Abstract
Moderate or Intense Low-oxygen Dilution (MILD) combustion is considered as one of the most promising novel combustion technologies, as it ensures high efficiency and very low emissions (NOx and CO). Because of the high level of dilution, the system reactivity is reduced and the chemical time scale is increased compared to conventional flames. Therefore, combustion models accounting for finite-rate chemistry are needed to study the characteristics of such flames. Reactor based models, such as the Partially Stirred Reactor and the Eddy Dissipation Concept models have been successfully used to model MILD combustion. This article describes recent progress and developments in the application of reactor based models for the simulation of the jet-in-hot-coflow burners that emulate MILD combustion. The main objective is to provide an overview about the current state of the art of reactor based models for turbulence-chemistry interactions in MILD regime and outline future prospects for the further development of such models. The literature acknowledges both Reynolds Averaged Navier Stokes and Large Eddy Simulations studies, with various operating conditions as well as different fuels. The results indicate that it is necessary to include both the mixing and chemical time scales explicitly in the combustion model formulation. Because of the distributed reaction area, according to recent investigations, Large Eddy Simulation grid can be sufficient to resolve the MILD combustion reacting structures. The present review underlines the importance of finite rate chemistry in MILD combustion simulations, as well as of providing reliable estimation of the characteristic time scales.
Introduction
Facing the current challenges of air pollution and energy shortage, it is urgent to develop fuel flexible, efficient and environmentally friendly combustion technologies. Novel combustion technologies with low emissions, high efficiency and fuel flexibility have become essential under these challenges (Li, 2019). In this context, one promising technology in energy production and manufacturing is Moderate or Intense Low oxygen Dilution (MILD) combustion (Wünning and Wünning, 1997; Cavaliere and de Joannon, 2004; de Joannon et al., 2012).
MILD combustion is established by diluting the fresh reactants with the combustion products and preheating the charge above the self-ignition temperature of the fuel (de Joannon et al., 2012). The high dilution is responsible for the widening of the reaction zone, while the mixing with hot exhaust gases ensures that reaction takes place also outside of the flammability limits. In terms of characteristic scales, the dilution lowers the oxygen concentration and smooths the temperature peaks. As a result, the chemical time scale increases and the strong interaction between chemistry reaction and mixing makes the investigation of such flames more challenging than conventional combustion regimes. The availability of validation data is then crucial to advance the current understanding of MILD combustion and further push its implementation in industrial applications. To decrease the influence of geometric complexity encountered in practical devices, simplified lab-scale axis-symmetric jet burners are generally used to emulate MILD conditions—for example, the jet-in-hot-coflow (JHC) burner (Dally et al., 2002; Oldenhof et al., 2010).
The JHC burner features a central jet and a secondary burner providing hot exhaust products as a coflow, reproducing the flue gas recirculation. The oxygen level in the coflow is highly diluted by adding nitrogen and it is generally controlled below 10 by mass or volume (Dally et al., 2002; Medwell et al., 2007; Oldenhof et al., 2010; Medwell and Dally, 2012; Ye et al., 2015a; Ye et al., 2017). The literature acknowledges a large number of studies on JHC burners, both experimental and numerical (Dally et al., 2002; Oldenhof et al., 2010; De et al., 2011; Medwell and Dally, 2012; Evans et al., 2015; Parente et al., 2015; Li et al., 2017). Both gaseous (Dally et al., 2002; Medwell et al., 2007; Oldenhof et al., 2010; Medwell and Dally, 2012; Mardani and Karimi Motaalegh Mahalegi, 2019), simple hydrocarbon fuels and liquid (Ye et al., 2015a; Ma et al., 2016; Ye et al., 2017;Ye et al., 2018; Mahalegi and Mardani, 2019), oxygenated or long-chain alkane fuels have been used as fuel. The liquid, long-chain alkane fuels show different characteristics (Ye et al., 2015a; Ye et al., 2017) under MILD conditions, compared to gaseous fuels.
Numerical investigations on the JHC burners were carried out mainly using Reynolds Averaged Navier-Stokes (RANS) (Christo and Dally, 2005; Frassoldati et al., 2009; Mardani et al., 2010; Shabanian et al., 2012; Parente et al., 2015; Evans et al., 2015; Aminian et al., 2016; Mardani, 2017; Evans et al., 2017; Chen et al., 2017; Li et al., 2017; Li et al., 2018b) simulation and Large Eddy Simulation (LES) (Afarin et al., 2011; Ihme and See, 2011; Ihme et al., 2012; Li et al., 2019). LES can capture more faithfully the flow features while RANS is still important for preliminary investigations, because of its reduced computational cost. However, despite high computational efficiency of RANS simulation, non-equilibrium phenomena are not well captured by steady-state assumptions. To this end, Large Eddy Simulation (LES) can provide superior results with respect to RANS.
Due to the presence of the coflow and the intense low-oxygen dilution, combustion in JHC burner is characterised by a relatively low Damköhler number the ratio of the mixing to chemical time scale). The interactions between the chemical reaction and fluid dynamics has therefore become more important. Finite-rate chemistry models with detailed chemical mechanisms are necessary. On one hand, combustion models for conventional flames usually rely on the assumption of time-scale separation (i.e., steady flamelets and related models). These models are computationally efficient, while the thermo-chemical space accessible in the numerical simulations is constrained. Furthermore, huge effort could be spent on the pre-processing steps, especially when dilution level in MILD regime and heat losses are considered (Lamouroux et al., 2014; Locci et al., 2014). On the other hand, the use of transported PDF methods (Haworth and Pope, 2010) appears still computationally prohibitive, especially for practical combustion systems. Reactor based models, including the Partially Stirred Reactor (PaSR) (Chomiak, 1990; Golovitchev and Chomiak, 2001) approach and the Eddy Dissipation Concept (EDC) (Magnussen, 1981; Gran and Magnussen, 1996; Magnussen, 2005) model, have been extensively used in the past years because of their affordable treatment of detailed chemistry, showing promise for modelling MILD combustion.
Reactor Based Models
Both the Partially Stirred Reactor (PaSR) approach and the Eddy Dissipation Concept (EDC) model are based on the assumption that each computational cell can be separated into two zones. Th reacting zone is generally referred to as “fine structures” (Magnussen, 1981; Chomiak, 1990; Gran and Magnussen, 1996; Golovitchev and Chomiak, 2001; Magnussen, 2005) while the non-reacting one is the “surrounding fluid”. Combustion takes place in the fine structures in contact with a surrounding fluid. The average reaction rate in a cell is then the results of an exchange between the fine structure and surrounding fluid. The fine structures are modelled as Perfectly Stirred Reactors (PSR) or Plug Flow Reactors (PFR). A conceptual drawing of the PaSR and EDC model is presented in Figure 1.
Figure 1
Eddy Dissipation Concept Model
The EDC model is based on a cascade model (Gran and Magnussen, 1996) providing the mass fraction of the fine structures, , and the mean residence time of the fluid within the fine structures , as a function of the flow characteristic scales:
In Eq. 1 and Eq. 2, ν is the kinematic viscosity, and are model constants (Gran and Magnussen, 1996). The mean reaction rate (source term in the species transport equation) is expressed as (Gran and Magnussen, 1996):
The term in Eq. 3 denotes the mean mass fraction of the species s between the fine structures and the surrounding fluid and is the mass fraction of species s in the fine structures. The reacting fraction of fine structures is denoted with χ and it is often set to unity (Lewandowski and Ertesvåg, 2018; Lewandowski et al., 2020b). In Eq 3, M = 2 and N = 3. In the other two version of the EDC model (Magnussen, 1981, 2005), the combinations M = 3/N = 3 and M = 2/N = 2 have been used.
Partially Stirred Reactor Model
In the PaSR model (Chomiak, 1990; Golovitchev and Chomiak, 2001), the parameter κ is used to represent the mass fraction of the reaction zone (fine structures) in the computational cell. It can be estimated as (Chomiak, 1990; Golovitchev and Chomiak, 2001; Kärrholm, 2008):where and are the characteristic chemical and mixing time scales in each cell, respectively. The mean source term provided to the species transport equation can be expressed as:where represents the residence time in the reactive structure. In order to get , a system of ODE is solved for the reacting structures.
Finite-Rate Chemistry Effects
In both models, the mean mass fraction is obtained by solving transport equations for the reactive scalars induced in the kinetic mechanism. The mass fraction of each species inside the fine structures is estimated assuming that the latter are Perfectly Stirred Reactors (PSR), which allows to use detailed chemical mechanisms (Gran and Magnussen, 1996):where is the formation rate of species s and is the species mass fraction in the surrounding fluid (Gran and Magnussen, 1996). Alternatively, the fine structures can be modelled as Plug Flow Reactors (PFR), evolving over a characteristic time equal to :
Discussion
Successful predictions of JHC flames are reported in the literature. In the context of RANS simulation using EDC model, over-predictions of temperature were observed using standard EDC constants of and , as reported by different authors (De et al., 2011; Aminian et al., 2012; Shabanian et al., 2012; Evans et al., 2015; Li et al., 2017; Mardani, 2017; Li et al., 2018b). Aminian et al. (Aminian et al., 2012) performed RANS simulation on the JHC burner with CH/H as fuel. In the experiments, the co-flow oxygen level was adjusted to 3, 6 and 9 with the addition of N. The fuel jet Reynolds number was 10,000. They found out that the standard EDC over-predicts the peak temperature downstream of the burner (120 mm from the burner exit). They concluded that residence time in the fine structures should be increased, to decrease the average reaction rate. To this end, the value of was increased to 1.5 and 3.0. It was finally concluded that was the best option for both co-flow oxygen levels, i.e., 3 and 9.
De et al. (2011) studied Delft JHC with Dutch natural gas as a fuel, numerically. Under-prediction of the flame lift-off height was observed with the EDC model, while the flame temperature were generally over-estimated. The authors proposed to change not only the time scale constant , as done by Aminian et al. (2012), but also , to directly impact the mass fraction of reacting structures. Based on a sensitivity analysis, the authors finally selected or ( and are not changed at the same time).
The modified constant value used by De et al. was adopted by Shabanian et al., (2012), when simulating the JHC burner fed with ethylene. The modified k-ϵ model reported by Aminian et al. (2012) was used. They compared three different combustion models in their research work: the steady flamelet model, EDC and transported PDF. Results obtained with the adjusted EDC coefficients and transported PDF showed superior performances compared to the standard steady flamelet model. Christo et al. (Christo and Dally, 2005) also compared three combustion models: ξ/PDF, flamelet model and EDC model in RANS, leading to the conclusion that conserved-scalar based models like ξ/PDF and flamelet models are inadequate for the JHC configuration. Chitgarha et al. (Chitgarha and Mardani, 2018) investigated the potential of flamelet modelling for JHC flames with RANS as well. Results with lower accuracy than the EDC model are obtained, though lower computational cost is required by the flamelet model. Chen et al. explored the tabulation of a Perfectly Stirred Reactor (PSR) covering the entire flammability range to model the JHC case. Good agreement is observed except for the CO prediction Chen et al. (2017). Furthermore, Kim et al. (Kim et al., 2005) explored the application of conditional moment closure (CMC) model on the JHC configuration, showing attractive predictions.
Evans et al. (Evans et al., 2015) carried out a systematic study and found out that adjusting the EDC parameters to and resulted in significantly improved predictions under different fuel compositions (, , ). Later in 2016, Tu et al. (Tu et al., 2016) adopted and to investigate the physical and chemical effects of CO addition to CH/H JHC Flames. The same was done by Li et al. (Li et al., 2018b), when evaluating the on the fly chemistry reduction and tabulation methods with detailed kinetic mechanisms. Additionally, the combination of and was also tested, showing satisfactory predictions of the Dutch natural gas and biogas JHC flames.
The studies discussed above focused on the modification of or , based on available experimental data. Different from the previous authors, Mardani et al. (Mardani, 2017) proposed a modification of the energy cascade parameters and , eventually affecting the values of and . In their work, the constant was changed from 0.0239 to 27 while was kept to its original value of 0.134. By this variation, the and coefficients in EDC model were varied between 0.0893–3.0 and 1.0–5.795, respectively. By these changes, the model was able to capture the flame lift-off of the reaction zone location and features.
Several sensitivity analyses have been carried out on the JHC burner using EDC model (De et al., 2011; Aminian et al., 2012; Shabanian et al., 2012; Tu et al., 2016; Mardani, 2017; Li et al., 2018a; Li et al., 2018b), to investigate the effect of different modelling choices, focusing on turbulent and combustion model parameters, as well as on the fuel and co-flow compositions. A comprehensive sensitivity study of the JHC burner was carried out by Li et al. (2017), using RANS. Their investigation includes the effect of turbulent combustion model formulations, boundary conditions, differential diffusion, turbulence model parameters and kinetic mechanisms on the results. Results showed that the reactors chosen, namely as Perfectly Stirred Reactors (PSR) or Plug Flow Reactors (PFR), to model the reaction fine structures, do not have a major impact on the results. Moreover, increased kinetic mechanism complexity does not lead to major improvements on the numerical predictions, suggesting that low-temperature oxidation mechanism are not relevant to the fuels and conditions investigated. While the inclusion of differential (molecular) diffusion and the appropriate choice of turbulent non-dimensional values such as Schmidt and Prandtl numbers helped to improve the prediction accuracy. The effect of molecular diffusion was also investigated by Salavati-Zadeh et al. (2018) with the conclusion that the inclusion of the molecular diffusion with proper Schmidt numbers for each species improves the prediction accuracy. The importance of differential diffusion was recognized by Christo and Dally, (2005) and Mardani et al. (2010) as well. In the work of Li et al. (2017), three different Eddy Dissipation Concept (EDC) model formulations were compared as well, showing their interaction with the choice of the constant in the turbulence model. Christo and Dally, (2005) also studied the interactions between combustion model and turbulence model under RANS framework, concluding that the variations of k-ϵ model like the renormalization group and the relizable k-ϵ models perform worse than the standard k-ϵ with the modified constant of
All EDC modifications presented above were based on a fitting procedure aimed at alleviating the temperature over-estimation observed with the standard EDC formulation. Parente et al. (2015) proposed a modification of EDC based on the revision of the energy cascade, taking into account the microscopically distributed features of MILD combustion. In particular, it was assumed that the reacting structures propagated with a turbulent flame speed expressed using the Damköhler formulation for high-intensity turbulence. With this assumption, the energy cascade coefficients and could be expressed as a function of the local turbulent Reynolds and Damköhler number of the flow. The estimation of the chemical time scale required in the Damköhler number was first based on a global reaction mechanism, and later extended to detailed chemistry by Evans et al. (2019). Beside the model constants and , Lewandowski and Ertesvåg (2018) proposed a modification on the reacting fraction χ of fine structures, which is often set to unity in other studies. They observed better agreement with experimental measurement when the reacting fraction is reduced below unity, with an improvement of both temperature and lift-off height predictions.
With the exception of the EDC formulation proposed by Parente et al. (2015) and Evans et al. (2019), the mass fraction of the fine structures and the residence time are solely determined by flow properties in EDC. However, MILD combustion is driven by the strong overlap of fluid dynamic and chemical scales. This suggests that the characteristic chemical time scale shall be included in the definition of the reacting structure features. This has pushed the investigation of PaSR approach for modelling MILD combustion. As indicated in Section 2.2, both chemical time scale and mixing time scale are used for the determination of κ, which is the factor accounting for non-perfect mixing (turbulence-chemistry interaction). Li et al. (2017) compared the EDC model with standard model constants to the standard PaSR model, for the simulation of JHC burner with C2H4, H2, H2O as fuel. The prediction of mean temperate and species mass fraction (including O and NO) improved significantly using PaSR, indicating its potential for MILD combustion. There are several ways to estimate the characteristic time scales required by the model (Li et al., 2018a; Ferrarotti et al., 2019). It was showed that appropriate choices of mixing and chemical time scales are crucial to ensure the prediction accuracy. Li and Ferrarotti et al. (Li et al., 2018a; Ferrarotti et al., 2019) reported that a dynamic evaluation for mixing time scale show superior performance for the configuration. The dynamic evaluation adopts mixing time scale as the ratio between the variance of mixture fraction and its dissipation rate, rather than global estimations based on Kolmogorov or integral mixing scales. Different approaches to evaluate chemical time scale were also compared, based on the species formation rates, the reaction rates and the eigenvalues of the formation rate Jacobian matrix (Li et al., 2018a). Various co-flow oxygen dilution levels and Reynolds numbers over a wide range of operating conditions (with C2H4, H2, H2O as fuel) were included in the validation work.
The JHC configuration was also investigated using LES, to improve the prediction of intermediate and minor species, as well as to capture the intermittency observed in some configurations when increasing the fuel jet Reynolds number (Parente et al., 2015). LES using a three-stream Flamelet Progress Variable (FPV) formulation (tabulated based model) was employed by Ihme et al. (Ihme and See, 2011; Ihme et al., 2012) to model the JHC flame. Satisfactory agreement with the experimental data were obtained for mean temperature and major species mass fractions. While the use of FPV-based approaches in MILD combustion is promising, the highly diluted feature of MILD combustion often leads to high dimensional tables (Lamouroux et al., 2014; Locci et al., 2014), whose generation can be challenging and time consuming. Afarin et al. (2011) used PaSR to investigate the reaction zone structure and the distribution of temperature and minor species mass fractions, showing satisfactory accuracy. Li et al. (2019) presented a detailed comparison between the conventional PaSR model and two implicit combustion models, in which the filtered source term comes directly from the chemical term, without inclusion of turbulence effects. Results indicated that all three models could accurately predict the combustion of CH/H in the JHC configuration, indicating that the sub-grid closure parameter κ played a minor role for the case under investigation. This suggested that, for low Damköhler number systems, the reacting structures can be potentially resolved on the LES grid.
A detailed comparison of different models applied on the JHC configuration is presented in Table 1 and the prediction errors on several axial locations is summarized in Table 2. The errors reported in Table 2 focus on the JHC burner with CH/H 50/50 as fuel and with 3 (HM1)/6 (HM2)/9 (HM3) oxygen levels in the coflow. It is worth mentioning that at around 100 mm downstream of the jet outlet, entertainment from the surroundings tunnel air starts to have an effect on the flame (Dally et al., 2002).
TABLE 1
| Models | Details | Findings | References |
|---|---|---|---|
| Standard EDC (RANS) | — | Over-predicted temperature at axial location = 120 mm | De et al. (2011); Aminian et al. (2012); Shabanian et al. (2012); Evans et al. (2015); Li et al. (2017); Mardani (2017); Li et al. (2018b) |
| Modified EDC (RANS) | Increased and/or decreased | Alleviated temperature over-prediction at axial location = 120 mm | De et al. (2011); Aminian et al. (2012); Shabanian et al. (2012); Evans et al. (2015); Tu et al. (2016); Li et al. (2018b) |
| Modified and | - and changed accordingly | Mardani (2017); Parente et al. (2015) | |
| - Capture the flame lift-off well | |||
| Dynamic EDC (RANS) | Model constants determined locally and reacting fraction in fine structures χ = 1.0 local model constants with reacting fraction χ variation | Satisfactory prediction at various axial locations and generalized models | Parente et al. (2015); Evans et al. (2019) |
| — | Improved prediction compared to unifom χ = 1.0 | Lewandowski and Ertesvåg (2018); Lewandowski et al. (2020b, a) | |
| PaSR model (RANS) | Mixing and chemical time scales included explictly | Prediction at axial location 60 mm improved significantly compared to standard EDC | Li et al. (2017) |
| Dynamic PaSR model (RANS) | Dynamic mixing time scale | Improved prediction compared to standard PaSR, especially the high turbulence case | Li et al. (2018a); Ferrarotti et al. (2019) |
| Flamelet model (RANS) | — | - Computationaly efficient | Shabanian et al. (2012); Christo and Dally (2005); Chitgarha and Mardani (2018) |
| - Inadequate for JHC configuration | |||
| Tabulated PSR model (RANS) | — | - Computationaly efficient | Chen et al. (2017) |
| - CO modelling is not satisfactory | |||
| Transported PDF (RANS) | — | - Provide superior results | Shabanian et al. (2012) |
| - Computationally inefficient | |||
| Conditional moment closure (RANS) | — | Conditional fluctuations of reactive scalars should be small enough for first-order closure | Kim et al. (2005) |
| Flamelet progress variable (LES) | — | - High workload for pre-processing | Ihme and See (2011); Ihme et al. (2012); Lamouroux et al. (2014); Locci et al. (2014) |
| - Largely increased table size with high dilution and heat losses | |||
| - Less CPU hour needed | |||
| PaSR (LES) | κ evaluated with mixing and chemical time scale | - No effort needed for pre-processing | Li et al. (2019) |
| - Over-all good agreement | |||
| - Compromise efficiency between tabulation based models and transported PDF model | |||
| Implicit models (LES) | κ = 1 | - Similar performance as PaSR model | Li et al. (2019) |
| - No need to estimate time scales |
Comparison of combustion models
TABLE 2
| Models | Description | 30 mm | 60 mm | 120 mm | Case | References |
|---|---|---|---|---|---|---|
| Standard EDC (RANS) | Standard constants | 0.8 | 4.8 | 27.4 | HM1 | Aminian et al. (2012) |
| 8.2 | 10.6 | 25.7 | HM2 | |||
| 5.8 | 8.2 | 20.2 | HM3 | |||
| Modified EDC (RANS) | = 1.5 | 3.0 | 0.7 | 13.3 | HM1 | Aminian et al. (2012) |
| 1.3 | 2.9 | 13.1 | HM2 | |||
| 2.0 | 0.5 | 9.5 | HM3 | |||
| = 3.0 | 7.9 | 6.0 | 0.4 | HM3 | Aminian et al. (2012) | |
| = 1.47, = 1.90 | 3.8 | 3.6 | 5.6 | HM1 | Parente et al. (2015) | |
| = 0.02 | 2.1 | 1.8 | 8.8 | HM1 | Mardani (2017) | |
| = 0.25 | 0.7 | 0.2 | 15.6 | HM1 | ||
| = 1.0 | 2.0 | 1.7 | 13.2 | HM1 | ||
| = 27 | 5.8 | 5.4 | 1.7 | HM1 | ||
| = 0.02 | 0.5 | 0.5 | 8.5 | HM3 | ||
| = 0.25 | 1.3 | 2.8 | 14.0 | HM3 | ||
| = 1.0 | 2.3 | 0.4 | 13.2 | HM3 | ||
| = 27 | 14.5 | 11.7 | 1.6 | HM3 | ||
| Dynamic EDC (RANS) | 3.48 | 1.75 | 9.35 | HM1 | Parente et al. (2015) | |
| 1.84 | 4.37 | 2.27 | HM2 | |||
| 8.56 | 10.49 | 3.08 | HM3 | |||
| Flamelet model (RANS) | 5.46 | 3.76 | — | HM1 | Chitgarha and Mardani (2018) | |
| 13.7 | 8.03 | — | HM3 |
Absolute errors for temperature predictions () reported in literature with various combustion models
Conclusion
The present paper reports a short review on the application of reactor based models to the simulation of a canonical MILD combustion system, the jet in hot coflow (JHC) burner. Successful predictions of JHC burner with finite-rate models were reported, with both Reynolds Average Navier-Stokes (RANS) simulation and Large Eddy Simulation (LES). The main conclusions drawn can be summarized as follows:
The Eddy Dissipation Concept (EDC) model and its modified versions (with modified model constants) were first considered by most authors (De et al., 2011; Aminian et al., 2012; Shabanian et al., 2012; Tu et al., 2016; Mardani, 2017; Li et al., 2018a; Li et al., 2018b) for the prediction of a jet-in-hot-coflow (JHC) burner with RANS simulation. However, there is a lack of generality.
The estimation of local EDC parameters based on the local turbulent Reynolds number and Damköhler numbers (Parente et al., 2015) could significantly improve the predictions in the context of MILD combustion, making the inclusion of chemical time scale (needed to estimate Damköhler number) important.
In PaSR, the turbulence-chemistry interaction factor is based on a more general definition which requires the estimation of both chemical and mixing time scales. The choice of such scales has a crucial impact on the model prediction (Li et al., 2018a).
A dynamic evaluation of mixing time scale in PaSR RANS formulation was proposed, presenting superior performance than the other globally defined mixing models under a wide range of JHC flame operating conditions (Li et al., 2018a; Ferrarotti et al., 2019).
The LES formulation of PaSR model along with other implicit combustion models show superior advantage for the prediction on the JHC flames (Li et al., 2019).
Future studies shall focus on the generalisation and unification of reactor-based approaches, using available DNS data in MILD combustion (Minamoto and Swaminathan, 2015; Doan and Swaminathan, 2019) together with machine learning and optimization algorithms.
Future studies should also target more complex fuels, such as oxygenated hydrocarbons and long-chain alkanes under MILD conditions. Because they have shown different characteristics, such as the appearance of visible flames and increased pollutant emissions, as indicated by (Weber et al., 2005; Saha et al., 2014; Ye et al., 2015b).
Funding
This project has received funding from the European Research Council, Starting Grant No. 714605, and the European Union’s Horizon 2020 research and innovation program under the Marie Skłodowska-Curie grant agreement No. 643134.
Statements
Author contributions
ZL wrote this paper. AP contributed to the paper conception and writing.
Acknowledgments
Some sentences in the article were rephrased from the PhD thesis (Li et al., 2019) of the first author ZL.
Conflict of interest
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
References
1
AfarinY.TabejamaatS.MardaniA. (2011). “Large eddy simulation study of H2/CH4 flame structure at MILD condition.” In Seventh mediterranean combustion symposium, Naples, Italy, September 11–September 15
2
AminianJ.GallettiC.ShahhosseiniS.TognottiL. (2012). Numerical investigation of a MILD combustion burner: analysis of mixing field, chemical kinetics and turbulence-chemistry interaction. Flow, Turbul. Combust.88, 597–623. 10.1007/s10494-012-9386-z
3
AminianJ.GallettiC.TognottiL. (2016). Extended EDC local extinction model accounting finite-rate chemistry for MILD combustion. Fuel165, 123–133. 10.1016/j.fuel.2015.10.041
4
CavaliereA.de JoannonM. (2004). MILD combustion. Prog. Energy Combust. Sci.30, 329–366. 10.1016/j.pecs.2004.02.003
5
ChenZ.ReddyV.RuanS.DoanN.RobertsW.SwaminathanN. (2017). Simulation of mild combustion using perfectly stirred reactor model. Proc. Combust. Inst.36, 4279–4286. 10.1016/j.proci.2016.06.007
6
ChitgarhaF.MardaniA. (2018). Assessment of steady and unsteady flamelet models for mild combustion modeling. Int. J. Hydrogen Energy43, 15551–15563. 10.1016/j.ijhydene.2018.06.071
7
ChomiakJ. (1990). Combustion: a study in theory, fact and application. New York, NY: Abacus Press/Gorden and Breach Science Publishers
8
ChristoF. C.DallyB. B. (2005). Modelling turbulent reacting jets issuing into a hot and diluted coflow. Combust. Flame142, 117–129. 10.1016/j.combustflame.2005.03.002
9
DallyB. B.KarpetisA. N.BarlowR. S. (2002). Structure of turbulent non-premixed jet flames in a diluted hot coflow. Proc. Combust. Inst.29, 1147–1154. 10.1016/s1540-7489(02)80145-6
10
DeA.OldenhofE.SathiahP.RoekaertsD. (2011). Numerical simulation of Delft-Jet-in-Hot-Coflow (DJHC) flames using the Eddy Dissipation Concept model for turbulence-chemistry interaction. Flow, Turbul. Combust.87, 537–567. 10.1007/s10494-011-9337-0
11
de JoannonM.SorrentinoG.CavaliereA. (2012). MILD combustion in diffusion-controlled regimes of hot diluted fuel. Combust. Flame159, 1832–1839. 10.1016/j.combustflame.2012.01.013
12
DoanN. A. K.SwaminathanN. (2019). Autoignition and flame propagation in non-premixed MILD combustion. Combust. Flame201, 234–243. 10.1016/j.combustflame.2018.12.025
13
EvansM.ChinniciA.MedwellP.YeJ. (2017). Ignition features of methane and ethylene fuel-blends in hot and diluted coflows. Fuel203, 279–289. 10.1016/j.combustflame.2018.12.025
14
EvansM. J.MedwellP. R.TianZ. F. (2015). Modelling lifted jet flames in a heated coflow using an optimised eddy dissipation concept model. Combust. Sci. Technol.187, 1093–1109. 10.1016/j.fuel.2017.04.113
15
EvansM.PetreC.MedwellP.ParenteA. (2019). Generalisation of the eddy-dissipation concept for jet flames with low turbulence and low damköhler number. Proc. Combust. Inst.37, 4497–4505. 10.1080/00102202.2014.1002836
16
FerrarottiM.LiZ.ParenteA. (2019). On the role of mixing models in the simulation of MILD combustion using finite-rate chemistry combustion models. Proc. Combust. Inst.37, 4531–4538. 10.1016/j.proci.2018.06.017
17
FrassoldatiA.SharmaP.CuociA.FaravelliT.RanziE. (2009). Kinetic and fluid dynamics modeling of methane/hydrogen jet flames in diluted coflow. Appl. Therm. Eng.30, 376–383. 10.1016/j.applthermaleng.2009.10.001
18
GolovitchevV.ChomiakJ. (2001). Numerical modeling of high temperature air flameless combustion. In The 4th international symposium on high temperature air combustion and gasification
19
GranI.MagnussenB. F. (1996). A numerical study of a bluff-body stabilized diffusion flame, part 2: influence of combustion modelling and finite-rate chemistry. Combust. Sci. Technol.119, 191–217. 10.1080/00102209608951999
20
HaworthD.PopeS. (2010). Transported probability density function methods for Reynolds-averaged and large-eddy simulations. New York, NY:(Springer)
21
IhmeM.SeeY. C. (2011). LES flamelet modeling of a three-stream MILD combustor: analysis of flame sensitivity to scalar inflow conditions. Proc. Combust. Inst.33, 1309–1317. 10.1016/j.proci.2010.05.019
22
IhmeM.ZhangJ.HeG.DallyB. (2012). Large Eddy Simulation of a Jet-in-Hot-Coflow burner operating in the oxygen-diluted combustion regime. Flow, Turbul. Combust.89, 449–46410.1007/s10494-012-9399-7
23
KärrholmF. P. (2008). Numerical Modelling of diesel spray injection, turbulence Interaction and combustion. Phd thesis, Chalmers University of Technology, Chalmers, Sweden
24
KimS. H.HuhK. Y.DallyB. (2005). Conditional moment closure modeling of turbulent nonpremixed combustion in diluted hot coflow. Proc. Combust. Inst.30, 751–757. 10.1016/j.proci.2004.08.161
25
LamourouxJ.IhmeM.FiorinaB.GicquelO. (2014). Tabulated chemistry approach for diluted combustion regimes with internal recirculation and heat losses. Combust. Flame161, 2120–2136. 10.1016/j.combustflame.2014.01.015
26
LewandowskiM. T.ErtesvågI. S. (2018). Analysis of the eddy dissipation concept formulation for mild combustion modeling. Fuel224, 687–700. 10.1016/j.fuel.2018.03.110
27
LewandowskiM. T.LiZ.ParenteA.PozorskiJ. (2020a). Generalised eddy dissipation concept for mild combustion regime at low local Reynolds and damköhler numbers. part 2: validation of the model. Fuel278, 117773https://doi.org/10.1016/j.fuel.2020.117773
28
LewandowskiM. T.ParenteA.PozorskiJ. (2020b). Generalised eddy dissipation concept for mild combustion regime at low local Reynolds and damköhler numbers. part 1: model framework development. Fuel278, 117743, 10.1016/j.fuel.2020.117743
29
LiZ.CuociA.ParenteA. (2019). Large eddy simulation of mild combustion using finite rate chemistry: effect of combustion sub-grid closure. Proc. Combust. Inst.37, 4519–4529. 10.1016/j.proci.2018.09.033
30
LiZ.CuociA.SadikiA.ParenteA. (2017). Comprehensive numerical study of the Adelaide Jet in Hot-Coflow burner by means of RANS and detailed chemistry. Energy139, 555–570. 10.1016/j.energy.2017.07.132
31
LiZ.FerrarottiM.CuociA.ParenteA. (2018a). Finite-rate chemistry modelling of non-conventional combustion regimes using a partially-stirred reactor closure: combustion model formulation and implementation details. Appl. Energy225, 637–655. 10.1016/j.apenergy.2018.04.085
32
LiZ.LewandowskiM. T.ContinoF.ParenteA. (2018b). Assessment of on-the-fly chemistry reduction and tabulation approaches for the simulation of moderate or intense low-oxygen dilution combustion. Energy and Fuels32, 10121–10131. 10.1021/acs.energyfuels.8b01001
33
LiZ. (2019). Sub-grid models for Large Eddy Simulation of non-conventional combustion regimes. Ph.D. thesis. Darmstadt(Germany): Université libre de Bruxelles and Technische Universitæt Darmstadt
34
LocciC.ColinO.MichelJ. B. (2014). Large Eddy simulations of a small-scale flameless combustor by means of diluted homogeneous reactors. Flow, Turbul. Combust.93, 305–347. 10.1007/s10494-014-9548-2
35
MaL.NaudB.RoekaertsD. (2016). Transported pdf modeling of ethanol spray in hot-diluted coflow flame. Flow, Turbul. Combust.96, 469–502. 10.1007/s10494-015-9623-3
36
MagnussenB. F. (1981). “On the structure of turbulence and a generalized Eddy Dissipation Concept for chemical reaction in turbulent flow.” In 19AIAA aerospace science meeting, St. Louis, Missouri
37
MagnussenB. F. (2005). “The Eddy Dissipation Concept a bridge between science and technology.” In ECCOMAS thematic conference on computational combustion, Lisbon, Portugal
38
MahalegiH. K. M.MardaniA. (2019). Ethanol spray combustion under a mild condition: a chemical kinetic study. Energy & Fuels33, 11861–11886. 10.1021/acs.energyfuels.9b02665
39
MardaniA.Karimi Motaalegh MahalegiH. (2019). Hydrogen enrichment of methane and syngas for mild combustion. Int. J. Hydrogen Energy44, 9423–9437. 10.1016/j.ijhydene.2019.02.072
40
MardaniA. (2017). Optimization of the Eddy Dissipation Concept (EDC) model for turbulence-chemistry interactions under hot diluted combustion of ch4/h2. Fuel191, 114–129. 10.1016/j.fuel.2016.11.056
41
MardaniA.TabejamaatS.GhamariM. (2010). Numerical study of influence of molecular diffusion in the MILD combustion regime. Combust. Theor. Model.14, 747–774. 10.1080/13647830.2010.512959
42
MedwellP. R.DallyB. B. (2012). Effect of fuel composition on jet flames in a heated and diluted oxidant stream. Combust. Flame159, 3138–3145. 10.1016/j.combustflame.2012.04.012
43
MedwellP. R.KaltP. A.DallyB. B. (2007). Imaging of diluted turbulent ethylene flames stabilized on a Jet in Hot Coflow (JHC) burner. Combust. Flame152, 100–113. 10.1016/j.combustflame.2007.09.003
44
MinamotoY.SwaminathanN. (2015). Subgrid scale modelling for MILD combustion. Proc. Combust. Inst.35, 3529–3536. 10.1016/j.proci.2014.07.025
45
OldenhofE.TummersM. J.van VeenE.RoekaertsD. (2010). Ignition kernel formation and lift-off behaviour of Jet-in-Hot-Coflow flames. Combust. Flame157, 1167–1178. 10.1016/j.combustflame.2010.01.002
46
ParenteA.MalikM. R.ContinoF.CuociA.DallyB. B. (2015). Extension of the Eddy Dissipation Concept for turbulence/chemistry interactions to MILD combustion. Fuel163, 98–111. 10.1016/j.fuel.2015.09.020
47
SahaM.DallyB. B.MedwellP. R.ClearyE. M. (2014). Moderate or Intense Low oxygen Dilution (MILD) combustion characteristics of pulverized coal in a self-recuperative furnace. Energy & Fuels28, 6046–6057. 10.1021/ef5006836
48
Salavati-ZadehA.EsfahanianV.Najafi]S. B. N.SaeedH.MohammadiM. (2018). Kinetic simulation of flameless burners with methane/hydrogen blended fuel: effects of molecular diffusion and schmidt number. Int. J. Hydrogen Energy43, 5972–5983. 10.1016/j.ijhydene.2017.11.149
49
ShabanianS. R.MedwellP. R.RahimiM.FrassoldatiA.CuociA. (2012). Kinetic and fluid dynamic modeling of ethylene jet flames in diluted and heated oxidant stream combustion conditions. Appl. Therm. Eng.52, 538–554. 10.1016/j.applthermaleng.2012.12.024
50
TuY.SuK.LiuH.ChenS.LiuZ.ZhengC. (2016). Physical and chemical effects of CO2 addition on CH4/H2 flames on a Jet-in-Hot-Coflow (JHC) burner. Energy and Fuels30, 1390–1399. 10.1021/acs.energyfuels.5b02499
51
WünningJ. A.WünningJ. G. (1997). Flameless oxidation to reduce thermal NO-formation. Prog. Energy Combust. Sci.23, 81–94. 10.1016/s0360-1285(97)00006-3
52
WeberR.SmartJ. P.vd KampW. (2005). On the (MILD) combustion of gaseous, liquid, and solid fuels in high temperature preheated air. Proc. Combust. Inst.30, 2623–2629. 10.1016/j.proci.2004.08.101
53
YeJ.MedwellP. R.EvansM. J.DallyB. B. (2017). Characteristics of turbulent n-heptane jet flames in a hot and diluted coflow. Combust. Flame183, 330–342. 10.1016/j.combustflame.2017.05.027
54
YeJ.MedwellP. R.KleinheinzK.EvansM. J.DallyB. B.PitschH. G. (2018). Structural differences of ethanol and dme jet flames in a hot diluted coflow. Combust. Flame192, 473 – 494. 10.1016/j.combustflame.2018.02.025
55
YeJ.MedwellP. R.VareaE.KruseS.DallyB. B.PitschH. G. (2015a). An experimental study on mild combustion of prevaporised liquid fuels. Appl. Energy151, 93–101. 10.1016/j.apenergy.2015.04.019
56
YeJ.MedwellP. R.VareaE.KruseS.DallyB. B.PitschH. G. (2015b). An experimental study on MILD combustion of prevaporised liquid fuels. Appl. Energy151, 93–101. 10.1016/j.apenergy.2015.04.019
Summary
Keywords
jet-in-hot-coflow burner, reactor-based models, MILD combustion, turbulence-chemistry interaction, finite rate chemistry
Citation
Li Z and Parente A (2020) A Review of the Numerical Investigations of Jet-In-Hot-Coflow Burner With Reactor-Based Models. Front. Mech. Eng. 6:512501. doi: 10.3389/fmech.2020.512501
Received
15 November 2019
Accepted
05 October 2020
Published
16 November 2020
Volume
6 - 2020
Edited by
Timothy S. Fisher, University of California, Los Angeles, United States
Reviewed by
Kan Zha, Sandia National Laboratories (SNL), United States
Amir Mardani, Sharif University of Technology, Iran
Updates
Copyright
© 2020 Li and Parente.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Zhiyi Li, zhiyi.li@ulb.ac.be; Alessandro Parente, alessandro.parente@ulb.ac.be
This article was submitted to Thermal and Mass Transport, a section of the journal Frontiers in Mechanical Engineering
Disclaimer
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.