A good tool: test bar

[Sanjay F]

I got this test bar sometime ago with a load of tooling I acquired; it look brand new, never used. I noticed when machining the crankshaft the other day the change in diameter across the length so dediced to check the lathe using this gadget.

It was 4 thou across the length of the test bar, which seems enormous to me? Anyway it has done its job and the reading is not fluctuating 

[petertha]

I'm a little bit confused by what you are testing there. If the MT end is fully engaged within the tail stock socket, and the TS happens to be displaced off of spindle center, then engaging the center drilled end at headstock can only attempt to bend the bar into compliance. OTOH a different style of test bar like an Edge Tool (easily replicated in the shop) engages both HS & TS on centers. Now you indicate on one collar, traverse the carriage to the other collar without disturbing DTI & compare the readings. Any difference shows the TS as being in/out and/or or up/down relative to spindle center. In/out is the typical misalignment that can creep in over time of if TS not set properly. A bit of 'tailstock high' is normal, like 0.001", but it does not have the same effect of cutting inadvertent taper on a bar.

My understanding - the tool you show is used to engage fully with a MT socket, be it TS or HS spindle. It identifies angular misalignment of what its engaged with (only) typically by traversing across bar with DTI. I have a few of these & offer a bit of warning. Especially if they originate out of India which for some reason they do. They are cylindrically ground & can be very precise in diameter. Mine are within tenths across any section anywhere up & down the bar. But sometimes, either out of the box or over time, they can 'banana'. Meaning still a circular section but bowed 0.004" on one end. I suspect their choice of steel and/or heat treating has something to do with this, but hey, what do we expect for 20$    If you have ever come across one made by a big name, they fetch 10X that used & rusty.

[Sanjay F]

That's interesting - I presumed it's checking the lateral displacement and if an MT tool in the TS is parallel with the HS which is wasn't. However now I looked again even there is a centre hole at each end of the bar so could be used as you have described using your Edge Technology device?

I'll give it another try tomorrow  :

[petertha]

I've wondered the same myself. The center drills on both ends are most likely required for rough turning the bar. If the bar was then heat treated & centerless ground as I've seen advertised, then the bar is supported by a different kind of work rest that holds it against the abrasive. Nothing to do with centers, hence 'centerless' which may well now be OFF center relative to finish ground surface. OTOH if it was ground between centers kind of like industrial traversing tool post grinder mode, then we could trust them. But my gut says on long skinny aspect ratio bars, that might would not be optimal.

[55fairlane]

Your spindle has a Morse taper in it, place test bar in the spindle, un supported. That is test 1
Place test bar in tail stock , that is test 2.
Use a center in the spindle (not the chuck) place test bar in tail stock, and supported by headstock center, that is test 3

You can check run out, trueness to the ways, tailstock alignment,  ect

Sorce,  what I was taught in my apprenticeship

[Chipswitheverything]

There is an assortment of factors that can lead to problems with turning acceptably parallel along the length of a workpiece, especially on the relatively lighter lathes that we tend to use in the home workshop. 
Has the lathe been levelled when it was first set up?, and checked now and then for any twist in the bed?, particularly if it is mounted on a wooden bench or stand. If the machine is on raising blocks, such as the ones that many Myford lathes ( including mine )are mounted on, then the adjustment is easy, via the jacking screws.  Otherwise shims may be needed under the bolted down feet.   If an accurate precision level can be used, or borrowed, then the job is easy, otherwise the procedure that the Myford handbook suggests, with a parallel bar, and DTI can be used.  Unless you are quite confident that a twist or strain has not been introduced into the lathe bed, then other adjustments are not really possible to accurately determine.

 The tailstock set over can be adjusted to neutral in various ways.  A test bar with centres, of known quality, supported by good centres in h'stock and t'stock, and checked with a DTI , is a useful way.  Or a centred BMS bar, say around 10" in length, 3/4 " dia., roughed down to leave a couple of collars upstanding , one near each end, can be supported by centres and driven by a "dog" , and very light cuts taken over the collars at the same undisturbed tool setting, to help adjust the tailstock set over to achieve equality at either end.  A quick check can also be made by mounting a DTI on a mag base on the lathe faceplate or catch-plate, and doing a turn-around test against a centre in the tailstock barrel.  Dave

[vtsteam]

Ditto on first checking if lathe bed is twisted, first, via levels.

If in doubt, then, you can check a pre-ground test bar for bow, and for center hole accuracy by supporting between centers, and putting DTI at each end in turn, and the middle. Rotate the bar for each test. Both ends should remain stable on the DTI. If not the bar's center at the end with the error is incorrectly drilled. If stable at both ends, but reading varies in middle, the bar is bowed. If all check out then do the longitudinal tests per 55fairlane and others above.

ps. use a known good dead center for the tailstock end -- there are some problematic live centers, I've found.