When 3D Printing Metal Actually Makes Sense
March 18, 2026
Additive manufacturing gets pitched as the answer to everything. Most of the time it is not. Printed metal parts are expensive, slow to produce in volume, and often worse than a machined or cast part for the same job. Our heat exchanger project for Meer Lab was one of the rare cases where it was the only answer.
The problem
Meer Lab's solar cogeneration system was producing more heat than it could get rid of. The receivers were heading past 150°C, and the operating limit was 80°C. The space available for cooling was fixed. We needed a lot more heat transfer without a bigger part.
Why conventional designs did not work
More cooling usually means more surface area. Fins, tubes, and plates add surface area in a fairly linear way, and they hit the size limit fast. We needed a geometry that packed far more surface into the same volume.
We landed on a gyroid, a type of triply periodic minimal surface. It is a continuous, curving structure that splits the volume into two interlocking channels with a huge amount of wall between them. It is also impossible to machine or cast. The only way to make it is to print it, so we specified 3D printed copper.
Proving it first
We ran thermal CFD on the design before anything was printed. The model predicted the system would settle around 70°C, comfortably under the limit. The printed part, made by Markforged, matched the prediction. As a bonus, the overall system came in about 30% cheaper than a standard concentrated solar heliostat setup.
How we decide now
This project gave us a simple test for when to print metal:
- Is the geometry impossible any other way? If it can be machined or cast, it probably should be.
- Does the geometry directly improve performance? Complexity for its own sake is just cost.
- Is the volume low? Printing makes sense for prototypes and small runs, not for ten thousand units.
- Can you prove it before you print it? Simulation is cheap compared to a failed copper print.
If the answer to all four is yes, printing is worth a serious look. If not, there is almost always a cheaper and better way.