Objective: Design a heat exchanger that makes use of Triply Periodic Minimal Surfaces (TPMSs), that mitigates the high pressure losses usually reported with these designs

Outcome: Developed several Python scripts to modify implicit TPMS functions to replace baffles and reduce pressure loss

Skills: Python, applied math, heat transfer, fluid dynamics

This is a personal project undertaken outside of school. It is ongoing.

Heat exchangers are ubiquitous across every field of engineering. Academic research indicates that TPMSs can make fantastic heat exchangers due to their continuously curved geometry, large shared surface area, and natural turbulence promotion. Building on research I encountered while working with TPMSs for a different purpose (see: Gyroids), I wanted to see if I could develop a heat exchanger that makes use of these interesting surfaces.

These surfaces were generated through Python scripts I wrote. This approach allows for more flexibility in implementation than if industry-standard software was used. For example, I have eliminated the flow-disrupting baffles usually used to separate the working fluids at the inlets/outlets by gradually phasing each subdomain in/out, something I believe industry-standard software cannot do.

I believe this approach allows for greater optimization and can be used to reduce the high pressure losses usually reported in the literature. I’m planning to develop a model to better characterize its performance, and once satisfied, have one 3D printed out of metal.

With 3D printing, it’s possible to create geometries that would be impossible through any other means, and adding intricate details to a design no longer increases manufacturing cost. For example, with this heat exchanger, I can functionally (read: smoothly and continuously) change any parameter such as wall thickness, relative volume of the domains, or unit cell size, as a function of X, Y, and Z coordinates. If a reliable heat exchange model could be developed, I could also change parameters based on any thermodynamic property of the working fluids. This could potentially unlock more compact and efficient heat exchangers than are currently possible. Industries such as aerospace and automotive, where size and weight of components are critical, would benefit from these designs.

3D printing is the only way to produce a TPMS (apart from maybe casting certain geometries). The metal 3D printing industry has suffered financially over the past few years due to overhyped prospects; 3D printing requires a very expensive machine and highly trained operator and is fairly slow (similar to machining parts, one of the most expensive metalworking processes). Most parts that can be made through alternative means, especially mass-produced ones, are probably best made using those other means. However, making a conventional brazed plate or shell and tube heat exchanger is also slow and labor intensive, making heat exchangers pretty expensive anyway. I believe if effective heat exchangers could be developed that require 3D printing, the cost could be justified.

One interesting application for these surfaces is the handling of working fluids at high pressures, e.g. sCO2/R-744 applications. TPMSs are remarkably strong due to their internal interconnectedness and lack of corners to concentrate stress (this was the original reason we were researching them at the Hixon Lab). I would also like to test stress survivability for this reason.

TPMS Heat Exchanger