Damping tiles are utilized aboard submarines to reduce their submerged acoustic signature by controlling interior vibration. Some structural examples include: ballast tanks, propulsion systems, machinery frames, bulkheads, decking and other surfaces. Current tile technology employs viscoelastic constraining layer damping due to its simplicity and robust nature. This research assessed the performance of piezoelectric shunted tiles using commercially available materials and components. The intended application was a replacement for Class I damping tiles. High modulus values and the inherent brittleness of piezoelectric crystal narrow its potential use. However, test data illustrates that under certain applications, piezo tiles show improved damping performance over a broader temperature range with a total reduced weight than viscoelastics.
Historically the use of piezoelectric materials in the passive mode has been relegated to sensors only. Changes in strain re-distribute charges in the material where they are collected at external electrodes. Time variations in strain however, generate time variations in charge. This electrical current can be dissipated using an external shunt resistor, resulting in mechanical loss. Piezoelectric crystals such as PZT are typically stiff and strain very little before fracture. These materials can be utilized effectively in applications where mechanical impedances and strain magnitudes are in parity, thereby optimizing electromechanical coupling. To quantify damping effectiveness, we measured the modal loss factor of a simple beam utilizing discrete piezoelectric tiles and an external shunting circuit. The beam geometry and test method were in substantial conformance to MIL-PRF-23653C, used for Navy damping tiles.
We developed a model of piezoelectric damping sufficient to characterize the phenomena for our test case. Assuming negligible loss in the native beam, the total system damping is a function of the respective material moduli, piezoelectric strain constant, piezo capacitance, and shunting resistance. While piezoelectric strain constants in the longitudinal (\textit{d33}) direction are typically the highest, exploiting this fact is problematic and most commercial materials are available with electrodes in the transverse (\textit{d31}) direction. Based on the model and available material data, we selected a series of the most promising materials.
Three steel beams were fabricated; one for each piezoelectric material. Since steel is electrically conductive, the beams were used as the common ground for the piezoelectric tiles. The tiles were fastened to the beam using an electrically conductive epoxy bond. Optimal mechanical coupling requires a stiff, thin bond layer free of voids. A thick bond layer would increase the distance between the piezo tile and the neutral axis of the beam, reducing mechanical coupling and therefore damping effectiveness. Therefore, careful bonding techniques were established to ensure the thinnest possible bond layer.
Test beams were suspended at the endpoints to simulate free-free boundary conditions while an electrodynamic shaker provided broadband excitation. Shunting resistances were localized on an external printed circuit board. Using a modal curve fit algorithm, we extracted loss factors for the first and second bending modes. Test results differed moderately from predicted loss. Non negligible damping due to the existence of the piezo tiles, wiring and epoxy could explain the disparity. Furthermore, piezoelectric suppliers admitted to a general lack of accurate strain coefficient values making accurate modeling difficult to achieve.
While shunted tiles may be less effective than some constraining layer viscoelastic technologies, the future use of passive peizoelectric materials shows promise. The strongest case for their use would be an application that requires either a high modulus damping material or more temperature stability than is afforded by viscoelastic materials.