Announcement

Collapse
No announcement yet.

How are precise screws made?

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • How are precise screws made?

    I don't want to hijack DICKEYBIRD'S thread on collets, so I'll ask this here.

    Many years ago I read an account on the making of the lead screws for the ruling engine used to make diffraction gratings with thousands of lines per inch. These gratings were used by Michaelson and Morley in some of their experiments. I think the article was in Scientific American, but my searching has not found it. Anyone have a link? Evan?
    Jim

  • #2
    One of the techniques I have read of involved multiple lead screws geared together and coupled by levers to average the errors of each to produce a screw that was more accurate than either. Repeating the process but involving 'child' screws lead to increasing accuracy.

    Comment


    • #3
      I believe I read the same SA article some years ago. I believe he has at least part of it correct in the use of multiple lead screws. Techniques like bridging across two nuts on parallel screws to average the errors. One screw is mounted as it was cut and the other one was reversed, end for end so those errors were in different locations.

      Also each screw can have two nuts on it with a linkage to average the errors between them. Or one long nut. This can help reduce random errors. Of course, you need to start with a longer screw and the second generation screws are somewhat shorter. Then the third generation screws are shorter still.

      Screws can be lapped with long nuts to average the errors along them.

      There are many techniques that were used and they are combined to make new generations of screws of ever increasing accuracy.

      Once you have a screw that is, perhaps one meter long, and are satisfied that it is as uniform as you can make it, then it can be compared with a standard meter. In the old days this was two marks scratched on a length of metal. A microscope would be use to observe those marks and to determine their centers. Today it would be done by counting wavelengths of light of a certain wavelength or some such method.

      For diffraction gratings to be useful for scientific purposes, the lines must be both uniform in their spacing and that spacing must be known to a high degree of accuracy. Oh, and of course they must be straight, again to a high degree of accuracy.
      Paul A.
      SE Texas

      Make it fit.
      You can't win and there is a penalty for trying!

      Comment


      • #4
        http://adsabs.harvard.edu/full/1951AJ.....56Q.120B

        Also http://www.opticsinfobase.org/ao/abs...uri=ao-1-4-415, if you have access to a college library or somewhere that has Applied Optics back issues.

        Unfortunately, all of the good papers on this require subscriptions, so i can't point to free versions. The first one noted above is really worth looking for as it describes the process of making the screw in detail. The second paper describes how the last residual errors in the screw were later compensated for by using an interferometer to control the carriage feed. The screw in the Mount Wilson engine was single point threaded (with an odd thread profile) then lapped with a long split nut. The screw was 20 inches long, the lapping and drive nuts were 8 inches long. After lapping the screw, the drive nut and screw were lapped together, this process took 145 hours over 5 months.
        Last edited by Optics Curmudgeon; 11-08-2013, 12:16 AM.

        Comment


        • #5
          Does anyone know how the generated accurate and uniform screws were used afterwards, because what I could tell there is no kind of control to the pitch of the screw to be exact, as it is cut & lapped? For example, after manufacturing the pitch is measured to be 1.001 mm. How one uses such a screw (to do something = what) ? This has always puzzled me.

          And there should be a book or something about this whole deal of starting from scratch to achieve flat, right angles, screws etc. You know, a method to produce various things that are now common enough to go unnoticed unless one thinks about them
          Amount of experience is in direct proportion to the value of broken equipment.

          Comment


          • #6
            Hmm- what is the el defacto standard length, and how was that determined? Is it the meter? How does the inch system fit in with this- after all there are measuring instruments still made to this day to a high degree of accuracy in the inch system. I know the conversion factors- I'm only asking about the standard lengths, and what is used or kept as the ultimate reference for them.
            I seldom do anything within the scope of logical reason and calculated cost/benefit, etc- I'm following my passion-

            Comment


            • #7
              Originally posted by darryl View Post
              Hmm- what is the el defacto standard length, and how was that determined? Is it the meter? How does the inch system fit in with this- after all there are measuring instruments still made to this day to a high degree of accuracy in the inch system. I know the conversion factors- I'm only asking about the standard lengths, and what is used or kept as the ultimate reference for them.
              Meter is define by light speed in vacuum, and inch is exactly 25.4 mm, so basically they are both defined from the same source. There is no standard (physical object) for either one to compare to.

              Edit: In essence, anything done in the inch system is basically done in metric measures, as the inch length is defined in meters.
              Amount of experience is in direct proportion to the value of broken equipment.

              Comment


              • #8
                The Meter was originally derived from a fraction of the Earth's circumpherence. Not a good source as that distance changes and kept being measured with more precision. Once a 'fixed' Meter was decided upon, a physical 'standard' Meter was produced and all other Meters derived from it via comparison. That standard still exists, but is no longer one Meter long beacuse it has been deteriorating. The Inch is now defined as a fraction of the Meter.

                Producing a reasonably accurate screw is quite simple and you don't need another screw to do it. You move the tool post along the bed with a wire wrapped around a drum mounted to the spindle.
                Paul Compton
                www.morini-mania.co.uk
                http://www.youtube.com/user/EVguru

                Comment


                • #9
                  Originally posted by J Harp View Post
                  I don't want to hijack DICKEYBIRD'S thread on collets, so I'll ask this here.

                  Many years ago I read an account on the making of the lead screws for the ruling engine used to make diffraction gratings with thousands of lines per inch. These gratings were used by Michaelson and Morley in some of their experiments. I think the article was in Scientific American, but my searching has not found it. Anyone have a link? Evan?
                  I believe it was the 1975 April 1975 "amateur scientist" column in Scientific American according to the index here:
                  http://amasci.com/amateur/sciamdx.html under "Ruling engine"

                  I have the CD with all of the collected Amateur Scientist articles (available here), I can verify that tonight if you like. I recall reading it ages ago, fascinating stuff.

                  Comment


                  • #10
                    Originally posted by Jaakko Fagerlund View Post
                    And there should be a book or something about this whole deal of starting from scratch to achieve flat, right angles, screws etc. You know, a method to produce various things that are now common enough to go unnoticed unless one thinks about them
                    In the Niven / Pournelle novel "Lucifer's Hammer", there is a cameo role of the scientist who saves civilisation in the aftermath of a catastrophic comet strike on the earth by caching all his books and reference works that contain enough knowledge to bootstrap an industrial-age technology base. Didn't the Conneticut Yankee at the Court of King Arthur bring an encyclopedia too?

                    I wonder if there is really such a doomsday technical archive somewhere? (I know there is a doomsday plant-seed vault in the frozen Arctic)

                    The great copyright libraries (British Library, Library of Congress etc) and the archives of the various patent offices would contain the information, but the problem would be finding it and sorting it out from all the other knowledge in those places.

                    Comment


                    • #11
                      Moore were famous for their ability to make super-accurate screws, as opposed to eg. SIP who preferred to use a linear cam to correct the pitch by moving the fiducial line.
                      I've never seen how Moore did it. Does anyone know or is it still a trade secret?

                      Comment


                      • #12
                        Originally posted by Jaakko Fagerlund View Post
                        Does anyone know how the generated accurate and uniform screws were used afterwards, because what I could tell there is no kind of control to the pitch of the screw to be exact, as it is cut & lapped? For example, after manufacturing the pitch is measured to be 1.001 mm. How one uses such a screw (to do something = what) ? This has always puzzled me.

                        And there should be a book or something about this whole deal of starting from scratch to achieve flat, right angles, screws etc. You know, a method to produce various things that are now common enough to go unnoticed unless one thinks about them
                        The screw, if known even but with a known error, can be used to guide a cutter to make a screw with that error compensated-out, using some method, perhaps as crude as a lever system. If you KNOW you are 0.01mm too long, you can devise a system to cause that error to be removed.

                        A screw of 1mm pitch arranged to make 0.01 turn while the "master" screw moves a nut the whole length, can be set up to move the tool back that 0.01mm evenly over the whole length of the master (but in error by 0.01mm) screw. Obviously if the compensating screw has a finer pitch it can move more of a turn, and likely be more accurate.
                        1601

                        Keep eye on ball.
                        Hashim Khan

                        Comment


                        • #13
                          Thanks for all the replies. I just thought it would be an interesting read. The subject of how precise threads were made with the equipment available in the past does come up here occasionally.
                          Jim

                          Comment


                          • #14
                            I believe Moore laps their screws with a nut.
                            An accuracy graph of sorts is generated by comparing
                            results from a series of linear measurements from a
                            master set of blocks (or whatever). The graph is kinda
                            like the SIP linear profile correction cam. Then the graph
                            is placed above the screw to be lapped. The operator
                            then pressures the nut either to lead or lag on the
                            screw, whichever way the graph calls for.
                            Then the screw is re-measured and another graph is
                            genreated, and the differential lapping is repeated
                            a few times.

                            --Doozer
                            Last edited by Doozer; 11-08-2013, 12:21 PM.
                            DZER

                            Comment


                            • #15
                              There's a thorough history of accurate screws, dividing engines, etc., to the end of the 19th century, available in the form of Randall Chapman Brooks' PhD thesis - The Precision Screw in Scientific Instruments of the 17th - 19th Centuries: With Particular Reference to Astronomical, Nautical and Surveying Instruments. It's available online as a big PDF, via this site:-

                              https://lra.le.ac.uk/handle/2381/8446

                              Comment

                              Working...
                              X