Finite Element Analysis of Bondline Effects on Multilayer Ultrasonic Transducers

Wu, Z., Mcrobbie, G., Cochran, S. , Knox, E. and Gallagher, S. (2005) Finite Element Analysis of Bondline Effects on Multilayer Ultrasonic Transducers. In: NAFEMS World Congress 2005: Engineering Simulation - Best Practices and Visions of the Future, St Julians, Malta, 17-20 May 2005,

Full text not currently available from Enlighten.

Abstract

The use of multiple active layers within an ultrasonic transducer permits the designer an additional degree of freedom compared with single active layer configurations. Amongst other advantages, this can be exploited to reduce the electrical impedance of the transducer or to increase the mechanical output for a given input voltage. However, unless a co-fired piezoceramic configuration is used, multiple active layers must be bonded together with adhesives and the bondlines will detract from the performance of the device. Hence, detailed studies of bondline behaviour and effects are of importance within this area of research. This paper reports one such study. The practical configuration under investigation comprised two layers of lead metaniobate piezoelectric ceramic bonded together with a hard-setting epoxy potting compound. This configuration was chosen rather than the more modern use of multilayer piezocomposite material because it is relatively easy to obtain high-quality bondlines between continuous ceramic and lead metaniobate suffers relatively little from interfering width-vibrational modes. Two basic finite element analysis model configurations were set up. The first included a uniform thickness bondline, with the actual thickness dimension set to 10.36 m in accordance with a fitting procedure described elsewhere. The second configuration included a bondline designed to match a measured surface profile. The measurement provided data with a 9.06 m peak-to-peak excursion sampled at 800 points/mm. For the purposes of modelling, these data were processed to provide a peak-to-peak excursion of 8 m sampled at 100 points/mm. Both models were run for three different bondline thicknesses, with the surface profile scaled to match the total thickness, generating electrical impedance data and surface displacements. Excellent agreement was found between both sets of modelled data and experimental measurements when the fitted thickness dimension was used. This provides an initial indication that surface roughness of levels similar to the total bondline thickness does not need to be taken into account in FEA modelling. However, when the bondline thicknesses and surface roughnesses were increased, it was found that significantly smaller surface displacements were obtained with the uniform bondline thickness model. The reasons for this require further investigation, but in the interim, the results indicate that care must be taken to ensure thin bondlines in multilayer piezoelectric configurations.

Item Type:Conference Proceedings
Keywords:Simulation.
Status:Published
Refereed:Yes
Glasgow Author(s) Enlighten ID:Cochran, Professor Sandy and Knox, Dr Esther
Authors: Wu, Z., Mcrobbie, G., Cochran, S., Knox, E., and Gallagher, S.
College/School:College of Science and Engineering > School of Engineering > Systems Power and Energy

University Staff: Request a correction | Enlighten Editors: Update this record