Inspector Vs CMM, Non Machined Plastic Injection Aerospace

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

  • Inspector Vs CMM, Non Machined Plastic Injection Aerospace

    We produce a large qty of non machined plastic parts (mold to deburr to shipping) for Aerospace applications. I'm starting to see a trend. FAIRS completed in open lay-up, table (standard measurement instruments) many moons ago, now being placed on CMM's in receiving inspection and getting kicked out.

    I'm now being summoned to a customer to get "beat on" for their CMM receiving inspection rejecting my parts that passed FAI. I need Pros and Cons for CMM inspection for these types of parts to help buffer my customer visit? Thank you in advance.

    For example, I read at times that folks try to keep plastic parts of CMM when ever possible due to surface irreg, micro structures, etc...

  • #2
    Rigidity.
    If the parts are rigid, there is no reason that the accuracy of CMM inspection should be any different than the accuracy of the inspection performed on a machined metal piece.
    If the plastic parts are flexible, thin wall, warped or distorted, then it will be challenging to get repeatable or accurate results with a CMM.
    If old-style (non-GDT) dimensioning is used, there are a great number of ways to interpret how a part can be set up. I only mention that because you noted that these parts had the FAIR completed "many moons ago".
    Lately, it occurs to me
    What a long, strange trip it's been.

    2017 R1 (Offline programming)

    Comment


    • #3
      mold "draft" angle could come into play
      Which one gets ridden today? MPH vs MPG..tough choice, both are FUN
      sigpic

      Starrett RGDC 4028-24 :alien:
      Demon vintages 3.7, 4.1, 4.2, 4.3, 2009

      Comment


      • #4
        Most molded plastic parts will distort, this makes it virtually impossible to get repeatable measurements. We had this problem here just his week, how can you get an accurate alignment on a surface that deviates 0.010 or more in some instances?
        Eh?

        Comment


        • #5
          +1 to all the above + my 2 coins of the realm, namely does the print have a callout for Free State or Constrained?

          thin walled/flexible/warped plastic parts are a PITA to start with - we had several special fixtures made to check critical dimensions on the plastic parts we use. We were never able to get repeatability using the CMM on these parts, even after purchasing a low force module
          sigpic
          Originally posted by Ironhoe
          I got something under my sporran for you, take care of it and you got my vote.

          Comment


          • #6
            The best thing you can do is make sure parts have sufficient time to cool. 24 hours by our studies done here is when the majority of shrinkage occurs. Are you waiting that long? Otherwise, your customers will get very different readings! If you can't wait 24 hours, then wait at least an hour.
            BeckyK
            Sheffield, B&S
            (1)1808-M, (4)D-28, (1)RS-70, (3)RS-150, (1)Mxcel PFX
            PC-Dmis 2009 MR1, 2013 MR1

            Comment


            • #7
              Simulated datum for plastic parts on a surface plate is more reliable than on a CMM. A CMM is less reliable due to the fact that it uses “points” to establish part reference frame, all features are check to established FCF (all 3 axes are mutually square to one another) which’s harder to achieve on a surface plate. If you know the parts well, you still can get very decent results on a CMM, it’s just a matter of knowing where to take the hits and how many to eliminate any potential projection errors

              Comment


              • #8
                Projection Errors?

                Comment


                • #9
                  Originally posted by Nicholas Kohanowski View Post
                  Projection Errors?
                  It means that if you part alignment is off by one degree due to bad hits…the location of a feature would be off by ~.017” per inch.

                  Comment


                  • #10
                    Originally posted by BeckyK View Post
                    The best thing you can do is make sure parts have sufficient time to cool. 24 hours by our studies done here is when the majority of shrinkage occurs. Are you waiting that long? Otherwise, your customers will get very different readings! If you can't wait 24 hours, then wait at least an hour.
                    +1

                    Here we use cmm holding fixtures to measure thin walled pastic parts, repeatability is not great but with tolerances of +/- 2mm in most cases, who cares?
                    Global S Blue 12.15.10 HH-AS-T2.5 / HP-S-X1h |PC-DMIS CAD++|Quindos7 CAD; Leitz Reference Xi 15.9.7

                    Comment


                    • #11
                      Another factor that is a dirty secret w/in the quality field & one that no one will admit to is the "fudge factor'! I used to see it all the time in almost every place I ever worked! It's something that only a human can apply. Prior to CMM's a human being measured the parts & recorded the data on a inspection report. How many inspectors "fudged" on the sample size? Maybe the sample size calls for a qty of 50. Some inspectors would check maybe 5 parts or even less & record the sample size as 50 as called for on the IR. At times the receiving or final inspector may find a dimension marginally in tolerance or even OOT. A supervisor,manager,or Q.E. would suggest that the dimension in question would be "OK" or "we always done it that way". The inspector would usually record in tolerance dims. on the IR. With the advent of the CMM sample sizes are checked according to spec. & marginal or OOT dims. are printed as hard data for every part run across the CMM. Now this is not something to bring to your customer but something to think about. Anyone with years experience in quality knows this but as stated will not admit it. It's a hard fact of life in the quality field. At most shops an inspector would be admonished for "taking too much time" if he actually tried to do the job per spec. checking the full sample. Along the same lines many inspectors back in the day were admonished for recording marginal or OOT dims on the IR. "Come on- we're only talkin' bout a few tenths"! The CMM doesn't care about what people think & records dims as measured.
                      sigpic
                      Southern Man don't need him around anyhow!

                      Comment


                      • #12
                        Draft (is the dim +draft or -draft?), warp, shrinkage, moisture content, free-state vs. restrained... it all factors in when measuring plastic. All of this can be overcome on the CMM if a competent programmer understands the proper application of GD&T and uses appropriate fixturing. I changed from the wonderful world of plastics to the exciting world of metal stamping. GEEZ... stamping is so much easier!

                        How do the parts measure on your cmm? Is the customer's complaint valid or are the measurment correlation issues? This is the first thing you need to address. If the cmm measurement data correlates between locations and on other measurement equipment, then it would be best to have a member of management address this concern with the customer.

                        The best case scenario is that the customer's cmm programmer isn't measuring the parts correctly. This also needs to be handled with kid gloves.

                        The concerns you mentioned about surface texture/structure are only valid based on the amount of tolerance that is allowed. The cmm will average surfaces with texture pretty well. In the world of ± 0.13mm, I've never had a significant problem due to surfaces.

                        Good luck.
                        To some the glass is half full
                        To some it is half empty
                        To me it is poorly designed by the Eng & Prod wants me to find a way to measure it in spec!

                        Comment


                        • #13
                          All good food for thought. Thanks folks!

                          The customer has what has been nick-named the 'super duper' CMM program. It inspects 30 features at once on this part. CMM reports 25 features out of print, eight of which "must haves" for next sub assembly fit. But with my old school tools or my CMM inspection, we are finding these features in. I've been summoned out to their local to figure out the discrepancies.

                          My CMM programmer says first glance at their program, he doesn't understand their clocking? Also they have lastest rev. pc-dmis, we have v4.1, I haven't been able to get them to save and forward for us to use their programming on our CMM. Trying to get from getting creamed when meet with their QC folks next week.

                          First thought was why take a $3 part and put it on the CMM.

                          Comment


                          • #14
                            Originally posted by george frick View Post
                            Another factor that is a dirty secret w/in the quality field & one that no one will admit to is the "fudge factor'! I used to see it all the time in almost every place I ever worked! It's something that only a human can apply. Prior to CMM's a human being measured the parts & recorded the data on a inspection report. How many inspectors "fudged" on the sample size? Maybe the sample size calls for a qty of 50. Some inspectors would check maybe 5 parts or even less & record the sample size as 50 as called for on the IR. At times the receiving or final inspector may find a dimension marginally in tolerance or even OOT. A supervisor,manager,or Q.E. would suggest that the dimension in question would be "OK" or "we always done it that way". The inspector would usually record in tolerance dims. on the IR. With the advent of the CMM sample sizes are checked according to spec. & marginal or OOT dims. are printed as hard data for every part run across the CMM. Now this is not something to bring to your customer but something to think about. Anyone with years experience in quality knows this but as stated will not admit it. It's a hard fact of life in the quality field. At most shops an inspector would be admonished for "taking too much time" if he actually tried to do the job per spec. checking the full sample. Along the same lines many inspectors back in the day were admonished for recording marginal or OOT dims on the IR. "Come on- we're only talkin' bout a few tenths"! The CMM doesn't care about what people think & records dims as measured.


                            ... pft, he don't know what he's talkin' 'bout... most of us don't even like fudge!
                            Eh?

                            Comment


                            • #15
                              Another factor that is a dirty secret w/in the quality field & one that no one will admit to is the "fudge factor'! I used to see it all the time in almost every place I ever worked! It's something that only a human can apply. Prior to CMM's a human being measured the parts & recorded the data on a inspection report. How many inspectors "fudged" on the sample size? Maybe the sample size calls for a qty of 50. Some inspectors would check maybe 5 parts or even less & record the sample size as 50 as called for on the IR. At times the receiving or final inspector may find a dimension marginally in tolerance or even OOT. A supervisor,manager,or Q.E. would suggest that the dimension in question would be "OK" or "we always done it that way". The inspector would usually record in tolerance dims. on the IR. With the advent of the CMM sample sizes are checked according to spec. & marginal or OOT dims. are printed as hard data for every part run across the CMM. Now this is not something to bring to your customer but something to think about. Anyone with years experience in quality knows this but as stated will not admit it. It's a hard fact of life in the quality field. At most shops an inspector would be admonished for "taking too much time" if he actually tried to do the job per spec. checking the full sample. Along the same lines many inspectors back in the day were admonished for recording marginal or OOT dims on the IR. "Come on- we're only talkin' bout a few tenths"! The CMM doesn't care about what people think & records dims as measured.
                              Quit the quality field twice over this issue.
                              Any more when someone says "it's only a couple of tenths" I just start asking for money. Lots of money.
                              They haven't given me the password to record rejections. I just email and leave the report on my leads desk. Usually don't find out what the decision was. Don't care either.
                              I like to get hammered on Friday night
                              Sometimes I can’t wait
                              So Monday’s alright

                              Lyle Henderson

                              Comment

                              Related Topics

                              Collapse

                              Working...
                              X