Flow, Turbulence and Combustion最新文献

The effect of non-parallel walls on the flame dynamics and the tulip flame formation in a narrow, closed volume channel is investigated experimentally using high-speed visualization techniques and pressure measurements. The angular channel is a (4^{circ}) planar diverging (D–C) or converging (C–C) channel ignited at one of the ends. The flame propagation is studied for a wide range of equivalence ratios ((phi=0.9-1.5)) of premixed LPG-air mixture in both the channel configurations, and the outcome is extended for a straight channel. The flame speed is higher in the D–C compared to the C–C and highest in the straight channel due to thermal expansion caused by the narrow volume and flame stretch in the varying cross-section. In addition, the pressure changes in the channels are recorded, and the results show that the rate of pressure varies with the flame growth and is not significantly affected by the flame surface area. Analysing the normalized position and time for the flame inversion for different (phi) and for all the channels, it is found that the flame inversion occurs prominently at halfway of the flame propagation and at about 1/3rd of the total time and it is true for both parallel and non-parallel walled channels at every equivalence ratio. Flame morphology with diverging and converging walls are presented. In particular, the flame deceleration process and tulip flame formation are visualized across different orthogonal cross-sections. Distorted tulip flame and the vortex flows are observed in the burned gas along with asymmetric tulip-shaped flame propagation at different equivalence ratios for the D–C and C–C.

Development of a generally applicable inlet condition generation method for Large Eddy Simulation (LES) is challenging and limits application to complex engineering flows. Inlet velocity time-series are required, at temporal/spatial resolutions consistent with LES numerics, covering the entire computational inlet plane, and for a time period allowing statistical stationarity. Ideally measurements would be used, but capture of large area, long duration time histories is problematic. Several generation techniques have been proposed, but compliance with measurements is normally guaranteed only for single point statistical data. The present work demonstrates how Stereoscopic Particle Image Velocimetry (SPIV) may be used to generate conditions simultaneously matching 1-point statistics, 2-point spatial correlations, and frequency spectra. A validation test case is selected containing complex flow structures typical of engineering applications. Linear Stochastic Estimation (LSE) and high-pass filtering are combined to match the LES inflow area with the smaller SPIV area required for accurate spatial resolution. A single synchronous velocity field is created from multiple non-concurrent SPIV fields. The method extends the inflow complexity that can be considered and provides improvement over existing methods.

In this work a new method of passive drag reduction on bluff bodies by tessellation is presented. Wind-tunnel measurements on tessellated spheres reveal that the variation of the drag coefficient is similar to other types of surface modifications with rotational symmetry such as dimples, manifested by a sudden decrease in the drag coefficient at a critical Reynolds number followed by a nearly constant drag value in the post-critical regime. However, tessellated spheres can achieve a further 10%-15% drag reduction compared to dimples without shifting the critical Reynolds number. To further investigate the underlying physics leading to this reduction we also conducted Direct Numerical Simulations of both tessellated and dimpled spheres at (Re=1.50times10^5). The predicted values of drag coefficient agree very well with the experiments and confirm the drag reduction. Analysis of the flow reveals that the tessellated panels introduce a smaller pressure “penalty” compared to dimples at the front part of the body. In addition, transition to turbulence occurs later and near the top of the body. As a result the boundary layer grows thinner and global separation is delayed by approximately (10^circ).

The underexpanded jet issuing from an elliptic convergent nozzle with an aspect ratio of 8 is quantitatively visualized using rainbow schlieren tomography. Flow visualizations are conducted at a nozzle pressure ratio of 4.0 to produce strong shocks within the jet plume. The Reynolds number, based on the equivalent diameter and flow properties at the nozzle exit, is 4.0(times)10(^5). Multi-view rainbow schlieren images with a horizontal filter setting are captured by rotating the nozzle around its central axis at an angular step of 5(^circ), ranging from 0(^circ) to 180(^circ). The jet density field is reconstructed at a spatial resolution of 13 (mu)m using the convolution back projection method. The near-field flow features of a shock-containing elliptic jet with a high aspect ratio are experimentally demonstrated for the first time. In addition, to examine the effects of angular steps on the reconstructed density field, the rainbow schlieren images taken at 10(^circ), 20(^circ), and 30(^circ) intervals are selected from a total of 37 rainbow schlieren images captured at 5(^circ) intervals. The density field is then reconstructed for these specific angular steps. The effects of angular steps are clarified on the two-dimensional density fields in the minor-axis and major-axis planes, as well as on the streamwise density profiles along the jet centerline and liplines in both the minor-axis and major-axis planes.

This study investigates the impact of (textrm{CO}_2) and (textrm{H}_{2}textrm{O}) dilution ratios on the characteristics of a premixed (textrm{C}_{3}textrm{H}_{8})/air turbulent flame in a swirled burner at atmospheric pressure. High-fidelity turbulence resolution is critical for capturing transient flame stabilization dynamics and pollutant formation in swirling flows. Therefore, Detached Eddy Simulation (DES) is employed to resolve large-scale unsteady turbulent structures, while the Eddy Dissipation Concept (EDC) models turbulence-chemistry interaction, incorporating a new reduced kinetic model made up of 36 species and 166 reactions. The study explores five volumetric fractions of (textrm{CO}_2) or (textrm{H}_{2}textrm{O}) dilution ((X_{textrm{CO}_2/textrm{H}_2textrm{O}} = 4%), 8%, 12%, 16%, and 20%), three swirl numbers ((Sn = 0), 0.6, and 1.05), and two equivalence ratios ((phi = 0.8) and 1). Validation against experimental data confirms the model’s accuracy in capturing flow fields and scalar distributions. The results show that (textrm{CO}_2) addition significantly lowers flame temperature and alters its shape, resulting in a major reduction in (textrm{NO}_{x}) concentrations at (X_{textrm{CO}_2} = 16%). Dilution by (textrm{H}_{2}textrm{O}) does not reduce (textrm{NO}_{x}) as noticeably, but still leads to somewhat more stable and thinner flames. Additionally, (textrm{CO}_2) is more effective than (textrm{H}_{2}textrm{O}) in suppressing flame flashback. This work provides interesting insights for optimizing swirl-stabilized flames.

Large uncertainties exist in the description of the turbulent premixed combustion regimes and their boundaries, especially the validity regime of flamelet assumptions, due to challenges involved in experimental studies. Our understanding of the physics of turbulent premixed combustion gets nebulous as the Karlovitz number increases (exceeds 500). A novel piloted burner design capable of generating high-intensity, homogeneous, axisymmetric turbulence is presented, and its flow characteristics are discussed. An impinging-jet injector is used to generate high-intensity turbulence. Axial flow measurements were made using hot-wire anemometry. Fluctuating rms velocities ranging from 1.4 m/s up to 30 m/s are reported for bulk velocities of 10 m/s to 60 m/s at the jet centerline. The measured velocity profiles were shown to be self-similar for all considered injector configurations. The limitations of hot-wire measurements in large turbulence and necessary corrections are discussed; the corrected length scales were compared to those obtained from scaling laws. Premixed methane-air flames exceeding 60 m/s of bulk velocities and equivalence ratios ranging from 0.65 to 1.0 can be stabilized using a large laminar pilot. The tested flame conditions span over two orders of magnitude in normalized turbulence velocities and three orders of magnitude in Karlovitz number. High-fidelity measurements of premixed flames over a broad range of turbulence can be realized using the burner.

The characteristics of a novel partially premixed pure hydrogen-oxygen combustion triple nozzle cluster unit are investigated by direct numerical simulation using a detailed chemical mechanism, mainly focusing on the mixing progress. Reactants are injected into the computational domain through a triple nozzle unit, with twin lateral oxidizer nozzles inclined inwards to impinge on a central fuel jet downstream of the inlets, creating a zone of intense turbulent mixing and heat release ignited by hot recirculated burnt products. The mixture fraction, normalized flame index and product-based reaction progress variable are analyzed to systematically describe reaction characteristics and the formation of premixed flame branches in the downstream neighborhood of the impingement point. Analysis of the transport budgets of elemental mixture fraction suggests that the improved overall mixing from jet impingement is due to an intensification of convective transport dispersing the fuel and oxidizer jets resulting in a larger diffusive interface. Turbulent combustion characteristics are explored for impingement angles of 60° and 75°, revealing a dominant effect on mixing intensity, heat concentration, flame length and alignment across premixed and non-premixed combustion regimes. These results shed light on the underlying reaction dynamics and parametric dependencies of the proposed multi-cluster configuration, providing a numerical reference for the development of advanced hydrogen partially premixed combustion systems utilizing jet impingement mixing.

Numerical experiments of targeted grid refinement are reported for stress-based wall-modeled large-eddy simulation (WMLES) of turbulent smooth-body separation. The flow over a Gaussian bump at a length-based (height-based) Reynolds number of 2 million (0.17 million) is simulated using an unstructured polyhedral solver on isotropic grids, and compared to high-fidelity Direct Numerical Simulation (DNS) data. The baseline coarse grid with 16 points per boundary-layer thickness in the upstream region does not capture smooth-body separation, while the baseline fine grid with double the resolution over half the boundary layer accurately captures the separated flow region indicating the importance of near-wall refinement. Refining the regions in the vicinity of the apex of the bump where the pressure gradient effects are expected to be dominant gave similar predictions to the baseline fine grid. Past DNS studies have noted the presence of an internal layer that begins to develop in the favorable pressure gradient region upstream of the bump apex, where its thickness is under 10% of the local boundary-layer thickness. However, the importance of sufficiently resolving the internal layer in predicting the flow separation accurately is unclear. Near-wall refinement just upstream of the apex targeted at the developing internal layer did not produce any flow separation, while near-wall refinement downstream of the apex in the vicinity of separation produced reasonable predictions, indicating that the latter is more critical for WMLES. While the present simulations contain uncertainties/errors due to the choice of grid topology, subgrid and wall model, these targeted grid refinement results provide useful insights to design optimal grids to capture turbulent smooth-body flow separation.

Based on the symmetry-theoretical analysis of the infinite hierarchy of multi-point moment equations, we rigorously derive the self-similar structure of velocity moments of arbitrary order in the turbulent round jet. Therein, we discover the dependency on a new parameter in the axial scaling laws. It has its roots in the statistical scaling symmetry which is connected to intermittency. This new parameter allows different states of self-similarity depending on the inflow condition, as postulated by George (in Arndt, R., George, W.K. (eds), Advances in Turbulence, pp. 39–73, 1989). However, the comparison with data fixes this parameter to zero, i.e. the classical scaling is recovered. Hence, intermittent effects are hidden when only considering the axial scaling laws. However, the influence of intermittency is still visible in the self-similar radial profiles. We find that a Gaussian-type curve fits the self-similar radial profiles of moments of arbitrary order of the axial velocity with high accuracy. The prefactors in the Gauss-function exponent exhibit a clear nonlinear dependency on the moment order, significantly deviating from a dimensional scaling. We attribute this to external or large-scale intermittency, which is likewise visible in the velocity PDF with increasing distance from the axis as we have found in a previous work (Nguyen and Oberlack, Phys. Rev. Fluids. 9(7), 074608, 2024). Furthermore, the statistical scaling symmetry reappears in the symmetry properties of the equations giving rise to the Gaussian profiles.

We perform Large-Eddy Simulations (LES) on the accelerating flow around streamwise-elongated rectangular cylinders with chord-to-depth ratios of 3:1 and 5:1 using Gaussian-type inflow accelerations of different intensities. The Reynolds numbers, defined with the freestream velocity and the crossflow dimension of the cylinder, range from Re = 17200 to Re = 65360. For both 3:1 and 5:1 rectangular cylinders the vortex shedding is characterized by constant-frequency time cells as observed in the literature for a square cylinder. For the 3:1 case, the Strouhal number variation range and the crossflow-force fluctuations within each time cell are the same for all cells. The results obtained under stationary inflow conditions for the rectangular 3:1 cylinder match well the statistical values computed in each time cell. On the other hand, for the 5:1 case, the cell-averaged recirculation region along the lateral side reduces in size during acceleration, leading to a narrower wake, decreased lift fluctuations, and higher Strouhal numbers. The shortening of the mean recirculation region with increasing Reynolds number for the 5:1 rectangular cylinder occurs at higher Reynolds numbers for accelerating inflows compared to stationary-inflow conditions. Finally, in agreement with what was observed for the square cylinder, for both considered aspect ratios the Strouhal number behaviors for accelerations of different severity collapse when plotted as a function of the Reynolds number.