Constrained optimization of a meander line EMAT generating S0 mode Lamb waves with genetic algorithms

Wang, Shen; Huang, Songling; Wang, Qing; Wang, Zhe; Zhao, Wei

Abstract

In this work a multiphysics model of a meander line electromagnetic acoustic transducer (EMAT) to generate S0 mode Lamb waves in an aluminum plate is built and the structural parameters of the transducer are optimized based on the model. Three approaches to solve the amplitude of the in-plane displacement component at an observation point are explored and the single frequency approach is selected because it is the fastest. In the optimization, the objective function to minimize is the negative amplitude of the S0 mode waves, and the design variables are the width and height of the magnet and liftoff values of the magnet and the coil from the top surface of the plate. The liftoff values form a linear constraint of the optimization problem. A genetic algorithm (GA) program capable of handling the linear constraint efficiently is developed. The internal status of the GA program is tracked carefully to avoid unnecessary evaluations of the objective function. With 30 runs of the program, the optimal set of variables leads to the ratio of the width of the magnet to that of the coil of 115.85%.

Citation

Wang, S., Huang, S., Wang, Q., Wang, Z., & Zhao, W. (2018). Constrained optimization of a meander line EMAT generating S0 mode Lamb waves with genetic algorithms. International Journal of Applied Electromagnetics and Mechanics, 58(4), 411-429. https://doi.org/10.3233/jae-180086

• Journal Article Type: Article
• Acceptance Date: Nov 10, 2018
• Online Publication Date: Dec 10, 2018
• Publication Date: Dec 10, 2018
• Deposit Date: Jan 4, 2019
• Publicly Available Date: Jan 16, 2019
• Journal: International Journal of Applied Electromagnetics and Mechanics
• Print ISSN: 1383-5416
• Electronic ISSN: 1875-8800
• Publisher: IOS Press
• Peer Reviewed: Peer Reviewed
• Volume: 58
• Issue: 4
• Pages: 411-429
• DOI: https://doi.org/10.3233/jae-180086
• Public URL: https://durham-repository.worktribe.com/output/1310603