Date of Award
Master of Science (MS)
Dr. Robert Prucka
Dr. Qilun Zhu
Dr. Benjamin Lawler
Hybridization of heavy-duty on-road vehicles presents an opportunity to significantly reduce internal combustion engine emissions in real-world operation. These gains can be realized through the coordination of the electric drive, engine, and aftertreatment systems. Accurate Multiphysics models of all powertrains sub-systems are required to achieve the goal of reduced emissions. This research aims to develop a model of a highly complex diesel engine aftertreatment system. This study focuses on utilizing transient data for calibration and validation of the aftertreatment system and reducing the run time when compared to real-time experiments. The calibration focuses on two physical phenomena, thermal behavior and chemical kinetics. Once a base model is set up, the calibration parameters are optimized using an accelerated genetic algorithm for factors that contribute to the reaction rates and the exhaust gas temperature. The research only utilizes data from transient engine experiments to better automate and speed-up the calibration process over traditional methodologies.
The model setup ensures that it is fast-running, with ten times speed-up as compared to real-time. The model is capable of predicting and matching combined error for and concentration on a cumulative basis under 9.8% and 1% for the experimental data for cold FTP and hot FTP, respectively. The results of the model also predict close trends with the temperature profiles and have a close match with the tailpipe emission species concentration over a cumulative basis but fails to capture some transient behavior. The model results are also evaluated to identify the leading cause for the error so the model can be improved for further development. The model has the capability to generate results for the aftertreatment for further research.
Srivastav, Uday, "Multiphysics Diesel Aftertreatment System Modeling for Reduced Emissions from Hybrid Electric Heavy-Duty Powertrains" (2023). All Theses. 4145.