In this paper the solution of a composite periodic longitudinal bar is investigated in the framework of the homogenization and stochastic techniques. A two scale asymptotic homogenization technique is used to provide the equivalent mechanical characterization of the bar. Such approach isolates a macro and micro-scale problem. The macro-scale problem describes the dynamics of the bar, while the micro-scale problem supplies the equivalent Young modulus and density. The micro-scale is made of two different materials: the portion of each material in the micro-scale is known in a statistical sense and a probability density function (pdf) is assumed to take into account such uncertainty. Consequently, the macro-scale material properties are random and their probability density functions are calculated in closed form. The random properties of the material represent the coefficients of the governing equation of the bar. The pdf of the solution (eigensolution) is obtained analytically. A Monte-Carlo simulation provides an alternative solution compared to the provided one.
Multi-scale stochastic method to perform a composite longitudinal bar / L., Guj; Culla, Antonio; Sestieri, Aldo. - STAMPA. - 1:(2010), pp. 4889-4902. (Intervento presentato al convegno ISMA2010 International Conference on Noise and Vibration Engineering including USD2010 tenutosi a LEUVEN nel 20-22 SETTEMBRE 2010).
Multi-scale stochastic method to perform a composite longitudinal bar
CULLA, Antonio;SESTIERI, Aldo
2010
Abstract
In this paper the solution of a composite periodic longitudinal bar is investigated in the framework of the homogenization and stochastic techniques. A two scale asymptotic homogenization technique is used to provide the equivalent mechanical characterization of the bar. Such approach isolates a macro and micro-scale problem. The macro-scale problem describes the dynamics of the bar, while the micro-scale problem supplies the equivalent Young modulus and density. The micro-scale is made of two different materials: the portion of each material in the micro-scale is known in a statistical sense and a probability density function (pdf) is assumed to take into account such uncertainty. Consequently, the macro-scale material properties are random and their probability density functions are calculated in closed form. The random properties of the material represent the coefficients of the governing equation of the bar. The pdf of the solution (eigensolution) is obtained analytically. A Monte-Carlo simulation provides an alternative solution compared to the provided one.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.