Shock-Wave/Boundary-Layer Interaction (SBLI) poses inherent challenges for supersonic/hypersonic aircraft, including flow separation, unsteady pressure loading, vibration, and engine unstart. A symmetric Double-Fin is a canonical configuration often used in high-speed aircraft components and subsystems, such as vertical tails and ramjet/scramjet inlets. Experiments to investigate SBLI induced by Double-Fin with fin angles of 10 degrees, 12 degrees, and 15 degrees, resulting in increasingly stronger interactions, are conducted at Mach 2. Using Scanning-Stereoscopic Particle Image Velocimetry (S-SPIV), three-dimensional, three-component velocity fields are acquired, facilitating a deeper exploration of highly three-dimensional multi-shock interactions. The evolution of this complex flow, such as the interaction of lambda-shocks generated by the fins, is comprehensively examined for improved understanding of the governing flow physics. In particular, the interaction between the front-foot of the reflected-shock and the fin-shock gives rise to a distinct, stronger shock referred to as Shock-Tangle, accompanied by a localized low-velocity region. Three-dimensional slip surfaces are identified at shock-shock intersections, and a streamwise counter-rotating vortex pair forms in the near-wall region downstream of the interaction. Through S-SPIV, key flow features are resolved in unprecedented detail, enabling quantitative comparison of the Double-Fin and Single-Fin flow topologies and revealing differences in quasi-conical symmetry and shock organization. The findings advance the fundamental understanding of fin-generated SBLI, with practical implications for current and next-generation high-speed vehicles.

Large-Eddy Simulations are used to study different active and passive control of a high-speed turbulent multi-stream shear flow. The primary goal is to mitigate the prominent tone in the baseline shear layer formed by the merging of Mach 1.23 and Mach 1 streams downstream of a thick splitter plate. The tone propagates into the rest of the domain, and is associated with shedding and Kelvin-Helmholtz instabilities. Vortices shed from the splitter plate trailing edge (SPTE) also perturb the oblique shock emanating from the shear layer. Active control is applied on the splitter plate trailing edge either by time-periodic blowing or by drawing a constant mass flux of ambient air into the actuator ducts, while passive control is examined with blind ducts. The mean and instantaneous features are analyzed followed by the spectral analysis of fluctuating pressure. Results show that the best results are obtained with constant blowing active control, either with multiple ducts or a single slot. The shear layer then displays finer scale downstream turbulence, with significant weakening of the tone as well as the oblique shock strength and unsteadiness. Periodic blowing is not effective; however, the peak mass flux examined was much higher than for constant blowing suggesting the need for a parametric study for a clearer conclusion. Although blind ducts displayed pressure waves moving back and forth in the duct due to downstream signals from the shedding process, the overall effect on the flow was minimal.

This work presents a novel approach for pressure field reconstruction from image velocimetry data using SIREN (Sinusoidal Representation Network), emphasizing its effectiveness as an implicit neural representation in noisy environments and its mesh-free nature. While we briefly assess two recently proposed methods—one-shot matrix-omnidirectional integration (OS-MODI) and Green’s function integral (GFI)—the primary focus is on the advantages of the SIREN approach. The OS-MODI technique performs well in noise-free conditions and with structured meshes but struggles when applied to unstructured meshes with high aspect ratio. Similarly, the GFI method encounters difficulties due to singularities inherent from the Newtonian kernel. In contrast, the proposed SIREN approach is a mesh-free method that directly reconstructs the pressure field, bypassing the need for an intrinsic grid connectivity and, hence, avoiding the challenges associated with ill-conditioned cells and unstructured meshes. This provides a distinct advantage over traditional mesh-based methods. Moreover, it is shown that changes in the architecture of the SIREN can be used to filter out inherent noise from velocimetry data. This work positions SIREN as a robust and versatile solution for pressure reconstruction, particularly in noisy environments characterized by the absence of mesh structure, opening new avenues for innovative applications in this field.

Active and passive control methods of a modern complex rectangular three-stream nozzle are experimentally analyzed and compared to its baseline condition. A main flow (Mach = 1.6) and a bypass flow (Mach = 1.0) are separated by a splitter plate until just downstream of the nozzle throat where they are allowed to mix. This mixing creates a Kelvin-Helmholtz like instability at the splitter plate trailing edge, which could negatively impact the performance of the system. A steady blowing jet array embedded in the splitter plate trailing edge, the active control method, is actuated parallel to the main and bypass flows. A similar passive control method using a span-wise slit on the top of the splitter plate to allow bleed from the main flow to interact with the splitter plate trailing edge, the same region as the steady actuation of the active control method. These control mechanisms aim to induce mixing between the two flows, diminishing the instability present in the splitter plate trailing edge region. Far-field acoustic measurements and schlieren imaging are used to define and investigate the baseline and controlled cases.

This work presents a novel approach for pressure field reconstruction from image velocimetry data using SIREN (Sinusoidal Representation Network), emphasizing its effectiveness as an implicit neural representation in noisy environments and its mesh-free nature. While we briefly assess two recently proposed methods - one-shot matrix-omnidirectional integration (OS-MODI) and Green's function integral (GFI) - the primary focus is on the advantages of the SIREN approach. The OS-MODI technique performs well in noise-free conditions and with structured meshes but struggles when applied to unstructured meshes with high aspect ratio. Similarly, the GFI method encounters difficulties due to singularities inherent from the Newtonian kernel. In contrast, the proposed SIREN approach is a mesh-free method that directly reconstructs the pressure field, bypassing the need for an intrinsic grid connectivity and, hence, avoiding the challenges associated with ill-conditioned cells and unstructured meshes. This provides a distinct advantage over traditional mesh-based methods. Moreover, it is shown that changes in the architecture of the SIREN can be used to filter out inherent noise from velocimetry data. This work positions SIREN as a robust and versatile solution for pressure reconstruction, particularly in noisy environments characterized by the absence of mesh structure, opening new avenues for innovative applications in this field.

In this article, a method for density tomography using a telecentric BOS setup coupled with a motorized, rotating nozzle will be described. The telecentric optics allow for the disentanglement of the ray path integral and density gradient operators through Leibniz’s integral rule because the rays seen by a telecentric setup are parallel in the Schlieren domain. This enables the usage of a fast gradient inversion solver based on the rotating parallel ray integral method, previously developed by the author, to solve for the ray-integrated density fields in two dimensions for each view individually, producing a set of ray-integrated density measurements similar to a set of images captured by a regular absorption or emission-based computed tomography setup. By leveraging well-known computed tomography algorithms, such as the inverse Radon transform, a tomographic reconstruction of the average density field is quickly found only requiring a simple 2D camera calibration. A demonstration experiment setup is used to test the technique, where a low-speed hot jet produced by different nozzle shapes is measured. The 3D density fields are then converted to temperature fields and the temperatures are compared to a reference measurement with a thermocouple scanner. The strong agreement between the two measurements creates confidence that this technique can be deployed to produce quantitative measurements of the density fields in complex flows, which can be combined with 3D PIV measurements to obtain the four time-averaged thermodynamic fields in compressible jet flows.

The effects of edge geometry and compressibility on the wake of a slanted afterbody model with rounded edges, relevant to cargo aircraft and high-speed train applications, are investigated through detailed experimental observations of the flowfield. The wake flow features are examined using centerline planar PIV measurements and novel Scanning-SPIV measurements to reconstruct the full mean volumetric velocity field. Planar PIV measurements at the model centerline reveal that increasing the Mach number reduces the shear layer growth rate, leading to decreased entrainment within the recirculation region. Consequently, the recirculation region increases in both length and height. Further downstream, the vortex circulation for the rounded-edge model remains nearly constant across both incompressible and compressible Mach numbers. Additional vortex properties are examined through the Reynolds-averaged vorticity transport equation applied to the volumetric flowfield measurements, revealing an increase in x-vorticity compression within the recirculation region. By Helmholtz’s vortex theorem, this increased vortex compression contributes to the growth of the recirculation region between Mach number conditions. Additionally, the dilatation term was explored, allowing for the delineation of compressibility effects on the recirculation region.

This study investigates shock wave-boundary layer interactions (SBLI) on both convex and concave surfaces of supersonic turbine blades using high-speed (50,000 Hz) shadowgraph imaging technique. The unsteadiness of separation bubbles and the resulting shock structures are examined and compared with both flat plate models and large eddy simulations. The results highlight the significant influence of surface curvature on SBLI behavior. On the convex surface, the impinging shock generates a large separated SBLI region, whereas a much smaller interaction forms on the concave surface. A bistable behavior is observed on the convex side, where the boundary layer alternates between an attached and a detached state. To isolate the role of curvature, a single-blade configuration was also tested, confirming that the impinging shock induces separation. In this configuration, the separation shock exhibited low-frequency bi-stability, intermittently jumping between two distinct locations.

In this work, we explore the possibility of reducing the number of cameras for tomographic particle tracking velocimetry (PTV) by using beamsplitters to project multiple viewpoints on a single camera sensor. This optical train allows a trade-off between particle seeding density and number of cameras, which enables the reduction of camera cost at the highest end of high-speed tomographic PTV applications where camera cost is considerable. Results are presented with synthetically-generated images for unmodified versions of state-of-the-art PTV algorithms (OpenLPT and DaVis 10.1), and a modification is implemented to the OpenLPT algorithm to include the knowledge that the views are multiplied, enabling the removal of more particles at every iteration of the iterative particle reconstruction (IPR) step. We observe that incorporating the knowledge of the image-multiplied projection brings the performance of the OpenLPT PTV algorithm very close to the baseline performance for equivalent particle seeding density values. However, when using real doubled experimental images we note a major challenge when performing volume self-calibration. The volume self-calibration process is significantly affected by the image doubling, and the knowledge of image doubling has to also be incorporated into the volume self-calibration process, which is left as future work.

The unsteady pressure dynamics in the wake of cargo aircraft is examined through the use of a simplified bluff body known as the 'slanted afterbody'. The current knowledge of the unsteady wake dynamics behind the slanted afterbody is expanded and further related to cargo aircraft through the addition of a rounded-edge afterbody at compressible flight conditions. The flowfield is examined over two Mach numbers, M-infinity = 0.3 and M-infinity = 0.6, and two Reynolds numbers, Re-D = 0.83E6 and Re-D = 1.47E6, to identify the independent effects of compressibility and Reynolds number. Unsteady pressure measurements are performed through the use of unsteady pressure transducers and unsteady pressure sensitive paint. Additionally, high-speed shadowgraph is used to examine the unsteady wake flow features and their evolution as they propagate downstream. The unsteady behavior is elucidated through various methods including modal analysis, to understand the effect of edge condition and slant angle on the unsteady dynamics. Principal among the flow features of the centerline separated state are downstream-convecting pressure waves at a frequency that varies little with respect to flow conditions or slant angle. However, a decrease in overall pressure fluctuations is seen in the rounded-edge geometry relative to the sharp-edge baseline. These waves are conjectured to be related to the vortex wandering modes witnessed on the sharp-edge model. In addition, the recently identified centerline attached vortex state displays unsteady phenomena that are significantly damped in magnitude and absent from spectral and modal analyses. The absence of these unsteady features stresses the importance of further flow analysis on this wake state.

In this work, we present a method to perform 2D and 3D omnidirectional pressure integration from velocity measurements with a single-iteration matrix inversion approach. This work builds upon our previous work, where the rotating parallel ray approach was extended to the limit of infinite rays by taking continuous projection integrals of the ray paths and recasting the problem as an iterative matrix inversion problem. This iterative matrix equation is now ``fast-forwarded'' to the ``infinity'' iteration, leading to a different matrix equation that can be solved in a single iteration, thereby presenting the same computational complexity as the Poisson equation. We observe computational speedups of $\sim10^6$ when compared to brute-force omnidirectional integration methods, enabling the treatment of grids of $\sim 10^9$ points and potentially even larger in a desktop setup at the time of publication. Further examination of the boundary conditions of our one-shot method shows that omnidirectional pressure integration implements a new type of boundary condition, which treats the boundary points as interior points to the extent that information is available.

We perform an assessment of pressure field reconstruction techniques for image velocimetry data. Three techniques are examined: a matrix omnidirectional integration method (OS-MODI), a Green's function integral approach (GFI) and a novel neural-network approach (NN). While the OS-MODI excels in noise-free conditions and structured meshes, the NN method, proposed in this study, offers superior performance in noisy environments, besides being mesh-free. For unstructured mesh analyses, challenges with ill-conditioned cells prompt sophisticated mesh generation strategies. Moreover, a simplistic application of the constant midpoint rule to compute the integrals of the boundary elements within the GFI method leads to issues with singularity. Despite obstacles, the OS-MODI and GFI show promise, with NN emerging as a robust solution for meshless noisy data.

Experimentally-measured pressure fields play an important role in understanding many fluid dynamics problems. Unfortunately, pressure fields are difficult to measure directly with non-invasive, spatially resolved diagnostics, and calculations of pressure from velocity have proven sensitive to error in the data. Omnidirectional line integration methods are usually more accurate and robust to these effects as compared to implicit Poisson equations, but have seen slower uptake due to the higher computational and memory costs, particularly in 3D domains. This paper demonstrates how omnidirectional line integration approaches can be converted to a matrix inversion problem. This novel formulation uses an iterative approach so that the boundary conditions are updated each step, preserving the convergence behavior of omnidirectional schemes while also keeping the computational efficiency of Poisson solvers. This method is implemented in Matlab and also as a GPU-accelerated code in CUDA-C++. The behavior of the new method is demonstrated on 2D and 3D synthetic and experimental data. Three-dimensional grid sizes of up to 125 million grid points are tractable with this method, opening exciting opportunities to perform volumetric pressure field estimation from 3D PIV measurements.

In this study, the slanted afterbody was modified to more closely represent cargo aircraft afterbodies through the addition of a basal edge fillet. Two slant angles ([Formula: see text] and [Formula: see text]) were explored over two Mach numbers ([Formula: see text] and [Formula: see text]) and two Reynolds numbers ([Formula: see text] and [Formula: see text]) using oil flow visualization and pressure-sensitive paint to identify surface flow phenomena, while the spanwise shadowgraph enabled the observation of mean wake phenomena. The [Formula: see text] rounded-edge afterbody surface measurements displayed a centerline-separated vortex state with flow features and regions of suction qualitatively similar to those of the baseline sharp-edge case, with some notable differences. For example, a separation bubble at the slant leading edge is nearly twice the size of the baseline sharp-edge model. This is accompanied by the presence of a “source point,” indicated by a region of high pressure where the flow reattaches, and a tighter counter-rotating vortex pair in comparison to the baseline. The rounded-edge prevented the previously seen compressible transition to a fully separated wake state within the Mach number range studied while showing an increase in separation bubble size with increasing Mach number. Also noteworthy was the existence of a new vortex state for the [Formula: see text] model, where the flow remains attached at the model centerline. This is in contrast to the previously observed flow topology of the centerline-separated vortex state in the sharp-edge model, which formed a separation bubble at the model centerline. The transition between the new centerline-attached vortex state and the traditional centerline-separated vortex state was shown to be dependent on both Mach number and Reynolds number, revealing the importance of boundary-layer development on the rounded afterbody flowfield. Overall, these observations indicate that the rounded afterbody produces flowfields with distinct features that vary from the previously described sharp-edge afterbody wake field under certain conditions.

The highly unsteady flowfield produced by impinging jets can result in loud noise and potentially harmful structural vibration in Short Take-Off and Vertical Landing applications. In the present study, a microjet-based control method is implemented on a Mach 1.5 supersonic nozzle to attenuate jet shear layer unsteadiness. Two microjet arrays are installed near the nozzle exit and the diverging section of the supersonic nozzle. Each microjet array consists of sixteen individually addressable microjets controlled by solenoid valves. Two electronic pressure regulators are used to adjust the supply pressure for each microjet array. A genetic algorithm is applied to optimize the microjet configuration to minimize overall sound pressure level. An optimization using a genetic algorithm is performed at a selected impingement distance (H/D = 3.75). The genetic algorithm solution reduced OASPL by 10 dB compared to the baseline condition, which is about 2 dB better than the microjet configuration where all microjets are fired with maximum pressure — this further noise reduction is achieved while reducing 26% of microjet mass flow rate. The optimal GA solution is applied to a range of impingement distances (0.75D – 10D), and it is revealed that this GA solution effectively reduces noise for the majority of impingement distances.

In this work, the slanted afterbody has been modified to more closely represent the intended application of cargo aircraft afterbodies through the addition of a basal edge fillet. Two slant angles (phi = 32 and phi = 45) were explored over two Mach numbers (M = 0.3 and M = 0.6) and two Reynolds numbers (ReD = 0.832E6 and ReD = 1.47E6) utilizing oil-flow visualization and pressure sensitive paint to identify surface flow phenomena while spanwise shadowgraph enabled the observation of unsteady wake phenomena. Mean surface measurements of the phi = 45 rounded-edge slanted afterbody vortex state displayed similar regions of suction to the sharp-edged case, corresponding to a large separation bubble at the slant leading-edge, and a point of higher pressure where the flow reattaches. The addition of a rounded-edge prevented the compressible transition to the wake state within the Mach number range studied while still showing an increase in separation bubble with increasing Mach number. Examination of the phi = 32 model revealed the existence of an additional vortex state flow topology with attached flow at the model centerline. This in contrast with the previously described flow topology of the vortex state in the sharp-edged model which includes a separation bubble at the model centerline. The dependency of this state on both Mach number and Reynolds number reveals the importance of boundary layer development within the rounded afterbody flowfield. Overall, these observations indicate the rounded afterbody produces varying effects within the previously described sharp-edge rounded afterbody wake field that are necessary to be studied further.

The performance of a co-flow jet airfoil, consisting of a modified NACA 6421 profile, is examined through a set of RANS numerical simulations and select fan/duct characterization experiments under various conditions to assess whether net power savings at cruise conditions is possible with this implementation. A realistic duct, transitioning from a circular cross-section to a rectangular cross-section, is designed and fully defined for both the straight, benchmark case, and a curved (CFJ) case. The system design then accounts for the full interaction between the external flow over the airfoil and the internal flow within the duct. It was found that a fan with an overall efficiency in excess of ~75% is required for a net-positive power savings during cruise.

Shock-Wave/Boundary-Layer Interaction (SBLI) is inevitable on supersonic and hypersonic air vehicles leading to undesired internal and external aerodynamic challenges such as flow separation, unsteady pressure loading, vibration, and unstart. In the present study, fully volumetric (3-dimensions, 3-components) mean flowfield of a symmetric Double-Fin (DF) SBLI with 10-degree fin angles at Mach 2 is revealed. Scanning Stereoscopic Particle Image Velocimetry (S-SPIV) technique is implemented to acquire snapshots of the instantaneous velocity field perpendicular to the incoming freestream. In the mean time, a traverse mechanism rigidly holds the laser, cameras, and optical elements together, and continuously scans along the streamwise direction. The scanning region covers the entire volume between the fins, and the acquired snapshots are utilized to construct a three-dimensional velocity field through spatio-temporal averaging. This approach reveals intricate details of the flow structures over a wide region of interest, including the development and interaction of oblique and -shocks generated by the sharp-fins. It is found that the interaction of the shocks leads to the creation of a Shock-Triangle and the subsequent formation of slip-lines. The study includes a detailed discussion of the topology of these flow structures.

In this study, a variation of the stacked stereoscopic PIV technique is proposed to perform fully volumetric (3-dimensions, 3-components) measurements of average flow fields within a single experiment through the usage of an automated traversing system that continuously scans the SPIV light sheet over a linear path. The simultaneous measurement of the traverse location and the laser Q-switch pulse enables the automated assignment of instantaneous PIV fields to known physical coordinates, enabling spatiotemporal averaging in post-processing to obtain volumetric measurements of a flow field. This method provides a trade-off between spatial resolution of the volume measurements and statistical convergence of the spatiotemporal averages, enabling volumetric measurements under challenging experimental conditions where only stereoscopic PIV is viable. A comparison with the more traditional temporal averaging method and planar PIV is presented to demonstrate the capabilities and limitations of this technique in realistic, challenging experimental setups. It is found that the spatiotemporal averaging convergence behavior differs slightly from the traditional temporal averaging for the wake of a bluff body model, however relative errors lower than two standard deviations can still be attained. Thus, this technique presents a viable alternative for rapid 3D reconstruction of averaged flow fields that can provide invaluable insight of various flow topologies.

An experimental assessment of thrust vectoring of an overexpanded, circular supersonic jet is performed by using an array of 64 individually addressable microjet-in-crossflow actuators located at four distinct streamwise locations. A genetic algorithm, using the pressure distribution measured by a rake of total pressure probes in front of the jet and having the goal of maximizing the thrust vectoring angle, achieved a maximum vectoring angle of 2 deg after exploring 900 distinct actuator configurations. The solution is also demonstrated to provide azimuthal control of the vectoring direction due to the circular cross section of the nozzle, which is useful for rocket control systems. Further analysis of the optimal solution indicates the genetic algorithm leveraged the production of an asymmetric shock structure at the diverging section of the nozzle to achieve shock vector control.