Final Report Summary - SUPERTRAC (Supersonic transition control)
The global objective of the SUPERTRAC project was to carry out fundamental, numerical and experimental investigations for evaluating the possibilities of laminar flow control on supersonic civil aircraft wings.
Laminar flow can be achieved by delaying the onset of laminar-turbulent transition on the wings using specific tools such as shape optimisation, suction or micron-sized roughness elements. Reducing the extent of turbulent flow is of considerable practical interest because it reduces the friction drag. It also contributes to satisfy the severe requirements on emission and noise, because drag reduction is directly related to the reduction of weight and size, as well as fuel burn and noise. Laminar flow control techniques have been widely tested for subsonic and transonic flows, but little is known about their extension to supersonic flows. This justifies the work undertaken within SUPERTRAC.
The SUPERTRAC project was divided in six Work packages (WPs):
- WP1: Specifications
- WP2: Laminar flow control by micron-sized roughness elements and anti-contamination devices
- WP3: Laminar flow control by suction
- WP4: Natural laminar flow control (numerical model)
- WP5: Results evaluation for further development.
In the WP1, the industrial partners (Airbus and Dassault aviation) provided a quantitative definition of the objectives as well as the preliminary definition of a fully three-dimensional wing which was used as a reference shape in WP4 (numerical model). Another objective was to make a review of the (few) available experimental data on supersonic transition in three-dimensional flow (swept wings). In particular, swept wing experiments performed at DLR, before the project starts, were re-analysed.
The first objective of WP2 was to define a simple model (swept wing of constant chord) equipped with micron-sized roughness elements and anti-contamination devices. This model was manufactured and tested in the S2 wind tunnel of the Modane-Avrieux ONERA centre. The analysis of the results was then shared between the partners.
WP3 was running in parallel with WP2. Another swept wing of constant chord, equipped with a suction panel in the leading edge region, was designed, manufactured and tested in the RWG wind tunnel of DLR Göttingen, then the results were analysed.
WP4 used the 'numerical' model defined in WP1. The objectives were:
i) to numerically investigate the concept of natural laminar flow control by shape optimisation;
ii) to analyse the compatibility of the different control techniques, in particular those of WP2 and WP3.
This resulted in the definition of the best compromise for skin friction drag reduction. The benefits which can be expected with the 'best' wing shape were estimated by comparing the performances of the optimised wing and those of a fully turbulent wing.
The results of WP2 to WP4 were summarised in WP5 by the industrial partners, with the objective of providing a quantification of the benefits and recommendations for practical applicability to future supersonic aircraft wings.
WP6 was devoted to the management and to the exploitation of the project.
The consortium used the most advanced numerical tools for the investigation of laminar flow control devices in supersonic conditions. These tools have been improved when necessary for solving the particular problems encountered within SUPERTRAC, and the knowledge of the partners has been substantially increased in many aspects.
The transition control by MSR is a new approach, which had never been validated in Europe, at least for supersonic conditions. The computations allowed a critical assessment of the capabilities of this concept. A strategy for the use of nonlinear Parabolised stability equations (PSE) was developed by the partners, so that systematic applications of this control technique are now possible, at least numerically. On the other side, the theoretical difficulties associated with nonlinear PSE have been identified; for instance, the spacing of the roughness elements can be determined but their height and their diameter remain unknown. Advanced receptivity theories (as used by DLR) provided some answers.
As far as suction is concerned, the numerical definition of the DLR model made use not only of classical tools (linear stability theory) but also of advanced optimisation methods (adjoint based) for the final design of the suction chamber.
The numerical optimisation of a fully three-dimensional supersonic wing for the purpose of natural laminar flow led to unexpected difficulties (for the partners, this was the first application of this concept to supersonic flows), which needed to propose a specific strategy. CIRA and ONERA developed a complete optimisation chain specific to the SUPERTRAC activities. The use of genetic algorithms was one of the most interesting achievements of the project.
Wind tunnel experiments allowed to judge the validity of the theoretical modelling and to answer (some of) the open questions. It must be pointed out that the MSR, ACD and suction experiments were the first ones of this type in Europe for supersonic conditions. All the results were made available to all the partners as electronic files and CD ROMS.
The MSR experiments did not lead to the expected positive results in term of transition control. A careful examination of the experimental data base and a detailed analysis of the results using advanced numerical tools allowed understanding the possible reasons for these disappointing results. There is no doubt that the knowledge gained within SUPERTRAC will be useful for future investigations on transition control by MSR.
Before the SUPERTRAC S2MA test campaign, the ACD concept had been studied at NASA in the early 90's. One of the tested devices seemed to work rather well in some conditions, but no detailed information was published concerning the optimum shape and size of the device. A similar device was considered within SUPERTRAC, but its performances were found to be very poor. Much better and quite spectacular results were obtained with a completely original shape, the size of which was estimated from RANS computations. It is guessed that the best SUPERTRAC device could be used at lower speed and could exhibit a better efficiency than the classical Gaster bumps.
SUPERTRAC also provided information of practical / industrial interest concerning the possibilities of laminar flow control at supersonic speeds. Some of them are extrapolations of results already established in the transonic regime. For instance, suction is more efficient for a wing designed for Natural laminar flow (NLF) than for a turbulent wing. Other practical results have been obtained for the first time. For instance, NLF is not compatible with the use of MSR.
As a final achievement of the project, the 'best' supersonic 3D wing has been defined and the expected benefits (in term of drag and fuel consumption reduction) have been estimated. It is clear that for the large sweep angle wing considered here, NLF alone is not sufficient for obtaining significant skin friction gains. However the application of a small amount of suction makes it possible to increase the laminar flow extent in a significant manner. Of course, many technological problems, such as the compatibility with leading edge high lift devices or the effect of surface imperfections need to be studied. These issues were out of the scope of the present project but could be addressed in future projects dealing with laminarity at high speed.
Laminar flow can be achieved by delaying the onset of laminar-turbulent transition on the wings using specific tools such as shape optimisation, suction or micron-sized roughness elements. Reducing the extent of turbulent flow is of considerable practical interest because it reduces the friction drag. It also contributes to satisfy the severe requirements on emission and noise, because drag reduction is directly related to the reduction of weight and size, as well as fuel burn and noise. Laminar flow control techniques have been widely tested for subsonic and transonic flows, but little is known about their extension to supersonic flows. This justifies the work undertaken within SUPERTRAC.
The SUPERTRAC project was divided in six Work packages (WPs):
- WP1: Specifications
- WP2: Laminar flow control by micron-sized roughness elements and anti-contamination devices
- WP3: Laminar flow control by suction
- WP4: Natural laminar flow control (numerical model)
- WP5: Results evaluation for further development.
In the WP1, the industrial partners (Airbus and Dassault aviation) provided a quantitative definition of the objectives as well as the preliminary definition of a fully three-dimensional wing which was used as a reference shape in WP4 (numerical model). Another objective was to make a review of the (few) available experimental data on supersonic transition in three-dimensional flow (swept wings). In particular, swept wing experiments performed at DLR, before the project starts, were re-analysed.
The first objective of WP2 was to define a simple model (swept wing of constant chord) equipped with micron-sized roughness elements and anti-contamination devices. This model was manufactured and tested in the S2 wind tunnel of the Modane-Avrieux ONERA centre. The analysis of the results was then shared between the partners.
WP3 was running in parallel with WP2. Another swept wing of constant chord, equipped with a suction panel in the leading edge region, was designed, manufactured and tested in the RWG wind tunnel of DLR Göttingen, then the results were analysed.
WP4 used the 'numerical' model defined in WP1. The objectives were:
i) to numerically investigate the concept of natural laminar flow control by shape optimisation;
ii) to analyse the compatibility of the different control techniques, in particular those of WP2 and WP3.
This resulted in the definition of the best compromise for skin friction drag reduction. The benefits which can be expected with the 'best' wing shape were estimated by comparing the performances of the optimised wing and those of a fully turbulent wing.
The results of WP2 to WP4 were summarised in WP5 by the industrial partners, with the objective of providing a quantification of the benefits and recommendations for practical applicability to future supersonic aircraft wings.
WP6 was devoted to the management and to the exploitation of the project.
The consortium used the most advanced numerical tools for the investigation of laminar flow control devices in supersonic conditions. These tools have been improved when necessary for solving the particular problems encountered within SUPERTRAC, and the knowledge of the partners has been substantially increased in many aspects.
The transition control by MSR is a new approach, which had never been validated in Europe, at least for supersonic conditions. The computations allowed a critical assessment of the capabilities of this concept. A strategy for the use of nonlinear Parabolised stability equations (PSE) was developed by the partners, so that systematic applications of this control technique are now possible, at least numerically. On the other side, the theoretical difficulties associated with nonlinear PSE have been identified; for instance, the spacing of the roughness elements can be determined but their height and their diameter remain unknown. Advanced receptivity theories (as used by DLR) provided some answers.
As far as suction is concerned, the numerical definition of the DLR model made use not only of classical tools (linear stability theory) but also of advanced optimisation methods (adjoint based) for the final design of the suction chamber.
The numerical optimisation of a fully three-dimensional supersonic wing for the purpose of natural laminar flow led to unexpected difficulties (for the partners, this was the first application of this concept to supersonic flows), which needed to propose a specific strategy. CIRA and ONERA developed a complete optimisation chain specific to the SUPERTRAC activities. The use of genetic algorithms was one of the most interesting achievements of the project.
Wind tunnel experiments allowed to judge the validity of the theoretical modelling and to answer (some of) the open questions. It must be pointed out that the MSR, ACD and suction experiments were the first ones of this type in Europe for supersonic conditions. All the results were made available to all the partners as electronic files and CD ROMS.
The MSR experiments did not lead to the expected positive results in term of transition control. A careful examination of the experimental data base and a detailed analysis of the results using advanced numerical tools allowed understanding the possible reasons for these disappointing results. There is no doubt that the knowledge gained within SUPERTRAC will be useful for future investigations on transition control by MSR.
Before the SUPERTRAC S2MA test campaign, the ACD concept had been studied at NASA in the early 90's. One of the tested devices seemed to work rather well in some conditions, but no detailed information was published concerning the optimum shape and size of the device. A similar device was considered within SUPERTRAC, but its performances were found to be very poor. Much better and quite spectacular results were obtained with a completely original shape, the size of which was estimated from RANS computations. It is guessed that the best SUPERTRAC device could be used at lower speed and could exhibit a better efficiency than the classical Gaster bumps.
SUPERTRAC also provided information of practical / industrial interest concerning the possibilities of laminar flow control at supersonic speeds. Some of them are extrapolations of results already established in the transonic regime. For instance, suction is more efficient for a wing designed for Natural laminar flow (NLF) than for a turbulent wing. Other practical results have been obtained for the first time. For instance, NLF is not compatible with the use of MSR.
As a final achievement of the project, the 'best' supersonic 3D wing has been defined and the expected benefits (in term of drag and fuel consumption reduction) have been estimated. It is clear that for the large sweep angle wing considered here, NLF alone is not sufficient for obtaining significant skin friction gains. However the application of a small amount of suction makes it possible to increase the laminar flow extent in a significant manner. Of course, many technological problems, such as the compatibility with leading edge high lift devices or the effect of surface imperfections need to be studied. These issues were out of the scope of the present project but could be addressed in future projects dealing with laminarity at high speed.
