Myford S7 Cutting Barrel Shaped Cylinders

I've been overhauling my Myford Super 7 and, having machined a 1" dia x 5" long test bar with 1" held in a collet chuck, I found (after many, many tests and adjustments) that it is cutting a barrel shape instead of a parallel cylinder.

I have adjusted the feet to ensure the ends of the cut are the same diameter, but having run the cutter from right to left each time using the slowest power feed, I found that the middle of the cylinder is consistently around 0.0015" (1.5thou) oversize. I ensured that the cutter was removing metal from the length of the bar on each pass.

Is this a common thing due to wear?

Would the wear be in the bed or the apron? Or both?

Does anyone have any workarounds for when accuracy is essential?

I should add that I measured the bed wear according to a recent article in MEW and it is just on the limit of needing a regrind, with 5thou of vertical wear and 2.5thou of horizontal wear across the full width of the bed.

Many thanks,

Mike

Edited By Mike Donnerstag on 23/05/2020 17:16:55

The standard test bar for setting a lathe to turn parallel is normally relieved in the middle just leaving collars at either end, it may be worth trying that to see what you get , saying that it looks like you have a wear problem as you stated, can't think of a work around for a barrel shape. I would perhaps also go bigger diameter on the test bar but that's being fussy & wouldn't get rid of your barrel shape.

Tony

Are we talking about a shaft or a bore

I'm thinking combunation of spindle bearings not set up correctly plus bed not level.

So acombination of work deflection which get smaller towards the chuck and a taper because you have adusted the bed to compensate for the deflection.

What does everyone else think?

regards Martin

You can try to see what you get if you turn a bar between centers or buy one of those ground test bars and run a DTI a long its length. Niko.

All sorts of weird thing happen when you induce twist. What he has going on is very unusual if a tailstock centre is not being used.

I'm in favour of the test bars and have two, both with MT2 at one end, one is about 7 1/2" long and the preferred one which was not available until well after I bought the first is about 11 1/2" long. They have centres in the ends as well as the MT. If you chuck up a bit of steel or brass bar and turn a 60 degree end on it and put a bar between that and a DEAD centre in the tailstock, an indicator can be used to check the heights as well as the alignments, and if the MT end is in the tailstock alone, it can check that out. They make it easy to check chucks as well.

Tony Pratt 1: Whenever I've set up the bed in the past (only had the lathe since Jan 2019) I used the collar method, but then I found discrepancies when turning a shaft of a shorter or longer length, hence why I tested the bed by machining the whole length of the shaft today under power at the slowest gearbox speed.

Martin Kyte: I've been testing by machining a shaft, not a bore. I've adjusted the bearings in the past and haven't found a big problem there.

speelwerk: I bought a test bar recently, so I will see if that confirms what I found when machining the (unsupported) shaft

Pete Rimmer: What do you think may be twisted to make the barrel shape?

old mart: I intended to align the tailstock only once I was happy with the unsupported shaft.

This is a classic symptom of compensating for an out-of line headstock by inducing twist in the bed.

Can you get your hands on a machinist's level to get the bed un-twisted?

This is a classic symptom of compensating for an out-of line headstock by inducing twist in the bed.

Can you get your hands on a machinist's level to get the bed un-twisted?

I doubt if the head stock is 'out of line' but it could be? I will await developments maybe a twisted bed is causing barreling but wear in the bed is the likely cause, note OP's wear figures

Tony

Mike what sort of tool are you using?

I use only a well honed HSS tool for machining test pieces after checking or resetting alignments. The type of tool and depth of cut are going to greatly influence the results. I limit the depth of cut to 0.05 mm or 0.002", this standardises the machining conditions.

I do think you are asking a lot of the Myford as regards the unsupported length, which I take to be 100 mm or 4". As you say you have 1" in the collet. On my Emco Maximat Super 11 which has a taper roller bearing headstock I only ever use a 75 mm protrusion of the test piece beyond the jaws of the chuck.

Personally I would start with the test bar in a spotless mandrel socket and run a clock along where a tool would normally pass and over the top. The latter test will tell you if the spindle is dipping or pointing skywards. The first test tells you whether the spindle is pointing down the bed centreline.

Provided you have not disturbed the headstock casting during your overhaul?

Then if the clock in the tool position is reading higher or lower towards the tailstock end then the bed is out of alignment.

To set this I either use an Engineers Level or a piece of 25 mm diameter silver steel kept specifically for the purpose. Grip one end in the 4 jawed chuck using protection pieces and clock to run true. Check the outer end and any high spot should be put at the top.

Return the clock to the tool position on the Silver Steel and clock from headstock to the tailstock end and adjust the bed feet until you get the same reading.

Remove this set-up and return to your test piece. It should not be that far out and minor adjustments to the bed feet at the tailstock end should put it right. Two collars on the test piece save a lot of time machining and preserve a HSS cutting edge.

Spot on concentric centres in the piece of Silver steel make setting the tailstock a doddle.

Regards

Gray,

How is the headstock located on the bed? Is it on the bolts or keyed between the shears? You wouldn't need much of a gap, or much of a bump on one corner of the key to cause a few thou out over a few inches.

Pete

Myfords have thrust screws to abutt the headstock to the rear shear before the 4 hold down bolts are tightened.

Personally I would ditch the collet in the original set up and hold a piece of free machining steel in the four jaw for rigidity and turn a collar test with a very sharp tool taking very little off at a time to set the bed up.

regards Martin

Well, it could be the cause then. I would certainly start with levelling the bed then turning a larger OD part in the chuck and mic it for taper, use those jack screws to pull the rear end round to adjust. Another thing I have noticed on the Myford lathes is that they tend to pull a little attol of metal up off the bed around the threaded holes for the headstock bolts. You really notice it when scraping. I would give those a healthy countersink so make sure none of them hold the headstock up slightly.

If the basics are ok such as no twist in the bed, reasonable bearings and no loose items, then I'd say it could well be attributal to wear of the vertical shears. You would only need ~ 0.0008" of movement to give a 0.0015" change in diameter, which would relate somewhat with your measurement of 0.0025" horizontal wear across the bed.

Get an accurate measurement of the full width of the bed in the worn area and also the width of the front way. Compare these readings with similar unworn areas of the bed to determine how much deviation exists at the inner front shear, which is the one likely to be most worn.

There's a conversion to 'wide gauge' guide that can be done to these lathes if it is wear that's causing your problem.

Martin.

I'd want to have more than 1" in the collet. Two inches at least to be sure it's solid. Also try it with four jaw chuck just to make sure your collets are not allowing movement etc.

Your wear figures are pretty borderline with the .003" horizontal and .005" vertical maximum that Myford recommended in their old literature on their bed regrinding service.

Is yours a post-1972 factory wide guide? If not, the critical measurement is not the full bed width but the width of the front shear where the narrow guide runs against it.

I think, at the stage of testing umpteen times I would test with cutting between centres. At least that would avoid most of the possibilities?

If it is the same diameter at both ends and different in the middle, you have a wear problem with the bed/carriage,I would suggest. Both may well need attention.

With upto 5thou wear in the bed how useful is an accurate level going to be as you could be resting on a worn surface one end and almost untouched at the tailstock end?

In a similar way although a test bar will give you a reading with zero force against the carrage if there is wear in the bed there is also likely to be some in the cross and top slides and any cutting force will push the tool away unlike the weak spring of a dti that just follows so unlikely to get two matching results. Hold dti in the tool post and pull the toolpost back while dti is resting against side of test bar.

If you are getting a barrel then do the bobbin test but with a 3rd large diameter piece in the middle, light cut first as Gray suggests then a heavy one to see if things change due to forces pushing the tool away.

Edited By JasonB on 24/05/2020 07:48:12

Or just take a continuous cut over the 4" length of the 1" diameter bar. It does not take that long. And gives a full reading of what is going on.

Yes, using a level on a worn bed is fraught with danger as 5 thou wear on one shear can throw a sensitive level out by a lot. Over the 4" width of the Myford bed that could throw the level out by 15 thou per foot! By the time you twist the bed to compensate for that, it could be all over the place.

NDIY's between centres turning test might be worth a shot. If you adjust the tailstock to turn the same diameter on the test bar at both ends but the middle is still barrel shaped, it would tend to confirm suspicions that the bed wear is the culprit.

Very hard to check headstock spindle alignment with a bed that worn. A ground test bar etc may sit dead true to the spindle axis but then when you wind the carriage along the worn bed, it is going to follow the worn section, so the bed overall may be in line with the spindle, but the worn section is not.  Maybe an extra-long test bar so you can run the carriage on the far end of the bed where wear is least? Problems then with long bar flexing etc.

Best shot may be to try the wide guide conversion if yours is a pre-1972 narrow guide machine. Your carriage then at least will follow the unworn rear shear vertical surface instead of the worn front shear vertical surface. But if yours is a later factory wideguide machine, you can't do the reverse because the tailstock will have worn the inner vertical faces of both shears, including the narrow guide surface.

Edited By Hopper on 24/05/2020 09:39:57

.

I agree completely, Hopper ... but isn’t that ‘where we came in’ ?

MichaelG.

________

Mike wrote :

I have adjusted the feet to ensure the ends of the cut are the same diameter, but having run the cutter from right to left each time using the slowest power feed, I found that the middle of the cylinder is consistently around 0.0015" (1.5thou) oversize. I ensured that the cutter was removing metal from the length of the bar on each pass.