The Statistical Interpretation of Fiber Counts in Yarn Cross-Sections

Abstract

Part I. In a model yarn spun from two components differing only in color. the factors affecting yarn irregularity may be divided into two classes, only one of which affects the regularity of blend proportions. The published tests for random mixing are reviewed and it is shown that the correct test for randomness is one based on the variance of blend proportions. Part II. When tests based on the theory of blending are applied to blends of dissimi lar fibers, the results must be interpreted with caution. Departures from randomness may be safely detected, but these departures may be caused by the physical differences between the fibers rather than by insufficient processing. Part III. The mixing of components within the cross-section must be considered separately, and special attention should be paid to effects in the yarn surface. The number of apparent groups along an arbitrary line in the section provides the most sensi tive test for departures from randomness, but information about the behavior of the individual components must be obtained by other means.