Size Effect of a Piezoelectric Patch on a Rectangular Plate with the Neural Network Model

Area of Science:

• Engineering
• Computational Mechanics
• Materials Science

Background:

• Artificial neural networks (ANNs) are increasingly utilized in engineering for predicting complex physical phenomena.
• Understanding the piezoelectric effect is crucial for vibration and noise reduction in structures.
• Previous studies have explored ANNs for various engineering predictions, but specific applications to piezoelectric patch effects on plate dynamics require further investigation.

Purpose of the Study:

• To develop and validate an artificial neural network (ANN) model for predicting the first-order natural frequency and displacement amplitude of a plate.
• To investigate the influence of piezoelectric patch size (length, width) and thickness on the plate's dynamic response.
• To assess the accuracy of the ANN model with varying input parameters, including detailed dimensions versus simplified representations.

Main Methods:

• Finite Element Method (FEM) simulations were performed to analyze a rectangular plate actuated by a piezoelectric patch under various patch sizes.
• The FEM-generated data was used to construct and train an artificial neural network (ANN) model.
• The ANN model's predictions for natural frequency and displacement amplitude were validated against FEM results.

Main Results:

• The ANN model accurately predicted the plate's natural frequency and displacement amplitude when provided with patch length, width, and thickness as inputs.
• A significant decrease in prediction accuracy was observed when the ANN model's inputs were simplified to patch size and thickness or patch volume.
• The study revealed a coupled influence of patch size and thickness on the first-order natural frequency, with the ANN model successfully predicting maximum and minimum values.

Conclusions:

• Detailed piezoelectric patch dimensions (length, width, thickness) are essential for accurate dynamic response prediction using artificial neural networks.
• Simplified input parameters lead to diminished accuracy in ANN models for analyzing plate vibrations.
• The findings underscore the capability of ANNs in modeling coupled phenomena in structural dynamics, specifically the interplay between piezoelectric patch geometry and plate vibrational characteristics.