Mechanical Engineering Seminar: "Spatio-Temporal Extraction of In-Cylinder Flow Characteristics in an Optical GDI Engine using PIV, LES, and Machine Learning" 2/17

02/15/2023
By John Mack

When: Friday, Feb. 17 1-2 p.m.
Where: Dandeneau 220
What: "Spatio-Temporal Extraction of In-Cylinder Flow Characteristics in an Optical GDI Engine using PIV, LES, and Machine Learning"
Who: David L.S. Hung (Associate Dean for Graduate Education, John Wu & Jane Sun Chair Professor of Mechanical Engineering, University of Michigan-Shanghai Jiao Tong University Joint Institute)

Seminar Abstract:

As a key powertrain technology for automobiles, gasoline direct-injection (GDI) engine enables optimal stratification of in-cylinder air charge motion with direct fuel injection which impacts fuel spray dispersion, fuel-air mixture formation, and flame development, resulting in higher thermal efficiency with improved fuel economy and lower exhaust emissions. Using time-resolved particle image velocimetry (PIV) measurement and numerical large eddy simulation (LES) method, researchers can study the ensemble in-cylinder flow dynamics of multiple cycles as well as their cycle-to-cycle variation characteristics. However, it is still a daunting task to develop robust techniques for capturing highly transient 3D or simultaneous multi-plane in-cylinder flow-field data particularly for large number of engine cycles. In this seminar, a novel approach utilizing machine learning algorithms based on multi-cycle and multi-operating conditions of flow datasets extracted from PIV and LES is presented. Large datasets of PIV flow field images and LES databases can be employed for training machine learning algorithms and establishing the flow characteristic models. Well-trained deep learning neural network models are not only capable of efficiently and accurately predicting the flow-field movement, but also improving the temporal resolution of PIV measurements without decreasing the spatial resolution. In summary, this research provides valuable knowledge for quantifying GDI engine in-cylinder temporal and spatial flow dynamics by combining experimental diagnostics and numerical methods with machine learning algorithms to reveal the fundamental physics-based fluid dynamic mechanisms.

Speaker Biography:

David L.S. Hung is the Associate Dean for Graduate Education and John Wu & Jane Sun Chair Professor of Mechanical Engineering at the University of Michigan-Shanghai Jiao Tong University Joint Institute, Shanghai Jiao Tong University. He conducts research on new energy vehicle technologies and sustainable energy systems, with special focus on internal combustion engines, alternative powertrain systems, & renewable fuels for ground vehicles. He has received research funding from agencies including China’s Ministry of Education and NSF-China. He also received numerous teaching awards for his contributions to the Joint Institute’s reform and innovation in training mechanical engineering students. He was named a Top Ten SJTU “KoGuan Professor” in 2015. Prior to working in China, he was an associate professor in the Department of Mechanical Engineering at Michigan State University. Previously he held professional positions in the US at General Motors, Delphi, and Visteon Corporations where he researched and developed powertrain components for use in advanced gasoline engines. His research projects were funded by the Department of Energy, Michigan Economic Development Corporation, and other agencies.

He is currently the Editor-in-Chief, Asia of Atomization and Sprays, the journal of the Institute for Liquid Atomization and Spray Systems - International (ILASS). He is also a voting member of the SAE Gasoline Fuel Injection Standards Committee. He has received numerous professional awards, including ASME Internal Combustion Engine Division Outstanding Speaker Award, SAE Myers Outstanding Paper Award, SAE Henry Souther Standards Award, and SAE/InterRegs Standards and Regulations Award for Young Engineers for his contributions to develop engineering standards and regulations for the global automotive industry. He is a Fellow of SAE and a distinguished visiting fellow at the University of Oxford with funding from the Royal Academy of Engineering, United Kingdom.

He earned his B.S. with Distinction in mechanical engineering from Iowa State University and his M.S. & Ph.D. in mechanical engineering from Carnegie Mellon University.