Data-driven chemical kinetic models of operationally relevant jet fuels

Kim, Keunsoo

Permalink

https://hdl.handle.net/2142/115927 Copy

Description

Title

Data-driven chemical kinetic models of operationally relevant jet fuels

Author(s)

Kim, Keunsoo

Issue Date

2022-07-14

Director of Research (if dissertation) or Advisor (if thesis)

Lee, Tonghun

Doctoral Committee Chair(s)

Lee, Tonghun

Committee Member(s)

• Allen, Cody Michael
• Feng, Jie
• Matalon, Moshe
• Temme, Jacob

Department of Study

Mechanical Sci & Engineering

Discipline

Mechanical Engineering

Degree Granting Institution

University of Illinois at Urbana-Champaign

Degree Name

Ph.D.

Degree Level

Dissertation

Keyword(s)

• Sustainable Aviation Fuels (SAFs)
• Data-driven
• Chemical Kinetics
• Autoignition
• Combustion
• ignition delay

Abstract

This dissertation assesses the autoignition characteristics of operationally relevant fuels and develops chemical kinetic mechanisms based on a data-driven approach. Recent interest in sustainable aviation fuels (civilian) and single-fuel policies (military) in the US necessitates a detailed understanding of the effect of various jet fuels' properties on auto-ignition behavior. Fuels can dramatically differ in combustion and emissions characteristics due to fuel properties and chemical composition differences. This variation can come from differences in geographical locations or even the introduction of new alternative fuel blends into the common pipeline and just being different fuels. It is common knowledge that the Cetane number is a representative metric that summarizes combustion characteristics across a range of fuel relevant properties. However, there is no concrete relationship between cetane number and chemical reactivity. To investigate the impact of various fuels on combustion characteristics, two different sets of cetane number (CN) varied fuels: F-24/ATJ blended fuels and varied CN controlled fuels, are evaluated. These two experiments can explore the effect of CN and chemistry on autoignition properties with minimized atomization and physical properties, respectively. Assessing the ignition behavior of fuels is done by analyzing zero-dimensional chemical ignition delay data for subsequent analysis of gas-phase reactivity from rapid compression machine (RCM) and shock tube experiments at a compressed pressure of PC = 2 MPa and three different equivalence ratios in synthetic dry air, between 700 K and 1250 K. These conditions are of interest because they can cover a wide range of practical combustion operations, including internal combustion engines and gas turbines. Typically, the reactivity is linearly proportional to the cetane number. However, the results showed that derived cetane number (DCN) could not be used as the sole predictor of ignition delay across the range of thermodynamic conditions relevant to propulsion systems. This dissertation will directly and with more detail show the connection between DCN and chemical composition effects on ignition. Combustion and reaction pathways are affected mainly by the structure of chemical components through fuel breakdown, hydrogen abstraction, and isomerization. The main components of jet fuels can be divided into three groups: normal-alkane, cyclo-alkane, and iso-alkane. In the high-temperature regime, the branched structure of iso-alkanes results in the decomposition of fuel into tertiary radicals easily, but its stable states hinder oxidization. In the intermediate and low-temperature regimes, H-abstraction and radical isomerization in branched structures retards the ignition delay due to its high number of primary C-H bonds. With regard to blending ratio, a non-linear relationship between blending ratio and ignition characteristics has been observed. The reactivity of blends likely tends to follow the component of the blended fuel with higher reactivity. When the more reactive fuel undergoes the chain branching step in the mixture system, it will sufficiently provide OH radicals and heat to the system. The variation and/or similarities in ignition behavior observed amongst fuels arises from the inherent chemical structure of fuel components. To obtain chemical kinetic mechanisms for the fuels in this study, data-driven optimization methodology is employed, and ARLMech-HC-F24ATJ and ARLMech-HC-VariedCN were introduced for numerical simulation of the ignition process herein. The mechanisms follow the HyChem approach, which includes fuel-dependent lumped reactions and detailed C0-C4 chemistry based on combustion chemistry. Rather than modifying detailed small hydrocarbon chemistry, fuel-dependent lumped reactions are optimized toward empirical ignition delay data by genetic algorithm optimization of reaction coefficients. To express the non-linearity blending effects under intermediate and low-temperature chemistry, an alkylation (also known as co-oxidation) reaction was considered for the F-24/ATJ blend mechanism. Unlike F-24 and ATJ mechanisms, mechanisms for varied CN fuels were started from scratch. The lumped chemical equations were formulated to satisfy the HyChem theory and chemical composition analysis from GCxGC results. This approach can be used as the mechanism development methodology for next-generation 'drop-in' fuels. The newly introduced data-driven mechanisms based on these techniques show a good performance and more accurately represent the ignition behavior than mechanisms created without using these techniques. The developed mechanisms were also used for solving practical combustion problems to examine the fuels' effects on combustion characteristics. The results provided a qualitative understanding and overall prediction of ignition behaviors and were able to validate other published experimental results. The experimental data in this dissertation will provide a guideline and insight regarding the autoignition characteristics of various fuels: conventional hydrocarbon fuels and alternative fuels. Moreover, the suggested data-driven based chemical kinetic mechanisms and development methodology can be applied to real combustion problems with computational fluid dynamic simulations to evaluate the combustion appropriateness of newly introduced fuels.

Graduation Semester

2022-08

Type of Resource

Thesis

Copyright and License Information

Copyright 2022 Keunsoo Kim. All rights reserved.

Owning Collections