Tribology & Surface Engineering

investigation of tribological characteristics to improve the lifespan and efficiency of next-generation internal combustion engine vehicles (NG-ICEVs) by lowering wear and friction coefficient in liner-piston interactions. The overall goal is to develop a comprehensive guideline for micro-texture characteristics (shape, size, density, etc...) depending on the working and ambient conditions of the ICEVs and micro-machining properties with Electrical Discharge Machining (EDM). CFD simulations, micro-EDM processes, and tribology experiments are being conducted concurrently.   

Deep Learning

development of neural network-based "heat transfer models" to utilize nucleate-boiling (NB) phenomenon in coolant systems for batteries. In theory, NB can enhance the heat transfer coefficient immensely, around 10^4～5 times, thanks to the air bubbles. However, the thermal interval for NB to work is limited, around 10-30°C, which needs a precise control algorithm for cooling systems. In addition, it is rather difficult to model these complex physical phenomena of multiphase fluid flow. This is where "deep learning" (DL) comes into the spotlight. The goal is to develop physics-based DL frameworks to predict heat flux during coolant flow and make necessary adjustments in the cooling system to ensure that the heat transfer is indeed in the NB regime. The physics-based model will be backed by visual data of air bubbles captured by high speed cameras during the experiments. By utilizing NB phenomenon, it is possible to scale down the size and weight of conventional cooling systems, which will eventually result in higher mileage for EVs and cut down the emissions for NG-ICEVs .