Estimation of Molecular Weight Distributions in Methyl Methacrylate Solution Polymerization

A new approach for estimating molecular weight distributions in methyl methacrylate solution polymerization is presented. This approach is based on a polymerization model described by a set of nonlinear coupled partial integro-differential equations. The model is discretized by orthogonal collocation on finite elements and implemented in a discrete extended Kalman filter. This makes it possible to predict the entire molecular weight distribution from simple measurements and periodic gel permeation chromatography data throughout the polymerization process. Experimental results and computer simulations are presented and discussed.