Study of methods for evaluating the probability of multiple site damage occurrences

Two methods currently available for evaluating the probability of Multiple Site Damage(MSD)occurrence were studied in this paper.One of the methods is a probabilistic analysis approach based on the statistical theory and fatigue characteristics of each structural detail,and the other is an approach which defines the initial damage scenario by means of Monte-Carlo simulation,and multiple initial crack scenarios are randomly generated.A modified method based on the Monte-Carlo simulation was proposed in this paper,in which the random fluctuation of the stress was considered to give more accurate evaluation results.In the presented method,the probability of MSD occurrence in a structural element containing multiple details was calculated based on the Monte-Carlo simulation and the p-S-N curve of a single structural detail.Fatigue tests were accomplished using specimens containing 21-similar-details to obtain the fatigue life corresponding to MSD occurrence.Tests on single-detail specimens and static calibration tests were also conducted to get the basic fatigue properties of the material and the degree of stress fluctuation.The aforementioned three methods were compared and validated via the test results.The influence of the stress random fluctuation degree on the probability of MSD occurrence and influence of the distribution types on evaluating the MSD occurrence probability were discussed.

A fatigue damage-cumulative model in peridynamics

While the present structural integrity evaluation method is based on the philosophy of assumed similitude, Fatigue and Damage Tolerance(F&DT) evaluations for next generation of air-vehicles require high-fidelity physical models within cyberspace. To serve the needs of F&DT evaluation in digital twin paradigm, a fatigue damage-cumulative model within peridynamic framework is proposed in this paper. Based on the concept of fatigue element block and damage accumulation law in form of Coffin-Manson relationship, the proposed model applies to both fatigue crack initiation and fatigue crack growth;fatigue crack growth rates under constant-amplitude and simple variable-amplitude block loading cases can be well predicted for three common structural materials without inputs of Paris law parameters. Additionally, the proposed model can also be easily extended to a probabilistic version;for verification, multiple-site-damage problems are simulated and the statistic nature of fatigue process in experiments can be well captured. In the end, main features of the proposed model are summarized, and distinctions from the other models are discussed. There may be a potential for the peridynamic damage-cumulative model proposed in this work to numerically predict fatigue problems in digital twin paradigm for future generations of aerospace vehicles.