Machine learning-assisted design automation workflow of non-uniform flow distributors for multipass crossflow device

Multipass crossflow devices are essential in the process engineering industry due to their broad range of applications. The growing use of additive manufacturing techniques offers opportunities to improve the efficiency of these devices by enabling non-uniform flow distribution, which can enhance heat transfer uniformity to avoid temperature gradients across individual channels in chemical engineering applications such as heat exchangers and packed bed reactors. This paper addresses the challenge of controlling non-uniform flow distribution by developing a novel design that locally varies channel diameters at predefined positions within the distributor, to achieve user-defined target flow distribution ratios. The proposed design combines machine learning-assisted design automation workflow with algorithmic modeling and computational fluid dynamics (CFD) simulations to manage the complexity of non-uniform flow distribution effectively. The results are further validated through an experimental study of an incompressible gas flow distributor, which supports the findings presented in this work. • Functional enhancement of flow distributors to control non-uniformity in multipass crossflow devices. • ML-driven automation workflow for design and optimization of flow distribution structures. • Experimental validation confirms practical applicability and effective flow control.