Preparation-induced bias in quantitative characterization of yarn cross-sections: Insights from modified paraffin cross-sectioning and nano-computed tomography

Reliable quantitative characterization of textile structures relies on accurate measurements of yarn cross-sectional geometry; however, the influence of specimen preparation on these measurements has not been systematically quantified. This study evaluated the effect of a modified paraffin embedding procedure on yarn cross-sectional geometry using high-resolution X-ray nano-computed tomography (nano-CT) as a nondestructive reference method. Tencel, polyester (PES), and polypropylene/cotton (PP/CO) yarns were analyzed using an identical image-processing workflow to compare fiber geometry and yarn structural parameters. The modified paraffin preparation systematically altered the geometry of all investigated fibers. The cross-sectional area increased by 31.75% for Tencel, 28.54% for PES, 52.76% for polypropylene, and 37.91% for cotton fibers, while changes in fiber circularity remained below 5%. These dimensional changes propagated directly to the yarn level, resulting in systematic overestimation of the effective yarn radius (8.62–13.03%) and effective packing density (17.91–28.30%) relative to nano-CT. The magnitude and spatial distribution of the deviations depended on fiber chemistry, reflecting water-induced swelling of cellulose fibers and solvent-induced swelling of synthetic fibers during specimen preparation. The results demonstrate that the modified paraffin cross-sectioning procedure is not geometrically neutral but introduces material-dependent systematic bias into quantitative measurements of fiber geometry and derived yarn structural parameters. X-ray nano-computed tomography provides an effective reference technique for validating conventional microscopic preparation methods and improving the reliability and comparability of quantitative structural characterization of fibrous materials.