IMPROVING THE ACCURACY OF COMPUTATIONAL METHODS FOR ANALYZING THE STABILITY OF STIFFENED SPHERICAL SEGMENTS USING EXPERIMENTAL RESULTS

Abstract

A correction function is introduced to improve the agreement between the theoretical and experimental data of representative sampling. The powers of the parameters of the function are determined based on the results of sequential minimization of standard deviation of a random value expressed as a ratio of the experimental critical pressure to the theoretical one. The standard deviations of a random value of a sampling for two design formulae, refined using the related correction functions, are shown to significantly decrease and reach practically the same values. A design formula comprising the product of raised design parameters is obtained that provides a substantial acceleration in synthesizing design variants of stiffened spherical segments. Several possible designs are remonstrated for relatively reducing the mass of the shell without affecting its reliability when using the design formula involving a minimum standard deviation value.

Key words: improving the accuracy of computational methods, stability, correction function, stiffened spherical segments.