NCJ Number
183389
Journal
Journal of Forensic Sciences Volume: 44 Issue: 2 Dated: March 1999 Pages: 283-291
Date Published
March 1999
Length
9 pages
Annotation
This article describes forensic classification of polyester fibers by infrared dichroic ratio pattern recognition.
Abstract
Using a statistical analysis of infrared dichroic ratio data and fiber morphology, 32 polyester fiber samples were classified into 22 unique individual fiber groups and five paired fiber groups. This classification was based on discriminant analysis of infrared dichroic spectra, differences in fiber diameters and variations in cross-sectional shapes. Using only the infrared dichroic data, the fibers were sorted systematically into 13 infrared groups. A model for discriminant analysis was derived and tested using 2,640 infrared dichroic ratios calculated from 5,280 quantitative absorbance measurements from single fiber spectra. A tentative protocol is recommended for fiber sorting. The 32 different polyester fibers that were analyzed represented five U.S. manufacturers. Because polyester fibers are the most common synthetic fibers, differentiating single fibers is important in comparisons of forensic trace evidence. Figures, tables, references