1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.
Dismiss Notice
You must be a registered member in order to post messages and view/download attached files in this forum.
Click here to register.

CNC Software Validation & Change Control

Discussion in 'Qualification & Validation (Also 21 CFR Part 11)' started by Candi1024, Aug 27, 2015.

  1. Candi1024

    Candi1024 Well-Known Member

    Joined:
    Jul 30, 2015
    Messages:
    129
    Likes Received:
    83
    Trophy Points:
    27
    Location:
    Pennsylvania
    Backstory: A part with "new" dimension was ordered by R&D to be ran on a CNC. It was then to be measured on a CMM. When the part was delivered to R&D, it was not the correct dimensions. Both operations were ran on an old revision of the part drawing.
    My boss's concern is that we currently do not have adequate control of the programs in these particular machines (and maybe others). We may program the CNC and name that program after a certain part number, but when a revision is made, this is not easily recognizable. In addition, when we run part number A, we may use program B, because this particular assembly of both parts are exactly the same.

    Valid concern.

    He feels we should be performing software validation.

    I'm not sure he using the correct terminology. I'd like some input on what we could do to prevent this.

    - IQOQ = Not sure if this has been performed or not. If not this should be completed, however I don't believe this will have any effect on this particular problem.

    - Software Validation = Not completed, but if the IQOQ was very robust, It may be covered as part of that document. However since this covers the equipment functions, this wouldn't affect part differences (assuming that it would currently pass software validation)

    - Process Q = We have not validated this particular process because it affects the function of our device, and we believe in our testing it is fully verified.

    - Part Qualification = Again I need to investigate this, but my guess is that every part had been checked on the CMM during design qualification at the very least. I think we may be getting closer to the issue here, in that during that part qualification, the program would be set, named, and the program backed up elsewhere.

    Change control seems like the issue here. I also need to investigate how one off parts for R&D are programmed and saved.

    What am I missing?
     
  2. Jan van der Kuil

    Jan van der Kuil New Member

    Joined:
    Aug 3, 2015
    Messages:
    4
    Likes Received:
    4
    Trophy Points:
    2
    Change control requires systems thinking: you must identify all elements (specifications, work instructions, software, validation requirements, et cetera) in case of an engineering change.

    When you want to change the dimensional specifications of a part it is not enough to change your drawing only, as that will not result in that part being manufactured. You also need to change the CNC program, need to install that program in the CNC machines (IQ), ensure that the correct program is being selected for that part by the operator, and verify that the program change results is the specified dimensions (PQ).

    The way to manage that all activities are being performed is to identify all these activities in an engineering change request. For traceability purpose it is needed to link CNC programs not only to a part ID (or multiple if applicable), but also to the revision number. CNC programs must also have a revision number for revision control purposes, so you will be able to audit/review if CNC programs have been updated when required by engineering changes.
     
    Candi1024 and Claes Gefvenberg like this.
  3. Claes Gefvenberg

    Claes Gefvenberg Moderator Staff Member

    Joined:
    Jul 31, 2015
    Messages:
    230
    Likes Received:
    208
    Trophy Points:
    42
    Location:
    Eskilstuna, Sweden
    Bingo!
    ...and also to the CMM prg in question.
     
    Candi1024 likes this.
  4. JAltmann

    JAltmann New Member

    Joined:
    Sep 2, 2015
    Messages:
    2
    Likes Received:
    2
    Trophy Points:
    2
    We have a decent program change control system here for the CNC stuff, but recently I separated the CMM stuff into its own system. The name is simple and easy to follow, it includes the P/N_Program Type_Op #_Machine_Program Revision. It makes it quite clear to the operators if they are running the correct program for the op/part #. We only keep the latest revision available at the local machines and do not allow the operators to access the controlled copies on the server. The programs revisions are controlled thru a change management system which ties to the part design and work instructions revisions.

    So if were not made aware of a part design change would be the only way we do not inspect to the right level. The previous system was cryptic and required tribal knowledge or "magic" decoder rings. The new system is a little more cumbersome overall, best the feedback has been very positive and helped keep from running the wrong program or passing invalid data to the next operation.