arrow
返回

Error-controlled kinetics reduction based on non-linear optimization and sensitivity analysis

delete2019-02-01
delete13
PRE
AI
T
Thomas Jaravel *
H
Hao Wu
M
Matthias Ihme
DOI:10.1016/j.combustflame.2018.11.007delete
delete原文链接
delete原文求助
delete分享
delete收藏
摘要

摘要

En 中文
Efficient reduction techniques are necessary for the construction of compact chemical kinetic mechanisms to enable the application to large-scale simulations. In this work, a new formulation is proposed to derive skeletal mechanisms by species elimination. The proposed method relies on a gradient-based non-linear optimization (NLO) method, allowing for direct error control on user-defined quantities of interest (Qols). The key idea consists in formulating the species elimination by relaxing an integer-optimization problem into a continuous optimization problem that is solved iteratively. A species-targeted sensitivity analysis (SA) formulation is presented that can be combined with the optimization method to reduce the overall computational complexity of the procedure. After illustrating its principle, the NLO procedure is applied to different fuels of increasing chemical complexity, including methane, n-dodecane, and a four-component kerosene surrogate. Direct comparisons are performed with the directed relation graph method with error propagation (DRGEP) and SA. For the reduction of a detailed methane mechanism, consistency with DRGEP and SA is demonstrated, and the NLO procedure is shown to generate smaller mechanisms compared to the other two methods. In application to larger hydrocarbon and multicomponent transportation fuels, it is shown that significantly smaller mechanisms are obtained with the NLO procedure compared to the other approaches. (C) 2018 The Combustion Institute. Published by Elsevier Inc. All rights reserved.
Keyword:
Reduced kinetics
Combustion
Optimization
AI总结

AI总结

对已上传原文的论文进行重点信息的提取,主要内容包括:简要概述、研究摘要、背景介绍、关键亮点、图文解析、展望与总结。

期刊

Combustion and Flame 封面图
Combustion and Flame
IF:
6.2
论文数:
9.7K
被引数:
4.2W

机构

S
Stanford University
学者数:
9.6W
论文数: 8.2W
被引数: 17.0W
引用论文

引用论文

Mixed-integer nonlinear optimization
err2013-04-02
err508
PREAI
errBelotti, Pietro; Kirches, Christian; Leyffer, Sven; Linderoth, Jeff; Luedtke, James; Mahajan, Ashutosh
err分享
err收藏
The experimental evaluation of a methodology for surrogate fuel formulation to emulate gas phase combustion kinetic phenomena
err2012-04-01
err419
PREAI
errDooley, Stephen; Won, Sang Hee; Heyne, Joshua; Farouk, Tanvir I.; Ju, Yiguang; Dryer, Frederick L.; Kumar, Kamal; Hui, Xin; Sung, Chih-Jen; Wang, Haowei; Oehlschlaeger, Matthew A.; Iyer, Venkatesh; Iyer, Suresh; Litzinger, Thomas A.; Santoro, Robert J.; Malewicki, Tomasz; Brezinsky, Kenneth
err分享
err收藏
err分享
err收藏
Dynamic stiffness removal for direct numerical simulations
err2009-08-01
err122
PREAI
errLu, Tianfeng; Law, Chung K.; Yoo, Chun Sang; Chen, Jacqueline H.
err分享
err收藏
学者 查看更多内容