Research
Ongoing projects
Machine learning models for convective heat transfer processes
Deep learning algorithms can readily identify the features embedded in unstructured data such as images, texts, and signals. Computational fluid dynamics (CFD) techniques are widely used for simulating convective heat transfer processes, but extensive computational resources and computation time are required to solve for the large-scale, spatiotemporal dynamic problems. If simulation and/or experimental data for transport processes are available, the deep learning models are expected to be trained to solve for the complex transport problems without necessitating significant computational costs, and serve as surrogate models.
3D printed heat exchangers
The current-state-of-the-art fabrication includes manual and long lead items such as tube bending, fit check tubing, hand welding, etching, priming, and epoxy bonding the tubing to the heat exchanger. Current 3D printing techniques will not only reduce the fabrication lead time and labor cost, also increase the effective heat transfer between the working fluid and the heat exchanger plate. Moreover, the designs of legacy heat exchangers were not derived by accounting for the geometric design freedom of 3D printing techniques. This project aims to optimize the heat exchanger designs by exploring a large design space that is enabled by the 3d printing techniques.
Thermal transport in carbon nanotube composites
There is a growing need to replace heavy copper (Cu) electrical and data wiring with lighter alternatives for the mobility and microelectronic industries. Cu-carbon nanotube (CNT) composites that integrate Cu with CNTs are considered as one of the copper’s substitutes due to the competitive density and the exceptional physical properties of CNTs. However, the electrical and thermal conductivities of experimentally synthesized Cu-CNT composites rarely exceeded those of Cu primary due to the large inherent interface resistances. Theoretical efforts are under way to understand the gap between the properties of CNT and CNT-based composites.
Previous projects
Wearable thermoelectric generators
Thermoelectric power generation offers a promising way to recover waste heat. A critical challenge in using thermoelectric generators (TEGs) for charging the portable or wearable electronics has been their limited outputs, as available temperature differential on human body is typically less than 10 K. We designed the TEG systems for wearable applications based both on theory and systematic experiments.
3D printed thermoelectric generators
The 3D printing process has been recognized as an advanced technology for directly producing such 3D cellular architectures with great geometrical complexity in a cost-effective manner. However, the complex 3D geometries of thermoelectric materials and modules have never been realized so far because the full functionality of 3D printing technology has yet not been applied to thermoelectric technology. We proposed the designs of 3D printed thermoelectric generators for efficient and durable power generation.