Topology optimized unit outperforms straight channel design in heat transfer, power density, and effectiveness.
There are few better examples of the synergy between computational design and advanced manufacturing than topology optimization and 3D printing. Indeed, it’s difficult to imagine getting the most out of one without utilizing the other. The latest example of this comes from the University of Wisconsin-Madison, where a team of engineers has created a new heat exchanger that outperforms traditional, straight channel designs by a significant margin.
“Traditionally, heat exchangers flow hot fluid and cold fluid through straight pipes, mainly because straight pipes are easy to manufacture,” mechanical engineering professor Xiaoping Qian in a press release. “But straight pipes are not necessarily the best geometry for transferring heat between hot and cold fluids.”
In conjunction with a technique he patented called projected undercut perimeter, Qian used topology optimization to create a design for hot and cold fluid channels that would maximize heat transfer. Dan Thoma, a professor of materials science and engineering, led the effort to 3D print Qian’s design using laser powder bed fusion.
While the new heat exchanger looks identical to a traditional, straight channel design from the outside, its internal core designs are markedly different. The optimized design has intertwining hot and cold fluid channels with intricate geometries and complex surface features to guide fluid flow in a twisting path that enhances the heat transfer.
UW-Madison mechanical engineering professor Mark Anderson conducted thermal-hydraulic tests on the 3D printed heat exchanger and compared it to the performance of a traditional design. What he found was that the optimized design was more effective in transferring heat and also achieved a 27.6% higher power density.
While previous research has used topology optimization to study two-fluid heat exchanger designs, according to Qian, this is the first instance of using topology optimization and manufacturability constraints together to ensure the design can be built and tested.
“Optimizing design on the computer is one thing, but to actually make and test it is a very different thing,” Qian said. “It’s exciting that our optimization method worked. We were able to actually manufacture our heat exchanger design. And, through experimental testing, we demonstrated the performance enhancement of our optimized design.”
The research is published in the International Journal of Heat and Mass Transfer.
