We investigate the Lagrangian rotational and deformation dynamics of long flexible fibers in a turbulent channel flow at a shear Reynolds number equal to 300, using one-way coupled Euler-Lagrange direct numerical simulations. The fibers, modeled as chains of sub-Kolmogorov rod elements with prescribed bending stiffness, span more than one order of magnitude in length and are analyzed in three wall-normal slabs covering the near-wall, log-layer and bulk regions. We compute autocorrelations of the tumbling and spinning rates, power spectra of the fiber’s end-to-end vector, and waiting-time distributions of intense tumbling events. Spinning decorrelates on a fixed multiple of the local Kolmogorov time, independent of fiber length, whereas tumbling develops a memory that grows with fiber length by factors of two to five, reversing the ordering of the two timescales between the shortest and longest fibers. Flapping spectra separate into a high-frequency regime in which all fiber lengths carry the same energy in absolute terms and a low-frequency regime in which the energy increases with fiber length. Intense tumbling exhibits approximately exponential, weakly clustered waiting-time statistics whose rescaled form is universal across slabs, fiber lengths and bending rigidities, whereas the event rate depends on both fiber length and rigidity.
In this study, we investigate the role of phase separation in transcritical oxygen-hydrogen mixing layers under liquid rocket engine-relevant conditions. High-fidelity simulations of a splitter-plate configuration are performed to compare two thermodynamic approaches, a single-phase (SP) model and a multiphase model based on the vapor–liquid equilibrium (VLE) concept. Three operating pressures are considered, spanning near-critical to highly supercritical conditions, with fixed stream Reynolds numbers so that pressure is the only varied parameter. The two models share the same overall flow topology but follow different thermodynamic paths within the two-phase region. The SP model predicts local temperatures below the oxygen injection value, a non-ideal mixing effect absent under VLE, whose mixing path increases monotonically with hydrogen content. Accounting for phase separation results in a thinner thermal mixing layer, reduced temperature fluctuations, a lower growth rate, and modified near-wake dynamics, with a longer recirculation region behind the lip. Conditional statistics of the scalar dissipation rate show that the VLE model predicts more intense molecular mixing at compositions inside the two-phase region, which encompasses the stoichiometric condition most relevant to ignition. The spectral analysis localizes the model effect at the small scales for velocity, while density is affected across all scales. All discrepancies are largest at near-critical pressure and attenuate as the operating pressure moves away from the critical point.
Accurate prediction of Flame-Wall Interactions (FWI) is crucial for the design of compact, high-efficiency combustors but remains a significant challenge for Reynolds-Averaged Navier-Stokes (RANS) simulations. This study presents the development and validation of a lightweight data-driven closure for the mean reaction rate, trained on high-fidelity Direct Numerical Simulation (DNS) data of a V-shaped turbulent premixed flame. To ensure numerical stability and physical consistency, the training methodology incorporates a logarithmic transformation of the reaction rate target and a novel error-function transformation of the wall distance, effectively localizing the model’s sensitivity to the near-wall quenching zone. A-priori validation demonstrates that the neural network accurately captures the mean reaction rate statistics and extrapolates successfully to unseen spatial regions. However, a-posteriori assessment in coupled RANS simulations reveals a critical sensitivity to input feature consistency. The full-feature model (7 input parameters), which includes turbulent kinetic energy and dissipation rate as inputs, fails to predict the correct quenching behavior due to significant discrepancies between the DNS-extracted turbulence quantities used for training and the corresponding quantities obtained from RANS. In contrast, a reduced-feature model (3 input parameters), relying exclusively on the mean progress variable, its gradient, and the transformed wall distance, demonstrates remarkable robustness. By decoupling the prediction from uncertain turbulence quantities, the model with 3 input parameters accurately reproduces the V-flame structure, achieving performance comparable to a manually tuned algebraic closure. Furthermore, the reduced model exhibits strong generalization capabilities, successfully predicting the transient physics of a Head-On Quenching (HOQ) configuration that was not present in the training dataset. These findings highlight that for data-driven RANS closures, selecting input features that remain consistent across fidelity levels, is paramount for achieving predictive stability and generalization.
We propose corrections to the diffusion equation providing solutions consistent with nano-scale and multi-scale physics. These corrections can be useful for various applications, including nano-technology, strong-gradient and rarefied transport. The problem is that the classical diffusion equation (which terms are in the core of fluid dynamics models) does not carry information about the structure of the fluid, and fails, for example, at the nano- and multi-scale. As we show here, this information was lost due to the way the original equation was derived. In this paper, we reconsidered Einstein’s derivation of the diffusion equation to eliminate truncation of higher terms. The results give the finite difference equation of mass transport where information about the fluid structure is not lost; this is because the finite-difference length is not a numerical discretization parameter but a mean free path, λ, a fluid property determined by the structure of the fluid and by molecular mechanisms of collision. For this reason, the (λ-discretized) finite difference diffusion equation becomes consistent with nano-scale and multi-scale physics. This is demonstrated by specific examples.
Surface roughness has a critical influence on the aerothermal and combustion behavior of gas turbine burners, particularly with the growing use of additively manufactured (AM) components that exhibit inherently rougher surfaces than conventionally machined parts. This study investigates the impact of surface roughness on flow and flame dynamics in a premixed jet burner (PJB) operating under both isothermal-air and reacting hydrogen-fired conditions. Experimental measurements using Laser Doppler Anemometry (LDA) and OH* chemiluminescence were complemented by Computational Fluid Dynamics (CFD) simulations employing Reynolds-Averaged Navier–Stokes (RANS) and Detached Eddy Simulation (DES) frameworks. Building upon existing equivalent sand-grain roughness (kₛ) models, a novel correlation incorporating both measured roughness parameters and bulk flow velocity was developed to improve predictive accuracy for resolved boundary layers (y+ ≈ 1), a regime in which ks+ > y⁺ is common and standard roughness-function formulations lose validity. The new correlation demonstrated strong agreement with experimental data in isothermal simulations. When extended to reacting hydrogen–air flames, RANS simulations successfully reproduced roughness-induced flow redistribution and flame shortening, although DES tended to overpredict bluff-body effects and underperform under lean conditions. Overall, the proposed ks correlation, validated under isothermal conditions, provides a computationally efficient framework for modelling surface-roughness effects, and its extension to reacting hydrogen flames, particularly under RANS, offers a promising basis for future development toward additively manufactured gas turbine components.
