Smart Compressed Sensing for Online Evaluation of CFRP Structure Integrity

This paper investigates a smart sensing technique for online and automatic evaluation for carbon fiber reinforced polymer (CFRP) structures' integrity. A compressed sensing (CS)-based algorithm framework integrating sampling and defect evaluation together is proposed enabling intelligence of an open-ended waveguide imaging system for the first time. Compared to traditional raster scan designs which require complete sampling, this smart CS technique can generate whole damage pattern while the scanning is conducting, the sensor intelligence is achieved by the proposed algorithm framework rather than seeking a hardware update. The CS enables accumulated downsampling and sparse recovery based on the frequency sparsity of impact damages on CFRP structures. Damage pattern in the reconstructed image can be detected using the proposed histogram threshold edge detection (HTED) algorithm when the image is stabilized enough. Edge-preserving smoothing is used to improve the stability of the reconstructed image. The experimental results illustrate time efficiency of the framework and more accurate damage localization using HTED. The proposed smart CS technique is attractive in quality control of CFRPs production. This technique can also be applied to situations where the sampled data is partly lost. Dispensing with hardware updates incurs minimum disruption and also benefits cost control and improves the productivity.