Prediction of heat transfer and air permeability properties of light weight nonwovens using artificial intelligence

Abstract

Effects of pore sizes and distribution of pore sizes of light weight spunlace nonwovens on the heat transfer and air permeability of these fabrics have been studied. Image analysis has been applied to extract the geometrical features of the cross-section of spunlace samples (pore sizes and distribution of pore sizes) at the different layers in the thickness direction. A neural network model is also developed for the prediction of heat transfer and air permeability with respects to structural properties of light weight nonwovens. Results show that the increase in pore sizes and distribution factor of pore sizes increases the air flow rate and heat transfer properties of the nonwoven fabrics respectively. The neural network model also predicts the air permeability and heat transfer of nonwovens in terms of the measured geometrical properties.