An eddy-viscosity based, subgrid-scale model for Large Eddy Simulations is derived from the analysis of the singular values of the resolved velocity gradient tensor. The proposed σ-model has by construction the property to automatically vanish as soon as the resolved field is either two-dimensional or two-component, including the pure shear and solid rotation cases. In addition, the model generates no subgrid-scale viscosity when the resolved scales are in pure axisymmetric or isotropic contraction/expansion. At last, it is shown analytically that it has the appropriate cubic behavior in the vicinity of solid boundaries without requiring any ad-hoc treatment. Results for two classical test cases (decaying isotropic turbulence and periodic channel flow) obtained from three different solvers with a variety of numerics (finite elements, finite differences or spectral methods) are presented to illustrate the potential of this model. The results obtained with the proposed model are systematically equivalent or slightly better than the results from the Dynamic Smagorinsky model. Still, the σ-model has a low computational cost, is easy to implement and does not require any homogeneous direction in space or time. It is thus anticipated that it has a high potential for the computation of non-homogeneous, wall-bounded flows.

The transient and quasi-steady flame structure of reacting fuel sprays produced by single-hole injectors has been studied using chemiluminescence imaging and Planar Laser-Induced Fluorescence (PLIF) in various constant-volume facilities at different research institutes participating in the Engine Combustion Network (ECN). The evolution of the high-temperature flame has been followed based on chemiluminescence imaging of the excited-state hydroxyl radical (OH *), and PLIF of ground-state OH. Regions associated with low-temperature chemical reactions are visualized using formaldehyde (CH 2 O) PLIF with 355-nm excitation. We compare the results obtained by different research institutes under nominally identical experimental conditions and fuel injectors. In spite of design differences among the various experimental facilities, the results are consistent. This lends confidence to studies of transient behavior and parameter variations performed by individual research groups. We present results of the transient flame structures at Spray A reference conditions, and include parametric variations around this baseline, involving ambient temperature, oxygen concentration and injection pressure. Key results are the observed influence of an entrainment wave on the transient flame behavior, model-substantiated explanations for the high-intensity OH * lobes at the lift-off length and differences with OH PLIF, and a general analogy of the flame structures with a spray cone along which the flame tends to locate for the applied parametric variations.

A detailed study on the spray local flow and flame structure has been performed by means of PIV and laser-sheet LIF techniques under Diesel spray conditions. Operating conditions were based on Engine Combustion Network recommendations. A consistent comparison of inert and reacting axial velocity fields has produced quantitative information on the effect of heat release on the local flow. Local axial velocity has been shown to increase 50 to 60% compared to the inert case, while the combustion-induced radial expansion of the spray has been quantified in terms of a 0.9 to 2.1 mm radius increase. As a result, the drop in entrainment rate has been quantified around 25% compared to the inert case. Streamline analysis also hints at a reduced entrainment under reacting conditions. A 1D spray model under reacting condition has been used, which confirms the modifications obtained in the main flow metrics when moving from inert to reacting conditions. When comparing the flow evolution with the flame structure, little effect of chemical activity on the spray flow upstream the lift-off length has been evidenced, in spite of the presence of formaldehyde in such regions. Only downstream of the lift-off length, as defined by OH LIF, has a strong change in flow pattern been observed as a result of combustion-induced heat release.

of subgrid-scale models with an LES-dedicated experimental database: the pulsatile impinging jet in turbulent cross-flow.1 Large-Eddy Simulation (LES) in complex geometries and industrial applications like piston engines, gas turbines or aircraft engines requires the use of advanced subgridscale (SGS) models able to take into account the main flow features and the turbulence anisotropy. Keeping this goal in mind, this paper reports an LES-dedicated experiment of a pulsatile hot-jet impinging a flat-plate in the presence of a cold turbulent cross-flow. Unlike commonly used academic test cases, this configuration involves different flow features encountered in complex configurations: shear/rotating regions, stagnation point, wall-turbulence, and the propagation of a vortex ring along the wall. This experiment was also designed with the aim to use quantitative and nonintrusive optical diagnostics such as Particle Image Velocimetry (PIV), and to easily perform an LES involving a relatively simple geometry and well-controlled boundary conditions. Hence, two eddy-viscosity-based SGS models are investigated: the dynamic Smagorinsky model 1 and the σ-model 2 . Both models give similar results during the first phase of the experiment. However, it was found that the dynamic Smagorinsky model could not accurately predict the vortex-ring propagation, while the σ-model provides a better agreement with the experimental measurements. Setting aside the implementation of the dynamic procedure (implemented here in its simplest form,i.e. without averaging over homogeneous directions and with clipping of negative values to ensure numerical stability), it is suggested that the mitigated predictions of the dynamic Smagorinsky model are due to the dynamic constant, which strongly depends on the mesh resolution. Indeed, the shear-stress near the wall increases during the vortex-ring impingement leading to a less refined mesh in terms of wall units, y + . This loss of resolution induces a poor damping of the dynamic constant, which is no longer able to adjust itself to ensure the expected y 3 -behaviour near the wall. It is shown that the dynamic constant is never small enough to properly balance the large values of the squared magnitude of the strain-rate tensor, 2S ij S ij . The experimental database is made available to the community upon request to the authors.

No abstract