threads???

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

  • threads???

    I realize this will probably be a stupid question but.... Measuring a 1/2-13unc-2b external thread on a pin with the datums being the centerline of cylinder of pin and the back plane of thread. Trying to get true position of thread correctly and it is being questioned so just double checking myslef to see that I am constructing this feature correct. Is using a barrel (shank) probe correct and if so does pitch need to be built in ?
    Last edited by atommonkeyz; 11-02-2015, 08:48 AM.

  • #2
    Originally posted by atommonkeyz View Post
    I realize this will probably be a stupid question but.... Measuring a 1/2-13unc-2b external thread on a pin with the datums being the centerline of cylinder of pin and the back plane of thread. Is using a barrel (shank) probe correct and if so does pitch need to be built in ?
    Shank probes, (also called barrel probes), are prone to errors if they are not perfectly aligned to the part. A round ruby is a better choice, the larger the better. Use the pitch function in autocircle. If you are only after location, you don't have to worry about where on the thread form the hits occur. If you are trying to also measure the thread major, you will want to make sure you are hitting as close as possible to the crest of the thread. Depending on your application and the length of the thread, you may want to use an autocylinder rather than an autocircle. In my experience 2 levels that are a multiple of the pitch apart works best. There are many, many threads about how to best measure threads, mostly internal, but the concept is the same either way. Lastly, if you have a scanning probe, you could use a spiral scan rather than the default measurement strategy.

    HTH
    sigpic"Hated by Many, Loved by Few" _ A.B. - Stone brewery

    Comment


    • #3
      This was what I was doing building two auto circles with pitch using 4 mm probe ball and constructing that to cylinder . I have a sales engineer in charge of dispostion of defective product. He doesn't believe in the results I am getting. What other hard gage method could I use to verify or measure this?

      Comment


      • #4
        Originally posted by atommonkeyz View Post
        This was what I was doing building two auto circles with pitch using 4 mm probe ball and constructing that to cylinder . I have a sales engineer in charge of dispostion of defective product. He doesn't believe in the results I am getting. What other hard gage method could I use to verify or measure this?
        What is defective?

        Location? - Surface plate layout.

        Major Diameter? - Micrometer.


        Watch closely as the CMM probes the thread, make certain the pitch is incrementing in the correct way, (you may need a negative pitch value or you may need to flip the ClockWise/CountgerClockWise option).



        sigpic"Hated by Many, Loved by Few" _ A.B. - Stone brewery

        Comment


        • #5
          defective is location........ surface plate layout???? As in indicator, vise.....on surface plate? Currently this pin has two cylinders of same size seperated by a collar 0.1 larger. The second cylinder is held 0.002 true position to the first as is the threads which are further away from datum as well. Part is made from inconel. I get on the second cylinder a out of tolerance condition of about 0.0016. On the threads I am getting a oot condition of about 0.0067. The sales guys idea is to put the part back into the cnc it was made with and run a indicator on it while in the machine and if it gives a near zero result then negate /ignore the cmm results. the circle is spinning in correct orientation and the pitch seems to be spacing(incrementing ) correctly.

          Comment


          • #6
            Originally posted by atommonkeyz View Post
            defective is location........ surface plate layout???? As in indicator, vise.....on surface plate? Currently this pin has two cylinders of same size seperated by a collar 0.1 larger. The second cylinder is held 0.002 true position to the first as is the threads which are further away from datum as well. Part is made from inconel. I get on the second cylinder a out of tolerance condition of about 0.0016. On the threads I am getting a oot condition of about 0.0067. The sales guys idea is to put the part back into the cnc it was made with and run a indicator on it while in the machine and if it gives a near zero result then negate /ignore the cmm results. the circle is spinning in correct orientation and the pitch seems to be spacing(incrementing ) correctly.

            Yes surface plate. Vise/v-block, height gage, dial test indicator, digital height gage, etc. Old School. Lacking the ability to do that, how do you ever double check the CMM results?

            By all means, let them put it back in the CNC and check it with an indicator, make sure you right there watching, it sounds like you might learn some things about manual inspection techniques, or you might spot a flaw in their methods. If the CMM is correct, they should show the same deviation, or very close when checking it on the machine. Although there is a very slight possibility the machine is not accurate/worn out/mis-aligned/etc. in such a way that checking the part in the machine does not reveal the error the machine cut into the part.
            sigpic"Hated by Many, Loved by Few" _ A.B. - Stone brewery

            Comment


            • #7
              You could try a series of points in a line up the side of the stud using paste with pattern with .010-.015" increments and paste that around the stud 4-6 places again with paste with pattern. Then construct a cylinder using the minimum circumscribed algorithm. That should get a fairly accurate size and location. It will take some time to measure, but it should be as accurate as you'll get on a CMM.
              PC-DMIS 2016.0 SP8

              Jeff

              Comment


              • #8
                You could try measuring the TP using the comparator.
                B&S One
                PC-DMIS CAD v2014

                Romer Infinity

                Comment


                • #9
                  Your measurements seem to be (if I'm reading correctly) getting worse and worse the further from the datum you go.

                  Does the cmm give you the same answer over and over with reloading the part in between runs? If not, I'd guess your alignment isn't coming out square.

                  If the cmm repeats with reloads, you should be able to check it on the surface plate, you shouldn't even have to look that close with as far off as the cmm is reporting.

                  If in the machine, they chuck the datum and tap the part in on the end, be sure to indicate the center. If the part is bent, the middle will run out between centers. If the center is good, indicate the datum (as far from the chuck as possible and you're looking for tenths), if that runs out, the two ends are running out.

                  Comment

                  Related Topics

                  Collapse

                  Working...
                  X