No announcement yet.

using a straight edge to check flatness

  • Filter
  • Time
  • Show
Clear All
new posts

  • using a straight edge to check flatness

    i picked up a 24" Starrett straight edge and finally got around to attempting to use it. i wanted to check the bed on a lathe. yes, i know a 48" straight edge would probably be better, but the exponential cost increase from 24" to 48" pretty much shot that idea down.

    anyway, how do you check the v-ways on a lathe? do you just kind of hold the straight edge at an angle so it is flat on the v-way bearing surface? are there any other tricks to it?

    i tried it on the old South Bend 11" i have and i can't fit any feeler gages under it. does this mean i'm good over 24" for 0.001" (or whatever the thinnest feeler gage would be)? on the old LeBlond lathe you can see just a bit of the remaining frosting on the end of the v-ways by the headstock. the rest of the bed is pretty much smooth. from the headstock out 24" (the length of the straight edge) again, no feeler gages fit under the straight edge. how deep does the scraping normally go?

    it just seems to me that there should be some measurable wear on the bed, but maybe i don't have a long enough straight edge to measure it. of course, if most of what i do is within 24" of the headstock, it would appear i am good to go.

    anyone have any thoughts, ideas, or suggestions?????

    andy b.
    The danger is not that computers will come to think like men - but that men will come to think like computers. - some guy on another forum not dedicated to machining

  • #2
    How about building one of these....
    Something like this will detect smoothed out dips in Vways over the entire length of your bed.Or for a cheap alternative, how about some taut piano wire mounted on equal blocks checked with an inside mic. That was (and maybe still is) a standard tool for checking alignment and wear for many years.Very cheap and with practice and/ or a cheap continuity tester can register accurate results.

    [This message has been edited by lklb (edited 03-31-2005).]

    [This message has been edited by lklb (edited 03-31-2005).]


    • #3
      oops! I edited out half of the last message.

      A couple years ago I helped a friend get his old lathe working. He had a straight edge about as long as yours . We could not detect substancial wear on the V ways with it. Still it didn't cut straight. We played with the tailstock ,still didn't help . He finally rigged a taut wire between centers.Got a junk microscopes to read the wire in 2 planes, first on the carrage then on some V blocks attached to make a half assed template.We found the wear with this- a lot more than he expected. He checked out regrinding costs,almost sh## his pants, put it back together,sold it for a tidyprofit. I don't know that much, but it seems to me that the relationship between 2 V ways can not be adequately inspected with just a straightedge. That is why I included the link to the alignment detector patent,and mentioned taut wire stuff. If it cuts straight don't worry about it.


      • #4
        Yes indeed; The link takes you to the patent description for the justly famous Kingway Alignment System, a box-o-parts that allows you to determine and quantily bow, twist, etc of any prismatic shape within its range. They have several advantages in time and convenience over lasers for quick in process checks. I've used this system many times in the past and found it to be almost infinitely adaptable.

        Rich King still sells them from his existing stock. Pricy though. Not. Cheap.

        [This message has been edited by Forrest Addy (edited 04-01-2005).]


        • #5
          Would it be possible to use prussian blue in this situation? I have used a straight edge and prussian a few times to find flatness on vise bottoms. I also use prussian on a surface plate and put the part on the plate and give a light rub to find high spots.

          Just an idea, never tried on lathe ways.
          CCBW, MAH


          • #6
            <font face="Verdana, Arial" size="2">Would it be possible to use prussian blue in this situation? I have used a straight edge and prussian a few times to find flatness on vise bottoms. I also use prussian on a surface plate and put the part on the plate and give a light rub to find high spots.
            Just an idea, never tried on lathe ways.
            Of course you can,but that still isn't gonna tell you anything about the relationships between the other surfaces. Just on top of some V-bed lathes, I count, What?, 7 surfaces that need to be in alignment with each other.That's where this thing or some home built rig comes into play.


            • #7
              I have got to get me one of those, or make me one of those. Eight lathes to level, and just one precision level to do it with, and a mile of wire I still have trouble figuring out at times...... This could save me hours
              CCBW, MAH


              • #8
                Can you give some details on using wires? I need to scrape my 9C and this sounds like something I could use. I see the basics of it, but what do you measure to as a reference? Just get some flats true with a straightedge then measure from them? If I make a small block to hold an indicator, I think I could tickle the underside of the wire for readings, instead of an inside mic like I've seen mentioned. Or is it the distance between the blocks that you watch?

                [This message has been edited by Smtty58 (edited 04-03-2005).]


                • #9
                  If your Starrett straight edge is like mine, it is not the correct tool to be using, it will give a close idea. The tool needed is a "scrapers" straight edge, some good photos are shown in the Reconditioning article starting in the Sep/Oct '04 HSM, photos #6 & 17. Look at photo #21, the inside flat way has been spotted with the 4' straight edge, in the lower corner is the headstock section, and the upper right hand is the other end of the straight edge, the gap in the middle is the "wear".
                  The Kingway Alignment System would have been a very welcome addition for checking progress, especially after at least one way was completed.


                  • #10
                    <font face="Verdana, Arial" size="2">Can you give some details on using wires?</font>

                    This is a taut wire jig,looks like it's used for big stuff (turbines and power plants). But the holders are very simple, adjustable for height and if you like cross travel.
                    The Mike has to be insulated on one side so you only get a signal in the headphones when you touch the wire.I would think a lighting device should work also.

                    You have to tension the hell out of the wire to eliminate sag or if you have a big machine,you have to compensate for catenary sag.
                    This is simple.. Get very fine piano wire and find out how much it weighs a foot.
                    Than go to machinery's( I got the 14 ed. in front of me ,page442-443)and using the conversion tables, determine how much weight you need to put on it to bring it near yield without breaking. Go to the site calculator below to get your sag.( careful it's in ft.)


                    Using this is pretty self evident, pick two reference points, and it easy to measure variations between them.....

                    [This message has been edited by lklb (edited 04-03-2005).]


                    • #11
                      Wow, thanks for all the pics! I've never heard of this method before. Looks a lot easier to make than the little gantry with the levels.


                      • #12
                        You can also check some of the alignment issues with a parallel test bar in the lathe (or mill) spindle. As long as it's parallel you can get it near center then measure your min & max and take the average. With that and some time you can figure out most of the error in a system.

                        The nice thing with that method is that you're usually checking with the lathe carriage, and so you're seeing with the tool will see. Sometimes that helps.

                        [This message has been edited by rkepler (edited 04-04-2005).]