CV% of Mass Variation and Imperfection Testing

CV% of mass variation (coefficient of variation of yarn linear density along length) is measured by integrating capacitive sensor signal variance over standard test length (1,000 m at 400 m/min = 2.5 min test, 10 replications per package for statistical validity) and computing CV% = (σ / µ) × 100 where σ = standard deviation and µ = mean signal. Imperfections are counted as threshold crossings: thin places when signal drops below −50% of mean for ≥1 cm (count/km), thick places when signal exceeds +50% of mean for ≥1 cm (count/km), neps when signal exceeds +200% of mean for ≥0.1 cm (count/km). USTER Statistics 2023 Ne 30 ring-spun cotton benchmarks: CV%: 5%ile = 12.0, 25%ile = 13.5, 50%ile = 14.8, 75%ile = 16.2, 95%ile = 18.5; thin places (−50%): 5%ile = 4, 25%ile = 15, 50%ile = 40, 75%ile = 85, 95%ile = 220 per km; thick places (+50%): 5%ile = 18, 25%ile = 55, 50%ile = 120, 75%ile = 240, 95%ile = 520; neps (+200%): 5%ile = 55, 25%ile = 105, 50%ile = 180, 75%ile = 290, 95%ile = 520. Spectrogram analysis (FFT of capacitive signal) identifies periodic defects: chimney peaks at specific wavelengths diagnose machine faults — roller eccentricity (wavelength = roller circumference × draft, e.g. 4.5 cm × 25 draft = 112 cm periodic thick place from 14 mm top roller eccentricity); comber wave (wavelength = mean fibre length × break draft zone draft, e.g. 27 mm × 1.35 = 36 mm); flat-top spectrogram (broadly elevated energy at short wavelengths 1–20 cm) indicates carding wire damage or high trash content. IPI total (thin + thick + neps) quality classification: IPI < 100 = export quality, IPI 100–250 = domestic first quality, IPI > 400 = second quality.