School of Mechanical Engineering Seminar
Wednesday, December 8, 2022, at 14:00
Wolfson Building of Mechanical Engineering, Room 206
"Failure Predictions of Bonded Composites Using Machine Learning Strategies"
Dr. Yuval Freed, Israel Aerospace Industries
Machine learning algorithms have become very attractive for design and substantiation of engineering applications, especially as available computational resources are rapidly increasing in the past few years. Several examples of structural-related usages of machine learning algorithms are presented in this seminar, with special focus of adhesively bonded joints.
Bonding of pre-cured composite parts is considered as an efficient assembly method, that can lead to significant weight and cost reductions as compared to standard joining techniques. While the advantages of bonding are clear, its main drawback is that it is very difficult to evaluate whether the bond had achieved its full strength or not. Following several accidents and past-experience from the late 1990s and early 2000s, the certification authorities require the applicant to determine the maximum allowed disbond size in the bonded assembly, with arresting features incorporated into the design to prevent such disbonds to propagate. In this study, a comprehensive testing campaign was conducted to obtain the mixed mode characteristics of bonded composites. Thousands of finite element simulations were automatically constructed, each with different combination of failure parameters (both VCCT and CZM were used as failure criteria). Machine learning strategies were then employed to determine the optimal failure parameters, with statistical analyses conducted to ensure B-Basis level reliability, accounting for the test results variability. Finally, the optimized failure parameters were validated with respect to several element-level bonded composite components. These failure parameters can now be used to determine the maximum allowed disbond size for adhesively bonded joints.
Short BIODr. Yuval Freed joined the Israel Aerospace Industries (IAI) in 2008 after awarded a PhD in Mechanical Engineering from Tel Aviv University. He was appointed as the head of the Fatigue Department in 2010, and as the manager of the Structural Analysis group in 2013. In this position, Dr. Freed leads ~40 Stress and Fatigue engineers. As of 2019, Dr. Freed also serves as the Deputy Director of IAI Design and Analysis Directorate. In this position he is administratively responsible on ~130 engineers. Dr. Freed is involved in many development programs in various fields, such as business jets development, aircraft conversions, special mission aircraft modification, unmanned air vehicles development, missiles and space. He has recently returned from a sabbatical year at the University of Washington in Seattle, where he studied usage of machine learning strategies for practical engineering application with emphasis on the aviation industry.
Join Zoom Meeting
https://tau-ac-il.zoom.us/j/4962025174?pwd=bVJUeElXRUUya3BERisyNllLOE9EZz09
