Computational modeling toward full chain of polypropylene production: From molecular to industrial scale. (5th April 2023)

• Record Type: Journal Article
• Authors: Yan, Wei-Cheng; Dong, Tao; Zhou, Yin-Ning; Luo, Zheng-Hong
• Is Part Of: Chemical engineering science. Volume 269(2023)
• Publication Date: 2023-04-05
• Subjects: Polymerization engineering -- Computational modeling -- Multi-scale modeling -- Polypropylene; ACO ant colony optimization -- AI artificial intelligence -- ANN artificial neural network -- BNPSO best-neighbor particle swarm optimization -- BP back propagation -- CFD Computational fluid dynamics -- CSTR continuous stirred-tank reactor -- DFT density functional theory -- DQMOM direct quadrature method of moments -- EMMS energy minimization multi-scale -- FBR fluidized bed reactor -- FNN Fuzzy neural network -- FPQMOM the fixed pivot quadrature method of moments -- GGA generalized gradient approximation -- GPSO global particle swarm optimization -- HACDE hybrid continuous ant colony differential evolution -- KTGF kinetic theory of granular flow -- LSSVM least squares support vector machine -- MAE Mean absolute error -- MAO methylaluminoxane -- MBM mechanism-based modelling -- MCM Monte Carlo method -- MD molecular dynamics -- MFI melt flow index -- MGM multigrain model -- ML Machine learning -- MLMCM multilayer Monte Carlo model -- MM molecular mechanics -- MN number average molecular weight -- MOM method of moment -- MRE Mean relative error -- MSmodel multiple active site kinetic model -- MW molecular weight -- Mw weight average molecular weight -- MWD molecular weight distribution -- OCS online correcting strategy -- ODE ordinary differential equation -- PBE population balance equation -- PBM population balance model -- PDI polydispersity index -- PFADM polymeric flow advection-dispersion model -- PFFDM polymeric flow Fick's diffusion model -- PFM polymer flow model -- PLS partial least squares -- PMGM polymer flow model and multigrain model coupled model -- PMLM polymeric multilayer model -- PP polypropylene -- PSD particle size distribution -- PSM Probabilistic and statistical method -- PSO particle swarm optimization -- QC quantum chemical -- QMOM quadrature method of moments -- RMSE Root-mean-squared error -- R2 Coefficient of determination -- SCM solid core model -- STD Standard deviation of absolute error -- SVM supported vector machines -- Sys-LS-SVM systematic least squares support vector machine -- TFM Eulerian-Eulerian two-fluid model -- TIC Theil's inequality coefficient; Chemical engineering -- Periodicals; Génie chimique -- Périodiques; Chemical engineering; Periodicals; Electronic journals; 660
• DOI: 10.1016/j.ces.2023.118448
• Languages: English
• ISSNs: 0009-2509

Abstract

Abstract: Since polypropylene was synthesized in 1954, tremendous breakthroughs have been achieved in transferring polypropylene from a discovery in the laboratory to an indispensable industrial product. One of the most difficult issue in polypropylene production is the precise control of the synthesis process to tailor the microstructure and the end-use properties, which needs deep understanding of the quantitative relationships among process, polymer structures and properties. However, semi-empirical correlations and experimental measurements are not able to capture the complex multi-scale characteristics of propylene polymerization process. In recent years, mathematical models have been intensively developed to quantitatively link the microstructure of polymer to final macroscopic properties at multi-scales. This review provides an overview of progress in computational modeling of polypropylene production from the perspectives of science and engineering aspects covering synthesis, structure–property relationship, reactor design, processing, composites, and applications. The developed mathematical models at various scales from molecular scale, particle scale and reactor scale toward plant scale throughout the full chain of production process are elaborated. The coupling strategies of models among different scales will be presented. In addition, model-based determination of quantitative relationships among process, apparatus, structure, and property for polypropylene are fully discussed including the recently developed emerging numerical approaches such as machine learning assisted modeling.