Find Nominals from CAD

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

  • Find Nominals from CAD

    If I am measuring with CAD and I do NOT have this on, are the results screwed? I measured the same part with it on and with it off, different results. Is this new with 4.0 on up? I don't remember seeing this button. To be fair, I've never used a lot of CAD before, either.


    TIA

  • #2
    Originally posted by John Kugler View Post
    If I am measuring with CAD and I do NOT have this on, are the results screwed? I measured the same part with it on and with it off, different results. Is this new with 4.0 on up? I don't remember seeing this button. To be fair, I've never used a lot of CAD before, either.


    TIA
    Unless they have completly bolluxed it up in that version, you do NOT need FIND NOMS marked.

    IF I under stand it correctly, there are 2 'find noms' in Pcdmis and they are used in different ways:

    FIND NOMS will make Pcdmis 'look' for nominal cad data for a manual touch, within a specified tolerance limit, ie no cad within 5mm, no nominals found.

    FIND NOMS ON EXECUTE will make Pcdmis re-pierce the cad model EVERY TIME you run the program, picking NEW nominals for each and every point. I would NOT use this ONLINE in any way, since NO machine positions perfectly, this will make all your points 'migrate' accross the cad model AND makes it a real pain to do any AXIAL SPC since the nominals will change for each part.
    sigpic
    Originally posted by AndersI
    I've got one from September 2006 (bug ticket) which has finally been fixed in 2013.

    Comment


    • #3
      There are actually 3. Find Noms and Find Noms on Execute in the F5 Setup.

      There is a Find Nominal from CAD mode button on the Probe Mode toolbar.

      Any ideas what the last one is? That is the one I'm playing with. I have the Find Noms on, but not on execute from F5.


      Thanks again.

      Comment


      • #4
        Used to use find nominals during execution with faro arm. If we reran program would go back and uncheck box in settings then F9 all features and change to nominals and home points other wise all points including those in alignments would constantly change nominals that was 3.5 don't know about 4 or above.
        sigpic
        3.7 MR4

        Comment


        • #5
          I just got off the phone with tech support. We were trying to fix an issue I am having and she wanted to know what boxes I had checked under F5. She said I SHOULD always have Find Nominals checked. Which it wasn't. Didn't really tell me why. Does anyone else use this?
          ~Josh~ Giles Rescue "Ut ceteri vivant"

          Comment


          • #6
            Originally posted by jkb_mcg View Post
            I just got off the phone with tech support. We were trying to fix an issue I am having and she wanted to know what boxes I had checked under F5. She said I SHOULD always have Find Nominals checked. Which it wasn't. Didn't really tell me why. Does anyone else use this?
            Here is THEIR reasoning behind this (this is a guess)

            IF you have FIND NOMS checked, when Pcdmis actually touches the part, it will not touch it exactly at the nominal location since ALL machines have SOME drift in them. By 'finding noms', it will 'look' for the piece of cad closest to the touch, grab the XYZIJK and calculate the actual point from there, thus eliminating the error from the drift of the machine.

            HOWEVER, this WILL cause your points to slowly move with time. EACH time you run the program, it will find slightly new nominals, which then become the new nominals for the next part. So, your X nominal can change with each part you check, same with Y and Z nominals. Do you want continually changing nominals? I DON'T!

            SNAP point will remove the drift from the results be projecting the actual touch point ONTO the nominal vector line of the point. THE ONLY thing this effects is the XYZ deviation values. It will NOT effect the T deviation value due to the way it is calculated. T deviation (Total vector deviation) is calculated by SQRT((Xdev*I)^2+(Ydev*J)^2+(Zdev*K)^2). This will work for SNAP or non-SNAP points T value. However, SNAP points will give you an easier formula (since the point is ON the nominal vector line) of SQRT(Xdev^2+Ydev^2+Zdev^2). However, SNAP or no SNAP, the T will be the same.

            Also, how trusting are you of the people at tech? Remember, these are the same people working on making the updates.
            sigpic
            Originally posted by AndersI
            I've got one from September 2006 (bug ticket) which has finally been fixed in 2013.

            Comment


            • #7
              Dammit, Matt... You should write a PC-DMIS for Dummies! Or at least a 'Hoedemans Help' file.

              ...bet the techs would have humongous use for it. Me too, by the way.
              PC-DMIS CAD++ 2o22.1 SP4

              Comment


              • #8
                Originally posted by vpt.se View Post
                Dammit, Matt... You should write a PC-DMIS for Dummies! Or at least a 'Hoedemans Help' file.

                ...bet the techs would have humongous use for it. Me too, by the way.
                I have written one, but the folks at B&S won't let it be on their website.

                Check here, sort by Pcdmis Software:

                http://www.*******.com/forum/forumdisplay.php?f=26

                The sort function is a drop-down at the bottom of the page.
                sigpic
                Originally posted by AndersI
                I've got one from September 2006 (bug ticket) which has finally been fixed in 2013.

                Comment


                • #9
                  Thanks a bunch!

                  The 500th post!
                  PC-DMIS CAD++ 2o22.1 SP4

                  Comment


                  • #10
                    Originally posted by vpt.se View Post
                    Thanks a bunch!

                    The 500th post!
                    what a puppy! happy about hitting 500. That is only 10% of MY level! Buuuuaaaaaaaahhhhhhhhhhhhhhhh
                    sigpic
                    Originally posted by AndersI
                    I've got one from September 2006 (bug ticket) which has finally been fixed in 2013.

                    Comment


                    • #11
                      Yeah, I know... Now I'm sad again.
                      PC-DMIS CAD++ 2o22.1 SP4

                      Comment


                      • #12
                        Originally posted by Matthew D. Hoedeman View Post
                        Here is THEIR reasoning behind this (this is a guess)

                        IF you have FIND NOMS checked, when Pcdmis actually touches the part, it will not touch it exactly at the nominal location since ALL machines have SOME drift in them. By 'finding noms', it will 'look' for the piece of cad closest to the touch, grab the XYZIJK and calculate the actual point from there, thus eliminating the error from the drift of the machine.

                        HOWEVER, this WILL cause your points to slowly move with time. EACH time you run the program, it will find slightly new nominals, which then become the new nominals for the next part. So, your X nominal can change with each part you check, same with Y and Z nominals. Do you want continually changing nominals? I DON'T!

                        SNAP point will remove the drift from the results be projecting the actual touch point ONTO the nominal vector line of the point. THE ONLY thing this effects is the XYZ deviation values. It will NOT effect the T deviation value due to the way it is calculated. T deviation (Total vector deviation) is calculated by SQRT((Xdev*I)^2+(Ydev*J)^2+(Zdev*K)^2). This will work for SNAP or non-SNAP points T value. However, SNAP points will give you an easier formula (since the point is ON the nominal vector line) of SQRT(Xdev^2+Ydev^2+Zdev^2). However, SNAP or no SNAP, the T will be the same.

                        Also, how trusting are you of the people at tech? Remember, these are the same people working on making the updates.


                        No, I don't want my NOMS to be changing all the time. The SNAP Point doesn't apply to edge points though does it?
                        ~Josh~ Giles Rescue "Ut ceteri vivant"

                        Comment


                        • #13
                          Originally posted by jkb_mcg View Post
                          No, I don't want my NOMS to be changing all the time. The SNAP Point doesn't apply to edge points though does it?
                          NO, however, there is a function that SORT of acts like a snap point on edge points. IF you are taking a surface sample (and you SHOULD to control your depth) and you are NOT reporting the S axis (S-surface deviation of edge point with a surface sample hit), you can use GAP ONLY which will ELIMINATE the surface deviation from the edge point result. I have not used it much, since I always (99.99% of the time) report the surface deviation and thus eliminate a point, I mean, sheet metal more or less REQUIRES a surface sample hit so I get TWO checks with one point, a surface point AND an edge point, I report XYZTS. So, since I have NOT used GAP ONLY much (it is part of the DIMENSION ~ LOCATION) I do not know if that causes it to 'snap' onto the nominal vector like a vector point snap does or if it only removes the surface deviation. I would THINK that it would snap it onto teh nominal vector line. You can NOT use GAP ONLY and report the S axis for an edge point, the S will always report ZERO since the GAP ONLY removes the surface variation.
                          sigpic
                          Originally posted by AndersI
                          I've got one from September 2006 (bug ticket) which has finally been fixed in 2013.

                          Comment

                          Related Topics

                          Collapse

                          Working...
                          X