Should We Mass Customize 3D Printed Heat Exchangers?

On the 3DPOD, JPL’s Doug Hofmann recently said that 3D printed heat exchangers were one of the applications he was most excited about in additive. Since he is part of a very advanced team that both builds gearboxes for Mars rovers, as well as systems to produce oxygen on Mars, this caused me to immediately revisit heat exchangers as a key application for 3D printing.

We’ve seen a lot of them over the years—3D printed heat exchangers. They typically seem intricate and needlessly complicated, designed to resemble an H.R. Geiger creation. With the world moving towards an admiration of the simple and the complex simultaneously, these are curious artifacts that may very well be a portend of an industrialized AM future. This is because heat exchangers are deceptively simple devices that use complex architectures to achieve a vital task: the passage of heat between two different fluids.

Courtesy of Paul Singh.

These fluids may be mixed in together or they may be separated. The most commonly used example is a car’s radiator. However, heat exchangers are used in just about any industrial process, a vast array of manufacturing processes, many other vehicles, and in a lot of processing applications, systems, and factories. Refrigeration, air conditioning systems, and much else besides, creating a $40 billion industry.

There are many different types of heat exchangers and many different designs. One common type is tubular, or shell-and-tube, heat exchangers. This design is made up of a network of fluid containing tubes that is heated, or cooled by an additional fluid outside of these tubes, but within the shell that surrounds them both. Plate heat exchangers feature a series of plates with tiny channels that are either permanently joined or stacked, but are able to be taken apart for maintenance. In this case, fluids pass in between the stacked plates.

A plate heat exchanger.

In parallel-flow heat exchangers, both fluids travel in parallel. While in counterflow, they each enter from a different side and, with crossflow, they cross each other. Generally, the counterflow design is better at uniform heat transfer, while crossflow may enable more compact designs. Theoretically, crossflow heat exchangers can reach 70% efficiency while counterflow designs could be 95% efficient. If you’d like a tad more depth, this is a nice explainer, but generally, “the proper selection of a heat exchanger is difficult, and becomes dependent upon both the products and the applications to be handled, either currently or in the future.”

This paper looks at a compact heat exchanger, 3D printed specifically for an aviation application.

“Modern passengers’ aircrafts fly at higher altitudes, where the air density is lower. Consequently, the installed air conditioning system needs larger airflow rates for a fixed amount of exchangeable heat power. In the framework of the heat transfer, CHXs represent a new frontier as they embody a perfect compromise of compactness and thermal efficiency. Compactness, expressed in terms of surface area density η=A/V, is the main feature. Extended internal surfaces are used to enlarge the surface area-to-volume ratio with different configurations, such as: plate-fin or tube-fin CHXs. The design of internal complex fins shape and arrangement is translated in terms of global minimization of the device volume, better temperature control and lower weight. Plate and fins CHXs are widely used for gas to gas applications. The purpose is to induce several sudden changes of the flow direction, generating an increase of the heat transfer coefficient but consequently, also of the pressure drop.”

The paper presents the business case for one particular gas-to-gas heat exchanger for one application, aircraft air conditioning, much better than I possibly could. Besides advantages in mass and weight reduction and increases of surface area, the device has been designed to create more turbulence. This details some of the important design considerations, but also simultaneously points to the fact that there is no such a thing as a perfect, one-size-fits-all heat exchanger. For each individual heat exchange case, there can be optimal design constraints depending on the use case, the liquids or gasses involved, and what needs to happen thermally. Is control more important than speed? Are we heating something up or cooling all the time? All of these constraints can be optimized in each specific instance.

For everything from cooling and heating to processing of a wide variety of products, heat exchangers are essential components. Through optimized flow, geometry, mixing, turbulence, separation and surface area, heat exchanger performance can be improved significantly. More compact heat exchangers mean more efficient or compact machines or devices, while weight savings can also reap rewards, especially in transport. Just like many complex shapes, heat exchangers can be significantly improved by optimizing geometry through 3D printing.

A Shell and tube heat exchanger.

This means that the math is daunting and we’d need some kind of automated computation fluid dynamics and finite element analysis operations but we could mass customize 3D printed heat exchangers eventually. This will be an incredible opportunity for us to take over this market. We would need to define all of the constraints and work through the design space in additive, but I’m convinced that it is possible. With everyone else designing standard solutions or making one-off solutions, it should be possible to make near-optimal heat exchangers with 3D printing near automatically by relying on specialty software tools. The heat exchangers may never be perfect, but they would outperform existing ones.

We can create turbulent textures, internal fins, twisting geometries, and fold things in around themselves more efficiently than other technologies. We can optimize shapes to a degree that others simply cannot. This paper outlines how an AM heat exchanger can outperform a traditionally made counterpart for an electric aircraft motor when it  “combines involute secondary surface and 3D lattice structure as tertiary fins to maximise flow mixing and surface area.” Here, we can learn of how our favorite things: lattices can be used in heat exchangers. In this journal article, a Prusa SL1 vat polymerization 3D printer is used to make airtight vessels to try to make a sorption heat exchanger.

Tubular heat exchanger.

This paper looks at heat exchangers for Micro Organic Rankine Cycles, whereby electricity can be generated from low heat. This could be used to harvest energy widely from industrial processes, if only it were efficient enough. Enter a counter flow heat exchanger. In this case, the part failed because of wall thickness, suboptimal support strategies, and design issues, but it’s a great lesson for someone wanting to make a similar part.

In this paper, a plate heat exchanger with 0.6mm wall thickness is imagined and one could easily see how we could optimize the internal structure and texture of such a part so that it could outperform traditional ones. Here, we can learn of how a  350-W polymer water-to-air heat exchanger can be made using Material Extrusion. The polymer FDM exchanger has a “220% and 125% improvement in heat flow rate over mass (Q/m) and heat flow rate over volume (Q/V), respectively, when compared to comparable state-of-the-art plate fins.” Thanks to Doug from JPL, and these papers, I too am now incredibly excited about 3D-printed heat exchangers.