Flow, Turbulence and Combustion最新文献

Gas turbine combustors commonly feature swirling flows. The swirl is usually characterized by the swirler geometry. The development of swirl-stabilized burners includes the experimental assessment of the resulting flow field and the quantification of the swirl, e.g. through Laser Doppler Anemometry (LDA). LDA accurately acquires flow velocity components in a small probing volume. Nevertheless, the measurement quality represents a trade-off between invested measurement time and spatial resolution. In this work, the potential of Physics-Informed Neural Networks (PINNs) to assimilate flow fields from sparse LDA measurements is investigated. A novel burner is employed in which the swirl is fluidically adjustable from a non-swirled jet to a fully swirled flow through a secondary air flow injection. Data is acquired through LDA within spatial measurement grids at multiple axial distances from the swirler. Assuming symmetry, axial, tangential, and radial velocity components are determined. A PINN is subsequently trained with the acquired data, creating a continuous and differentiable flow field representation by evaluating RANS equations. By systematically reducing the training data while evaluating the physical validity of the reconstructed field, a minimum training data requirement is identified. As a result, for three operating conditions, the flow field is adequately characterized by a minimum of measurement points.

We present a data-driven approach to Reynolds-averaged Navier-Stokes (RANS) turbulence closure modelling in magnetohydrodynamic (MHD) flows. In these flows the magnetic field interacting with the conductive fluid induces unconventional turbulence states such as quasi two-dimensional (2D) turbulence, and turbulence suppression, which are poorly represented by standard Boussinesq models. Our data-driven approach uses time-averaged Large Eddy Simulation (LES) data of annular pipe flows, at different Hartmann numbers, to derive corrections for the (k)-(omega) SST model. Correction fields are obtained by injecting time averaged LES fields into the MHD RANS equations, and examining the remaining residuals. The correction to the Reynolds-stress anisotropy is approximated with a modified Tensor Basis Neural Network (TBNN). We extend the generalised eddy hypothesis with a traceless antisymmetric tensor representation of the Lorentz force to obtain MHD flow features, thus keeping Galilean and frame invariance while including MHD effects in the turbulence model. The resulting data-driven models are shown to reduce errors in the mean flow, and to generalise to annular flow cases with different Hartmann numbers from those of the training cases.

Inside the engine combustor, addition of the coolant from the wall makes the physics of flame-wall interaction (FWI) even more complex. Considering the application relevance, wall heat flux is analyzed and modeled. Under various flow conditions, the model predictions satisfactorily match the direct numerical simulation (DNS) results. The effects of coolant on the entrained flame and head-on flame are clearly different. Statistics of the near-wall flame orientation and curvature are sensitive to the coolant blowing ratio (BR). The entrained flame is more likely to be swept away, while the head-on flame is more stable. Both the model and simulation indicate consistently that an increase in BR, although quantitatively small, will greatly reduce the wall heat flux induced by the head-on flame. In contrast, the change of wall heat flux induced by the entrained flame is much smaller. Since most of the near-wall flame is head-on, the BR effect is significant. Additionally, in an a priori large eddy simulation (LES) study, the model predictions show better consistency with DNS, in comparison with the most commonly used turbulence sub-grid models.

We present a machine learning–based framework for blending data-driven turbulent closures in the Reynolds-Averaged Navier–Stokes (RANS) equations, aimed at improving their generalizability across diverse flow regimes. Specialized models (hereafter referred to as “experts”) are trained via sparse Bayesian learning and symbolic regression for distinct flow classes, including turbulent channel flows, separated flows, and a near sonic axisymmetric jet. These experts are then combined intrusively within the RANS equations using weighting functions, initially derived via a Gaussian kernel on a dataset spanning equilibrium shear conditions to separated flows. Finally, a Random Forest Regressor is trained to map local physical features to these weighting functions, enabling deployment in previously unseen scenarios. We evaluate the resulting blended model on three representative test cases: a turbulent zero-pressure-gradient flat plate, a wall-mounted hump, and a NACA0012 airfoil at various angles of attack, ranging from fully attached to near-stall conditions. Results for these 2D flows show that the proposed strategy adapts to local flow characteristics, effectively leveraging the strengths of individual models and consistently selecting the most suitable expert in each region. Notably, the blended model also demonstrates robustness for flow configurations not included in the training set, underscoring its potential as a practical and generalizable framework for RANS turbulence modeling.

In this paper we report on Large Eddy Simulations of natural convection in water pools with evaporation across the free surface and at the hard turbulence regime. The free surface is approximated as a free-slip top boundary. The loss of water is estimated via a dynamic and inhomogeneous evaporation model. Also, the descent of the free surface is accounted for by regularly reducing the computational domain and applying a remeshing procedure. We present results for 4 different Rayleigh numbers, ranging from (boldsymbol{Ra = 1.35 times {10^8}}) to (boldsymbol{Ra{ = 10^{10}}}). Our simulations predict a slow decrease of the free-surface temperature and evaporation rate over time. This may be attributed to the descent of the free surface due to evaporation which tends to reduce the intensity of turbulent motions. On the other hand, the flow structure remains the same throughout the duration of the simulations. More specifically, the flow is organized in a large scale circulation aligned in a diagonal plane with smaller convective rolls near the corners of the domain. Also, the absence of a top hydrodynamic boundary layer enhances turbulent mixing and convective heat transfer near the free surface. This enhancement is manifested by a shift of the profile of the mean surface temperature towards the upper part of the domain, with the shift becoming more pronounced as the turbulence intensity increases. Herein we also provide results for the Nusselt number (boldsymbol{Nu}) and present a new (boldsymbol{Nu - Ra}) scaling for convection in pools and cavities that covers a large range of turbulence intensities.

The presence of obstacles in confined spaces results in high overpressure from premixed flame combustion and the specific obstacle configurations significantly affect flame dynamics. Although linear plate-type obstacles had been extensively explored for flame acceleration, the present study focused on obstacles with volume blockage. This paper investigated the effects of different shapes of such obstacles and their configurations on the flame propagation characteristics inside a partially confined geometry. Four different flame surface density models were tested: Algebraic Flame Surface Wrinkling model, Turbulent Flame Speed Closure, Algebraic model and Transport model. The Transport model by Weller was selected with the dynamic k-equation Large Eddy Simulation model for turbulence modelling. Four shapes of obstacles, triangular, rectangular, elliptical and circular were examined. The effect of separation distance (standard-100 mm, spaced-out-150 mm and squeezed-in-70 mm) between the obstacles was investigated, along with their number and configuration (in-line and staggered). The results revealed that for standard separation, the overpressure peak is maximum for triangular and minimum for circular obstacles. Staggering the obstacles reduced the peak overpressure. Further, the overpressure peak reduced with both increasing and reducing separation compared to the standard case for triangle, ellipse, and circle-shaped obstacles, whereas it increased with greater separation for rectangular obstacles. The most significant reduction across all cases was observed upon reducing the separation distance. Oscillatory pressure behaviour owing to combustion in unburnt mixture pockets is reported for rectangle and triangle obstacles, attributed to their minimal sphericity. The flame surface area, representative of the turbulence generated, is observed to be directly correlated with the peak overpressure value across the dataset.

Large Eddy Simulations (LES) are increasingly used in industry due to their superior accuracy compared to traditional statistical methods like Reynolds-Averaged Navier-Stokes (RANS) simulation. However, their high computational cost remains a major obstacle to performing daily parametric studies in engineering design offices. The objective of this work is to improve the efficiency of LES-based parametric studies through multi-fidelity surrogate modeling. Taking into account the computational cost of each turbulence modeling approach, multi-fidelity technic propose to combine limited number of LES results with more numerous RANS simulations. To achieve this, we use Artificial Neural Networks (ANN), which are particularly effective at capturing complex relationships between fidelity levels and handling discontinuities. To further reduce computational cost, we propose a new adaptive sampling strategy that selects high-fidelity LES points based on an estimation of interpolation error. This approach enhances the accuracy of the multi-fidelity method by efficiently allocating computational resources where they are most needed. The proposed strategy is first validated on an analytical test case before being applied to the study of the lift coefficient as a function of the angle of attack for a NACA0012 airfoil. We demonstrate that with only five LES evaluations, our method accurately captures the main features of this function, including the stall angle. This work paves the way for more efficient LES-based parametric studies.

A scaling relation has been derived to link the fractal dimension of a flame surface with the ratio of the normalised 3D flame surface area to its 2D counterpart. This derivation assumes an isotropic distribution of angles between the measurement plane and the flame’s normal vector, as well as a uniform distribution of angles between the principal direction and the flame’s tangent vector. The validity of the newly derived relation was assessed using an existing Direct Numerical Simulation (DNS) database of statistically planar turbulent premixed flames, encompassing a range of different Karlovitz numbers. The DNS data-based assessment revealed that the newly derived relations are reasonably accurate for the thin reaction zones regime flames, with the precision of predictions based on isotropy improving, as the Karlovitz number increases. Moreover, 2D measurements of the flame surface fractal dimension and the flame wrinkling factor can be effectively used to predict the actual 3D flame wrinkling factor for flames with Karlovitz numbers much greater than unity. Alternatively, the ratio of the 3D wrinkling factor to its 2D counterpart can provide a reasonable estimate of the 3D fractal dimension for flames in the thin reaction zones regime. The newly derived relations provide an estimation for the value of fractal dimension in the limit of high Karlovitz number using an alternative route.

An extension to multi-species and reacting flows of Doak’s “Momentum Potential Theory of Energy Flux carried by Momentum Fluctuations” is proposed as a general and comprehensive framework for thermoacoustic characterization of combustor systems. This framework is applied here for the first time in its extended form to analyze simulation data relative to the flow in a bluff-body stabilized combustor, in stable operating conditions. The proposed thermoacoustic model is able to: (i) unambiguously separate turbulent, acoustic, thermal, and mixture fluctuations; (ii) effectively describe the interaction between turbulent, acoustic, thermal, and mixture dynamics; (iii) highlight the main characteristics of the combustion noise emitted by the systems. By means of the performed analysis, the thermal phenomena are found to dominate the dynamics interaction. All convective quantities interact in the shear layer at the flame border and feature a similar, low-frequency spectral behavior. As expected, the acoustics does not couple directly with the convective quantities, due to the considered stable conditions. Although, the acoustic spectrum is strongly characterized by three peaks, which can be attributed to secondary, high-frequency thermal fluctuations. The modes related to these peaks can be seen, therefore, as a representation of the combustion noise emitted by the flame. The new terms related to the mixture do not seem to effectively contribute to the dynamics interaction and to the acoustic production, at least for the considered configuration and operating conditions.