Final Report Summary - MARS (Manipulation of Reynolds Stress for Separation Control and Drag Reduction)
Executive Summary:
The MARS project has been developed under the umbrella of the 7th Research Framework programme of the European Commission (EC) and the International Science & Technology Cooperation Program of Civil Aviation, sponsored by Ministry of Industry and Information Technology of the People's Republic of China (MIIT).
This international collaborative effort was a successful outcome of the networking activities within the AEROCHINA2 project. MARS (Manipulation of Reynolds Stress for Separation Control and Drag Reduction) is one of the research projects generated as an outcome of the networking activities.
MARS is devoted to the analysis and understanding of the phenomena involved on the birth of the aerodynamic drag. Drag reduction and flow separation control are the high-level aims, while the project focus on the turbulence phenomena, and effect of Reynolds stress on the drag force.
The project was defined with the agreement of the European industry; namely Airbus, Alenia and Dassault Aviation, who considered the need to fully understand the initiation of the drag force in order to be able to deal with methods and techniques to reduce the skin friction on the aircraft.
The consortium, led by CAE (Chinese Aeronautical Establishment, China) in the Chinese side and CIMNE (International Centre for Numerical Methods in Engineering, Spain) in the European side, is compound by 10 Chinese partners and 12 European partners. The technical coordination is responsibility of ARI (Aeronautical Research Institute, China) and USFD (University of Sheffield, UK).
The objectives of the project can be summarized as:
1. To control turbulent wall bounded shear flow effectively from a more fundamental level by investigating directly the behaviour of Reynolds stresses and their response to the manipulation of the dynamic components of the flow
2. To conduct both experimental tests and computational simulations to extract reliable flow physics, including Reynolds stresses, for a number of active flow control devices
3. To explore large scale unsteadiness (unsteady jets, wakes and vortices) produced from the capability of these devices for to provide effective control of the dynamic components, and hence the Reynolds stresses, within the shear layers associated with large scale wakes and separations
4. To identify key strategies that enable efficient control of dynamic fluid structures within shear layers and to design and optimise these devices for separation control (higher shear) and drag reduction (lower shear)
5. To demonstrate the most promising device/devices at relevant scale, the ability to efficiently increase or decrease the Reynolds stresses
6. To investigate the application of this capability within an aircraft context
7. To foster further collaboration between European and Chinese researchers in the key technology area of flow control for the benefit of the civil aircraft industries on both sides.
The project has been planned for 3 years, with a final effective duration of 3.5 years.
Project Context and Objectives:
Drag reduction and separation control are directly related to more efficient air transportation and less emission of harmful gases into the environment. While the aerospace industry is striving to have more and more optimised designs, it is still some way away from the targets set out in the ACARE 2020 vision for 50% reduction in aircraft emissions. Separation control and drag reduction contribute directly towards this target and active flow control could plays an important role in achieving it. After many decades of development, the highly optimised aircraft designs make further large improvements difficult without a ‘game-changing’ technology such as active flow control. Active flow control provides an additional dimension for further improving aircraft performance, in particular, for performance at different operational points, e.g. at both cruise and take-off/landing conditions.
The importance of reducing skin friction on meeting the global fuel burn targets is obvious. At cruise approximately one half of the total drag of a modern commercial transport aircraft is attributed to skin friction. The importance of improved separation control is less obvious unless we acknowledge the influence of aircraft mass on fuel burn. In fact the sensitivity of fuel burn to mass is greater than that to skin friction. Therefore if the structural mass of the aircraft can be reduced by more efficient low speed configurations or improved load alleviation off-design then this carries a significant direct benefit on cruise fuel burn.
The turbulence Reynolds stress is the most important dynamic quantity affecting the mean flow as it is responsible for a major part of the momentum transfer in the wall bounded turbulent flow. It has a direct relevance to both skin friction (for a turbulent boundary layer) and flow separation (occurs when skin friction drops to zero). The near wall region for a turbulent boundary layer can be divided into the viscous sub-layer, where the mean viscous stress is important and the approximately constant Reynolds stress region, where the viscous stress drops to zero and the Reynolds stress peaks. As the Reynolds number increases, the peak Reynolds stress approaches the value of the viscous stress at the wall. Therefore active manipulation of the Reynolds stress can directly lead to changes in the viscous stress at the wall so as to effectively control the flow for effective flow control.
However, there is a lack of current understanding of the inter-relationship between the various flow control devices and the Reynolds stresses in the flow field they produced. An improved understanding can potentially significantly improve the effectiveness of flow control as the Reynolds stresses are closely related to the flow behaviour at the surface for effective separation control or drag reduction. A variety of control devices are available and new ones are invented but which one for what purpose is an open question yet to be fully answered.
The majority of prior work has focused on the introduction of changes to the mean flow that resulted in changes to the Reynolds stress. In the present proposal we propose to reverse that process and consider the long term goal of controlling dynamic structures that then influence the Reynolds stress that in turn changes the mean flow. This radical approach recognises that we are still some way away from hardware to implement the concept at flight scales but if successful, would establish a first important step towards our ultimate ambition.
The focus of the present project will be on the effects of a number of active flow control devices on the discrete dynamic components of the turbulent shear layers and the Reynolds stress. From the application point of view, the current proposal provides a positive and necessary step in the right direction wherein it will demonstrate the capability to control individual structures that are larger in scale and lower in frequency compared to the richness of the time and spatial scales in a turbulent boundary layer. In order to focus the current project with the limited duration and budget, the current project excludes laminar flow control for transition delay, another area of extensive previous and current research. Furthermore, the project will put more effort on active flow control means rather than passive controls.
To explain our basic strategy in manipulating Reynolds stresses through the dynamic components of the turbulent shear layers, it is helpful to start with the triple decomposition proposed by Reynolds and Hussain (1972) for an instantaneous velocity, U
U = Ū + ũ + uˈ
The first term on the RHS is the time averaged mean velocity. If we attempt to control this via flow control then most devices offer little gain in efficiency on a global energy basis i.e. change in energy out equals energy in.
The second term on the RHS is the periodic/dynamic component of the flow and for some specific flow scenarios this can be shown to be dominant in determining the flow state and characteristics. The stresses produced from this term are referred as the periodic stresses or “apparent stresses”. It offers some interesting opportunities for demonstrating the way in which to deploy flow control technologies for dynamic environments (responsive environments, smart inputs and sensible control). This also implies that, for statistically steady flows, where the second term disappears, artificial introduction of the periodic term may be necessary for effective control.
The final term on the RHS represents the broadband ‘random’ turbulent fluctuations, from which the Reynolds stresses are defined. While direct control of the “random” components is the ultimate goal, the current project aims to investigate the control of the periodic stresses, the dynamic components of the flow, in order to manipulate the Reynolds stress for the benefit of flow control.
With the basic strategy for MARS established the global aims of the project can be defined. They are:
• To use the periodic flows embedded in the two identified flow cases as platforms in which direct control of discrete dynamic structures to manipulate the Reynolds stress can be observed, measured and simulated.
• To measure, simulate and understand the impact of certain actuators upon discrete structures in a turbulent shear layer and to identify candidate actuators for further development for skin friction reduction and flow separation control at flight scales.
An important characteristic of this proposal is that in MARS we wish to explore the possibility of influencing the mean flow via the Reynolds stress by direct manipulation of discrete structures that contribute to the stress. This is in contrast to most previous work in which the actuators change the mean flow directly. Therefore, the MARS approach offers the potential for higher system gain.
The proposed project is based on the mutual understanding of the EU and Chinese partners, mostly from the AeroChina and AeroChinaII consortia funded by the EU. In AeroChinaII, a Working Group on Flow Control was established due to a strong interest in the area from both sides. A series of EU-China Flow Control workshops were held, which hugely improved our understanding of the capabilities in both experiments and simulations and pacing problems in industrial applications. While the MARS consortium is based on the key members of this working group, some new EU and Chinese partners were invited in the current project in order to further strengthen the current consortium.
Project Results:
MARS project was organized in 5 work packages; 3 of them including technical tasks. The analysis of the flow control devices, and their capacity to control and modify the Reynolds stress was done mainly in WP2 and WP3. WP2 grouped the experimental tasks, WP3 grouped the numerical simulation tasks, and both of them fed the WP4, where the down-selection of the best performing devices and the optimization analysis of the selected ones were planned.
The project defined a set of test cases which have been analysed experimentally and numerically. The list of test cases is:
- Profile NACA 0015
o Without control device
o With Plasma actuators
o With Synthetic Jets actuators
o With Micro-Blowing/Suction actuators
o With Pulsed Jets actuators
- Backward Facing Step
o Without control device
o With Plasma actuators
o With Oscillating surface actuators
o With Micro-Blowing/Suction actuators
o With Synthetic Jets actuators
o With Spanwise Vortex Generators
Regarding the experimental test cases, the primary objective of these experiments was to explore the possibility of influencing the mean flow downstream a backward-facing step or in the wake of a NACA0015 model by using a variety of actuators. The present investigations demonstrate the influence of these devices on the development of the free shear layer. Some of these actuators can provide a high mass injection or velocity ratio (synthetic jets, pulsed jets,…) when others can only promotes small amplitude perturbations in the natural flow (plasma actuators, piezoelectric actuators,…). For the latter actuators, the flow dynamic cannot be drastically changed without fine control strategies as developed in this project.
Regarding the proposed experimental researches as defined in the DoW, most of the proposal content has been conducted. Development of innovative actuators has been conducted, the dynamical behaviour of the separated boundary layer has been analysed by mean and time-resolved measurements, the wall signature of periodic structures existing in the separated layer has been identified, the most effective location for the actuator has been defined regarding a reduction or increase in turbulence, lock-in of coherent flow structures has been demonstrated.
Regarding the simulation tasks, the main objectives of WP3 are the comparison and validation of numerical results with respect to experimental data. Numerical computations will focus on the flow control mechanisms, especially the manipulation of the Reynolds stresses and the suppression of flow-separation.
The task in WP3 is to numerically study the basic flow (without flow-control devices) and to compare the results obtained with experimental data in order to verify the ability of the computational tools to reproduce the main features of the flow.
In the present study, the mesh requirements were first considered for simulating the uncontrolled “base” flows accurately in order to detect possible unsteady turbulent flow physics, such as the characteristic frequencies related to shear layer instability, flow separation or vortex shedding, etc. In the simulation of flow field with oscillating control devices, as with the oscillating surfaces and synthetic/pulse jet, moving mesh strategies were taken to simulate accurately the unsteady flow and the effects of oscillating control devices on the flow field characteristics.
In WP3, the flow fields with active control devices were simulated as exactly the same as the experiments outlined in WP2, to validate the code and to identify the basic flow unsteady components and unsteady components from flow control devices. High accuracy numerical techniques have been developed to simulate the flow field with control devices in it, especially the techniques to simulate high frequency unsteady flow induced by flow control devices. From WP3 simulation the relationships of the unsteady property between the flow field and the frequency of devices were explored. The present consortium also investigated the interaction between the basic flow unsteady components and the unsteady components from the flow control devices to evaluate the proper control strategy for drag reduction as well as for separation control.
Integrating the contributions from all the partners, the conclusions related to the experimental analysis could be drawn as follows:
A. Measurement technique
Beside the ordinary measurement techniques, the advanced techniques, such as PIV, especially TOMO PIV, Global skin friction measurement and Data processing, play a very important role:
TOMO PIV, deployed by DLR for joint experimental campaign, provides the most detailed 3 dimensional information of turbulent flow and evaluates accurately the effectiveness of different actuators
Global skin friction measurement is based on fluorescent oil film and can provide detailed information of skin friction on a surface qualitatively and quantitatively. This is especially useful for present project.
Data processing requirement for present project is to extract Reynolds stress and individual organized motion in the controlled shear layer. This has been done by phase average technique and energetic decomposition such as proper orthogonal decomposition.
B. Development and deployment of actuators
Eight different flow control actuators are deployed for BFS and/or NACA0015. The effectiveness from experimental results of present project indicate the some promising actuators.
New type synthetic jet is based on nonlinear gas oscillation in a pipe and emitted from a slot. This synthetic jet can deliver the more energy in a more effective manner and result in more effective flow control both for BFS and airfoil.
Spanwise vortex generator induces an organized motion into and interacts with the separated shear layer from the upper corner of the step. The reattachment point moves upstream and the skin friction downstream increases.
SDBD plasma actuator induces a wall jet flow with periodic fluctuations that would influence the flow fields of BFS. The reattachment point can be reduced by 20% due to strong influence of the discharge on the shear layer development and resulting Reynolds stresses.
Pulsed jets cause convection of large eddies in the wake of NACA airfoil. Observation of the jet deployment process and jet removal process indicate significant fluctuation that may influence the Reynolds stress.
Piezoelectric actuators, used in present project, cause only minor fluctuation that is not enough to influence the flow fields. On the other hand, the location of the actuators may not be appropriate.
Suction is a traditional flow control technique. However present investigation verifies an important fact that unsteady disturbances come into effect only in the shear layer. The more is the shear rate, the more effective is the control.
C. Flow control over backward facing step
Flow control over backward facing step has been experimentally investigated with the aim of deep insight into mechanism of Reynolds stress manipulation by inducing periodic disturbances in to the separated shear layer. The relationship of the Reynolds stresses and skin friction with flow separation and reattachment has been clarified. The increased Reynolds shear stress, caused by different actuators, will result in shortened recirculation zone and reattachment point movement upstream.
D. Flow control of NACA0015
Experimental results provide a very detailed data base for modelling of turbulent flows, for their statistics of natural wakes in separated and attached situations, the two extreme situations of the unsteady process. These data can be used as test cases for steady RANS. In addition, transient behaviors are provided, being particularly suited for URANS approaches.
E. Further work
Although the proposed research work has been conducted and innovative achievements have been obtained, comparing with the original expectation of the project, some further activities may be complemented.
Measurement techniques:
TOMO PIV is a powerful tool, but time consuming. Further improvement of efficiency is needed.
Skin friction measurement has not been recognized by every partner. More accurate and convenient calibration method is to be developed.
Specific decomposition methods have to be developed for extracting the turbulent contributions in BFS.
Actuators:
Improving the existent actuators or developing new actuators for practical applications is a challenging task necessary to increase the TRL of flow control systems.
Mechanism analysis:
Detailed mechanism analysis for flow control in present project is not satisfactory because of limited schedule. Combining with CFD, further analysis may provide a deep insight into the interaction mechanisms.
On the other hand, the conclusions from the numerical analysis could be summarized as:
Preliminary numerical simulations of the flows without the flow control devices were performed by different partners with different solver and different turbulence approaches. For the backward-facing step case, numerical results show that the computational domain height in the wall normal direction is a key computational parameter for the accurately prediction of the reattachment length. For the NACA0015 airfoil with an angle of attack of 11 deg., the separation of the boundary layer on the upper surface of the airfoil could only be reproduced with the DDES-LLR or IDDES-SA turbulence approach.
Preliminary synthetic jet control computations were performed for backward-facing step to show a significant decrease in the reattachment length. Various control strategies were taken for the NACA0015 airfoil case. The numerical results show a reduction of the size of the recirculation bubble leading to a significant increase in lift and decrease in drag.
Taking into account the described conclusions, the consortium has selected 4 cases from the above list, two for the NACA profile and two for the Step:
- Profile NACA 0015
o With Synthetic Jets
o With Pulsed Jets
- Backward Facing Step
o With Plasma actuator
o With Synthetic Jets
These four selected cases have been used as the optimization analysis set-up.
A particular case to be mentioned is the set-up of an experimental optimization analysis. It has been defined with the experimental set-up by Poitiers, and the optimizer provided by CIMNE. The results obtained has been very promising thanks to the fast coupling between the wind tunnel and the optimizer, and the fast convergence of the optimization solution.
Potential Impact:
MARS was defined as an up-stream research project aimed to understand the mechanisms of drag generation. A better understanding of the physical phenomena was sought. In parallel, a secondary objective was to assess a set of flow control devices in order to identify a not only the best one with the best configuration but also to foresee the next generation of flow control devices. The research performed merged the experimental analysis with the numerical simulation in order to improve the understanding of the physical phenomena. In order to do so, additional research has been performed about the implementation of measurement techniques, cross-comparison of the flow control devices, cross-comparison of the experimental and numerical simulations, and finally, the definition of optimization analysis of the flow control devices applied to the test case geometries.
At a preliminary stage, the impact from the project could be summarized as:
- Improvement of experimental techniques for flow control device analysis: application of PIV and TOMO PIV techniques to the measurement of velocities and vorticity of the flow under no-control and control conditions.
- Better understanding of the physical phenomena, leading to a better set-up of the numerical simulations and the accurate definition of the most suitable numerical techniques.
- Setting up high demanding optimization analysis with experimental and numerical solvers.
- Industrial assessment on the applicability and the further research to be performed on this field.
List of Websites:
The European Website:
www.cimne.com/mars
The Chinese website:
http://ch-eu.cae.ac.cn/en/ProjectIndex.aspx?id=Mars
The MARS project has been developed under the umbrella of the 7th Research Framework programme of the European Commission (EC) and the International Science & Technology Cooperation Program of Civil Aviation, sponsored by Ministry of Industry and Information Technology of the People's Republic of China (MIIT).
This international collaborative effort was a successful outcome of the networking activities within the AEROCHINA2 project. MARS (Manipulation of Reynolds Stress for Separation Control and Drag Reduction) is one of the research projects generated as an outcome of the networking activities.
MARS is devoted to the analysis and understanding of the phenomena involved on the birth of the aerodynamic drag. Drag reduction and flow separation control are the high-level aims, while the project focus on the turbulence phenomena, and effect of Reynolds stress on the drag force.
The project was defined with the agreement of the European industry; namely Airbus, Alenia and Dassault Aviation, who considered the need to fully understand the initiation of the drag force in order to be able to deal with methods and techniques to reduce the skin friction on the aircraft.
The consortium, led by CAE (Chinese Aeronautical Establishment, China) in the Chinese side and CIMNE (International Centre for Numerical Methods in Engineering, Spain) in the European side, is compound by 10 Chinese partners and 12 European partners. The technical coordination is responsibility of ARI (Aeronautical Research Institute, China) and USFD (University of Sheffield, UK).
The objectives of the project can be summarized as:
1. To control turbulent wall bounded shear flow effectively from a more fundamental level by investigating directly the behaviour of Reynolds stresses and their response to the manipulation of the dynamic components of the flow
2. To conduct both experimental tests and computational simulations to extract reliable flow physics, including Reynolds stresses, for a number of active flow control devices
3. To explore large scale unsteadiness (unsteady jets, wakes and vortices) produced from the capability of these devices for to provide effective control of the dynamic components, and hence the Reynolds stresses, within the shear layers associated with large scale wakes and separations
4. To identify key strategies that enable efficient control of dynamic fluid structures within shear layers and to design and optimise these devices for separation control (higher shear) and drag reduction (lower shear)
5. To demonstrate the most promising device/devices at relevant scale, the ability to efficiently increase or decrease the Reynolds stresses
6. To investigate the application of this capability within an aircraft context
7. To foster further collaboration between European and Chinese researchers in the key technology area of flow control for the benefit of the civil aircraft industries on both sides.
The project has been planned for 3 years, with a final effective duration of 3.5 years.
Project Context and Objectives:
Drag reduction and separation control are directly related to more efficient air transportation and less emission of harmful gases into the environment. While the aerospace industry is striving to have more and more optimised designs, it is still some way away from the targets set out in the ACARE 2020 vision for 50% reduction in aircraft emissions. Separation control and drag reduction contribute directly towards this target and active flow control could plays an important role in achieving it. After many decades of development, the highly optimised aircraft designs make further large improvements difficult without a ‘game-changing’ technology such as active flow control. Active flow control provides an additional dimension for further improving aircraft performance, in particular, for performance at different operational points, e.g. at both cruise and take-off/landing conditions.
The importance of reducing skin friction on meeting the global fuel burn targets is obvious. At cruise approximately one half of the total drag of a modern commercial transport aircraft is attributed to skin friction. The importance of improved separation control is less obvious unless we acknowledge the influence of aircraft mass on fuel burn. In fact the sensitivity of fuel burn to mass is greater than that to skin friction. Therefore if the structural mass of the aircraft can be reduced by more efficient low speed configurations or improved load alleviation off-design then this carries a significant direct benefit on cruise fuel burn.
The turbulence Reynolds stress is the most important dynamic quantity affecting the mean flow as it is responsible for a major part of the momentum transfer in the wall bounded turbulent flow. It has a direct relevance to both skin friction (for a turbulent boundary layer) and flow separation (occurs when skin friction drops to zero). The near wall region for a turbulent boundary layer can be divided into the viscous sub-layer, where the mean viscous stress is important and the approximately constant Reynolds stress region, where the viscous stress drops to zero and the Reynolds stress peaks. As the Reynolds number increases, the peak Reynolds stress approaches the value of the viscous stress at the wall. Therefore active manipulation of the Reynolds stress can directly lead to changes in the viscous stress at the wall so as to effectively control the flow for effective flow control.
However, there is a lack of current understanding of the inter-relationship between the various flow control devices and the Reynolds stresses in the flow field they produced. An improved understanding can potentially significantly improve the effectiveness of flow control as the Reynolds stresses are closely related to the flow behaviour at the surface for effective separation control or drag reduction. A variety of control devices are available and new ones are invented but which one for what purpose is an open question yet to be fully answered.
The majority of prior work has focused on the introduction of changes to the mean flow that resulted in changes to the Reynolds stress. In the present proposal we propose to reverse that process and consider the long term goal of controlling dynamic structures that then influence the Reynolds stress that in turn changes the mean flow. This radical approach recognises that we are still some way away from hardware to implement the concept at flight scales but if successful, would establish a first important step towards our ultimate ambition.
The focus of the present project will be on the effects of a number of active flow control devices on the discrete dynamic components of the turbulent shear layers and the Reynolds stress. From the application point of view, the current proposal provides a positive and necessary step in the right direction wherein it will demonstrate the capability to control individual structures that are larger in scale and lower in frequency compared to the richness of the time and spatial scales in a turbulent boundary layer. In order to focus the current project with the limited duration and budget, the current project excludes laminar flow control for transition delay, another area of extensive previous and current research. Furthermore, the project will put more effort on active flow control means rather than passive controls.
To explain our basic strategy in manipulating Reynolds stresses through the dynamic components of the turbulent shear layers, it is helpful to start with the triple decomposition proposed by Reynolds and Hussain (1972) for an instantaneous velocity, U
U = Ū + ũ + uˈ
The first term on the RHS is the time averaged mean velocity. If we attempt to control this via flow control then most devices offer little gain in efficiency on a global energy basis i.e. change in energy out equals energy in.
The second term on the RHS is the periodic/dynamic component of the flow and for some specific flow scenarios this can be shown to be dominant in determining the flow state and characteristics. The stresses produced from this term are referred as the periodic stresses or “apparent stresses”. It offers some interesting opportunities for demonstrating the way in which to deploy flow control technologies for dynamic environments (responsive environments, smart inputs and sensible control). This also implies that, for statistically steady flows, where the second term disappears, artificial introduction of the periodic term may be necessary for effective control.
The final term on the RHS represents the broadband ‘random’ turbulent fluctuations, from which the Reynolds stresses are defined. While direct control of the “random” components is the ultimate goal, the current project aims to investigate the control of the periodic stresses, the dynamic components of the flow, in order to manipulate the Reynolds stress for the benefit of flow control.
With the basic strategy for MARS established the global aims of the project can be defined. They are:
• To use the periodic flows embedded in the two identified flow cases as platforms in which direct control of discrete dynamic structures to manipulate the Reynolds stress can be observed, measured and simulated.
• To measure, simulate and understand the impact of certain actuators upon discrete structures in a turbulent shear layer and to identify candidate actuators for further development for skin friction reduction and flow separation control at flight scales.
An important characteristic of this proposal is that in MARS we wish to explore the possibility of influencing the mean flow via the Reynolds stress by direct manipulation of discrete structures that contribute to the stress. This is in contrast to most previous work in which the actuators change the mean flow directly. Therefore, the MARS approach offers the potential for higher system gain.
The proposed project is based on the mutual understanding of the EU and Chinese partners, mostly from the AeroChina and AeroChinaII consortia funded by the EU. In AeroChinaII, a Working Group on Flow Control was established due to a strong interest in the area from both sides. A series of EU-China Flow Control workshops were held, which hugely improved our understanding of the capabilities in both experiments and simulations and pacing problems in industrial applications. While the MARS consortium is based on the key members of this working group, some new EU and Chinese partners were invited in the current project in order to further strengthen the current consortium.
Project Results:
MARS project was organized in 5 work packages; 3 of them including technical tasks. The analysis of the flow control devices, and their capacity to control and modify the Reynolds stress was done mainly in WP2 and WP3. WP2 grouped the experimental tasks, WP3 grouped the numerical simulation tasks, and both of them fed the WP4, where the down-selection of the best performing devices and the optimization analysis of the selected ones were planned.
The project defined a set of test cases which have been analysed experimentally and numerically. The list of test cases is:
- Profile NACA 0015
o Without control device
o With Plasma actuators
o With Synthetic Jets actuators
o With Micro-Blowing/Suction actuators
o With Pulsed Jets actuators
- Backward Facing Step
o Without control device
o With Plasma actuators
o With Oscillating surface actuators
o With Micro-Blowing/Suction actuators
o With Synthetic Jets actuators
o With Spanwise Vortex Generators
Regarding the experimental test cases, the primary objective of these experiments was to explore the possibility of influencing the mean flow downstream a backward-facing step or in the wake of a NACA0015 model by using a variety of actuators. The present investigations demonstrate the influence of these devices on the development of the free shear layer. Some of these actuators can provide a high mass injection or velocity ratio (synthetic jets, pulsed jets,…) when others can only promotes small amplitude perturbations in the natural flow (plasma actuators, piezoelectric actuators,…). For the latter actuators, the flow dynamic cannot be drastically changed without fine control strategies as developed in this project.
Regarding the proposed experimental researches as defined in the DoW, most of the proposal content has been conducted. Development of innovative actuators has been conducted, the dynamical behaviour of the separated boundary layer has been analysed by mean and time-resolved measurements, the wall signature of periodic structures existing in the separated layer has been identified, the most effective location for the actuator has been defined regarding a reduction or increase in turbulence, lock-in of coherent flow structures has been demonstrated.
Regarding the simulation tasks, the main objectives of WP3 are the comparison and validation of numerical results with respect to experimental data. Numerical computations will focus on the flow control mechanisms, especially the manipulation of the Reynolds stresses and the suppression of flow-separation.
The task in WP3 is to numerically study the basic flow (without flow-control devices) and to compare the results obtained with experimental data in order to verify the ability of the computational tools to reproduce the main features of the flow.
In the present study, the mesh requirements were first considered for simulating the uncontrolled “base” flows accurately in order to detect possible unsteady turbulent flow physics, such as the characteristic frequencies related to shear layer instability, flow separation or vortex shedding, etc. In the simulation of flow field with oscillating control devices, as with the oscillating surfaces and synthetic/pulse jet, moving mesh strategies were taken to simulate accurately the unsteady flow and the effects of oscillating control devices on the flow field characteristics.
In WP3, the flow fields with active control devices were simulated as exactly the same as the experiments outlined in WP2, to validate the code and to identify the basic flow unsteady components and unsteady components from flow control devices. High accuracy numerical techniques have been developed to simulate the flow field with control devices in it, especially the techniques to simulate high frequency unsteady flow induced by flow control devices. From WP3 simulation the relationships of the unsteady property between the flow field and the frequency of devices were explored. The present consortium also investigated the interaction between the basic flow unsteady components and the unsteady components from the flow control devices to evaluate the proper control strategy for drag reduction as well as for separation control.
Integrating the contributions from all the partners, the conclusions related to the experimental analysis could be drawn as follows:
A. Measurement technique
Beside the ordinary measurement techniques, the advanced techniques, such as PIV, especially TOMO PIV, Global skin friction measurement and Data processing, play a very important role:
TOMO PIV, deployed by DLR for joint experimental campaign, provides the most detailed 3 dimensional information of turbulent flow and evaluates accurately the effectiveness of different actuators
Global skin friction measurement is based on fluorescent oil film and can provide detailed information of skin friction on a surface qualitatively and quantitatively. This is especially useful for present project.
Data processing requirement for present project is to extract Reynolds stress and individual organized motion in the controlled shear layer. This has been done by phase average technique and energetic decomposition such as proper orthogonal decomposition.
B. Development and deployment of actuators
Eight different flow control actuators are deployed for BFS and/or NACA0015. The effectiveness from experimental results of present project indicate the some promising actuators.
New type synthetic jet is based on nonlinear gas oscillation in a pipe and emitted from a slot. This synthetic jet can deliver the more energy in a more effective manner and result in more effective flow control both for BFS and airfoil.
Spanwise vortex generator induces an organized motion into and interacts with the separated shear layer from the upper corner of the step. The reattachment point moves upstream and the skin friction downstream increases.
SDBD plasma actuator induces a wall jet flow with periodic fluctuations that would influence the flow fields of BFS. The reattachment point can be reduced by 20% due to strong influence of the discharge on the shear layer development and resulting Reynolds stresses.
Pulsed jets cause convection of large eddies in the wake of NACA airfoil. Observation of the jet deployment process and jet removal process indicate significant fluctuation that may influence the Reynolds stress.
Piezoelectric actuators, used in present project, cause only minor fluctuation that is not enough to influence the flow fields. On the other hand, the location of the actuators may not be appropriate.
Suction is a traditional flow control technique. However present investigation verifies an important fact that unsteady disturbances come into effect only in the shear layer. The more is the shear rate, the more effective is the control.
C. Flow control over backward facing step
Flow control over backward facing step has been experimentally investigated with the aim of deep insight into mechanism of Reynolds stress manipulation by inducing periodic disturbances in to the separated shear layer. The relationship of the Reynolds stresses and skin friction with flow separation and reattachment has been clarified. The increased Reynolds shear stress, caused by different actuators, will result in shortened recirculation zone and reattachment point movement upstream.
D. Flow control of NACA0015
Experimental results provide a very detailed data base for modelling of turbulent flows, for their statistics of natural wakes in separated and attached situations, the two extreme situations of the unsteady process. These data can be used as test cases for steady RANS. In addition, transient behaviors are provided, being particularly suited for URANS approaches.
E. Further work
Although the proposed research work has been conducted and innovative achievements have been obtained, comparing with the original expectation of the project, some further activities may be complemented.
Measurement techniques:
TOMO PIV is a powerful tool, but time consuming. Further improvement of efficiency is needed.
Skin friction measurement has not been recognized by every partner. More accurate and convenient calibration method is to be developed.
Specific decomposition methods have to be developed for extracting the turbulent contributions in BFS.
Actuators:
Improving the existent actuators or developing new actuators for practical applications is a challenging task necessary to increase the TRL of flow control systems.
Mechanism analysis:
Detailed mechanism analysis for flow control in present project is not satisfactory because of limited schedule. Combining with CFD, further analysis may provide a deep insight into the interaction mechanisms.
On the other hand, the conclusions from the numerical analysis could be summarized as:
Preliminary numerical simulations of the flows without the flow control devices were performed by different partners with different solver and different turbulence approaches. For the backward-facing step case, numerical results show that the computational domain height in the wall normal direction is a key computational parameter for the accurately prediction of the reattachment length. For the NACA0015 airfoil with an angle of attack of 11 deg., the separation of the boundary layer on the upper surface of the airfoil could only be reproduced with the DDES-LLR or IDDES-SA turbulence approach.
Preliminary synthetic jet control computations were performed for backward-facing step to show a significant decrease in the reattachment length. Various control strategies were taken for the NACA0015 airfoil case. The numerical results show a reduction of the size of the recirculation bubble leading to a significant increase in lift and decrease in drag.
Taking into account the described conclusions, the consortium has selected 4 cases from the above list, two for the NACA profile and two for the Step:
- Profile NACA 0015
o With Synthetic Jets
o With Pulsed Jets
- Backward Facing Step
o With Plasma actuator
o With Synthetic Jets
These four selected cases have been used as the optimization analysis set-up.
A particular case to be mentioned is the set-up of an experimental optimization analysis. It has been defined with the experimental set-up by Poitiers, and the optimizer provided by CIMNE. The results obtained has been very promising thanks to the fast coupling between the wind tunnel and the optimizer, and the fast convergence of the optimization solution.
Potential Impact:
MARS was defined as an up-stream research project aimed to understand the mechanisms of drag generation. A better understanding of the physical phenomena was sought. In parallel, a secondary objective was to assess a set of flow control devices in order to identify a not only the best one with the best configuration but also to foresee the next generation of flow control devices. The research performed merged the experimental analysis with the numerical simulation in order to improve the understanding of the physical phenomena. In order to do so, additional research has been performed about the implementation of measurement techniques, cross-comparison of the flow control devices, cross-comparison of the experimental and numerical simulations, and finally, the definition of optimization analysis of the flow control devices applied to the test case geometries.
At a preliminary stage, the impact from the project could be summarized as:
- Improvement of experimental techniques for flow control device analysis: application of PIV and TOMO PIV techniques to the measurement of velocities and vorticity of the flow under no-control and control conditions.
- Better understanding of the physical phenomena, leading to a better set-up of the numerical simulations and the accurate definition of the most suitable numerical techniques.
- Setting up high demanding optimization analysis with experimental and numerical solvers.
- Industrial assessment on the applicability and the further research to be performed on this field.
List of Websites:
The European Website:
www.cimne.com/mars
The Chinese website:
http://ch-eu.cae.ac.cn/en/ProjectIndex.aspx?id=Mars