In photography the Zone System is used to fully understand the peculiarities of equipment and the way they are used. It involves taking rolls of pictures (less applicable to digital) to not only determine optimum film speed, developing times, exposure times et. cetera but to be able to anticipate the end product.

Have any individuals developed such a routine for understanding their lathe? Of course there are a lot of "depends on" but it seems to me that a systemic approach would be more productive than continually doing a baby's step before every measurement. It seems to me that pulling the tool away to measure introduces more error than trusting what you know of the machine's response.

I imagine that having a method for working around the short comings of the machine would be very much part of this strategy.

It would be very helpful to find out how some of you were able to "know" or discover the capabilities of your machine.

Raymond

Ray, not to burst your bubble but the DRO is a solution to the measurement problems.

It will compensate for all kinds of wear in the leadscrews.

You make one measurement, reset your zero point and away you go to finished dimension.

Subsequent parts can just be machined without checking and checking.

Of course on the finest work it still makes sense to tread lightly and carefully and not proceed like a bull in a china shop.

-Matt

Hi Raymond,

If $750 "Trav-A-Dial", or $1,500 DRO is not an option, set a $35 a 1" Dial Indicator on a bracket with an extended rod against the cross slide and you have a poor man's "high-tech" system in place.

It works like a charm, guaranteed!

Here is a dial indicator on the crosslide taper attachment tounge on a sliding mount, an electonic gage head can be mounted in it also.
Travadial for the carriage.

http://www.metalworking.com/DropBox/.../Don%20007.jpg

Thank you for the response. I am aware of the DRO solutions (the Heidenhain 780 is slow in being released and the Positip 880 is expensive) but I was thinking more in the line "How do you experienced people push the limits?".

I don't expect all your hard earned secrets just the ones that you are willing to share so that I can attempt to apply them to my situation. Of course others will also benefit from the information.

Raymond

I would agree that a high precision DRO goes a long ways to helping discover what the lathe is capable of doing, but the DRO does not create a technique simply by its prescence on the machine.

On a manual machine used for cutting steels, the tool edge exists in a state of variable wear. Learning to compensate for wear simply takes lots of practice and if your cutting methods are standardized, you can discover a typical wear life. This is done on cncs all the time.

What you really need to determine on your manual machine is its repeatability. If you do exactly the same thing over and over, you should expect to see exactly the same results (with a little variation due to tool wear) every time. If your machine is not repeatable in results, then it needs some work done on it to fix whatever causes the random errors.

Other than that, I generally adopt the procedure of using one tool for roughing and another for finishing on the manual lathe. On cnc, this is not strictly necessary, because with better coolant application, better spindle speed control and higher speeds generally attainable, a single tool can be used to produce moderately accurate surfaces right from rough to finish, for quite a few parts.

For high precision over many, many parts, a seperate finishing tool still reduces the headaches of locating the point and breaking in a new tool even on cnc.

On manual lathes, one source of random error arises from low accuracy quickchange toolposts. For this reason, I prefer the 40 position quick change toolpost, as the repeatability is excellent. I've not used the more common 4 sided toolpost enough to determine the repeatability rate. Needless to say, if you have to change tools, and you cannot trust the location after the change, then you have to baby step every cut to reestablish your trust in that tool position, even with a DRO.

Other than that, learn to take finish cuts that are consistent at one depth of cut, perhaps .010 to .050" depth, but pick a value you (or the job) are comfortable with and stick with it. On the first part, deliberately plan your cuts to leave exactly enough material for two finishing cuts of exactly the same depth. On the first part, cut and measure after the first finish cut. Hopefully, you will follow this with another cut of exactly the same depth, or very close. At this time, you can reset your DRO so that is reading correctly, then take the final cut. The size should be correct.

On the second part, now you can dispense with the two finish cuts and take only one, but still, the last rough cut must accurately leave the same amount to finish as you measured before the final finish cut on the first part.

Its like an experiment: keep all conditions the same, you should get the same results. If not, the machine itself may have a problem that needs attention. Either fix it, or take the baby steps approach to each finish cut.

Well, the only other "hard earned secrets just the ones that you are willing to share " is to retrofit your a-axis (cross-slide) to CNC.

Ooops, read an x-axis.

I suppose in photography you may be referring to a 9 negative series.
In holding close tolerance on one part. Dont know just how you do that on a CNC machine fiddling around with tool wear,spindle and work deflection and all that great stuff that they do for more than one item.
I can tell you that the local National lab uses manual machines for the one off part.
Myself never needing to produce much more than one item or tool.
Observe that a part held in a chuck and extended just a couple of inches, it will be larger on the outer end from deflection from the cutting forces.
So the answer would be to reduce the cutting forces that push the work away from the tool.
Might depend on your requirment, Even taking a depth of cut of .010" will deflect a part several tenths in 2".
Also depends on the manual machine, the one pictured above is considered a pretty good one. Mounting a DRO or even a CNC control, will not enhance the accuracy for making one part.
As far as secretes are concerned, the indusry Iam in is full of secretes, the main one is, those that have the secretes, are hiding what they dont know.

Don

Knowing your eguipment is one thing but knowing your set up is another.

There are a number of variables that affect the reliability of a machine. Is the toolbit ground right, set up right (on center) and is it sharp. Say "NO" to all three of those conditions and I'll make your best lathe perform like crap.

Another variable is the rigidness of the workpiece. For example, a 1" diameter workpiece extending 2" from the chuck or collet is one thing. But, a 1/4" diameter piece extended 2" is another. A live center supporting the end helps the end but how about the middle?

Okay, let's say all cutting conditions are optimized with sharp tools and a rigid set up. It is up to the lathe and the operator now. One must know how much material to remove to "HIT" a certain dimension within the desired tolerance. Some lathes can remove .020 in one pass and hit +/-.0002 while others can only remove .005" and hit the same target diameter. Some lathes even less. A person needs to run some tests of the amount of feed and depth of cut to obtain desired results.

I have a DRO on my lathe and love having it. However, when working to less than .001", I rely on the lines on the feed dial.
Jim

One item I found of importance in what Iam doing( holding a .001 overall tolerance in M/C engines) is a consistant tool, and angles to suit the material.
In this cutter grinders workhead is a simple vise and a lathe finishing tool being ground.
With care a tenth or better can be held in the Monarch EE lathe on a 5/8" or larger piece of toolsteel in a 2.5" length. For taper and diameter.
The depth of cut not more than .00075. Feed at .0005 rpm close to 4000.
Soft carbide (hard carbide will chip and spoil the work)
Lube, a grease like paste made from beeswax, moly dee and 111 tricloro.

thats what I do, if it dont work for you ( oh well)
http://www.clubchopper.com/photopost...cat=500&page=2

Don

"Have any individuals developed such a routine for understanding their lathe?"....
it's just like shooting pictures....
cut lot's of steel......take lot's of notes....
because like photography, every setup is differant....so the best you can do is grow in your knowledge of the "jumping off" point's
.
F8 and be there baby!

Thank you all. Here are a few responses all grouped together.

HuFlungDung:
"...adopt the procedure of using one tool for roughing and another for finishing"
This eluded me yet it makes sense to keep the sharpest tools for finishing i.e. recycle the finishing tool to the roughing bin - or do you recommend using a different insert for finishing than roughing.

"...error arises from low accuracy quickchange toolposts" Darn it, I just bought an Aloris CXA to replace a four position turret (thought the BXA might not be stiff enough). My intent is to machine a jig for consistent tool placement within the holder. What make of 40 position holder are you favouring?

"...finish cuts that are consistent at one depth of cut" Yes, that should help consistency.


Don:
" Observe that a part held in a chuck and extended just a couple of inches, it will be larger on the outer end from deflection from the cutting forces." Don this is exactly what I am refering to. I realize that it would be a pain to do parallel machining with the taper attachement set at an impossibly small angle but it might resolve the problem. Probably there are easier ways but quite often the biggest headache is identifying the problem.

"a consistant tool, and angles to suit the material." I had visited your web site and I like your sharpening setup (vises on a post is clever). Of course I'll try to copy as much as I understand.

Jim:
"A person needs to run some tests of the amount of feed and depth of cut to obtain desired results." I understand the other variables you mention, a person will still gain knowledge by running careful tests even though there is a bit of fuzzyness with the results.

"However, when working to less than .001", I rely on the lines on the feed dial." I'm not certain I follow this. Do you advocate using a dial indicator measure Carriage/Cross-Slide/Compound movement? I thought a DRO would measure to .0002"

wippin' boy:
"...take lot's of notes..." You raise a fundamental procedure.

Raymond

"I realize that it would be a pain to do parallel machining with the taper attachement set at an impossibly small angle but it might resolve the problem."

Methinks if you are turning taper you have a headstock to bed misalignment problem or wear on the bed.

A lathe like the 10EE will turn sub thousandths but if you talk to owners they are almost a separate class of machinery Most lathes will work to .001" or .0005" with care. Smaller than that and you are really into grinding or lapping IMHO.

It is also my opinion that workpiece deflection that causes taper can be minimized in the setup. Turn between centers and the part gets 48X as rigid as it is when cantilevered (check beam deflection equations for point load cantilevered vs. point load double-supported for confirmation I'm not just blowing smoke).

-Matt

It might seem simple but you can use ordinary callipers ,jenny callipers and micrometer screw guages to get the accuracy level at which most machinists need to get bits to fit together. DRO is very nice and cnc is also great for the production of accurate work.
when the job is for a hobby and no profit other than satisfaction is required then the simple stuff might be your challenge for the day.
On the other hand your discussion is a really good way to talk about it instead of going out into the shed and winding up the machine.
Regards,
Bobby.

Raymond,
I'm not advocating that the 4 way dovetail toolpost is not repeatable, it is just a fact that I have never tested this aspect. I just so happened to buy a used lathe years ago, that had the 40 position toolpost on it, and I fell in love with it, and soon after, converted all my lathes to that style They are a little pricey ( even at www.shars.com ), but if you need 'em, you need 'em.

If you have trouble turning a true cylinder (with a sharp tool) on a part that is hanging unsupported out of the chuck (for a reasonable distance), check that your chuck jaws are a snug fit in their slideways, and perhaps regrind the jaws so as to eliminate the bellmouth jaw taper that is common to a well used chuck. If the part is solidly held, this will help a great deal in getting straight turned cylinders.

Finally, if you have checked that out, but still have a little bit of taper, I would suggest that you adjust the levelling screws on the tailstock end of the lathe bed. If the taper of the part is larger at the outer end, then adjust the foot screw on the operator side of the lathe so as to lift the bed on the front side. IMO, this type of test cut is the first thing I do after relevelling a lathe bed, as "level" means nothing if the lathe will not cut true cylinders. After this adjustment, then you can fiddle and fuss with the tailstock setover for turning cylinders between centers.

Hu F.D.
Actually I've had some challenges holding 1.5" shafting in my 3 jaw chuck. The shaft is bored and reamed to 1" which appears to make it flexible even at 3" when mildly attacked with .085" Sowa carbide parting tool. I was trying to cut some .040" spacers and it felt more like a random number generator. There was somekind of pattern i.e. the carriage movement was gradually increased (up to .005")as the cutter moved towards the chuck and a delicate starter cut. I have been looking for a good 10" 4 jaw - now the motivation is even higher.

Raymond

turning round in a 4jaw will add much time to your process for indicating every new hold...
more info on the .005 change please....
do not quite understand what you mean
are we talking linier taper or overall length changes in the spacer?

Gradually increasing the amount the carriage was moved (as it cut each spacer going towards the chuck) is what gave some kind of thickness consistency. My guess is that the chuck does not hold well and that even though the starting cuts were light the shaft may have moved slightly. The shaft run out was about 1 thou. It looks like the chuck is due for somekind of tuneup.

Raymond

"Turn between centers and the part gets 48X as rigid as it is when cantilevered "

I get 16X. Gotta dig out the books and see which of us is wrong.