Digital Image Analytics for Evaluation of Nanoparticle Impregnation in Woven Fabrics

Abstract

Surface and infiltrating treatments (impregnation) of woven fabrics using nanoparticles have been proven to improve their mechanical, optical, and aesthetic performance in diverse applications. The accuracy, uniformity, and repeatability of the nanoparticle impregnation process determine the quality of the functionalities bestowed on such flexible hybrid composites. Current evaluation techniques like various types of spectroscopy or microscopy are either resource-intensive or are practical only for very small sample sizes. Therefore, there is a critical need for a fast, full-field, cost-effective, non-destructive and reliable technique to evaluate the impregnation quality for woven fabrics. To this end, the development of a new approach using digital image analytics (DIA) to evaluate the uniformity and accuracy of nanoparticle impregnation in woven fabrics is explored in this study. In this study, the impregnation of 80 nm spherical silica nanoparticles into Kevlar 29 style 745 fabric is considered as the test case for all treatment levels. Three trackable heuristic markers are identified from grayscale value histograms of digital images to extract statistically-relevant data sets for neat and treated samples that enable correlation with nanoparticle impregnation. Heuristic feature shifts in grayscale value histograms and spatial frequency profiling correlate with different treatment levels, serving as quantitative metrics to assess uniformity and accuracy of nanoparticle impregnation in the fabric. Further, scanning electron microscopy and energy dispersive X-ray spectroscopy are used to validate the predictions from DIA, thus demonstrating its capability to act as a powerful new tool to evaluate hybrid materials that are seeing greater use in modern technologies.