ABSTRACT

It is known that the various means available for measuring fiber strength of cotton generate values which vary in their relationship with yarn strength. This relationship may also vary with different cotton varieties. Recent investigations into the causes for these inconsistencies have led to development of new designs and modifications to HVI lines aimed at improving the precision of measurements and the ability to predict yarn strength.

A preliminary study was conducted using bale samples from the USDA-AMS annual survey of leading cotton varieties for the 1990 and 1991 crops. The fiber and yarn strength data reported by USDA-AMS was supplemented with fiber strength data determined by Spinlab (Model 900A), an MCI-HVI modified to apply increased specimen brushing to reduce the effects of crimp, and a newly developed universal strength tester which combines fundamental measurements of force and mass with high speed breaks. Data for varieties which appeared in both years of the survey were analyzed to detect the main effects of variety and instruments and their interaction on fiber strength. Also, correlation analysis was performed to evaluate the relationship between the various fiber strength measurements and yarn strength (skein strength and single end tenacity of rotor spun and ring spun yarns).

Results revealed significant differences among varieties and instruments for fiber strength as expected. Analysis is also detected a significant variety x instrument interaction, indicating that the fiber strength ranking of varities may vary depending upon the instrument used to measure it. Also, the magnitude of the difference in strength among varieties may vary with different instruments. These effects are important considering the fact that the fiber strength measurement has a direct impact an the perceived value of a bale of cotton.

Correlations of the various fiber strength measurements with yarn strength resulted in a range of R-square values from 51.2 to 87.2 percent. Stelometer, Spinlab- KVI, and the universal strength tester produced the highest correlations with yarn strength with average values of 73.6, 77.3, and 75.9 percent, respectively. The MCI-HVI and the MCI-HVI modified to apply increased specimen brushing correlated at levels of 61.3 and 66.3 percent, respectively.

These results bear more significance when one considers that each data point for the universal strength tester represents the mean of only eight breaks, while the MCI-HVI and stelometer data result from 20 breaks and the Spinlab and modified MCI data result from the mean of 12 breaks. Apparently, the universal strength tester generates measurements which correlate well with yarn strength and are relatively precise.

While these results are meaningful, it must be noted that certain assumptions for a valid analysis of variance could not be met due to the nature of the study. These results will be used only to form a basis of support for continued research in this area.