Multivariate Regression Modeling of Parameters Affecting Punching Force during Random Velour Needling

Abstract

Manufacturing of random velour needled fabrics comprises two separate needling operations. In the first stage a carded fibrous assembly is consolidated using a barbed needle. The consolidated assembly is opened-up due to action of fork needles of the second needling stage. In this special needling operation the mono-structure consolidated fibrous assembly is converted to a bi-structure fibrous assembly, comprised of base and pile layers. The extent of imparted structural changes influences physical and mechanical properties of the resultant random velour fabric. Fork needles which are the culprit of the structural alteration are subjected to relatively large forces. In this work the effect of random velour needling parameters and fiber characteristics on force exerted on the fork needle were investigated. Force measurement and data analysis were carried out using a purposely designed twin load cell force measuring unit and required software. Sixty four samples of polyester fabrics were prepared. Separate identification of the effect of principle independent parameters on the exerted force on individual fork needles was achieved using a linear multivariate regression analysis based on SAS 9.1 statistical software. Estimation of total average force (Frms) and peak average force (FPA) were used as response variables of the model. Results manifested that force exerted on the fork needle is profoundly affected by a large number of parameters. It was also found that, total fiber displacement “ TFD“, occurring during random-velour needling was the foremost factor directly influencing force on the fork needles.