Actuator and sensor design and modeling for structural acoustic control

Abstract

The use of a high-fidelity finite element model is investigated for the design and closed loop performance prediction of shaped and distributed sensors and actuators for structural acoustic control. Sensor and actuator design was found to be sensitive to nodeline discrepancies between the model and experiment caused by moderate manufacturing defects and/or boundary condition uncertainties. Relying on the finite element model for sensor shaping or distribution results in a slight difference between the desired and achieved sensor performance. The modeshape sensitivity is compounded when both the actuator and sensor are shaped or distributed, as is the case with a distributed sensuator design. This results in an unacceptable difference between the desired and achieved distributed sensuator performance. Since the advantages of shaping and distribution can be gained either at the system input (actuators) or output (sensors), modeshape information from a correlated analytic model should be used for one or the other, but not both. Experimental verification of critical modeshapes is also recommended to reduce sensor and actuator performance loss. The finite element model was also used to predict achievable closed-loop acoustic performance for various sensor and actuator pairs for transmission and reflection control. Since a finite element model is generally not accurate enough to be used as the basis for high performance compensator design, the predicted performance was compared to experimental results of compensators designed with an accurate data-fit model using the same control design methods. Good correlation was achieved between predicted and implemented results for linear behavior of the system. Finally, a comparison was made between a modally shaped PVDF sensor/PZT actuator design and a single wafer PZT sensuator. Both have desirable open-loop characteristics and comparable predicted performance. The predicted performance could not be implemented for the sensuator design due to a severe amplitude non-linearity. The PVDF sensor design is very linear, and the implemented performance slightly exceeded that predicted using the finite element model. Due to the implementation difficulties of the sensuator, the PVDF sensor/PZT actuator design is the better choice for acoustic transmission control.