Machine learning-assisted discovery of flow reactor designs

Abstract

Additive manufacturing has enabled the fabrication of advanced reactor geometries, permitting larger, more complex design spaces. Identifying promising configurations within such spaces presents a significant challenge for current approaches. Furthermore, existing parameterizations of reactor geometries are low dimensional with expensive optimization, limiting more complex solutions. To address this challenge, we have established a machine learning-assisted approach for the design of new chemical reactors, combining the application of high-dimensional parameterizations, computational fluid dynamics and multi-fidelity Bayesian optimization. We associate the development of mixing-enhancing vortical flow structures in coiled reactors with performance and used our approach to identify the key characteristics of optimal designs. By appealing to the principles of fluid dynamics, we rationalized the selection of design features that lead to experimental plug flow performance improvements of ~60% compared with conventional designs. Our results demonstrate that coupling advanced manufacturing techniques with ‘augmented intelligence’ approaches can give rise to reactor designs with enhanced performance. Identifying the optimal geometry of continuous flow reactors is a major challenge due to the large available parameter design space. Here the authors combine a machine learning-assisted methodology with computational fluid dynamics and additive manufacturing for the design of more efficient, complex coiled-tube reactors.