Different Regimes Of Combustion Research Articles

The Formation of Perovskite during the Combustion of an Energy-Rich Glycine-Nitrate Precursor.

The effect of different regimes of combustion of glycine–nitrate precursors on the formation of perovskite phases (LaMnO3 and LaCrO3) without additional heat treatment was studied. The following three combustion regimes were compared: the traditional solution combustion synthesis (SCS), volume combustion synthesis (VCS) using a powdered precursor, and self-propagating high-temperature synthesis (SHS) using a precursor pellet. The products of combustion were studied using a series of physicochemical methods (attenuated total reflection infrared spectroscopy (ATR FTIR), X-ray diffraction (XRD), high-resolution transmission electron microscopy (HRTEM), and thermal analysis). SHS was found to be the most productive regime for the formation of perovskite because of its ability to develop high temperatures in the reaction zone, which led to a reduced content of the thermally stable lanthanum carbonate impurities and to an increased yield and crystallite size of the perovskite phase. The reasons for the better crystallinity and purity of LaCrO3 as compared with LaMnO3 is also discussed, namely the low temperatures of the onset of the thermolysis, the fast rate of combustion, and the favorable thermodynamics for the achievement of high temperatures in the reaction zone.

Evaluation of Flame Area Based on Detailed Chemistry DNS of Premixed Turbulent Hydrogen-Air Flames in Different Regimes of Combustion

Precise evaluation of flame surface area plays a pivotal role in the fundamental understanding and accurate modelling of turbulent premixed flames. This necessity is reflected in the requirement for the instantaneous flame area evaluation of the turbulent burning velocity (by making use of Damkohler’s first hypothesis). Moreover, the information regarding flame area is required in the context of flame surface density based modelling, and for determining the wrinkling factor or estimating the efficiency function. Usually flame surface areas in experiments and Direct Numerical Simulation (DNS) analyses are evaluated differently and the present analysis aims at comparing these approaches by making use of a detailed chemistry DNS database of turbulent, statistically planar flames. It has been found that the flame surface area evaluation is sensitive to the choice of scalar quantity and the isosurface level, and this holds particularly true for two-dimensional evaluations. The conditions, which provide a satisfactory agreement between experimental and numerical approaches in the flame area evaluation, have been identified by a detailed comparative analysis of the usual postprocessing techniques.

Combustion characteristics of gasoline DICI engine in the transition from HCCI to PPC: Experiment and numerical analysis

Both numerical simulations and experiments were conducted in a heavy-duty DICI engine, with PRF81 as a gasoline surrogate, to investigate how the fuel stratification, auto-ignition and combustion are affected by the start of injection (SOI). The intake air temperature was adjusted to keep the combustion phasing constant when the SOI was swept from −100 to −20oCA ATDC, covering different regimes of combustion, from HCCI to PPC. It is found that in the HCCI regime the combustion process is less sensitive to the variation of SOI since the fuel/air mixture is fairly homogeneous. The fuel/air mixture is under fuel-lean condition and the required intake temperature for a constant CA50 is the highest. In the PPC regime there is an optimal SOI window, within which the required intake temperature is the lowest to maintain a constant CA50 and the engine thermal efficiency is the highest. The optimal operation window starts at the SOI when all fuel is injected into the piston bowl and ends when the fuel injection is towards the bottom wall of the piston bowl, which results in a high heat transfer losses. The SOI window for optimal engine operation is expected to be fuel injector and piston bowl geometry dependent. During the transition regime, the fuel is injected towards the piston head in the squish region. The combustion process is highly sensitive to SOI due to the high sensitivity of fuel distribution in the cylinder to SOI. The engine thermal efficiency is the lowest due to the incomplete oxidation of fuel in the squish region.

Experimental and numerical investigation of shrub combustion

The use of physics-based models to simulate vegetation fires has been steadily increasing over the last decade. These models offer the possibility to study quickly and at a low cost the combustion of vegetation according to a wide set of parameters which are difficult to control experimentally. However, the validation and improvement of these models is still ongoing and coupled experimental / numerical approaches are needed.In a previous work, a validation of the numerical code WFDS was done for the spreading of fires across pine needles fuel beds. It highlighted the ability of WFDS to reproduce the combustion dynamics and the related physical quantities (heat release rate, mass loss rate, rate of spread) during the growing and steady-state phases of the fire. The objective of the present work is to test WFDS for the prediction of the combustion of a natural shrub, which is a more complex fuel than a fuel bed. As a first step, 3 shrubs of rockroses (Cistus monspeliensis) were characterized. The distribution of the different classes of particles that compose the shrubs was determined and used as input for the shrub model. A series of 41 fire tests were conducted on the shrubs to investigate their flammability. Ignition was performed using radiant panel. The measurements of the main physical quantities were also used as a comparison basis for the numerical approach. A numerical model of shrub was developed from the characterization results. Then, the combustion of the shrub was simulated with WFDS. The grid mesh resolution needed for the simulation was estimated from the extinction length within the shrub crown and from the heat release rate (HRR) during the fire tests. Finally, experimental measurements and predictions were compared to determine the capability of WFDS to reproduce the combustion of the shrubs.The experimental investigation of the shrub flammability showed four different regimes of combustion governed mainly by the location of ignition. Simulations were run for the flammability regime that exhibits the best reproducibility. The comparison between simulation and experimental results demonstrated a good ability of WFDS to assess the most relevant fire properties, namely the peak of HRR, flame duration and consumption rate. The main differences between the predictions and observations were identified during the ignition and extinction phases. A plateau of HRR is predicted shortly after the ignition, limiting the fire growth. In addition, the duration of the predicted smoldering phase is too short which causes a total extinction of the fire few seconds after the end of the flaming phase.

Generalized flame surface density transport conditional on flow topologies for turbulent H2-air premixed flames in different regimes of combustion

The generalized flame surface density (FSD) transport conditional on local flow topologies in premixed turbulent flames has been analyzed based on a detailed chemistry direct numerical simulation database of statistically planar turbulent hydrogen-air premixed flames with an equivalence ratio of 0.7 representing the corrugated flamelets, thin reaction zones and broken reaction zones regimes of combustion. The local flow topologies have been categorized by the values of the three invariants of the velocity gradient tensor and the statistical behaviors of the generalized FSD and different terms of its transport equation conditional on these flow topologies have been analyzed in detail for different choices of the reaction progress variable. The qualitative behavior of the different terms of the generalized FSD transport equation has been found to be similar for different choices of reaction progress variable but the statistical behaviors of the tangential strain rate term and its components have been found to be affected by the regime of combustion. The topologies, which exist for all values of dilatation rate, contribute significantly to the generalized FSD transport in premixed turbulent flames for all regimes of combustion. An unstable nodal flow topology, which is representative of a counter-flow configuration, has been found to be a dominant contributor to the FSD transport for all regimes of combustion irrespective of the choice of reaction progress variable. Moreover, a focal topology which is obtained only for positive values of dilatation rate, has been found to contribute significantly, especially to the curvature and propagation terms of the FSD transport equation for all regimes of combustion including the broken reaction zones regime. However, the contributions of the flow topologies to the turbulent transport and tangential strain rate term, which are obtained only for positive dilatation rates, have been found to weaken from the corrugated flamelets to the broken reaction zones regime.

Simulation of moving boundaries interacting with compressible reacting flows using a second-order adaptive Cartesian cut-cell method

A high-order adaptive Cartesian cut-cell method, developed in the past by the authors [1] for simulation of compressible viscous flow over static embedded boundaries, is now extended for reacting flow simulations over moving interfaces. The main difficulty related to simulation of moving boundary problems using immersed boundary techniques is the loss of conservation of mass, momentum and energy during the transition of numerical grid cells from solid to fluid and vice versa. Gas phase reactions near solid boundaries can produce huge source terms to the governing equations, which if not properly treated for moving boundaries, can result in inaccuracies in numerical predictions. The small cell clustering algorithm proposed in our previous work is now extended to handle moving boundaries enforcing strict conservation. In addition, the cell clustering algorithm also preserves the smoothness of solution near moving surfaces. A second order Runge–Kutta scheme where the boundaries are allowed to change during the sub-time steps is employed. This scheme improves the time accuracy of the calculations when the body motion is driven by hydrodynamic forces. Simple one dimensional reacting and non-reacting studies of moving piston are first performed in order to demonstrate the accuracy of the proposed method. Results are then reported for flow past moving cylinders at subsonic and supersonic velocities in a viscous compressible flow and are compared with theoretical and previously available experimental data. The ability of the scheme to handle deforming boundaries and interaction of hydrodynamic forces with rigid body motion is demonstrated using different test cases. Finally, the method is applied to investigate the detonation initiation and stabilization mechanisms on a cylinder and a sphere, when they are launched into a detonable mixture. The effect of the filling pressure on the detonation stabilization mechanisms over a hyper-velocity sphere launched into a hydrogen–oxygen–argon mixture is studied and a qualitative comparison of the results with the experimental data are made. Results indicate that the current method is able to correctly reproduce the different regimes of combustion observed in the experiments. Through the various examples it is demonstrated that our method is robust and accurate for simulation of compressible viscous reacting flow problems with moving/deforming boundaries.

A Burke–Schumann analysis of diffusion-flame structures supported by a burning droplet

A Burke–Schumann description of three different regimes of combustion of a fuel droplet in an oxidising atmosphere, namely the premixed-flame regime, the partial-burning regime and the diffusion-flame regime, is presented by treating the fuel and oxygen leakage fractions through the flame as known parameters. The analysis shows that the burning-rate constant, the flame-standoff ratio, and the flame temperature in these regimes can be obtained from the classical droplet-burning results by suitable definitions of an effective ambient oxygen mass fraction and an effective fuel concentration in the droplet interior. The results show that increasing oxygen leakage alone through the flame lowers both the droplet burning rate and the flame temperature, whereas leakage of fuel alone leaves the burning rate unaffected while reducing the flame temperature and moving the flame closer to the droplet surface. Solutions for the partial-burning regime are shown to exist only for a limited range of fuel and oxygen leakage fractions.

Statistics of vorticity alignment with local strain rates in turbulent premixed flames

The instantaneous alignment of the vorticity vector with local principal strain rates is analysed for statistically planar turbulent premixed flames with different values of heat release parameter and global Lewis number spanning different regimes of combustion. It has been shown that the vorticity vector predominantly aligns with the intermediate principal strain rate in turbulent premixed flames, irrespective of the regime of combustion, heat release parameter and Lewis number. However, the relative alignment of vorticity with the most extensive and compressive principal strain rates changes based on the underlying combustion conditions. Detailed physical explanations are provided for the observed behaviours of vorticity alignment with local principal strain rates. It has been shown that heat release due to combustion significantly affects the alignment of vorticity with local principal strain rates. However, the mean contribution of the vortex-stretching term in the transport equation of enstrophy remains positive for all cases considered here, irrespective of the nature of the vorticity alignment.

A novel principal component analysis-based acceleration scheme for LES–ODT: An a priori study

A parameterization of the composition space based on principal component analysis (PCA) is proposed to represent the transport equations with the one-dimensional turbulence (ODT) solutions of a hybrid large-eddy simulation (LES) and ODT scheme. The 1D ODT solutions are embedded in the 3D LES domain and solve for thermo-chemical scalars; while, the LES governing equations solve for the flow. An a priori validation of the proposed approach is implemented based on stand-alone ODT solutions of the Sandia Flame F, which is characterized by different regimes of combustion starting with pilot stabilization, to extinction and reignition and self-stabilized combustion. The PCA analysis is carried out with a full set of the thermo-chemical scalars’ vector as well as a subset of this vector. The subset is made up primarily of major species and temperature. The results show that the different regimes are reproduced using only three principal components for the thermo-chemical scalars based on the full and a subset of the thermo-chemical scalars’ vector. Reproduction of the source term of the principal components represents a challenge, because of the inherent non-linearity of reaction rates’ expressions. It is found that using the subset of the thermo-chemical scalars’ vector both minor species and the first three principal components source terms are reasonably well predicted.

Unstable combustion induced by oblique shock waves at the non-attaching condition of the oblique detonation wave

The instability of oblique shock wave (OSW) induced combustion is examined for a wedge with a flow turning angle greater than the maximum attach angle of the oblique detonation wave (ODW), where archival results rarely exist for this case in previous literatures. Numerical simulations were carried out for wedges of different length scales to account for the ratio of the chemical and fluid dynamic time scales. The results reveal three different regimes of combustion. (1) No ignition or decoupled combustion was observed if a fluid dynamic time is shorter than a chemical time behind an OSW. (2) Oscillatory combustion was observed behind an OSW if a fluid dynamic time is longer than a chemical time behind an OSW and the fluid dynamic time is shorter than the chemical time behind a normal shock wave (NSW) at the same Mach number. (3) Detached bow shock-induced combustion (or detached overdriven detonation wave) was observed if a fluid dynamic time is longer than a chemical time behind a NSW. Since no ignition or decoupled combustion occurs as a very slow reaction and the detached wave occurs as an infinitely fast reaction, the finite rate chemistry is considered to be the key for the oscillating combustion induced by an OSW over a wedge of a finite length with a flow turning angle greater than the maximum attach angle for an ODW. Since this case has not been previously reported, grid independency was tested intensively to account for the interaction between the shock and reaction waves and to determine the critical time scale where the oscillating combustion can be observed.

Computation of premixed turbulent combustion using the probability density function approach

Fundamental properties of premixed flames in different regimes of combustion (i.e., flamelet and distributed combustion), including the respective region of transition, are compared by means of a unified stochastic-deterministic field theory. The calculations are based on the Monte Carlo solution of a modeled transport equation for the probability density function (pdf) of a reaction progress variable. In the pdf equation, it is essentially the molecular diffusion term that is different for the two types of combustion and, consequently, different models for the term have to be employed. The problem is overcome by considering stochastic and deterministic contributions to the dissipation of the fluctuations of the progress variable. Using a newly developed “mixing model”, the pdf transport equation provides a consistent description of both the laminar flamelet and the distributed regime. The close relationship between the modified pdf approach and a flamelet model based on the field equation of a scalar G( x, t ) is briefly discused. Furthermore, we study the influence of turbulence and stretch on the turbulent burning velocity and flame structure and compare the numerical results with experiments.

Plif imaging of hypersonic reactive flow around blunt bodies

Qualitative OH planar laser-induced fluorescence (PLIF) imaging experiments of supersonic reactive flows over a 40° wedge and a 19-mm-diameter blunted cylinder are reported. Hydrogen-, ethylene-, and methane-based mixtures have been used in this work. Three different free stream conditions. M∞=6.7, 5.2 and 4.2, were used to study the combustion modes around the wedge and blunted cylinder. Separate schlieren imaging results were also obtained for several of the test cases, providing complementary shock wave visualization. In several cases for the wedge, shock-induced combustion was observed a finite distance behind the nonreacting oblique shock wave. For the most sensitive hydrogen mixture, an oblique detonation occurred in which the shock wave and combustion front were fully coupled. In the experiments with the blunted cylinder, two different regimes of combustion were observed. When the velocity of the flow was larger than the CJ velocity of the mixture, the mode was that of an adiabatic shock followed by a smooth combustion front. For cases of flow velocity similar to the CJ velocity or less, regular oscillations were observed on the image and in the stagnation pressure history. These observations are qualitatively consistent with previous work.

Supersonic H2-air combustions behind oblique shock waves

In order to study the mechanisms of initiation and stabilization of H2-Air combustions (stoechiometric mixture initially atT 0=293 K andp 0=0.5 bar) in supersonic flow conditions behind an oblique shock wave (OSW), an original technique is used where OSW is generated in this mixture by the lateral expansion of the burnt gas behind a normal CJ gaseous detonation propagating into a bounding reactive mixture. Four Mach numberM of propagation of OSW are considered in the study, namelyM=7.7-6.1-4.4 and 3. Depending on the Mach numberM and inclinaison angleθ of OSW different regimes of combustion may occur in the driven mixture. For high values ofM (6.1 and 7.7) delayed steady overdriven oblique detonation waves (SODW) were obtained with a near CJ detonation wave as the critical regime. It was found that SODW obtained correspond quite well to prediction of the polar method. When thermal conditions behind the OSW are lower, either for high Mach number 6.1 and 7.7 for smaller angleθ than the previous case, or for lower Mach number, 4.4 and 3, the flame initiated at the apex is stabilized as a turbulent oblique flame behind the OSW. With much lower conditions, no combustion appears in the H2-Air mixture.