Reliability Based Design Optimization for Selective Excitation of the Vibration Modes of a Cantilever Spring

Abstract

This paper is devoted to procedures for the reliability-based optimization methods of engineering structures combining measurement and sensitivity technique, for the purpose of the better sensitivity in force-gradient detection. In the experiment part of this study, the mica muscovite cantilever beam clamped-free is used. The excitation of a cantilever beam with several small sheets of piezoelectric polymer adequately glued to it selects one high-frequency vibration mode of the cantilever. The proposed strategy is design into a framework that allows the solution of optimization problems involving a several number of design parameters that characterizes the systems, including dimensional tolerance, material properties, boundary conditions, loads, and model predictions, considered to be uncertainties or variables. The proposed methodology directly supports quality engineering aspects enabling to specify the manufacturing tolerances normally required to achieve desired product reliability. Within this context, the robust design obtained is optimal over the range of variable conditions because it considers uncertainties during the optimization process. The large number of exact evaluations of problem, combined with the typically high dimensions of FE models of industrial structures, makes reliability-based optimization procedures very costly, sometimes unfeasible. Those difficulties motivate the study reported in this paper, in which a strategy is proposed consisting in the use of reliability-based optimization strategy combined with measurement and sensitivity technique specially adapted to the structures of industrial interested.