Mathematical Modeling of ARGET Copolymerization of Styrene and Acrylonitrile. II: Efficient Calculation of Bivariate Distributions using Parallel Computing

PINTOS, E.; FORTUNATTI, C.; BRANDOLIN, A.; SARMORIA, C.; ASTEASUAIN, M.

Bahia Blanca

Conferencia; ICPR Americas 2020; 2020

Universidad Nacional del Sur

The probability generating function (pgf) technique in combinationwith parallel computing is a very valuable tool for modeling bivariatedistributions in polymerization systems. Models with this level of detail usuallyinvolve a very large system of stiff differential equations that demands efficientsolution techniques. Part I of this work presented the parameter estimation of amathematical model of the Activators ReGenerated by Electron Transfer ? AtomTransfer Radical Polymerization (ARGET-ATRP) of styrene and acrylonitrile(SAN). A good agreement between model and experimental data was achieved.Part II focuses on the modeling of the bivariate molecular weight distribution(MWD)-copolymer composition distribution (CCD) of the copolymer, as well asthe overall MWD and CCDs. Julia, an open-source language specificallydesigned for scientific computing, was used to implement the mathematicalmodel. The parallelized implementation of the code in Julia allowed computingthe bivariate distributions in relatively short times.