Dynamics of laminar ethylene lifted flame with ozone addition

doi:10.1016/j.proci.2020.09.027

Proceedings of the Combustion Institute

Volume 38, Issue 4, 2021, Pages 6773-6780

https://doi.org/10.1016/j.proci.2020.09.027 Get rights and content

Abstract

The effect of ozone (O₃) addition on ethylene (C₂H₄) non-premixed jet flames is investigated. Stable C₂H₄ lifted flames are established with oxygen/nitrogen co-flow at reduced oxygen content conditions. It is observed in the experiments that after O₃ addition, the flame liftoff height could either increase or decrease, depending on the initial value of the flame liftoff height before O₃ is added. Formaldehyde (CH₂O) planar laser-induced fluorescence (PLIF) measurement shows that prompt ozonolysis reaction between C₂H₄ and O₃ produces large amounts of CH₂O upstream of the flame. In contrast to previous studies of O₃ addition on lifted flames—with saturated hydrocarbon fuels in which O₃ decomposition dominates—the ozonolysis reaction between C₂H₄ and O₃ changes the chemical composition of flow even at room temperature. Such chemical reaction causes the simultaneous increase of both the triple flame propagation speed of lifted flame and the axial jet velocity along the stoichiometric contour, which also therefore changes the dynamic balance between these two values to stabilize the flame. While the increase of triple flame propagation speed tends to decrease the flame liftoff height, the increase of axial jet velocity along the stoichiometric contour tends to have the opposite effect. A competing kinetic-dynamic process forms, and the final location of the flame depends on the degree of ozonolysis reaction, which is determined by the initial flame liftoff height before O₃ is added.

Introduction

Ozone (O₃) addition has great potential in actively enhancing and controlling combustion [1], [2], [3]–4]. Considerable increases in flame propagation speeds have been observed with O₃ addition. These experiments include those using both non-premixed laminar lifted flame [5] and premixed Bunsen burner [6], [7], [8], [9]–10] configurations with saturated hydrocarbon fuels, such as methane (CH₄), propane (C₃H₈) and syngas (H₂/CO). It is well accepted that the production of O atoms through O₃ decomposition within pre-heat zone plays a vital role in the observed flame speed enhancement, while the reactions between O₃ and saturated fuels are virtually negligible. However, with unsaturated hydrocarbon fuels, the addition of O₃ could lead to an entirely different outcome, owing to the prompt exothermic reaction between O₃ and fuel, i.e., the ozonolysis reaction. The difference between the reaction rate constants of alkene and alkane fuels with O₃ could be as large as $k_{alkene, O_{3}} / k_{alkane, O_{3}} \sim 1 0^{6}$ [11]. Recently, Gao et al. [11] observed direct auto-ignition using a non-premixed ethylene (C₂H₄) jet with O₃ addition in an oxidizer co-flow at room temperature and pressure conditions without an external ignition source. Significant enhancement on flame propagation by continuous formation of auto-ignition kernels upstream was observed. This unique phenomenon indicated that the ozonolysis reaction could be considered as an alternative means to modify combustion processes at low-temperature conditions. Different from conventional fuel oxidation which is confined in the flame zone, the high reactivity between unsaturated hydrocarbon and O₃ renders immediate reaction and heat release far upstream upon mixing, and therefore affecting the flame both kinetically and dynamically. Such a feature (combustion region) has not been studied before.

The influence of O₃ addition on laminar flame speed (S_L) of C₂H₄ was investigated in premixed configurations [10,12,13]. It was reported that at atmospheric conditions, adding O₃ would decrease S_L, while at sub-atmospheric conditions, enhancement in S_L was measured. Based on kinetic simulations, Gao et al. [10] concludes that the reaction of C₂H₄/O₃ has a positive effect on S_L by producing reactive species such as formaldehyde (CH₂O) and hydrogen (H₂). Meanwhile, the premixed configuration induces continuous heat loss to the burner wall which decreases the total enthalpy of the system, thus lowering S_L. At low-pressure, the ozonolysis reaction is suppressed given a limited residence time upstream of the flame zone. Therefore, S_L would be increased via O₃ decomposition, similar to the case with alkanes. Nevertheless, the effect of the ozonolysis reaction on flame dynamics has not been investigated in detail.

In this work, the effect of O₃ addition on C₂H₄ flame dynamics is studied using a non-premixed laminar co-flow jet flame burner. The lifted laminar jet flame has been studied extensively [14], [15], [16], [17], [18], [19]–20]. The stabilization of lifted flames is achieved by the dynamic balance between the propagation speed of the triple flame (S_tri), which is located at the base of the lifted flame, and the axial flow velocity (u_st) just upstream of the flame along the stoichiometric contour. Therefore, the liftoff height (H_L) of the flame could be a sensitive indicator of the change in either S_tri or upstream condition. In our experiments, by tuning the co-flow composition, a stable C₂H₄ lifted flame is obtained in the laminar regime. Various flame dynamic behaviors are observed as a function of fuel jet velocity (u_f). Specifically, in the low u_f region, H_L decreases with O₃ addition, while instead it increases in high u_f region. Planar laser-induced fluorescence (PLIF) of CH₂O and numerical simulations are performed to understand the highly coupled kinetic and dynamic process.

Section snippets

Experiment setup

Figure 1 schematically illustrates the experimental system, which consists of a laminar jet flame burner and the PLIF set-up. All experiments are undertaken at room temperature and pressure (T = 298 K, P = 101 kPa). The burner used here is the same one used in the study by Ombrello et al. [5,21] comprising a slender, aerodynamically shaped quartz fuel nozzle with 0.69 mm inner diameter at the exit on the tip, and a cylindrical quartz shield with inner diameter of 90 mm and length of 460 mm to

Computational framework

Numerical simulations are conducted using ANSYS Fluent [22] with the SIMPLEC algorithm. An axisymmetric 2-D computational domain is adopted with a size (axial × radial) of 450 mm × 45 mm. A gravitational acceleration of 9.79 m/s² is given in the negative axial direction so that buoyancy effect could be accounted for. The built-in CHEMKIN-CFD solver of laminar finite-rate chemistry is enabled. The detailed USC Mech II model [23] is firstly reduced (for shorter computational time) using Global

Stable lifted flames without O₃ addition

Stable C₂H₄ laminar lifted flames are obtained with pure fuel jet and the O₂/N₂ mixture as co-flow. Figure 2 shows a series of photographs of stable C₂H₄ lifted flames at different u_f but constant co-flow velocity, U_CO = 0.013 m/s, and constant oxidizer composition of 12.7% O₂ + 87.3% N₂. The u_f changes from 2.98 m/s as an attached diffusion flame to finally 5.23 m/s.

As the flame detaches from the fuel nozzle, a clear triple flame structure appears with a relatively weak lean premixed wing and

Conclusions

In this study, the effect of O₃ addition on non-premixed C₂H₄ lifted jet flames is investigated, using experiments, including PLIF of CH₂O, and numerical simulations. Stable, laminar lifted C₂H₄ flames are established in a co-flow geometry and a wide range of liftoff height is obtained. Numerical simulations and experimental results are in qualitative agreement. With O₃ addition in the co-flow, the change of flame liftoff height shows sensitivity on its initial value. If the initial flame

Declaration of Competing Interest

None.

Acknowledgments

WS acknowledges Prof. Yiguang Ju from Princeton University for the loan of the burner. This work is supported by the Air Force Office of Scientific Research (FA9550-16-1-0441)

References (31)

T. Ombrello et al.
Combust. Flame
(2010)
Z.H. Wang et al.
Combust. Flame
(2012)
X. Liang et al.
Int. J. Hydrogen Energy
(2013)
T.M. Vu et al.
Combust. Flame
(2014)
X. Gao et al.
Combust. Flame
(2015)
S. Chung et al.
Combust. Flame
(1991)
B. Lee et al.
Combust. Flame
(1997)
Y.-C. Chen et al.
Combust. Flame
(2000)
S. Ghosal et al.
Combust. Flame
(2001)
S. Chung
Proc. Combust. Inst.
(2007)

T. Ombrello et al.

Combust. Flame

(2010)

X. Gao et al.

Combust. Flame

(2016)

A. Mohammadi, H. Kawanabe, T. Ishiyama, M. Shioji A. Komada, Study on Combustion Control in Natural-Gas PCCI Engines...

H. Nishida et al.

J. Propuls. Power

(2006)

I. Ekoto et al.

SAE Int. J. Fuels Lubr.

(2018)

Cited by (9)

Knock probability determination employing convolutional neural network and IGTD algorithm
2023, Energy
This study presents a novel method based on the convolutional neural network to evaluate knock probability. In this way, lots of data sets are extracted from the real driving conditions of a turbocharged gasoline vehicle. The trained network shows 92 % accuracy in knock detection and 87 % accuracy in normal combustion recognition. This trained network is then used to forecast the knocking probability of the vehicle in different operating conditions. The results revealed that the exhaust gas temperature sensor considerably affects the knock recognition precision by 3 %. Most of the network errors in knock probability determination would be associated with the full load-high speed region and the network errors in normal combustion detection would be related to the low-end torque zone. It is also found that the knock probability decreases with the vehicle's speed increase and its probability at the constant vehicle speed would be higher at low engine rotational speed. This new approach can predict engine knock zones for different vehicle parameters and can be implemented by the engine control unit to suppress knock before happening in an active manner rather than the current passive method.
Robust key parameter identification of dedicated hybrid engine performance indicators via K-fold filter collaborated feature selection
2023, Engineering Applications of Artificial Intelligence
Dedicated hybrid engine technology using auxiliary electronic components has been proven as an energy-saving solution to public concerns about energy consumption and carbon emissions. This paper proposes a generic approach of K-fold filter-collaborated feature selection (KFFC-FS) to robustly identify the key parameters of three engine performance indicators, i.e., volumetric efficiency, thermal efficiency, and fuel consumption. By using this approach, five filters are collaborated to provide a robust rank of feature importance and avoid the feature overestimation caused by the single filter. Meanwhile, the K-fold cross validation method is introduced to avoid random precision issues and overfitting, further enhancing the robustness of key parameter identification for the independent engine performance indicators. In this research, the modelling data is collected from an experimental test bench with a BYD 1.5L gasoline engine. Under the basics of the studied three engine performance indicators by using a multiple-layer perceptron network, the proposed approach further reduces by at least 10.3% root-mean-square error (RMSE) and at least 30% reduction of the model inputs.
Characteristic time scale as optimal input in Machine Learning algorithms: Homogeneous autoignition
2023, Energy and AI
Considering temporally evolving processes, the search for optimal input selection in Machine Learning (ML) algorithms is extended here beyond (i) the readily available independent variables defining the process and (ii) the dependent variables suggested by feature extraction methods, by considering the time scale that characterizes the process. The analysis is based on the process of homogeneous autoignition, which is fully determined by the initial temperature T(0) and pressure p(0) of the mixture and the equivalence ratio $ϕ$ that specifies the initial mixture composition. The aim is to seek the optimal input for the prediction of the time at which the mixture ignites. The Multilayer Perceptron (MLP) and Principal Component Analysis (PCA) algorithms are employed for prediction and feature extraction, respectively. It is demonstrated that the time scale that characterizes the initiation of the process $τ_{e} (0)$ , provides much better accuracy as input to MLP than any pair of the three independent parameters T(0), p(0) and $ϕ$ or their two principal components. Indicatively, it is shown that using $τ_{e} (0)$ as input results in a coefficient of determination R $^{2}$ in the range of 0.953 to 0.982, while the maximum value of R $^{2}$ when using the independent parameters or principal components is 0.660. The physical grounds, on which the success of $τ_{e} (0)$ is based, are discussed. The results suggest the need for further research in order to develop selection methodologies of optimal inputs among those that characterize the process.
Combustion machine learning: Principles, progress and prospects: Combustion machine learning
2022, Progress in Energy and Combustion Science
Citation Excerpt :
Convolutional and recurrent layers in these networks enable the direct use of complex datasets such as visual images and time-series data as feature sets. Kuzhagaliyeva et al. [638] reported that deep LSTM networks make accurate (75%) predictions of engine preignition when fed with time-series data extracted from low-cost pressure sensors. In the same work, 1D convolutional networks produced more accurate (79%) predictions when fed with datasets preprocessed via PCA.
Progress in combustion science and engineering has led to the generation of large amounts of data from large-scale simulations, high-resolution experiments, and sensors. This corpus of data offers enormous opportunities for extracting new knowledge and insights—if harnessed effectively. Machine learning (ML) techniques have demonstrated remarkable success in data analytics, thus offering a new paradigm for data-intense analyses and scientific investigations through combustion machine learning (CombML). While data-driven methods are utilized in various combustion areas, recent advances in algorithmic developments, the accessibility of open-source software libraries, the availability of computational resources, and the abundance of data have together rendered ML techniques ubiquitous in scientific analysis and engineering. This article examines ML techniques for applications in combustion science and engineering. Starting with a review of sources of data, data-driven techniques, and concepts, we examine supervised, unsupervised, and semi-supervised ML methods. Various combustion examples are considered to illustrate and to evaluate these methods. Next, we review past and recent applications of ML approaches to problems in combustion, spanning fundamental combustion investigations, propulsion and energy-conversion systems, and fire and explosion hazards. Challenges unique to CombML are discussed and further opportunities are identified, focusing on interpretability, uncertainty quantification, robustness, consistency, creation and curation of benchmark data, and the augmentation of ML methods with prior combustion-domain knowledge.
Introduction
2022, Artificial Intelligence and Data Driven Optimization of Internal Combustion Engines
A comprehensive review of mechanical fault diagnosis methods based on convolutional neural network
2024, Journal of Vibroengineering

View all citing articles on Scopus

View full text

Dynamics of laminar ethylene lifted flame with ozone addition

Abstract

Introduction

Section snippets

Experiment setup

Computational framework

Stable lifted flames without O3 addition

Conclusions

Declaration of Competing Interest

Acknowledgments

Combust. Flame

Combust. Flame

Int. J. Hydrogen Energy

Combust. Flame

Combust. Flame

Combust. Flame

Combust. Flame

Combust. Flame

Combust. Flame

Proc. Combust. Inst.

Combust. Flame

Combust. Flame

J. Propuls. Power

SAE Int. J. Fuels Lubr.

Stable lifted flames without O₃ addition