General rules for sampling inspection of bulk materials with uniform size
Some standard content:
National Standard of the People's Republic of China
General rules for sampling inspection of hulk materials with uniform size1 Subject content and scope of application
This standard specifies the sampling inspection method, procedure and general principles for hulk materials with uniform size. CB/T 13732 92
This standard is applicable to the estimation and inspection of the delivery batch or batch average quality of bulk materials with uniform size (such as grain, raw sugar, fertilizer, etc.). 2 Reference standards
GB3358
Statistical terms and symbols
GB4086
Statistical distribution value table
GB4091 Conventional control chart
3 Terms and symbols
3.1 Terms
3.1.1 Bulk materials
Refers to materials that are continuous and loose and difficult to distinguish as individuals or sampling units: such as a pile of coal, a pile of grain, a pile of fertilizer, etc. 3.1.2 Bulk materials with uniform particle size
Bulk materials with uniform particle size and a small coefficient of variation that meet the requirements. 3.1.3 Delivery batch
Refers to the bulk materials of various types provided at one time agreed upon by the producer and the user. A delivery batch can consist of one batch or multiple batches. 3.1.4 Batch
The unit that constitutes a delivery batch.
3.1.5 Bulk
The bulk weight of a delivery lot or batch, or the total number of subpacks or samples contained in it. 3.1.6 Subpacks
The non-intersecting parts of a delivery lot or batch specified for sampling. Packaging can be actual (such as packaged bulk) or artificially divided (such as bulk bulk! 3.1.7 Bulk bulk
Bulk materials that do not have a clear actual packaging. For example, a shipload of unpackaged raw sugar, a carload of packaged grain, etc., and the packaging can be divided artificially. 3.1.8 Packaged bulk
Bulk materials that have a clear actual packaging. Such as packaged fertilizers, packaged grain, etc. 3.1.9 Sample
A certain number of materials extracted from a larger amount of bulk. 3.1.10 Sample quantity
The single quantity of a single sample,
Approved by the State Administration of Technical Supervision on October 7, 1992 and implemented on June 1, 1993
3.1.11 Collected samples
GB/T 13732—92
A composite sample of incremental samples of dried tea taken from bulk materials. 3.1.12 Bulk sample
A collection of all incremental samples and collective samples taken from a delivery batch or lot. 3.2.13 Test sample
A sample suitable for testing the quality characteristics of bulk materials. 3.1.14 Reduction
The process of preparing a bulk sample, collective sample or incremental sample into one or more test samples. 3.1.15 Conveyor belt sampling
When the delivery batch or the material in the batch is transported by a belt conveyor, the sampling of incremental samples of a certain length across the entire belt width is cut from the conveyor belt or the drop point.
3.1. 16 Hold sampling
Sampling of additional samples from the hold or loading and unloading tools during loading and unloading when the delivery batch or batch is transferred on board. 3.1.17 Truck sampling
Sampling of additional samples from the truck or loading and unloading tools when the delivery batch or batch is transferred on the truck. 3.1.18 First batch inspection
Sampling inspection for the first time on a certain bulk material, or although the bulk material has been inspected before, there is no reference data on its quality fluctuation,
3.1.19 Non-first batch inspection
Sampling inspection for a certain bulk material that is not the first time, and there is historical data on the quality fluctuation of the bulk material for reference. 3.2 Symbols
Product: The average value of a certain quality characteristic of the delivery batch or batch. X: Estimated value .
T: Precision of estimation. It describes the precision that the sample estimate of the delivery batch or batch average quality expressed as one standard deviation must achieve. www.bzxz.net
: Variance between subpackaging. It describes the variance of a quality characteristic between subpackaging or layers within a delivery batch or batch. S: Estimated value of.
: Variance within subpackaging. It describes the variance of a quality characteristic between subpackaging or layers. S%: Estimated value of,
: Sampling variance. It describes the error caused by using the quality of the sample to represent the quality of the entire delivery batch or batch. Its square root is used to express sampling precision. S;n Estimated value of.
: Intra-batch variance. It describes the variance of a quality characteristic between subpackaging or layers within a delivery batch or batch. S;& Estimated value of.
: Standard deviation of x, which represents the precision of the sample estimate of the delivery batch or batch average quality expressed as standard deviation. S; estimated value of x.
%: test variance. It describes the size of the error caused by the test sample. S: estimated value.
Reduction variance. It describes the size of the error caused by the combination of factory samples and the preparation of samples by reduction method. S's estimated value.
nNumber of samples taken from the delivery batch or batch. : Number of subpackaging taken from the delivery batch or batch GB/T 13732-92
n: Number of samples taken from each subpackaging taken. t: Number of initial sample collections.
N: Batch.
UCI. : Control chart, control limit.
CI: Center line of control chart.
LCL: Lower control limit of control chart.
R: Range.
SmThe maximum value of samples that may appear in a series of delivery batches or batches of the same bulk material. S: the maximum value of samples that may appear in a series of batches or batches of the same type of bulk material. Sm: the maximum value of samples that may appear in a series of batches or batches of the same type of bulk material. : standard normal distribution P quantile.
? Confidence level.
Half the length of the confidence interval on the effect side,
a: producer risk.
3: user risk.
Product: acceptable flux level.
Construction: limit quality level.
: upper acceptable quality level.
: lower acceptable quality level.
: upper limit quality level,
Ratio: half of the lower limit quality level.
Start: judgment constant.
c\standardized distance.
4 Calculation of delivery batch or batch average quality
4. 1 Selection of predetermined estimation accuracy T
Before estimating the delivery batch or batch average quality, the value of T must be predetermined. The selection of T should take into account factors such as the batch size, sampling costs, the preciousness of the inspected materials, and the purpose of estimating the average quality, and can be determined by negotiation between the manufacturer and the user. 4.2 First batch inspection
4.2.1 Packaged bulk materials
4.2.1.1 Number of additional samples
For the first batch inspection where additional samples are easy to add, the procedure for estimating the delivery batch or batch average quality is as follows (see Figure 1): a
GB/T 1373292
Determine the number of initial nest samples to be obtained, n is generally not less than 8. Delivery batch or batch
;m sub-packaging
several groups (m sub-packaging--group)
ni collective samples
nr test groups
observation values
From the delivery batch or batch, m sub-packagings are selected, generally m-6, and then several incremental samples are selected from each sub-packaging (the selection of incremental sample size see 6, and the same number of incremental samples are selected from each sub-packaging. Note, 1) Here the sub-packaging sizes are assumed to be equal. If the sub-packaging sizes vary greatly, proportional incremental sampling C
collective samples should be selected.
Randomly divide the mnt subpackages into groups or m subpackages, and get n groups. Mix the subpackages in the same group to get nd. Make each of the samples into a test specimen independently. Test each specimen once, and get observation values in total, recorded as X,,X.,..,X.
Calculate xs*
Calculate n1
If n≤, go to step 1.
If>, go to step 2.
GB/T 13732—92
（4）
h. Draw m(n, -n,) more packages from the delivery batch or batch, and then draw samples from these packages and group, reduce and test them in the same way as in step bcd. Finally, a total of observations (including n observations in step d) are obtained, recorded as XX…,X. 1. Calculate x, S
Calculate #
k, if n≥n+20%n, set n; n+=n and return to step 1. Otherwise, go to step 1. 1. Calculate Sx
The average quality of the delivery batch or batch is estimated to be Sx, and the estimated accuracy is Sx. The length of half of the bilateral confidence interval of the product reliability is m.
+（9)
d SUa+n
Under the confidence level, the lower confidence limit of the average quality of the delivery batch or batch is another d, and the upper confidence limit is ×+d, which can be abbreviated as (-d, x++d) (WeChat level 7).
4.2.1.2 Where it is difficult to add samples
For the first batch inspection where it is difficult to add samples, the procedure for estimating the average quality of the delivery batch or batch is as follows (see Figure 2): h.
GB/T 13732-92
Based on experience or other methods, estimate the S value of this bulk material. Delivery batch or batch
, 10 subpacks
, 10 incremental samples
, 8 industrial samples
, 8 test samples
, 8 full values
, use the secondary sampling method, select 10 subpacks from the delivery batch or batch, and then select 10 incremental samples from each subpack (the selection of incremental samples is shown in Figure 62.
nnw should satisfy:
)S+ S%+ Ss + S
and it is specified to be a multiple of 8. (11) where N is the batch size of the subpack (11)
, which satisfies (11): There are many combinations, and 10 economically reasonable combinations should be selected. At the same time, considering that it is difficult to add samples, as many subpacks as possible should be selected.
S and S can be obtained in advance by the method in Appendix B of this standard. c. Randomly divide the sampled packages into 8 groups of equal size, mix all the sampled packages in the same group, and get 8 samples. Make a test specimen for each of the 8 samples, and test each specimen once. Get 8 observation values in total, recorded as X,,-++t, X..
Calculate,
-（12
(13）
GB/T 13732—92
If E2. 0, then Sm+SwM, S and S% used in formula (11) are considered valid. Otherwise, repeat the above sampling, merging, dividing and testing procedures (see Figure 2) in the selected subpackaging, and obtain 8 more observations. Together with the 8 observations obtained in the previous round, calculate the arithmetic average of the 16 values, and record the average value as tt. ", and use tt as the estimate of the average mass of the delivery batch or batch, and its estimation accuracy shall not be less than half the length of the double-sided confidence interval of the confidence level of Tg.Unin
The average mass of the delivery batch or batch is (tt-d, x+d) (confidence level 7), 4.2.2 Bulk materials
4.2.2.1 Where it is easy to add samples
For the first batch inspection where it is easy to add samples, the procedure for estimating the average mass of the delivery batch or batch is as follows (see Figure 3): Delivery batch or batch
Same as 4.2.1. 1, step a.
Draw mn, increments from the delivery batch or lot, and take m-6 increments.
: group (four increments in one group)
ni sets of samples
ni test samples
r measured values
(15)
Randomly divide the drawn m; increments into a group of m increments, and obtain
. Mix the increments in the same group to obtain
same as step d of 4.2.1.1.
same as step e of 1.2.1.1.
same as step f of 4.2.1.1.
same as step d of 4.2.1.1.
same as step 3 of 4.2.1.1.
GB/T 13732:92
Take mn (n) incremental samples from the delivery batch or batch, and then group, reduce and test these incremental samples as in step cd, and finally obtain observations (including the observations in step d). Recorded as Xx.Xi.
Same as step i of 4.2.1.1.
Same as step k of 4.2.1.1.
Same as step 1 of 4.2.1.1.
Same as step 4.2.1.1. 4.2.2.2 Where it is difficult to add additional samples For the first inspection where additional samples are difficult to add, the procedure for estimating the average quality of the delivery batch or lot is as follows (see Figure 1): Delivery batch or lot
4 incremental samples
1 nest group
8 test specimens
"Individual average value
Calculate the value of the bulk material based on experience or other methods. Take at least n samples from the delivery batch or lot: T=1-+
Determine the position of the sample
(the formula is based on the test sample as a single grain
GE/T 13732-92
S and S can be obtained by the method in Appendix B of this standard. Randomly divide the n samples into 8 groups of equal size, mix the samples in the same group, and get 8 sets of samples. Make a test sample for each of the 8 samples independently, and test each sample once, and get a total of 8 observation values. Record them as X, X, .., X..d. Same as step d of 4.2.1.2.
If E≤2.0, then S, S and S used in formula (16) are considered valid. Otherwise, repeat the above reduction and testing procedures (see Figure 4) on the 8 sets of samples obtained, and get 8 more sets of samples. Observed values, together with the 8 observed values obtained in the previous round, calculate the arithmetic mean of the 16 values and record the average as .
f. Same as step f of 4.2.1.2.
g. Same as step g of 4.2.1.2.
4.3 Non-first batch inspection
4.3.1 Packaged bulk materials
Non-first batch inspection of packaged bulk materials and the procedure for estimating the average quality of the delivery batch or batch are as follows (see Figure 5): 10000 incremental samples
100000 collective samples
100000 complete samples
Four required values
Use historical data to obtain the estimate of the average quality of the delivered batch or batch, S and S, or use the methods in Appendix B and Appendix C of this standard to obtain Sw, S and S.
b Draw 10000 packages from the delivery batch or batch, and then take 1000000 packages from each of the drawn packages (see 6 for the selection of the incremental sample size): The sum should satisfy:
GB/T13732-92
and n is an even number, (18) where N is the batch size in units of subpackaging (18)
c. The samples in each subpackaging are numbered in the order of taking. The odd-numbered samples and the odd-numbered samples are mixed separately. Two samples are obtained. For each sample, 1 sample is independently prepared. Each sample is tested independently to obtain four observations, recorded as: xX, X, t. Two control charts are made to test the stability of the variance. The first control chart tests the stability of the test variance. The center line and lower limit of the control chart are as follows: UCL = 3. 686SM
CL1.1285x
If R,-[X,一X, and 2一IX,一X, are all within the control limits, it means that the test variance is stable, and the second control chart test is carried out: otherwise, it means that the test variance is unstable, and the test method itself should be improved or the number of repeated tests should be increased. The second control chart tests the stability of the product and %. The center line and upper and lower control limits of the control chart are as follows: UC,= 3. 686
Cl, = 1. 128
X+X2X,+X,
fall within the control limits - that is, the packaging variance, the intra-packaging variance and the sub-packaging variance are stable, and the second batch of sales is carried out: otherwise, it means that the inter-packaging variance or (dose) intra-packaging variance or (sum) sub-packaging variance is unstable, and the first batch of inspection procedures in Article 4.2 should be changed. 。 The average mass of the delivery batch or batch is estimated as: X=X+X+X+x
, and its estimated accuracy is
f. Half the length of the double-sided confidence interval of the confidence level is d=T,Uuina
, and the average mass of the delivery batch or batch is (-, +)(confidence level). 4.3.2 Bulk materials
For non-first batch inspection of bulk materials, the procedure for estimating the average quality of the delivery batch or batch is as follows (see Figure 6):.(21)
·(22)
GB/T 13732--92
Five incremental samples
“one collective sample
one certificate
four existing samples
Use the historical data to obtain, estimate the product -SS and S; or use the method in Appendices B and C of this standard to obtain S, S and S
b, take m incremental samples from the delivery batch or batch (the selection of incremental samples is shown in 6). n should meet the following requirements: =-
H specified n is an even number.
-(23）
C. Number the n samples drawn in the order of drawing, mix the odd-numbered samples and the false-numbered samples respectively, and get two sets of samples. For each set of samples, independently prepare a sample, and conduct two independent tests on each sample to get four observation values, which are recorded as X,, X, X, and X,. d.Make two control charts to test the stability of variance. The first control chart tests the stability of test variance. The center line and upper and lower control limits of the control chart are as follows: UCL - 3.686S.
CL—1.1285
If R,= X,一X,! and R,=1X,一x, are both within the control limits, it means that the test variance is stable. Perform the second control chart test. Otherwise, it means that the test variance is unstable. The test method itself should be included or the number of single and multiple tests should be increased. The second control chart tests the stability of the sum. The center line and upper and lower control limits of the control chart are as follows: 1UCL=3.686
CL — 1.128 V3
X+x,X+X
)X+$+sign
falls within the control limit, indicating that the variance between the samples and the variance of the subsamples are stable, and the step e is entered. Otherwise, it indicates that the variance between the samples and (or) the variance of the subsamples are unstable, and the first batch inspection procedure in 4.2 should be used instead. GB/T 13732-92
The average quality of the delivery batch or batch is estimated as: X- X++X, +X++ X
and its estimation accuracy is T.
The half of the length of the bilateral confidence interval of the confidence level is: d=T.Dania
The average quality of the delivery batch or batch is (x-d, x+d confidence level). 5 Sampling acceptance with average value as the indicator
5. 1 Determination of α, β,,,,
Before sampling acceptance, α, β,,,, must be determined in advance. -(26)
This standard stipulates that α is 0.05, α is 0.10, that is, when the delivery batch or batch average quality is equal to or better than the standard, the delivery batch or batch is accepted with a probability of at least 95%; when the delivery batch or batch average quality is equal to or worse than the standard, the delivery batch or batch is accepted with a probability of at most 10. For the case of unilateral upper specification limit, α is used as the upper limit; for the case of unilateral lower specification limit, α is used as the lower limit acceptance quality level, α is used as the lower limit acceptance quality level, α is used as the upper ...
5.2 First Batch Inspection
5.2.1 Packaged Bulk Materials
5. 2. 1. 1 Where Additional Samples Are Easy For the first batch inspection where additional samples are easy to add: the sampling procedure is as follows: a.
Same as step 4. 2. 1. 1.
Same as step 4. 2. 1. 1. b.
Same as step 4. 2. 1. 1. c.
Same as step 4. 2. 1. 1. d.
Same as step 4. 2. 1. 1.
For the one-sided lower specification limit, calculate
For the one-sided upper specification limit, calculate
According to the d\ value, find the number of samples from the table of the required population size n in Appendix A. 11.g.
Same as step 4.2.1.1.
Same as step 4.2.1.1.
Same as step 1i of 4.2.1.1.
For the one-sided lower specification limit, calculate
-(29)
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.