After two years of studying the link between cotton origin and spinning performance a U.S. mill announced it will not accept any Georgia-grown cotton after this year.

According to Matthew Thomas, technical director, Frontier Spinning Mills, Inc., the company’s research in 2003 and 2004 showed a greater occurrence of “ends down” in cotton grown in Georgia, versus cotton grown in four other states. Thomas spoke at the 18th annual Engineered Fiber Selection Systems Conference, in Memphis.

Thomas said in 2003, the first year of the study, Frontier spun 20 percent Georgia cotton in all its mills and dropped it to 7.5 percent to date for 2004. “We project that for 2005, we are going to be at a 3 percent rate, based on pre-existing contracts. Going forward, our management has made the decision that on all new contract inquiries, they will be for zero-percent Georgia cotton.”

Frontier was founded in 1996, with corporate offices in Sanford, N.C. It currently operates nine mills, three in Alabama, one in Georgia and five in North Carolina. Frontier is primarily an open-end spinning mill, and consumes over a million of bales of cotton annually.

For the study, the mill chose a company plant that consumes about 90,000 bales of cotton annually. The 100 percent open-end spinning plant spins 100 percent cotton for the knit apparel market.

The study looked for a correlation between ends down and the origin/origins of the cotton being spun. Ends down, according to Thomas, “is an interruption in the spinning process that occurs when the spinning tension becomes greater than the tensile strength of the yarn. An ends down is not only lost efficiency, but it’s lost revenue.”

According to Thomas, as yarn counts get finer, ends down generally go up. Also the knit apparel market spins at a much lower twist than for a weaving product, therefore ends down becomes more critical. Strength is not a critical factor in the knit apparel market.

In 2003, the top five states supplying cotton to the plant were evaluated for ends down rate. Mississippi received a benchmark zero value. Arkansas had 6.5 percent fewer ends down than Mississippi; Louisiana, 11 percent higher; Alabama, 13.8 percent higher, and Georgia, 40.9 percent higher.

The study showed that in 2003, at the mean of 244 ends down per 1,000 rotor hours, the cotton consumed consisted of 40 percent Georgia cotton, 24 percent, Mississippi, 18 percent, Arkansas, 8 percent, Louisiana, and 6 percent, Alabama.

“When we performed the best (the lowest number of ends down of 134 per 1,000 rotor hours) we consumed 1 percent Georgia cotton. Primarily, the bulk of the cotton during that time period came from Mississippi, 36 percent, and Arkansas, 49 percent.

“As you move up in the number of ends down, the consumption of Georgia cotton consumed also moved up. The worst 10 percent of the performance was 470 ends down per 1000 rotor hours, when we consumed 84 percent Georgia cotton.”

Fiber properties of the cotton consumed were also studied. “For micronaire, we were stable, except in the worst 10 percent for ends down, when the micronaire was a little lower. It’s just the opposite of what we would expect. Normally, when mike goes down, we achieve better spinning performance.”

Uniformity and short fiber index are the most important properties for performance, according to Thomas. “As ends down moves up, uniformity values are going down, and the short fiber index starts going up — primarily because short fiber is a function of uniformity.”

In 2004, the mill made some significant changes in its cotton consumption. “We put a base of 80 percent uniformity on all the Georgia cotton we consumed. We also limited the volume of the Georgia cotton that we consumed.”

For 2004, the mean ends down was only slightly lower and the amount of Georgia cotton consumed was only 14 percent. “The larger volume of cotton is coming from Alabama, Arkansas and Mississippi. The studies showed that with Mississippi as the zero benchmark, Louisiana had 34 percent more ends down, and Georgia was unchanged.”

While spinning efficiency seems related to uniformity and short fiber index, “we also had higher ends down levels in 2004 versus 2003 at the same short fiber index level. That was unexpected. It shows that using the uniformity value and the short fiber index to correlate ends down rates, doesn’t tell the whole story.”

Thomas said that according to regression analysis, “if we ran 100 percent Georgia cotton in 2003, our projected ends down rate would be 446 ends down per 1,000 hours. For 2004, the regression model would say that we would have 714 ends down per 1000. Again, these are regression models, so we decided to actually try it.

“We made a laydown of 100 Georgia cotton and spun it. We had 800 ends down per 1,000 hours and 70 percent spinning efficiency. For us, 200 ends down is our target. When we start getting above that, we start seeing degradation in efficiency.”

e-mail: erobinson@primediabusiness.com