High Airspeed Testing of the Sikorsky X2 Technology TM Demonstrator ABSTRACT

Transcription

1 High Airspeed Testing of the Sikorsky X2 Technology TM Demonstrator D. Walsh, S. Weiner, K. Arifian, T. Lawrence, M. Wilson, T. Millott and R. Blackwell Sikorsky Aircraft Corporation Stratford, CT 665 ABSTRACT The Sikorsky X2 Technology TM demonstrator aircraft is nearing completion of its flight test evaluation. The demonstrator is a 6-lb-class vehicle powered by a single LHTEC T8-LHT-8 engine. It employs a fly by wire control system, counter-rotating four-bladed stiff-inplane main rotors and a six-bladed auxiliary propulsor to achieve cruise speeds in the 25 knot range. The demonstrator is designed to combine helicopter operational strengths such as hover efficiency, low speed maneuverability and autorotation capability with expanded hot day high altitude hover performance, high speed maneuverability and cruise speed capability. To date the fully instrumented aircraft has flown to 25 knots true airspeed and has met its design requirements in the areas of aerodynamic performance, flying qualities, acoustics and vibration. Rotor and control system loads and tip clearance have remained within limits without any requirements for modification. A brief overview of the demonstrator aircraft and a recap of the hover/low speed data presented in a previous paper (Ref. ) is provided. The high speed forward flight envelope expansion process is described. Results in the areas of flying qualities, vibration, performance, structures and stability of the main rotors and propeller are presented. INTRODUCTION The X2 Technology demonstrator (X2TD) uses a suite of advanced technologies to provide 25 knot cruise speed capability with low vibration, low noise and reduced pilot workload. These advanced capabilities are achieved in ways that do not compromise conventional helicopter attributes such as excellent hover performance, low speed maneuverability and autorotation. The test program for this aircraft has been constructed to provide an incremental development path for the components and systems needed to provide this unique combination of attributes. The unique character of some of the technologies of the X2TD required that new test methods be devised to ensure that the resulting vehicle possessed levels of reliability and safety required of an experimental aircraft. TEST PROGRAM The X2 Technology path to 25 knots is outlined in Fig.. The development program included an extensive ground test effort followed by envelope expansion flights broken into 4 phases. A major aircraft update was included following Phase I of the flight testing. Envelope expansion * Director, Ground and Flight Test; dwalsh@sikorsky.com. Presented at the 67th Annual Forum of the American Helicopter Society International, Virginia Beach, VA, May 4, 2. Copyright 2 Sikorsky Aircraft Corporation, published with permission. to 25+ knots was accomplished in 7 flights within a 2 year period following st flight. ENVELOPE EXPANSION TO 25 KTS Ground Runs Hover Low Speed Phase I Hover to 4 kts X2 TD st 2 nd 3 rd Flight Update: Hub, Fairings, SAS & Prop Integration, Landing Gear Retraction Prop Ground Run 4 th Flight Prop & SAS Validation Flights Move To WPB Phase II 4 to 2 kts 5 th, 6 th, 7 th 8 th, 9 th Flight Helicopter Mode th, th, 2 th Flight Prop Speed Transitions Phase III 2 to 8 kts 3 th, 4 th, 5 th 6 th, 7 th Flight Figure. Phases of X2TD envelope expansion testing. Speed Dash Phase IV 8 to 25 kts The Key Performance Parameters (KPPs) for the program were to exceed an airspeed of 25 knots and demonstrate low vibration, low pilot workload and low noise. The technologies needed to meet these goals were investigated in logical steps as the phased flight program progressed. The test program uses Sikorsky s extensive experience in design, test and evaluation of new concepts, which has, over the years, culminated in a standard method of testing

2 by the Sikorsky Test and Evaluation branch. The X2 Technology program, like others at Sikorsky, has a specific group of test engineers dedicated solely to completing the development testing on the aircraft. In addition to the experience-based test methods, the unique character of some of the technologies of the demonstrator required that new test methods be devised to ensure that the resulting vehicle possessed the required levels of reliability and safety for an experimental aircraft. Ground Testing The test program began with extensive component level ground testing, including a main rotor blade fatigue test, main gearbox no load lubrication test, Fly-By-Wire (FBW) hot bench development/integration and a main rotor servo temperature/vibration safety of flight qualification test, to name a few. Figure 2. X2TD on the tiedown pad for bare-head runs. The testing then progressed to an aircraft level tie down ground test where full main rotor and pusher prop thrust plus substantial main rotor control moments were explored. Initial engine starting and control testing were conducted with the engine driving into the combiner portion of the main gearbox. These runs completed the integration of the digital engine control unit with the X2TD FBW flight control system. The basic engine integration runs were followed by extensive bare head (no blades) testing of the aircraft (Fig. 2) focused on the main transmission, lubrication, cooling, rotating component instrumentation and basic electrical system integration. Initial blades-on ground runs were completed with the blades fixed at flat pitch, allowing the team to evaluate the new blades without the FBW system operating (Fig. 3). The FBW main rotor servos were then mechanically installed and rigging checks completed. Initial ground runs verified proper operation of the FBW control system and gathered data for main rotor track and balance. The operating envelope was incrementally expanded to include full main rotor thrust plus cyclic, pedal and differential pilot inputs and power-on operation from 85% to 5% rotor speed. A 25 hour pre flight endurance test followed this envelope expansion, establishing gear patterns for the drive system and confirming that no infant mortality issues remained. Following a detailed visual inspection, the aircraft entered the flight test phase of the program. The complete ground test program is detailed in Ref.. Figure 3. X2TD on tiedown pad for main rotor blades-on testing. Flight Test Phase I Phase I of the flight test program was conducted at the Sikorsky/Schweizer facility in Elmira, NY. Continuing the incremental build up philosophy, the propeller driveshaft and the main rotor fairings were not installed and the main landing gear was fixed in the down position for the st three flights. Additionally, the rate damping portion of the FBW control system was not active for these initial flights. The first three flights of Phase I were conducted without the prop turning. st flight occurred on August 27, 28 (Fig. 4). These initial flights expanded the flight envelope to include takeoff and landing to/from hover IGE/OGE, sideward flight/rearward flight, hover turns and forward air taxi to 2 knots with 2 degree angle of bank turns. Several adjustments to the FBW control sensitivity were made between flights resulting in stable aircraft with low pilot workload. Further details of the st three flights are included in Ref.. 2

3 the prop not operating. Acceleration and deceleration profiles using the prop were developed. The limit cycle roll oscillation observed during Phase I was eliminated with the adjusted gains provided from simulation. Flight 6 was the first out of the yard flight, with the aircraft flying an extended pattern. The initial part of the takeoff was completed using main rotor thrust only. As the aircraft accelerated, prop pitch was increased resulting in a levelbody acceleration to a rotation speed of 6-7 knots. Climb was completed between 65 and 75 knots with a nominal climb rate of 3 fpm. Figure 4. X2TD undergoing initial flight testing. A major aircraft update was completed following the first three flights. Modifications included completion of the propeller drive and FBW control system installation of the landing gear retraction system installation of the MR hub and pylon fairings improvements to the main gearbox lubrication system activation of the Stability Augmentation System (SAS) rate damping plus other FBW software updates based on Phase I flight test data and simulation Following the update a ground run checkout of the modified systems was performed on the tie-down pad. A hour endurance run was completed prior to flight 4. Flight 4 was the st flight with the propeller operating. Pitch and roll handling qualities were not significantly affected by the addition of spinning propeller, but the yaw trim position required significantly more left pedal. The SAS reduced pilot workload significantly. The initial roll SAS gains were too high resulting in a small-amplitude.5 Hz limit cycle oscillation. SAS data from this flight were used to update the simulation model and new gains were determined for the following flights. The flight 4 testing included hover, side flight to knots, forward air taxi to 4 knots with the prop at thrust and forward air taxi to 2 knots using the prop. At each point control steps and pulses were used to evaluate handling qualities and performance of the SAS. The aircraft was then moved to Sikorsky s Development Flight Center (DFC) in West Palm Beach, Florida for the high speed envelope expansion. Flight Test Phase II The initial flights in Phase II expanded the forward air taxi envelope to 6 knots using the prop and to 73 knots with Subsequent Phase II flights expanded the forward speed envelope to 3 knots. FBW maps for propeller thrust, differential collective (yaw) washout and differential lateral cyclic were activated and optimized. The differential collective washout schedule was adjusted, finally settling on a linear washout between low and high speed. Rudder control travel was increased by 3% to improve aircraft yaw control in the 6 to knot speed range. The differential lateral cyclic map was also adjusted to provide desired main rotor blade tip clearance versus airspeed. Following each flight handling qualities and flight control data were reviewed and the simulation model updated. Improvements to the flight control system were made between each of the flights in Phase II. The main rotor control phase angle was changed, resulting in purer pitch and roll responses to longitudinal and lateral stick inputs and increased control margins. The SAS authority was increased and several adjustments were made to the pitch and roll gains. The landing gear was retracted for the first time on flight 8 resulting in no noticeable change in aircraft handling qualities with the gear retracted. Flight 8 also marked the first use of the Active Vibration Control (AVC) system. At the end of Phase II, the speed envelope was expanded to 3 knots, aircraft handling qualities were good with the pilot able to perform many maneuvers hands off the controls and overall vibration levels were reasonable. Flight Test Phase III The pitch and roll SAS gains required continued optimization as the speeds increased during Phase III envelope expansion. Gain schedules versus airspeed were adjusted to maintain low pilot workload and acceptable dynamic stability. Also during Phase III the climb profile was adjusted to provide better main gearbox cooling, the AVC software was adjusted for improved 4P vibration reduction and the main rotor track was adjusted to help reduce the P and 2P levels. Operations at reduced main rotor speeds were evaluated by beeping the rotor down to %. This evaluation was in preparation for flights at speeds above 2 knots where the rotor speed is decreased 3

4 to maintain an advancing tip Mach number of.9. The speed envelope was expanded to 8 KTAS on flight 2. phase margins. Additional FBW software updates were made to extend the SAS gain scheduling to 25+ knots and to update the differential lateral cyclic versus airspeed map. While the SAS gain adjustments improved dynamic stability the aircraft exhibited reduced pitch axis static stability at speeds above 6 knots. Low frequency longitudinal stick inputs of between 5 and % were required to maintain the desired pitch attitude. It was obvious that modifications to improve pitch static stability would have to be implemented. The airspeed envelope was expanded to 225 KTAS on flight 4. The combination of the secondary tail and the modified pitch SAS logic resulted in excellent dynamic and static pitch stability. The pilot was able to fly the aircraft hands off for extended periods of time. Figure 6 compares the longitudinal stick motions for high speed test points before and after the SAS improvements were made and the secondary horizontal tail was installed. In a period of six weeks following flight 2 a secondary horizontal tail was conceived, designed, manufactured, installed on the aircraft and checked during light on the wheels ground runs. The simulator was used to determine the secondary tail area (7 ft2) required to provide adequate static stability out to 25+ knots. The new tail, installed near the bottom of the vertical tail, is shown in Fig. 5. (a) Figure 5. Secondary horizontal tail. The secondary horizontal tail was evaluated on flight 2b (there was no flight 3). A complete set of pilot input pulses, doublets and frequency sweeps were obtained at 4-5 knots. Pitch static stability was significantly improved. Aircraft response to pilot inputs was generally deadbeat. Trim attitudes and power were unchanged from previous flights at this speed. Forward cyclic required during low speed transition was also unchanged. Turns to 3 degrees angle of bank were demonstrated at 5 knots. This flight brought Phase III to a close. (b) Figure 6. Longitudinal stick motions (a) before and (b) after improvements to SAS and addition of secondary horizontal tail. Flight Test Phase IV Review of data from Phase III showed that, at higher airspeeds, the required pitch rate damping could not be obtained while maintaining adequate pitch SAS gain and phase margins. The basic structure of the pitch SAS was changed from a proportional plus lag rate logic to a lead/lag compensator type of logic. Simulation showed much improved SAS performance with good gain and A level flight speed sweep from 4 to 22 knots was completed. Aircraft performance was evaluated at reduced Nr s, down to %. Additionally, the roll stick shaping was modified to improve the roll jitter observed on flight 4. Flight 5 4

5 expanded the airspeed envelope to 235 knots. The roll stick shaping improved the roll axis performance and pitch stability continued to be good at these higher speeds. X2 TECHNOLOGY DEMONSTRATOR Test Summary With the FBW/SAS optimized and overall vibration in the cockpit, on the engine and at equipment locations at satisfactory levels, it was time to continue the envelope expansion toward the goal of 25 KTAS. Flight 7 was conducted in the early morning of September 5, 2. The takeoff was the most aggressive to date with the pilot accelerating to 5 knots in a level attitude before rotating into a cruise climb at 4 knots and 8 fpm. Level off was at ~7 ft. Hd. Prop pitch was increased to trim the aircraft at 2 knots, 22 knots and 23 knots. The prop pitch was then increased a few degrees while the collective was lowered to ~% above the flat pitch setting. The aircraft trimmed at 253 KTAS in level flight. The chase aircraft showed a true airspeed of 255 knots. After taking data at 25+ knots the pitch attitude was lowered by 3-4 degrees, with the same prop and collective settings, and the aircraft entered an 8 fpm descent. Speed increased to 263 KTAS. After recovery from the dive, the aircraft was slowed to a very comfortable 2 knots. A three perpendicular-leg GPS airspeed calibration was completed at 2 knots. The aircraft returned to base for an uneventful landing. 8.3 hours Total Run Time 3.2 hours Blades On 96. hours FBW Active 67. hours Prop Operating 2.6 hours AVC Operating 255 Engine Starts 2 Flights Totaling 8.6 hours 25 Knots achieved on Sept 5 th Figure 7. X2 test program summary. PILOTED SIMULATION One of the major contributors to the success of the X2 program was the use of a fixed base piloted simulator. A Sikorsky GenHel flight dynamics model of the XH-59A was initially used as a baseline since flight test data for correlation was available. From a handling qualities perspective, the initial focus was on matching the transition pitch-up phenomena so that the aircraft could subsequently be evaluated in this critical regime. After validation of the model and duplication of the original transition flight dynamics, an X2 GenHel model was created. To assure the usefulness of the simulator results a duplicate of the aircraft FBW cyclic controller was installed in the simulator cockpit along with cockpit displays that replicated those of the aircraft. Additional Flight Testing Another Key Performance Parameter for the program was to demonstrate the aircraft s low acoustic signature. Data for the low speed helicopter mode flying was obtained by removing one section of the prop driveshaft. This simulated the normal operation for an X2TM type of aircraft equipped with a prop drive clutch. The aircraft was extremely quiet operating in this mode. Additional data was obtained with the prop turning and showed that the noise levels at 2 knots were the same as similar size and weight production helicopters flying at knots. Initial simulation sessions were devoted to pilot familiarization with the cockpit controls and with the expected flight characteristics of the vehicle. Once flight testing started, the simulation model was compared to the test data and updated to better reflect the actual vehicle. During development a number of typical problems arose including FBW controller sensitivity, oscillations in pitch and roll, low directional control power and insufficient pitch stability at high speed. The simulator became very valuable in resolving these issues as stick sensitivity, control law gains, and airframe modifications were all evaluated before being introduced to the aircraft. In particular, several different solutions could be flown to select the best one for flight testing, saving significant amounts of test time and resources. Having established a comfortable cruise envelope, several demonstration flights were completed at Sikorsky s Development Flight Center. These flights speak volumes to the maturity of this prototype aircraft after only 7 engineering development flights. A summary of the ground and flight test program is presented in Fig. 7. Another valuable aspect of the simulator was the ability to assess different failure modes, both in terms of aircraft response and cockpit annunciations and in terms of pilot response. This allowed for the development and practice of emergency safety procedures. It also provided more 5

6 confidence to the pilot as he knew what actions to take during various failures. While initial simulations were conducted using generic visual databases, an upgrade to the West Palm Beach Development Flight Center database was installed. This proved to be immensely valuable as the piloted simulator sessions allowed the pilot to select landmarks for turns and evaluate different flight paths for different tests. By the end of the program the simulator was being used extensively for training and to maintain pilot proficiency between the flights. AIRCRAFT PERFORMANCE Figure 8. Comparison of L/De predicted for the X2TD with XH-59A and S-92 test data. By using two rigid, counter-rotating, coaxial sets of rotor blades, and adjusting the retreating-side blade pitch to minimize drag, the X2TD eliminates the dissymmetry of lift constraint imposed on conventional helicopters. The design also eliminates the need for a tail rotor, since the counter-rotating rotors provide their own torque reaction. This leaves the power that would normally be used for the anti-torque tail rotor to be used for lift from the main rotors in low speed flight, or for thrust from the propeller in high speed flight. In addition to the design of the X2TD blades, the integration of the rotor system into the overall vehicle was carefully considered in order to meet the overall performance goals for the aircraft. The high rigidity of the blades allows the two rotors to be placed closer together without the resultant risk of collision between blades at high speed and during maneuvering. This close stacking reduces overall drag. The integrated hub and mast fairings, made possible by the limited degree of individual blade movement, gives the X2TD a practical, low-drag, coaxial rotor system implementation. In the past, these issues have created the main performance limitation of other stackedrotor configurations for helicopters with conventional, articulated blades. The X2TD rotor blade design incorporates unique planform, twist and airfoil combinations in a stiff, allcomposite blade. These blades are rigidly attached to the large diameter rotor hubs/shafts, and only have a single degree of freedom (pitch) to adjust blade angle of attack. Flapping and lead-lag forces are reacted by small deflections of the blade structure itself eliminating the need for blade dampers, mechanically simplifying the rotor hub. The overall performance of the aircraft exceeded expectations. In fact, the primary goal of 25 KTAS in cruise level flight was exceeded without use of the interrotor shaft fairing system. Figure 9 compares the main rotor and aux prop power required in relation to power available versus true airspeed. In high speed flight, the rotor system works more efficiently with the discs inclined slightly nose up relative to the free stream. Therefore, above 2 knots, the aircraft was generally flown at a positive pitch attitude from 2-5 degrees. This allowed the rotor to absorb relatively little power, yet maintain the ability to lift the weight of the aircraft. Most of the power generated was then used to drive the auxiliary propulsor for the thrust required to overcome the drag on the vehicle at high speed. The unique design of the X2TD rotor blades produces a lift to drag ratio of the rotor system that is twice that of a conventional rotor at its most efficient speed, and, at 25 knots, is double that achieved on the XH-59A. Figure 8 shows the preflight prediction for rotor lift to drag ratio for the X2TD compared to a conventional single rotor helicopter (S-92 ) and flight data from the XH-59A. Three key points are apparent. First, at 25 knots, the X2TD doubled the L/De of the XH-59A. Secondly, the peak L/De of the X2TD rotor system, is approximately double that of a conventional helicopter, and occurs at a true airspeed of 2-22 knots, nearly knots greater than that of the conventional helicopter. The excellent performance of these patented blades is the basis for the increased operational envelope of the X2TD. 6

7 Power Required* at 4K Density Altitude 2 Tip Clearance vs TAS for Flights 9-7 Total Power Required (Test Data) 8 Prop Power Required (Test Data) Power Available (MCP) Rotor Power Required (Test Data) 6 25 Predicted Installed Power Available 4K ISA 4 2 T ip Clearance (in) 2 SHP 8 Total Power Propeller Power Flight Test Observed Tip Clearance Limit Rotor Power True Airspeed (KTS) Figure 9. Power required vs. power available (Test data normalized to 4K ISA) True Airspeed (KTS) Figure. Tip clearance vs. KTAS during level flight. A unique feature of this rotor system is the ability to control the lateral lift offset on each rotor. As the speed increases, the center of lift acting on each rotor migrates laterally outboard to the advancing side of each disk. Left unchecked, this would result in high /rev structural loads and blade deflections. However, the X2TD employs differential lateral cyclic control to mitigate these consequences of the rigid rotor system. The trade-off, though, is increasing drag on the rotor system with increasing lift demand from the retreating side of each disk. As such, it was necessary to develop a schedule of differential lateral cyclic to be employed with increasing airspeed to maintain adequate blade structural margin and tip clearance, while imposing minimal drag penalty. One unique feature of the coaxial rotor configuration is the de-swirl effect created by the lower rotor which serves to straighten the incoming flow from the upper rotor and provide improved rotor thrust even at high overall thrust levels. Figure 2 illustrates the aircraft response to a vertical climb maneuver initiated from a hover (prop disengaged). An approximate 9% increase in collective pitch at the rotor and resultant 5% increase in engine power results in a vertical climb rate of nearly, feet per minute. X2TD - VROC Performance (Prop Disengaged) Collective Position (%) Engine Torque (%) Rate of Climb (fpm) Rate of Climb (fpm) Collective Position and Engine Torque (%) Figure shows the lateral lift offset with increasing airspeed during flight test. The resulting tip clearance during level flight is given in Fig.. A tip clearance limit of inches was observed during the development test program to ensure margin during initial envelope expansion. This control schedule allowed the X2TD to achieve good rotor efficiency at high speed while more than safely managing structural and tip clearance limits. 45 Time (sec) Lateral Lift Offset vs TAS for Flights 9-7 Figure 2. Vertical rate of climb performance from hover (Prop not powered). 25 Lateral Lift Offset (%) 4 2 DYNAMICS 5 Design characteristics The combination of the X2TD s 25 knot performance, operation at varying rotor speeds, and very rigid main rotor blades presented unique challenges in the dynamics area. The choice of main rotor blade natural frequencies, number of blades per rotor, the crossover positions of upper and lower blades and the design of the active vibration control system all impacted the achievement of low vibration and True Airspeed (kts) Figure. Lateral lift offset achieved during level flight vs. airspeed. 7

8 avoidance of aeroelastic response issues. An additional challenge was to incorporate the 6-bladed pusher prop ensuring that it remained stable and did not experience or cause high vibratory loads. The auxiliary propulsor provided by AeroComposites Inc (which was adapted from fixed wing applications) performed without issue throughout the program. blade flapwise or torsion response, fuselage motion or the automatic flight control system was noted at any test condition. The rotor was extensively instrumented to capture blade flapwise and chordwise bending and torsion moments at several radial stations and moments applied to the hubs. An interesting result was how little the blade moments varied with airspeed. Figure 5 shows upper rotor root flapwise bending moment and pushrod load time histories for airspeeds from 2 to 25 knots. With the exception of increasing /rev flap bending as the rotor carried more lift on the advancing blade, the blade moments do not change significantly with airspeed. Insofar as the upper and lower rotors are not each required to operate with zero roll moment, the increased control loads associated with retreating blade stall are avoided and no significant increases in pushrod or servo loads were seen as airspeed was increased. Added to those design considerations are conventional issues such as design of the fuselage to avoid amplification of vibration, engine-drive system integration, design of the main rotor pylon to avoid tail buffet, design of the empennage to avoid excessive response to main rotor n/rev wake impingement and design of a rudder system which remains stable at high forward speeds. Fortunately none of these more conventional areas turned out to cause problems during envelope expansion. For example the engine-drive-fuel control system operated without issues, no tail buffet was ever experienced and no structural limits were approached on the empennage. 4/rev hub loads The predecessor XH-59A which flew in the 97 s and 98 s was free of aeroelastic stability issues but, as discussed in Ref. 2, that aircraft which had three blades per rotor and no vibration control system experienced vibration well above desirable levels. Review of blade harmonic loads and fuselage vibration data from the XH-59A made it clear that high levels of 2/rev blade root flap moments were the dominant sources of the 3/rev vibration. Although the active vibration control (AVC) system planned for the X2TD was expected to significantly attenuate vibration, it was not clear that the program goal of less than of.3 ips of cockpit n/rev vibration at the 25 kt cruise condition could be met by an AVC system of reasonable weight with 3-bladed rotors and as a result (as discussed in Refs. 3 and 4) the decision was made to use four-blade rotors. Bending moments measured at the roots of the blades and on the two shafts allowed the vibratory moments applied to the fuselage to be tracked during envelope expansion and the trends compared with trends of fuselage 4/rev vibration. Examination showed that the 4/rev blade root moments are higher than the 3 and 5/rev moments, but none of the loads build up dramatically with airspeed. By comparison, the 2/rev root flap moments, which would contribute to 3/rev vibration on a 3-bladed rotor, were significantly larger than the 3/rev moments so the decision to use four rather than three blades per rotor appears to have been the correct one. It was demonstrated during testing of the XH-59A (Ref. 2) that the azimuthal positioning of the upper and lower rotors affects the degree of addition or cancellation of n/rev hub loads applied by the two rotors to the fuselage. It was also apparent from that aircraft that the 3/rev roll and pitch moments were the dominant sources of fuselage vibration. Indexing of the two rotors such that each has a blade pointing in the downstream position at the same time promotes cancellation of the n/rev roll and yaw moments and side forces while causing the n/rev vertical and longitudinal forces and pitch moments of the two rotors to add. Positioning of two 4-bladed rotors such that blades cross at the 22.5 deg azimuth position promotes cancelation of 4/rev vertical and longitudinal forces and pitch moments and allows the 4/rev roll and yaw moments and side forces of the two rotors to reinforce one another. For the X2TD, the decision was made to index the rotors to promote cancellation of the roll, yaw and lateral inputs (although the option was present to change that during the flight program) because it was believed that the positions in the fuselage which were available for placing the active Blade Natural Frequencies and Response Characteristics The natural frequencies of the blades, which were confirmed by rap tests, are shown in Fig. 3. The first flapwise and first chordwise mode frequencies are in the.4 to.6/rev range for the operational rotor speed range. The second flap mode is placed between 4 and 5/rev and, as a result of the inherent stiffness of the inboard portion of the blade, the first torsion mode frequency is above 9/rev. As a stiff-inplane design, the X2TD rotor is not subject to ground or air resonance. The only aeroelastic response of note was the transient chordwise response of the rotor following sharp control inputs which were applied to track stability of the system during envelope expansion. Figure 4 provides a typical example. Damping ratios remained in the 3 percent of critical damping range throughout the flight envelope. No coupling of chordwise response with 8

9 vibration control system force generators would be more effective in dealing with 4/rev pitching moments than they would be for controlling 4/rev roll moments. Measured vibratory moments generally confirmed that net 4/rev roll moments were mitigated by the indexing choice. Figure 6 shows upper, lower and total 4/rev roll and pitch moments (normalized by the maximum value of upper rotor 4/rev roll moment) versus airspeed. At high airspeeds the total roll moment is about half of that applied by either rotor while the total pitch moment is about equal to the sum of the pitch moments from the two rotors. Long. Stick Position, % Upper Rotor Root Edgewise Bending, The vibratory forces from the two rotors were low enough that reasonably good 4/rev vibration (generally less than.5 in/sec) was achieved for airspeeds up to 75 knots without any vibration control. Once the aircraft had expanded the envelope to 75 kts the AVC system discussed below was activated so that low vibration was maintained up to the 25 knot cruise speed. The trends of 4/rev moments shown in Fig. 6 suggest that the vibratory hub loads present at 25 knots are not significantly higher than those present at 75 knots, a fact which bodes well for achieving low vibration on an aircraft with an X2 rotor system using an AVC system of modest weight Upper Rotor Root Flatwise Bending Upper Rotor Pushrod Load Aircraft Pitch Rate Figure 3. Main rotor blade natural frequencies vs. percent rotor speed Time, sec Figure 4. Transient response to longitudinal stick pulses. 9

10 Blade Root Normal Bending Pushrod Load Blade Root Normal Bending KTAS Pushrod Load 6 KTAS lb 2 KTAS 2 KTAS Figure 5. Upper rotor blade root normal bending and pushrod load for four airspeeds. UPPER ROTOR 4P ROLL MOMENT LOWER ROTOR 4P ROLL MOMENT TOTAL 4P ROLL MOMENT True Airspeed, kts True Airspeed, kts LOWER ROTOR 4P PITCH MOMENT UPPER ROTOR 4P PITCH MOMENT True Airspeed, kts True Airspeed, kts TOTAL 4P PITCH MOMENT True Airspeed, kts True Airspeed, kts Figure 6. Upper rotor, lower rotor and total 4/rev normalized roll and pitch moments

11 Active Vibration Control Experience on the XH-59A showed that effective control of N/rev vibration is needed to achieve high speed flight on an aircraft with a rigid coaxial rotor system. An active vibration control (AVC) approach was selected due to its weight efficiency and its ability to track changes in rotor speed without performance degradation. The X2 AVC system is based on the off-the-shelf system utilized on the S-92ATM (Ref. 5) and the UH-6M BLACK HAWK (Ref. 6) production helicopters. AVC achieves vibration attenuation by applying 4/rev forces to the airframe to counteract those produced by the rotor system. AVC implements a closed-loop feedback control algorithm utilizing accelerometers as the feedback sensors and airframe-mounted force generators (FGs) as actuators. A tachometer (Nr) sensor is used to provide a reference signal to allow AVC to swiftly and accurately track changes in rotor speed. The architecture of the AVC system is described in Ref. 2. Vertical, Roll & Pitch Longitudinal, Lateral & Yaw Forward Figure 7. Force generator installation. The AVC system was installed and checked out in the hangar and during bare-head ground runs to verify its functionality, the structural integrity of the AVC support structure and the electromagnetic compatibility of the AVC system with critical electronic systems. Although the AVC system was cleared for use prior to envelope expansion flight testing, the system remained deactivated during the initial flights while other systems were being evaluated. Envelope expansion revealed that the uncontrolled 4/rev vibration was not excessive so the system was not activated until the aircraft achieved an airspeed of approximately 75 knots at which point the system was activated and remained on for the rest of the test program. The vibration control approach used on the X2TD differs from the approach used in the S-92A and UH-6M applications in two fundamental ways. Firstly, the X2TD does not use a hub-mounted vibration absorber to attenuate in-plane hub loads but relies completely on the AVC system to control the 4/rev vibration. Secondly, instead of distributing FGs at remote locations throughout the airframe like in the S-92A and UH-6M applications, the FGs are placed close to the locations at which the struts that support the transmission connect to the fuselage so as to cancel the 4/rev rotor excitation forces before they enter the fuselage. A trade study was conducted during preliminary design to establish the locations of the feedback accelerometers and best arrangement of FGs to provide control of 4/rev vibrations in high-speed forward flight. The AVC system achieved very significant reductions in the 4/rev vibration levels during high-speed flight. It maintained levels in the pilot compartment below the UH6M vibration specification throughout the entire airspeed envelope. Figure 8 contains plots of cockpit floor vertical, lateral and longitudinal 4/rev vibration versus airspeed. Measurements of 3/rev vibration from the XH-59A (which had three blades per rotor) are provided for comparison. As noted previously, the AVC system was not activated until airspeeds exceeded of about 75 knots. For airspeeds at which AVC was activated, the 4/rev vertical and lateral vibration levels did not exceed.5 ips out to the maximum level flight airspeed of 25 knots and the program goal of maintaining vibration below.3 ips at 25 knots was met. The result of the trade study was the FG arrangement shown in Fig. 7. Three FGs produce vertical forces (providing capability to null vertical, roll and pitch excitation from the rotors), two produce longitudinal forces and one produces lateral forces. It was anticipated that some of the force generators could be removed based on flight test results but the initial installation included the capability for nulling the three forces and three moments applied by the rotors. Four-per-rev longitudinal vibration levels reached.5 ips at 25 knots. Based on pilot comments, the AVC system placed higher weighting on vertical and lateral 4/rev vibration, so the slightly higher level of longitudinal vibration is not surprising. For reference, the 4 per-rev longitudinal vibration specifications for a modern conventional helicopter cockpit are.54 ips up to 45 knots, and.7 ips for airspeeds above 45 knots. Six accelerometers were placed in the cockpit area, two near the engine and two on the auxiliary propulsion gearbox. The AVC control algorithm allows weighting of the individual sensors to be adjusted to get best overall vibration. It s noteworthy that the 4/rev vibration shown in Fig. 8 for airspeeds above 22 knots was achieved with only

12 three FGs working to suppress the vibration. It is believed that with more time to optimize the system a vibration level below.3 ips can be achieved for airspeeds up to and beyond 25 kts with a subset of the six installed force generators. In addition to the low 4/rev vibration present in the cockpit, vibration levels on the T-8 engine, on installed electrical, avionics and hydraulic equipment and on the empennage remained within the design or qualification levels for those components with no design changes required. Acknowledgements Pilot N/rev Vertical Vibration 4. The authors would like to acknowledge the contributions of the entire X2 Team in completing the design, manufacture and testing of this aircraft. We would also like to express our admiration for Mr. Kevin Bredenbeck, the X2 pilot and Sikorsky s Chief Test Pilot, for his skill, courage and dedication to the X2 program N/rev 2. (ips) XH-59A Pure Helo Mode.5 XH-59A Aux Propulsion Mode. References X2. Walsh, D., Development Testing of the Sikorsky X2 TechnologyTM Demonstrator, American Helicopter Society 65th Annual Forum, Grapevine TX, May Airspeed, kts Abbe, J, Blackwell, R. and Jenney, D., Advancing Blade Concept (ABC)TM Dynamics, American Helicopter Society 33rd Annual Forum, Washington D.C., May 977. Pilot N/rev Lateral Vibration Blackwell, R and Millott, T., Dynamics Design Characteristics of the Sikorsky X2 TechnologyTM Demonstrator Aircraft, American Helicopter Society 64 th Annual Forum, Montreal, Canada, April 29-May, XH-59A Pure Helo Mode N/rev 2. (ips) XH-59A Aux Propulsion Mode.5 4. Bagai, A., Aerodynamic Design of the X2TM Technology Demonstrator Main Rotor Blade, American Helicopter Society 64th Annual Forum, Montreal, Canada, April 28. X Airspeed, kts Goodman, R.K. and Millott, T.A., Design, Development, and Flight Testing of the Active Vibration Control System for the Sikorsky S-92, American Helicopter Society 56th Annual Forum, Virginia Beach, VA, May 2. Pilot N/rev Longitudinal Vibration Millott, T.A., Goodman, R.K., Wong, J.K., Welsh, W.A., Correia, J.R., and Cassil, C.E., Risk Reduction Flight Test of a Pre-Production Active Vibration Control System for the UH-6M, American Helicopter Society 59th Annual Forum, Phoenix, AZ, May N/rev 2. (ips) XH-59A Pure Helo Mode.5 X Airspeed, Kts 2 25 Figure 8. Comparison of X2TD and XH-59A cockpit floor N/rev vibration levels vs airspeed. 2