In this paper, a new tuning method for shunt damping with a series resistance, inductance and negative capacitance is proposed and its validity is investigated. It is based on the measured electromechanical impedance of a piezoelectric system, which is represented through an equivalent electrical circuit that takes into account the characteristics of the piezoelectric transducer and the host structure. Afterwards, an additional circuit representing the shunt is connected and the Norton equivalent impedance is obtained at the terminals that represent the mechanical mode of interest. During the tuning process, the optimal shunt parameters are found by minimizing the maximum absolute value of the Norton equivalent impedance over a defined frequency range through a numerical optimization. Taking benefit from the analogy between electrical impedance and mechanical admittance, the minimization of different mechanical responses (displacement, velocity or acceleration) is also proposed and the different optimum shunt parameters obtained are compared. In view of real technical applications, this method allows the integration of a real negative capacitance circuit, i.e., a negative impedance converter, rather than an ideal component. It is thus possible to use the impedance of this circuit and optimize the individual component values. Since this method is based on one simple measurement, it can be applied to arbitrary structures without the need of complex dynamic tests or expensive finite elements calculations. Finally, an experimental analysis is carried out in order to compare the damping performance of the proposed method and the conventional analytical method that minimizes a mechanical frequency response function.
Tuned vibration absorbers have become common for passive vibration reduction in many industrial applications. Lightly damped absorbers (also called neutralizers) can be used to suppress narrowband disturbances by tuning them to the excitation frequency. If the resonance is adapted in-operation, the performance of those devices can be significantly enhanced, or inertial mass can be decreased. However, the integration of actuators, sensors and control electronics into the system raises new design challenges. In this work, the development of adaptive-passive systems for vibration reduction at an industrial scale is presented. As an example, vibration reduction of a ship engine was studied in a full scale test. Simulations were used to study the feasibility and evaluate the system concept at an early stage. Several ways to adjust the resonance of the neutralizer were evaluated, including piezoelectric actuation and common mechatronic drives. Prototypes were implemented and tested. Since vibration absorbers suffer from high dynamic loads, reliability tests were used to assess the long-term behavior under operational conditions and to improve the components. It was proved that the adaptive systems are capable to withstand the mechanical loads in an industrial application. Also a control strategy had to be implemented in order to track the excitation frequency. The most mature concepts were integrated into the full scale test. An imbalance exciter was used to simulate the engine vibrations at a realistic level experimentally. The neutralizers were tested at varying excitation frequencies to evaluate the tracking capabilities of the control system. It was proved that a significant vibration reduction is possible.
To solve a wide range of vibration problems with the active structures technology, different simulation approaches for several models are needed. The selection of an appropriate modeling strategy is depending, amongst others, on the frequency range, the modal density and the control target. An active system consists of several components: the mechanical structure, at least one sensor and actuator, signal conditioning electronics and the controller. For each individual part of the active system the simulation approaches can be different. To integrate the several modeling approaches into an active system simulation and to ensure a highly efficient and accurate calculation, all sub models must harmonize. For this purpose, structural models considered in this article are modal state-space formulations for the lower frequency range and transfer function based models for the higher frequency range. The modal state-space formulations are derived from finite element models and/or experimental modal analyses. Consequently, the structure models which are based on transfer functions are directly derived from measurements. The transfer functions are identified with the Steiglitz-McBride iteration method. To convert them from the z-domain to the s-domain a least squares solution is implemented. An analytical approach is used to derive models of active interfaces. These models are transferred into impedance formulations. To couple mechanical and electrical sub-systems with the active materials, the concept of impedance modeling was successfully tested. The impedance models are enhanced by adapting them to adequate measurements. The controller design strongly depends on the frequency range and the number of modes to be controlled. To control systems with a small number of modes, techniques such as active damping or independent modal space control may be used, whereas in the case of systems with a large number of modes or with modes that are not well separated, other control concepts (e.g. adaptive controllers) are more convenient. If other elements (e.g. signal amplifiers or filters) in the signal paths have a significant influence on the transfer functions, they must be modeled as well by an adequate transfer function model. All the different models described above are implemented into one typical active system simulation. Afterwards, experiments will be performed to verify the simulations.