Predicting the Tensile Properties of UV Degraded Nylon66/Polyester Woven Fabric Using Regression and Artificial Neural Network Models. Issue 1 (March 2015)

• Record Type: Journal Article
• Authors: Moezzi, Meysam; Ghane, Mohammad; Semnani, Dariush
• Is Part Of: Journal of engineered fibers and fabrics. Volume 10:Issue 1(2015)
• Publication Date: 2015-03
• Subjects: UV degradation -- color values -- fabric tensile behavior -- regression analysis -- artificial neural network; Nonwoven fabrics -- Periodicals; Fibers -- Periodicals; Fibers; Nonwoven fabrics; Periodicals; 677.6
• DOI: 10.1177/155892501501000101
• Languages: English
• ISSNs: 1558-9250

Abstract

UV radiation is the major source of radiation in environmental conditions and affects many characteristics of exposed fabrics. Physical properties of polymeric material and fabric deteriorate when exposed to environmental conditions. The study of the changes in physical properties has been a major issue for researchers. Color absorption of degraded fabrics varies with exposure time. The main aim of this work was to predict the physical properties of UV degraded woven fabrics at different levels of exposure time. Samples of plain-woven fabrics were selected. The fabric consisted of nylon 66 as weft yarns and polyester as warp yarns. A UV light source was used to induce controlled degradation at different exposure times. The samples were dyed in identical conditions and the color values for all samples were measured using a spectrophotometer. The samples were then tested with tensile testing machine and stress-strain curves were obtained. Six parameters were considered in stress-strain curves. Exposure time caused the differences in the color values of the samples. This was used to evaluate the tensile behavior. Regression and artificial neural network methods were used to correlate each of the six parameters of tensile properties with color values. To validate the methods, experimental samples were tested with the tensile testing machine and the results were compared with the predicted values. The results show a good agreement between the experimental and predicted models.