On Friday 08 September 2017 15:32:59 John Kasunich wrote: > On Fri, Sep 8, 2017, at 02:19 PM, Gene Heskett wrote: > > Greetings all; > > > > > > I have measured the x axis of this lathe quite a few times, and > > arrived at an X scale that moves it 1.000 inches according to a dial > > indicator. > > > > However, when carving metal it doesn't seem to be correct. > > Part of the problem is almost certainly the difference between > unloaded moves and loaded (cutting) moves. Tools flex, toolposts > flex, machine flexes, etc.
True, but I'd go so far as to say its a heck of a lot stiffer today than when it was shipped to the naval yard circa 1952. No willow tree compound exists now as it was badly damaged when I took possession. So the computer drive is my compound. > When you are trying to set the "scale" of an axis, you want to use the > longest travel that you can accurately measure. That minimizes the > effect of things like lash, flex, etc. > > You can't consider the scale on an axis to be correct until you can > command multiple different moves and have them all be correct. > Command a 0.1", and the dial indicator better say 0.100". Command > a 1" move, and the dial indicator better say 1.000". To this point it does John, but thats the travel length of my biggest dial. But using the stinger of 3 or 4 of the $30 6" calipers all seem to agree that a 1" move command from axis, is less than an inch. Small disagreements of course. Humm, I also have a digital hole depth gage with 6" of screw on probe adapters. But that would be the last thing I'd trust for a real reading unless I take the time to calibrate a 6 pack of screw in extensions. So I'll first see if I can mount the 8" scale I was going to put on the tailstock and get some good readings from it. > Command a 6" > move, and dial calipers or indicator plus a 6" gage block or whatever > you can come up with better say 6.000". And you should be able to > make and measure a whole sequence of 1.000" moves all the way > down the length of your axis. [...] > There are probably two things at play here - flex and touch-off error. > (Scale error could be happening too, but IMHO you shouldn't cut > metal until you've verified scale with multiple measurements as > described above.) > > I find it hard to do a really accurate touch-off on the inside of a > bore. And even on the outside (where I gradually back the tool > away from the part until a 0.250" dowel pin slips thru the gap), > the touch-off is done with no load on the tool and is subject to > change due to flex. > > In your situation, I would do the following: > > 1) Bring the tool into the 0.750" nominal old bore until it just takes > a tiny chip. Touch off X to a radius of 0.375, diameter 0.750. I > know this is probably wrong, but it gets me close. > > 2) Set X to 0.385 (0.010 depth of cut), which should give me a > diameter of 0.770. Take a cut on the bore. Get the tool clear and > measure the bore as accurately as I can (bore gage, telescoping gage > and mic, whatever). That measurement tells me what the machine will > do when cutting 0.010 deep. It will be different for heavier or > lighter cuts, but I'm planning to finish my job with a 0.010" cut. > > 3) Reposition the tool back to X = 0.385 (using MDI). Use the > touch-off button to reset the diameter to the exact measured bore > diameter. Now I'm corrected for flex and anything else, and as long as > the tool doesn't wear or something else bad happen I should be good. > > 4) Make roughing cuts at whatever depth of cut I can do until I get > close. Final roughing cut should be at X = 0.4137, that is 0.020 (two > passes) undersize on the radius. Measuring the bore after the last > roughing cut isn't critical, but I usually do it just as a check. > > 5) Make my first finishing cut, with X = 0.4275, target diameter is > 0.855". This cut is 0.010" deep, just like my calibration cut, so the > flex should be the same and it should come out pretty darn close. > > 6) Measure the bore as accurately as I can. This one really counts. > Maybe the tool wore a bit during roughing, maybe something > moved, or maybe the flex or screw error is different since we're > at a different spot on the axis. This measurement will let me fix > that. > > 7) Move X to 0.4275, exactly where it was when I made the cut. > If the measured diameter isn't exactly 0.855", hit the touch off > button and set X to exactly half of the _measured_ diameter. > This should set me up for exactly one more pass, cutting 0.010, > which should have the same flex and everything else as the > previous pass. > > 8) Make my final finishing cut, with X = 0.4375, target diameter > 0.875. > > 9) Measure and hope I didn't screw it up. ;-) > This procedure is NOT appropriate for production - I stopped and > measured twice during the job, and probably modified the g-code > a couple times (in some cases I don't even bother with a program > except for the roughing, I make the first and final cuts with MDI). > > But for a one-off like Gene is describing, I've found this process > to be the best way to get the diameter I'm aiming for. Which is essentially what I did. And got the wrong size. :( In the right direction so it was eventually correctable to a fit given the rust on the shaft as it was outdoors for 40 years at the local horsepistol. Cheers, Gene Heskett -- "There are four boxes to be used in defense of liberty: soap, ballot, jury, and ammo. Please use in that order." -Ed Howdershelt (Author) Genes Web page <http://geneslinuxbox.net:6309/gene> ------------------------------------------------------------------------------ Check out the vibrant tech community on one of the world's most engaging tech sites, Slashdot.org! http://sdm.link/slashdot _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
