Target Method for the Statistical Processing of the Results of Tensile Testing of Defective Pipes

We develop statistical methods for the comparative analysis of the accuracy of relations used for the determination of the residual strength of plastic pipes with rectangular surface defects. In the existing approaches, the accuracy of relations is established for the entire set of experiments as a single generalized value. At the same time, in the present work, the corresponding values are established separately for every fixed collection of the sizes of defects (points of interest). For each test, we determine its weight, which is not larger than one at every point of interest. The statistical analysis of accuracy is performed separately for each point of interest, which enables us to smooth out the experimental errors and determine the error of the formula itself more precisely.