Aero structure damage growth prediction using age-based state transition models. (April 2021)

• Record Type: Journal Article
• Authors: Yousuf, Waleed Bin; Khan, Tariq Mairaj Rasool; Shah, Aqueel
• Is Part Of: Engineering failure analysis. Volume 122(2021)
• Publication Date: 2021-04
• Subjects: Prediction algorithms -- Fatigue -- Reliability engineering -- Aerospace NDT -- Prognostics and health management; System failures (Engineering) -- Periodicals; Fracture mechanics -- Periodicals; Reliability (Engineering) -- Periodicals; Pannes -- Périodiques; Rupture, Mécanique de la -- Périodiques; Fiabilité -- Périodiques; Fracture mechanics; Reliability (Engineering); System failures (Engineering); Periodicals; Electronic journals; 620.112
• DOI: 10.1016/j.engfailanal.2020.105186
• Languages: English
• ISSNs: 1350-6307

Abstract

Highlights: Dynamic prediction density based Particle Filter algorithm. Three dimensions Crack growth prediction of aero-structure. Actual Degradation data used for prediction. Maintenance planning/scheduling. Abstract: Flaw growth in aero-structures is a stochastic phenomenon as the structure is subjected to random stresses and varying environmental conditions. Accurate prediction of flaw growth enables maintenance managers to undertake correct decision making leading to timely repair/ replacement actions. The flaw growth rate varies with respect to age of the structure making it difficult to predict using classical prognosis techniques. A statistical framework based on sequential Monte Carlo technique known as particle filter (PF) is used in this study for predicting flaw growth. PF implementation involves degradation prediction step and update step at each instant. The selection of a suitable probability density function as a state transition model for predicting flaw growth plays a vital role in prediction accuracy. Different prediction densities represent varying rates of flaw growth during the structure's lifecycle. The use of a single prediction density at every prediction stage throughout the life of the aero structure is inappropriate and leads to inaccuracy. Therefore, dynamic flaw prediction density is proposed to accurately predict the flaw growth. The suitable candidate for dynamic prediction density is Weibull density as it is equivalent to different densities by changing the value of shape parameter. The shape parameter is estimated by applying the Maximum Likelihood Estimation (MLE) technique on the available database of measurements at each life stage, spanning 1000 loading cycles, throughout the life. A case study of crack growth around the countersunk rivet holes of an in-service passenger aircraft's wing structure is used to validate the accuracy of the proposed technique. The database is divided into training and validation segments. The proposed prognosis technique and classical prognosis technique (with single/constant prediction density throughout the lifecycle) are compared to showcase the efficacy of the proposed technique for prognostics of aero-structures.