Assessment of Structural Integrity Using Machine Learning
| Autor: | V. S. Bui, Q. H. Nguyen, V. D. Nguyen, V.-X. Tran, Truong-Vinh Hoang |
|---|---|
| Rok vydání: | 2021 |
| Předmět: |
Strain energy release rate
Artificial neural network Computer science business.industry Fracture mechanics Machine learning computer.software_genre Finite element method Surrogate model Structural health monitoring Artificial intelligence Uncertainty quantification business computer Reliability (statistics) |
| Zdroj: | Research in Intelligent and Computing in Engineering ISBN: 9789811575266 |
| Popis: | For structural integrity assessment, when dealing with problems such as uncertainty quantification, real-time structural health monitoring, and reliability analysis, fracture mechanics parameters, e.g. the energy release rate, are required to be evaluated for a large amount of crack configurations. Using finite element (FE) models directly for this task is generally time consuming. This work follows a machine learning-based method aiming at a fast estimation of the fracture mechanics parameters. Towards this end, the offline data are first created by performing the FE simulations for a set of crack configurations from which these parameters can be extracted and then a machine learning algorithm (e.g. artificial neural network (ANN)) is used to build a surrogate model with the obtained data. Using this surrogate model, the above mentioned tasks can be performed with a significantly reduced computational cost while assuring accurate results. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|