Edge Point

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

  • Edge Point

    OK, I know I am going to be in for a lot of ribbing on this one, but here goes:

    I have a part (no fixture yet) that I have to do an iterative alignment on. It is basically a flat plate with a 45 degree bend in it that is also at 45 degrees to the datum edge (compound angle) The trim edge on one side of the compound angled surface is one of my datums, the top flat being A, a slot being B, then this trim edge. The problem is that the trim edge is NOT 90 degrees to the surface, it is not trimmed at 90 degrees to the surface, but at just about 45 degrees so that the edge will be 'square' to the top flat. So, I have a bend in a part that can (and WILL and DOES) vary and has profile tolerance as well, but I have to make a touch on the trim edge to align this part. If I use EDGE POINT from either surface ('top', 'bottom', or the 'edge'), Pcdmis automatically make the edge vector 90 degrees to teh surface I just picked. I CAN get the actual vector from the edge surface, but I can not measure it as a vector/surface point because of the part deviation/variation.

    OK, hopefully, you understand the situation, so here is my question:
    1) Can I use EDGE POINT, picking on the 'top' surface, then manually enter the vector for the edge into place for the edge point and have it work correctly?

    Basically, I absolutely have to ahve a surface sample to control the 'depth' touch position, but the surface and edge vectors are not 90 degrees to each other.

    Imaging THIS |\ is how the part looks, I have to touch on the lefts side and control the touch on the angled right side. Will changing the vector in the edge point screw it up and give me bad data. I also have to check several points on this edge and I will pretty much have to use the same method for each point, but the method will HAVE to work.
    sigpic
    Originally posted by AndersI
    I've got one from September 2006 (bug ticket) which has finally been fixed in 2013.

  • #2
    Hello Matt,
    If I understand it right, the angled surface is a datum edge. I know it is a little more work, but two lines on the flat and on the angle, with constructed points simulating the edge will be more accurate and represent your actual part better. Just my .02 cents.

    Have a Great Day!
    A.Gore
    sigpicA.Gore

    Comment


    • #3
      I had some similar points (not quite 45 deg though). I tried manipulating with the edge points to get accurate points. I eventually gave up and did it like ag162 suggests. Construct intersection lines and do temporary alignments. You can get right where you want to be with that.
      Bill
      Bill Jarrells
      A lie can travel half way around the world while the truth is putting on its shoes. - Mark Twain

      Comment


      • #4
        Originally posted by ag162 View Post
        Hello Matt,
        If I understand it right, the angled surface is a datum edge. I know it is a little more work, but two lines on the flat and on the angle, with constructed points simulating the edge will be more accurate and represent your actual part better. Just my .02 cents.

        Have a Great Day!
        A.Gore

        Originally posted by Wingman View Post
        I had some similar points (not quite 45 deg though). I tried manipulating with the edge points to get accurate points. I eventually gave up and did it like ag162 suggests. Construct intersection lines and do temporary alignments. You can get right where you want to be with that.
        Bill
        Well, the surface itself does move, which complicated matters and the edge surface, due to sheet-metal trimming only have about 0.3 to 0.5mm of 'shear' edge, the rest is breakage. The problem is getting the probe to touch on the shear and not scoot over the surface or hit in the breakage.

        I have been playing with it and have discovered that changing the EDGE vector to mathc the actual vector of the edge does NOT work as Pcdmis uses the SURFACE vector for most of it's comp calculations. Changing the surface vector of the edge point will give a bad surface deviation, but I am not interested in the surface, just getting good touches on the edge and the results match using surface points on the surface and surface points on the edge which are measured in relation to the surface points on the surface, if ya follow. If the edge shear was larger, maybe I could use the 2-line method, but there are addition issues (still trying to get them resolved with the customer) due to the design and requirement that this trim be cut at a 45 degree angle to the surface and that is that it is curling the edge of the part, thus making it even more complicated.

        BUT, while waiting for help, I did discover that you can NOT get a good result for the TRIM deviation by using the actual surface vector and actual edge vector for an edge that is not 90 degrees to the surface.

        It does appear that if you use the actual vector for the trim edge for the edge vector and a vector 90 degrees from that for the surface vector that the edge will come out right, but of course the surface is wrong, but since I am looking for the edge and not surface and am only using the surface to control the edge touch, I think this is the way I will have to go. If anyone has any other ideas, I'll be glad to hear them.
        sigpic
        Originally posted by AndersI
        I've got one from September 2006 (bug ticket) which has finally been fixed in 2013.

        Comment


        • #5
          Hhhhmmmm,
          Ok, do you have a check fixture in which you can secure the part and align off of the fixture? If not, then the part needs to be corrected. If the datums are out, due to the edge rolling then regardless of the rest of the geometry, the part is bad. Sorry I can't be more help.

          A.Gore
          sigpicA.Gore

          Comment


          • #6
            Originally posted by ag162 View Post
            Hhhhmmmm,
            Ok, do you have a check fixture in which you can secure the part and align off of the fixture? If not, then the part needs to be corrected. If the datums are out, due to the edge rolling then regardless of the rest of the geometry, the part is bad. Sorry I can't be more help.

            A.Gore
            No fixture (yet), they SHOULD be coming in (should have been here by now). Even with the surface getting rolled due to the trimming condition, the trim is still 'right' and where it needs to be. As for correcting the part, there really isn't much that can be done about it. The trim angle (45 degrees to the surface) is dictated by the customer and we have complained about it since they forced the design on us. The surface of the part does absolutely nothing, the edge of the part, on both the 45 degree surface the the flat surface sit flat on a mating part, sort of like taking your hand, bending your fingers to a 45, then placing the edge of your hand and fingers on the desk. They have sent a picture of the assembly and nothing, absolutely nothing, touches the 45 degree surface, just the edge of it, where the crappy trim condition is, it is more like a kick-stand on a bike, there to support the flat surface and brace it up.
            sigpic
            Originally posted by AndersI
            I've got one from September 2006 (bug ticket) which has finally been fixed in 2013.

            Comment


            • #7
              Can you make a temporary fixture by using parallels or something of that nature (I do not know how big your part is)?
              sigpic
              Xcel 15-20-10 - PFXcel 7-6-5 - Merlin 11-11-7 - Romer Absolute 7525SI
              PCDMIS 2012
              Windows Office XP

              Comment


              • #8
                I hate it when I don't have an answer!!! ARGH!!! ok,,,do you have cad? Can you do what I call a reverse alignment,,,using other points to get you in the ball park,,,so that you can go back and pick the datums up and do a correct alignment? It breaks the rules a little, aligning off of points that are not datums,,,but it allows you to do a proper alignment a little easier in some cases.

                A.Gore
                sigpicA.Gore

                Comment


                • #9
                  Originally posted by rangerboat72 View Post
                  Can you make a temporary fixture by using parallels or something of that nature (I do not know how big your part is)?
                  No, not without a lot of machining, and compound drilling/reaming in the mix. I looked at it, tried to figure out something, someway, but it looks like any temp fixture would be just as complicated as a finished fixture.
                  sigpic
                  Originally posted by AndersI
                  I've got one from September 2006 (bug ticket) which has finally been fixed in 2013.

                  Comment


                  • #10
                    Originally posted by ag162 View Post
                    I hate it when I don't have an answer!!! ARGH!!! ok,,,do you have cad? Can you do what I call a reverse alignment,,,using other points to get you in the ball park,,,so that you can go back and pick the datums up and do a correct alignment? It breaks the rules a little, aligning off of points that are not datums,,,but it allows you to do a proper alignment a little easier in some cases.

                    A.Gore
                    Well, that's pretty much what I was doing by using a surface point on the surface, close to the edge, then doing another surface point on the edge, and using RELATIVE measure to tie it to the point on the surface, basically, an edge point 'the hard way'. This gave me a hit on the edge, and controlled the depth from the surface so that it would contact the shear everytime, but I didn't want to do all that screwing around for all the edge points I have to check on that surface. However, after playing with it (for 3 hours so far) I have found that if I:

                    1) Use edge point and pick the edge data in the cad file
                    then
                    2) Put the surface vector in the edge vector place and the edge vector in the surface vector place it will give me the same result as using the 2 surface points with relative measure.

                    This method will give you the same values/nominals/results IF you are working on a square edge. If you pick the surface for the edge or the edge surface, the vectors are just swapped from edge to surface and vise-versa. So, in this case, it gives me the correct probe comp for the edge touch but the wrong comp for the surface touch, but it controls the edge touch and keeps it on the shear. And, since the edge vector is 0.995 for the datum it is for, I think it will be close enough for now, at least until we can get the fixture. I told them from DAY 1 that I would ahve to have the fixture to check these parts but they still have not been able to get the fire lit under the customer who is having the fixtures made.
                    sigpic
                    Originally posted by AndersI
                    I've got one from September 2006 (bug ticket) which has finally been fixed in 2013.

                    Comment


                    • #11
                      It sounds like that is the only way for now,,,,,,good luck.

                      Have a Great Day!
                      A.Gore

                      Is it Miller Time yet????
                      sigpicA.Gore

                      Comment


                      • #12
                        Originally posted by ag162 View Post
                        It sounds like that is the only way for now,,,,,,good luck.

                        Have a Great Day!
                        A.Gore

                        Is it Miller Time yet????
                        Man, I know I am ready for a few frosty sensations right now.
                        sigpic
                        Originally posted by AndersI
                        I've got one from September 2006 (bug ticket) which has finally been fixed in 2013.

                        Comment


                        • #13
                          Originally posted by Matthew D. Hoedeman View Post
                          ...frosty sensations...
                          Sounds like something Craiger would be into.

                          Comment


                          • #14
                            Originally posted by Goodluck View Post
                            Sounds like something Craiger would be into.
                            Where do you think I got it from?!?!?!?
                            sigpic
                            Originally posted by AndersI
                            I've got one from September 2006 (bug ticket) which has finally been fixed in 2013.

                            Comment


                            • #15
                              Originally posted by Matthew D. Hoedeman View Post
                              Where do you think I got it from?!?!?!?
                              You got a frosty sensation from Craig?!

                              Comment

                              Related Topics

                              Collapse

                              Working...
                              X