Cotton quality as affected by changes in spindle speed

Authors: Baker, Kevin; Hughs, Sidney; Foulk, Jonn

Submitted to: Applied Engineering in Agriculture
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 2/3/2010
Publication Date: 6/30/2010
Citation: Baker, K.D., Hughs, S.E., Foulk, J. 2010. Cotton quality as affected by changes in spindle speed. Applied Engineering in Agriculture. 26(3):363-369.

Interpretive Summary: Spindle picking has become the preferred method of harvesting most cotton in the U.S. Improvements to the cotton harvester have primarily focused on increased speed in order to reduce the cost of harvesting and reducing head weight. As the spindle speed has increased, cotton fibers can wrap more tightly around the spindle. As spindle diameter decreases, cotton fibers will wrap around the spindle more and will also become tighter on the spindle. As spindle length decreases, cotton plants must be further compressed as they pass through the picking zone. These changes have resulted in a general decrease in cotton fiber quality, particularly regarding spindle twists, preparation, and neps. Data from this study confirms that the amount of trash, spindle twists and neps increases as the spindle speed is increased.

Technical Abstract: Three cotton varieties were grown under furrow-irrigated conditions in southern New Mexico and hand-harvested in a way that kept individual bolls intact. The cotton bolls were conditioned in a controlled atmosphere and then subjected to a single cotton picker spindle operating at a speed of 1000 to 3000 rpm. Two spindle designs were studied, a 12.7 mm (½”) round, tapered, barbed spindle and a 4.8 mm (3/16”) square spindle that was straight and smooth. Mass measurements were taken to determine the portion of seed cotton not picked and the portion that would fly off and not stick to the spindle. A force gauge was used to determine the peak force that was needed to pull the seed cotton from the spindle. Moisture content of the bolls was 9 to 10 % d.b. Results showed that the smaller, straight spindle was more aggressive in removing cotton from the boll. There was approximately twice as much fly-off from the barbed spindle than from the smaller straight spindle at any given speed. Fly-off also increased exponentially for each spindle type as the speed was increased for both spindle types. The peak force required to remove the seed cotton from the spindle ranged from 50 to 100 % more for the smaller straight spindle than from the barbed spindle. For both spindles, the peak force requirement was approximately doubled each time the speed was increased by 1000 rpm, indicating an exponential relationship between speed and wrap tightness. Additionally, field tests were conducted for the 2005 and 2006 crop years by the USDA, Agricultural Research Service, Southwestern Cotton Ginning Research Laboratory in Mesilla Park, New Mexico. Three cotton varieties were grown under furrow-irrigated conditions in southern New Mexico and harvested with a modified 1-row cotton picker each year using a ground speed of 0.85 m/s and spindle speeds of 1500, 2000, and 2400 rpm for the 2005 crop and spindle speeds of 2000, 3000, and 4000 rpm for the 2006 crop. The tests were replicated 4 times. Stalk losses in the field were significantly greater at a spindle speed of 1500 rpm than for speeds of 2000 rpm or greater for all varieties. This indicates that a spindle speed of at least 2000 rpm is needed for the picker to adequately function. Stalk losses were greater with speeds of 3000 and 4000 rpm than for a speed of 2000 rpm with the Pima variety. The number of spindle twists in a 1000 g seed cotton sample and the percent of seed cotton that had spindle twists was greater for the 3000 and 4000 rpm spindle speeds than for the 2000 rpm spindle speed. Both measurements of spindle twists in seed cotton nearly doubled when spindle speed increased from 2000 to 3000 rpm and then increased more when spindle speed increased to 4000 rpm. The increase in spindle twists makes preserving fiber quality while ginning a greater challenge. HVI classing data also showed no significant differences among treatments except for upland lint samples collected before lint cleaning. In these samples, there were higher levels of trash with spindle speeds of 3000 and 4000 rpm than with a speed of 2000 rpm. The differences were no longer significant for samples collected after lint cleaning. Differences were significant for AFIS nep count and short fiber count in the raw stock from the bale with all three varieties. Both neps and short fiber content increased when spindle speed was increased from 2000 to 3000 rpm and from 3000 to 4000 rpm. These nep count and short fiber differences were diminished, but did not disappear as the fiber was further processed. For AFIS dust count and trash count, significant differences were noted in the raw stock, with higher levels of dust and trash at speeds of 3000 rpm or greater. Differences in dust count and trash count disappeared as the fiber was further processed.