Manual Alignment

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

  • Manual Alignment

    hi, all . Im a beginner to cmm using v4.1 so i will probably be asking a lot of basic questions for a while.
    We check a lot of batches of parts mostly cyclindrical parts with milled profiles etc.
    we put each part on table and do a manual alignment and then run program to check part. Usually plane ,line , point alignment. Does in matter which direction we go when doing manual alignment for plane and line etc .Or when doing any feature . Can anyone give me any info on how pcdmis is reading features during manual alignment.

    many thanks in advance


  • #2
    Lines will require you to measure them in the same direction, every time, especially for alignment features. When you rotate, lets say, the XPLUS axis to a line, then the XPLUS axis will point in the same direction as the line does, and the line will point frmo the first hit to the last hit. So, if you measure it left-to-right the first time, the right-to-left the second time, it will completly change the alignment by 180 degrees.
    sigpic
    Originally posted by AndersI
    I've got one from September 2006 (bug ticket) which has finally been fixed in 2013.

    Comment


    • #3
      Like Matt said you can get hosed when a guy measures in the wrong direction on a line. My favorite for manual alignments is simply a readpoint then go DCC and find the part. If you must/want to do manual alignment with features then you can monkey proof it:

      Code:
                  COMMENT/OPER,"WHEN MEASURING THE FACE HITS MUST BE TAKEN FROM LEFT TO RIGHT"
      REDO_LN1   =LABEL/
      LN1        =FEAT/LINE,RECT,UNBND
                  THEO/-2,-2.9,-13,1,0,0
                  ACTL/-1.7177,-2.8972,-13.0506,0.9999975,-0.0022291,0
                  MEAS/LINE,3,WORKPLANE
                  HIT/BASIC,-2,-2.9,-13,0,-1,0,-2.0002,-2.9,-13
                  HIT/BASIC,-1,-2.9,-13,0,-1,0,-1.0002,-2.9,-13
                  HIT/BASIC,0,-2.9,-13,0,-1,0,-0.0002,-2.9,-13
                  ENDMEAS/
                  IF/LN1.I < 0
                  COMMENT/OPER,"YOU MUST MEASURE THE FACE FROM LEFT TO RIGHT"
                  GOTO/REDO_LN1
                  END_IF/
      <internet bumper sticker goes here>

      Comment


      • #4
        Like Kegger, er Craiger says, put some standards into place. For manual alignment (here) we always measure in a specific order:

        LEFT TO RIGHT
        FRONT TO BACK

        Always, then name the features something that will tell you WHAT to measure

        REAR_EDGE
        FRONT_EDGE
        LEFT_EDGE
        RIGHT_EDGE

        So, if you are measure the RIGHT_EDGE, you know to measure it FRONT TO BACK. FRONT_EDGE (or REAR_EDGE) would be LEFT TO RIGHT.

        It does not make any difference for a plane which or where or what order you measure the hits. R(ight)R(ear), R(ight)F(ront), L(eft)R(ear), R(ight)R(ear) or LF, LR, RF, RR or RR, LR, LF, RF, it makes no difference. The lines DO make a difference.
        sigpic
        Originally posted by AndersI
        I've got one from September 2006 (bug ticket) which has finally been fixed in 2013.

        Comment


        • #5
          Like craig said. Readpoint & then straight into dcc. MOST can place the probe were you tell them.
          sigpic.....Its called golf because all the other 4 letter words were taken

          Comment


          • #6
            Originally posted by craiger_ny View Post
            Like Matt said you can get hosed when a guy measures in the wrong direction on a line. My favorite for manual alignments is simply a readpoint then go DCC and find the part. If you must/want to do manual alignment with features then you can monkey proof it:

            Code:
                        COMMENT/OPER,"WHEN MEASURING THE FACE HITS MUST BE TAKEN FROM LEFT TO RIGHT"
            REDO_LN1   =LABEL/
            LN1        =FEAT/LINE,RECT,UNBND
                        THEO/-2,-2.9,-13,1,0,0
                        ACTL/-1.7177,-2.8972,-13.0506,0.9999975,-0.0022291,0
                        MEAS/LINE,3,WORKPLANE
                        HIT/BASIC,-2,-2.9,-13,0,-1,0,-2.0002,-2.9,-13
                        HIT/BASIC,-1,-2.9,-13,0,-1,0,-1.0002,-2.9,-13
                        HIT/BASIC,0,-2.9,-13,0,-1,0,-0.0002,-2.9,-13
                        ENDMEAS/
                        IF/LN1.I < 0
                        COMMENT/OPER,"YOU MUST MEASURE THE FACE FROM LEFT TO RIGHT"
                        GOTO/REDO_LN1
                        END_IF/
            I like the goto/redo_ln1 that's a good safety feature.
            sigpic

            Comment


            • #7
              Originally posted by bw_bob View Post
              Like craig said. Readpoint & then straight into dcc. MOST can place the probe were you tell them.


              Ive wanted to put this into programs for a while now. If a part moves on the cmm bed (i.e. it isnt on a permenant fixture) what would the code look like if, for example, I wanted the probe manually moved above a datum hole; then
              kick straight into a DCC alignment?

              I hope this makes sense. TIA.





              F
              Recently jumped from 3.5 Mr 2 CAD
              to 2012 CAD++

              Comment


              • #8
                Originally posted by Farmer View Post
                Ive wanted to put this into programs for a while now. If a part moves on the cmm bed (i.e. it isnt on a permenant fixture) what would the code look like if, for example, I wanted the probe manually moved above a datum hole; then
                kick straight into a DCC alignment?

                I hope this makes sense. TIA.





                F
                Ctrl + G is your shortcut for a readpoint. It basically reads the position of the ruby and creates a point there. Then you X, Y, Z origin on it and go from there.


                Best bet is to prompt the operator to position the probe over a corner of a part and using a wide open prehit/retract find the edges. Every time you find a feature do an alignment. Or you could prompt to position the ruby in a hole. Make sure the initial feature (the hole he places it in) is a the smaller of your features. You want the features that you are searching for to be big and forgiving.
                <internet bumper sticker goes here>

                Comment


                • #9
                  If you don't want to do the readpoint and then straight to DCC as suggested and you are worried about lines being measured in the wrong direction, eliminate the lines from your alignment. Three planes work just as well as plane, line, line. It just takes more manual hits.

                  I typically insert a digital picture of the part setup. Then all I need is to set the x, y and z orgin. I usually use the highest point on the part and an internal circle (holes are easier to measure than studs). As long as the part is set on the table facing the correct direction and relatively square, all runs fine.

                  To square parts to the table, I have a couple of studs threaded into the holes on the table. I have 2 foot steel rule that I run between the studs and just bump parts up against it before I clamp them. Not perfect but close enough.
                  Last edited by #2#; 02-26-2007, 01:15 PM.

                  Comment


                  • #10
                    COMMENT/OPER,"WHEN MEASURING THE FACE HITS MUST BE TAKEN FROM LEFT TO RIGHT"
                    REDO_LN1 =LABEL/
                    LN1 =FEAT/LINE,RECT,UNBND
                    THEO/-2,-2.9,-13,1,0,0
                    ACTL/-1.7177,-2.8972,-13.0506,0.9999975,-0.0022291,0
                    MEAS/LINE,3,WORKPLANE
                    HIT/BASIC,-2,-2.9,-13,0,-1,0,-2.0002,-2.9,-13
                    HIT/BASIC,-1,-2.9,-13,0,-1,0,-1.0002,-2.9,-13
                    HIT/BASIC,0,-2.9,-13,0,-1,0,-0.0002,-2.9,-13
                    ENDMEAS/
                    IF/LN1.I < 0
                    COMMENT/OPER,"YOU MUST MEASURE THE FACE FROM LEFT TO RIGHT"
                    GOTO/REDO_LN1
                    END_IF/
                    very nice Craig. I like that. points to you
                    Dale
                    v2011 MR1 CAD ++

                    Comment


                    • #11
                      Thanks guys , that has cleared up a few nagging doubts that i have been having. I will try to be more systematic from now on.

                      Comment


                      • #12
                        Cheers Guys, this is exactly what I need. I can now create some generic programs. This will save me lots of time.
                        Recently jumped from 3.5 Mr 2 CAD
                        to 2012 CAD++

                        Comment

                        Related Topics

                        Collapse

                        Working...
                        X