Prof. Weiwei Cai

School of Mechanical Engineering, Shanghai Jiao Tong University, China

E-mail: cweiwei@sjtu.edu.cn

Bio

Professor Weiwei Cai is a professor at the School of Mechanical Engineering, Shanghai Jiao Tong University. He received his Ph.D. in Mechanical Engineering from Clemson University in 2010. Before joining SJTU in 2015, he was a Marie Curie Research Fellow at the University of Cambridge and a postdoctoral researcher at Virginia Tech. His research interests lie at the intersection of computational optics and combustion diagnostics, developing novel optical measurement systems with multi-dimensional, high spatiotemporal resolution, multi-physics, and miniaturization capabilities. His work has been published in top-tier journals including Science (3 papers), Nature Electronics, Progress in Energy and Combustion Science (2 papers), Journal of Fluid Mechanics, Combustion and Flame, and many others, with over 100 papers in total. Prof. Cai has received numerous prestigious awards, including the Masao Horiba Award Nomination, the Wu Zhonghua Outstanding Young Scholar Award from the Chinese Society of Engineering Thermophysics, and the Erlangen Young Scientist Award from the Max Planck Institute for the Science of Light. He has been consecutively listed in Stanford's Top 2% Global Scientists ranking for six consecutive years (2020-2025). He currently serves as an editorial board member for Measurement Science and Technology, Engineering Thermophysics, and other journals.

Title

Computational Metrology for Flow and Combustion Diagnostics

Abstract

This talk outlines our latest advancements in computational metrometry for flow and combustion diagnostics, highlighting a shift towards intelligent, computation-powered measurement techniques. We focus on three key research thrusts. First, we present a miniaturized computational spectrometer, implemented on a fully integrated 576-Kb memristor chip. This hardware-software co-designed system achieves software-equivalent spectral reconstruction accuracy while offering 26.5 times faster speed and 162.7 times greater energy efficiency than state-of-the-art alternatives. Second, we introduce a tensor decomposition-based, four-dimensional background-oriented schlieren tomography technique. By organizing time-resolved data into an X-Y-Z-T spatiotemporal tensor and integrating neural networks, this method enables high-speed, high-fidelity reconstruction of volumetric refractive index fields with significantly reduced memory usage and processing time. Third, we demonstrate a Neural Refractive Index Field (NeRIF) approach for volumetric flow visualization. NeRIF uses a specially trained neural network to implicitly represent the flow field, substantially improving reconstruction accuracy and spatial resolution while reducing computational costs by orders of magnitude. Collectively, these developments provide unprecedented capabilities for probing complex phenomena in aerospace propulsion and energy systems.