 |
||
|
||
Home > TERATEC FORUM > Workshop 4
Big data, multiscale and materials
Machine learning based optimisation of composite materials in structural applications
Arnaud FROIDMONT, NOESIS
This paper presents a new accurate and efficient reliability technique for designing and optimizing composite materials structures based on a Self-Organizing Map Adaptive Sampling algorithm (SOMBAS). This approach allows the computation of reliability estimates with better efficiency and accuracy. Uncertainty quantification is essential for composite materials design where the current challenge is to make them cost effective and competitive with metals.
An industrial composite wing structure is used as a test case to demonstrate the validity of the reported method in terms of accuracy, computation time and applicability.
