Final Report Summary - ROD (Rotorcraf Drag Reduction)
Executive Summary:
The first part of the activity of the project was related to the design and production of the model. Although some parts were already existent and have been delivered by GRC2 the model required a completely new internal structure including the rotor hub driving system with the swash plate for the blade pitch control. Thus the model was designed, manufactured, assembled and instrumented. The dimensioning was based on the loads evaluation coming from the CFD simulations carried out in the frame of this project.
The model could be installed on the wind tunnel pylon in both upside-down and upright mode. The upside-down tests allowed to study the effect of devices layed on the fuselage lower surface (as the vortex generators) avoiding pylon disturbances. On the other hand the upright tests allowed to evaluate the performance of the helicopter including the rotor-hub (with blade stubs).
Furthermore the optimized components (vortex generators, gear sponsons, rotor hub caps and blade root fairings) have been completely designed and manufactured.
A comprehensive wind tunnel campaign was performed to assess the aerodynamic effectiveness of CFD-based optimized helicopter components for drag reduction. Several configurations at different angle of attack and side-slip have been tested to assess the performances. Moreover, a purposely designed and developed setup was used to perform several stereo PIV surveys in the region just after the back ramp and in the region just before the fin. The acquired flow fields (as well as the measured steady and unsteady pressures) were quite useful in the physical interpretation of the results.
A CFD activity supported all the design and experimental activity and was particularly important for the evaluation of the wind tunnel effects corrections.
Project Context and Objectives:
Due to the expected growth of the rotorcraft traffic for passenger transport, the rotorcraft contribution to environmental impact, negligible today, would become more significant in next decade unless a major initiative succeeds in keeping it under control. As the power required to fly a helicopter in forward flight (and consequently the fuel burn and CO2 emissions) is strongly dependent on the aerodynamic drag of its airframe and of non-lifting rotating parts, the design of this parts needs to be optimized with the aim of the drag reduction.
The ROD project pertains to the wind tunnel testing of a common helicopter platform of the heavy-weight class. The aim of the testing activity is the evaluation of the effectiveness (mainly in terms of drag reduction) of the shape optimisation performed by GRC2 consortium on several components. Both original and optimised models were tested to assess the optimisation effectiveness by comparison. The tests was complemented by a CFD simulation activity for the evaluation of the wind tunnel corrections and finally a physical interpretation of the obtained results was presented.
The project was essentially based on a wind tunnel activity split into two different test campaigns: the first one regarding the original helicopter configuration and the second one regarding the version optimized by GRC2. The differences between the performances with the two versions allowed to estimate the effectiveness of the optimization. Furthermore flow-field and pressure surveys were carried out to outline some physical interpretation of the results producing more generally usable information. The tests results were corrected for wind tunnel effects using direct measurements and the results of CFD simulations.
Project Results:
The tests with the model in upside-down configuration were addressed to evaluate the effect of the VG array optimised by ONERA simulations avoiding the interference of the strut wake. The drag measurements carried out with the model at cruise angle of attack showed that the smaller counter-rotating VG array provides the higher drag reduction (see deliverable D3.2). The best VG configuration (small counter-rotating) was therefore selected to be subject of a detailed investigation. In particular, the functioning of the best VG array was investigated in the α-sweep range. The experimental tests results show an appreciable reduction of drag due to VG in the range -4° < α < 10° with highest drag reduction obtained for α = 2°. The comparison of CFD simulations by ONERA and wind tunnel results shows a very good agreement between them in terms of the drag coefficient difference evaluated with the best VG, with only exception of the lowest angle of attack considered for the CFD simulation (see deliverable D5.1).
Static pressure measurements and PIV surveys carried out at the back-ramp region compared to CFD results enabled to achieve a detailed insight about the physics related to the functioning of the best VG array (smaller counter-rotating). A fundamental effect of the VG array, when located downstream the pronounced upsweep characterizing the blunt fuselage, is twofold. First in triggering a smooth increase of the adverse pressure gradient at the backdoor and second in re-energising the boundary layer and thus preventing or limiting flow separation. The second effect is clearly visible from the comparison between the velocity field measured in the back-ramp region with and without the VG. In particular, the highest benefit introduced by the small counter-rotating VG in terms of drag reduction was found for the angle of attack α = 2°, as clearly shown by both CFD and experiments.
The analysis of the flow field around the ramp region evaluated making use of PIV surveys and CFD calculations showed that for the clean geometry configuration (without the VG) the flow close to the ramp is characterised by a large separation. The extent of this separated flow region in both longitudinal and span-wise direction is clearly indicated by the negative values of the stream-wise velocity component measured on X-Z and Y-Z planes. On the other hand, PIV surveys for the geometry equipped with the VG do not show back-flow region in the volume of investigation. A general quite good agreement between CFD and PIV results can be observed on the Y -Z planes, even if time-averaged PIV results show a more symmetrical structure of the separated region with respect to CFD. The flow topology at the rear-ramp being very sensitive to pressure fluctuations, it is numerically experienced for this geometry that after a transient phase the flow symmetry at the rear-ramp is lost once the flow starts separating. For the model geometry equipped with the VG, CFD results show that the flow is still slightly separated close to the backdoor/tail boom junction. Moreover, the comparison of the vertical velocity component shows that this VG configuration was not designed to control the longitudinal vortex structures issued by the blunt fuselage.
Static pressure measurements confirm the potential benefit of VG in reducing drag by increasing the pressure field and therefore by limiting the suction effect responsible for pressure drag. Very small effects of the VG arrays are visible for both the experiments and CFD in the region before the VG arrays. On the other hand, on the backdoor surface downstream the VG array, a clear increase of the pressure can be observed from both experiments and CFD where the flow-field is further decelerated. Note that the pressure rise between the clean geometry and VG is quite similar in the experiments and in the simulations.
The unsteady pressure measurements carried out over the back-ramp surface provide a comparison of the unsteadiness level of the flow in this region between the clean geometry case and the model configuration with the VG. Coherently with the PIV velocity field surveys, the measurements carried out by Kulite transducers located on the mid-span plane of the back-ramp downstream the VG array with the clean geometry show a rather higher level of pressure fluctuations, confirming the higher level of unsteadiness that characterise the separated flow region without the VG.
The tests with the model in upright configuration with the rotating hub were mainly addressed to evaluate the performance of the different optimised hub configuration in terms of drag reduction. In order to obtain an accurate estimate of the contribution to the aerodynamic performance, the tests were performed by adding all the optimised components starting from the original to the final optimised configuration. The first hub-cap and fairings provided by ONERA are respectively called in the following small hub-cap and stubs fairings. The second hub-cap tested was manufactured starting from the external optimised shape of a full hub fairing optimised by DLR. Differently from the original design by DLR, this component presents a large open underside necessary to avoid interference with the actual rotor hub. In the following this hub-cap is called large hub-cap. DLR provided also the shape of the optimised set of sponsons (namely new sponsons).
The best performance between the hub-caps was found with the small hub-cap. The blade stubs attachments fairings produce a small increase of the performance with the small hub cap, while they produce a contained decrease of the large hub-cap. Therefore, the wind tunnel activity showed that an overall maximum drag reduction of 6.1% can be obtained at cruise attitude with respect to the original geometry.
PIV surveys were carried out to investigate the possible tail-shake effect of the different optimised hub-caps due to the rotor hub wake. Thus, the stereo PIV surveys were carried out in a measurement volume just before the fin. A quantitative evaluation on the extent of the rotor hub wake was obtained by the comparison of the streamwise velocity component profiles extracted from the measurement volume at model mid-span closest to the fin. PIV results for the original hub-cap configuration show that the velocity defect region is confined in the lower part of the measurement volume close to the tail boom. Thus, the rotor hub wake influences only the lowest part of the fin. On the other hand, a wider velocity defect region can be observed from the PIV results obtained with both the large and small optimised hub-caps. In particular, the area with the higher velocity defect is more extended for the large hub-cap configuration. This means that the optimised hub caps do not deflect the wake enough to avoid the collision with the fin. Moreover, the blade stubs attachments fairings do not produce appreciable effects on the rotor hub wake.
The unsteady pressure measurements carried out on the fin provide interesting information about the unsteadiness of the rotor hub wake for the different hub caps tested. The pressure RMS are higher on the port side corresponding to the upper side of the fin airfoils that is more sensible to the instantaneous incidences variations. In particular, the RMS of the pressure signals measured by the highest Kulite transducers KF1 and KF2 is lower for the original hub cap configuration with respect to both the small and large hub cap. The RMS value of the pressure signals measured without the hub cap are similar to the ones obtained with the optimised hub caps. The RMS values confirm that the flow impinging the higher part of the fin presents an higher level of unsteadiness for the rotor hub configuration with the large hub cap. A further confirmation of this feature is given by the highest amplitude of the spectrum peak corresponding to the rotor 4-per-rev frequency obtained from the KF1 and KF2 transducers for the large hub cap configuration. A similar level of pressure fluctuation was observed from the measurements carried out by the lower KF3 transducer for the three hub cap configurations tested and for the configuration without the hub cap. The lowest transducer KF4 shows an higher value of the signal RMS for the original hub cap configuration.
Potential Impact:
The present project is in the frame of CleanSky Project and namely in the frame of Green Rotorcraft (GRC) Integrated Technology Demonstrators. The aim of GRC is to develop new technologies for the reduction of helicopters noise and emissions. The remarkable reduction in the drag is a significant indication about the direction to follow in order to build less polluting helicopters. Furthermore all the data acquired and the physical observations can be very useful for helicopter designing, including some safety items (as the problem of the tail shaking).
The Project web page is on-line since the 26 March 2013. Due to the nature of single participant project this site is more devoted to dissemination than to project management.
Up to now, the following publications have been produced/prepared:
1. G. Gibertini, F. Auteri, G. Campanardi, G. Droandi, D. Grassi, A. Le Pape and A. Zanotti. A Test Rig to Assess the Effectiveness of Drag Reduction Devices on a Heavy-Class Helicopter. 41th ERF, September 1–4, 2015, Munich, Germany.
2. G. Gibertini, J.C. Boniface, A. Zanotti, G. Droandi, F. Auteri, R. Gaveriaux and A. Le Pape. Helicopter drag reduction by vortex generators. Aerospace Science and Technology, Vol. 47, pp. 324-339, 2015.
3. G. Gibertini. Drag and Drop. To appear in Aerospace Testing International.
4. A. Zanotti, G. Droandi, G. Gibertini, F. Auteri, J.C. Boniface, R. Gaveriaux and A. Le Pape. Wind Tunnel Assessment of the Computational Framework for Helicopter Fuselage Drag Reduction Using Vortex Generators. Abstract submitted for the AHS 72nd Annual Forum, West Palm Beach, Florida, USA, May 17–19, 2016.
5. A. Zanotti, G. Droandi, G. Gibertini, D. Grassi, G. Campanardi, F. Auteri, A. Aceti and A. Le Pape. Wind Tunnel Tests of a Heavy-Class Helicopter Optimised for Drag Reduction. Submitted to Aeronautical Journal.
6. A. Zanotti, G. Droandi, F. Auteri, G. Gibertini and A. Le Pape. Robustness and Limits of Vortex Generators Effectiveness in Helicopter Drag Reduction. Submitted to Journal of the American Helicopter Society.
Furthermore, taking into account the problem of confidentiality, the test activity has been shown to the graduated students of Aeronautical Engineering of Politecnico di Milano.
List of Websites:
www.aero.polimi.it/rodproject
The first part of the activity of the project was related to the design and production of the model. Although some parts were already existent and have been delivered by GRC2 the model required a completely new internal structure including the rotor hub driving system with the swash plate for the blade pitch control. Thus the model was designed, manufactured, assembled and instrumented. The dimensioning was based on the loads evaluation coming from the CFD simulations carried out in the frame of this project.
The model could be installed on the wind tunnel pylon in both upside-down and upright mode. The upside-down tests allowed to study the effect of devices layed on the fuselage lower surface (as the vortex generators) avoiding pylon disturbances. On the other hand the upright tests allowed to evaluate the performance of the helicopter including the rotor-hub (with blade stubs).
Furthermore the optimized components (vortex generators, gear sponsons, rotor hub caps and blade root fairings) have been completely designed and manufactured.
A comprehensive wind tunnel campaign was performed to assess the aerodynamic effectiveness of CFD-based optimized helicopter components for drag reduction. Several configurations at different angle of attack and side-slip have been tested to assess the performances. Moreover, a purposely designed and developed setup was used to perform several stereo PIV surveys in the region just after the back ramp and in the region just before the fin. The acquired flow fields (as well as the measured steady and unsteady pressures) were quite useful in the physical interpretation of the results.
A CFD activity supported all the design and experimental activity and was particularly important for the evaluation of the wind tunnel effects corrections.
Project Context and Objectives:
Due to the expected growth of the rotorcraft traffic for passenger transport, the rotorcraft contribution to environmental impact, negligible today, would become more significant in next decade unless a major initiative succeeds in keeping it under control. As the power required to fly a helicopter in forward flight (and consequently the fuel burn and CO2 emissions) is strongly dependent on the aerodynamic drag of its airframe and of non-lifting rotating parts, the design of this parts needs to be optimized with the aim of the drag reduction.
The ROD project pertains to the wind tunnel testing of a common helicopter platform of the heavy-weight class. The aim of the testing activity is the evaluation of the effectiveness (mainly in terms of drag reduction) of the shape optimisation performed by GRC2 consortium on several components. Both original and optimised models were tested to assess the optimisation effectiveness by comparison. The tests was complemented by a CFD simulation activity for the evaluation of the wind tunnel corrections and finally a physical interpretation of the obtained results was presented.
The project was essentially based on a wind tunnel activity split into two different test campaigns: the first one regarding the original helicopter configuration and the second one regarding the version optimized by GRC2. The differences between the performances with the two versions allowed to estimate the effectiveness of the optimization. Furthermore flow-field and pressure surveys were carried out to outline some physical interpretation of the results producing more generally usable information. The tests results were corrected for wind tunnel effects using direct measurements and the results of CFD simulations.
Project Results:
The tests with the model in upside-down configuration were addressed to evaluate the effect of the VG array optimised by ONERA simulations avoiding the interference of the strut wake. The drag measurements carried out with the model at cruise angle of attack showed that the smaller counter-rotating VG array provides the higher drag reduction (see deliverable D3.2). The best VG configuration (small counter-rotating) was therefore selected to be subject of a detailed investigation. In particular, the functioning of the best VG array was investigated in the α-sweep range. The experimental tests results show an appreciable reduction of drag due to VG in the range -4° < α < 10° with highest drag reduction obtained for α = 2°. The comparison of CFD simulations by ONERA and wind tunnel results shows a very good agreement between them in terms of the drag coefficient difference evaluated with the best VG, with only exception of the lowest angle of attack considered for the CFD simulation (see deliverable D5.1).
Static pressure measurements and PIV surveys carried out at the back-ramp region compared to CFD results enabled to achieve a detailed insight about the physics related to the functioning of the best VG array (smaller counter-rotating). A fundamental effect of the VG array, when located downstream the pronounced upsweep characterizing the blunt fuselage, is twofold. First in triggering a smooth increase of the adverse pressure gradient at the backdoor and second in re-energising the boundary layer and thus preventing or limiting flow separation. The second effect is clearly visible from the comparison between the velocity field measured in the back-ramp region with and without the VG. In particular, the highest benefit introduced by the small counter-rotating VG in terms of drag reduction was found for the angle of attack α = 2°, as clearly shown by both CFD and experiments.
The analysis of the flow field around the ramp region evaluated making use of PIV surveys and CFD calculations showed that for the clean geometry configuration (without the VG) the flow close to the ramp is characterised by a large separation. The extent of this separated flow region in both longitudinal and span-wise direction is clearly indicated by the negative values of the stream-wise velocity component measured on X-Z and Y-Z planes. On the other hand, PIV surveys for the geometry equipped with the VG do not show back-flow region in the volume of investigation. A general quite good agreement between CFD and PIV results can be observed on the Y -Z planes, even if time-averaged PIV results show a more symmetrical structure of the separated region with respect to CFD. The flow topology at the rear-ramp being very sensitive to pressure fluctuations, it is numerically experienced for this geometry that after a transient phase the flow symmetry at the rear-ramp is lost once the flow starts separating. For the model geometry equipped with the VG, CFD results show that the flow is still slightly separated close to the backdoor/tail boom junction. Moreover, the comparison of the vertical velocity component shows that this VG configuration was not designed to control the longitudinal vortex structures issued by the blunt fuselage.
Static pressure measurements confirm the potential benefit of VG in reducing drag by increasing the pressure field and therefore by limiting the suction effect responsible for pressure drag. Very small effects of the VG arrays are visible for both the experiments and CFD in the region before the VG arrays. On the other hand, on the backdoor surface downstream the VG array, a clear increase of the pressure can be observed from both experiments and CFD where the flow-field is further decelerated. Note that the pressure rise between the clean geometry and VG is quite similar in the experiments and in the simulations.
The unsteady pressure measurements carried out over the back-ramp surface provide a comparison of the unsteadiness level of the flow in this region between the clean geometry case and the model configuration with the VG. Coherently with the PIV velocity field surveys, the measurements carried out by Kulite transducers located on the mid-span plane of the back-ramp downstream the VG array with the clean geometry show a rather higher level of pressure fluctuations, confirming the higher level of unsteadiness that characterise the separated flow region without the VG.
The tests with the model in upright configuration with the rotating hub were mainly addressed to evaluate the performance of the different optimised hub configuration in terms of drag reduction. In order to obtain an accurate estimate of the contribution to the aerodynamic performance, the tests were performed by adding all the optimised components starting from the original to the final optimised configuration. The first hub-cap and fairings provided by ONERA are respectively called in the following small hub-cap and stubs fairings. The second hub-cap tested was manufactured starting from the external optimised shape of a full hub fairing optimised by DLR. Differently from the original design by DLR, this component presents a large open underside necessary to avoid interference with the actual rotor hub. In the following this hub-cap is called large hub-cap. DLR provided also the shape of the optimised set of sponsons (namely new sponsons).
The best performance between the hub-caps was found with the small hub-cap. The blade stubs attachments fairings produce a small increase of the performance with the small hub cap, while they produce a contained decrease of the large hub-cap. Therefore, the wind tunnel activity showed that an overall maximum drag reduction of 6.1% can be obtained at cruise attitude with respect to the original geometry.
PIV surveys were carried out to investigate the possible tail-shake effect of the different optimised hub-caps due to the rotor hub wake. Thus, the stereo PIV surveys were carried out in a measurement volume just before the fin. A quantitative evaluation on the extent of the rotor hub wake was obtained by the comparison of the streamwise velocity component profiles extracted from the measurement volume at model mid-span closest to the fin. PIV results for the original hub-cap configuration show that the velocity defect region is confined in the lower part of the measurement volume close to the tail boom. Thus, the rotor hub wake influences only the lowest part of the fin. On the other hand, a wider velocity defect region can be observed from the PIV results obtained with both the large and small optimised hub-caps. In particular, the area with the higher velocity defect is more extended for the large hub-cap configuration. This means that the optimised hub caps do not deflect the wake enough to avoid the collision with the fin. Moreover, the blade stubs attachments fairings do not produce appreciable effects on the rotor hub wake.
The unsteady pressure measurements carried out on the fin provide interesting information about the unsteadiness of the rotor hub wake for the different hub caps tested. The pressure RMS are higher on the port side corresponding to the upper side of the fin airfoils that is more sensible to the instantaneous incidences variations. In particular, the RMS of the pressure signals measured by the highest Kulite transducers KF1 and KF2 is lower for the original hub cap configuration with respect to both the small and large hub cap. The RMS value of the pressure signals measured without the hub cap are similar to the ones obtained with the optimised hub caps. The RMS values confirm that the flow impinging the higher part of the fin presents an higher level of unsteadiness for the rotor hub configuration with the large hub cap. A further confirmation of this feature is given by the highest amplitude of the spectrum peak corresponding to the rotor 4-per-rev frequency obtained from the KF1 and KF2 transducers for the large hub cap configuration. A similar level of pressure fluctuation was observed from the measurements carried out by the lower KF3 transducer for the three hub cap configurations tested and for the configuration without the hub cap. The lowest transducer KF4 shows an higher value of the signal RMS for the original hub cap configuration.
Potential Impact:
The present project is in the frame of CleanSky Project and namely in the frame of Green Rotorcraft (GRC) Integrated Technology Demonstrators. The aim of GRC is to develop new technologies for the reduction of helicopters noise and emissions. The remarkable reduction in the drag is a significant indication about the direction to follow in order to build less polluting helicopters. Furthermore all the data acquired and the physical observations can be very useful for helicopter designing, including some safety items (as the problem of the tail shaking).
The Project web page is on-line since the 26 March 2013. Due to the nature of single participant project this site is more devoted to dissemination than to project management.
Up to now, the following publications have been produced/prepared:
1. G. Gibertini, F. Auteri, G. Campanardi, G. Droandi, D. Grassi, A. Le Pape and A. Zanotti. A Test Rig to Assess the Effectiveness of Drag Reduction Devices on a Heavy-Class Helicopter. 41th ERF, September 1–4, 2015, Munich, Germany.
2. G. Gibertini, J.C. Boniface, A. Zanotti, G. Droandi, F. Auteri, R. Gaveriaux and A. Le Pape. Helicopter drag reduction by vortex generators. Aerospace Science and Technology, Vol. 47, pp. 324-339, 2015.
3. G. Gibertini. Drag and Drop. To appear in Aerospace Testing International.
4. A. Zanotti, G. Droandi, G. Gibertini, F. Auteri, J.C. Boniface, R. Gaveriaux and A. Le Pape. Wind Tunnel Assessment of the Computational Framework for Helicopter Fuselage Drag Reduction Using Vortex Generators. Abstract submitted for the AHS 72nd Annual Forum, West Palm Beach, Florida, USA, May 17–19, 2016.
5. A. Zanotti, G. Droandi, G. Gibertini, D. Grassi, G. Campanardi, F. Auteri, A. Aceti and A. Le Pape. Wind Tunnel Tests of a Heavy-Class Helicopter Optimised for Drag Reduction. Submitted to Aeronautical Journal.
6. A. Zanotti, G. Droandi, F. Auteri, G. Gibertini and A. Le Pape. Robustness and Limits of Vortex Generators Effectiveness in Helicopter Drag Reduction. Submitted to Journal of the American Helicopter Society.
Furthermore, taking into account the problem of confidentiality, the test activity has been shown to the graduated students of Aeronautical Engineering of Politecnico di Milano.
List of Websites:
www.aero.polimi.it/rodproject