

Andy

[dazz]

Hi

I had had a look at the link. Some elements, like the gearbox, look the same as a TDS. The headstock is completely different but the welding in the base is just as bad.

I did a bit of experimentation today. I found that if I increase the cut and feed to load up the lathe, the chattering stops. The 4-jaw chuck gave a slight improvement but solved nothing. I am wondering if the single phase motor is the source of the vibrations that are damped when there is enough load to keep the drive train fully engaged. Even with back-gear out, there is a fair bit of backlash.

[Triac whizz]

just another thought..

Is the tool at centre height? My guess it's a bit high as it works better with a bigger cut. I guess you have some top rake on that tool?

just another 2 centimes worth

[dazz]

Hi

A good thought. I aim to get the tool dead centre, except for the parting off tool, which is about 10thou below centre. I get the same behaviour regardless of which tool I use.

[Triac whizz]

hmm, curiouser and curiouser....

any chance the chuck is not seating properly?
what's the steel you're using some old sh*t out the scrap box? try another piece, try some ali, try some putty

[dazz]

Hi

Yes, I am using New Zealand mild steel. This is about 80% scrap so composition is variable and uncertain. With the piece I have, I tried to face the end. It just kept ripping the cutting edge off my tool on contact. The end had been hardened by the suppliers cut-off saw. I had to hacksaw the end off the bar to get back to steel that wasn't hardened.

I am aiming to get some decent free cutting stuff next week. Rubbish steel shouldn't cause chatter, just rubbish finish.

The chucks seem to fit true and tight. Nothing I put a DTI to shows any sign of wobble or backlash. There is backlash in the drive train, but no sign of wear. The backlash would have been there when the lathe was made. If backlash is the problem, then it has probably been agravated by fresh slippery lubrication.

[dazz]

I have just made a series of measurements to try and figure out if anything is bent. I measured wobble (W) and ecentricity (E) with a DTI. The results are as follows:

Measurements of W on the spindle register face were zero. Measurements of E at the start of the morse taper were zero.

With a 3 jaw chuck fitted, the E at the backplate was 0.01mm, W on the face was 0.01mm. No problems there.

With a 200mm face plate fitted, the W at the outer edge of the face was 0.06mm. E around the rim was 0.06mm. The Hi/Lo points of E and W weren't at the same places. The Hi/Lo points coincided with the line of slots so it's possible the faceplate is bowed slightly.

With a 4-jaw chuck, E was 0.09mm measured on the backplate, not the chuck. W was 0.09mm. The Hi/Lo points were different to the faceplate. They would line up if the spindle was bent.

Backlash in the drive train was measured at 2 degrees with backgear out, and 1 degree with backgear in.

The results indicate small inaccuracies but there is nothing to indicate a problem with the spindle. None of the errors "line-up" which they would do if the spindle was bent.

Based on inspection when the lathe was stripped, there is no evidence of any fault with the bearings. There is no backlash in the bearings. I think I have sufficient, but not excessive pre-load.

Locking the apron to the bed, and over-tightening the gibs (temporarily) has no effect on chatter. There doesn't seem to be a problem with tool stability.

Chatter is worse with back-gear out than in, but this could be due to the change in speed away from the chatter frequency. It could be that the backlash combined with a pulsing single phase motor is the source of chatter. Given that this lathe was only ever available with single phase motors, it shouldn't be a problem. I would need to fit a VSD and 3 phase motor to find out if the single-phase motor is the cause.

I still need to do some more experimentation with tooling to make sure my grinding is not the cause. I am getting good chips and no chatter with heavy cuts. I also need to try cutting quality steel instead of what I have.

So far, no obvious cause of chatter. It would be easier if something was mal-adjusted, worn or bent, but there isn't that I can find.

[Triac whizz]

hmm, post a piccy of the setup with a bit of your lovely turning

might give us a clue, if not there's always more people that look at posting with piccys