Computational Shape Optimization Of Microreactors Based On Cfd Simulation And Surrogate Model Driven Optimization

Runzhe Liang,Zhihong Yuan

DOI: https://doi.org/10.1016/B978-0-12-823377-1.50155-5

2020-01-01

Abstract:Pharmaceutical industry has become an emerging industry worldwide. However, the automation level and the production efficiency of pharmaceutical manufacturing are relatively low because of the large-scale utilization of batch mode. In the context of the rise of microdevices, this paper investigated the shape optimization of Y-shape microreactors for the production of the optically pure chiral product (S)-1-phenylethylamine. We established the computational fluid dynamics (CFD) based automatic simulation platform to obtain a large number of simulation data sets. We built the data-driven surrogate model based on support vector machine (SVM). Then, we formulated the optimization framework to obtain the optimal shape parameters. By shape optimization of microreactors, the yield of the main product can be increased to about 4.3 times as high as that of the initial microreactor.