Bayesian Calibration and Physics-Informed Models for Enhanced Piezoelectric Energy Harvesting on Fan-Folded Structures.

• Grant number: 23/15931-1
• Support Opportunities: Scholarships abroad - Research Internship - Scientific Initiation
• Start date: April 08, 2024
• End date: July 25, 2024
• Field of knowledge: Engineering - Mechanical Engineering - Mechanics of Solids
• Principal Investigator: Paulo Sergio Varoto
• Grantee: Catarina Luz Moura Ibiapina Barros
• Supervisor: Rafael de Oliveira Teloli
• Host Institution: Escola de Engenharia de São Carlos (EESC). Universidade de São Paulo (USP). São Carlos , SP, Brazil
• Institution abroad: Institut Franche-Comté Electronique Mécanique Thermique Et Optique - Sciences Et Technologies, France
• Associated to the scholarship: 22/09350-3 - Design and Optimisation of Piezoelectric Energy Harvesting Using Evolutionary Methods, BP.IC

Abstract

The current research project aims to primarily apply Bayesian calibration and physics-informed models in the design of piezoelectric energy harvesters with multiple degrees of freedom and fan-folded geometry. The study falls within the domain of smart structures utilizing piezoelectric materials, which can be employed for both structural control techniques and sensing for dynamic signal monitoring. The project's scope aligns with the theme of Vibration Piezoelectric Energy Harvesting, intending to devise electromechanical devices utilizing piezoelectric materials capable of converting structural vibration signals from a given dynamic environment into useful electrical voltage. This is particularly relevant in scenarios where the main spectral components of vibrational signals lie in the frequency range of 0-100 Hz, which commonly encompasses the majority of vibrational phenomena termed environmental.Specifically, Bayesian calibration of the fan-folded model will be conducted using experimental results from tests on components at the Dynamics Laboratory of EESC-USP, representing fan-foldeds with 2 to 5 beams and thicknesses of 1 and 2 mm. Subsequently, the calibrated model will be employed to generate data for training a metamodel, also informed by the physics of the system. This metamodel will serve as a surrogate model to reduce the high computational cost that would be required if the numerical model were used as the optimization fitness function.Additionally, to achieve practical application for the study, there is a proposal to test a structure subjected to vibrations, with energy generation through the use of fan-foldeds.