TCM 1 1554-(R) – Howmet Fixations Simmonds – St Cosme en Vairais (FR) PQ223

QAR: Click Here
Conditions for approval required by:  Julien Gauther

4 Replies to “TCM 1 1554-(R) – Howmet Fixations Simmonds – St Cosme en Vairais (FR) PQ223”

  1. FPI: mfg batch size is 50999 parts. As per EN3152, tables 3 and 4, AQL=0,065%. A down arrow is marked against [35001 – 150000] mfg batch size range. This down arrow means “Use sampling plan below (sample size and Ac)”. The sampling plan and Acceptance number referred to in this case are 800 parts and Acceptance 1 defect.
    This issue has been raised in ASD-CERT Mechanical Workshop minutes, dated February 12th, 2019.
    Many thanks in advance
    Best regards Julien

  2. Response to TC’s comment about sample size.

    There is no dispute about the tables 3 and 4 to be used in the EN3152:2004 a derivative of the ISO 2859-1, tables 2-A and 6-A.

    Table 3 indicates the “Classification of defects”.
    In the most left column “Category of defects” the division into Major and Minor is made.
    The second column “AQL” (Acceptance quality limit) indicates which AQL applies in Table 4 “Sampling plans for visual inspections and dimensional characteristics”.

    For example, table 3 indicates that for the Characteristic, Presence of locking elements and the Surface discontinuities the 0.0065% AQL applies.
    For Thread size, Perpendicularity of head-bearing surface, Coating and so on, an AQL of 1.0 % applies.
    Then there is also an AQL of 2.5% for the Wrenching size and configuration etc.
    The AQL of 4.0% is for All other dimensions etc. as can be read.
    So much for Table 3

    Table 4 here in the EN3152:2004 is a combination of two tables.
    a. The first two columns are the “Sample size”, which is the number of items in the sample related to the Batch size.
    b. And the third column shows the “Acceptance number (Ac) and limiting quality (LQ10) in accordance with the acceptance quality limit (AQL)”
    AQL is de quality level that is the worst tolerable process average when a continuing series of lots is submitted for acceptance sampling. Divided into Ac (acceptance quality) limit and LQ a quality level which for the purposes of sampling inspection is limited to a low probability of acceptance.

    The Batch size indicates how many products are involved in the Sample size is determined according to a statistical sampling scheme.
    This is a stand-alone table, where, for example, 35001 to 150000 products give a sample size of 500 products.
    That’s all it is. This is the sole purpose of the first two columns.

    Let’s go to the remaining column: Acceptance number (Ac) and limiting quality (LQ10) in accordance with the acceptance quality limit (AQL).
    In this table we see the AQLs which we have determined in table 3.

    Let’s go back to the 35001 to 150000 products where it has been statistically determined that a sample size of 500 applies.
    If we now take AQL 0.065%, then there should be no deviations in the lot up to and including 315 pieces. If we drop to the lot with 500 and 800 products, 1 product may be found there during inspection with deviation to declare this lot AC (acceptable).

    The up and down arrows in these columns have absolutely nothing to do with the previously determined “Sample size”.
    For 35001 to 150000 products, it remains 500 and never becomes 800. The up and down arrows have nothing to do with the sample size, but with the AC in the AQL 0.065%. How would you explain the previous position with an AQL of, for example, 2.5%
    Again, the first two columns have nothing to do with the arrows in the other columns.

    Hope with this explanation I have made it clear that the number of parts (sample size) to be tested should be 800 pieces is not correct and that the correct sample size (500 pieces) has been used by the manufacturer!

    For example, if a “batch size” is 100 pieces. That gives a sample size of 20 pieces. Then, based on the downward arrows, a minimum of 3201 products would have to be produced to reach the 200 pieces, otherwise I would not be able to use the table. This does not seem to me to be the reason why they have created single sampling plans for normal inspections as specified in ISO 2859-1

    In addition, in terms of ‘the numbers, it is stated in the paragraph “Substantiation by analogy” in this report that this Re-qualification can be seen as by analogy and that can be used to reduce the amount of required testing.
    See for the statements above the EN9133, point 8.1.

    Unfortunately, I do not have the aforementioned ASD-CERT Mechanical Workshop minutes, dated February 12th, 2019. I would greatly appreciate it if I could receive this from you. I cannot imagine that this deviates from my explanation above.
    Thank you in advance for this!

    Best Regards
    Kees

    1. Julien Gauther
      05/12/2023 AT 8:46 AM
      FPI: mfg batch size is 50999 parts. As per EN3152, tables 3 and 4, AQL=0,065%. A down arrow is marked against [35001 – 150000] mfg batch size range. This down arrow means “Use sampling plan below (sample size and Ac)”. The sampling plan and Acceptance number referred to in this case are 800 parts and Acceptance 1 defect.
      This issue has been raised in ASD-CERT Mechanical Workshop minutes, dated February 12th, 2019.

      Genevieve’s Reply : See the new TR updated enclosed . The sampling was completed to achieve the required quantity requested in EN3152 .

  3. Dear all,
    Many thanks for taking into account this sampling plan.
    Approved from my side.
    Have a nice day
    Best regards Julien

Leave a Reply

Your email address will not be published. Required fields are marked *

Form restricted to registered, logged in users. Please return to the home page and use the Create Account option.