Extending Applicable Ranges of Regression Equations for Yarn Strength Forecasting

Abstract

Using recently developed technology for rapid instrument testing of cotton fiber, it is possible to characterize large numbers of bales rapidly and accurately. Many com mercial spinners of cotton and cotton blend yams have used such data in combination with standard statistical techniques to generate regression equations for forecasting product quality. Because of the mill-specific nature of such forecasting equations, each mill or each product must be studied as a separate case. This report describes work that permits extending the applicable ranges of these regression equations to include the effects of changes in yarn count, twist multiple, combing noils, and blend level. Mathematical models quantitatively describing these parameters are presented in a form that can be easily incorporated into existing forecasting equations.