Shock Turbulent Boundary Layer Interaction Research Articles

Comparative study of turbulence models in CFD for transonic flow over the ONERA M6 wing

Transonic flight presents significant challenges due to complex phenomena such as shock waves and turbulent boundary layer interactions, often leading to flow separation and instabilities. This study aims to evaluate the performance of various turbulence models in predicting these phenomena using Computational Fluid Dynamics (CFD) with ANSYS-Fluent 2020R1. The ONERA M6 wing model is analyzed under specific flow conditions (angle of attack, α = 3.06°, Mach number, M∞ = 0.8395, and Reynolds number, Re = 11.72 × 10^6) to assess five turbulence models: Spalart–Allmaras, k-ε Standard, k-ε Realizable, k-ω Standard, and k-ω SST. Results from the simulations closely matched experimental data, with all models demonstrating a margin of error within 5% compared to NASA’s CFD benchmarks. Notably, the Spalart–Allmaras and k-ω SST models showed superior accuracy in predicting shock formation and pressure distribution, highlighting their potential for enhanced predictive capabilities in aerodynamic analysis. This study demonstrates the effectiveness of ANSYS-Fluent in aerodynamic performance evaluation and suggests pathways for advancing aircraft design efficiency and safety through improved turbulence modeling.

Aerodynamic heating in hypersonic shock wave and turbulent boundary layer interaction

In hypersonic flight the shock wave and turbulent boundary layer interaction (STBLI) sharply increases wall heat transfer that intensifies the aerodynamic heating problems. In this work the STBLI is modelled by compression ramp flow with a Mach number of 5, a Reynolds number based on momentum thickness of 4652 and a wall to recovery temperature ratio of 0.5. The aerodynamic heat generation and transport mechanisms are investigated in the interaction based on theoretical analysis and direct numerical simulation (DNS) that agrees with previous studies. A prediction correlation of wall heat flux in STBLI is deduced theoretically and validated by some representative data including the present DNS, which improves the prediction accuracy and can be applied to a wider $Ma$ range compared with the canonical Q-P theory. The correlation indicates that the sharp increase of wall heat transfer in the STBLI can be explained by the boundary layer compression and the convection transport enhancement. Based on the DNS results, the aerodynamic heat generation and transport mechanisms are revealed in the separation, recirculation and reattachment zones in the STBLI. From this perspective, the peak heat flux can be further explained by the enhancement of near-wall turbulent energy dissipation, compression aerodynamic heat generation and the near-wall turbulent transport. The generation and transport of compression aerodynamic heat reveal the underlying mechanism of the strong correlation between the peak heat flux ratios and the pressure ratios in STBLIs.

Structure and Dynamics of Bleed-Controlled Impinging Shock/Turbulent-Boundary-Layer Interactions

Porous boundary-layer bleed alleviates the detrimental effects of shock/turbulent-boundary-layer interactions (SBLIs) in high-speed vehicles. However, many underlying mechanisms through which bleed influences SBLI dynamics remain poorly understood, including unsteady separation mitigation, effects on low-frequency spectra, and turbulent kinetic energy (TKE) modulation. Proper resolution of the dynamics of these phenomena motivates the use of wall-resolved large-eddy simulations. The configuration is based on experiments at Mach 2.5 and a momentum-thickness Reynolds number of 2800, on which a shock of flow-deflection angle of 8 deg impinges. Porous bleed patches are considered with resolution of the flow structure in individual holes. Two different suction strengths are examined, denoted “half-” and “full-bleed” cases based on the sonic mass flow coefficient. Although both bleed cases localize and reduce the mean reversed flow region, only the full-bleed case successfully reduces unsteadiness, TKE, wall-pressure loading, and overall distortion due to the SBLIs. These distinctions are reflected in various local and global effects of bleed, which are examined with three-dimensional modal analysis. Among the key findings is energy shifting from the low-frequency separation behavior to a higher-frequency signature, which modulates the bleed-associated shocks and expansions. Overall, the study highlights the complex interactions between individual bleed-hole flow structures that cumulatively yield the observed overall effects, which are important in optimizing bleed systems for various design objectives.

Extremely high wall pressure events in shock wave and turbulent boundary layer interactions using DNS data

This study investigates high-amplitude Extreme Wall Pressure fluctuation Events (EWPEs) in Shock wave/Turbulent Boundary Layer Interactions (STBLIs) through the conditional sampling of direct numerical simulation databases. The aim is to evaluate the effect of STBLIs and their strength on the statistical properties and associated turbulent structures of EWPEs using the conditional-averaging and clustering method. The temporal statistical results show that the occurrence probability and contribution ratio of EWPEs decrease downstream of strong STBLI, but their duration and interval time increase. Regarding two-dimensional wall pressure structures, the large population of small-scale structures becomes more elongated, but strong interactions induce a greater number of large-scale structures. The pairing of wall pressure events with a higher occurrence probability is verified by the joint probability density functions. Conditional analysis reveals that, as the interaction strength increases, the ejection motion associated with positive events occurs farther downstream and the spanwise vortex core locating above negative events is lifted up along the wall-normal direction. Moreover, analysis associates the paired wall pressure events with the sweep, ejection, and swirl motions in STBLIs, where hairpin eddies play an important role in the formation of positive–negative paired wall pressure structures.

Effects of herringbone riblets on shock-wave/turbulent boundary-layer interactions

In the experiment reported in this paper, a spanwise array of herringbone riblets strips is applied upstream of the interaction region between an impinging shockwave and a turbulent boundary layer developing along a flat plate boundary layer at a freestream Mach number of 1.85. High-speed schlieren photography, surface oil flow visualization and PSP are used to examine the effect of herringbone riblets on the characteristics of the shockwave-boundary layer interactions. It shows that despite that the height of the riblets is less than 2 % of the local thickness of the undisturbed boundary layer, a profound modification to the shockwave-boundary layer interaction zone is observed, and the overall size of the separation bubble is decreased. The observed control effects are attributed to the large-scale secondary flow structures produced by the directional riblets. To the best knowledge of the author, this is the first study evaluating the control effect of herringbone riblets on shock wave and turbulent boundary layer interactions in supersonic flow using the experimental method.

Wall pressure fluctuations in wall heated and cooled shock wave and turbulent boundary layer interactions

Wall pressure fluctuations in shock wave and turbulent boundary layer interactions on heated and cooled walls are investigated by performing direct numerical simulations. The objectives of this study are to identify flow dynamics causing low- and high-frequency wall pressure fluctuations and evaluate wall heating and cooling effects on the wall pressure fluctuations. The wall pressure spectra highlight two characteristic low-frequency spectral peaks at the separation point and below the expansion waves. Conditional analysis reveals that the two spectral peaks originate from the oscillations of the reflected shock and expansion waves, coupled with the expansion and contraction motions of the separated turbulent shear layer. Regarding wall temperature effects, wall cooling enhances the magnitude of the wall pressure fluctuations regardless of the low- and high-frequency components. Furthermore, influences of the expansion waves are also enhanced by wall cooling, which increases the wall pressure fluctuations locally after the separation.

Flow and surface loading induced by a supersonic jet over an aft deck

Aircraft designs may include an engine that exhausts over the upper fuselage of the aircraft. This design has many benefits including reduced noise below the aircraft. However, the scrubbing of the high-speed turbulent jet flow over the fuselage has the potential to cause structural fatigue. This paper describes a joint experimental/computational study of the surface pressure fluctuations induced by the supersonic jet flow. The model configuration consists of a rectangular convergent-divergent nozzle that exhausts over a flat plate. The plate is a continuation of the lower lip of the nozzle. Pressure measurements are made at several locations on the plate surface. In addition, schlieren visualization is used to identify the jet's shock cell structure. Numerical simulations are performed using Large Eddy Simulation (LES). Results are compared between the simulations and the experiments. The importance of shock turbulent boundary layer interaction to the surface loading is identified.

Flow features of swept shock/turbulent boundary-layer interaction due to a gap beneath a sharp fin

A numerical study of the shock/turbulent boundary-layer interaction (STBLI) generated by a sharp fin with a 60deg leading-edge sweep mounted above a flat plate at Mach 2.5 was performed. Incipient separation due to this highly swept fin was determined by varying the fin angle to the incoming freestream. The effect of the fin/plate gap on STBLI incipient separation was established. Sweeping the fin increases the deflection required for incipient separation. Sweeping the fin reduces the interaction strength. A gap causes flow leakage under the fin, further reducing the interaction strength on the windward side. The incipient angle of attack is increased when the gap is large enough. In this study, all gaps investigated produced incipient separation. For the leeward side of the fin, a vortex at the fin leading edge was generated due to the pressure difference across the fin leeward and windward sides. In the cases with gaps, an additional vortex is generated at the fin leading edge and a vortex at the plate surface is also produced. It was found that the effect of the gap is to change the location and shape of these additional vortices.

Evaluation of novel wall function approach for supersonic flow problems involving separations induced by geometry and shocks

Supersonic combustion ramjet engine involves complex flow physics including shock-turbulent boundary layer interaction (STBLI) and shock-induced separations. The general approach while using an averaged Navier-Stokes equations to capture these flow separation, due to the flow physics and/or geometry accurately, is to use a low-Reynolds turbulence model. This requires the placement of first grid point in the viscous sublayer ( y + ≃ 1 ), which implies a need of finer mesh near the wall leading to an overall increase in number of cells in the computational domain and thus an increase in overall computational time. The wall-function method is a substitute to this approach which allows for the use of a coarser mesh in the region near to the wall. However, the standard wall-function methods are generally not very successful in capturing flow separations. In this work, we have employed a generalized wall function approach which incorporates the effect of both pressure gradient and compressibility in its formulation and have evaluated its performance for complex supersonic flow cases involving shock-turbulence boundary layer interactions (STBLI) and shock induced separations, which have not been done previously. We have shown that this method captures the flow physics accurately even with much coarser meshes near the wall. Further, insensitivity of the method to the placement of first grid point has also been demonstrated.

Hypersonic shock wave and turbulent boundary layer interaction in a sharp cone/flare model

A direct numerical simulation of hypersonic Shock wave and Turbulent Boundary Layer Interaction(STBLI) at Mach 6.0 on a sharp 7° half-angle circular cone/flare configuration at zero angle of attack is performed. The flare angle is 34° and the momentum thickness Reynolds number based on the incoming turbulent boundary layer on the sharp circular cone is Reθ = 2506. It is found that the mean flow is separated and the separation bubble occurring near the corner exhibits unsteadiness. The Reynolds analogy factor changes dramatically across the interaction, and varies between 1.06 and 1.27 in the downstream region, while the QP85 scaling factor has a nearly constant value of 0.5 across the interaction. The evolution of the reattached boundary layer is characterized in terms of the mean profiles, the Reynolds stress components, the anisotropy tensor and the turbulence kinetic energy. It is argued that the recovery is incomplete and the near-wall asymptotic behavior does not occur for the hypersonic interaction. In addition, mean skin friction decomposition in an axisymmetric turbulent boundary layer is carried out for the first time. Downstream of the interaction, the contributions of transverse curvature and body divergence are negligible, whereas the positive contribution associated with the turbulence kinetic energy production and the negative spatial-growth contribution are dominant. Based on scale decomposition, the positive contribution is further divided into terms with different spanwise length scales. The negative contribution is analyzed by comparing the convective term, the streamwise-heterogeneity term and the pressure gradient term.

Characteristics of reattached boundary layer in shock wave and turbulent boundary layer interaction

The reattached boundary layer in the interaction of an oblique shock wave with a flat-plate turbulent boundary layer at Mach number 2.25 is studied by means of Direct Numerical Simulation (DNS). The numerical results are carefully compared with available experimental and DNS data in terms of turbulence statistics, wall pressure and skin friction. The coherent vortex structures are significantly enhanced due to the shock interaction, and the reattached boundary layer is characterized by large-scale structures in the outer region. The space-time correlation of fluctuating wall shear stress and streamwise velocity fluctuation reveals that the structural inclination angle exhibits a gradual decrease during the recovery process. The scale interactions are analyzed by using a two-point amplitude modulation correlation. A possible mechanism is proposed to account for the strong amplitude modulation in the downstream region. Moreover, the mean skin-friction is decomposed to understand the physically informed contributions. Unlike the upstream Turbulent Boundary Layer (TBL), the contribution associated with the Turbulence Kinetic Energy (TKE) production is greatly amplified, while the spatial growth contribution induced by the pressure gradient largely inhibits skin-friction generation. Based on bidimensional empirical mode decomposition, the turbulence kinetic energy production contribution is further split into different terms with specific spanwise length scales.

Influence of separation structure on the dynamics of shock/turbulent-boundary-layer interactions

Shock/turbulent-boundary-layer interactions (STBLIs) are ubiquitous in high-speed flight and propulsion applications. Experimental and computational investigations of swept, three-dimensional (3-D) interactions, which exhibit quasi-conical mean-flow symmetry in the limit of infinite span, have demonstrated key differences in unsteadiness from their analogous, two-dimensional (2-D), spanwise-homogeneous counterparts. For swept interactions, represented by the swept–fin-on-plate and swept–compression–ramp-on-plate configurations, differences associated with the separated shear layers may be traced to the intermixing of 2-D (spanwise independent) and 3-D (spanwise dependent) scaling laws for the separated mean flow. This results in a broader spectrum of unsteadiness that includes relatively lower frequencies associated with the separated shear layers in 3-D interactions. However, lower frequency ranges associated with the global “breathing” of strongly separated 2-D interactions are significantly less prominent in these simple, swept 3-D interactions. A logical extension of 3-D interaction complexity is the compound interaction formed by the merging of two simple interactions. The first objective of this work is therefore to analyze the more complex picture of the dynamics of such interactions, by considering as an exemplar, wall-resolved simulations of the double-fin-on-plate configuration. We show that in the region of interaction merging, new flow scales, changes in separation topology, and the emergence of lower-frequency phenomena are observed, whereas the dynamics of the interaction near the fin leading edges are similar to those of the simple, swept interactions. The second objective is to evolve a unified understanding of the dynamics of STBLIs associated with complex configurations relevant to actual propulsion systems, which involve the coupling between multiple shock systems and multiple flow separation and attachment events. For this, we revisit the salient aspects of scaling phenomena in a manner that aids in assimilating the double-fin flow with simpler swept interactions. The emphasis is on the influence of the underlying structure of the separated flow on the dynamics. The distinct features of the compound interactions manifest in a centerline symmetry pattern that replaces the quasi-conical symmetry of simple interactions. The primary separation displays topological closure to reveal new length scales, associated unsteadiness bands, and secondary flow separation.

Characteristics of wall-shear stress fluctuations in shock wave and turbulent boundary layer interaction

The wall-shear stress (WSS) fluctuations in the interaction of an oblique shock wave with a flat-plate turbulent boundary layer are investigated by means of direct numerical simulation (DNS) at Mach 2.25. The numerical results agree very well with previous experiments and DNS data in terms of turbulence statistics, wall pressure, and skin friction. The fluctuating WSS characteristics, including probability density function (PDF), frequency spectrum, space–time correlation, and convection velocity, are analysed systematically. It is found that the positively skewed PDF shape of the streamwise WSS fluctuations is significantly changed due to the presence of a separation bubble, while the PDF shape of the spanwise component is slightly affected, exhibiting a symmetric behaviour across the interaction. The weighted power-spectrum density map indicates that the low-frequency unsteadiness associated with the separated shock - exhibits little influence on the spectrum for either component, and no enhancement of the low-frequency energy is observed. A significant reduction in the spatial extent of the two-point correlation is observed, causing spanwise elongated coherence for the streamwise WSS fluctuations in the separation region. Moreover, the elliptic behaviour of the space–time correlations is essentially preserved throughout the interaction, and this is accompanied by a sudden reduction of the convection velocity in the separation bubble.

STATISTICAL CHARACTERISTICS OF PRESSURE FLUCTUATION IN SHOCK WAVE AND TURBULENT BOUNDARY LAYER INTERACTION

Shock wave and turbulent boundary layer interaction widely exists in the internal and external flow of high-speed aircraft. The aerodynamic performance and flight safety of aircraft are seriously affected by the strong pressure fluctuation in the interaction region. To investigate statistical characteristics of fluctuating pressure, the interaction between an incident shock of 33.2° and a spatially developed Mach 2.25 turbulent boundary layer is analyzed by means of direct numerical simulation (DNS). The numerical results have been carefully validated against with previous experiment and DNS at similar flow conditions in terms of mean velocity profile, turbulence intensity and wall pressure distribution. Statistics at the wall and in the outer layer, including fluctuation intensity, power spectral density, two-point correlation and space-time correlation, are quantitatively compared. The differences between them are analyzed in detail. It is found that the effect of the shock interaction on the wall-pressure fluctuation and the fluctuating pressure in the outer layer are utterly different. Based on the analysis of the power spectra density, the fluctuations in the separated region are both characterized by the low-frequency content, but in the reattachment region, the peak frequency of outer pressure fluctuations quickly shifts to higher frequency, with the low-frequency energy of wall-pressure fluctuation still being predominant. It is identified that the two-point correlations of pressure fluctuation at the wall and in the outer layer are both more elongated in the spanwise direction than that in the streamwise direction. The integral scale at the wall is generally increased, while the one in the outer layer increases sharply after passing the shock and then gradually decreases. The analysis of space-time correlation indicates that the iso-correlation contours are similar to the elliptical distribution and the convection velocity deduced by the correlation is dramatically decreased. Downstream of the interaction, the convection velocity in the outer layer is higher than that of wall-pressure fluctuation.

Anisotropic SST turbulence model for shock-boundary layer interaction

Menter SST k-ω is a Reynolds-averaged Navier–Stokes based two-equation turbulence model routinely used in industry for predicting aerodynamic flows. It shows excellent performance for low-speed flows, but gives inconsistent predictions for high-speed shock-induced separated flows. The model assumption of using a constant value of 0.31 for the structure parameter contradicts experimental observations. The model is also unable to predict Reynolds stress anisotropy generated by shock waves. In this work, we augment the SST model with quadratic eddy viscosity formulation of an explicit algebraic Reynolds stress model. A new relation for the structure parameter is proposed, making it a function of the local strain-rates and is no longer a constant in the regions of shock/turbulent boundary layer interaction (SBLI). Additional shock-physics is introduced using (Sinha et al., 2003) shock-unsteadiness model and an upper limit to the value of structure parameter is set in regions of shock waves. The new model, termed as SUQ-SST, is validated using a number of SBLI test cases ranging from supersonic to hypersonic speeds and near-incipient to fully-separated flows. Results show that the modifications do not alter the boundary layer prediction capability of the SST model. On the other hand, the new model gives significant improvement in predicting Reynolds stress anisotropy, flow separation, and surface properties in a wide range of SBLI flows.

Reflected shock/turbulent boundary layer interaction structure analysis based on large eddy simulation

In order to understand the physical phenomenon of the reflected shock/turbulent boundary layer interaction, the Large Eddy Simulation (LES) is conducted to investigate shock wave and turbulent boundary layer interaction in a 12° compression ramp with inlet high Mach number of 2.9. Rescaling/recycling method is used as inflow turbulence generation technique and validated on a supersonic flat plate turbulent boundary layer. The flow field of recycling plane in the plate computation domain is obtained to give the inlet boundary condition for the LES computation. This paper focuses on the reflected shock/turbulent boundary layer interaction region, where the fine flow structure and instantaneous flow field are analyzed in detail. It is found that the unsteady motion of the shock wave leads to the increase of wall pressure fluctuation.

Experimental study on shock and turbulent boundary layer interactions under different incident shock waves

In this paper, the interactions between shock wave and the turbulent boundary layer of an incoming wall were experimentally studied. The experiments were performed in a Mach 3.4 supersonic low-noise wind tunnel at the unit Reynolds number of 6.30×106 m−1. Under different incident shock wave conditions, fine flow field structure, wall pressure, and wall temperature distribution were determined. Relationship was also analyzed. Instantaneous fine structures of typical flow fields with shock wave and turbulent boundary layer interactions were investigated using the nano-tracer planar laser scattering technique, and the transient flow images were analyzed statistically. The oscillation position of the induced shock wave satisfies a normal distribution. Simultaneously, the wall pressure distribution under different incident shockwave conditions of the interaction region was determined using the flush air data sensing system, and the relationship between flow the structure of the separation region and the wall pressure distribution with different incident shock wave intensities is analyzed. Using a temperature sensitive paint technology, the wall temperature distribution under different incident shockwave conditions of the interaction region model was determined. Also, the relationship between the strip structure caused by Gortler vortices and the incident shock wave intensity in the shockwave/turbulent boundary layer interaction was analyzed.

Experimental Study on Unsteady Characteristics of Shock and Turbulent Boundary Layer Interactions

In this paper, the experimental study on the shock wave and turbulent boundary layer interactions was performed in the Mach 3.4 supersonic low-noise wind tunnel. The angle of the shock generator was θ = 15°, and the unit Reynolds number of 6.30 × 106 m–1. The wall temperature and pressure distribution during the disturbance process of the shock wave and the turbulent boundary layer were obtained based on the temperature-sensitive paints technique and the flush air data sensing system, and the basic flow field of the interactions region was partitioned. Meanwhile, based on the nano-tracer planar laser scattering technique, and the instantaneous fine structures in the interaction region were obtained and the spatiotemporal evolution characteristics of the flow structure were analyzed. The flow visualization images showed that the oscillation position of the induced shock wave satisfied the normal distribution. Compared with the flow visualization images and the temperature results, the correlation between the flow structure of the interactions region and the wall temperature change was obtained. At the same time, the wall fluctuation pressure of the center surface of the shock wave and turbulent boundary layer interactions region was measured by the high-frequency pulsating pressure sensor. The power spectrum density results showed that under the action of the shock wave incident by the shock generator, there were two characteristics frequency signal of 12 and 30 kHz in the induced shock oscillation interval. For the signal of 12 kHz, the frequency value and the amplitude were increased from the turbulent boundary layer to the separation bubble, and the oscillation energy of the induced shock wave was enhanced. The amplitude of the peak signal of each measurement point gradually decreased from the separation bubble to the reattachment zone, and the energy was gradually attenuated. For the high frequency signal of 30 kHz, the frequency variation of each channel was relatively small, relatively stable, and the energy was concentrated.

Dynamics of strong swept-shock/turbulent-boundary-layer interactions

The mechanisms of unsteadiness in nominally two-dimensional (2-D) shock/turbulent-boundary-layer interactions (STBLIs) cannot be directly extended to three-dimensional (3-D) STBLIs, because of differences in interaction structure; swept 3-D interactions, including the sharp-fin and swept-compression-ramp configurations, are of particular interest in this work. Complications arise from the observation that the separation length employed to scale low-frequency unsteadiness in 2-D (spanwise homogeneous) interactions is not a global property of 3-D (swept) interactions, due to the quasi-conical symmetry of the latter. Also, flow separation in 3-D interactions is topologically different, in that closure of the primary separation cannot occur without breaking the quasi-conical symmetry of the interaction – consequently, the unsteady properties of the separation are different. To address these points, large-eddy simulations are performed to assess unsteadiness in 3-D interactions, with the aim of understanding key differences relative to analogous 2-D interactions, the former of which have received less attention in the literature. The mechanism underlying the prominent band of low-frequency unsteadiness (two decades below the characteristic boundary-layer frequency) is shown to be significantly muted in swept interactions. An interesting scaling for the band of mid-frequency unsteadiness is uncovered (at least one decade below the characteristic boundary-layer frequency). This is a consequence of the observed connection between coherent fluctuations in the separated shear layer and local mean-flow gradients, indicating a mix between competing 2-D and 3-D free-interaction scaling laws. In contrast, high-frequency fluctuations largely retain the 2-D scaling introduced by the incoming turbulent boundary layer. The spatial structure of the mid-frequency coherence in 3-D STBLIs is isolated, revealing the significant influence of these convective coherent structures on shock rippling/corrugation, as well as a spanwise dependence of coherence size consistent with the 3-D mean-flow similarity scaling. Finally, the dynamic linear response of a representative 3-D interaction is compared to that of a representative 2-D interaction; the absolute instability present in the 2-D interaction is not present in the 3-D interaction. The coincident absence of both the absolute instability and associated band of low-frequency unsteadiness in 3-D STBLIs underscores the significance of this absolute instability in facilitating low-frequency unsteadiness in 2-D interactions.

Computational Analysis of Transverse Sonic Injection in Supersonic Crossflow Using RANS Models

AbstractTransverse injection at sonic speed from a rectangular slot into a supersonic crossflow is numerically explored with an indigenously developed parallel three-dimensional (3D) Reynold-averaged Navier–Stokes (RANS) solver for unstructured grids. The RANS models used for turbulence closure are the one-equation Spalart–Allmaras (SA) model and the two-equation shear stress transport (SST) model. For each model, the influence of compressibility corrections is assessed. Due to the presence of shock-turbulent boundary layer interaction (STBLI) in the flow, various STBLI corrections are assessed for both the models. Most of the simulations are two-dimensional (2D), but three-dimensional simulations are also performed to investigate the mismatch between the experimental dataset and the numerical results. The SA model is less sensitive to STBLI corrections, but some improvement in its prediction of the separation distance is found with the compressibility corrections. The SST model results are insensitive to the compressibility corrections, but the STBLI correction improves its results. Improved agreement with the experimental dataset is found when simulations are done in 3D, suggesting that the experiments were not so close to 2D as previously believed.