Contribuição ao uso de metamodelos de expansão em polinômios de caos em problemas de confiabilidade estrutural

Abstract

Advances in the field of Structural Engineering have provided leverage to anthropic activity in the most diverse scenarios, ranging from the extraction of oil and gas in ultra-deep wells to aerospace exploration. This progress is underpined by the application of methodologies capable of emulate loads on structures, the behavior of its materials, its operating conditions, among others. However, many of these artifices are computationally costly and may become prohibitive, as the complexity of the analysis is increased. Simulations involving reliability that takes into account the uncertainties of the variables of the problem and proposes to estimate the probability of failure of structures, are examples of applications that require thorough evaluation of their limit state equations. This way, there is a tendency to incorporate metamodeling to engineering problems, so that analyses are carried out on simpler and faster models that still satisfactorily represent the original problem. This work aims to evaluate the use of metamodels obtained by Polynomial Chaos Expansion (PCE) in structural reliability applications. Computational implementations are made in Python. The results obtained are compared and validated with benchmark problems that are available in the literature related to the theme.