2017-11-172022-04-262022-04-262017-09-15QUEIROZ, L. R. S. Abordagem estocástica de máquinas rotativas utilizando os métodos hipercubo latino e caos polinomial. 2017. 119 f. Dissertação (Mestrado em Modelagem e Otimização) - Universidade Federal de Goiás, Catalão, 2017.http://repositorio.ufcat.edu.br/tede/handle/tede/7974Mechanical systems may suffer with uncertainties that can appear from non-precise data and due the dynamic nature of the problem. Different methods have been used to deal with uncertainty propagation, such as the Latin Hypercube sampling and Polynomial Chaos. Latin hypercube allows to obtain the solution of the random process, from sampling using some probability distribution, over the process domain data. In its turn, the polynomial chaos expansion allows to separate the stochastic components from the deterministic ones of the random solution by using orthogonal polynomials in conformity with the probability distribution of the random variables representing uncertainties. In this work, we apply the Latin hypercube and the polynomial chaos in the quantification of uncertainties. In the beginning some simple mechanical systems were considered, for the purpose to validate the methodology and, then, we studied the effects of uncertainties on a rotor supported by hydrodynamic bearings.application/pdfAcesso AbertoIncertezasHipercubo latinoCaos polinomialMáquinas rotativasUncertaintiesLatin hypercubePolynomial chaosRotating machineENGENHARIASDissertação