A weighted averaging method for tensile data on vulcanizate, has been dopted in JIS. This is based on the fact that the distribution of tensile data is considered to be a kind of doubly exponential theory by Kase. But the author thinks the median method agree with the results of the tests than above mentioned one by the improvement of testing process. The author has proved inductively the truth of normal type of the frequency-distribution by many experiments and moreover, deductively from the established facts of many other substances; i.e.
(1) The main factor of the rupture of specimens is not always attributed to the largest flaw or the smallest cross-section in the parallel part. So that the tensile strength should be calculated by dividing the breaking force by the median cross sectional area as the representative value.
(2) In the case of bending the bench marks at break, we must measure the mean distance, excepting the maximum between the marks.
Hereby, if we adopt median value, the distribution of the data may be nearly normal. So that the calculation is very simple and convenient on the statistical analysis of the data. As the result, it is more efficient and moreover, the value is closer to the true value.
View full abstract