For various inspections of textiles, in fact, most of them are carried out in the form of testing textiles sampling.
what is sampling in textile testing?
Under normal circumstances, the totality of the tested objects is always relatively large, and most of them are destructive, and it is impossible to test all of them. Therefore, sub-samples are usually taken from the totality of the tested objects for quality inspection.
The extent to which the results of the textile sampling test can represent the characteristics of the overall test object depends on the size of the sub-sample and the sampling method. In textile products, there are more or less quality differences between the unit products within the population. The larger the sample size, that is, the more individuals contained in the sample, the closer the measured result is to the overall result (true value). The amount of sample required to achieve the required credibility of the test results can be determined by statistical methods. But no matter how large the sample is taken, how accurate the instrument is, if the sampling method itself lacks representativeness, the test results are also unreliable. To ensure the representativeness of the sample to the population, it is necessary to adopt a reasonable sampling method, that is, to avoid systematic errors in sampling as much as possible, not only to eliminate tendency sampling, but also to minimize random errors. For this reason, random sampling methods should be used.
Specifically, there are mainly four textile sampling methods as follows.
1. Pure random sampling
Take several samples (sub-samples) from the population so that each unit product in the population has the same chance of being sampled. This kind of sampling is called pure random sampling, or simple random sampling. Pure random sampling does not go through any grouping and queuing of the population, and draws from it entirely by chance. Theoretically speaking, pure random sampling best conforms to the random principle of sampling, so it is the basic form of sampling.
Although pure random sampling is in theory the most consistent with the principle of randomness, in practice it has great contingency, especially when the overall variation is large, the representativeness of pure random sampling is not as strong as that of grouped re-sampling.
2. Isometric sampling
Equidistant sampling is to first queue the population according to a certain mark, and then sample at equal distances.
Compared with pure random sampling, equidistant sampling can make the sub-samples more evenly distributed in the whole, and can make the sub-samples have better representativeness. However, if the product quality fluctuates regularly and coincides with equidistant sampling, systematic errors will occur.
3. Representative sampling
Representative sampling is to use statistical grouping method to divide the population into several representative type groups, and then use pure random sampling or equidistant sampling within the group to sample from each group separately, and then combine the sub-samples of each part into one sub-sample . In representative sampling, the number of samples in each group can be determined by the following method: determine the degree of variation within each group, take a little more for the group with large variation, take less for the group with small variation, and there is no uniform proportion; or The proportion of the total to determine the number of groups should be taken.
4. Periodic random sampling
Phased random sampling is to take a part of the sub-sample from the whole, and then take a sample from this part of the sub-sample. Obtaining samples from a batch of goods can be divided into three stages: batch samples, samples, and samples.
(1) Batch sample: Obtain a certain number of packages (or boxes) from the entire batch of goods to be inspected.
(2) Sample: Use appropriate methods to reduce the batch sample to a sample for the laboratory
(3) Sample: From the samples in the laboratory, obtain samples for testing of various physical and mechanical properties and chemical properties according to a certain method.
Summary: When testing textiles, a large number of samples or laboratory samples must be taken first, and then made into samples.
If you want to know more about
textile testing related issues, please leave us a message: email:info@qinsun-lab.com