Reliability analysis of smart laminated composite plates under static loads using artificial neural networks

Area of Science:

• Materials Science
• Mechanical Engineering
• Reliability Engineering

Background:

• Smart structures with piezoelectric elements are increasingly used for sensing and actuation.
• Manufacturing uncertainties in geometry and material properties can compromise structural reliability.
• Reliability analysis is crucial for ensuring the safe and effective performance of these advanced materials.

Purpose of the Study:

• To perform a reliability analysis of a smart laminated composite plate with a piezoelectric fiber-reinforced composite (PFRC) actuator layer.
• To investigate the impact of uncertainties in material and geometric properties on stresses and displacements.
• To determine the most significant parameters affecting structural reliability under static electrical and mechanical loads.

Main Methods:

• Developed a coupled finite element (FE) model in COMSOL Multiphysics.
• Created and trained an artificial neural network (ANN) model using FE data to handle parameter uncertainties.
• Employed Monte Carlo Simulation (MCS) and First- and Second-Order Reliability Methods (FORM/SORM) for reliability assessment.

Main Results:

• The piezoelectric stress coefficient was identified as the most critical factor influencing nondimensional stresses and displacements.
• Parameter variation and resulting uncertainty increase under applied electrical load.
• A threshold of approximately 3% input parameter variation was found, beyond which failure probability rapidly and non-linearly increases.

Conclusions:

• Reliability analysis is essential for smart composite structures, particularly considering piezoelectric properties.
• Electrical loads exacerbate the impact of parameter uncertainties on structural performance.
• The study provides a framework for predicting failure probability in PFRC-integrated smart structures, highlighting critical design thresholds.