Recent IBC Flight Test Results from the CH-53G Helicopter

Uwe T.P. ArnoldDirector R&D

ZF Luftfahrttechnik GmbHuwe.arnold@zf.com

Abstract•

Since December 2001, ZFL has conducted Individ-ual Blade Control (IBC) flight tests with a CH-53Gtestbed of the German Federal Armed Forces Engi-neering Center for Aircraft. The experimental IBCsystem used has been designed, manufactured,installed, certified, and tested by ZFL. This papershows a comprehensive overview of the results fromthe concluded open loop and the just started closedloop campaigns. Vibration reduction turned out to behighly successful, especially with respect to thesmall required IBC authority. Without being explicitlyoptimized with respect to amplitude or phase, IBCwas able to reduce the vibrations by more than 90%in a single axis or by more than 60% in all spatialdirections in certain cases.

The noise experiments were focused on the reduc-tion of BVI in descent flight conditions. With 2/revIBC of 0.67 deg amplitude, noise reductions of up to3 dB have been recorded. Following the trendsknown from prior wind tunnel experiments, the in-creased amplitudes, which are used during theclosed loop campaign, should allow for even higherreductions. The positive impact of IBC on rotor per-formance at high forward speed was also success-fully demonstrated. The simultaneous effect of IBCon both the rotor power and the flight condition cor-responds to net power savings of up to 6% at 130ktsforward speed. This means the application of opti-mum IBC not only reduces the rotor torque but alsotends to increase forward speed and let the rotor-craft climb. The lower frequencies primarily used toreduce noise and power required also have consid-erable effect on the control system loads. In theoptimum case encountered during the open looptests, 2/rev IBC of the right phase was able to re-duce the vibratory pitch link loads by more than30%. Fortunately, it turned out that at high speed,the IBC input for optimum rotor performance alsodecreases the pitch link loads.

The last part of the paper gives an overview of thesystem modifications which had to be implementedfor the closed-loop flight tests. It will be shown whatcontrol system architecture is available and whatalgorithms have been tested so far.

NotationAFCS Automatic Flight Control

SystemAccHGx,-y, -z

g acceleration at maingearbox in x-, y-, z-direction

AccLadx,-y, -z

g acceleration at cargo com-partment in x-, y-, z-direction

AccPilx,-y, -z

g acceleration close to pilotseat in x-, y-, z-direction

An deg n/rev IBC amplitudeBVI Blade Vortex InteractionFMEA Failure Mode Effect AnalysisG(…) = (x2+y2+z2)1/2 unweighted cost functiong m/s2 9.81, gravity constantHHC Higher Harmonic ControlIBC Individual Blade ControlJ - Cost function to be

minimizedN 6, number of bladesT1, T2 g/deg

g/deg2IBC to vibration responsetransfer matrix (linear andnonlinear)

(M)TOW (Maximum) Take-off WeightP kW PowerQ Nm Rotor TorqueVIAS kts Indicated Air Speedu deg vector of higher harmonic

control inputs (cos, sin,compon.)

Wi Weighting matrix withrespect to i

z g vector of vibrations andcontrol loads (cos, sincompon.)

∆ϑ IBC=Σ Ancos(nψ−ϕn) nominal pitch angle due toIBC

γ deg Flight path angleϕn deg n/rev IBC control phase

angleψ deg rotor azimuth angleΩ rad/s Rotor Speed

37-1

1 IntroductionZF Luftfahrttechnik GmbH (ZFL) has designed,manufactured, certified, and tested a whole family ofIndividual Blade Control (IBC) systems using hy-draulic blade root actuators. The most remarkableapplications have been a flight-worthy system usedstill today on Eurocopter's BO-105 S1 testbed, seeRef. [7], as well as two powerful experimental sys-tems used for full-scale wind tunnel tests of BO-105and UH-60 rotors at NASA Ames, see Refs. [4] and[11] respectively. These campaigns have yielded asound data base demonstrating the positive effect ofIBC on vibrations, noise, rotor power required, andloads.

Based on the success of these programs ZFL wasawarded a contract from the German Federal Officeof Defense, Technology, and Procurement (BWB) todesign, manufacture, install, qualify and flight test anIBC system for the medium weight transport heli-copter CH-53G. With respect to the rotor designparameters this program is perfectly suited to com-plement the earlier IBC activities. The six-bladedfully articulated rotor of the CH-53G features a com-parably high LOCK number but a small hinge offsetand therefore differs considerably from the rotorsinvestigated previously.

The test campaign is carried out at the GermanFederal Armed Forces Technical and AirworthinessCenter for Aircraft (WTD 61) in Manching, where thecertification authority is located, too. This facilityoperates the testbed, Fig. 1, and provides the fullspectrum of technical and personnel support for datagathering, downlink, and recording as well as opera-tion of the telemetry ground station. After the pro-gram had reached the last certification step with thefinal flight clearance, the open loop campaign hadstarted December 19, 2001 with the first IBC inputsapplied in flight.

Figure 1: IBC CH-53G testbed operated by WTD 61

In the German Armed Forces the medium lift VTOLcapacity is based on a fleet of approximately 100CH-53G helicopters. These rotorcraft now are inservice for approximately 30 years and are sched-

uled to fly (however in reduced number) for another25 years. This provides the motivation to considersuitable upgrades that can help to improve the per-formance and to contain the growing operation andmaintenance costs for these ageing aircraft. Theprimary fields addressed and positively altered bythe IBC for this aircraft are:• Component fatigue and failure induced by high

vibratory stress• Unscheduled maintenance cost due to high vi-

bration level• CRTs of dynamical components determined by

high (control system) loads• Enabling of high TOW and/or high speed opera-

tion currently prohibited with regard to compo-nent life time constrains

Other benefits that can be expected through theapplication of IBC include:• Reduction of power required or improvement of

equivalent lift-to-drag ratio at high forward speed• Significant reduction in noise radiation• Automated in-flight tracking

The experience gained from the previous and thecurrent test campaigns bolsters ZFL’s view that IBCis a practical and valuable solution. It has retrofitcapability and promises a high benefit-to-cost ratio.

2 IBC System

2.1 Basic System ArchitectureFor this IBC demonstrator program, ZFL has de-signed and manufactured all major hardware com-ponents as well as the control hard- and software.Ref. [9] gives a detailed description of the systemlayout. Key components of the IBC system are thesix servo-hydraulic actuators mounted between therotating swashplate and the blade pitch horns, com-pare Fig. 2 and 3.

HYDRAULICSLIPRING

Figure 2: Integration of the IBC-System into theCH-53G Testbed

37-2

Figure 3: IBC Actuator Mounted between Swashplateand Blade Pitch Horn Replacing the RigidPitch Rod

Hydraulic power and electrical control signals aretransferred via slip rings through the rotor shaft fromthe fixed frame to the hub. With regard to easy han-dling during the flight test operation and for minimalinterference with the testbed, most of the hydraulicand control components are located in the cargocompartment of the fuselage. The hydraulic powersupply for the IBC system is provided by two pumpsmounted on the main gearbox input shafts. Thisfollows from the requirement of complete separationof the experimental system from all basic aircraftsystems.

2.2 Servo-hydraulic ActuatorKey component of the IBC System is the servo-valvecontrolled hydraulic blade root actuator. It replacesthe conventional rigid pitch rod and imposes thedesired relative motion between rotating swashplateand blade pitch horn with high bandwidth and accu-racy. The basic specifications are given in Tab. 1.

max. piston stroke +/- 6.7 mmhard stops +/- 1.27° blade pitchusable amplitude 1.1° blade pitchmax. piston velocity@ design load

0.39 m/s

controlled frequencies 2/rev ... 7/revmax. PLL dynamic/static 18.9 kN / 28.8 kNlock-out force > +/- 14 kNbreak out force > +/- 40 kNlock-out time < 100 msactuator mass 10.2 kgsystem pressure 207 bar / 3,000 psihydraulic power available 49 kWelectrical power 0.6 kW

Table 1: System Specifications

Although the IBC System had to be sized to fit intothe helicopter without major alterations, it needed tobe powerful enough in terms of actuator amplitudesand forces to allow for full exploration of the IBCbenefits.

2.3 Safety ConceptDuring the conceptual design special emphasis wasput on the safety architecture. The IBC system hasbeen designed to show fail-safe behavior. TheFMEA has clearly proven the reliability required forcertification. Although the components required toactively introduce IBC are partly of simplex designand driven by only one hydraulic power supply, thesafety lock-out system is entirely duplex. This safetyconcept has also demanded a complete separationof the lock-out system. It is therefore mechanicallyas well as electrically separated from the rest of theIBC system. The sources used to trigger the lock-outprocedure in case certain limits are violated include(a) IBC system parameters as position errors of theactuators, etc., (b) the aircraft reactions to IBC asblade flap and lead-lag angles, pitch link loads,gearbox and tail boom accelerations, etc. or (c)manual switches operated by the flight crew, seeFig. 4.

Foundation of Safety Concept1

AutomaticRecognition ofIBC System

Malfunction orEnvelopeViolation

(Cause)

2

Monitoring ofAircraft

Reactionon IBC

and Checkfor Limit

Violations

(Effect)

3

ManualOperation of

Lock-OutTrigger byFlight Crew

(HumanJudgement)

Fast Simultaneous Reliable Actuator Lock-Out

Figure 4: Three Column Safety Philosophy

When triggered, the control system is forced to fadeout the periodic control functions and to commandall actuators into their neutral position. More impor-tant, however, by opening a set of redundant valvesin the hydraulic manifold a rapid depressurizing ofthe hydraulic system is assured. At all six actuatorsthis leads to the release of the two independentsafety locking pistons which mechanically catch theworking piston and center it in its neutral position.The locking pistons are spring loaded and kept clearof the working piston by the hydraulic pressure itself.So any loss of system pressure - intentionally oraccidentally - releases both locking pistons. Thetime from the trigger signal to the complete lock-outwas repeatedly determined as one of the importantparameters in each flight test stage. It typically

37-3

stayed below 90 ms corresponding to approximately90 deg rotor azimuth for all flight conditions.

2.4 TestbedThe CH-53G testbed 84+02 is operated by the WTD61 since many years and had been used for a broadspectrum of test and certification issues for that heli-copter type. Therefore, a comprehensive instru-mentation did already exist with regard to flight me-chanical parameters. Some 80 parameters are con-tinuously measured with sampling rates of 119 or476 Hz and recorded as PCM data stream on a DATrecorder.

With the installation of the IBC system some 100additional parameters had to be processed. Theystem from various sensors in the rotating framewhich monitor the rotor operating condition (pitch,flap, lag angles, shaft and blade bending moments,control system loads, etc.) or the proper function ofthe IBC actuators (piston travel, pitch link and scis-sors loads). Additional sensors are placed in thefuselage which monitor the condition of the hydraulicsystem and the accelerations at different locations.Moreover, several internal states of the digital IBCcontroller boards in the rotor hub and in the fuselageare continuously recorded for safety reasons. Allthese signals are processed by the IBC measure-ment system, which is integral part of the IBC com-puter. The sampling rate is precisely synchronized tothe rotor azimuth using 128 samples per revolution.Separate measurements are automatically or manu-ally triggered and extend over a predefined period oftime. All data are written to a flash disk for easy off-line processing.

Both systems continuously exchange data via ananalog interface. The IBC measurement systemreceives all parameters processed in the Data Gath-ering system while sending most of its own pa-rameters to that system. This allows all safety rele-vant parameters of the IBC system to be included inone of the two data streams which are transmitted tothe ground station for online monitoring.

2.5 Qualification and CertificationBefore its integration into the testbed, the key IBCcomponents as well as later the complete systemunderwent the usual series of development, qualifi-cation and certification tests. For this purpose ZFLhad set up various test rigs, where the followingissues have been experimentally investigated:• Fatigue strength of IBC actuator and its scissors

(five specimens, >107 load cycles each)• Piston seal and bearing life under centrifugal

loads (>1,250h)• Functional reliability and endurance of actuator

lock-out system (three actuators, >78,000 se-quences over all)

• Lock-out time at low and high temperatures• Reliability of shut-off manifold and hydraulic

system

In a special component test rig a complete CH-53Gflight control system had been set up. It consisted ofan original swashplate, swashplate guide, IBC ac-tuators and flap/lag hinges attached to a dummyrotor hub. Aerodynamic and inertial/centrifugal loadson the actuators were simulated by added springsand masses. The interaction of the IBC actuatorswith the primary control system and the change ofcontrol loads due to IBC operation was thoroughlyexamined. All components were equipped with straingauges as still are the flying components for moni-toring the in-flight loads.

After the qualification of the separate IBC compo-nents, the whole IBC system was tested on an ironbird. It consisted of a complete CH-53G quickchange unit, i.e. main gear box, swashplate, rotorhead assembly, and the associated hydraulic sys-tem. This system test has provided valuable insightinto the dynamical behavior of the hydraulic system,especially the pressure pulsation during harmonicactuator operation and system lock-out. The testingof the fully equipped quick change unit also includedthe final acceptance test before it was installed inthe testbed.

The whole certification process was guided by theAirworthiness Center for Military Aircraft at the WTD61. The IBC system has been designed according tothe applicable paragraphs of FAR Part 29, MILT8679, MIL H8501 A, MIL Std 1629A, ADS-33D,and VG-95370. In addition, the following standardshad to be followed for all components that constitutethe lock-out system: MIL H5440 H, RTCA-DO160D,VG-95373. Flight clearance was granted step bystep for progressive IBC authorities based on theexperimental results of the preceding step.

3 Open Loop Flight Test CampaignAfter the complete ground test certification proce-dure was accomplished, flight clearance for the firstflight with IBC was granted in December 2001. In thefirst test phase hard stops were implemented thatallowed for max. IBC amplitudes of 0.15 deg. Theflights covered various system validation tests andthe first investigation of the IBC benefits. Althoughprimarily intended to serve as a certification mile-stone and to prove the basic IBC functionality itturned out that the effectiveness of IBC in affectingvibrations was so strong that valuable results couldalready be gathered with this small authority. Laterduring the test program the hard stops were step-wise changed to allow for amplitudes of 0.67 andfinally 1.1 deg. Table 2 gives an overview of thevarious test conditions covered up to now.

37-4

All flight tests are conducted in close cooperationbetween WTD 61 and ZFL. WTD 61 provides allnecessary support for the basic aircraft. This in-cludes maintenance personnel, the complete flightcrew, and the airport with its proven flight test infra-structure.

P = Primary Phases: 0°, 60°, 120°, 180°, 240°, 300°, 359° HF = Horizontal Level FlightS = Secondary Phases: 30°, 90°, 150°, 210°, 270°, 330° DC = Descend FlightT = Test Phases: 0°, 90°, 180°, 270°

IBC Inputs (max. Amplitudes)FlightCondition 2/rev 3/rev 4/rev 5/rev 6/rev 7/rev

HOGE 0.67° 0.67° 0.67° 0.67° 0.67° 0.67°

40ktsHF

0.67° 0.67° 0.67° 0.67° 0.67° 0.67°

60ktsHF

0.67° 0.67° 0.67° 0.67° 0.67° 0.67°

90ktsHF

0.67° 0.67° 0.67° 0.25°T

0.25°T

0.25°T

Certification Flights EMC-Tests Lock-Out Interaction

IBC/ AFCS

130ktsHF

0.67° 0.67° 0.67° 0.25°T

0.25°T

0.25°T

60ktsHF

0.83° 0.67°P, S

0.15°P

0.15°P, S

0.15°P

90ktsHF

1.10°P

0.67°P, S

0.40°P, S

0.25°P, S

0.25°P, S

0.25°P, S

VibrationsPitch Link Loads

120ktsHF

1.10°P, S

0.83°P, S

0.50°P, S

0.25°P, S

0.25°P, S

0.25°P, S

95ktsHF

0.33°P

Noise

65ktsDC -6°

0.67°0°, 30°, 60°,90°, 120°,240°

Power Required 130ktsHF

0.67°0°, 180°, 240°, 300°

Table 2: Overview of IBC Flight Test Conditions

3.1 Open Loop ResultsThe results presented hereafter are derived from thedata of the first two phases where actuator strokescorresponding to max. 0.15 and then 0.67 deg couldbe used. The standard procedure during the openloop tests was to pick one single frequency, keepthe amplitude constant and vary the correspondingIBC input phase angle in 60 deg increments. Refer-ence data points were taken before and after eachphase sweep with the actuators both locked andunlocked but commanded to neutral position.

3.1.1 Actuator to Blade Transfer FunctionThe IBC system precisely controls the actuatorstroke. Depending on the applied frequency theresultant blade pitch motion differs with regard toamplitude and phase from the geometrically ex-pected value. Figure 5 compares the actuator stroke(expressed in terms of the corresponding nominalblade pitch angle) and the actually measured blademotion at that frequency.

It can be seen, that for 3/rev and 4/rev the blademotions tend to be higher than expected from astrictly geometric transfer of the actuator stroke,whereas at 5/rev and 7/rev the blade motions areconsiderably smaller. Moreover, phase deviationsare present in all cases. The fact that at the higherfrequencies the transfer function is quite independ-ent from the IBC phase suggests that the impedanceseen by the actuator is primarily of inertial nature. Atthe lower frequencies, however, the magnitude

strongly depends on the IBC phase, showing astrong impact of the aerodynamics. In the most ex-treme case, a 2/rev actuator motion imposed with300 deg IBC phase just cancels the 2/rev blade pitchmotion component obviously present at this flightcondition without IBC, whereas for 120 deg IBCphase that inherent 2/rev blade motion is more thandoubled.

−0.3 −0.2 −0.1 0 0.1 0.2

−0.1

−0.05

0

0.05

0.1

0.15

0.2

0.25

2/rev

cos−Comp.

sin−

Com

p.

−0.1 0 0.1 0.2−0.15

−0.1

−0.05

0

0.05

0.1

3/rev

cos−Comp.

sin−

Com

p.

−0.1 0 0.1 0.2

−0.1

−0.05

0

0.05

0.1

0.15

0.24/rev

cos−Comp.

sin−

Com

p.

−0.1 0 0.1

−0.1

−0.05

0

0.05

0.1

5/rev

cos−Comp.

sin−

Com

p.

−0.1 0 0.1

−0.1

−0.05

0

0.05

0.1

6/rev

cos−Comp.

sin

−C

om

p.

−0.1 0 0.1

−0.1

−0.05

0

0.05

0.1

7/rev

cos−Comp.

sin−

Com

p.

Figure 5: Blade Pitch Angle Response to IBC withConstant Actuator Stroke (2/rev ... 6/rev @120 kts and 7/rev @ 90 kts)

3.1.2 Vibration ReductionThe diagrams presented hereafter show the relevant6/rev component of the measured parameter (ascomputed from FFT over 16 rotor revolutions) eitherby SIN/COS polar plots, by magnitude versus IBCphase angle plots or by frequency spectra. All datapoints were taken in straight level flight at approxi-mately 4000 ft density altitude; the aircraft weightranged from 75% to 85% of the CH-53G's 42,000 lbsMTOW. This compilation is solely intended to high-light some of the interesting effects and by nomeans comprehensive.

Before the first data had been evaluated it was as-sumed that the initial authority corresponding to +/-0.15 deg nominal blade pitch variation would hardlyyield any usable data, since the impact on the vibra-tions was expected to be small. The data then proc-essed, however, showed a very high effectiveness

37-5

of IBC with respect to the vibrations, which was notmodeled by any of the rotor codes used for predic-tion. Moreover, the recorded data are of high con-sistency even though the number of repeated datapoints was restricted. To make the data applicable toamplitude extrapolation and multiple frequency (i.e.mixed mode) predictions they were used to identify asecond order T-matrix model (T2-fit) according to thefollowing relationship.

021 )( zuTudiaguTz ++=

More details are given in Ref. [10]. This model hasproven its ability to capture also most of the inherentnonlinear effects when using non-HHC frequencies(2/rev, 3/rev and 4/rev in the case of the six-bladedCH-53G). It is also used within one of the closedloop control algorithms

In the following diagrams Figs. 6 through 8 the 6/revaccelerations at two different sensor locations re-corded during 0.15 deg 5/rev IBC phase sweeps arecompared to the corresponding reference data with-out IBC.

x y z

x y z

Figure 6: Measured 6/rev Accelerations at MainGearbox and Pilot Seat due to 0.15 deg5/rev IBC (120kts)

The added solid curves in Fig. 6 and 7 represent theidentified T2-fit. The values predicted by the identi-fied model and the actual data points are connectedby short lines representing the identification errors. Itcan be seen that all data can consistently be repre-sented by the T2-fit and the identification errors stayvery small in most cases. Moreover, although notevaluated by the identification algorithm even theIBC-off values of the model match the measuredreference data quite well. In this 5/rev case one findsfour examples where the applied small IBC ampli-tudes were already sufficient to cancel a particularacceleration component. It must be noted, however,that for the simultaneous reduction of multiple vibra-

tion components at different locations the required5/rev contribution to an optimized mixed mode signalcan be somewhat higher.

x y z

x y z

Baseline

Baseline

Figure 7: Measured 6/rev Accelerations at MainGear-box and Pilot Seat due to 0.15deg5/rev IBC vs. Phase Angle (120kts)

It should also be noted that in most cases from plotslike Fig. 7 no final conclusions can be drawn withrespect to the achievable vibration reduction and theoptimum IBC phase angle unless the used ampli-tude happens to be the optimal one. Since the am-plitudes have never explicitly been optimized duringthe open loop campaign, such impressive resultslike the almost complete cancellation of the vertical6/rev component at the pilot seat as shown in Fig. 8were recorded more by chance rather than by inten-tion.

without IBC

with IBC

Figure 8: Effect of 0.15 deg 5/rev IBC on z-VibrationSpectrum at Pilot Seat (120kts)

Figures 9 and 10 show the influence of the two adja-cent frequencies, 4/rev and 6/rev. It is worthwhilementioning that 4/rev and 6/rev show similar effec-

37-6

tiveness (for the definition of the effectiveness seeRef. [10]), although 4/rev is not a classical HHCfrequency for the six-bladed rotor and would beranked ineffective by linear first order analysis.

x y z

x y z

Figure 9: Measured 6/rev Accelerations at MainGearbox and Pilot Seat due to 0.15 deg4/rev IBC (120kts)

−0.1 −0.05 0 0.05−0.05

0

0.05

0.1

0.15

ref_7

←φ=360

←φ=60

←φ=121φ=181→

φ=299→φ=240→

AccHGx [g] <− 6/rev

sin−

Com

p.

εz=12.54[%] E=0.40[g/°]

−0.02 0 0.02 0.04 0.06 0.080

0.02

0.04

0.06

0.08

0.1

0.12

ref_7

←φ=360

←φ=60

←φ=121

←φ=181

←φ=299

φ=240→

AccHGy [g] <− 6/rev

sin−

Com

p.

εz=5.54[%] E=0.26[g/°]

−0.1 −0.05 0

−0.15

−0.1

−0.05

0

ref_7

φ=360→φ=60→

φ=121→

φ=181→φ=299→

φ=240→

AccHGz [g] <− 6/rev

sin−

Com

p.

εz=5.66[%] E=0.35[g/°]

0 0.02 0.04 0.060

0.02

0.04

0.06

0.08

ref_7

←φ=360

←φ=60

←φ=121

←φ=181

←φ=299←φ=240

AccPilx [g] <− 6/rev

sin−

Com

p.

εz=5.92[%] E=0.12[g/°]

−0.08−0.06−0.04−0.02 0 0.020

0.02

0.04

0.06

0.08

0.1

0.12

ref_7

φ=360→φ=60→

φ=121→

φ=181→

φ=299→

φ=240→

AccPily [g] <− 6/rev

sin−

Com

p.

εz=4.32[%] E=0.13[g/°]

−0.1 −0.05 0

0

0.05

0.1

ref_7

φ=360→

φ=60→φ=121→

φ=181→

φ=299→

φ=240→

AccPilz [g] <− 6/rev

sin−

Com

p.

εz=8.10[%] E=0.18[g/°]

Figure 10: Measured 6/rev Accelerations at MainGear-box and Pilot Seat due to 0.15 deg6/rev IBC (120kts)

Figures 11 and 12 give an overview of the measuredIBC effectiveness for all applied frequencies. Espe-cially at the main gearbox the typical behavior canbe observed that 5/rev and 7/rev IBC primarily af-fects the in-plane accelerations while 6/rev is moreeffective in the vertical axis. The fact that the non-HHC frequency 4/rev has considerable impact onthe vibrations is again underlined. This frequencywill later be shown to serve as valuable supplementin mixed mode cases. In general, the calculated IBCeffectiveness with respect to the vibration reductiontask reaches values of more than 0.6 g/deg, which isconsiderably higher than all values measured for theBO-105 or UH-60 helicopters.

2/rev 3/rev 4/rev 5/rev 6/rev 7/rev

zy

x0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

IBC

Effe

ctiv

enes

s [g

/deg

]

Figure 11: Effectiveness of Different IBC Harmonicsat Main Gearbox (6/rev Acceleration perdeg n/rev IBC; 120kts, 7/rev: 90kts)

2/rev 3/rev 4/rev 5/rev 6/rev 7/rev

zy

x0

0.1

0.2

0.3

0.4

0.5

0.6

0.7IB

C E

ffect

iven

ess

[g/d

eg]

Figure 12: Effectiveness of Different IBC Harmonicsat Pilot Seat (6/rev Acceleration per degn/rev IBC; 120kts, 7/rev: 90kts)

As can be seen from the previous pictures, in manycases the optimum IBC amplitudes and phase an-gles for the three orthogonal sensor orientations ordifferent sensor locations do not coincide. This be-comes obvious as soon as the x, y, and z signalsare combined into unweighted cost functions repre-senting the spatial accelerations at the different lo-cations, see Fig. 13.

Even though the deviating requirements for the dif-ferent sensor directions and locations diminish theover-all results, the presented example shows that insome cases single harmonic IBC was able to lowerthe spatial vibrations at one location by more than60% (cargo compartment) while the vibrations at theother observed locations were also reduced, how-ever to a smaller extent. In other cases single har-monic IBC had divergent effects on different loca-tions.

37-7

Figure 13: Effect Spatial 6/rev Accelerations at FourSensor Locations Due to 0.15 deg 6/revIBC (60kts)

3.1.3 Predicted Multiharmonic EffectAs shown in the previous section single harmonicIBC becomes less effective if one needs to reducevibrations at different locations and multiple axessimultaneously. Therefore, the identified nonlinear T-matrix models were used to calculate optimummixed mode inputs considering frequencies from4/rev to 7/rev. In the following example the costfunction consisted of the three equally weightedacceleration components x, y, z at the main gearboxor the pilot seat, respectively. Figure 14 shows therelative effect of different frequency combinations onthe cost function as yielded by the numerical optimi-zation.

The calculations confirm earlier findings partly al-ready known from previous HHC and IBC programs.• The important role IBC with (N-2)/rev can play is

underlined (4/rev in this case)• Applying more than one control frequency helps

to tackle different sensor axes and/or locationssimultaneously

• For the CH-53G the over-all authority requiredfor the vibration reduction task seems to staybelow 1 deg even for the suitable multi-harmonicinputs

1

0.8

0.6

0.4

0.2

0Opt

im. A

mpl

itude

s [°

]Vi

brat

ion

Red.

[%] 100

80

60

40

20

0

5/re

v 4/re

v 7/re

v 6/re

v

5, 6

/rev

4, 6

/rev

4, 6

, 7/re

v

5, 6

, 7/re

v

5/re

v

5/re

v

4/re

v

7/re

v

6/re

v 6/re

v

6/re

v4/re

v

4/re

v6/

rev

7/re

v5/

rev 7/

rev

6/re

v

Main Gearbox

1

0.8

0.6

0.4

0.2

0

100

80

60

40

20

0

4/re

v 5/re

v

7/re

v

6/re

v

4, 6

/rev

5, 6

/rev

5, 6

, 7/re

v

4, 5

, 6/re

v

4/re

v

5/re

v 7/re

v

6/re

v

6/re

v

6/re

v4/

rev 6/

rev

4/re

v6/

rev

5/re

v

4/re

v

5/re

v

7/re

v5/

rev

Pilot Seat

Figure 14: Predicted Spatial Vibration Reduction forMain Gearbox and Pilot Seat (120kts)

It can be seen that for this nonlinear extrapolation amixture of three frequencies is sufficient to almostcompletely cancel the considered vibration compo-nents. These theoretical results will shortly be vali-dated when optimized mixed mode IBC will be ap-plied during the closed loop campaign.

3.1.4 Noise ReductionAs it has been shown in various investigations (e.g.Refs. [6], [7], [14]), noise radiation caused by theblade vortex interaction (BVI) phenomenon can belowered by using 2/rev and/or 3/rev IBC. Thus, theeffect of 2/rev IBC on the noise measured on groundwas studied during the open loop campaign. As ithas been shown in the previous section, 2/rev con-trol was found to have only small impact on vibra-tions for this six-bladed rotor. Hence, 2/rev controlwas expected to be applicable without any severedrawbacks to the vibration reduction task.

The noise measurements were performed in closeco-operation with the WTD 61 in a setup accordingto the ICAO regulations. Three microphones wereplaced in a line perpendicular to the flight path.Marks on the ground and a mobile GPS receiversupported the pilot in flying as exact as possibleover the center microphone in the desired altitude.The flight path was tracked and recorded at aground station by optical means. These data wereused to give the crew instantaneous feedback on thequality of the flight path and later to correct themeasured sound pressure levels according to theresidual deviations. In addition to the ground meas-urements two external microphones had beenmounted to the right sponson of the aircraft. It wasintended to use their signals as an on-board indica-tor for the noise radiation of the rotor.

37-8

The following diagrams refer to the descent flightexperiments only, for additional results from thehorizontal fly-overs see Ref. [19]. Figure 15 showsthe geometry of the descent flight path. A variationof ±10 m in altitude and ±25 m lateral deviation wastolerable (due to GPS and the ground marks, how-ever, lateral deviations were much smaller). Other-wise, the flight was repeated. The altitude of thehelicopter above the microphone line was used tocorrect the sound pressure level in order to get con-sistent data. A measurement distance of 500 mahead and behind the microphone line had to bepassed with the flight condition kept as steady aspossible. The aircraft was expected to cross themicrophone line in an altitude of 120 m. The IBCsignal was faded in before the helicopter entered themeasurement zone and was kept constant through-out the measurement. Based on preceding refer-ence flights at different descent rates a flight pathangle of -6° was found to produce the strongest BVIeffect and therefore chosen for the IBC trials. Thiswas in good agreement with the results from the IBCflight testing of the BO-105.

Top View(not to scale)max. Lateral Deviation: ±25m

Side View(not to scale)max. Vertical Deviation: ±10m

Microphone #1

Microphone #2

-γ

500m 500m

End ofMeasurement

Start ofMeasurement

Distance 1000m

150m

150m

FlightPath

120m Altitude atMicrophones Microphone

Line

Microphone #3

Figure 15: Descent Flight Path Geometry for NoiseReduction Flight Tests

Figure 16 shows the impact of 0.67 deg 2/rev IBC onthe averaged sound pressure levels at the threemicrophones (center, retreating, advancing side) at65 kts for various IBC phase angles. Without IBC thecenter microphones had recorded the highest levels.The maximum reduction was achieved at 30 degIBC phase angle, when the sound pressure level atthe center microphone was reduced by about 3 dB.This is an impressive result, given the relativelysmall amplitude of only 0.67 deg. Simultaneously,the sound pressure levels at the other two micro-phones were also reduced. This is important, since itclearly indicates an alleviation of the BVI effect,whereas for 60 deg phase angle, the sound pres-sure level was reduced at the center microphone,but increased on the retreating side pointing towardsa change in directivity rather than an over-all reduc-tion.

-3.5

-3

-2.5

-2

-1.5

-1

-0.5

0

0.5

1

1.5

0 30 60 120 240 300 360

IBC-Phase 2/rev [°]

Soun

d Pr

essu

re L

evel

Diff

eren

ce [d

B]

Micro 1 (centre)

Micro 2 (Retr. Side)

Micro 3 (Adv. Side)

Figure 16: Effect of 0.67 deg 2/rev IBC on MeasuredNoise in Descent Flight Condition (65kts,6deg Glide Slope)

From several studies it is known that higher levels ofnoise reduction can be expected, if higher ampli-tudes are applied. The analysis of IBC wind tunneltests with respect to noise reduction given in [14]clearly shows falling sound pressure levels for in-creased IBC amplitudes. This implies that lowersound pressure levels can be expected also for theCH-53G when the noise measurements are re-peated with the maximum authority of 1.1 deg.

Finally, Fig. 17 shows the comparison of soundpressure level time histories without and with IBC(A2 = 0.67 deg, ϕ2 = 30 deg). As can be seen, thesound pressure level is not only being reduced at allmicrophones, but also throughout almost the com-plete time history. This again points at an over-allreduction of the BVI strength.

1031009794918885

Mic #2 (Retr. Side)

1031009794918885

Mic #1 (Center)

VIAS = 65kts, γ = -6°

Non

-cor

rect

ed S

ound

Pre

ssur

e Le

vel [

dB]

1031009794918885

Mic #3 (Adv. Side)

-10 -8 -6 -4 -2 0 2 4 6 8 10

Time [sec] (t=0 at moment of highest sound pressure level)

Reference, no IBCA2 = 0.67°, ϕ2 = 30°

Figure 17: Effect of 2/rev IBC on Non-corrected SoundPressure Level in Descent

3.1.5 Reduction of Power RequiredIt has long been accepted that 2/rev IBC can benefi-cially affect the rotor power required, see Ref. [5]

37-9

beside others. Therefore, 2/rev flight tests at highforward speed have also been evaluated with re-spect to this IBC application. Moreover, specialflights have been devoted to this particular question.The general problem of resolving small powerchanges from flight test measurements also appliesto the IBC experiments. Two approaches were usedto assess the IBC effect on the main rotor powerconsumption. First, IBC phase sweeps were con-ducted in close sequence so that small changes ofthe flight condition introduced by the IBC inputsthemselves had to be tolerated. Second, on-off se-quences were performed, while the pilot was askedto retrim the aircraft as precisely as possible.

Figure 18 shows data taken according to the firstapproach. It can be seen that beside the measuredshaft power also speed and altitude changed fromdata point to data point. This can be caused eitherby the varying IBC effect on torque, lift, and/or pro-pulsive force or by the pilot’s reaction on trimchanges. The shown measurements were takenapproximately 5 sec apart from each other and showvery consistent trends. A second phase sweep withIBC phase angles shifted by 30 deg (ϕ2 = 30, 90,150, 210, 270, 330 deg) was used to supplement theshown first sequence.

4500

4000

3500Alti

tude

[ft]

#1 #2 #3 #4 #5 #6 #7 #8Data Points

135

130

125

120VIAS

[kts

]

Level Flight VIAS = 125kts, A2 = 0.67°w/o IBC ϕ2 = 0° = 60° = 120° = 180° = 240° = 300° = 359°

2600

2500

2400

2300

Pow

erQ

·Ω [k

W]

Figure 18: Flight Performance Relevant Parametersduring 2/rev IBC Phase Sweep

Both runs consistently showed a power variation inthe order of ±2 deg with the minimum at phase an-gles of approximately 240 deg, compare upper dia-gram of Fig. 19. Then, to consider the variation ofthe flight condition, correction factors have beencomputed, which translate the relevant deviationsinto equivalent power changes. The parametersused for these corrections comprise (a) the forwardspeed affecting the parasitic power, (b) the flightpath angle representing the exchange of potentialenergy, and (c) the speed change representing thevariation in kinetic energy. Simple energy relation-ships turned out to yield very similar results com-pared to more sophisticated analysis with a compre-hensive helicopter performance code.

15

10

5

0

-5

-10

-15

∆Pco

rrec

ted [

%] More Power Required

Less Power Required-6%

0 60 120 180 240 300 360IBC-Phase Angle ϕ2 [°]

15

10

5

0

-5

-10

-15

∆Pun

corr

ecte

d [%

]

-2%

Figure 19: Rotor Power Reduction due to 0.67 deg2/rev IBC (125 kts)

Figure 20 shows the corresponding power correc-tions that had to be applied to the shaft powermeasurements. The resulting power variation versusIBC phase angle is shown in the lower diagram ofFig. 19. At the optimum phase angle of now 210 dega power reduction of more than 6% was found.

1510

50

-5-10-15∆P

alte

red

spee

d [%

]

Speed/Drag Effect

1510

50

-5-10-15∆p

clim

b/de

scen

t [%

]

Heave Effect

∆pac

cele

r./de

cele

r. [%

]

1510

50

-5-10-15

0 60 120 180 240 300 360IBC-Phase Angle ϕ2 [°]

Acceleration Effect

Figure 20: Power Corrections due to Flight ConditionChanges

37-10

It is worthwhile mentioning that although the correc-tion factors in some cases did develop differentlyduring both phase sweeps (see phase angles 0, 30,60 deg, e.g.) the trend of the corrected power is veryconsistent.

The second approach, used as benchmark only atthe presumably optimum IBC phase angles, hasyielded very similar power reductions with 3.1% at180 deg and 4.2% at 240 deg.

3.1.6 Load ReductionThe last IBC benefit that has been investigatedbased on the conducted flight tests concerns thereduction of different loads. One known effect is thechange of pitch link or rather actuator loads due toIBC, compare Ref. [5]. But many more componentsof the dynamic system are affected by the IBC in-puts. For the ageing fleet of CH-53Gs this applica-tion will be of increased importance since ap-proaching retirement times for rare and/or highpriced components start to become a major concern.

It is well known that at higher forward speeds thepitch link loads are composed of significant portionsof higher harmonic components, which considerablycontribute to the peak-to-peak values relevant forthe component life time. Therefore, the impact of thelower IBC frequencies on the corresponding compo-nents of the pitch link loads were studied. Again, theprimary attention was put on single harmonic 2/revresults since the aerodynamic effects, which couldbeneficially be altered by IBC, were known to bemost dominating at this frequency. Figure 21 showsthe recorded pitch link load amplitudes versus theIBC phase angle.

-3000-2000-1000

0100020003000400050006000

7000

0 60 120 180 240 300 360Phase Angle [deg]

Pitc

h Li

nk L

oad

[N]

Amplitude with IBC Ref. Amplitude Locked Actuator Ref. Amplitude Unlocked Actuator Mean Value with IBC Ref. Mean Value Locked Actuator Ref. Mean Value Unocked Actuator

0

1000

2000

3000

4000

5000

6000

0 60 120 180 240 300 360

Amplitude with IBC, A2=0,66°

Amplitude w/o IBC

Figure 21: Effect of 0.66 deg 2/rev IBC on PitchLink/Actuator Load Amplitudes (125kts)

At 270 deg phase angle the pitch link load amplitudewas reduced by more than 25%. It should also benoted that at 210 deg, where the rotor power re-quired was optimally reduced, the control systemloads were still more than 15% below the reference

value. The following time history plots underline theimpressive IBC effect. Figure 22 shows the rise ofthe actuator axial forces after the beneficial 2/revIBC signal was intentionally faded out. Figure 23directly compares the control load time histories fortwo comparable data points without and with optimalIBC. In this case the load amplitude is reduced bymore than 30%.

Blade #1

ϑ IB

C [°

]F P

L [kN

]

Figure 22: Effect of 0.66 deg 2/rev IBC on Pitch Link/Actuator Load Time History (OptimumPhase, IBC Fade-Out, 130kts)

w/o IBC

MA

G (F

PL) [

kN]

IBC on:

revs [-]

2

Figure 23: Effect of 0.66 deg 2/rev IBC on Pitch Link/Actuator Load Time History (over 2 RotorRevolutions, Optimum Phase, 130kts)

As mentioned earlier, also other components in therotor and control system load paths can be affectedby IBC. Higher IBC frequencies tend to increase theloads due to the dominance of inertial effects. Fortu-nately, the application of such frequencies needed

37-11

never to be restricted with regard to the load issuebecause of the small amplitudes required for thevibration reduction task. The lower frequencies, incontrast, could deliberately be used to reduce loadsacting on certain critical components. Figure 24shows the relative variation of three more sensorsignals due to 2/rev and 3/rev IBC. Since the opti-mum phase angles do not coincide for all compo-nents, the cost function to be used by a feasiblecontroller will have to combine different load meas-urements in an appropriate manner.

Pitc

h Li

nkLo

ads

[%]

Forc

e R

otat

ing

Scis

sors

[%]

Flap

ping

Mom

ent [

%]

Shaf

t Ben

ding

Mom

ent [

%]

2/rev, A2=0.67° 3/rev, A3= 0.26°

ϕ 2 [°] ϕ 3 [°]

-30

0

30

-30

0

30

60

-30

0

30

-30

0

30

0 90 180 270 360

-30

0

30

-30

0

30

60

-30

0

30

-30

0

30

0 90 180 270 360

Figure 24: Relative Effect of 0.67 deg 2/rev and 0.26deg 3/rev IBC on Different Rotating SystemLoad Parameters

4 Closed Loop Flight Test CampaignAs expected from the beginning, the optimum IBCinputs considerably depend upon the flight condition.The predominant parameter was found to be theforward speed. Therefore, it will not be sufficient touse predefined sets of IBC inputs in a productionversion of an IBC system. Some years ago ZFL hadstarted to investigate different candidates of controlalgorithms to be used for closed loop control, seeRef. [8]. Due to the success of the initial open looptests ZFL was awarded an extension of the originalcontract which covers the development and flighttesting of such a closed loop IBC system.

4.1 Control System ArchitectureTo retain the open loop core system with all itssafety features a separate control computer wasadded to the existing open loop system which formsan additional outer control loop and allows to testdifferent frequency domain algorithms. Thus, theclosed loop hardware added to the existing IBCsystem extends the existing cascade control struc-ture of the present IBC system by an additional con-trol loop, as shown in Fig. 25.

bladedynamic

IBCActuatorP

Inner Control Loop(Position Control)

blade iblade N

blade 1

Helicopter

Ai, ϕi

Ai,refϕ i,ref

FFT

ψ

Amplitude/Phase error

CompensationIFFT

Intermediate Control Loop(for each blade)

OuterController

Outer Control Loop

vibration, control load,

noise,rotor powermeasurands

Figure 25: Control System Structure of Closed LoopIBC System

The open loop core system was altered only wherenecessary to communicate with the closed loopcontrol computer. This approach was essential inminimizing the certification effort, since the inputs ofthe controller can now be treated very similar to themanual inputs of the former human IBC operator.The sensor signals required for feedback were al-ready available in the data recording portion of theopen loop IBC system.

The main interface between the closed loop hard-ware and the open loop system is the link whichcontinuously transmits the commanded amplitudeand phase values of the chosen IBC frequencies.What was done manually by the flight test engineerduring the open loop flight tests is now provided bythe closed loop control computer. To control the testprocedure during the flight an additionalman/machine interface has been designed. Theclosed loop control system consists of a modulardSPACE hardware system, a TFT-Touch screenand a Host-PC. The controller design is performedunder Matlab/Simulink and automatically imple-mented on the dSPACE real-time system from blockdiagram level. The modified IBC system is capableof generating single and mixed mode IBC inputs thatconsist of arbitrary combinations of harmonics from2/rev – 7/rev. The highest control update rate feasi-ble with the extended IBC system is 4/rev.

The closed loop control algorithms used to form theouter control loop for the vibration and control loadreduction tasks are based on the well-known fre-quency domain approach, compare Refs. [8] and[15] through [18]. The frequency domain approachassumes a quasi-static linear relationship betweenthe outputs (measured vibrations and control loads)and the corresponding sets of IBC inputs. This rela-tionship is formed by the so-called T-matrix model.The inputs are characterized by the cosine and sineparts of the 2/rev – 7/rev components of the IBC-input. The vibrations are composed of the NΩ com-ponents of accelerations measured at different loca-tions (pilot seat, main gearbox, cargo compartmentand tailboom). For the reduction of control loadssuitable frequency components of measured pitchlink and booster loads are used instead.

37-12

The over-all outer control loop used for the vibrationand control load reduction consists of two maintasks. A system identification task used to estimatethe T-matrix model and a controller task used tocalculate IBC inputs to accomplish the desired con-trol goal. The system identification can be performedusing different Recursive Least Square methods(standard RLS, RLS with forgetting factor, differentstabilized RLS methods) or Kalman filter implemen-tations. The system identification task is able to es-timate de-coupled T-matrix sub-models. If for exam-ple vibration and control loads shall simultaneouslybe addressed, the impact of 2/rev IBC on all vibra-tion components can be neglected in the identifica-tion algorithm. Using this feature it is possible toneglect small cross-coupling effects in the controllerdesign. Therefore selected IBC frequencies can beused to control specified outputs. Within the con-troller task the computation of IBC inputs is realizedby minimizing

nTnn

Tnnz

Tn uWuuWuzWzJ ∆∆++=

∆ϑϑ

along with the identified T-matrix model

( )111ˆ

−− −+= nnnn uuTzz

where the subscript n denotes the current time step.The solution of the above optimization problem iseither formulated using the feedback of measuredvalues of zn or using feedback of the identified refer-ence values. The IBC-frequencies 4 through 7/revare preferred for the vibration reduction and 2/revIBC is reserved for the control system load reduc-tion. Due to the different implementations of thesystem identification and the controller tasks theouter control loop can be realized with the followingstructures: non-adaptive closed loop, adaptiveclosed loop, non-adaptive feed forward; and adap-tive feed forward.

4.2 Closed Loop Ground Test ResultsIn order to validate the function of the closed loophard- and software, special tests have been per-formed not only with the system test rig but also withthe testbed helicopter during ground test runs. Be-cause no useful acceleration signals were availableduring these tests (rigid test rig or fuselage on theground), a servo (booster) force of the primary con-trol system was chosen as the parameter to be con-trolled by 6/rev IBC. Figures 26 and 27 show controlsequences as recorded during these tests.

In the first case, Fig. 26, the plotted time historiescomprise both the identification and the control timeframes. For this test, an artificial 6/rev disturbancesignal was superimposed to the measured servoforce. It can clearly be seen, how this force compo-nent reacts on the automated phase sweep which is

invoked as part of the identification process. Then,within less than 4 sec the controller has convergedto the optimum 6/rev amplitude and phase valueswhich perfectly cancels the fictitious servo load.

0.00.10.20.3

A6 [

°]

0120240360

ϕ 6 [°

]6/

rev

Aft S

ervo

Fo

rce

[N] 1000

0 10 20 30 40 500

200400600800

Time [sec]

System IdentificationClosedLoop

Figure 26: Closed Loop Test Carried out at Full ScaleSystem Test Rig (Artificial 6/rev Aft ServoForce Suppressed by 6/rev IBC), from [19]

For the ground tests shown in Fig. 27, no dummysignals had to be added, since sufficient 6/rev signalcontent was present. In this case only the closedloop time frame is shown. Again, it can be seen, howfast and reliable the algorithm converges, before atthe end of the sequence IBC is switched off and thecost function jumps up to the initial reference value.Based on these successful tests a stable and effi-cient controller performance is expected also for theupcoming flight tests.

6/re

v A

ft Se

rvo

Forc

e [N

]A

6 [°]

ϕ 6 [°

]

IBC off

x 10-3

Time [sec]

Closed Loop Control

Figure 27: Closed Loop Test Carried out during aGround Test Run (6/rev Aft Servo ForceSuppressed by 6/rev IBC)

37-13

5 Conclusions and OutlookThis ongoing flight test campaign on the CH-53Ghas already demonstrated the potential of IBC toimprove the rotor properties with respect to vibra-tions, noise, power required, and control loads. Theresults presented above are very encouraging andconfirm the findings in [13]. The mixed mode ex-trapolations are still to be validated during the up-coming closed loop flight tests. The effectiveness ofIBC with respect to the vibration reduction task hasclearly surpassed prior expectations. The majorresults can be highlighted as follows:• More than 60% vibration reduction at the cargo

compartment through 5/rev single mode IBC hasbeen demonstrated with simultaneous, howeversmaller, improvements at all other evaluated lo-cations

• Local vibration reductions of more than 90%have been predicted by the non-linear T-matrixmodel for the application of mixed mode IBC

• The authority required for optimum mixed modevibration reduction is not expected to exceed 1deg

• In landing approach noise reductions of up to3dB have been shown using 0.67 deg 2/rev IBC

• In high speed level flight the effective rotorpower required was reduced by up to 6% againusing 0.67 deg 2/rev IBC

• At the same speed pitch link load reductions ofapproximately 30% were measured for appro-priate 2/rev IBC inputs

• Simultaneous power and control load reductioncan be realized because the optimum 2/rev IBCphase angles were found to be sufficiently close

One primary advantage of the IBC concept is thatdifferent deficiencies of a helicopter rotor operatingin tangential flow can be treated simultaneously byone single system. It is obvious, however, that anIBC retrofit kit will have to be optimized with respectto weight, cost and installation effort compared tothe described experimental system. The latter wasprimarily designed to be an experimental tool formaximum flexibility and minimum interference withthe testbed. The design goals for a production ver-sion are low weight, low power consumption, andsmall installation space as well as reliable functionand autonomous operation. ZFL has pursued sev-eral design studies for different helicopters, not onlyby varying details of the mechanical integration butalso under consideration of alternative methods forthe power supply.

One preliminary design for the CH-53G shows ahighly integrated solution weighing below 1% of thehelicopter MTOW. It features the complete integra-tion of all mechanical and hydraulic components inthe rotating frame, see Fig. 28. This architecture

removes the need for a hydraulic slipring. The pumpis driven by the rotor itself and does not need anextra power pickup.

Actuator

Heat Exchanger

Gearbox /Pump Unit

Control Panel

Controller

Manifold &Accumulators

PressureReservoir

Sensors

Figure 28: Possible Layout Variant of IBC Retrofit Kit

A further simplification of the IBC system could berealized if the design would consider the IBC-specific load / piston travel characteristics. Flight andwind tunnel test data have shown that the averagemechanical power consumed by each actuator iscomparably low, because the energy flow reversesduring a considerable part of each rotor revolution.Thus, the power demand of IBC could be drasticallyreduced if a "regenerative" IBC system was de-signed that allows for power recovery.

There are several technical solutions which followthis idea. One solution, the so-called IBC displace-ment system, is currently being tested at ZFL. Thissystem does not rely on the servo valve principle butdirectly connects a variable displacement pump withspecialized actuators. This setup enables the de-sired bi-directional energy flow between the actuatorand the pump. The regenerative IBC displacementdemonstrator has meanwhile proven the validity ofthis concept and is now being optimized in detail.This technical approach can greatly simplify the IBCsystem layout and may help to introduce IBC intoexisting or new helicopters.

AcknowledgementThe work presented in this paper has primarily beenfunded by the German Federal Office of Defense,Technology, and Procurement (BWB). The authoralso likes to thank the staff of the German FederalArmed Forces Technical and Airworthiness Centerfor Aircraft in Manching (WTD 61) for their continu-ous support of this successful IBC campaign.

37-14

References[1] P. Richter, H.-D. Eisbrecher, V. Klöppel, “De-

sign and First Tests of Individual Blade ControlActuators”, 16th European Rotorcraft Forum,1990.

[2] W. R. Splettstoesser et al., “BVI ImpulsiveNoise Reduction by Higher Harmonics PitchControl: Results of a Scaled Model Rotor Ex-periment in the DNW”, 17th European Rotor-craft Forum, Berlin, 1991.

[3] D. Teves, V. Klöppel, P. Richter, “Developmentof Active Control Technology in the RotatingFrame, Flight Testing and Theoretical Investi-gations”, 18th European Rotorcraft Forum, Avi-gnon, 1992.

[4] S. A. Jacklin, A. Blaas, S. M. Swanson, D. Te-ves, “Second Test of a Helicopter IndividualBlade Control System in the NASA Ames 40-by-80 feet Wind Tunnel”, 2nd AHS InternationalAeromechanics Specialists Conference, 1995.

[5] U.T.P. Arnold, M. Müller, P. Richter, “Theoreti-cal and Experimental Prediction of IndividualBlade Control Benefits”, 23rd European Rotor-craft Forum, Dresden, 1997.

[6] S.M. Swanson, S.A. Jacklin, A. Blaas, G. Niesl,R. Kube, “Acoustic Results from a Full-ScaleWind Tunnel Test Evaluating Individual BladeControl”, 51st AHS Annual Forum, Fort Worth,1995.

[7] D. Schimke, U.T.P. Arnold, R. Kube, “IndividualBlade Root Control Demonstration - Evaluationof Recent Flight Tests”, 54th AHS Annual Fo-rum, Washington D.C., 1998.

[8] D. Morbitzer, U.T.P. Arnold, M. Müller, “Vibra-tion and Noise Reduction through IndividualBlade Control Experimental and TheoreticalResults”, 24th European Rotorcraft Forum,Marseille, 1998.

[9] O. Kunze, U.T.P. Arnold, S. Waaske, “Devel-opment and Design of an Individual Blade Con-trol System for the Sikorsky CH-53G Helicop-ter”, 55th Annual Forum of the American Heli-copter Society, Montreal, 1999

[10] M. Müller, U.T.P. Arnold, D. Morbitzer, “On theImportance and Effectiveness of 2/rev IBC forNoise, Vibration and Pitch Link Load Reduc-tion”, 25th European Rotorcraft Forum, Rome,1999

[11] A. Haber, S.A. Jacklin, G. deSimone, “Devel-opment, Manufacturing, and Component Test-ing of an Individual Blade Control System for aUH-60 Helicopter Rotor”, AHS Aerodynamics,Acoustics, and Test and Evaluation TechnicalSpecialists Meeting, San Francisco, 2002

[12] P. Richter, W. König, “Einzelblattsteuerung(IBC) für den Transporthubschrauber CH-53G“,Deutscher Luft- und Raumfahrtkongress, Ham-burg, 2001

[13] U.T.P. Arnold, G. Strecker, “Certification,Ground and Flight Testing of an ExperimentalIBC System for the CH-53G helicopter”, 58thAnnual Forum of the American Helicopter Soci-ety, Montreal, 2002

[14] S.A. Jacklin, A. Haber, G. de Simone, et al.,“Wind Tunnel Test of a UH-60 Individual BladeControl System for Adaptive Performance Im-provement and Vibration Control“, 58th AnnualForum of the American Helicopter Society,Montreal, 2002

[15] W. Johnson, W., “Self-Tuning Regulators forMulticyclic Control of Helicopter Vibrations”,NASA Technical Paper No. 1996, 1982.

[16] G. Lehmann; R. Kube, “Automatic VibrationReduction at a Four Bladed Hingeless ModelRotor – A Wind Tunnel Demonstration”, VerticaVol. 14, No. 1, pp. 69-86, 1990.

[17] T. Millot, W. Welsh, “Helicopter Active Noiseand Vibration Reduction”, 25th European Ro-torcraft Forum, Rome, 1999.

[18] D. Fürst, T. Auspitzer, M. T. Höfinger, B. G. vander Wall, “Numerical Investigation of VibrationReduction Through IBC for a 20to HelicopterRotor Model”, 28th European Rotorcraft Forum,Bristol, 2002

[19] C. Kessler, D. Fürst, U. Arnold; “Open LoopFlight Test Results and Closed Loop Status ofthe IBC System on the CH-53G Helicopter”,59th Annual Forum of the American HelicopterSociety, Phoenix, Arizona, 2003

37-15