Statistical characterization of microcellular polyurethane foams microstructure based on 2D and 3D image analysis

This paper presents protocols developed to quantitatively characterize the cellular microstructure of microcellular polyurethane foams, from scanning electron microscopy (2D) and X-ray micro-computed tomography (2D and 3D) data. The objectives are to provide, for both techniques: (i) a detailed description of the analysis steps based on open source Python algorithms; (ii) a method for automatic, robust and objective detection of the cells to limit user's biases; (iii) a statistical description of fraction, size, shape and spatial distribution of cells. The study considers 12 samples with densities ranging from about 400 to 600 kg m-3 and pore sizes from a few micrometers to several hundred micrometers. In addition, the database obtained is used to investigate the reliability of 2D measurements to describe the cellular microstructure statistics.