PDA

View Full Version : Is my lathe broken?



loply
10-09-2012, 04:57 PM
Hi folks,

I picked up this Denford Viceroy lathe a few months ago and have just finished replacing the transmission with a belt drive and installing a VFD. I was planning on converting it to CNC (an ambitious undertaking, I know!).

I don't know exactly what I'm looking at, but, I'm getting some odd behavior from the spindle. I've uploaded a video to YouTube showing my various tests.

Firstly, the lathe has not yet been levelled, and is cutting a 0.06mm taper over 50mm, but, I don't know if either of these facts impacts on the problem shown in the video.

The 'register' is indicating less than 0.01mm of runout, but the internal taper (and any center held in it) are running out massively, and the register on the chuck backplate that came with the lathe has massive run out, yet the face of it is running true.

I've no idea what's going on here but I'm hoping I don't have a bent spindle.

If I chuck up a straight part - like a chrome rod or a piece of drill rod - I get loads of runout, but, that might just be because the chuck is knackered, and I only have one that fits this lathe.

I would really appreciate it if anybody who knows a thing or two could look at the video and let me know what to do/check next, as I'm beginning to worry I have a dead horse on my hands.

If the internal taper on the spindle is simply rough/damaged causing the dead center to run off, then why is the register on the chuck backplate (shown at the end of the video) showing massive runout too?

I'm tempted to order a new chuck backplate and a 4-jaw chuck and just see how close I can dial a part in, but, I don't want to spend any money until I know what I'm looking at.

Spindle bearings are adjusted with a little preload here, and the spindle taper has been cleaned religiously.


http://www.youtube.com/watch?v=ETSzzPiXe3o

Cheers,
Rich

RLWP
10-09-2012, 05:04 PM
How about reboreing the taper in the spindle?

Richard

loply
10-09-2012, 05:10 PM
I know the video is a bit long but if you watch it to the end you'll see that the chuck backplate is also running out, so I think the issue goes beyond the spindle taper.

Euph0ny
10-09-2012, 05:11 PM
The outside of your spindle seems to be running true, as is the edge of the register, since your faceplate isn't running out, but the inside of the spindle seems out of kilter when measured directly or with a centre inserted.

Could it be that the internal taper of the spindle has been damaged or incorrectly reamed out? I would get a known good male taper, blue it up and offer it to the spindle taper to see what kind of contact you're getting - that should give you an idea if the spindle taper is out of whack. It might be something as simple as a burr that needs flattening.

MetalMunger
10-09-2012, 05:38 PM
Get yourself started on the right path by begging, borrowing or stealing a good Dial Test Indicator with an extra long probe.
Clean the inside taper of the spindle then run your finger around inside and feel for burrs or rough spots. Indicate the inside taper and mark and record the high spots, then indicate the exposed bearing race that is turning in the video mark and record the high spots again.
Report back.
Has the spindle been removed at some point and reinstalled if so the bearing can be "cocked" or have the bearings been replaced?

winchman
10-09-2012, 06:20 PM
I don't like seeing the runout get worse as you get further from the bearing. That makes me think the spindle is bent, but hopefully there's a happier explanation.

Gunsmithing
10-09-2012, 06:27 PM
Look at your bearing. It most likly in bearing not the spinlde. Look for any chips some lathes will get chip in the roller/ball

Dave

flylo
10-09-2012, 06:46 PM
Are you supposed to beable to see inside the bearing like it shows near the end of the video? Seems like it should be sealed or covered. I now see where a cover must bolt on.

loply
10-09-2012, 06:48 PM
Hi folks,

Just been out and had another look at things.

First off - one problem solved - it turns out the bore on the back of the chuck backplate is 3 thou oversized, which explains the large amount of chuck runout. I checked the spindle register and it is precisely 1.5000 inches but the backplate bore is 1.5003 so that can hit the bin. I could rebore and sleeve it I suppose...

Now that I seem to have solved that I can presumably get a chuck mounted properly, so I'm hopeful I simply need to run a reamer through the spindle taper. I need to get a new backplate to find out for sure though, I guess.

With regards to the bearings, the spindle has been out, but the front bearing remained in place on the spindle, and the rear one is pushed back in place by tightening the locking nuts that would preload the bearings, then slackening them off again. I'm confident that bit isn't the problem because neither bearing races were removed from the headstock, and you can exert great force to reset the rear bearing on the spindle by tightening the locking nuts.

I notice that the DTI on the taper doesn't swing constantly like it would with a bend, it stays static for 90 degrees at a time then shifts suddenly, as it the taper has been squared out a bit... Perhaps by having things bashed in it to remove dead centres or suchlike.

flylo
10-09-2012, 06:56 PM
The name on the video is quite catchy, sounds like an Army training film!

Arcane
10-09-2012, 07:42 PM
Get yourself started on the right path by begging, borrowing or stealing a good Dial Test Indicator with an extra long probe.
Clean the inside taper of the spindle then run your finger around inside and feel for burrs or rough spots. Just a reminder directed at any new people on the board...never do this under power, always by turning the spindle by hand. Even if your finger is highly unlikely to get caught in the taper that is probably a #3 MT, it's a bad habit to get into, especially if you happen to do it with a #2 MT. :eek: Indicate the inside taper and mark and record the high spots, then indicate the exposed bearing race that is turning in the video mark and record the high spots again.
Report back.
Has the spindle been removed at some point and reinstalled if so the bearing can be "cocked" or have the bearings been replaced?
Hi folks,

Just been out and had another look at things.

First off - one problem solved - it turns out the bore on the back of the chuck backplate is 3 thou oversized, which explains the large amount of chuck runout. I checked the spindle register and it is precisely 1.5000 inches but the backplate bore is 1.5003 so that can hit the bin. I could rebore and sleeve it I suppose...

I don't think that affects it at all. The registers on my chuck backplates are all considerably oversized and they all thread on and register properly every time, unless there is a small bit of swarf in the threads. It appears that the 60 degree angle on the threads is all that is necessary to center the chuck properly. This has been discussed in the past also, btw.

Also in your video, I didn't notice you indicating the vertical flat where the bolt holes are located that is the surface the chuck body seats against. The large flat on the face of the backplate that you did show doesn't touch the chuck body, there's quite a gap there at least on mine there are.

Now that I seem to have solved that I can presumably get a chuck mounted properly, so I'm hopeful I simply need to run a reamer through the spindle taper. I need to get a new backplate to find out for sure though, I guess.

With regards to the bearings, the spindle has been out, but the front bearing remained in place on the spindle, and the rear one is pushed back in place by tightening the locking nuts that would preload the bearings, then slackening them off again. I'm confident that bit isn't the problem because neither bearing races were removed from the headstock, and you can exert great force to reset the rear bearing on the spindle by tightening the locking nuts.

It certainly appears that you can see the balls in the front bearing. That means there's a huge hole where swarf can enter the bearing. That's a most serious concern!

I notice that the DTI on the taper doesn't swing constantly like it would with a bend, it stays static for 90 degrees at a time then shifts suddenly, as it the taper has been squared out a bit... Perhaps by having things bashed in it to remove dead centres or suchlike. Hope this helps.

MetalMunger
10-09-2012, 08:01 PM
I notice that the DTI on the taper doesn't swing constantly like it would with a bend, it stays static for 90 degrees at a time then shifts suddenly, as it the taper has been squared out a bit... Perhaps by having things bashed in it to remove dead centres or suchlike.

This is the type of indicator you need to properly measure your spindle:

http://nano-machinery.com/catalog/images/Dial%20Test%20Indicator_Mitutoyo_513-515T.jpg

J Tiers
10-09-2012, 11:18 PM
Yes.....

The "register" is not intended to be a wringing fit, which is the only way it could possibly do any meaningful centering.... I have several chucks and not a one has a close fit there, some have 25 thou clearance, but they all repeat nicely, and chuck up work without more than normal runout.

In any case,,,,,

What I think I am seeing is....

1) essentially no radial runout on the unthreaded exterior of the spindle

2) essentially no runout on the FRONT FACE of the chuck backplate (axial movement) Incidentally that is NOT the chuck contact surface...... or should not be....

3) a significant runout of the accessible part of the spindle taper, and adapters etc put into it.

4) significant radial runout of the chuck backplate

5) a missing cover on the front of teh headstock.

6) all runout is fairly smooth and "sinusoidal", not really jerky or limited to one spot, although there is a suggestion that the indicaator is losing contact with the surface part of the time in some tests

OK.... #6 almost eliminates an issue of crud in the bearing... that would be jumpy and associated with one place, not fairly smooth.

I am going to say that #2 eliminates most bent spindle thoughts.... a bend should have made the backplate "wabble".

#4 is interesting, but COULD be a function of the backplate itself..... it isn't a for-sure property of the spindle. it does not seem to be the same AMOUNT as inside the spindle....so isn't necessarily related

#1 means the spindle itself, AND THE BEARING must be reasonably decent on the OD and as far as rotating smoothly around a fixed axis without wobbles.

#3 suggests that the rest of the spindle taper needs to be investigated with a lever arm type indicator similar to that illustrated.

I don't quite see how the taper can be bad and the exterior be good, unless the thing was made badly originally, or unless the taper was bored out to a different one, done badly. The suggestion of a 'flat spot" is interesting, although I think the effect was loss of contact and not a "geometric feature".... Not sure how the flat spot " could occur without more damage.

darryl
10-10-2012, 12:03 AM
I think what I'm seeing is the spindle being fairly true, but the threaded end being off-center and probably a bit out of round. Maybe it had a crash that didn't bend the spindle proper, just the threaded end. That would account for the runout in the taper and the anomaly in the eccentricity. Because the threads are responsible for centering the chuck, it leaves the chuck eccentric, but the register keeps most of the wobble out. That would explain why a faceplate could run fairly true.

You would be able to make a setup whereby you can true up the taper, using the crosslide, but you won't be able to improve the threads without re-cutting them. If you did that, you would be reducing the diameter of the threaded portion by roughly 10 thou by the time you had all of the thread flank area 'accuratized'. I'm guessing at this figure by the amount of runout you are showing- could be more, or could be less. Either way you might still have enough 'overlap' between the spindle threads and all the threaded adapters and back plates you have so that nothing is compromised by the re-cutting.

Of course you also have to consider how hard it would be to re-cut the threads. If they are hardened, your best option might be to grind them, but it's unlikely you'd have the setup to do that. If you started with a regular single-point threading operation and it failed because the cutter couldn't stand up to the job, you might be left with more damage than is already there. But I think with the right cutter and a rigid setup, you probably would be able to clean up those 'dented' threads.

So far I've been assuming that I'm right about only the last inch or so of the spindle being significantly out of whack. The test for this is to indicate further into the taper, and right into the small end if possible.

winchman
10-10-2012, 01:49 AM
I'd like to see you check the axial runout of both the shoulders near the DTI at the start of the video. The chuck seats against one of those shoulders as you tighten it into place, and any runout there will affect the chuck.

I think the threaded part of the spindle is bent, and the flat inside the taper is on the side which is closer to the spindle axis.

lakeside53
10-10-2012, 02:27 AM
Hi folks,



First off - one problem solved - it turns out the bore on the back of the chuck backplate is 3 thou oversized, which explains the large amount of chuck runout. I checked the spindle register and it is precisely 1.5000 inches but the backplate bore is 1.5003 so that can hit the bin.



That's normal.... The registration is to the THREADS, not the shoulder. The precision square face one the spindle locks the chuck to the threads. Don't mess with your spindle until your are 1000% sure it's not something else.

Jaakko Fagerlund
10-10-2012, 04:01 AM
And if it comes to turning case hardened (the spindle), invest in a ceramic insert. Goes like hot knife to butter when run at high speed and no coolant.

RLWP
10-10-2012, 04:04 AM
1.5003" is three tenths of a thou - I'd say that was pretty good, or is it a typo?

I wouldn't ream the headstock, I'd bore it out. You should get a better result with a single point tool

And before all of that, I think I'd have the spindle AND the bearings out, just to make sure I hadn't done anything silly when I had it apart

Richard

loply
10-10-2012, 05:01 AM
Hi folks,

Thanks a lot for the replies.

I didn't realise the chuck was centered by the threads, I presumed it was by the register behind them, and that the threads merely pulled it back onto this.

Incidentally the CI backplate that came with my chuck has a stripped thread for about 1/4 turn at the rearmost part, though because it buts against a shoulder I can't see how that's from overtorquing, it looks more to me like it was banged with something whilst off the lathe.

I was worried by the replies that the spindle was bent or crushed on this threaded section, but luckily a local bloke once donated a wide foot Mitutoyo dial indicator to me, so I popped it on the threaded section and took the video below.

I carefully adjusted the screw on the indicator stand until it was reading smoothly so I know the indicator foot was riding on three threads at any point in time.

Initially the needle was jumping all over due to various dings and burrs on the threads. I very carefully filed these down and now I'm left with the video below - Seems about 0.02mm (slightly under a thou) of runout on the threads.

The needle movement does seem slightly 'jerky' to me, as if the threads are not quite circular, rather than being circular but spinning eccentrically.

Bearing in mind that my cheap import indicator appeared to detect some runout on the spindle register - possibly about 0.005mm (2 tenths) - I wonder if I have a slight bend? Or are the threads just a bit unevenly worn? Is 0.02mm runout on the threads acceptable?


http://www.youtube.com/watch?v=JrWTrFaKTkg

Cheers,
Rich

PS - There is a nice alloy bearing cover which wasn't in place on the previous vids, just because I had been oiling the front bearing.

J Tiers
10-10-2012, 08:10 AM
There is very little that suggests a bend or out of round on the OD.....

have you re-checked the backplate for radial runout? The burrs etc which you removed could have contributed to the originally observed runout, which itself did not appear correlated to the ID issues.

If the backplate was tightened against the shoulder of the spindle, the face runout would have been (and was) minimal, but radial runout could have been caused by burred-up and dinged threads.

Your runout problem does seem to get down to an issue with the taper socket. A survey of it with a long-nose DTI should give more info on that.

Jaakko Fagerlund
10-10-2012, 09:00 AM
Just a WAG, but from the looks of the indicator it seems like someone had banged the spindle so it has malformed from one point quite good. Dropped?

rkepler
10-10-2012, 09:46 AM
It sure looks to me that the outside of the spindle is in decent enought shape, but that there's some issue with the inside taper. If you have a good quality taper (test would be best but a newish dead center or even a good quality taper end on a drill will work) you might consider blueing the male and inserting into the spindle and seeing where the contact is by transfer. A second best strategy would be to mark the end of the male taper with a wide point marking pen (something for a dry erase white board is OK too) and inserting into the spindle and marking both for location, then lightly rotating the male taper a few degrees in the spindle before removing - that will leave the high points in the spindle marked into the male taper just removed (there is very little transfer with markers).

In either case the location and degree of transfer should help give an idea of the shaper of the spindle taper.

It sounds to me as though someone had a crash involving the spindle taper forward of the shoulder behind the register - this would allow chucks to center fairly well by butting against the shoulder and being pulled to center on the threads (so only a small error being added by the threads being off) while anything in the taper would have axial runout.

Be very careful if you try to correct the spindle taper. The only choices here are grinding or boring, don't think that it could be corrected with a reamer. With the small cut I would even think that boring is out unless the spindle is relatively soft, (in my experience) a cermet needs some minimum depth to maintain it's cut. Also keep in mine that a small change in diameter on a morse taper socket is a pretty big setback with the male tool in the taper - taking out .005 will set an inserted taper back .050 to .100 at a guess (too lazy to run the numbers on one cup of coffee).

Blackadder
10-10-2012, 11:19 AM
Rich

as to the chuck back plate run out just machine it true in situ do the front recut the chuck register ( note that there jaws the register is the outer flange ) give the location register a little play , dont worry about the back face it only sees fresh air and your eyes

remount the chuck put a piece of Silver steel in clock it up if it has runout slack off the mounting screws a touch and bump the body of the chuck into position retighten up the screws ( now you have a set tru chuck )

as to the spindle a reamer would only follow the error and they are sold as finishing reamers anyway .

If it was me I would forget about the spindle taper , as others have said a thou 0.001 will move that taper a long way into the spindle much more and you will run out of room on the tooling

just my 2 cent or should it be groats

Stuart

center of the UK

loply
10-10-2012, 11:59 AM
Hi folks,

Thanks again for the replies. I carefully checked everything out again just now.

I definitely have 0.02mm (potentially a bit less) runout on the exterior of the threads, but I'm thinking this may just be down to various attempts over the years at removing dings or suchlike.

I remounted the backplate after a very thorough cleaning and having rectified all possible dings on the spindle threads and, though the face doesn't wobble, I still have 0.09mm of radial runout on the backplate's chuck register.

At this point my guess is that the backplate was machined whilst not mounted properly, and the spindle is more or less fine, as I can't possibly think how else it ended up so wrong.

I plan to remachine the backplate and fit a smaller chuck, which I have spare, thus rendering the original one unmountable - no big loss as it's clearly past it's best anyway.

Once I've done this I'll chuck up some drill rod and do some further testing, but I think I'm in the clear.

As to the internal taper on the spindle, I think I will just leave it as-is. If a situation arises where I need to put something in there and can't turn a center in a chuck, then I'll consider taking a skim with a boring bar. My guess is that it's just out of shape due to mishandling/wear/people putting things in it to bang out a stuck center, etc.

Thanks for all the help folks, will let you know once the outcome is known.

Cheers,
Rich

Peter.
10-10-2012, 12:45 PM
Why don't you just face the backplate, turn the OD register down a bit then fit up the chuck and bump it to remove the runout then clamp it up tight?

BTW, I have a MT3 reamer you can borrow if you'd like to pay the postage.

rode2rouen
10-10-2012, 01:31 PM
Buy a tripod for your camera!!


Rex

loply
10-16-2012, 05:21 PM
Hi everybody,

Just a quick update on this. All is resolved (more or less!).

I can't imagine why but it turns out the chuck backplate that came with the lathe was just machined reallllly badly. The register on it was 0.1mm off center.

I bought two new chucks with new backplates and, after carefully removing various dings on the spindle nose, they both run totally true.

So true in fact that tonight I leveled the lathe (just roughly!) and turned a 6" by 2" test bar and measured it 4 times at 4 positions along it's length...

Absolutely ZERO detectable taper. Absolutely bang on 47.19mm both half an inch from the chuck and 5.5 inches from the chuck! So I'm very, very happy.

The only bad news is that on close inspection both the tailstock and headstock Morse tapers are knackered. The headstock is real bad and will need reboring if I ever want to use it, and the tailstock looks just a bit rough but is causing centres to stick up high. Luckily a helpful forum member here has promised to loan me his MT3 reamer so I can at least sort the tailstock out.

Thanks for all the advice.

Rich

J Tiers
10-16-2012, 10:12 PM
For the headstock taper, IGNORE IT.

Turn a piece of stock of around 20mm or so with a slightly smaller section long enough to grip with a chuck. Turn a center on teh free end of it in-place. No need to machine more of it than you have to.

Now use that center without removing it, setting the driving dog against a chuck jaw. Next time you use the center, re-skim it first , as it won't have gone back in chuck the same way.

The shoulder between the smaller section and the OD is to fit against the front of the chuck jaws and take the thrust.

dazz
11-07-2012, 06:35 PM
Hi
I have one of these lathes. If you go to the Denford website, you can see posts on its rebuild and some of the work I have done.
When set up correctly, these are very accurate lathes. The hard chrome bed effectively eliminates wear.

Getting the lathe level isn't as important as getting any twist out of the bed. The bed sits on 4 height adjustable bolts. It is possible to have a twisted bed that is compensated by an offset tailstock. Your results indicate the bed is straight.

If you are not happey with your spindle, you should consider making one. Drawings are available on the Denford site. Denford do not have spare parts. The same applies to the tailstock spindle. They are not complex items. They would just need care to make.

Your headstock bearings should be greased, not oiled.

These are good lathes so I hope you enjoy yours as much as I have mine.

Dazz