Multiple regression model of mineral wool fibre thickness on a double-disc spinning machine

摘要

This paper describes a multiple regression model for calculating the thickness of mineral wool fibre manufactured on a double disc spinning machine. The aim of the paper was to set up a model which would allow process control of mineral wool production. Based on the selection of the key parameters that define mineral wool fiberisation and with the aid of dimension less analysis, characteristic numbers were defined and a regression model was set up. Parametric regression model constants were determined from experimental results obtained on a real double disc spinning machine. Regression model results showed high correlation with the multiple regression coefficient r~2 higher than 0.87. The multiple regression model developed can be further applied to multi-disc spinning machines. It would permit mineral wool production process control since all the key parameters in the model can be measured on a real production process.