An Experimental Case Study in Combining Topology Optimization and Additive Manufacturing Techniques to Generate Novel and High-Performance Compact Heat Exchanger Designs

Abstract The recent industrialization of additive manufacturing processes has led to research interest in alternative, more complex two-fluid heat exchanger designs, such as those constructed from repeating triply periodic unit cell geometries like gyroids, Schwarz minimal surfaces or other complex geometries. These types of designs are particularly convenient because their performance can be scaled up or down by including more or fewer periodic unit cells in the final heat exchanger lattice. Despite their complexity, repeating periodic unit cell-based heat exchanger designs have not been shown to have performance advantages over more traditional designs and therefore have yet to find widespread industrial application. This failure is partially due to the fact that the periodic unit cells used to build these heat exchangers are not optimized for thermo-fluid efficiency, as well as the fact that accurate simulation of a heat exchanger containing thousands of complex unit cells is extremely challenging, making the design of the overall heat exchanger difficult. This paper presents the results of a project to develop a heat exchanger design method which can overcome these deficiencies. In particular, we design, build, and test a crossflow two-fluid heat exchanger based on a periodic unit cell designed using a proprietary topology optimization algorithm. The heat exchanger is manufactured using additive manufacturing. Analysis of the unit cell design showed that it generated secondary flows to increase the effectiveness of the heat transfer while also minimizing unnecessary loss-generating flow structures. The performance of the entire heat exchanger was simulated computationally using RANS-based CFD simulations on a single unit cell. The results of this simulation were used to build a reduced-order model to predict the heat transfer and pressure drop across the entire heat exchanger at two representative operating conditions. After numerical simulation, the final design was manufactured from stainless steel using a Renishaw 500Q machine and its performance was measured experimentally. The experimental results are favorably compared against the numerical predictions to validate both the CFD simulations on the periodic unit cell as well as the reduced order model of the entire heat exchanger. This indicates that the topology optimization design process can be confidently applied to design scalable heat exchangers.