Demonstration of Prognostics Health Monitoring (PHM) in Adhesive Lap Joints using Simulated Studies

Abstract

Adhesively bonded joints are increasingly used in structural applications due to many advantages over classical mechanical fasteners. However, they are susceptible to fatigue damage due to the hostile working environment. It is essential to detect, quantify the damage and estimate the remaining useful life (RUL) under fatigue loading. This paper presents prognostics health monitoring (PHM) framework that can quantify the damage and estimate the RUL in adhesive lap joints. A PHM framework is the synthesis of four disciplines: damage diagnostics, predictive modeling, uncertainty quantification, and uncertainty propagation. Damage diagnostics provide the damage evolution in terms of damage growth rate as input to PHM system. In this work, a new diagnostic method based on ultrasonic Lamb waves is proposed for in-situ measurements of crack length in a single lap joint (SLJ). The idea is to excite single mode using two piezo transducers and extract the wave packet reflected from the crack tip to estimate crack length. The proposed method is verified using computational simulations. A predictive model must be capable of simulating the damage growth physics and can govern the growth rate using model parameters. In this study, the cohesive zone model (CZM) is used to simulate crack growth in SLJ. Uncertainty quantification methods require the evaluation of the predictive model for large parameter sets. To achieve this, setup is built using Python script to run crack propagation simulations in ABAQUS. Convergence issues and computational challenges associated with the setup is addressed. Finally, the procedure to estimate RUL using diagnostics data, predictive model and uncertainty quantification methods is discussed.