Search

• A novel n-heptane reform gas combustion is investigated using LES and t-PDF model. • The n-heptane reform gas is shown to suppress the ignition and increase the liftoff. • The reforming gas weakens the NTC ignition behavior of n-heptane. • The combustion process involves inner diffusion flame and outer lean premixed flame. • The premixed flame in the surrounding reforming gas/air mixture enhances soot and suppresses NOx emission. This paper presents a large eddy simulation study of n-heptane spray combustion in an n-heptane low temperature reform (LTR) gas environment in a constant volume combustion chamber, under conditions relevant to single-fuel reactivity controlled compression ignition (RCCI) combustion engines. The LTR gas is made up of partially oxidized intermediate species from rich n-heptane/air mixture in an external constant temperature reformer. It is found that a higher reform temperature results in a longer ignition delay time of the n-heptane spray and a higher liftoff length, due to the chemical effect of the LTR gas and the difference in the reaction zone structures. A significantly different spray flame structure is identified in the RCCI case from that of single-fuel spray combustion. After the onset of high temperature ignition, a double-layer flame structure is established in the RCCI case, with a diffusion flame layer and a lean premixed flame layer. The lean premixed flame affects the flow field, which significantly suppresses the mixing around the spray tip. As a result, the RCCI case exhibits a lower NOx formation but a higher soot formation than the single-fuel case. [ABSTRACT FROM AUTHOR]

The flame dynamics in a novel, ultra-lean non-premixed model gas turbine (GT) burner flame was numerically studied using large eddy simulation (LES) coupled with probability density function (PDF) based on the Euler stochastic field (ESF) method. One non-reacting case and three reacting cases with global equivalence ratio ϕ glob = 1.0 , 0.6 , 0.3 were simulated. Comparison of mean flow fields and OH distributions between numerical and experimental results was conducted. Flame dynamics including flame stabilization, structures and transitions of combustion modes were investigated. The simulation results were in close agreements with the experimental measurements showing the capability of PDF-ESF LES model in predicting the flame behaviors. At higher ϕ glob , the fuel jet velocity was higher, which yielded higher scalar dissipation rate, χ , near the burner exit, leading to local extinction and flame lifted-off. In the extinction region, a series of relatively low to medium temperature reactions were active, providing favorable conditions for re-ignition downstream where χ is lower than the critical scalar dissipation rate for flame extinction, χ crt . In addition, with ϕ glob decreasing, the flame height decreased due to a smaller jet velocity and χ , and thus the mechanism of flame stabilization changed from the swirl-stabilized to the bluff-body stabilized. [ABSTRACT FROM AUTHOR]

In the present study, large eddy simulation is used to investigate combustion recession for the Engine Combustion Network Spray A flame at two ambient temperatures (850 K and 800 K) and two injection pressures (100 MPa and 50 MPa). The present numerical results are able to capture different combustion recession phenomena after the end-of-injection (AEOI). With an injection pressure of 100 MPa , the model predicts a "separated" combustion recession at the ambient temperature of 850 K and no combustion recession at the ambient temperature of 800 K , in which both predictions correspond to the measurements. The combustion recession is mainly controlled by the auto-ignition process at the ambient temperature of 850 K. At the ambient temperature of 800 K , the local temperature within the fuel-rich region is not high enough to promote the high-temperature ignition process. As time progresses, the mixture within the fuel-rich region rapidly transitions to become an overly fuel-lean mixture, which further hinders high-temperature ignition to occur. Nonetheless, it is shown that lowering the injection pressure to 50 MPa causes the combustion recession to occur at the ambient temperature of 800 K. This is likely attributed to the low injection case having a lower air entrainment rate AEOI, which causes the mixtures upstream of the lift-off position to transition slower from fuel-rich to fuel-lean mixtures. • LES model captures the evolution of flame structure during combustion recession. • At high P i n j of 100 MPa, combustion recession occurs at 850 K but not at 800 K. • Lowering P i n j to 50 MPa leads to combustion recession in the 800 K spray case. • Air entrainment after EOI affects the fuel–air mixing and combustion recession. • Lowering P i n j leads to lower entrainment rate after EOI and relatively richer mixture. [ABSTRACT FROM AUTHOR]

This paper presents the numerical results of combustion and soot recession in the Engine Combustion Network (ECN) Spray A and D flames using large eddy simulations (LES). The nominal injector nozzle diameters for the ECN Spray A and D are 90 μ m and 186 μ m. A two-equation soot model is implemented to model the soot formation and oxidation processes. The numerical model is validated by comparing the simulated and measured data in terms of ignition delay time (IDT), lift-off-length (LOL), and temporal evolution of soot mass. The combustion and soot recession processes are analyzed after the end-of-injection (AEOI). The combustion recession processes are first driven by auto-ignition wave propagation followed also by the convective flow in both the Spray A and D flames. A separated high-temperature flame structure is observed in the Spray A flame due to having small favorable mixture regions for the auto-ignition to occur upstream of the quasi-steady lift-off position AEOI. In contrast, a spatially-continuous high-temperature flame structure is formed in the Spray D flame due to having more ignitable mixture regions upstream of the quasi-steady lift-off position and a higher heat release. Soot recession is observed in the Spray D flame but not in the Spray A flame. This is attributed to the mixture upstream of the quasi-steady state soot region becoming favorable to promote soot formation in the Spray D flame, but it becomes fuel-lean AEOI in the Spray A flame. • Effects of nozzle diameters on combustion and soot recession are studied. • ECN Spray A shows separated high-T flame structure after end of injection. • ECN Spray D shows spatially-continuous high-T flame structure after end of injection. • Soot recession is present in the ECN Spray D flame, but not in the ECN Spray A flame. [ABSTRACT FROM AUTHOR]

With the increase of computing power, large eddy simulation (LES) within the Lagrangian–Eulerian two-phase flow framework has evolved as an useful numerical tool to gain insights to the spray combustion processes in advanced internal combustion engines. However, the inherent effect of grid/filter size (G/FS) on the physics revealed by LES is not fully understood. In this paper, we present several new viewpoints on this by analyzing the results from a Lagrangian–Eulerian LES study of Engine Combustion Networks Spray A with a newly developed multi-region adaptive mesh refinement method in OpenFOAM. It is found that G/FS affects the predicted spray characteristics, the modes of spray combustion, the ignition delay time and the liftoff length in different ways. First, owing to the nature of Lagrangian particle-in-cell approach, the spray liquid penetration length converges as the G/FS approaches to the nozzle diameter. The convergence behavior is rather independent of the spray model parameters. Second, it is critical to have a well resolved spray liquid region; the main spray combustion characteristics, e.g., the mean pressure rise profile, the onsets of the cool flame and hot flame, and the main flame structure, are shown to be similar and independent of G/FS once the liquid region is properly resolved. However, the detailed flame structures at the liftoff position are sensitive to G/FS; the upstream auto-ignition events, the low temperature ignition assisted flame propagation and the hot flame propagation, are shown to partially rely on the adopted G/FS. Finally, it is found that a finer G/FS predicts a slower fuel/oxidizer mixing and a shorter flame liftoff length, yielding a higher soot mass. • We proposed a novel multi-region adaptive mesh refinement (AMR) method for ECN Spray A flames. • Using the multi-region AMR method can obviously improve the grid size convergence of flame characteristics. • The spray ignition position and timing are insensitive to the grid size employed (from 250 μ m to 62. 5 μ m). • The flame liftoff and soot formation are closely related to the grid size employed. • More fruitful combustion modes at flame liftoff position are captured with finer grid size. [ABSTRACT FROM AUTHOR]

In this study, large eddy simulations, coupled a two-equation soot model, are performed to investigate the effects of ambient carbon dioxide (CO 2) and water (H 2 O) additions on the soot formation and oxidation processes in an n -dodecane spray flame. In the soot model, acetylene (C 2 H 2) is soot precursor and surface growth species, while hydroxyl radical (OH) and oxygen (O 2) are soot oxidizers. The effect of ambient CO 2 and H 2 O additions on soot formation/oxidation can be separated into thermal and chemical effects. For the thermal effects, the ambient CO 2 and H 2 O additions increase C 2 H 2 but reduce OH formation by lowering the flame temperature. This leads to a higher soot mass formed. On the contrary to the thermal effects, the ambient CO 2 and H 2 O additions reduce the soot formation due to their chemical effects. The reaction CH 2 ∗ + CO 2 ↔ CH 2 O + CO is found to be responsible for reducing C 2 H 2 formation. The ambient H 2 O addition results in a higher OH but lower the C 2 H 2 mass formed owing to the reverse reactions H 2 + OH ↔ H 2 O + H and OH + OH ↔ H 2 O + O. Furthermore, the chemical effects is more significant than the thermal effects under the tested conditions. This leads to a lower soot mass formed when adding ambient CO 2 and H 2 O. • Soot suppression effect due to ambient CO 2 and H 2 O additions is captured using LES. • Thermal effects of CO 2 and H 2 O promote soot, but their chemical effects inhibit soot. • The mechanisms of soot suppression due to ambient CO 2 and H 2 O additions are revealed. [ABSTRACT FROM AUTHOR]

• Transient evolution of soot mass is captured using LES. • Underlying mechanisms behind the formation of soot spike are proposed. • Before QSS, LES predicts a larger soot formation region as compared the URANS model. Numerical simulations using large eddy simulation (LES) and Unsteady Reynolds Averaged Navier–Stokes (URANS) are carried out to identify the underlying mechanisms that govern the early soot evolution process in an n -dodecane spray flame at 21% O 2 by molar concentration. A two-equation phenomenological soot model is used here to simulate soot formation and oxidation. Both ignition delay time (IDT) and lift-off length (LOL) are found to agree with experimental measurements. The transient evolution of soot mass, in particularly the soot spike phenomenon, is captured in the present LES cases, but not in the URANS cases. Hence, a comparison of numerical results from LES and URANS simulations is conducted to provide a better insight of this phenomenon. LES is able to predict the rapid increasing soot mass during the early stage of soot formation due to having a large favorable region of equivalence ratio (ϕ > 1.5) and temperature (T > 1800 K) for soot formation. This favorable region increases and then decreases to reach a quasi-steady state in the LES case, while it continues to increase in the URANS simulation during the early time. In addition, the soot spike is a consequence of the competition between soot formation and oxidation rates. The time instance when the total soot mass reaches peak value coincides with the time instance when the total mass of soot precursor reaches a plateau. The soot spike is formed due to the continuous increase of oxidizing species in the LES case which leads to a more dominant oxidation process than the formation process. [ABSTRACT FROM AUTHOR]

Experimental and numerical studies were performed to understand the stabilization of lean premixed natural gas/air flames in a gas turbine model combustor which was equipped with a swirl burner, known as the CECOST burner, designed to replicate the flow and flame structures in an industrial gas turbine engine. The operability range, flame stabilization, and flashback were investigated employing simultaneous OH– and CH 2 O-PLIF, and high-speed chemiluminescence imaging. Large eddy simulation (LES) was carried out for analysis of the vortex breakdown structures under non-reacting conditions. It was found that the vortex breakdown structures under isothermal conditions were insensitive to the Reynolds number (Re) for Re ≥ 10000; however, the stability of the flames and operability range of the burner were highly sensitive to Re as well as to equivalence ratio (ϕ). The equivalence ratio was varied at various Reynolds numbers to observe different regimes of the flame ranging from the lean blowout (LBO) limit to the flashback limit. The LBO limit was found to be mainly a function of equivalence ratio while being nearly independent of the Reynolds number, whereas the occurrence of flashback showed distinct characteristics for different ranges of the Reynolds number. At low and moderate Reynolds numbers, (Re ≤ 17000), flashback occurred when increasing ϕ from lean towards stoichiometric conditions. The coupling between the flow field and heat release induces vortex breakdown in the mixing tube and initiates flashback. In contrast, at higher Reynolds numbers (Re > 17000) no flashback was observed even when ϕ was increased to stoichiometric conditions. At these conditions with high Re, the increase in the bulk flow velocity affects the vortex breakdown structure, pushing the vortex breakdown downstream, which in turn prevents the flame from flashing back into the mixing tube. [ABSTRACT FROM AUTHOR]