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.

Mold qualification on different machines

Discussion in 'Qualification & Validation (Also 21 CFR Part 11)' started by jojo63, Jan 11, 2016.

  1. jojo63

    jojo63 Member

    Joined:
    Oct 1, 2015
    Messages:
    6
    Likes Received:
    0
    Trophy Points:
    1
    Location:
    France
    Hello,

    I'm looking your feedback in the activity sector of plastic injection.

    For validation, we perform an IQ-OQ-PQ on the couple injection moulding machine/mold. If we switch the mold on another machine, we perform a new IQ-OQ-PQ.

    This is very long. How do you do ? What advice can you give me?

    The machines are not completely identical, an equivalence relation is impossible.

    Thanks you for your feedback.

    Jojo
     
  2. ncwalker

    ncwalker Well-Known Member

    Joined:
    Sep 21, 2015
    Messages:
    261
    Likes Received:
    168
    Trophy Points:
    42
    Location:
    North Carolina
    This perspective is from die casting, which is similar: first I am unclear why am equivalence relation would be impossible. There are two metrics which suitably describe the ability of a molding machine to make a part . One is clamp tonnage and the other is injection power. Normally when you qualify a given mold in a given machine, it is then qualified in every other machine that is equal or better in clamp tonnage and injection power. Should you need to move it to a lesser machine, it must be requalified. Clamp tonnage is easy - that's just the machine size. Injection power is described by a p-v diagram that should be in the maintenance book of your machine. It is generated during the machine runoff phase and is typically redone every time the front end is rebuilt. Basically the strategy is the part runs in the good machine, therefore it will run in the better machine.

    This does NOT mean you can skip your first shot quality checks. You can definitely make a bad part in the better machine if the settings aren't right, and they will be different. But you do not have to go through the long process of demonstrating the cell can make the part from a statistical-type standpoint (IQ-OQ-PQ). You can just start up and dial it in.

    If this IQ-OQ-PQ is a customer driven requirement and their representative doesn't understand molding processes or p-v diagrams, explain them to them and (gently) remind them they pay you to be the experts on the subject.
     
    Atul Khandekar likes this.
  3. jojo63

    jojo63 Member

    Joined:
    Oct 1, 2015
    Messages:
    6
    Likes Received:
    0
    Trophy Points:
    1
    Location:
    France
    Hi,

    I am in back.

    This is unfortunately not so simple. The parameters mentioned are not important (only according to the mold of capacity). However, the diameter of the plasticizing screw is important, and even if it is identical, the injection parameters may be slightly different between two injection moulding machine.

    Nobody has experience in this area?

    Thanks for your comments.

    Jojo
     
  4. ncwalker

    ncwalker Well-Known Member

    Joined:
    Sep 21, 2015
    Messages:
    261
    Likes Received:
    168
    Trophy Points:
    42
    Location:
    North Carolina
    Your alternate is to demonstrate the OUTPUT is equivalent. You should have variables you are monitoring (injection speed, from measuring screw speed; mold temperature; etc). These are results of your settings. If I know that if I run a job on machine A and get a certain table of results, if I put this job on machine B and get the SAME table of results, I can demonstrate equivalence. You'd want to do more than one stroke, for sure. But ... if the behaviors of the setup on Machine A can be duplicated on Machine B then you have equivalence.
     
  5. jojo63

    jojo63 Member

    Joined:
    Oct 1, 2015
    Messages:
    6
    Likes Received:
    0
    Trophy Points:
    1
    Location:
    France
    I will explain what we are practicing.

    For QO, we perform 3 small runs on a machine A: 1 run with minimum settings, 1 run with nominal settings and 1 run with maximum settings. The goal is to demonstrate that the manufactured products conform using extreme settings.

    For the QP, we produce 3 standards runs to nominal settings on the machine A and we check the conformity of products.

    A product may be manufactured on a machine B where parameters can be slightly different.

    for not to manufacture 3 new standards runs on machine B, I thought proceed this way:

    Realization of 3 small runs on machine A (mini, maxi and nominal parameters)
    Realization of 3 small runs on machine B (mini, maxi and nominal parameters)
    These runs are included in a QO.

    For the QP, realization of 3 standards runs (eg 2 runs on machine A and 1 run on machine B) and verification of product compliance.

    What do you think?
     
  6. ncwalker

    ncwalker Well-Known Member

    Joined:
    Sep 21, 2015
    Messages:
    261
    Likes Received:
    168
    Trophy Points:
    42
    Location:
    North Carolina
    I think that's an excellent approach. I am actually impressed that your QO on machine A is tested in the tails as well. That's good.

    So your minimum and maximum settings on the QO/Machine A run demonstrate the process window that Machine A can achieve. You're QP run demonstrates that Machine B can make the nominal part. The only hole in the analysis is that the Machine B has NOT demonstrated a comparable process window to Machine A.

    You should do this, but here's the thing - you really do NOT have to do this on each and every part. What you need to demonstrate is that B performs "like" A. If you do this on one part and demonstrate it, then qualify a new part on machine A, by superposition, it is qualified on B.

    Your battle will be your customer. An unreasonable customer who doesn't understand how injection molding works will say "Sorry, I want it done on MY parts." Sadly, this is where the dog and pony show comes in to play. If you do your equivalency study on a selected part and put it together in a very nice, clean professional presentation, that will sway a customer engineer who is not familiar with the process. You may have to repeat such a study, say, every 3 years, to demonstrate that one of the machines hasn't drifted.

    But what you have laid out to me makes sense, I would buy in to it.
     
    jojo63 likes this.
  7. jojo63

    jojo63 Member

    Joined:
    Oct 1, 2015
    Messages:
    6
    Likes Received:
    0
    Trophy Points:
    1
    Location:
    France
    NcWalker,

    I will work in this direction. Thank you for this discussion. :)