ARMY MIL-HDBK-109-1960 STATISTICAL PROCEDURES FOR DETERMINING VALIDITY OF SUPPLIERS' ATTRIBUTES INSPECTION《用于决定承制厂属性检查有效性的统计程序》.pdf

1、6 MAY 1960 Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,- AcE?i%s VM - OFFICE OF THE ASSISTPNT SECRETARY OF DEFENSE WASHINGTON 25, D. C. SUPPLY AND LOGISTICS St.t?-:istical Procedures for Determining Validity of Suppliers Attributes Inspection H-10

2、9 This (interim) handbook was prepared by the Chemical Coqs Materiel Conmaad, I ,-,ah + onh“ “L., bLisArzyucJ iurii“ UYt-LJ i1,ygestion:; for improvernent of tliis handbook should be addressed to Jfice of tt.e ter;o foi- n-r-J Is- - L. 2 5. CORKECTIVE ACTION 5.1 Statistical Criteria Indicate “Warnin

3、g- 19 5.2 Statistical Criteria Indicate “Action“- 19 Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-APPENDIX :a!ile i Limits for Determining Diccrepancits Between Suppliers and Aibe YT Check Ratings for Paired Attribute Sampling Inspections Consumer

4、s Paired Attributes Sampling Inspections -. obie TI: Cricicai Limics for Cuinuiative Check “Katings r.11- TTV -L-L;IJ*- ,c A - -_- As u _- -d. . LP., c L LUUUALAL UA CILLCLUIILE WL nuiiiuaciicLLy i Each f K Lei- ficatio: Trials fQr 3 Ypy-if:eY occeptu:cc pr ,c I, WILL Ur JUUt;CU. Each Lndivida: gill

5、 be Lfispected by the spp:ie :a accordance with the sampling standard. A second sample, proportional j.n size to the prescribed sample taken by the supplier will be drawn from the same lot by the consumer. From Table I, for a specified ratio of sample -T; i zt.s, consumer accercails an Il-g 4 Provid

6、ed by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-everything that the conaumer8 inspector would do. Under such conditions, the consumer can rely upon the euppliers inspection records to support acceptance of supplies offered to him. The amount of sunplng ins

7、pection performed by the consumer can tho8 be reduced to the minimum required to assure a cw of mieriai meeting acceptance quality standards. 2.3 STATISTICAL CRITISRIA FOR PAIRED ATTRIBUTE SAMPLINGS. For con- venence in referring in general terns to the quantities required for the harnogeneity test,

8、 the resulta of paired attribute anplings are represented symbolically in Table A: TABLE A Notation for Two-Sample Test for Homogeneity Defect ive Effect ive Sample Size Suppliers Sele *I “s - ds “S n C n - dc C C Consumers Sample d These data are recorded to decide whether the results of inspecting

9、 two samples, one the suppliers of size n8 and the other, the consmers of size nc, which are found to contain d, and dc defectives (or defects), respectively, are significantly different. The recorded data can be vniently tescd for statistical significance at approximtiy the 5 per cent level (one ch

10、ance in 20 that observed differences might be due to the “luck of the draw“) by means of Table I, which provides the critical values for various ratios of ns/nc : r. The critical limits of Table I are the “action“ numbers for d, when associated uith an observed d, for specified ratios of ns/nc equal

11、 to 1, 2, 3, 5, and 8, respectively. 5 Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-2.3.1 EXAMPLE OF USE OF TABLE I. Suppose a specification prescribes lot sampling inspection in accordance with Military Standard 105 at inspection level II for AQL

12、 (Major) = 1.0 per cent defective and AQL mi ri or 2.5 per cent deictiv. Ii the suppiir prducd iots of i, units each 2nd used single sm-cling, the apprnprigte snapling pi,gfi fr major defects would be sample size = 110 and acceptance number = 3. The supplier performs sunpling inspection and records

13、that he found 2 defec- .;Js in the sample of 110 units inspected for majors. After completing his inspectionj the sippler submits the lot with his inspection record the consumers inspector. Since this is the first lot, the inspector might appropriately choose to perform product verification inspecti

14、on using the #il sarie sw-ple size as the supplier. The :nspeetr perfl5 bg ifispec- cion and finds 5 defectives in his smle of 110 units inspected, The sampling results are displayed as follows: Defect ive Ef fective Sample Size L 08 LLW 1 rn n .,-I- -I- aupprrr n aaunpLe Consumers Sample 5 LOS 110

15、Referring to Table I, for r = 1 (since the ratio of the suppliers sample size to the consumeris sample size equals i), the inspector finds LL-C -d- a-,. c- ,La 9 c-ca- -L. _ 1 L a-L- I - la- CL., LILUL, LULLC:3)1UIIUAII LU LIIS L UFACLLAVCI UYUCLVIZU U;I LIIS OUY)1AJ.CA, LLIC “action“ limit for the

16、number of defectives observed in the consumers sample is 7. Since the consumer noted only 5 defectives in his verifica- tion sample, he has no evidence at this time to doubt the validity of the suppliers inspection result for the lot in question. 6 Provided by IHSNot for ResaleNo reproduction or net

17、working permitted without license from IHS-,-,-After the consumers inspector has performed verification inspec- *a- I- LL- C,-L LIVIL Lor Lne 1;irsc far lots and MS iound no significant discrepancies hetween aimnfiery cpelts and CL+, -si. -A- i.:- -I- -J- y-rr - mJ ASUUCS LLLP PaUiyLC JLCC to one-ha

18、lf that of the supplier. (More definite guidance for reducing verification sampling is available from the operating characteristics of the test as explained in the next section of this handbook.) CLtF UICLLUL L) LCLULU AUL he lt verified hed 3 dfctivs in a sampie of 110 units, but the consumers innp

19、ecrar faund a defectives in I PPElP nf Now, suppose ta.* +.-.-*-I- ,-A c, 55 units from the same lot. Referring to Table i for r = 2 (ratio of euppilers sample size to consumers sample = 2) with the entry ds = 3 d “= -tri +L,. LL-L d, z 3 AIO9 net exceed Che cnrresponding critica? value, dc(a) : 5.

20、indication that ruidan variation is not responsible for the differences Again, from the single test performed, there is no pronounced bened in th pair9 eunplingi frw the same l. 2.4 cWINTTnN OF TEST- FOR Hn3Cr.ENETY. !#hen the i?pplh?r Lnscitiites reduced inspection as provided by MIL-STD 105, the e

21、xpected number of defec- tives in his samplee may be very small (average less than 1.0 per sample). LiKewise, the expected number o eieciives in the consumers sampes, *II minrgined t 3 fixed size retio (cuy r : 2) t rhe Contractors smnles, will be even smaller. Since the power of the test to detect

22、discrepancies in inspection depends in large measure on the total number of defectives found as well as the division of sample sizes, some way must be found to As :or,g CL -+I- I, UUILU up the pb-e f the test. LIIC LULLU A) ?s i systsmi. xcdig the tlgceinll limit, the consumer is a8vised to take cor

23、rective 8 Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-.,-tarr- -A b- -i- - -L- aLcAvu aatu ItuL Lu the next 2 lots provided strong evidence that corrective action was required. Both lots yielded identical sampling resulta, ds 1 and dc = 4, which

24、are significant in accordance with th cririIi f TatLe i. of 5.4Q egh for thece results derived from Table JI is also substantial, The heek iing and, if sunined, -4ould exceed the ”warning“ critical value of Table III for 3. l GENERAL PROCEDURES. At the start of production, when little is known about

25、 a suppliers quality history or the validity of his inspection records, it may be appropriate to perform product verification inspection at the same inspection levei as that required in he spifiatin. Fr -w-.rl =iWYfr*-, if the specific=tFsn required sampling inspection in accordance with MIL-S”D 105

26、 at inspection level I;, the consumers inspector could use that same level for his initial verification inspection. men the consumers inspector has validated the suppliers inspection records after the first few lets, he m,ighc decrease his sample or verification purposes to one-half or one-third of

27、the suppliers sample. As production continues and the process average stabilizes between the upper and lower limits For the ii Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-prescribed AQL, the consumers inspector may shift to a sample size of one-i

28、iith or even one-eignich oi Che suppiiers sample. m,en the scpplier ia puth=riz prior results of product verification to determine the incidence of defects as well as the check ratings obtained. The relative occurrence of defects will L Y 8,J“L- -L-.* onr-fifth the rate previous:y eitafneb for nt-1

29、single ql of 14 .T?nbll_rdi Cnnsequencly, if individual lot ccl!Bporinona are deaired; ehe sampi.e slze ratio (r) should return to 1 or 2. An alternate procedure, more pwerful in differentiating effects of disparate inspection systems, is to arrange For accumihiion or defec- tives ti; 2 vi mie cnscc

30、Ut?ve sSq?es, yr=iding the =q?c =Fze rc,tie remains; constant. Thus, for the results of Table CI where the sample size ratio (r) is maintained equal to 3, the pooled resulta of the first 2 lots (ds = 2 and dc = 3), are significant, whereas the individual lots results are fioc, as judged by ihr Criti

31、ca: va:es of Table l. li; Using the methu.U OC: ponling dota, the nimiber af romyrisons to be included in the group s;iouid be determined in advance, and it should be restricted to only a far (say 5 comparisons in order not to delay decisive action. resuits can be treated as a one iot entiicy for as

32、signment oi chcic ratinga The pooled -L-L T-Llri TT -A +LA a.-l+nA rhorL rstinr, cmn LL11UUgii LUULC: L.L, mglu LLIE LCLUIULQLCU LLLCLiR roLriia bmn. he cmhiEed With ether ratings bearing an the validation o for can over-all evaluation againsr the crtical values of Table III. 3.2 PROCEDUPES BASED ON

33、 OPERATING Ci ARACITRISTICS (O-C) CURVES. Ariaj.op,ous to t5e U-C Curves of acceptance SamplLrig pians, the suopliers inspection records ent. pratiliij .L,CLL,C-L,.=- -h-eov 2 3 4 5 5 6 7 7 8 8 9 LO 10 11 11 12 i2 3 13 14 14 is 15 15 16 7 d n r=5 U, i j i 2 3 3 4 4 5 7 a 6 5 6 6 6 7 7 ? 7 8 8 8 n 9

34、9 9 10 10 I? r=8 a,iAj 1 2 2 3 3 3 3 4 4 4 4 5 5 5 5 5 6 6 6 i, 6 ? 7 7 7 7 0 Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-TABLE I (Continued) A Ub 28 29 3.0 3i 32 33 34 35 r=l b, (; 42 43 45 49 47 48 49 51 r=P u, (A) 23 24 24 25 25 26 27 27 r=3 A

35、 I. mJ 16 17 i7 is 18 18 19 19 r:5 11 11 ii i2 12 12 13 13 r=8 d. : Number of defectives (or defects) observed in the I supplier s sample. = Number of defectives (or defectsj observed in Che consumers sample. dC dc!A! - “Crian“ Limct far CIc. m-on thio rider ie reached or exceeded in the consumers s

36、ample, a course of action is adopted on the premise that a discrepancy does exist. 7. 6 Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-rn o o -l -7 u ri O u. Provided by IHSNot for ResaleNo reproduction or networking permitted without license from I

37、HS-,-,-d,O 12 3 4 5 6 7 8 ij 10 11 12 13 1415 c O 1 1.70 0.69 0.34 0.17 0.09 0.05 0.03 0.01 0.0 0.00 0.00 U.0 .00 0.00 0.000.00 2 2.58 1.24 0.69 0.40 0.23 0.14 0.08 0.95 0.03 0.02 0.01 0.00 0.00 0.00 0.00 0.00 3 3.4i i.b3 i.il .6Y .46 0.28 .i7 .?i 0.07 0.04 .2 ,ui 0.Ui 0.W U.U U. 4 4.20 2.43 1.57 !O

38、6? 0.20 0:oa 0.03 0.02 c.01 o*oo 0?00 oao ci-00 o*oo 0.00 0.00 0.00 Or30 000 8 7.25 4.86 3.65 2.75 2.9 :.59 :.7: u.92 0.69 0.52 0.38 0.78 0.2: u.l3 G.: v.v7 10 1 t“ i3 14 15 i6 ? 18 19 In LV 21 22 23 24 26 27 28 29 IT -.0 30 31 32 33 34 35 36 3; - 38 39 6.25 4.78 3.73 2.94 2.33 1.84 1.45 1.14 0.6 0.

39、64, 0.53 0.41 0.31 0.23 0.17 6.81 5.34 4.24 3.4 2.53 .i9 i.55 i. i.ii U.88 .69 U.54 0.42 .32 .24 7.76 5.91 (r.94 46 1-22. 1.01 8.07 6.89 5.90 5.07 4.35 3.73 3.20 2.74 2.34 2.00 1.69 1.43 1.20 7.42 6.39 5.52 4.77 4.12 3.56 3.07 2.64 2.26 1.94 1.65 1.40 7.96 6.99 5.99 5-21 4.53 3.93 3.41 2.95 2.55 2.2

40、2 ?.ne L.bl 7.41 6.47 5.65 4.94 4.31 3.76 3.26 2.85 2.47 2.14 1.84 7.93 6.95-6.10 5.36 4.70 4.13 3.61 3.16 2.76 2.40 2.08 7.44 6.56 5.79 5.11 4.50 3.96 3.48 3.06 2.67 2.34 7.94 :.3 6.23 5.32 4.89 b.32 3.82 3.35 2.96 2.60 ?$51 6-68 5,rq 5,28 4.69 4-16 3.69 326 2.88 7.99 7.13 6.37 5.68 5.07 4.51 4.02

41、3.57 3.17 7.59 6.80 6.09 5.45 4.88 4.36 3.89 3.46 8.b 7.25 .5i 5.85 5.25 4.7i 4.i 3.i 7.70 6.91: 6.25 5.63 5.7 1.55 1.n9 7.37 6.66 6.02 5.43 4.90 4.41 7.81 7.08 6.41 5.81 5.25 4.75 I .A 6.82 6.L9 5.62 5,m 7.03 7.13 6.58 5.99 5.44 7.65 6.98 6.37 3.80 i.79 .i4 b.i4 7.5b 5.32 7.94 7.30 7.70 7.38 6.75 6

42、. Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-8888888;? 888888;oaa ?88888% ?8? SZSs; a888?g:;o“: ?xx?: gggg;?;$JX 8888?;?;:? 00000OObQO 8%-88888Z9 o 0 0oSaD 00 0 0000000000 0000000000 O000 -N-h 0000300000 ooooo-m.n 0000000000 oooa oooooooooo 0000

43、000000 -.-*o 0000000000 OOOO-N*hN.)i 0000000000 ?94?1 0000000000 0000000000 588oodg2E;: gggo-.gb=“- c;o0o8ooOoo 0000000000 -nn s8So; 333OD3131c2 000N.1W4UQN oooooo-non 0000000000 P 3 o n I N Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-