Quantitative dispersion analysis of inclusions in polymer composites.

The state of dispersion plays an important role on the performance of polymer nanocomposites. Dispersion is usually assessed based on the qualitative evaluation of microscopy micrographs. In this paper, a quantitative algorithm is introduced for analyzing the dispersion of inclusions in polymer composites using image analysis. In a binary image, on-pixels are considered as particle elements while off-pixels stand for matrix elements. To quantify the dispersion, the mean distance value between any matrix elements to their corresponding nearest neighboring particle element is measured. A dispersion index, DI, is then defined by comparing the image of interest with the associated uniformly dispersed case. Synthetic models were utilized to examine the sensitivity of the algorithm to various dispersity scenarios such as the effect of particle size, clustering, and cluster distribution. Optical micrographs of carbon nanotube modified epoxy with different states of dispersion were also employed to assess the applicability and functionality of the algorithm to real micrographs.