Lathe accuracy

I don't think .0002" is necessarily exceptional for spindle runout. My Chinese 12 x 36 is at least that good. The question of variation in

12" between centers would probably depend more on how well your tailstock center is aligned with your headstock and how accurate your centers are. You should be able to turn something between centers to the accuracy of your spindle TIR which in the case of the Prazi should be .0002".

Center line-up probably is the biggest practical factor, although it shouldn't be hard to get them perfectly lined up, assuming the lathe has normal tailstock adjustments.

In a cheap lathe another factor is straightness of the bedways. A little sag or vertical arch to the bed will have little effect, but a curve in the X-axis will have a pronounced effect. This is easy to detect by turning a sample piece and charting its actual diameter every inch or so along its length. Straight taper is a tailstock alignment problem; anything else is a bedway problem.

Ed Huntress

I hear this question every so often, at work and here.

I don't think I've ever seen a lathe incapable of producing accurate results (such as a lathe should be able to produce). I think they only time this really comes up is when the left end of the ways are worn due to the carriage staying near the chuck.

Other than that, I think it's a foolish question. My standard answer at work is something to the extent of, "it will make parts as accurately as you are capable of making them".

Regards,

Robin

as is my 4015 3n1. well, close, i could only guess, it was less than

1/2 a thou. what impressed me was the the perimeter of the chuck wasn't much worse. now, i cannot actually remember the 3 jaw accuracy, but it was better than i expected, certainly better than the rigidity expected from the lathe. but that is moot, right? i did not have any really "round" stock for testing the chuck.

what, steel? rigidity seems to be the limit on my Grizzly 4015, which weights 380lbs or so. the lathe base is approx the same size as the 9x20 chinese imports. that Prazi does look nice, pricey little devil. --Loren

This is one of those cases where hobbyists have an advantage over commercial machinists. For 100 years, it's been traditional in home shops with limited lathes to lap parts that require extremely accurate diameters. Commercial machinists don't have the time for that kind of stuff. With lapping, you can produce a consistent diameter of 0.0001 in. or better.

And, today, we have those nifty lapping films that make it so much easier.

Ed Huntress

That specification is the spindle runout. They are claiming the spindle runout is .0002 or less. That doesn't mean you can turn with that accuracy.

Factors that would affect your accuracy in real life conditions would be tooling and setting thereof, hardness of the work, adjustment of the machine, feed and speed (does a prazi have power feed?), & coolant use.

Typically when I need to make gauges on my Southbend I use a toolpost grinder because it is too difficult to take off .0001 with a single point tool, and its hard to hit the dimension dead on the first time.

Even in my production days we were expected to turn to tenths with a Hardinge turret machine, and it took a lot of tinkering and adjustment. With turret tooling you make you adjustments one time and lock it in. Plus we had to let the machine run for awhile and let the bearings reach operating temperature.

Ahh yes the warm-up period. When I worked at an aerospace contractor all the union guys wanted to work in grinding or the toolroom. They would get to work at 6 am, turn on the spindles and go have coffee for 45 minutes so their machines could "warm up". I think they pushed it just a bit.

ds

It's very possible they did, but modern machines that work in the region around one tenth or less often need a lot more time than that to warm up. IIRC, Wasino's S-mu lathe settled down at around two hours, which is considered very fast at that level of accuracy (guarenteed +/- 20 millionths). And a C-frame mill of high-accuracy pretensions, such as the Makino extreme toolroom machines of a decade or so ago, can require hours more.

Ultimate toolroom machines, like the discontinued Wasino Wing Ace $300,000+ surface grinder, should be left running all the time.

Ed Huntress

The spindle and spindle bearings are of course going to be vastly different between the Prazi-in-question, and your hardinge example! That right there makes up most of the difference in behavior.

I think the Prazi spec is basically quoting what a dial gage will read, if it clocks the OD of the spindle - which of course is a nonsense number. The real number is, how much does the dial move when you put a broomstick into the spindle hole, and push on it with 30 lbs of force?

The hardinge turret tooling sure does come in 'tenths' micro-adjustable versions - one sees it all the time on ebay.

Jim

================================================== please reply to: JRR(zero) at yktvmv (dot) vnet (dot) ibm (dot) com ==================================================

Turning a 12 inch long shaft with 2 tenths taper on a small machine is a pretty tough job. I highly doubt I could achieve this with my myford or my rockwell. Especially if the shaft has a small diameter.

chuck

One thing often overlooked is that an beginning machinist trying to make accurate parts needs a lathe with more intrinsic accuracy than some old fart who has learned to live with the inaccuracies of his lathe and still make good parts.

brownnsharp

Ummm, no. When I was grinding, the machine wouldn't settle down sometimes for as long as a couple of hours. Not much of a factor on production, but in a tool room situation where tolerances were tighter, (much) the variation would have been intolerable. Most times, minutes saved turned into hours wasted.