Some standard content:
The People's Republic of China
Guohao Standard
Plastic Resin Sampling Method
GB 2547 —81
(Confirmed in 1989)
This standard covers how to select some products from a batch of plastic resin products to form a representative sample during sampling inspection. It uses the principles of mathematical statistics to determine the sample size and uses random methods to interpolate. I. Determination of Sample Size
1. In order to make the estimated value of the average value of the overall quality of the product obtained by the sample reflect the actual situation of the population satisfactorily, it is necessary to obtain an appropriate number of sampling units (i.e., the smallest package) from the population. The sample size can be obtained by the following formula n=(A00/E)z
Where: - Sample size, i.e., the number of sampling units aa-
Estimated value of the standard deviation of the overall quality of the product! The maximum permissible error between the estimated value of the average quality of the product obtained from the sample and the average quality of the product obtained by the corresponding method for each sampling unit; A--intercept coefficient, which indicates the corresponding probability that the error between the estimated value of the average quality of the product obtained from the sample and the average quality of the product obtained by the corresponding method for each sampling unit exceeds the maximum permissible error E. Formula (1) can be transformed into Formula (2), which is sometimes used for convenience, n=(AV./e)2
where V. =α./X--the estimated value of the coefficient of variation of the overall quality of the product, e=E/X-
--the maximum permissible error expressed as a percentage of the text! X--the average quality of the product.
....(2)
(1), or V. is obtained:
a, based on the historical data of the same product. The standard deviation or coefficient of variation of several batches of products with equal or similar sample sizes are calculated using the following formulas.
SVE(-E/n-
Wherein: S-
Standard deviation of the sample of the batch:
-Single measurement value
Indicates--Arithmetic mean of single measurement value--Sample size of the batch
V!--Coefficient of variation of the batch.
Then calculate their average value.
=S or =/ respectively as the estimated value of or.
Promulgated by the State Administration of Standards
The Ministry of Chemical Industry of the People's Republic of China proposed.....(3)
Implemented on March 1, 1982
Ministry of Chemical Industry, Zui Guang Chemical Research Institute, etc. Drafted
Wherein: 1 is the batch number.
GB 2547-81
Note: When calculating の. Or according to a, generally speaking, the larger the sample size π, the larger the batch number ", the more accurate the result, but in practical application, if n\ is larger, the batch number 1 can be smaller, if the number is smaller, the batch I should be larger. For example, when n' is greater than 20; 1 can be 4 to 6. When h is about 10, 1 is better than 10.
b. If there is no such historical data available, you can follow the principle of "Note" in a and start the work of accumulating data in order to estimate the or V that meets the requirements.
Sufficient samples must be drawn regularly for separate inspections in order to continuously correct o or V. (2) Determination of the maximum allowable error E or e. The maximum allowable error E or e can be specified according to needs and possibilities. The so-called "need" refers to the accuracy required for the estimated value of a certain quality characteristic, which should be considered based on the impact of the quality characteristic on the application of the product. If a slight change in a certain quality characteristic will cause product transformation, or have a great impact on molding processing and product application, then the accuracy of the characteristic estimate obtained from the sample should be higher, that is, E or e should be specified to be smaller, otherwise E or e can be specified to be larger.
The so-called "possibility" refers to whether the manpower and material resources required to test the sample size are appropriate. According to formula (1) or (2), the sample size is inversely proportional to the square of the maximum allowable error E or e. If E or e is unnecessarily set too small, it will become too large and the inspection cost will be very high, which is often uneconomical. Therefore, if the specified E or e is too large, E or e can be adjusted (increase E or e: that is, reduce the accuracy of the estimated value) to obtain a smaller n. In short, when determining the maximum allowable error E or ε, the issue to be considered is to achieve an appropriate balance between the required accuracy of the estimated value and the cost of obtaining an estimated value of such accuracy. (3) Determination of the probability coefficient A.
The probability coefficient can be determined based on the credible interval required for the result. In industrial production, it is generally sufficient to set it to 1.96. At this time, the probability that the error between the estimated value of the overall quality mean of the product obtained from the sample and the average value of the overall quality of the product measured for each sampling unit exceeds the maximum allowable error E or is 5%. The A value corresponding to other probabilities can be obtained from the normal distribution table as needed, such as:
(4) For plastic resin products, there are usually several quality characteristics. The number n required for each quality characteristic can be calculated separately, and then the largest one can be taken as the sample size of the inspection batch. The \ number can also be calculated using the one with the largest coefficient of variation among the key quality characteristics related to the main use of the product.
. Sampling: Selection of sampling units 2. The sample size # calculated according to formula (1) or (2) should be randomly selected from the product population. The specific steps can be carried out according to one of the following two methods:
(1) Random sampling method
, the total number of sampling units N of the product, are numbered consecutively in a certain (or production) order, starting from 1 to N. b. Use a random number table to determine the number of the sampling units to be selected (see Appendix 1 for the random number table and its use). (2) Systematic sampling method
a. Divide the total number of sampling units N of the product by the sample size, and take the integer part of the quotient as the sampling interval. b. Randomly determine a sampling unit from the 1st to the th sampling units, and then take a sample from every other sampling unit. Note: If sampling at the discharge port is convenient or the product is in the process of moving, the systematic sampling method can be used. 3. Sampling method
GB 2547-81
III. Sampling
Sampling should be taken from the sampling unit determined in 2 (1) b or 2 (2) b. The sampling tools and sampling methods used should ensure that representative samples can be taken from the sampling unit. This is especially true for products that may be uneven in packaging or transportation (such as high scores for large and small items, different moisture content, etc.). In this case, it is appropriate to use a suitable size of sampler to take samples from different parts (top, middle, bottom, center, outer circle, etc.). For products that are uniform in the package, an inch sampler is appropriate. 4. Sampling quantity
(1) If the purpose of sampling is only to obtain the average quality of the product as a whole, the samples taken from each package can be mixed for testing. The total amount of samples taken should be at least twice the amount of the test tape. Take samples of roughly equal size from the selected sampling units and mix them evenly. Then divide them into two parts, one for testing and the other for storage in a sealed container that will not contaminate the product. Each part must be marked with the product name, batch number, production date, sampling date, etc. (2) If the purpose of sampling is to obtain the quality dispersion of each sampling unit in the whole batch of products, the samples taken must not be mixed and must be tested separately. In this case, the amount of sample taken from each bundle of sampling units should be twice the amount required for the test. After mixing them evenly, divide them into two parts, one for testing and the other for storage in a sealed container that will not contaminate the product. Each part must be marked with the product name, batch number, production date, sampling date, etc. (3) For tests with very small sample quantities, samples several or several dozen times the test quantity should be taken from the determined sampling units (with the principle of taking out representative samples). After taking out, use the cone quartering method to evenly shrink the sample (see Appendix 2) until the appropriate amount is obtained. Some granular materials are large in size. After being fined to a certain degree, they can be crushed into small particles by mechanical pulverization, and then the sample is reduced until the appropriate amount is obtained.
Note: When mechanically crushing, be careful not to overheat the sample to prevent degradation. 5. For plastic tree currency, there are many cases of obtaining the average quality, so in daily inspection, mixed tests can be carried out. At this time, during the product transmission process or during the product packaging process, continuous sampling with an automatic continuous sampler is also a reasonable sampling method. However, in order to understand and master the data on the dispersion of quality within the batch, it is necessary to regularly draw samples of appropriate sizes for separate tests. The sensitivity data accumulated in this way can be used for the determination and control of α, or to improve production processes Appendix 1
GB 2547-81
Random number table and its use in random sampling The random number table is composed of numbers from 0 to 9 arranged randomly. This appendix provides two pages, each page has 50×50=2500 numbers arranged in 50 rows (horizontally) and 50 columns (woven, one number per column), which can be used as any number of digits as needed. The usage of the random sampling table is as follows: 1.1. Determine the page number. Close your eyes and place the pencil on the random number table. If the number the pencil tip lands on is an odd number, then you will choose to take the number from page 1. If the number the pencil tip lands on is an even number, then you will choose to take the number from page 2. 2. Determine the starting point. Close your eyes and place the pencil on the random number table. Use the two-digit number the pencil tip lands on to determine the row. Use the same method to determine the column (when the two-digit number the pencil tip lands on is 51-99 or 00, subtract or add 50 to make it 01-49 or 50). 3. Advance direction. Starting from the page, row, and number determined by 1 and 2, if you are taking a single or double digit number (the column determined as the starting point is the first digit, and the one to the right is the second digit), then move from left to right. When you reach the right, move to the left end of the next row and continue to the right. If you are taking a number with more than three digits, then take it from top to bottom. When you reach the bottom, move to the top of the next column and continue to take it from the bottom. The number of digits taken depends on the total number of sampling units N. 4. When the number taken out is less than V, take the original number. When it is greater than N, divide it by N and take the remainder. Arrange them in order and remove the repeated numbers until n numbers are determined. This is the selected sampling unit number. Example: 10 samples are randomly taken out from a batch of 200 sampling units. We decide to start from the 47th row of page 2 and the first column according to the methods 1 and 2. The three-digit number taken out is 798. Then, from top to bottom, we take 988, 034, 055, 761, 499, 606, 965, 601, 971, 248, - after processing (remainder after dividing by 200), we get 198, 188, 34, 55, 161, 99, 6, 165, 1, 171, 48. There are no repeated numbers, so the first 10 numbers are the selected sampling unit numbers. 03 47 43 73 86
97 74 24 67. 62
16 76 62.27 66
12 56 85 99 26
5559563564
1622 77 94 39 | |tt | 75
23 42 40 64 74
526281995
37 85 94 35 12
70 29 17 12 13
66 62 16 37. 36
99 49 57 22 77
16 08 15 4 72 | |tt | 24
16 90 82 66 59
11 27 94 75 06
35 24 10 16 20
3823168638
31 96 26 91 47
66 67 40 67 14
14 90 84 45 11
6805:511800
20 46 78 73 90
64 19 58 97 79
0526937060
07 97 10 88 23
68 71 86 85 85
2699616553
14 65 52 68 75
17 63 77 58 71
90. 26 59 21 19
41 23 52 55 99 | |tt | 14
90 96 23 70 00 | |tt | 54 82
57 24 56 06 88
16 95 55 67 19
78 64 56 07 82.
09 47 27 98 54
44 17 16 58 09
84 16 07 44 99 | |tt | | 82 97 77 77 81 | 45 72 | |tt | 32 20 30 | |tt | 49 13 | |tt | 15 13
09 98 42 99 64
64 87 66 47 54
58 37 78 80 70
87 59 36 22 41,
71 41 61 50 72
2352233312
31 04 49 69 96
31 99 73 68 68
94 58 28 41 36
98 80 33 00 91
73:81 53 94 79
73 82 97 22 21
22 95 75 42 49
39 00 03 06 90
GB 2547--81
random Number table (nest 1 page)
46 98 63 71 62
42 53 32 37 32
32 90 79 78 53
0503729315
31 62 43 09 90
17 379323 18||tt| |77 04 74 47 67
98 10 50 71 76
52 42 07 44 8
49 17 46 09 62
79 83 86 19 62
8311 463224
07 45 32 14 08
00 56 76 31 38
42 34 07 96 88
13 89 51 03 74
97 12 25 93 47
1664361600
45 59 34 68 49
20 15 7 00 49
44 22 78 84 26
71 92 38 67 54
9657 69 36 10
77 84 57 03 29
53 75 91 93 30
6719007174
02 94 37 34 02
79 78 45 04 91
87 75 65 81 41
34· 86 82 53 91|| tt||11 05 65 09 68
52 27. 41 14 86
07 60 62 93. 55
04 02 33 31 08
01 90 10 76 06
92 03 61 58 77
61 71 62 99 15
7332081112
4210 50 6742
26 78 63 06 55
12 41 94 96 26
96 93 02 18 39
1047484588
35 81 33 03 76
45 3759 0309
0977931982
33 62 46 86 28
0503272483
39 32 82 22 49
55 85 78 38 36
33 26 16 80 .45
27 07 36 :07 51
13 55 38 58 59
57 12 10 14 21
06 18 44 32 53
87 35 20 96 43
21 76 33 50 25
12 86 73 68 07
15 51 00 13 42
90 52 84 77 27
06 76 50 03 10
20 14 85 88 45
3298940772
8022025353
54 42 06 87 98
17 76 37 19 04
70 33 24 03 54
04 43 18 66 79
12 72 07 34 45
52856660 44
04 33 46 09 52
13 58 18 24 76
96469242.45
10 45650426
34 25 20 57 27
60 47 21 29 68
76 70 90 '30 86
16 92 53 56 16
40 01 74 91 62
0052434885|| tt||76 83 20 37 90
22 98 12 22 08
59 33 82 43 90
3954164936
40 78 78 89 62
59 56 78 06 83
06 51 29 16 93|| tt||44. 95 92 63 16
32 17 55 85 74
13 08 27 01 50
44 95 27 36 99
07 02 18 36 07
13 41 43 89 20
24 30 12 No. 4 60
90. 35 57 29 12
74 94 80 04 04
08 31 54 46. 31
72 89 44 06 60
02 48 07 70 37
94 37 30 69 32
60. 11 14 10 95
24 51 79 89 73
88 97 54 14 10
88 26 49 81 76
23 B3 01 30 30
84 26 34 91 64
892. 12 06 76
44, 39 52 38 79
99 66 02 79 54
08 02 73 43 2B
55 23 64 05 05
10 93 72 88 71
93 86 79 10 75
-86 60. 42 04 53
35 85 29 48 39
07 74 21 19 30
97 77 46 44 80
94 77 24 21 90
99 27 72 95 14
38 68 88 11 80
6B 07 97 06 57
16 54 56 95 52
97 60 49 04 91
11 04 96 67 24
40 48 73 51 92
02 02 37.03 31
38 45 94 30 38
027550 959B
48 51 84 08 32
2765268962
57 16 00 11 66
07 62 74 95 80
49 37 38 44 59
47 95 93 13 30
02 67 74 17 33
52 91 05 70 74
580577 0951|| tt||29 56 24 29 48
94 44 67 16 94
15 29 39 39 43
02 96 74 30 3
259932 7023
97 17 14 49 17
1899107234||tt ||82 62 54 65 60
45 07 31 66 49
5394133847
35 80 39 94 88
16 04 61 67 87
9089007633
53 74 23 99 67
633806·8654
35305821 46
63 43 36 82 69
98 25 37 55 26
02 63 21 17 69||tt| |64 55 22 21 82
85 07 26 13 89
5854162415
34 85 27 84 87
0392 1827 46
62 93 30 27 59
08 45 93 15 22
07 08 55 18 40
0185899566
72 84 71:14 35
88 78 28 16 84
45 17 75 65 57
96 76 28 12 54| |tt||43 31 67 72 30
50 44 66 44 21
22 66 22 15 86
96 24 40 14 51
3173916119
78 60 73 99 84|| tt||84 37 90 61 56
36 67 10 08 23
07 28 59 07 48
10 15 83 87 60
55 19 68 97 65
5381291339
51 88 32 68 92
36 91 70 29 13
37 71 67 95 13
93 66 13 83 27
0296084565
49 83 43 48 35
84 60 71 62 46
18 17 30 88. 71
79 69 10 61 78
75 93 36 57 83
3830922903
51 295010 34
21 31 38 86 24
29 01 23 87 88 | | tt | | 9533952200 | 65 2694
06 72 17 10 94
65 51 18- 37 88
. 01 91 82 81 46
71 50 0 8 56
48 22 28 06 00||tt| |01 10 07 82 04
51 54 44 82 00
61 48 64 56 26
57 .99 16 96 58
37 76 41 66 48
60 21 75 46 91
45 44 76 13 90
51 10 19 34 88
GB 2547—81
Random number table (page 2)
94 62 67 86 24
02 82 90 23 07
25 21 31 75 96
61 38 44 12 45
74 71 12 94 97
3815701148
61 54 13 43 91
59 63 69 36 03
62 B1 65 04 69||tt ||90 18 ± 48 13 26
30 33 72 85 22
88 97 80 61 45
98 77 27 85 42
24 94 96 61 02
15 84 97 19 75
19 11 58 49 26
13 52 53 94 53
284019 7212
22 01 11 94 25
24 02 94 08 63
66 06 58 05 62
26 63 75 41 99
23 22 30 88 57
60 20 72 93 48
43 89 94 36 45
7010239805
98 93 35 0886
89 64 58 89 76||tt| |79 24 31 66 66
0373521656
35 01 20 71 34
33 98 74 66 99
80 03 54 07 27
20 02 14 95 94||tt ||92 79 64 64 72
13 05 :00 41 84
82 88 33 69 96
40 - 80 81 30 37
44 91 14 88 47
71 32 76 95 62
56 20 14 82 11
06 28 81 39 38
31 57 75 95 80
37 79 81 53 74
58 02 39 37 67
18 74 72 00 18
24 36 59 87 38
54 97 20 56 96
38 46 82 68 72
0822237177
50 11 17 17 76
75 45 69 30 96
25 12 74 75 67|| tt||71 96 16 16 88
38 32 36 66 02
68 16 54 35 02
58 42 36 72 24
96 67 47 29 83
9857072369|| tt||56 69 47 07 41
85 11 34 76 60
99 29 76 29 81
83 86 62 27 89
21 48 24 06 93
00 53 56 90 27
62 33 74 82 14
40 14 71 94 58
96 94 78 32 66
64 85 04 05 72
28 54 96 53 84
93 07 54 72 59
72 36 04 19 76
34 59 23 05 38
89 23 30 63 15
87 00 22 58 40
74 21 97 90 65
82 25 06 8 63
51 97 02 74 77
73 24 16 10 33
42 10 14 20 92
38 79 58 69 32
82 07 53 89. 35
15 74 80 08 32
32 14 82 99 70
91 01 93 20 49
98334119 95
79 62 67 B0 60
49 28 24 00 49
32 92 85 88 65
24 02 71 37 07|| tt||43 40 46 86 98
8278122329
69 11 15 83 80
38 18 65 18 97
37 70 15 42 57
84 64 38 56 98||tt ||23 53 04 01 63
28 88 61 08 84
57 55 66 83 15
12 76 39 43 78
86 31 57 20 18
73 89 65 70 31
60 40 60 81 19
6864367445
69 36 38 25 39
42 35 48 96 32
58 37 52 18 61
94 69 40 06 07||tt ||65 95 39 69 58
90 22 91 07 12www.bzxz.net
76 48 45 34 60
33 34 91 58 93
30 14 78 58 27
91 98 94 05 49
3342293887||tt ||53 73 19 09 03
4594193881
50 95 62 74 33
01 32 90 76 14
48 14 52 98 94
2145 5709 7
47 45 15 18 60
25 15 35 71 30
56 34 20 47 89
92. 54. 01 75 25
96 42 68 63 86
61 29 08 93 67
76 15 48 49 44
52 83 90 94 76
16 55 23 42 45
Name 1 76 80 26 92
96 35 23 79 18
1646705080
8060471897||tt ||82 96 59 26 94
47 63 53 38 09
75 91 12 81 19
55 65 79 78 07
5434818535
03 92 18 68 75||tt ||00 83 26 91 03
06 66 24 12 27
13 29 64 19 28
85 72 13 49 21
65 66 80 39 07
99 01 30 98 64
45 76 08 64 27
69 62 03 42 73
73 42 37 11 61
64 63 91 08 25
95 60 78 46 75
99 17 49 48 76
24 62 01 61 16
19 59 50 88 92
4803 4515 22
14 52 41 52 48
03 37 18 39 11
18 16 36 78 86
56 80 30 19 44
7835340872
01 64 18 39 96
63 14 52 32 52
86 6359 B0 02
01 47 59 38 00||tt| |22 13 88 83 34
56 54 29 56 93
14 44 99 81 07
13 80 55 62 54.
53 89 74 80 41
66 07 93 :89 30
19 48 56 27 44
82 11 08 95 97
88 12 57 21 77
99 82 93 24 98
43 11 71 99 31
74 54 13 26 84
04 32 92 08 09
18 55 63 77 09
70 47 14 54 36
54 96 09 11 06
82 80 84 25 39
05 98 90 07 35
67 72 16 42 79
63493021 30
66 39 67 98 60
Appendix II
GB 2547-1
Push This method can be used to divide the sample equally and reduce the sample volume to the minimum required amount. The specific method is as follows: 1. Pour the mixed sample into a 2. Use a shovel to scoop the sample into a pile to form a cone. Each shovel should be scooped from the bottom and poured onto the top of the cone, allowing the sample to flow evenly around the cone.
3. Use a clean flat plate to flatten the top of the cone into a frustum until the height of the frustum is one quarter of the height of the cone. "4. Use a clean ruler to divide the circular sample into four equal parts through its center. 5. Randomly discard the two opposite parts, put the remaining two parts together, and repeat the above operation until until the required sample volume is obtained.
Tip: This standard content only shows part of the intercepted content of the complete standard. If you need the complete standard, please go to the top to download the complete standard document for free.