Flow control in a transonic diffuser through mass and vorticity injection to mitigate massive separation
Authors
Gartner, Jeremy
ORCID
Loading...
Other Contributors
Amitay, Michael
Letchford, C. W.
Gandhi, Farhan
Sahni, Onkar
Vaccaro, John C.
Letchford, C. W.
Gandhi, Farhan
Sahni, Onkar
Vaccaro, John C.
Issue Date
2017-05
Keywords
Aeronautical engineering
Degree
PhD
Terms of Use
Attribution-NonCommercial-NoDerivs 3.0 United States
This electronic version is a licensed copy owned by Rensselaer Polytechnic Institute, Troy, NY. Copyright of original work retained by author.
This electronic version is a licensed copy owned by Rensselaer Polytechnic Institute, Troy, NY. Copyright of original work retained by author.
Full Citation
Abstract
An experimental flow control investigation was conducted in two test articles: an S-duct and a diffuser, both at high-subsonic conditions.
The highest pressure recovery achieved by the presence of the vortex generators was when they were erected into the flow at the height corresponding to the local boundary layer thickness (h⁄δ=1). The sweeping jets array achieved similar performance as the pulsed jets arrays. Moreover, they outperformed the 2-D jets, located at the same streamwise location, both in pressure recovery and in symmetry of the flow. The 2-D jet and the segmented jet located closer to the separation point (at x⁄L=0.25) resulted in a more efficient flow reattachment, resulting in a higher pressure recovery than for the 2-D jets located at the beginning of the ramp. Also, the effect of mass flow ratio and momentum coefficient was investigated, leading to the conclusion that both parameters are necessary to compare the effectiveness of flow control actuators. Furthermore, the 2-D jets and the segmented jet were investigated under two unsteady conditions - with a low RMS unsteady injection and a high RMS unsteady injection. Independently to the actuator, the steady jet consistently resulted in a slightly asymmetric flow field. Also, the low RMS unsteady jet performed slightly better than the steady jet, whereas the high RMS outperformed significantly the steady jet and the low RMS unsteady jet, independently to its actuation frequency. Moreover, the performances of the high RMS unsteady jet were consistently higher at an actuation frequency of 200 Hz, which corresponds to the shedding frequency of the separated flow. In addition, it was found that the 2-D jets located at the beginning of the ramp under steady and low RMS unsteady conditions, resulted in a lower pressure recovery compared to the baseline, and an asymmetric flow field. Note that these actuators were also tested at M = 0.4 resulting in a different pressure recovery but in a similar flow field as at M = 0.7.
The interaction of multiple flow control actuators with the flow field was investigated at two Mach numbers of M=0.7 and 0.4, and with and without the presence of side-walls suction. Here, the wall suction was used to delay the formation of the secondary flow structures. The flow control actuators included vortex generators, sweeping jets array, pulsed jets array, segmented jet, and 2-D steady and unsteady jets, which were placed at different streamwise locations relative to the separation point, and different throat widths in order to compare the effect of the momentum coefficient over the effect of the mass flow ratio. Note that the segmented jet and the 2-D jets were investigated under steady and unsteady injection with two different RMS values (low and high RMS).
Due to the asymmetry, only one vortex was observed near the ramp surface, and a second one was observed in the opposite corner in the bulk flow. Measurements of the low RMS jet showed a slightly better uniformity along the span compared to the steady jet, and phase-locked measurements showed that the vortices shed from the jet orifice advected by the free stream, and decayed relatively fast without enough strength to reattach the flow. The high RMS unsteady jet showed a symmetric flow field of mostly attached flow. Furthermore, the rotation of the streamwise counter-rotating vortices flipped compared to the baseline flow field. It is due to the flow being pulled upward from the floor, which leads to a rotational motion when it encounters the ramp and therefore formed these secondary structures in the opposite direction. Finally, it is hypothesized that the high-RMS unsteady jet reattached the flow through two mechanisms: (1) momentum addition near the curved surface which kept the jet attached to the ramp through the Coanda effect, which in turn “pulled” flow upward, towards the ramp surface; and (2) actuating the jet with a sinusoidal waveform at frequencies that commensurate with the naturally unstable modes (frequencies) of the separated mixing layer led to larger vortical structures that increased the mixing, and tilted the flow; thus, mitigating the separation.
As a result of the study on the compact S-duct inlet, it was decided to decouple the two mechanisms dominating the flow field – separation due to an adverse pressure gradient and secondary structures due to a radial pressure gradient. This, therefore, facilitated an investigation at a more fundamental level of the interaction of various flow control actuators with the simplified flow field.
The flow pattern in S-ducts with aggressive curvature has been shown to be, in some cases, asymmetric at the Aerodynamic Interface Plane (AIP). It was hypothesized that the interaction between two mechanisms, flow separation and the presence of secondary structures, can lead to the asymmetry. In the present work, a two-dimensional honeycomb mesh was added upstream of the curved duct to create a pressure drop across it, and therefore to an increased velocity deficit in the boundary layer. This velocity deficit led to a stronger streamwise separation, overcoming the contribution of the secondary structures and eliminated the flow field asymmetry. The experiments were performed at Mach numbers of M = 0.2, 0.44 and 0.58. Steady and unsteady surface and AIP pressure measurements, together with Particle Image Velocimetry (PIV), were used to explore the effect of inserting a honeycomb into the flow field by increasing its height from 0 to 2.2 times the local boundary layer thickness. Using the honeycomb, flow symmetry was achieved for the specific geometrical configuration tested with a negligible decrease of the pressure recovery.
Next, it was decided to take this study a step further, by enabling the secondary structures to develop. This was achieved by removing the side-walls suction, and conducting detailed stereoscopic particle image velocimetry (SPIV) experiments on selected actuators at M = 0.4. For this purpose, the 2-D Jet, located at the beginning of the ramp, was chosen since its efficacy was either detrimental or constructive, depending on the methods of its operation (i.e., steady jet, low RMS unsteady jet or high RMS unsteady jet). These experiments showed that without flow control (baseline case), the flow field exhibited a slight asymmetry where there was an initial spanwise velocity component with a magnitude on each side of ~2% of the freestream velocity, which could be due to secondary structures caused by the corners in a turbulent flow. Moreover, two counter-rotating streamwise vortices were developed in the separated region due to the radial pressure gradient. The steady jet results showed that one side near the ceiling stayed attached, leading the flow to separate at midspan and develop asymmetry. It is assumed that the local flow reattachment is due to a combination of the spanwise velocity as shown for the baseline, and a slightly non-uniform spanwise distribution of the steady jet at its exit plane. Therefore, it causes a surplus of momentum in the upper corner leading to an entrainment of the flow field toward this direction.
The highest pressure recovery achieved by the presence of the vortex generators was when they were erected into the flow at the height corresponding to the local boundary layer thickness (h⁄δ=1). The sweeping jets array achieved similar performance as the pulsed jets arrays. Moreover, they outperformed the 2-D jets, located at the same streamwise location, both in pressure recovery and in symmetry of the flow. The 2-D jet and the segmented jet located closer to the separation point (at x⁄L=0.25) resulted in a more efficient flow reattachment, resulting in a higher pressure recovery than for the 2-D jets located at the beginning of the ramp. Also, the effect of mass flow ratio and momentum coefficient was investigated, leading to the conclusion that both parameters are necessary to compare the effectiveness of flow control actuators. Furthermore, the 2-D jets and the segmented jet were investigated under two unsteady conditions - with a low RMS unsteady injection and a high RMS unsteady injection. Independently to the actuator, the steady jet consistently resulted in a slightly asymmetric flow field. Also, the low RMS unsteady jet performed slightly better than the steady jet, whereas the high RMS outperformed significantly the steady jet and the low RMS unsteady jet, independently to its actuation frequency. Moreover, the performances of the high RMS unsteady jet were consistently higher at an actuation frequency of 200 Hz, which corresponds to the shedding frequency of the separated flow. In addition, it was found that the 2-D jets located at the beginning of the ramp under steady and low RMS unsteady conditions, resulted in a lower pressure recovery compared to the baseline, and an asymmetric flow field. Note that these actuators were also tested at M = 0.4 resulting in a different pressure recovery but in a similar flow field as at M = 0.7.
The interaction of multiple flow control actuators with the flow field was investigated at two Mach numbers of M=0.7 and 0.4, and with and without the presence of side-walls suction. Here, the wall suction was used to delay the formation of the secondary flow structures. The flow control actuators included vortex generators, sweeping jets array, pulsed jets array, segmented jet, and 2-D steady and unsteady jets, which were placed at different streamwise locations relative to the separation point, and different throat widths in order to compare the effect of the momentum coefficient over the effect of the mass flow ratio. Note that the segmented jet and the 2-D jets were investigated under steady and unsteady injection with two different RMS values (low and high RMS).
Due to the asymmetry, only one vortex was observed near the ramp surface, and a second one was observed in the opposite corner in the bulk flow. Measurements of the low RMS jet showed a slightly better uniformity along the span compared to the steady jet, and phase-locked measurements showed that the vortices shed from the jet orifice advected by the free stream, and decayed relatively fast without enough strength to reattach the flow. The high RMS unsteady jet showed a symmetric flow field of mostly attached flow. Furthermore, the rotation of the streamwise counter-rotating vortices flipped compared to the baseline flow field. It is due to the flow being pulled upward from the floor, which leads to a rotational motion when it encounters the ramp and therefore formed these secondary structures in the opposite direction. Finally, it is hypothesized that the high-RMS unsteady jet reattached the flow through two mechanisms: (1) momentum addition near the curved surface which kept the jet attached to the ramp through the Coanda effect, which in turn “pulled” flow upward, towards the ramp surface; and (2) actuating the jet with a sinusoidal waveform at frequencies that commensurate with the naturally unstable modes (frequencies) of the separated mixing layer led to larger vortical structures that increased the mixing, and tilted the flow; thus, mitigating the separation.
As a result of the study on the compact S-duct inlet, it was decided to decouple the two mechanisms dominating the flow field – separation due to an adverse pressure gradient and secondary structures due to a radial pressure gradient. This, therefore, facilitated an investigation at a more fundamental level of the interaction of various flow control actuators with the simplified flow field.
The flow pattern in S-ducts with aggressive curvature has been shown to be, in some cases, asymmetric at the Aerodynamic Interface Plane (AIP). It was hypothesized that the interaction between two mechanisms, flow separation and the presence of secondary structures, can lead to the asymmetry. In the present work, a two-dimensional honeycomb mesh was added upstream of the curved duct to create a pressure drop across it, and therefore to an increased velocity deficit in the boundary layer. This velocity deficit led to a stronger streamwise separation, overcoming the contribution of the secondary structures and eliminated the flow field asymmetry. The experiments were performed at Mach numbers of M = 0.2, 0.44 and 0.58. Steady and unsteady surface and AIP pressure measurements, together with Particle Image Velocimetry (PIV), were used to explore the effect of inserting a honeycomb into the flow field by increasing its height from 0 to 2.2 times the local boundary layer thickness. Using the honeycomb, flow symmetry was achieved for the specific geometrical configuration tested with a negligible decrease of the pressure recovery.
Next, it was decided to take this study a step further, by enabling the secondary structures to develop. This was achieved by removing the side-walls suction, and conducting detailed stereoscopic particle image velocimetry (SPIV) experiments on selected actuators at M = 0.4. For this purpose, the 2-D Jet, located at the beginning of the ramp, was chosen since its efficacy was either detrimental or constructive, depending on the methods of its operation (i.e., steady jet, low RMS unsteady jet or high RMS unsteady jet). These experiments showed that without flow control (baseline case), the flow field exhibited a slight asymmetry where there was an initial spanwise velocity component with a magnitude on each side of ~2% of the freestream velocity, which could be due to secondary structures caused by the corners in a turbulent flow. Moreover, two counter-rotating streamwise vortices were developed in the separated region due to the radial pressure gradient. The steady jet results showed that one side near the ceiling stayed attached, leading the flow to separate at midspan and develop asymmetry. It is assumed that the local flow reattachment is due to a combination of the spanwise velocity as shown for the baseline, and a slightly non-uniform spanwise distribution of the steady jet at its exit plane. Therefore, it causes a surplus of momentum in the upper corner leading to an entrainment of the flow field toward this direction.
Description
May 2017
School of Engineering
School of Engineering
Department
Dept. of Mechanical, Aerospace, and Nuclear Engineering
Publisher
Rensselaer Polytechnic Institute, Troy, NY
Relationships
Rensselaer Theses and Dissertations Online Collection
Access
CC BY-NC-ND. Users may download and share copies with attribution in accordance with a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 License. No commercial use or derivatives are permitted without the explicit approval of the author.