This article presents an optimization method to determine the arc constants used in the mathematical model which concerns the critical flashover voltage of a polluted insulator. These constants must be standard for many insulators, hence the establishment of a model that very accurately simulates the experimental results. The optimization method based on Particle Swarm Optimization (PSO) that resolves a problem by iteratively trying to improve a candidate solution with regard to a given measure of validity. It can be seen that the calculated results by application of PSO has allowed to define the constants of the arc where the establishment of a model that simulates the experimental and analytical results of other researchers.

This paper aims to investigate and analyze the effect of uncertainties and fluctuations on the structural reliability of filament wound cylindrical composite gas storage tank using Monte Carlo Simulation method (MCS). A performance function using Tsai-Wu failure criterion, with seven random variables, has been developed wherein the random variables are the three elastic constants of the material (longitudinal modulus E1, the transverse modulus E2 and shear modulus G12), the thickness of the laminate, the radius of the tank, the winding angle and the internal pressure. The safety margin distributions in terms of the coefficients of variation (COVs) of the composite gas storage tank are obtained. The results show that the internal pressure and the winding angle are the main parameters that affect the structural reliability of the tank and the probability of failure increases especially when all the parameters are treated as random. Furthermore, high values of coefficients of variation cause the shrink of the safety margin and can induce the failure of the axisymmetric structure.