358 CHAPTER 12: Acceptance Sampling

The ES

PI Ps

97 98 99 100 101 102 103

X

FIGURE 12.11 The distribution of a quality characteristic and the proportion outside
of specifications (p).

but the proportion of units out of specifications. Meanwhile, when there is a double
specification uses procedure 2 or method M, which we will exemplify because it is the
which applies in both cases.
If the standard deviation,σ, of the batch or of the process is known and stable, the plans result-
tes son los más económicos y eficientes. Cuando σ it is unknown it can be used the diversion-
standard deviation or the range of the sample. However, the range method requires a sample
bigger and not very advisable.
MIL STD 414 consists of four sections: section A is the general description of
sampling plans, including definitions, letter codes for sample sizes and
CO curves for the different sampling plans. Section B provides various plans of
sampling based on the sample standard deviation for the case whereσ is unknown
The section presents sampling plans based on the range of the sample.
It provides sampling plans by variables for the case where the standard deviation is
known. Next, the use of section B is studied.

Steps to design a MIL STD 414 plan

1. Determine the batch size.
2. Specify the NCA (or AQL).
3. Choose the inspection level (usually level IV, which can be changed if the situation...
the justification). The higher the level of inspection, the stricter the plan is (more
quickly falls its CO curve.
4. In table 12.16 and according to the lot size and the level of inspection, find the
sample size code letter.
[Link] la tabla 12.17, de acuerdo con la letra código y el NCA, se busca el plan simple para
normal inspection, which consists of a sample size, and the value M, which is the
maximum percentage of defective items tolerated in the batch.
6. In the same table 12.17, starting from the NCAs that are at the bottom, one finds the
plan that would be employed under strict inspection, with its corresponding paranyM values.
 Acceptance sampling by variables (MIL STD 414) 359

Unknown variability:
Unknown variability:
deviation method Known variability
rank method
standard

Specification limit Specification limit

on one side from both sides

Procedure 1 Procedure 2 Procedure 2

Methods MethodM MethodM

FIGURE 12.12 Organization of MIL STD 414.

7. Randomly select a sample of size, and each piece of it is measured.

–
the quality feature. With the obtained data calculate the mean X and the standard deviation
sample standard, S.
8. Of the following two indices, according to the type of specifications that the feature has.
quality characteristic, calculate for one or both.

ES X
Z ES for higher specification (ES)
S

X EI
Z EI for lower specification (EI)
S
–
Note that both indices: ZEIyZESis the distance between the sample mean, X,
and the corresponding specification, expressed in units of standard deviation of the
show, S. Therefore, the value of these indices will be large if S is small and if X tends to
–
to be at the center of the specifications.
9. Estimate the proportion of defective units in the lot. To do this, refer to table 12.18, in the
column of ZEIyZESlocate its corresponding value and move along that row until
the sample size column of the inspection plan, n. The value found in the
The intersection value of Zyn corresponds to the estimate of the percentage of defective items.
lower side lot, pi, or at the top side,s, respectively.
10. Decision of acceptance or rejection:

•For variables with only lower specification. Accept the sip lotiis less than or equal to
value of M (found in step 6). Otherwise, reject it.
•For variables only with upper specification. Accept the sip batchs is less than or equal to
M. Otherwise, reject it.
For variables with double specification. Accept the batch if the sum of the percentage infected...
higher than the superior,= pi+ ps, is less than or equal to M. Otherwise, reject the
batch.
360 CHAPTER 12: Acceptance Sampling

TABLE 12.16 Letter codes for sample size for MIL STD 414 (sampling for variables).

INSPECTION LEVELS

LOT SIZE I II III IV V

3 to 8 B B B B C
9to15 B B B B D
16 to 25 B B B C E
26 to 40 B B B D F
41 to 65 B B C E G
66 to 110 B B D F H
111to180 B C E G I
181 to 300 B D F H J
301 to 500 C E G I K
501 to 800 D F H J L
801 a1300 E G I K L
1301 a3200 F H J L M
3201 a8000 G I L M N
8001 to 22000 H J M N O
22001 to 110000 I K N O P
110001 to 550000 I K O P Q
550001 and more I K P Q Q

EXAMPLE 12.6
In an auto parts factory, there have been problems. From a particular lot, they are selected randomly
with the dimension of a certain steel bar at the moment 40 bars are measured for their length. With the 40 data.
to assemble it. The ideal length of the bar is 100 it is calculated
mm, with a tolerance of 2 mm. It was decided to implement- – = 100.15 yS= 0.8.
X
to take an internal acceptance sample for the purpose of
to avoid allowing batches to pass to the assembly stage and with these, the indices are calculated:
with very poor quality.
The batch size for these bars is 3000. 102 100.15
ZES 2.31
In accordance with the background and the purposes, the election is made 0.8
an NCA (AQL) of 1.0%. In this way, batches with 1.0%
bars outside the specifications will have a high
probability of being accepted for assembly. The level of 100.15 98
ZEI 2.6875
The inspection that will be used is the usual one (level IV). From the ta- 0.8
bla 12.16 finds that the code letter for the size
from the sample isL. Let's assume that the des- With these values, in the den column= 40 from the table
standard deviation of the process and that it is going to be estimated 12.18 and taking the value of ZESequal to the nearest value
conS. Therefore, according to step 6, in the table (2.30), it is estimated that the percentage of bars in the batch that
12.17 it is found that the normal inspection plan is exceeding the upper specification is equal tos= 0.888%
In the same way, it is found that the percentage of ba-
n = 40,M= 2.71% rivers that have a length shorter than the (EI) espi= 0.236%.
and the severe inspection plan is Therefore, the total percentage that is estimated to be outside of
specifications esp= 0.888+ 0.236= 1.124%, which is
n= 40,M= 1.88%
less than 2.71%, so the lot is accepted.
 Use of statistical software 361

TABLE 12.17 Table for normal and severe inspection (unknown variability,
standard deviation method), method M.
LYRICS
CODE ACCEPTABLE QUALITY LEVEL: AQL OR NCA (NORMAL INSPECTION)
SIZE
DEL
OFTHE
SIZE
SAMPLE
OF THE 0.04 0.065 0.10 0.15 0.25 0.40 0.65 1.00 1.50 2.50 4.00 6.50 10.0 15.0
SAMPLE

M M M M M M M M M M M M M M

B 3 7.59 18.86 26.94 33.69 40.47

C 4 1.53 5.50 10.92 16.45 22.86 29.45 36.90

D 5 1.33 3.32 5.83 9.80 14.39 20.19 26.56 33.99

E 7 0.422 1.06 2.14 3.55 5.35 8.40 12.20 17.35 23.29 30.50
F 10 0.349 0.716 1.30 2.17 3.26 4.77 7.29 10.54 15.17 20.74 27.57

G 15 0.099 0.099 0.312 0.503 0.818 1.31 2.11 3.05 4.31 6.56 9.46 13.71 18.94 25.61
H 20 0.135 0.135 0.365 0.544 0.846 1.29 2.05 2.95 4.09 6.17 8.92 12.99 18.03 24.53
I 25 0.155 0.156 0.380 0.551 0.877 1.29 2.00 2.86 3.97 5.97 8.63 12.57 17.51 23.97

J 30 0.179 0.179 0.413 0.581 0.879 1.29 1.98 2.83 3.91 5.86 8.47 12.36 17.24 23.58
K 35 0.170 0.170 0.388 0.535 0.847 1.23 1.87 2.68 3.70 5.57 8.10 11.87 16.65 22.91
L 40 0.179 0.179 0.401 0.566 0.873 1.26 1.88 2.71 3.72 5.58 8.09 11.85 16.61 22.86

M 50 0.163 0.163 0.363 0.503 0.789 1.17 1.71 2.49 3.45 5.20 7.61 11.23 15.87 22.00
N 75 1.147 0.147 0.330 0.467 0.720 1.07 1.60 2.29 3.20 4.87 7.15 10.63 15.13 21.11
O 100 0.145 0.145 0.317 0.447 0.689 1.02 1.53 2.20 3.07 4.69 6.91 10.32 14.75 20.66

P 150 0.134 0.134 0.293 0.413 0.638 0.949 1.43 2.05 2.89 4.43 6.57 9.88 14.20 20.02
Q 200 0.135 0.135 0.294 0.414 0.637 0.945 1.42 2.04 2.87 4.40 6.53 9.81 14.12 19.92

.065 0.10 .15 .25 .40 .65 1.00 1.50 2.50 4.00 6.50 10.00 15.00

Acceptable quality levels: NCA or AQL (severe inspection)

Use of statistical software

To obtain a CO curve to characterize the plan. Next, we will see how to do this.
Once a sampling plan by attributes has been obtained by any method, it is important

in Excel and Statgraphics.

Type A curve, plans (N, n, c)

Type A curves are based on the hypergeometric distribution (see Chapter 3), and are applied
when 10n>N. It is important to remember that a CO curve for the plan (N,n,c) graphsPcon-
trap, where the values of the proportion of defective items in the batch are and P is the probability
to accept them. Therefore, with Excel it is necessary to calculate Pa:

Pa= P ( 0 ) +… + P(c)
The calculation of the probability that a certain number of defective items, x, will appear in the sample,
it is made with:
[Link](x,n,K,N)
362 CHAPTER 12: Acceptance Sampling

TABLE12.18Tabletoestimatethepercentageofdefectiveitemsinthelot(piops) paraZEIoZES
using the standard deviation method.
ZESo SAMPLE SIZE
ZEI
3 4 5 7 10 15 20 25 30 35 40 50 75 100 150 200

0 50.00 50.00 50.00 50.00 50.00 50.00 50.00 50.00 50.00 50.00 50.00 50.00 50.00 50.00 50.00 50.00
0.1 47.24 46.67 46.44 46.26 46.16 46.10 46.08 46.06 46.06 46.05 46.04 46.04 46.03 46.03 46.02 46.02
0.3 41.63 40.00 39.37 38.87 38.60 38.44 38.37 38.33 38.31 38.29 38.28 38.27 38.25 38.24 38.22 38.22

0.35 40.20 38.33 37.62 37.06 36.75 36.57 36.49 36.45 36.43 36.41 36.40 36.38 36.36 36.35 36.33 36.33
0.45 37.26 35.00 34.16 33.49 33.23 32.92 32.84 32.79 32.76 32.74 32.73 32.72 32.68 32.67 32.66 32.65
0.50 35.75 33.33 32.44 31.74 31.37 31.15 31.06 31.01 30.98 30.96 30.95 30.93 30.90 30.89 30.87 30.87

0.55 34.20 31.67 30.74 January 3029.64 29.41 29.32 29.27 29.24 29.22 29.21 29.19 29.16 29.15 29.41 29.13
0.60 35.61 30.00 29.05 28.32 27.94 27.72 27.63 27.58 27.55 27.53 27.52 27.50 27.47 27.46 27.45 27.44
0.70 29.27 26.67 25.74 25.03 24.67 24.46 24.38 24.33 24.31 24.29 24.28 24.26 24.24 24.33 24.21 24.21

0.75 27.50 25.00 24.11 23.44 23.10 22.90 22.83 22.79 22.76 22.75 22.73 22.72 22.70 22.69 22.68 22.67
0.80 25.64 23.33 25.51 21.88 21.57 21.40 21.33 21.29 21.27 21.26 21.25 21.23 21.22 21.21 21.20 21.20
0.85 23.67 21.67 20.93 20.37 20.10 19.94 19.89 19.86 19.84 19.82 19.82 19.80 19.79 19.78 19.78 19.77
0.90 21.55 20.00 19.38 18.90 18.67 18.54 18.50 18.47 18.46 18.45 18.44 18.43 18.42 18.42 18.41 18.41

0.95 19.25 18.33 17.86 17.48 17.29 17.20 17.17 17.15 17.14 17.13 17.13 17.12 17.12 17.11
November 17 17.11
1.00 16.67 16.62 16.36 16.10 15.97 15.91 15.89 15.88 15.88 15.87 15.87 15.87 15.87 15.87 15.87 15.87
1.05 13.66 15.00 14.91 14.77 14.71 14.68 14.67 14.67 14.67 14.67 14.68 14.68 14.68 14.68 14.68 14.68
1.10 9.84 13.33 13.48 13.49 13.50 13.51 13.52 13.52 13.53 13.54 13.54 13.54 13.55 13.55 13.56 13.56

1.15 0.29 11.67 October 1212.27 12.34 12.39 12.42 12.44 12.45 12.46 12.46 12.47 12.48 12.49 12.49 12:30
1.20 0.00 10.00 10.76 11.10 11.24 11.34 11.38 11.41 11.42 11.43 11.44 11.46 11.47 11.48 11.49 11.49
1.25 0.00 8.33 9.46 9.98 10.21 10.34 10.40 10.43 10.46 10.47 10.48 10.50 10.52 10.53 10.54 10.56
1.30 0.00 6.67 8.21 8.93 9.22 9.40 9.48 9.52 9.55 9.57 9.58 9.60 9.63 9.64 9.65 9.66

1.35 0.00 5.00 7.02 7.92 8.30 8.52 8.61 8.66 8.69 8.72 8.74 8.76 8.79 8.81 8.82 8.83
1.40 0.00 3.33 5.88 6.98 7.44 7.69 7.80 7.86 7.90 7.92 7.94 7.97 8.01 8.02 8.04 8.05
1.45 0.00 1.67 4.81 6.10 6.63 6.92 7.04 7.11 7.15 7.18 7.21 7.24 7.28 7.30 7.31 7.33
1.50 0.00 0.00 3.80 5.28 5.87 6.20 6.34 6.41 6.46 6.50 6.52 6.55 6.60 6.61 6.64 6.65

1.55 0.00 0.00 2.87 4.52 5.18 5.54 5.69 5.77 5.82 5.86 5.88 5.92 5.97 5.99 6.01 6.02
1.60 0.00 0.00 2.03 3.83 4.54 4.92 5.09 5.17 5.23 5.27 5.30 5.33 5.38 5.41 5.43 5.44
1.65 0.00 0.00 1.28 3.19 3.95 4.36 4.83 4.62 4.68 4.72 4.75 4.79 4.85 4.87 4.90 4.91
1.70 0.00 0.00 0.66 2.62 3.41 3.84 4.02 4.12 4.18 4.22 4.25 4.30 4.35 4.38 4.41 4.42

1.75 0.00 0.00 0.19 2.11 2.93 3.37 3.56 3.66 3.72 3.77 3.80 3.84 3.90 3.93 3.95 3.97
1.80 0.00 0.00 0.00 1.65 2.49 2.94 3.13 3.24 3.30 3.35 3.38 3.43 3.48 3.51 3.54 3.55
1.85 0.00 0.00 0.00 1.26 1.09 2.56 2.75 2.85 2.92 2.97 3.00 3.05 3.10 3.13 3.16 3.17
1.90 0.00 0.00 0.00 0.93 1.75 2.21 2.40 2.51 2.57 2.62 2.65 2.70 2.76 2.79 2.82 2.83

1.95 0.00 0.00 0.00 0.65 1.44 1.90 2.09 2.91 2.26 2.31 2.34 2.39 2.45 2.48 2.50 2.52
2.00 0.00 0.00 0.00 0.43 1.17 1.62 1.81 1.91 1.98 2.03 2.06 2.10 2.16 2.19 2.22 2.23
2.05 0.00 0.00 0.00 0.26 0.94 1.37 1.56 1.66 1.73 1.77 1.80 1.85 1.91 1.94 1.96 1.98
2.10 0.00 0.00 0.00 0.14 0.74 1.16 1.34 1.44 1.50 1.54 1.58 1.62 1.68 1.71 1.73 1.75

2.15 0.00 0.00 0.00 0.06 0.58 0.97 1.14 1.24 1.30 1.34 1.37 1.42 1.47 1.50 1.53 1.54
2.20 0.000 0.00 0.000 0.015 0.437 0.803 0.968 1.061 0.120 1.161 1.192 1.233 1.287 1.314 1.340 1.352
2.25 0.000 0.00 0.000 0.001 0.324 0.660 0.816 0.905 0.962 1.002 1.031 1.071 1.123 1.148 1.173 1.186
2.30 0.000 0.00 0.000 0.000 0.233 0.538 0.685 0.769 0.823 0.861 0.888 0.927 0.977 1.001 1.025 1.037
(continues)
 Use of statistical software 363

TABLE12.18Tabletoestimatethepercentageofdefectiveitemsinthelot(piops) paraZEIoZES
using the standard deviation method (continued).
ZESo SAMPLE SIZE
ZEI
3 4 5 7 10 15 20 25 30 35 40 50 75 100 150 200

2.35 0.000 0.000 0.000 0.000 0.163 0.435 0.571 0.650 0.701 0.736 0.763 0.779 0.847 0.870 0.893 0.905
2.40 0.000 0.000 0.000 0.000 0.109 0.348 0.473 0.546 0.594 0.628 0.653 0.687 0.732 0.755 0.777 0.737
2.45 0.000 0.000 0.000 0.000 0.069 0.275 0.389 0.457 0.501 0.533 0.556 0.589 0.632 0.653 0.673 0.684
2.50 0.000 0.000 0.000 0.000 0.041 0.214 0.317 0.380 0.421 0.451 0.473 0.503 0.543 0.563 0.582 0.592

2.55 0.000 0.000 0.000 0.000 0.023 0.165 0.257 0.314 0.352 0.379 0.400 0.428 0.465 0.484 0.502 0.511
2.60 0.000 0.000 0.000 0.000 0.011 0.125 0.207 0.258 0.293 0.318 0.337 0.363 0.398 0.415 0.432 0.441
2.65 0.000 0.000 0.000 0.000 0.005 0.094 0.165 0.211 0.243 0.265 0.282 0.307 0.339 0.355 0.371 0.379
2.70 0.000 0.000 0.000 0.000 0.001 0.069 0.130 0.171 0.200 0.220 0.236 0.258 0.288 0.302 0.317 0.325

2.75 0.000 0.000 0.000 0.000 0.000 0.049 0.102 0.138 0.163 0.182 0.196 0.216 0.243 0.257 0.271 0.277
2.80 0.000 0.000 0.000 0.000 0.000 0.035 0.079 0.110 0.133 0.150 0.162 0.181 0.205 0.218 0.230 0.237
2.85 0.000 0.000 0.000 0.000 0.000 0.024 0.060 0.088 0.108 0.122 0.134 0.150 0.173 0.184 0.195 0.201
2.90 0.000 0.000 0.000 0.000 0.000 0.016 0.046 0.069 0.087 0.100 0.110 0.125 0.145 0.155 0.165 0.171

2.95 0.000 0.000 0.000 0.000 0.000 0.010 0.034 0.054 0.069 0.081 0.090 0.103 0.121 0.130 0.140 0.144
3.00 0.000 0.000 0.000 0.000 0.000 0.006 0.025 0.042 0.056 0.065 0.073 0.084 0.101 0.109 0.118 0.122
3.05 0.000 0.000 0.000 0.000 0.000 0.004 0.018 0.032 0.043 0.052 0.059 0.069 0.083 0.091 0.099 0.103
3.10 0.000 0.000 0.000 0.000 0.000 0.002 0.013 0.024 0.034 0.041 0.047 0.056 0.069 0.076 0.083 0.086

3.15 0.000 0.000 0.000 0.000 0.000 0.001 0.009 0.018 0.026 0.033 0.038 0.046 0.067 0.063 0.069 0.072
3.20 0.000 0.000 0.000 0.000 0.000 0.001 0.006 0.014 0.020 0.026 0.030 0.037 0.047 0.062 0.057 0.060
3.25 0.000 0.000 0.000 0.000 0.000 0.000 0.004 0.010 0.015 0.020 0.024 0.030 0.038 0.043 0.048 0.050
3.30 0.000 0.000 0.000 0.000 0.000 0.000 0.003 0.007 0.012 0.015 0.019 0.024 0.031 0.035 0.039 0.042

3.35 0.000 0.000 0.000 0.000 0.000 0.000 0.002 0.005 0.009 0.012 0.015 0.019 0.025 0.029 0.032 0.034
3.40 0.000 0.000 0.000 0.000 0.000 0.000 0.001 0.004 0.007 0.009 0.011 0.015 0.020 0.023 0.027 0.028
3.45 0.000 0.000 0.000 0.000 0.000 0.000 0.001 0.003 0.005 0.007 0.009 0.012 0.016 0.019 0.022 0.023
3.50 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.002 0.003 0.005 0.007 0.009 0.013 0.015 0.018 0.019

3.55 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.001 0.003 0.004 0.005 0.007 0.011 0.012 0.015 0.016
3.60 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.001 0.002 0.002 0.004 0.006 0.008 0.010 0.012 0.013
3.65 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.001 0.001 0.001 0.003 0.004 0.007 0.008 0.010 0.010
3.70 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.001 0.002 0.002 0.003 0.005 0.006 0.008 0.008

3.75 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.001 0.001 0.002 0.002 0.004 0.005 0.006 0.007
3.80 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.001 0.001 0.002 0.003 0.004 0.005 0.006
3.85 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.001 0.001 0.001 0.002 0.003 0.004 0.004
3.90 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.001 0.001 0.002 0.003 0.003 0.004

where x is the number of defects in the sample= 0,...,c), it's the sample size, Nes
the batch size and the possible number of defective items in the batch, which is obtained
conK= ROUND(p*N). Therefore, it is necessary to calculate:
Pa= [Link](0,n,K,N)+… + [Link](Ac,n,K,N)
for different dependent values. Hence, it is necessary to generate a broad sequence of possible
dep values, and for each of them calculate K= ROUND(p*N) and calculate P with the
hypergeometric distribution.