I'm needing a one inch lathe test bar. Finding ground and polished one inch material is a piece of cake. The bar needs centers on each end. If I just chuck the bar in a one inch collet and bore, I'll pass on the 0.0005 inch runnout of my collet to the test bar.
How would you do this? Or do I just buy the test bar?
Karl
Karl, I'd start with a bar over desired size and put centers in both ends. Then machine/ grind it (adjusting the tailstock to get your diameter true end to end). Once done, you've set your tailstock true AND made your own alignment bar for use the next time it is needed.
"Karl Townsend" wrote in message news:SwIZe.3111$ snipped-for-privacy@newsread1.news.pas.earthlink.net...
Best chance of success is to grind the test bar after the centers are drilled. Precision work is generally accomplished in that fashion unless there's a good reason to pursue other options.
Assuming you buy ground and polished material, be advised that it isn't always what it's cracked up to be. It's not uncommon for long pieces of material to be run through centerless grinders or sanders. Unless they are dead straight, they often are not perfectly round, nor true to size. If you're worried about a few tenths, and you should be, I really don't think you're going to be happy with a test bar that isn't ground as I suggested.
Harold
If you find a good good true round bar and center it and it ends up with your 0.0005 runout you could extend it out of your chuck and use your compound to bore in the center on the angle after you have tapped the bar true. Jim "Karl Townsend" wrote in message news:SwIZe.3111$ snipped-for-privacy@newsread1.news.pas.earthlink.net...
If you have a good chuck and soft jaws you could bore it size and try that. Or you could try to bore a soft 5c collet. If you try that bore it about .001 undersize don't go oversize or it will not be true. Jim "Karl Townsend" wrote in message news:SwIZe.3111$ snipped-for-privacy@newsread1.news.pas.earthlink.net...
I guess my first question is, 'what are you going to be testing with the bar?'
Is it going to have a taper in the end, for checking the spindle of the machine? Or is it to be mounted between centers, for dialing in the tailstock?
The trouble with manufacturing a test fixture like that, on the machine it will be used on, is that the machine probably cannot easily impart the degree of accuracy needed (I would like to see nearly a whole extra decimal point - ie the test fixture should be measurabley accurate to 0.0001 if you are using it to dial the machine in to 0.001) and it sounds like that would be tough for you to achieve, given the machine.
Jim
If I may "go off half cocked" and without giving it a lot of thought (popular reaction lately on RCM): It seems that the 0.0005" might be absorbed, so to speak, by making the drilled center slightly larger but with a correctly shaped 60 deg. cone. In other words, the point of the center drill would drill larger by the amount of runout and then guide the cone in correctly. I know, intuitively, this is not right - but help me out here.
Bob Swinney
Uh oh. You just pushed Harold's hot button!
I'm pretty sure I know what he's gonna say here....
Jim
Think about something here. If you were to take and chuck a piece of stock in a four jaw chuck and deliberately offset it by an 1/8 of an inch, then drilled a hole in the end, how close to center do you suppose the hole would be?
If you answered 1/8 of an inch, now consider why it is not closer to center than that.
Once you can answer that question you should see the error of your logic.
(How did I do Harold?)
Run it in a steady rest and shave a thin chip off the 60 degree center hole with the compound. Even if it isn't exactly centered it should rotate around its central axis.
jw
:-)
Yep! When you're looking for that level of precision, the center becomes all important. That's one of the reasons center laps were made, aside from re-locating centers when a thou made the difference between a good part and scrap.
What Bob suggested works, but only to a degree. The biggest problem is it's not predictable.
What Jim said is on the money. You can't make determinations to tenths when a test bar has more error than the one you're looking for. Standard practice in the tool room where I was groomed was to have a tolerance of only 10% of the product tolerance for tooling. It wasn't uncommon to find dimensions with only a tenth tolerance. It makes sense if you want proper results.
Harold
Frankly, quite well. Nothing magical comes from machining----you get out what you put in----rarely achieving a silk purse from a sow's ear. In order to work with precision results, you *must* use precision methods, otherwise you're at the mercy of the luck of the draw. A blind squirrel finds the occasional nut, but one with keen eyesight eats heartily.
Harold
Rarely would that yield the desired results, although it would generally improve the overall condition of the typical center. Grinding the part using lapped centers assures proper results. Anything less is a crap shoot.
Harold
Harold sez;
"> What Bob suggested works, but only to a degree. The biggest problem is it's
Story of my life.
Bob Swinney
It would be almost exactly at the center!
Of the lathe spindle's axis, that is. (the only reason it would not be is because the centerdrill would wander as it started out.
The trouble happens when the test bar is then *used* between centers. First off it would have to be trued up so the external part of the bar "agreed with" the centers that didn't line up. This would mean putting it in the same lathe between centers and taking a cut on it. The degree of tailstock offset is still unknown so it becomes a bit of trial and error to get the outer surface truly cylindrical. If the tailstock is low or high that makes it even tougher.
And in the end he would wind up with a the centers on the test bar not bearing fully on the fixture it is mounted in, so in spite of the outer surface of the test bar being in *principle* truly cylindrical, the end result will be far from useable.
A thought experiment like this where you deliberately do something far out of alignment to see 'what happens then' is a nice way to understand how those parameters actually affect the final accuracy of teh finished item.
Jim
How long is the bar?
I've used a mandrel as a test bar before finding good cylinder squares. I bought two cheap at auctions because none of the machinists present knew what they were.
The mandrel is tapered but you can reverse it to indicate the same diameter at both ends. Plus they are useful to turn wheels and pulleys.
jw
PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.