Composite Dimension Seperation Help Needed

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

  • Composite Dimension Seperation Help Needed

    Hello all,

    We have a flatness callout which also has unit flatness.

    https://www.pcdmisforum.com/image/pn...AASUVORK5CYII=

    This creates duplicate labels in QC-Calc database named "FCFFLAT1DTMA_SCN_FCFFLAT1 M" for both flatness. Everything runs good until we run a short version of the program and as we do, it creates a new QCC file because of duplicate feature error. Prolink said they won't be able to help in this as the labels have to be unique.
    So I am trying to report this dimension into two separate axis with Generic Points
    "FCFFLAT1.MEAS" assigned to 'X' value of generic point gives overall flatness value. How to get the MEAS value from second row into the 'Y' value of the generic points?

    F1 =GENERIC/POINT,DEPENDENT,CARTESIAN,$
    NOM/XYZ,<0,0,0>,$
    MEAS/XYZ,<FCFFLAT1.MEAS,FCFFLAT1.MEAS,-7>,

    Any help will be appreciated.

  • #2
    Isn't the new QCC file created from the different number of features you are inspecting from your "shorter" program ? You are talking QC-CALC here right ? A little confused how this is only a problem on your "shorter" program. Are you running from a marked sets program or an actual separate program ? If the later is the case you just need to change the output name. Just throwing out ideas to keep it simple.. Report plain old FLATNESS with Legacy dimensioning ?

    Comment


    • sim4191
      sim4191 commented
      Editing a comment
      Thanks for your reply Schlag.

      QC-Calc has a setting which let's it adapt to the feature labels being sent to it. So, in case of less number of dimensions it matches the incoming labels to existing labels and reports it (empty dimension points are shown with golden dot). So the two duplicate labels already existent in the QCC file confuse it and it ends up creating a new file with same duplicate labels again.
      We are not using marked sets, instead using loop commands (if-goto) to jumps over couple of features.
      Part print calls out for 'UNIT FLATNESS' as a major feature and hence I cannot use legacy dimensioning.

    • Schlag
      Schlag commented
      Editing a comment
      I hear you. The legacy Flatness would have not covered the per unit area and only been a work around it would have been in addition to. Can you dimension them after the loop ? It will be added code but then you would be re-creating the dimension for each part and they would be named different.

    • sim4191
      sim4191 commented
      Editing a comment
      These dimensions are actually not part of the loop. It's just the way QC-Calc is set up and there is no way around that.
      It's like if we keep sending the data in same number of features and order, everything is fine. The moment QC-Calc gets something different, it gets confused because of identical labels.
      Now from my above description it looks like a QC-Calc is the issue. However if PC-DMIS had an option to separate out dimensions, this could be avoided.

      Just looking for a way around and I really appreciate your replies.

  • #3
    QC-CALC does this for all Xact Measure dimensions. Why is this one the only issue ? This is one of the reasons I avoided xact for quite some time. It always out puts the 1st dim and then the 2nd has the "M1" behind it ? But this is only an issue on the shortened program ? I am truly lost and have nothing.

    Comment


    • sim4191
      sim4191 commented
      Editing a comment
      I totally understand and it's pain because it's only happening in two programs which have composite dimensions (we have other 10 programs running fine with short programs logic too).
      It puts the names for both Overall flatness and Unit flatness as "FCFFLAT1PLN_OCSFCFFLAT1 M" and hence gives the duplicate label error. Wouldn't be an issue if it put "M1" for 2nd dimension. Also, I wouldn't say it's an issue because of short program but rather because of different number of dimensions.
      We do run short programs on other operations of the part and they show dimensions in QC-Calc by adapting to it and showing 'golden dot' for features with no output to QC-Calc.

    • Schlag
      Schlag commented
      Editing a comment
      I don't have a lot of "short" or "inprocess" programs. The ones I do I just delete features and save them as separate programs just for that reason of the qc-calc file output. I'm sure this will get alot of cringes from some programmers but it works and I never took the time to come up with I would totally agree are much better options.

    • sim4191
      sim4191 commented
      Editing a comment
      I totally understand what you mean and I actually thought of doing that too. But they way we are set-up and actually have lot of programs, it kinda doesn't make sense for us to have separate programs and also we need to have control over how often operators run short programs.
      I do appreciate you taking the time to help me figure out this issue.

  • #4
    I'm curious what file you're sending out of pc-dmis? And what version of pc-dmis? I'm using the xml format and was pleased when I didn't have problems on a composite per unit line profile tolerance. "Per unit" got added for the lower tier. We use qc-calc also.

    Comment


    • sim4191
      sim4191 commented
      Editing a comment
      Thanks for commenting DJAMS.
      We use PC-DMIS 2015.0 but I am not sure what do you mean by the file which we send out?
      I thought PC-DMIS outputs to QC-Calc in .TMP extension? isn't that so?
      If not, how can I change that?

    • DJAMS
      DJAMS commented
      Editing a comment
      Sorry for being so brief. See below for details. QC-Calc supports multiple output types from pc-dmis. (at least 2).

  • #5
    Here's more details. We're running 2014 in production. If you've got this or greater, it looks like switching your output file to the Datapage+ XML format will solve your problem. You'll have to reconfigure your "machine" in QC-Calc as well.

    I mentioned that this was handled OK for pre unit line profile, and I just created a quick test for per unit flatness and it's handled the same way. For the lower tier, the feature name is replaced with "PER UNIT". So as long as you've got QC-Calc configured to include both dimension and feature name in the QC-calc ID's, this takes care of the problem.

    Here's what the xml output looks like:
    Code:
    <?xml version="1.0"?>
    <!--PC-DMIS DataPage+ Data Output File-->
    <ExecutionTransaction DateTime="131655194525092008">
      <PartData PartName="test" RevisionNumber="" SerialNumber="" Units="INCH" StatsCount="1" MinusTolerancesShowNegative="False" DataSource="" />
      <TraceFieldCmd Name="PATHSWITCH" Value="Z:\QC-Calc_Database\1234567" />
      <TraceFieldCmd Name="DRAWINGNUMBER" Value="1234567-0" />
      <TraceFieldCmd Name="DRAWINGREVISION" Value="f" />
      <TraceFieldCmd Name="PROGRAMNAME" Value="1-QPCS-1234567-0" />
      <TraceFieldCmd Name="PROGRAMREVDATE" Value="01/01/1701" />
      <TraceFieldCmd Name="SERIALNUMBER" Value="n/a" />
      <TraceFieldCmd Name="WORKCENTER" Value="0" />
      <TraceFieldCmd Name="CMM" Value="0" />
      <TraceFieldCmd Name="WORKORDER" Value="0" />
      <TraceFieldCmd Name="OPERATOR" Value="0" />
      <CommentCmd CommentType="DOCUMENTATION" Comment=" " />
      <DimensionCmd OutputMode="BOTH" DimensionID="FCFFLAT2" Type="184" GDT_SYMBOL="FLATNESS" FeatureID="SCN1" Axis="M" Nominal="0" Measured="0" PlusTol="0.005" Max="0" Min="0" />
      <DimensionCmd OutputMode="BOTH" DimensionID="FCFFLAT2" Type="184" GDT_SYMBOL="FLATNESS" FeatureID="PER UNIT" Axis="M" Nominal="0" Measured="0" PlusTol="0.001" Max="0" Min="0" />
    </ExecutionTransaction>
    Stats command looks like this:
    Code:
    STATS/ON,DATAPAGE+,,OFFLINE=YES
    RECORD=DIMENSION,TRACE,COMMENT,FCF PROFILE,

    Comment


    • sim4191
      sim4191 commented
      Editing a comment
      Thanks again for your reply DJAMS.
      So this is how our programs look like:
      *Towards the start of the program*
      STATS/ON,DATAPAGE,$
      DIRECTORY=,$
      READ=10,WRITE=20,MEMPAGES=64,DIMENSION NAME,CONTROLCALC OFF,$
      STATS/END
      TRACEFIELD/NO_DISPLAY,LIMIT=15 ; Operator : OP
      TRACEFIELD/NO_DISPLAY,LIMIT=15 ; Serial Number : SN
      TRACEFIELD/NO_DISPLAY,LIMIT=15 ; Machine : MC
      TRACEFIELD/NO_DISPLAY,LIMIT=15 ; Fixture : FX
      TRACEFIELD/NO_DISPLAY,LIMIT=15 ; CMM Number : CN
      TRACEFIELD/NO_DISPLAY,LIMIT=15 ; Pallet Number : PN

      *Right before end*

      STATS/TRANSFER,DIRECTORY=C:\SPCDATA

      Now, one difference which I saw was that during STATS/ON command "DATAPAGE" is selected but during STATS/TRANSFER command it shows "DATAPAGE+" selected after pressing F9. Could this be an issue? (P.S. We don't have any other issue with regard to dimensions).

      Also, another thing which I checked was that QC-Calc only allows 32 characters for feature label and since name of feature in QC-Calc is this "LOC (78) FCFFLAT1PLN_OCSFCFFLAT1 M", will shorting the name of the dimensions from FCFFLAT1 to FLAT help?

    • DJAMS
      DJAMS commented
      Editing a comment
      Since the root of your problem is that you're getting a duplicate ID in QCC, I think that shortening the name will only create another pair of duplicate labels.

  • #6
    You know, you had a reasonable workaround in your initial post - using the generic feature, I totally forgot about that. Here are some threads that show how to get the lower tier measured value.

    https://www.pcdmisforum.com/forum/pc...e-profile-data

    https://www.pcdmisforum.com/forum/pc...te-profile-fcf


    Using the generic feature is probably a better way than re-doing your whole spc setup.

    Comment


    • DJAMS
      DJAMS commented
      Editing a comment
      You're welcome. I hope that method works out. I've avoided Xact dimensioning but was forced into using it on this composite per unit profile tolerance. So I don't have any familiarity with the assignment used in that thread.

      If it gets to the point where you think you want to try the XML output from pc-dmis, just post saying so.

    • sim4191
      sim4191 commented
      Editing a comment
      Thanks again DJAMS, the links you posted helped and I was able to make it work.
      Sorry for the late post.

    • DJAMS
      DJAMS commented
      Editing a comment
      You're most welcome. Glad you got it sorted out.

Related Topics

Collapse

Working...
X