Efficient Modeling of the Bivariate Molecular Weight Distribution - Copolymer Composition Distribution in SAN Copolymerization Using Parallel Computing

PINTOS, ESTEBAN; FORTUNATTI, CECILIA; BRANDOLIN, ADRIANA; SARMORIA, CLAUDIA; ASTEASUAIN, MARIANO

Boston - virtual

Congreso; AICHE Annual Meeting 2021; 2021

American Institute of Chemical Engineering (AICHE)

The implementation using parallel computing of a model based on the probability generating function technique proved to be a very valuable tool for modeling multivariate distributions in polymer systems. A mathematical model ofthe styrene-acrylonitrile (SAN) copolymerization that predicts the bivariate molecular weight distribution (MWD)-copolymer composition distribution (CCD) of the copolymer, as well as the overall MWD and CCD was programmed in Julia. Julia is an open-source language specifically designed for scientific computing that has native support for parallel computing. The parallelized implementation of the code in Julia allowed solving a huge system of equations to compute the bivariate distributions in relatively short times. A valuable insight on the polymer microstructure under different operating conditions was achieved.