STOCHASTIC ANALYSIS OF NONLINEAR MECHANICAL SYSTEM
Tatiane Nunes da Costa1; Alcione Borges Purcina1; Romes Antonio Borges1
1 Universidade Federal de Goiás - Regional Catalão
doi:10.20906/CPS/COB-2015-2423
Resumo
The majority of engineering problems and other areas are subject to certain types of uncertainties, and these affect the responses. The study about the stochastic methods have gained featured in recent years, with the intention of obtain answers that are more reliable. Among the existing techniques, to consider the uncertainties on problems, stand out the Monte Carlo method and Latin Hypercube method. It is worth mentioning that, the Latin Hypercube method is a variant of the Monte Carlo method and shows itself efficient for obtaining solution, even with a large number of random variables. Thus, this method has the advantage of reducing the required number of simulations to the obtainment of the results, having a smaller computational cost, when compared with the Monte Carlo. For this reason, this work has for objective the application of the Latin Hypercube method to a system with Nonlinear Dynamic Vibration Absorber and analyze the results. For this, initially will be described about the functioning of Latin Hypercube method. The following, is modeled a system of two degrees of freedom and presents a study of their behavior through the application of Fourth Order Runge-Kutta Method. Finally, with the application of the Latin Hypercube method, value ranges (Minimum, Medium and Maximum) of problem are obtained and analyzed, highlighting that, as closer is the minimum range of the maximum, the better the result.
Palavras-chave: Uncertainty; Monte Carlo Method ; Latin Hypercube Method; Mechanical Systems