Posted in Operations and Supply Chain Terms, Total Reads: 1565
Definition: Variable Sampling
Variable sampling is used for measurement of hardness , tensile strength , resistance and other physical properties of the desired sample. The variables used for measurement are on a continuous scale and the limits for rejection are clearly demarcated for the entire range. The sample which is tested from the complete lot is said to be its representative.
It is a much more descriptive of the type of defect and the gives the measure of deviation from the accepted standard. The reason for declaring the product to be defective can easily be inferred from the result . It is preferred to an attribute sampling where the product is categorized as ok or defective . The attribute sampling does not indicate the exact reason for failure of a sample. It usually takes a large sample size so that there is at-least a couple of defective pieces and the failure rate can be expressed as a percentage. It is considered to be a wasteful activity because after 100% testing we end up classifying them as tested ok or faulty.
The sample size selected for testing is between 4-25 . The sample size is kept low because a homogenous quality is ensured during designing and unnecessary expenditure is not done on testing. All the indicative parameters have to be recorded for the selected samples which makes it all the more time consuming. In the sample set for variable sampling the original values are taken as reference values and the measured values are compared with them and the acceptance limits is decided by the confidence interval restricted by the normal distribution curve and all the values lying outside them are rejected . The curve is divided into three zones .Zone 1 and 3 are short zones and Zone 2 is known as long zone . The upper limits and lower limits of deviation are kept as +3 sigma and -3 sigma respectively . Hence it not only saves a lot of inspection cost it also keeps a check on controlling all the risk levels .