Abstract

Adhesive bonded structures are gaining attention in engineering and research communities due to their advantages over conventional joining methods. Non-destructive testing and health monitoring of adhesive bonded structures are challenges requiring focused research. Piezoelectric transducers are used for the actuation and sensing purposes in structural health monitoring procedures. These transducers which are adhesive bonded, get disbonds from the host structure during their service period. Presence of a transducer disbond between the transducer and host structure can be inferred as structural disbond and may produce false alarms. It is necessary that both the types of disbonds are distinguished from each other so that an integrated health monitoring procedure can be developed. This paper presents the use of electromechanical admittance technique for the integrated health monitoring of adhesive bonded beams using surface bonded piezoelectric patches. Electromechanical admittance model for one degree of freedom system is revisited and used as a governing model for the adhesive bonded beams. The analytical results are validated with simulations and experimental results. Conventional non-destructive techniques like X-ray and ultrasonics testing are also employed to justify the use of the electromechanical admittance scheme for disbond detection in the adhesive bonded structures. The electromechanical admittance values (both real and imaginary parts) for three levels of transducer and structural disbonds along with the combination cases are collected from the precision impedance analyzer in a frequency range of 1–30 kHz. Numerical study of coupled-domain harmonic analysis is utilized to study the disbond cases. It is shown that the directional shifting of the electromechanical admittance spectrum distinguishes both the types of disbonds. In addition, artificial neural networks are also employed on electromechanical admittance data from simulations and experiments to predict disbond type and the severity levels.

Full Text
Paper version not known

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call