Forecasting End-Uses of Nonwovens by Integrating Measured Data and Human Expertise

Abstract

This paper, which deals with the forecasting of nonwovens end-uses, is divided in two parts. The first part presents optimized methods for measuring the structures of nonwovens. The raw data are extracted directly from 3D images of the accurate topographic surfaces of the materials and also from other instruments. Next, data analysis techniques are applied to select relevant structural parameters and forecast the expected end-uses of nonwovens. Relevant physical features are selected by integrating measured data and the knowledge of experts. The effectiveness of these methods has been shown through a number of nonwoven products designed for filtration.

Modelling nonwovens using artificial neural networks

“…Ann was used to model the relationship between the structural parameters and the filtration property of spun bonded nonwovens (Koehl et al, 2005). The input structural parameters were fibre orientation ratio (MD/CD) in machine direction (MD) and cross-machine direction (CD), thickness, basis weight, fibre density, binder rate, total pore volume and basis weight uniformity.…”