Damage characterization of embedded defects in composites using a hybrid thermography, computational, and artificial neural networks approach

Article Materials Science, Characterization & Testing

Multi-characteristic combination based reliability enhancement of optical bidirectional thermal wave radar imaging for GFRP laminates with subsurface defects

Jinlong Gong et al.

Summary: This paper quantitatively investigated the detection reliability of optical bidirectional thermal wave radar imaging for glass fiber reinforced polymer laminates with subsurface defects, using various imaging algorithms. A multi-characteristic combination method was proposed, which showed enhanced reliability in detecting GFRP laminate defects compared to other algorithms.

NDT & E INTERNATIONAL (2021)

Add to Collection

Article Materials Science, Characterization & Testing

Variational level set and fuzzy clustering for enhanced thermal image segmentation and damage assessment

Zijun Wang et al.

Summary: This study introduces an infrared image segmentation algorithm that combines variational level set and fuzzy clustering algorithms to enhance thermal image quality for damage assessment. The algorithm incorporates local Shannon entropy and fuzzy membership functions to improve robustness against noisy images, while also using higher-order derivative edge detection operator and internal penalty energy for optimization. Experimental results demonstrate the efficiency and robustness of the proposed methodology.

NDT & E INTERNATIONAL (2021)

Add to Collection

Article Polymer Science

Enhanced Defect Detection in Carbon Fiber Reinforced Polymer Composites via Generative Kernel Principal Component Thermography

Kaixin Liu et al.

Summary: In this study, a novel GKPCT method is proposed for defect detection in CFRP composites. The method uses spectral normalization generative adversarial network to augment thermograms and applies KPCT method for feature mapping to differentiate defects and background. Experimental results demonstrate the feasibility and advantages of the proposed GKPCT method.

POLYMERS (2021)

Add to Collection

Article Mechanics

A micromechanics and machine learning coupled approach for failure prediction of unidirectional CFRP composites under triaxial loading: A preliminary study

Jiayun Chen et al.

Summary: This study proposes a hybrid method based on ANN and micro-mechanics for predicting the failure of composite lamina under triaxial loading. The ANN algorithm shows robustness in predicting failure with over 97.5% accuracy under different triaxial loading conditions for UD lamina. The approach can be used for modifying existing failure criteria and proposing new ANN-based failure criteria.

COMPOSITE STRUCTURES (2021)

Add to Collection

Article Instruments & Instrumentation