The development of a computer simulation model of the manufacture of semi-worsted wool yarns

The development of a computer-based simulation model of the production of semi-worsted spun yarns from scoured wool blends is described. The scoured wool properties used as input variables include mean fibre diameter, fibre diameter CV, mean fibre length, fibre length CV, bundle strength, loose wool bulk, medullation, vegetable matter content, colour (brightness, Y) and whiteness, Y - Z), and degree of entanglement (subjectively assessed). Modelling the consequences of stock-dyeing on bundle strength and colour is based on empirical data and accounts for five processing variables - dye-liquor pH and temperature, dyeing time, packing density, and liquor circulation pressure. The consequences of carding are based on previously published theoretical models to explain fibre breakage on opening and the derivation of carding power. A wide range of machine variables is accommodated besides the regain and fatty-matter content of the wool. Gilling variables are restricted to the number of passages and the faller-pin density. The properties of carded and gilled slivers are predicted. Spinning involves yarn twist and sliver and yarn linear density, along with processing variables. A theoretical model, suitably modified, has been incorporated for determining yarn tensile properties. The model accounts for changes in the properties of scoured wool and processing variables over a range of values common to the commercial semi-worsted processing of New Zealand wools. The production variables of primary interest are yarn irregularity, breaking load, extension at break, and bulk; ends down in spinning; fibre loss as card waste; and card-mixing. The necessary conversion equations are based almost entirely on published empirical and theoretical literature, which has been extensively reviewed. Univariate and multi-variate simulation runs with the model are reported for validation and to illustrate the model's potential usefulness. The model appears to satisfactorily determine the relative importance of the various exogenous variables in relation to yarn properties and processing efficiency, except for the prediction of the extension at break