Program question (alignment & constructed features)

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

  • Program question (alignment & constructed features)

    i have a question about a program i’m attempting to run. i’m not sure if the alignments are good because when the part is moved about on the CMM some of the features will not adjust with the location move.

    Other issue with this part is that the constructed features don’t appear to take affect once another part is ran. The program seems to run fine once i go through the program manually. Also, the hole measurement locations will change.

    Greatly would appreciate input on the alignment and why constructed features causing me problems.

    Thank you
    Attached Files
    Brandnew
    Senior Member
    Last edited by Brandnew; 10-12-2010, 09:03 PM.
    Brown & Sharpe / PC-DMIS 3.7
    Eph 2:8-9 / 1 Thes 5:21 - Who or What are you living for?

  • #2
    Alignment should be done in perhaps a more systematic way.
    What I do:
    Create a plane from points.
    Level to plane.
    Origin XYZ on plane.
    Create features for rotation. 2 pnts or a couple of circles.
    Rotate to line created by above features.
    Origin in 2 axes one or more of those features.

    This is done manually.
    Then I go into DCC and either repeat what I have already done, or create new features.
    Make new alignment from DCC features.

    Repeat the DCC stuff.
    Lately, it occurs to me
    What a long, strange trip it's been.

    2017 R1 (Offline programming)

    Comment


    • #3
      Thanks John, i'll try this tomorrow. i'll find out if this solves my constructed feature dilemma as well.
      Brown & Sharpe / PC-DMIS 3.7
      Eph 2:8-9 / 1 Thes 5:21 - Who or What are you living for?

      Comment


      • #4
        Construct the features in the alignment you will be inspecting them in.
        sigpic GDTPS - 0584

        Comment


        • #5
          Hi again,

          So i did make the alignments based on manual points to create the plane and line to rotate to. However i find that afterwards the constructed features are still not adjusting to the next part. i placed an overview of the part and the two constructed points that are not adjusting (datum A and B). For these i measure a line on each side and construct a midline. Is this the incorrect method to origin to Datum A and B?
          i'm thinking i shouldn't use these in terms of the alignment for the parts to be measured, but rather just origin to these after the part is measured to determine the required measurements.
          Attached Files
          Brown & Sharpe / PC-DMIS 3.7
          Eph 2:8-9 / 1 Thes 5:21 - Who or What are you living for?

          Comment


          • #6
            Make sure the constructed features have the correct vectors. Also- I have constructed cylinders from inside circles that magicaly flip into outside cylinders when constructed... hmmm
            Global Performance
            5-7-5, PH10T, 4.2-1
            Micro Xcel 7-6-5
            PH6, 4.2-1

            Comment


            • #7
              Yeah, i did actually catch that the vectors for the manual alignment hits were reading off and thus i fixed that and the alignments would come up good.

              In regards to constructing in the alignments on how they will be inspected, the alignments don't change...its just that i use these constructed features as part of the alignment themselves.

              Again, i place the part and run the program (after closing it) and once again the constructed features aren't adjusted automatically. But after i F9 them again, things run peachy but i'm hesitant to think that measurements are accurate.

              i guess my main question with this, are my constructed features not working b/c of the alignment or is this something like when i create a line (not construct) i have to click the F3 to activate the line and so the same thing with the construct features?
              Brown & Sharpe / PC-DMIS 3.7
              Eph 2:8-9 / 1 Thes 5:21 - Who or What are you living for?

              Comment


              • #8
                Are the constructed features marked? (Blue)

                Comment


                • #9
                  Originally posted by Chally72 View Post
                  Are the constructed features marked? (Blue)
                  No, they come up in black text...unlike the other features which i can see it being blue and know they are not active.

                  Let me run this as an example to get your opinion:

                  Say the part program is created; alignments are made with the origin going to constructed features. Run the whole program, no problem. Place another part, run the program and then the constructed features don't adjust. i F9 them and see that the actual measurements are adjusted and so the alignment adjust as appropriate and run the program and it runs smoothly.

                  Are the values i get something i can say is reading accurately or not?

                  i don't have a set fixture for the part, but i figure that the alignments should adjust for this every time i place a new part (i'm not placing the part from one end of the cmm to the other)...or is this the problem i'm having?
                  Or (which i hope is not the case) do i have something wrong with the software now?

                  thanks
                  Brown & Sharpe / PC-DMIS 3.7
                  Eph 2:8-9 / 1 Thes 5:21 - Who or What are you living for?

                  Comment


                  • #10
                    Here's are some other questions; but they may or may not have any bearing on your issue.
                    • Are the constructed features constructed before or after a DCC alignment?
                    • How much "off" (non-adjustment when running next piece) are we taliking about? If the next part is placed an inch away from the previous one, are the constructed features an inch off, or only a few thousandths? Depending, that could be two different causes.
                    • Are any of the blue (marked) features used in any alignment?
                    I've said many times - most of the errors I see in my programs during testing come from whacked alignments.



                    Added:
                    I looked back at the code you posted (you may have since changed it) and I am looking at the last bit -

                    Code:
                    [LEFT]PNT_BK_A =AUTO/VECTOR POINT,SHOWALLPARAMS = NO
                    THEO/0,2.35,0.705,0,0,1
                    ACTL/0.0001,2.35,0.6857,0,0,1
                    TARG/0,2.35,0.705,0,0,1
                    PNT_BK_B =AUTO/VECTOR POINT,SHOWALLPARAMS = NO
                    THEO/0,2.24,0.705,0,0,1
                    ACTL/0.0001,2.24,0.6873,0,0,1
                    TARG/0,2.24,0.705,0,0,1
                    PNT_BK_C =AUTO/VECTOR POINT,SHOWALLPARAMS = NO
                    THEO/0,2.1,0.705,0,0,1
                    ACTL/0.0001,2.1,0.6884,0,0,1
                    TARG/0,2.1,0.705,0,0,1
                    PNT_FR_A =AUTO/VECTOR POINT,SHOWALLPARAMS = NO
                    THEO/0,-2.1,0.705,0,0,1
                    ACTL/0.0001,-2.0998,0.6957,0,0,1
                    TARG/0,-2.1,0.705,0,0,1
                    PNT_FR_B =AUTO/VECTOR POINT,SHOWALLPARAMS = NO
                    THEO/0,-2.24,0.705,0,0,1
                    ACTL/0.0001,-2.24,0.6952,0,0,1
                    TARG/0,-2.24,0.705,0,0,1
                    PNT_FR_C =AUTO/VECTOR POINT,SHOWALLPARAMS = NO
                    THEO/0,-2.35,0.705,0,0,1
                    ACTL/0.0001,-2.35,0.6935,0,0,1
                    TARG/0,-2.35,0.705,0,0,1
                    ASSIGN/V1 = (PNT_BK_A.Z+PNT_BK_B.Z+PNT_BK_C.Z)/3
                    ASSIGN/V2 = (PNT_FR_A.Z+PNT_FR_B.Z+PNT_FR_C.Z)/3
                    [COLOR=red][B]ASSIGN/V3[/B][/COLOR] = (V1+V2)/2
                    [COLOR=green][B]F1[/B][/COLOR] =GENERIC/POINT,DEPENDENT,RECT,$
                    NOM/XYZ,0,0,1,$
                    MEAS/XYZ,-0.0024,0.1943,[COLOR=red][B]V3[/B][/COLOR],$
                    NOM/IJK,0,0,1,$
                    MEAS/IJK,0,0.0001,[COLOR=red][B]V3[/B][/COLOR]
                    A6 =ALIGNMENT/START,RECALL:A5, LIST= YES
                    ALIGNMENT/LEVEL,ZPLUS,PLN_CIRCLE
                    ALIGNMENT/ROTATE,XPLUS,TO,LIN_LH_RH,ABOUT,ZPLUS
                    ALIGNMENT/TRANS,XAXIS,DATUM_B
                    ALIGNMENT/TRANS,YAXIS,DATUM_A
                    ALIGNMENT/TRANS,ZAXIS,[COLOR=green][B]F1[/B][/COLOR][/LEFT]
                    ALIGNMENT/END
                    According to the code:
                    The Z value in generic point F1 used in the A6 alignment is based off a variable (V3), and V3 is created via half of the sum of the averages of the Z hits of two sets of three each vector points, correct? YOWZER - there must be a reason for that . . . anyway -

                    This might be OK to get a point to use as a Z origin, but I don't believe it will work to use to same number also as the K vector in the generic point. Something just doesn't sound right about that . . .

                    Like I said, wacky alignments wreak all kinds of havoc.
                    ir a programmer
                    PC-DMIS demi-Guru
                    Last edited by ir a programmer; 10-15-2010, 01:56 PM. Reason: Added some stuff
                    sigpic

                    Comment


                    • #11
                      Your alignment doesn't have Datum -A- as your primary.

                      Your sketch looks like -A- should be a Midplane as should -B-.

                      If it is a forging and you don't have good surfaces to measure as planes, I'm not sure what you can do.
                      Sorry.

                      B&S CHAMELEON/PCDMIS CAD++ V2011

                      There are no bugs, only "UNDOCUMENTED ENHANCEMENTS!"

                      sigpic

                      Comment


                      • #12
                        [QUOTE=ir a programmer;304623]Here's are some other questions; but they may or may not have any bearing on your issue.
                        • Are the constructed features constructed before or after a DCC alignment?
                          Both - i did the program over where the plane, line and origins were manual or dcc and still same issue.
                        • How much "off" (non-adjustment when running next piece) are we taliking about?
                          - This actually might be the issue; i place the part "relatively" in the same location but then i would see the holes off by .050" or so. But this is where i thought the alignment would adjust.
                        • Are any of the blue (marked) features used in any alignment?
                          - No blue in the align, except when i construct midlines based on measured lines thats the only blue but this doesn't go into the alignment.


                        I've said many times - most of the errors I see in my programs during testing come from whacked alignments.
                        - One thing i realize, how much i don't know and need to learn...so i don't doubt that i have this.


                        According to the code:
                        YOWZER - there must be a reason for that . . . anyway -

                        i got rid of the generic and have an offset point based on the assigned variables. Reason for this, the drawing calls for .705BSC off the center of these 4 holes - but the .705BSC is off the forging of the part and only based on two of the forging areas out of 4. So i have 6 points taken (3 for one forging and 3 for another - average these out and based on two measurements i average these two as well. i do this just to try to get a relatively decent number off the forging to base my true position from.

                        thanks for the input
                        Brown & Sharpe / PC-DMIS 3.7
                        Eph 2:8-9 / 1 Thes 5:21 - Who or What are you living for?

                        Comment


                        • #13
                          Are you measuring your manual points with the "slow" button on? For some reason B&S told us we had to on one of our machines. Our numbers weren't correlating from machine to machine. I never would have thought that would affect the DCC alignment, but it did.

                          Comment


                          • #14
                            [QUOTE=Brandnew;304655]
                            Originally posted by ir a programmer View Post
                            Here's are some other questions; but they may or may not have any bearing on your issue.[LIST][*]Are the constructed features constructed before or after a DCC alignment?
                            Both - i did the program over where the plane, line and origins were manual or dcc and still same issue.[*]How much "off" (non-adjustment when running next piece) are we taliking about?
                            - This actually might be the issue; i place the part "relatively" in the same location but then i would see the holes off by .050" or so. But this is where i thought the alignment would adjust.[*]Are any of the blue (marked) features used in any alignment?
                            - No blue in the align, except when i construct midlines based on measured lines thats the only blue but this doesn't go into the alignment.[/LIST]

                            I've said many times - most of the errors I see in my programs during testing come from whacked alignments.
                            - One thing i realize, how much i don't know and need to learn...so i don't doubt that i have this.


                            According to the code:
                            YOWZER - there must be a reason for that . . . anyway -

                            i got rid of the generic and have an offset point based on the assigned variables. Reason for this, the drawing calls for .705BSC off the center of these 4 holes - but the .705BSC is off the forging of the part and only based on two of the forging areas out of 4. So i have 6 points taken (3 for one forging and 3 for another - average these out and based on two measurements i average these two as well. i do this just to try to get a relatively decent number off the forging to base my true position from.

                            thanks for the input
                            Aside from any alignment issues -

                            I see a RED FLAG

                            Anything blue is a marked (skipped) item and is NOT an active feature/measurement/dimension/whatever. If, later on in the code, the program is using an item that is or was constructed from anything blue, this will cause errors. The error depends on what it's used for. So what are the constructed midlines used for?

                            Generally, if I have things marked during testing (like testing two different methods to construct a feature), once I determine the one I want, and that the other will not be needed, I delete it, so that at some future point someone else does not unkowingly unmark all and turn the "other" back on.
                            sigpic

                            Comment


                            • #15
                              Originally posted by rfarr View Post
                              Are you measuring your manual points with the "slow" button on? For some reason B&S told us we had to on one of our machines. Our numbers weren't correlating from machine to machine. I never would have thought that would affect the DCC alignment, but it did.
                              Yeah the "slow" button is turned on...but i never actually paid attention when doing manual hits...i'll watch for this next time i run manual points...thanks
                              Brown & Sharpe / PC-DMIS 3.7
                              Eph 2:8-9 / 1 Thes 5:21 - Who or What are you living for?

                              Comment

                              Related Topics

                              Collapse

                              Working...
                              X