Brown Bag Seminar
Friday, April 21
11:00 a.m. -12:00 p.m.
Weber, Classroom II
Presenters:
Hugh (Ka Yui) Chen
Michael Keraga
Yuga Ojima
Hugh (Ka Yui) Chen
Title:
Tunable Diode Laser Spectroscopy
Abstract:
The Tunable Diode Laser Spectroscopy Lab is an experimental branch for the graduate aerospace propulsion lab, specifically focused on diode laser spectroscopy and other relevant applications. The lab offers students the opportunity to participate in designing the experiment procedure through the construction of brand-new diode laser testing rigs, calibration of equipment, data acquisition along with analysis, and more. Students are able to gain a comprehensive understanding of the fundamental theorem of diode laser technologies, and its various applications in aerospace.
Advisor:
Professor Adam Steinberg
Michael Keraga
Title:
Determination of Flame Lift-Off Length and Ignition Delay Time of a Direct-Injection Diesel Spray Under Various Chamber Conditions
Abstract:
Under quiescent conditions, the flame of a direct-injection diesel spray will stabilize at an axial location downstream of the injector face. The distance from the flame stabilization point to the injector face is defined as the flame lift-off length. Additionally, the time interval between the start of injection and combustion is defined as the ignition delay time. With the growing demand for low emission engines, intensive research into the auto-ignition combustion process in diesel engines suggests that these two quantities play a significant role in emission formation processes. Recent studies have shown that measuring ultraviolet light emitted by OH* (OH in its excited state) chemiluminescence is the optimal imaging technique in determining lift-off length. Using a CCD video camera, this study will use line-of-sight data of OH* chemiluminescence to determine lift-off length and ignition delay time of a direct-injection diesel spray in a fixed-volume combustion chamber at various chamber conditions.
Advisor:
Professor Tim Lieuwen
Yuga Ojima
Title:
Machine Learning Approaches for Categorizing Daily Interior Temperature Trends in Campus Buildings
Abstract:
The research presents an algorithm pipeline that employs unsupervised machine learning and supervised deep learning techniques to detect daily interior temperature trends in campus buildings. The lack of a reliable classification approach for daily interior temperature data from diverse campus buildings had hampered advancements in sustainable infrastructure. The study's findings provide valuable insights into interior temperature behavior, including seasonal changes, and can help optimize building energy efficiency and sustainability. The pipeline's implementation could be extended to different kinds of buildings to uncover distinctive temperature patterns and opportunities for energy efficiency improvements. The research demonstrates the potential of machine-learning techniques and time-series data to address classification problems in infrastructure management and pave the way for future research in this area.
Advisors:
Dr. Jung-Ho Lewe, Professor Dimitri Mavris