Data fusion and non-destructive testing of damaged fiber-reinforced laminates
Abstract
Efficient strategies for non-destructive testing of damaged composite laminates modeled from homogenization of fiber-reinforced polymers could in- volve fused data. Here, those are from electromagnetic and infrared thermographic modalities, for which semi-analytical models of the interaction are available. Focus is on inter-layer delaminations. Paths forward are outlined, mostly within the realm of Bayesian approaches and of convolutional neural networks, both of wide breadth and the second ones not involving too many prior regularization factors.
| Origin | Files produced by the author(s) |
|---|