Abstract

A general frame work is developed for the sensitivity analysis and design optimization of smart laminated composite beams with capability to suppress the vibration under random excitations. The smart structure consists of a host laminated composite beam with embedded/surface bonded piezoelectric sensors/actuators. A layerwise displacement model including the electro-mechanical coupling is utilized to account for the strong inhomogeneities through the thickness and to develop the finite element model. To perform the sensitivity analysis of the smart structure for different design parameters, analytical gradients based on developed layerwise finite element model for both static and dynamic problems are proposed. The developed sensitivity gradients provide an efficient way to predict the behavior of responses of smart structure without re-analysis. A design optimization algorithm based on developed analytical gradients and layerwise finite element model is then developed to determine the optimal design of the smart laminated beams for a variety of objective and constraints functions, including interlaminar stresses, weight and natural frequencies. The smart laminated beam design based on the developed optimization algorithm is used in the dynamic analysis and vibration control. An optimal control strategy, Linear Quadratic Regulator (LQR) is used to design the feedback control gain and to control the vibration response of system under deterministic loads. Effects of laminate configuration and sensor/actuator location are investigated in controlled response. An in-house experimental set-up is designed to demonstrate the performance and functionality of the proof-of-concept of smart composite beam. In many practical applications, including aerospace and automotive industries, smart laminated structures are exposed to random loading, considering this, an optimal control algorithm is developed to suppress the vibration response of the smart beam under random excitations. Different types of random loadings, including Gaussian white-noise, band limited and narrow-band excitations are investigated.