Seminar Title
Performance enhancement of functionally graded mems accelerometer for condition monitoring using artificial neural network
Seminar Type
Departmental Seminar
Department
Mechanical Engineering
Speaker Name
Uttam Kumar Kar (Roll No:- 519me1017)
Speaker Type
Student
Venue
Seminar Hall, Mechanical Engineering Department
Date & Time
26 May 2023 11:00 AM
Contact
Dr R K Behera, 06612504
Abstract
The functionally graded micro electro mechanical system (FG-MEMS) sensors are widely used in various industrial applications due to its low power consumption, high sensitivity and high resolution. Recently, these sensors are being used in monitoring the condition of the machinery, dynamic certification of any intended structural components, fault prediction, structural ageing issues, and various other structural dynamics research and diagnostics. The design of such devices is challenging for engineers, since its performance depend on many parameters. To solve this problem, the present work investigated the optimal design
parameters of FG-MEMS sensor to improve their performance, resolution and reliability in condition monitoring. In order to obtain the effective material and geometrical parameters of FG-MEMS sensor, a surrogate optimization methodology with artificial neural network (ANN) function estimation model is proposed. The material properties are graded using generalized power law model and modified couple stress theory is employed to capture the size effects. Finally, a case study of beam crack detection using FG-MEMS accelerometer is demonstrated.