LAUSR

Abstract Laminated composites remain an important family of advanced materials playing essential roles in the development of high-performance components. A classic challenge with this composite family is delamination. Lately, vibration-based methods augmented by dynamic response signal processing and soft computing techniques, have become central for the non-destructive detection of delamination. Many researchers have contributed significantly to tackling the problem. However, the concurrent determination of the presence and severity of delamination remains a challenge for structures more complicated than simple beam-like structures. This paper employs a framework based on the combination of random forests (RF) and natural frequency-shift damage assessment methods for the simultaneous predictions of severity and location of delamination in composite plates. The study commences with the establishment of an accurate finite element (FE) procedure for the inclusion of delamination in composite plates. The FE procedure was validated with good agreement against published results. Four RF models with different architectures are developed and compared (including a comparison with a deep neural network model). Relying only on four natural frequencies, the predictive strength of the RF model with optimally tuned hyperparameters trees is assessed via the FE-generated and experimentally obtained datasets. Results and the model performance metrics indicate that the proposed RF-based method is capable of accurately predicting five parameters quantifying the location and severity of delamination in composite plates. The best RF model produces a coefficient of correlation as high as 0.996 for the location and size parameters and up to 90% accuracy for classifying the delaminated interface.