This thermal switch project was one of the highlights of my master’s program, offering me invaluable experience in FEM modeling and hands-on experimentation.

The aim was to create a passive thermal switch that toggles between ON and OFF states based on temperature. At high temperatures, it switches ON, and at low temperatures, it switches OFF. Such a device could be used in temperature control modules like a battery thermal management system (BTMS) in electric vehicles (EVs) or spacecraft.

The heart of this thermal switch was a shape-memory alloy (SMA) spring, which behaves uniquely based on temperature: below a certain temperature, the SMA spring is soft and pliable; above a certain temperature, it becomes rigid and strong.

OFF state: Regular spring overcomes the SMA spring, compress the bottom part, resulting in isolated top and bottom.

ON state: SMA spring overcomes the regular spring, extend the bottom part, resulting in contacted top and bottom.

Simulating the thermal switch in FEM software was no walk in the park. The challenges included:
Coupled thermal-structural modeling: The SMA material’s mechanical properties change with temperature, requiring precise phase-change behavior definitions.
Contact modeling: Defining thermal and mechanical contact between components was tricky. Even when the geometry looked visually connected, the FEM software couldn’t detect the contact without proper settings.

After many iterations, we developed a working transient simulation. The system was exposed to an ambient temperature increase from 30℃ to 80℃ within 2 seconds. As the temperature rose, the SMA spring became rigid, enabling contact between the middle and top plates. The simulation showed a switching ratio of 13.95, meaning the heat conductance in the ON state was nearly 14 times greater than in the OFF state.

After many iterations, we developed a working transient simulation. The system was exposed to an ambient temperature increase from 30℃ to 80℃ within 2 seconds. As the temperature rose, the SMA spring became rigid, enabling contact between the middle and top plates. The simulation showed a switching ratio of 13.95, meaning the heat conductance in the ON state was nearly 14 times greater than in the OFF state.

We built an experimental setup to test the functionality of our switch. From top to bottom, this setup consists of liquid cooler, quartz (for heat flux calibration), thermal switch and heat source. By controlling the heat flux with heat source power, we managed to show OFF state (with small gap) and ON state (with gap closed).

We used IR camera to detect the vertical thermal pathway (shown below). By calculating the thermal conductance in both OFF and ON states, we measured the experimental switching ratio to be 1.406 — a number that was way below our expectations 13.95 — which was a little disappointing.

Despite the lower-than-expected results, this project taught me a lot about the intricacies of thermal-structural design and experimentation. While my journey with the thermal switch ended as I moved on to more pressing PhD projects, it remains a valuable chapter in my academic story.

Written by JJ on Nov 1st, 2024