Detailed modeling of polycyclic aromatic hydrocarbon and soot formation employing theory and experiments
Final Report Abstract
My project on detailed modeling of polycyclic aromatic hydrocarbon and soot formation lasted 24 months. I focused on computational work, because this is the area of expertise, I could contribute best to my host group’s research efforts. A major effort in the Green Group is the development of the Reaction Mechanism Generator (RMG) open-source software. This includes improving the RMG database with key thermochemical and rate parameters for modeling PAH formation in acetylene and Jet A1 surrogate pyrolysis. While the focus of these additions was on PAH formation pathways, initiation reactions for acetylene pyrolysis were also considered. Notable conclusions include the importance of acetone in accurately predicting formation of odd-carbon species, the remarkably low contribution of acetylene dimerization to vinylacetylene or diacetylene, and the dominance of the HACA mechanism in the formation pathways to all PAH species in the acetylene model. Therefore, this work represents an important milestone for the RMG software in automatically generating a mechanism predicting up to pyrene. The results of this work are used by our industry partner Mitsubishi Heavy Industries to design advanced propulsion technology. However, there are still improvements which can be made to both the RMG software and the chemical mechanism. Notably, the need for more kinetics data for elementary reaction pathways to PAHs and the generation of PAH formation models is still computationally challenging due to the sheer number of potential species and reactions which RMG evaluates. With the concerted efforts outlined in this work, we also have identified critical bottlenecks and provided scalability solutions to make computer-aided reaction mechanism construction a tool with practical usefulness in the generation of predictive detailed reaction mechanisms. It was shown that bottlenecks for the overall performance of RMG simulations are dynamic in nature and depend on the state of the simulation and the methods used for the calculation of chemical data. A three-pronged approach consisting of code optimization, algorithm heuristics, and parallel computing was employed to accelerate the essential phases of the rate-based enlargement procedure. The parallel computing option allows RMG to make use of several CPU’s within a shared memory machine and therefore allows, together with other scalability strategies, for speeding up a mechanism generation simulation by up to an order of magnitude. Even though the scalability strategies created new opportunities for detailed modeling of diverse real-world processes, generating a chemical kinetic mechanism with RMG including formation pathways for three and four ring PAHs for combustion processes was still challenging. Therefore, we focus on generating chemical kinetic mechanisms for pyrolysis, neglecting combustion chemistry involving oxygen. Due to the work that still has to be done on generating a combustion mechanism including PAHs up to four rings with RMG, I decided to use an established chemical kinetic mechanism to investigate the use of oscillating jet flames in a co-flow for developing soot models for practical applications, in collaboration with the University of Adelaide. The potential for soot model improvement is analyzed by correlating the transient gasphase species, polycyclic aromatic hydrocarbon, and soot formation or destruction behavior in a co-flowing jet flame. A need for improved soot nucleation, condensation, and oxidation formulations is identified. It is further found that the soot number density distribution in mixture fraction space is comparable to the transitional turbulent flame regime dominated by Kelvin-Helmholtz rollers. The study identifies shortcomings of a high-fidelity soot model that is developed in steady laminar flames but applied for engines working in unsteady and often turbulent combustion environments. Therefore, this study provides a stepping stone to improve high-fidelity soot models towards reliable application in real-world applications like aero engines.
Publications
- Scalability strategies for automated reaction mechanism generation, Computers and Chemical Engineering, 131 (2019) 106578
A. Jocher, N. M. Vandewiele, K. Han, M. Liu, C. W. Gao, R. J. Gillis, W. H. Green
(See online at https://doi.org/10.1016/j.compchemeng.2019.106578)
