Compression Ramp Interaction Research Articles

Self-supervised transformers for turbulent flow time series

There has been a rapid advancement in deep learning models for diverse research fields and, more recently, in fluid dynamics. This study presents self-supervised transformers' deep learning for complex turbulent flow signals across various test problems. Self-supervision aims to leverage the ability to extract meaningful representations from sparse flow time-series data to improve the transformer model accuracy and computational efficiency. Two high-speed flow cases are considered: a supersonic compression ramp and shock-boundary layer interaction over a statically deformed surface. Several training scenarios are investigated across the two different supersonic configurations. The training data concern wall pressure fluctuations due to their importance in aerodynamics, aeroelasticity, noise, and acoustic fatigue. The results provide insight into transformers, self-supervision, and deep learning with application to complex time series. The architecture is extendable to other research domains where time series data are essential.

Spanwise unsteadiness in the sidewall-confined shock-wave/boundary-layer interaction

Three-dimensional effects of sidewalls on the low-frequency unsteadiness of the shock-wave/boundary-layer interaction (SBLI) are of academic and practical importance but not yet well understood. Considerable attention has been paid to the viscous effect of sidewalls, whereas the potential inviscid confinement effect of sidewalls has received little attention. The present work provides experimental evidence of multiscale spanwise travelling waves crossing the separation front under the confinement of sidewalls. Global pressure measurements were made for a sidewall-confined 24 $^\circ$ compression ramp interaction in Mach-2.83 flow using fast-responding pressure-sensitive paint. The unsteady pressure in a statistically two-dimensional intermittent region suggests that in addition to the canonical streamwise oscillation, the separation front exhibits significant low-frequency, multiscale spanwise distortion. Modal analysis further reveals that multiscale spanwise unsteadiness has higher intensity and frequency than the streamwise oscillation. Such strong spanwise unsteadiness calls attention to the low-frequency unsteadiness in previous sidewall-confined SBLI experiments and encourages further study on the mechanism of the confinement effect.

Influence of jet spacing in spanwise-inclined jet injection in supersonic crossflow

In separation control with air-jet vortex generators in supersonic flow, the spacing of the jets in a row array has a crucial effect on the control effectiveness. Previous experimental studies have revealed the overall influence of jet spacing on air-jet vortex generator controlled shock-induced separation zones, focusing mostly on the separation regions. There is, however, a gap in knowledge regarding the mechanisms leading to these changes in control effectiveness, particularly on the interaction between multiple jets and the details of downstream flow evolution. Therefore, the objective of the current study is to provide detailed information on the underlying flow dynamics associated with the injection of a row of spanwise-inclined jets into a supersonic turbulent boundary layer – and in particular the effects of different jet spacings. Four different spacings were studied with large-eddy simulations. In addition, we performed oil-flow and schlieren visualizations of separation control in a 24$^\circ$compression–ramp interaction with different jet-spacing configurations to validate and discuss our conclusions regarding the effects of jet/jet interactions on the separation-control effectiveness. We analyse the influence of jet spacing on the flow topology, induced vortical structures, and boundary-layer statistics. The jet-induced major vortex pair is the dominant flow structure energizing the near-wall boundary layer. The paper details how interactions amongst adjacent vortex pairs differ for varying jet spacings, thus influencing the momentum transfer and eventually the control efficiency. The dynamic behaviour of the flow was analysed using a three-dimensional dynamic mode decomposition. The resulting insights are key to the development of efficient control set-ups.

Fast PSP measurement of three-dimensional low-frequency unsteadiness in sidewall-confined shock wave/turbulent boundary layer interaction

The presence of sidewalls exerts three-dimensional (3D) effects on shock wave/turbulent boundary layer interaction (SBLI), which results in experimental challenges due to the highly unsteady flow with complex 3D patterns. The low-frequency large-scale flow patterns of a 24° sidewall-confined compression ramp interaction were investigated using fast pressure-sensitive paint. The experiment was conducted in a Mach 2.5 wind tunnel, and the unit Reynolds number was Re = 1.12 × 107/m. The sampling rate of the fast pressure-sensitive paint was 4 kHz, and the spatial resolution was 0.175 mm/pixel. In-situ calibration via several pressure transducers was used to determine the time-resolved global pressure distributions. Correlation analysis and proper orthogonal decomposition analysis were used to extract the large-scale unsteady flow patterns. The time-averaged results provide a global view of the typical flow structure, which consists of primary interaction zone and corner interaction zones. The time-resolved results show the asymmetry of the instantaneous pressure distribution induced by the sidewalls, and spanwise differences of the pressure fluctuation amplitudes in the interaction zones are observed. In addition to the common streamwise breathing mode, an interesting spanwise antisymmetric mode is revealed that provides clear evidence of the strong 3D effects in sidewall-confined SBLI.

Control of Shock-Induced Separation of a Turbulent Boundary Layer Using Air-Jet Vortex Generators

An experimental analysis was carried out to study the control effectiveness and turbulence behavior of an array of steady air-jet vortex generators on a shock-induced boundary-layer separation. Control was applied to a 24° compression-ramp interaction at Mach 2.52 and . A parametric study varying the jet spacing revealed that an adequate interaction between the vortices is essential for effective control, and the most favorable control effect was achieved with a jet spacing of . This case was analyzed in more detail using particle image velocimetry. The results reveal a strong corrugation of the separation zone due to the injection of the jets with a substantial reduction in its area along the entire spanwise extent of the air-jet array, compared with the uncontrolled case. Measurements of velocity fluctuations along the interaction region also show that the jet injection results in a reduction in turbulent intensity after reattachment, contributing to an overall reduction in amplification factor across the interaction region. The expected increase in turbulent mixing, which is essentially responsible for the control effect, was confirmed in the region in-between jet injection and separation shock. Farther away from the surface, the flow is not affected, which reduces potential drag penalties.

Low-frequency dynamics of shock-induced separation in a compression ramp interaction

The low-frequency dynamics of the shock-induced separation region in a Mach 2 compression ramp interaction is investigated by performing high-speed particle image velocimetry (HSPIV) measurements, at a rate of 6kHz, in a streamwise–spanwise plane. The HSPIV measurements made in the upstream turbulent boundary layer indicate the presence of spanwise strips of elongated regions of uniform streamwise velocity that extend to lengths greater than 30δ, validating previous results based on planar laser scattering measurements obtained by Ganapathisubramani, Clemens & Dolling (J. Fluid Mech., vol. 585, 2007, p. 369). At a wall normal-location of y/δ=0.2, a surrogate for separation based on a velocity threshold is found to fluctuate over a streamwise range of ±1.2δ, consistent with previous studies. The amplitude of unsteadiness has contributions from at least two sources that are related to the incoming boundary layer. First, the velocity threshold based surrogate separation line exhibits large-scale undulations along the spanwise direction that conform to the passage of elongated low- and high-speed regions in the upstream boundary layer. This motion is classified as the local influence of the upstream boundary layer. Second, the spanwise-averaged surrogate separation is found to respond to the overall change in streamwise velocity in the incoming boundary layer and is classified as the global influence of the upstream boundary layer. However, this global influence includes the contributions from the elongated low- and high-speed regions. Preliminary findings based on statistical analysis suggest that the local influence contributes nearly 50% more than the global influence. Regardless, the low-frequency unsteadiness of the separation-region can be attributed to the local and global influences of the incoming boundary layer.

Effects of upstream boundary layer on the unsteadiness of shock-induced separation

The relationship between the upstream boundary layer and the low-frequency, large-scale unsteadiness of the separated flow in a Mach 2 compression ramp interaction is investigated by performing wide-field particle image velocimetry (PIV) and planar laser scattering (PLS) measurements in streamwise–spanwise planes. Planar laser scattering measurements in the upstream boundary layer indicate the presence of spanwise strips of elongated regions of uniform momentum with lengths greater than 40δ. These long coherent structures have been observed in a Mach 2 supersonic boundary layer (Ganapathisubramani, Clemens & Dolling 2006) and they exhibit strong similarities to those that have been found in incompressible boundary layers (Tomkins & Adrian 2003; Ganapathisubramani, Longmire & Marusic 2003). At a wall-normal location ofy/δ=0.2, the inferred instantaneous separation line of the separation region is found to oscillate betweenx/δ=−3 and −1 (wherex/δ=0 is the ramp corner). The instantaneous spanwise separation line is found to respond to the elongated regions of uniform momentum. It is shown that high- and low-momentum regions are correlated with smaller and larger size of the separation region, respectively. Furthermore, the instantaneous separation line exhibits large-scale undulations that conform to the low- and high-speed regions in the upstream boundary layer. The low-frequency unsteadiness of the separation region/shock foot observed in numerous previous studies can be explained by a turbulent mechanism that includes these elongated regions of uniform momentum.

Relationship Between Upstream Turbulent Boundary-Layer Velocity Fluctuations and Separation Shock Unsteadiness

Particle image velocimetry and high-frequency response wall pressure measurements have been used to investigate the relationship between upstream turbulent boundary-layer properties and the unsteady separation shock behavior in a Mach 5 unswept compression ramp interaction. No correlation is found between variations in the incoming boundary-layer thickness and the separation shock foot position, as has been suggested in earlier work. However, themean velocity proe le, conditioned on theseparation shock foot position, exhibits a subtly fullershape when the shock is downstream than when it is upstream. More signie cantly, a clear correlation is observed between positivestreamwisevelocity e uctuations in thelowerthird of the upstream boundary layer and downstream shock motions, and vice versa. The strongest correlations are found for velocity e uctuations with frequencies of about4‐10 kHz, which is signie cantly lowerthan the frequencies that characterize the large-scale structures in the boundary layer (40 kHz), although spatial limitations in the transducer array may limit the instrument sensitivity to this lower range. These results are qualitatively consistent with the simple physical principle that a fuller velocity proe le imparts increased resistance to separation to the boundary layer and, hence, causes downstream shock motion, whereas a less-full velocity proe le is associated with lower resistance to separation and, hence, upstream shock motion.

High-Speed Turbulent Separated Flows: Consistency of Mathematical Models and Flow Physics

A large fraction of the experimental database used for validation of computational fluid dynamics (CFD) predictions of shock-induced turbulent separated flows consists of time-averaged measurements. If meaningful conclusions are to be drawn from comparisons of these time-averaged data with computations, it is important to understand how the data are generated physically. This can be difficult and time consuming but is an essential element in the selection of an appropriate mathematical model for the computation. For the separated compression ramp interaction, which is widely used as a CFD test case, global unsteadiness is a dominant phenomenon. Without modeling the unsteadiness, accurate predictions of the time-averaged wall pressure, outgoing velocity profiles, and other parameters will likely remain elusive, irrespective of the turbulence model.

Effects of Sweepback on Unsteady Separation in Mach 5 Compression Ramp Interactions

Fluctuating wall pressure measurements have been made upstream of the corner line in Mach 5 compression ramp interactions generated by unswept and 10-, 20-, 25-, 30-, 40-, and 50-deg swept models. The streamwise ramp angle was 28 deg in all cases. The results show the following: 1) In highly swept interactions the rms distributions of pressure fluctuations as well as the mean distributions are essentially quasiconically symmetric. The rms levels decrease globally with increasing sweep as does the maximum rms generated by the translating separation shock. 2) The length of the intermittent region, over which the separation shock foot translates, decreases with increasing sweep. In a given interaction, the length of the intermittent region grows spanwise. 3) Dominant separation shock frequencies increase from about 0.3-0.5 kHz in unswept flows to about 2-7 kHz in highly swept flows. In a given interaction, shock frequencies decrease spanwise. 4) The higher frequencies are shown to be a direct result of the decrease in the length scale of the separation shock motion.

Computation of turbulent, separated, unswept compression ramp interactions

Examination of the literature shows that the comparison between experiment and computation for highly separated unswept compression ramp flows is generally poor, irrespective of the turbulence model used. In general, the upstream influence is not correct, the wall pressure rise through separation is too steep, and the pressures under the separated shear layer are too high. In the current study, the objective is to determine if these discrepancies might be attributed more to other factors such as flowfield unsteadiness or three-dimensionality, rather than to inadequate turbulence modeling. To examine this possibility, multichannel wall pressure fluctuations were measured under the unsteady separation shock wave in a 28-deg unswept compression ramp flow at Mach 5. The results show that the large scale, low frequency separation shock unsteadiness controls the distribution of time-averaged surface properties and that neglect of the unsteadiness is probably the primary cause of the discrepancy between experiment and computation.

Correlation of separation shock motion with pressure fluctuations inthe incoming boundary layer

Fluctuating wall pressures have been measured simultaneously in the incoming undisturbed turbulent boundary layer and under the unsteady separation shock in a Mach 5 compression ramp interaction. Conditional sampling algorithms, a variable-window ensemble averaging technique, and the variable interval time averaging technique have been used to investigate the possibility of a correlation between pressure fluctuations in the incoming flow and the separation shock wave motion.

Effect of periodic blowing on attached and separated supersonic turbulent boundary layers

Periodic blowing at frequencies up to 5 kHz was used to control the unsteadiness of two-dimensional shockwave/turbulent boundary-layer interactions. Two separate experiments were performed. In the first case, periodic blowing was introduced through a spanwise slot in the wall to produce an unsteady shock-wave/boundarylayer interaction boundary layer on the tunnel wall. In the second case, periodic blowing was introduced into the shock-induced separation bubble formed by a 24-deg compression corner interaction. The incoming flow conditions for both experiments were My. = 2.84, Rejl = 6.5 x 10 7/m, and 80 = 26 mm. Measurements of the fluctuating mass flux and wall pressure were made, and the unsteady flowfield was visualized through stroboscopic schlieren videography. The measurements were conditionally sampled based on the phase of the blowing cycle. The results suggest that (at least in this case) the naturally unsteady shock motion in the compression ramp interaction does not contribute significantly to the turbulence amplification, as had been previously thought. Instead, there is strong evidence to suggest that large-scale motions associated with the maxima in the angular momentum profiles in the downstream boundary layer are responsible for the large mixing observed.

Unsteady wave structure near separation in a Mach 5 compression rampinteraction

Fluctuating wall-pressure measurements have been made under the unsteady separation shock and the separated shear layer in a Mach 5 compression ramp-induced turbulent interaction. The freestream unit Reynolds number was 49.6x10 6 m -1 and the turbulent boundary layer developed on the tunnel floor under approximately adiabatic wall-temperature conditions. Conditional sampling and «variable-window» ensemble-averaging techniques have been used to determine ensemble-averaged pressure distributions for different separation shock-wave positions.