IIIIIIIIIIIIIIfllfllfl IIIIIIIIIII - dtic. · PDF fileas a means to reduce the levels of...

AD-AO98 794 BELL HELICOPTER TEXTRON FORT WORTH TX F/6 1/3FLIGHT TEST EVALUATION OF A NONLINEAR HUB SPRING ON A UIJI .H EL--ETC(U)APR A1 P J HOLLIFIELO, L A DOOLEY OAAJ02-77-C 006B

UNCLASSIFIED USAAVRADCOM TR Al0 D27 NI

I EUE~hh

IIIIIIIIIIIIIIfllfllflIIIIIIIIIII.II

&611 118 113 5liilino

IWA 11420i IIi

1.2 flfI -AI 11111.6

MICROCOPY RESOLUTION TEST CHART

LEVE.I*AAVRADcOM-T-6.D2

AD A098794 W

FIJ TEST EVALUATION OF A NONUNEAR HUB SPRINGO. A UWIN HELICOPTER"

P. J. s , L W. Doole, J. R. Van GosbsP. a', I410 osFort *Ort. Tan 76101 ,,S

'~ Apri1981 ~C

Final Repor

Approved for public release;distribution unlimited.

A~ftIED TECHNOLOGY LABORATORYU. & AJJMj Y RIESEARCH AND TECHNOLOGY LABORATORIES (AVRADCOM)

. g wk, Va. 23604

81 ~ ~ (p4

7.

APPLIED TECHNOLOGY LABORATORY POSITION STATEMENT

This report documents the engineering analysis, development, arnd flight test of a non-linger hub restraint mechanism for the UH-1H helicopter. The hub restraint consist ofan siastomeric spring unit which imparts a restoring moment to the teetering rotor hubwhen deflected through the flapping angles. In all flight conditions tested the hubspring Is shown to reduce the probability of teetering rotor hub contact with tt~4torshaft or mast by reducing main rotor f lapping.

The results contained in this report effectively support the nonlinear hub spring deviceas a means to reduce the levels of teetering rotor flapping that could lead to mastbumping and/or airframe blade strike& The report further identifies a proposed mastdesign criterion to ensure that mast loads can be sustained and structural deformationmade improbable during possible inflight flapping stop contact under operational ormaterial failure conditions which the aircraft could otherwise be reasonably expected tosurvive.

Mr. G. T. White of the Aeromechanics Technical Area, Aeronautical Technology Divisioo,served as project engineer for this effort.

DI5CLAIMRS

The Wlng In this eort e not to be constued a an ofiial Departmnt of the Anmy poition unles sod1 s es by oilie mthorizad document

Whe Govrmnt dravinMs esfione. or 96he dmt e used for amy purpos oter the n In coetionwI adob* mem Gennetprocuremnt oeraton. lie Un~e base GovarrAnent thev lanr no

* ~~~roeponslblt nor any oblgation ediee mer: ed o fist Owat Ow vmmmn may hav formiead. fumlld** ~~~or In any m plebd the sad drarangs. waelficeons. or othe dam Is not to be regade by ipoatom a'

adhrseU a In any mnner Ilcnsing the holde or a Ote pason or onxoetlon, or conveyig any rofp orperission, to manufature ua, or sai any patete Invnton tha may In any wayb he wrld teo

Trad name 01We In this reor do not constitt an official endorsement or appro-i of the ms of rmdicmeldherdwwor or software.

DISPOSITION ItSf~c~N

Detw this woar vosw nolonger nosasd Do no retun It to th evoIgInato.1

.. 9

UNCLASSIFIEDSECURITY CLASSIFICATION or ris PIAGE (tm ON- 5rW-d _______________

REPOI ,W MENTATION PAGE 53 POlR COMPL&MIG PORM

~~~~I PErAuq-p 2wau.a GOVTJM ACRUO NO 10. R9ROGRAMLECATALOET TASK9

Bell- HeicperTetonD-j 620A1L62For Wot ea 611___________

41. PFFRMN OWEP RA11.1111111

Aproe for .... puli releaseCT distribution unlimited.

7. JISTRIUUiON SlAT dN 20s YA.A 2... mee 7i 65S.Ifdjetf

Restraeit, at Leoad niuainHl#pe iultoFapn

marti for th exa- eiote speetd Atog r s a60

neped oredditol tesortng, the .huAsri is hown to poIdeaiResedargi&,ecnolsoey Lbyratrein -ain rotor flpinMnl

condition tested 158f

(ARDCM Fort Eustis, Virg LISS mis

1MNv CLASSIFICAION OF THIS PAGE1UM as sbus

As a part of this effort, a hybrid computer program was vori-fied as able to predict mast loads due to flapping stop contact.Using this program a parametric study of mast loads as a func-tion of rotor flapping was performed in order to develop adesign criteria to ensure that mast loads can be sustained duringin-flight flapping stop contact.

In addition, a comparison of the main rotor flapping predictedby the hybrid computer and the digital computer C81 programusing elastic blades is shown. Also, an evaluation of U. S.Army helicopter tactics to determine which NOE maneuvers aresusceptible to high main rotor flapping is presented.

UNCLASSIFI3OSSCUOITY CLASSIPICATIO OF TI4I PAOEGbmml DIM w

PREFACE

The work reported herein was performed by Bell HelicopterTextron (BHT) under Contract DAAJ02-77-C-0064, "Proposal forCriteria and Concept for Improved UH-1 Mast Bumping SafetyMargin," awarded in September 1977 by the Applied Techno-logy Laboratory (ATL) of the U.S. Army Research and Techno-logy Laboratories (AVRADCOM).

Technical program direction was provided by Mr. G. T. Whiteof the ATL. Principal Bell Helicopter Textron personnelassociated with the program were Messrs. P. J. Hollifield,L. W. Dooley, J. R. Van Gaasbeek, J. D. Honaker, J. Carr,K. G. McEntire, and J. P. Norvell. - .. ..

Aoession For

NTIS 0"1&DTIC TAB Q

Unianouncedjust tflcatio

Distribu ton/

Availability CodesA;i.!{1 and/o-

Dist SpocIa 1

3

TABLE OF CONTENTS

Page

1. INTRODUCTION .......................................... 14

1.1 OBJECTIVES ........ * ........ ......................... 141.2 TECHNICAL APPROACH ............................... 15

2. MAST LOADS SIMULATION .................................. 17

2.1 PROGRAM MODIFICATIONS..............................192.2 PROGRAM VERIFICATION ............................. 212.3 UH-lH SIMULATIONO................................. 31

2.3.1 UH-lH Model ................................ 31

2.3.2 Preflight Simulation ...................... 33

2.4 UH-IH HUB SPRING POST-FLIGHT ANALYSIS ............ 39

2.4.1 Definitions and Sign Conventions .......... 392.4.2 Measured versus Computed Mast Bending

Moments ................................... 42

3. HUB SPRING DESIGN ...................................... 51

3.1 DESIGN CHARACTERISTICS ........................... 513.2 FAILURE MODES ..................................... 62

4. DESIGN LOAD CRITERIA ............................. 63

4.1 PARAMETRIC STUDY ................................. 63

4.1.1 Variation of Linear and Nonlinear Hub Re-straint ......................... .............. 63

4.1.2 Variation of Pylon Mount Spring Rate ...... 654.1.3 Variation of Mast Stiffness ............... 694.1.4 Summary of Variation of Parameters ........ 69

4.2 STOP CONTACT LOAD PREDICTION ..................... 69

4.3 MAST LOAD CRITERIA ............................... 75

5. ELASTIC ROTOR EFFECTS .......... ..................... 78

S5 .1 STEADY FLIGHT CONDITIONS ................ 785.2 TRANSIENT CONDITIONS ............................. 79

5

now -

TABLE OF CONTENTS (Continued)

Page

6. TACTICS EVALUATION .................................... 89

6.1 EVALUATION OF TACTICAL LITERATURE ................ 896.2 UH-1 NOE COURSE...................................896.3 TACTICS EVALUATION SUMMARY ....................... 91

7. FLIGHT TEST PROGRAM ................................... 92

7.1 PROGRAM OBJECTIVE ................................ 927.2 EFFECT OF THE HUB SPRING ON MAST BUMPING SAFETY

MARGIN ........................................... 92

7.2.1 Main Rotor Flapping Comparisons DuringSteady Flight Conditions .................. 94

7.2.2 NOE Maneuver Flapping Comparison .......... 997.2.3 Steady High Flapping Conditions ........... 1027.2.4 Low-g Flight Control Power ................ 1077.2.5 Summary of Improved Mast Bumping Safety

Margin Tests .............................. 107

7.3 EFFECT OF HUB SPRING ON HANDLING QUALITIESCHARACTERISTICS .................................. 114

7.3.1 Control Margins during Steady Flight ...... 1147.3.2 Static Longitudinal Stability ............. 1147.3.3 Maneuvering Stability ..................... 1147.3.4 Controllability ............................1197.3.5 Dynamic Stability .......................... 1197.3.6 General Handling Qualities Summary ........ 127

7.4 EFFECT OF HUB SPRING ON VIBRATIONCHARACTERISTICS .................................. 132

7.4.1 Center-of-Gravity Vibration Characteris-tics ...................................... 132

7.4.2 Crew Station Vibration Characteristics.... 1327.4.3 Summary of Vibration Characteristics ...... 132

7.5 EFFECT OF HUB SPRING ON MAIN ROTOR COMPONENTFATIGUE LIVES .................................... 137

7.5.1 Fatigue Damage During Normal Operation .... 1377.5.2 Fatigue Evaluation of Complete Test

Program ................................... 1417.5.3 Fatigue Evaluation Summary................ 1467.5.4 Recommendation ............................ 146

6

TABLE OF CONTENTS (Concluded)

Page

7.6 PROBLEMS ENCOUNTERED AND SOLUTIONS ............... 147

7.6.1 Structural Bond Failure ................... 1477.6.2 Abrasion and Tearing of Elastomerics ...... 1477.6.3 Excessive Stress in Side Member ........... 1487.6.4 Fretting of Mast and Yoke ................. 1487.6.5 Transmission Structural Analysis .......... 148

7.7 SUMMARY OF FLIGHT TEST RESULTS ................... 149

8. CONCLUSIONS AND RECOMMENDATIONS ....................... 150

8.1 CONCLUSIONS ...................................... 1508.2 RECOMMENDATIONS .................................. 151

9. REFERENCES ............................................ 152

10. LIST OF ABBREVIATIONS AND SYMBOLS ..................... 153

APPENDIX A - TABULATION OF ARHF01 INPUTS FOR THE UH-1HWITHNONLINEARHUB SPRING .................... 155

APPENDIX B - QUALITATIVE PILOT FLIGHT REPORT ON THE UH-lNONLINEAR HUB SPRING ........................ 157

7

... .. .- ,- . .... .. -.

LIST OF ILLUSTRATIONS

Figure Page

1 Rotor blade-element aerodynamic coefficientsused in ARHF01 .................................. 18

2 Rotor model on test stand ....................... 22

3 Rotor model and flapping stops .................. 25

4 Analytical model for determination of flappingstop spring rate ................................ 26

5 Comparison of measured and computed upper mastbending moment, model rotor, 075=00, Bi=7.50 .... 27

6 Comparison of measured and computed upper mastbending moment, model rotor, 075-30, B1=7.5 0 .... 28

7 Comparison of measured and computed upper mastbending moment, model rotor, 75=00.............. 29

a Comparison of measured and computed upper mastbending moment, model rotor, 075=30 ................ 30

9 Total hub restoring moment due to hub spring .... 34

10 Computed oscillatory mast bending moment atflapping stop contact point, UH-lH main rotor,2500-pound thrust, hover ........................ 35

11 Computed oscillatory mast shear at flappingstop contact point, UH-lH main rotor, 2500-pound thrust, hover ............................. 36

12 Computed oscillatory mast bending moment atflapping stop contact point, UH-lH main rotor,8500-pound thrust, hover ......... ............... 37

13 Computed oscillatory mast shear at flappingstop contact point, UH-lH main rotor, 8500-pound thrust, hover ............................. 38

14 Main rotor mast bending moment sign convention.. 41

15 Measured and computed main rotor mast Station16 parallel bending moments for the tie-downtest ............................................ 44

8

LIST OF ILLUSTRATIONS (Continued)

Figure Page

16 Measured and computed main rotor mast Station46.75 parallel bending moments for the tie-down test ....................................... 45

17 Measured and computed main rotor mast Station 16parallel bending moments at 66 KTAS ............. 47

18 Measured and computed main rotor mast Station46.75 parallel bending moments at 66 KTAS ....... 48

19 Measured and computed main rotor mast Station 16parallel bending moments at 118 KTAS ............ 49

20 Measured and computed main rotor mast Station46.75 parallel bending moments at 118 KTAS ...... 50

21 Hub moment provided by nonlinear hub spring ..... 53

22 Drawing of UH-lH nonlinear hub spring ........... 55

23 UH-iH nonlinear hub spring......................56

24 UH-IH nonlinear hub spring attached to mainrotor yoke ................... ........................ 58

25 UH-lH nonlinear hub spring installation ......... 60

26 Loads at mast contact point for baseline UH-H.. 64

27 Loads at mast contact point for baseline UH-lHwith proposed nonlinear hub spring .............. 66

28 Loads at mast contact point for baseline UH-lHwith nonlinear hub spring (half proposed springrates) ............................................... 67

29 Loads at mast contact point for baseline UH-lHwith nonlinear hub spring (double proposedspring rates) .................. ...................... 68

30 Loads at mast contact point for baseline UH-lHwith half pylon mount stiffness ................. 70

31 Loads at mast contact point for baseline UH-1Hwith double pylon mount stiffness ............... 71

9


Figure Page

32 Loads at mast contact point for baseline UH-1Hwith half mast stiffness ........................ 72

33 Loads at mast contact point for baseline UH-lHwith double mast stiffness ...................... 73

34 Effects of elastic main rotor blades on heli-copter parameter predictions during simulatedlevel flight .................................... 80

35 Effects of elastic main rotor blades on heli-copter parameter predictions during a simulatedpull-up ......................................... 84

36 Effects of elastic main rotor blades on heli-copter parameter predictions during a simulatedpushover ...................................... 87

37 Bell Helicopter Textron UH-iH ................... 93

38 Effect of the hub spring on main rotor flappingin rearward and forward flight at the most cri-tical loading condition ......................... 95

39 Effect of the hub spring on main rotor flappingin sideward flight .............................. 96

40 Effect of the hub spring on main rotor flappingduring autorotation ............................. 97

41 Effect of the hub spring on main rotor flappingduring MCP climbs ............................... 98

42 Effect of the hub spring on main rotor flappingduring maneuvering stability testing ............ 100

43 Effect of the hub spring on main rotor flappingand helicopter response during a right sidewarddash ............................................ 103

44 Effect of the hub spring on main rotor flappingand helicopter response during a level flightroll reversal entered at 80 knots ............... 105

10

7=7 .. .


Figure Page

45 Effects of the hub spring on main rotor flappingand helicopter response during a roller coastermaneuver to low-g ............................... 109

46 Effect of the hub spring on main rotor flappingand helicopter response following a leftcontrol step at low-g ........................... 111

47 Effect of the hub spring on lateral sensitivityat g-levels less than 1.0 ....................... 113

48 Effect of the hub spring on longitudinal controlpositions during level rearward and forwardflight at the most critical loading condition... 115

49 Effect of the hub spring on control positionsduring level sideward flight .................... 116

50 Effect of the hub spring on static longitudinalstability in level flight ....................... 117

51 Effect of the hub spring on maneuvering stabil-ity at 0.6VNE and VNE ........................... 118

52 Effects of the hub spring on main rotor flappingand helicopter response to an aft cycliccontrol step at OGE hover ....................... 120

53 Effects of the hub spring on main rotor flappingand helicopter response to a right lateralcyclic control step at OGE hover ................ 122

54 Effects of the hub spring on main rotor flappingand helicopter response to a left lateral cycliccontrol step at 0.6VNE .......................... 124

55 Effect of hub spring on control sensitivity anddamping ......................................... 126

56 Effects of the hub spring on main rotor flappingand helicopter response to a left lateral cycliccontrol pulse at 0.6VNE ......................... 128

57 Effects of the hub spring on main rotor flappinghelicopter response to an aft cyclic controlpulse at 0.6VNE ................................. 130

~11

LIST OF ILLUSTRATIONS (Concluded)

Figure Page

58 Effect of the hub spring on center-of-gravity2-per-rev vibration acceleration in levelflight .......................................... 133

59 Effect of the hub spring on crew station2-per-rev vibration acceleration in levelflight .......................................... 135

12

.. --.. T ' "

+ _ ___i__ _ . -. . ~. ' . .Ty- ...-- -77- . -

LIST OF TABLES

Table Page

1 MODEL CHARACTERISTICS ............................. 23

2 UH-lH CHARACTERISTICS ............................. 32

3 PARAMETERS OF TEST CONDITIONS SIMULATED ......... 40

4 DESIGN FLAPPING SPECTRUM .......................... 52

5 EFFECTIVE FLAPPING STOP SPRING RATE .............. 77

6 MODE SHAPES ..................................... 78

7 GROSS WEIGHT AND CENTER-OF-GRAVITY CONDITIONS... 79

8 U.S. ARMY TACTICAL MANUALS ........................ 90

9 EFFECT OF THE HUB SPRING ON MAIN ROTOR FLAPPINGAND AIRCRAFT RESPONSE DURING NAP-OF-THE-EARTH(NOE) MANEUVERS ................................... 101

10 STEADY HIGH FLAPPING CONDITIONS ................. 108

11 SUMMARY OF MAIN ROTOR BLADE FATIGUE DAMAGEDURING FLIGHT WITHIN OPERATIONAL ENVELOPE ....... 138

12 SUMMARY OF MAIN ROTOR MAST FATIGUE DAMAGEDURING FLIGHT WITHIN OPERATIONAL ENVELOPE ....... 139

13 SUMMARY OF ROTOR PITCH HORN FATIGUE DAMAGEDURING FLIGHT WITHIN OPERATIONAL ENVELOPE ....... 140

14 SUMMARY OF MAIN ROTOR BLADE FATIGUE DAMAGEDURING HUB SPRING TEST PROGRAM ................... 142

15 SUMMARY OF YOKE FATIGUE DAMAGE DURING HUBSPRING TEST PROGRAM .............................. 143

16 SUMMARY OF MAIN ROTOR MAST FATIGUE DAMAGEDURING HUB SPRING PROGRAM ........................ 144

17 SUMMARY OF MAIN ROTOR PITCH HORN FATIGUEDAMAGE DURING HUB SPRING PROGRAM ................. 145

13

.. . . . .* ..

1. INTRODUCTION

Since 1972, when a mast bumping problem was identified by theU.S. Army, studies have been conducted to better define theproblems of high main rotor flapping, to determine the effects ofthe operational envelope limits of flapping, and to design andtest concepts to improve the mast bumping safety margin of UH-1 helicopters. In 1975, the U.S. Army awarded Bell HelicopterTextron (BHT) a contract (DAAJ02-75-C-0030) to investigate anddefine rotor blade flapping criteria. During this investiga-tion, computer simulation was used to determine flight condi-tions, pilot control inputs, and helicopter characteristicsinfluencing main rotor flapping. The findings and recommenda-tions of this investigation are presented in Reference 1. Afollow-up contract (DAAJ02-76-C-0043) was awarded to BHT bythe U.S. Army in 1976 to study and report on the effect ofoperational envelope limits on teetering rotor flapping. Thisstudy was to verify the accuracy of flapping predictions givenby computer simulation, to investigate the effect of operationoutside the recommended flight envelope, and to investigatemethods of extending this envelope. Reference 2 presents thefindings and recommendations of this study. Following the re-commendations of Reference 2, a third contract (DAAJ02-77-C-0064) was awarded to BHT by the U.S. Army in 1977. This studywould seek to develop criteria and a concept for improving theUH-1 helicopter's mast bumping safety margin. The results ofthis work are presented in this report.

1.1 OBJECTIVES

The objectives of this contracted study were developed fromthe results of previous work on the subject of main rotorflapping. These objectives were:

- Verify, by comparison with rotor model tests, the abilityto calculate loads due to flapping stop contact.

'Dooley, L. W., ROTOR BLADE FLAPPING CRITERIA INVESTIGATION,Bell Helicopter Textron, USAAMRDL Technical Report 76-33,Eustis Directorate, U.S. Army Air Mobility Research andDevelopment Laboratory, Fort Eustis, Virginia, December 1976,AD A034459.2Dooley, L. W., and Ferguson, S. W. III, EFFECT OF OPERA-TIONAL ENVELOPE LIMITS Off TEETERING ROTOR FLAPPING, BellHelicopter Textron, USARTL Technical Report 78-9, AppliedTechnology Laboratory, U.S. Army Research and TechnologyLaboratories, Fort Eustis, Virginia, July 1978, AD A059187.

14

- Design and procure a nonlinear hub spring for the UH-1helicopter.

- Formulate a design load criteria for flapping stop con-tact loads.

- Evaluate elastic rotor effects on flapping predicted bycomputer simulation.

Evaluate current U.S. Army Nap-of-the-Earth (NOE) tacticsfor possible high flapping maneuvers.

Verify by flight test the UH-1 nonlinear hub spring as aconcept to improve the mast bumping safety margin by re-ducing main rotor flapping and increased flapping-stop-to-mast clearance.

1.2 TECHNICAL APPROACH

In order to define the load generated during flapping stopcontact, the BHT rotor/pylon hybrid computer program (ARHM20)was modified and used to investigate the loads on both modeland full-scale masts during simulated conditions of in-flightflapping stop contact. In order to obtain flapping and mastcontact loads data at simulated flight conditions, aeroelasticmodel rotor whirl-stand tests were conducted. Correlationbetween the aeroelastic model rotor and the rotor/pylon hybridcomputer program was demonstrated.

After establishing that the hybrid program could adequatelycorrelate measured and predicted mast loads, it was used toconduct a parametric analysis of stiffness parameters in theU.S. Army UH-lH helicopter rotor and pylon system. From thisanalysis, a design load criterion reflecting inadvertent flap-ping stop contact was defined.

A nonlinear hub spring was subsequently designed for the UH-lHhelicopter rotor system. Design considerations includedflapping value for flapping stop snubber engagement, shape ofthe snubber, shear pad, snubber spring rates, and springmounting arrangement. A structural analysis of the main rotormast, hub, and blades, which included moments introduced bythe nonlinear hub spring, followed to assure a positive struc-tural margin of safety. After U.S. Government approval, thedesign was released for competitive procurement.

Prior to defining a flight test plan, the following tasks wereaccomplished. First, comparisons were made between the hubflapping predicted by the "rigid blade" hybrid version of theRotorcraft Flight Simulation Computer Program (C81) and the

digital C81 with elastic blades in order to determine theeffect of elastic blades on predicted main rotor blade flap-ping. Also, current Army helicopter tactical instructionmanuals were analyzed to determine the impact of these tacticson excessive flapping. Finally, an observation flight in anArmy UH-lH helicopter over the nap-of-the-earth course at FortRucker, Alabama, was made to determine if any typical UH-lHoperational NOE maneuvers are susceptible to high flapping.

A flight test program was then developed that would determinethe effectiveness of a nonlinear flapping restraint in improv-ing the UH-l mast bumping safety margin. The test aircraftwas instrumented to measure main rotor mast, hub, and bladeloads, handling qualities parameters, and vibration levels.

Results of the mast loads simulation are presented in Section2, the hub spring design in Section 3, and the design loadcriteria in Section 4. The effects of the elastic rotor onmain rotor hub flapping prediction are presented in Section 5and the evaluation of NOE tactics in Section 6. The flighttest program and the results of that program are presented inSection 7.

16

2. MAST LOADS SIMULATION

Calculation of mast loads due to rotor flapping stop contactrequires a complex simulation model. An existing BET hybridcomputer program (ARHM20) was modified to the necessary degreeof sophistication and was subsequently used to compute mastbending moment and transverse shear during flapping stopimpact. The modified program (named ARHF01) simulates ahelicopter in either hover or forward flight, and consists ofelastic rotor, elastic fuselage, and elastic landing gearmodels, which are dynamically coupled. The models used in theprogram and an input guide are given in Reference 3.

Modeling Considerations

The rotor model in ARHF01 uses three-quarter radius aerody-namics and a maximum of three rotor elastic modes. The blade-element lift and drag coefficients are functions of angle ofattack and are built into the program using limiter circuitry.These two relationships are presented in Figure 1.

Rotor dynamics are accounted for by using a rigid body flap-ping mode, a coning mode, and an inplane mode. The latter twoare represented by a rigid blade articulated at separate, off-set, coning, and inplane hinges. The natural frequency ofthese two modes is computed using the nonrotating frequencyand the Southwell coefficient for each mode. All three modesare input as uncoupled modes (i.e., the out-of-plane mode hasno inplane deflection) and are coupled in the equations ofmotion through the aerodynamics and inertia loads. Rotorelastic torsion is ignored. Coupling between the rotor andthe pylon structure is also accounted for in the program.Undersling, precone, mast tilt, and 63 are included in therotor model.

Up to six elastic modes can be used to represent the airframe.The airframe modes include five components of displacement atthe hub so that coupled rotor-pylon dynamics are modeled (theyaw displacement at the hub is ignored). Mode shape compo-nents at selected points in the airframe can also be input sothat fuselage vibration can be computed. Control coupling isincluded as is aerodynamic coupling arising from blade veloc-ity components due to pylon motion. The landing gear modesare dynamically coupled with the airframe, and this model canbe used for studying landing dynamics.3VanGaasbeek, James R. and Landry, L. Matthew, Jr., HYBRIDCOMPUTER PROGRAM ARHF02, Technical Data Report 699-099-015,to be published by Bell Helicopter Textron, Fort Worth,Texas.

17

4Ji

* 1.0

/U

-40 '30 --20 10 0 0 20 30 40

Angle of Attack, degrees

%4 1 .5 w .aS0.5-

-40 -30 -20 -10 10 20 30 40

* Angle of Attackc, degrees

Figure 1. Rotor blade-element

aerodynamic

coefficients used in AREF01.

.18

.

2.1 PROGRAM MODIFICATIONS

Several modifications were made to the ARHN20 program toenhance the simulation model and to facilitate program utili-zation. In the former category, the ability to calculate mastloads at up to three points was incorporated, and a nonlinearflapping spring model was added. One of the three mast pointsfor load calculations is the flapping stop contact point.The loads are:

S = (-H cos + Y sin )+ aFSRM(la)

M = (-H cos 4 + Y sin t) (XLFS) -MTOT - MFS (lb)

where S and M are the mast shear and bending moment in therotating system. H and Y are the hub inplane shears, includ-ing inertia loads due to rotor hub accelerations, in the non-rotating system, and MTOT is the externally applied hub flap-ping moment due to flapping restraint. XLFS is the distancefrom the teetering axis to the flapping stop contact point.

MFS is the externally applied moment due to the flapping stopcontact, which is modeled as a sprinq that is engaged when P(hub flapping) exceeds PS (the flapping stop angle). The re-sulting moment is

MFS'= KFS P* (2a)

where

P* = - PS} sign (P), 0 > PS (2b)

=0 p < PS (2c)

In this model, the hub actually "penetrates" the mast bYhaving P > PS Sign conventions and geometry are given

in the following sketch.

19

Teetering (-H Cos 0 + Y sin#)

s

MFS ----- SM

S

In like manner, the mast moment at any user-selected pointsbelow the flapping-stop contact point is computed as

N = MTOT + MFS + (-H cos + Y sin*) X (3)

in which MTOT and MFS are as defined for Equations (la) and(lb), and X is the distance from the teetering axis to thepoint at which the moment is computed.

The flapping spring model in the ARHF01 program was modifiedto add a nonlinear spring to the linear spring already in theprogram. The nonlinear model chosen was that in the digitalversion of the Rotorcraft Flight Simulation Program C81(Reference 4). The total flapping spring moment is

MTOT ~KL II< j (4a)

=KL 0+ KNL (P 0r-1 ~B r) PIi > jp ,j (4b)

4 VanGaasbeek, James R., ROTORCRAFT FLIGHT SIMULATION COMPUTERPROGRAM C81, WITH DATAMAP INTERFACE, Volume I - User's Man-ual, to be published by the United States Army Research andTechnology Laboratories, Applied Technology Laboratory, FortEustis, Virginia.

20

- 7777 7r 7-

where

KNL is the nonlinear spring rate

OB is the flapping angle at which the nonlinear springis engaged

r is the order of the nonlinearity

In addition to these analytical enhancements, the program wasmodified to improve its ability to run without saturating theanalog amplifiers by allowing scaling changes to be made asinputs instead of requiring a recompilation of the FORTRAN pro-gram on the digital computer. The program was also modifiedto permit the user to specify a piecewise-linear or sinusoidaltime history of the controls. Lastly, the ARHF01 program wasmodified to output certain time-history variables on a digitaltape for tabulation or CALCOMP plotting at a later time.

2.2 PROGRAM VERIFICATION

A series of whirl-stand'tests were conducted with a model rotorto provide data for verification of the ARHF01 program. Asthe model rotor was assembled from components fabricated forprevious test programs, it was not an exact scale model of anyparticular production rotor. Certain components were modifiedto attempt to make the model representative of a UH-1-typerotor. The model characteristics are summarized in Table 1.Figure 2 shows the rotor mounted on the test stand.

The model rotor hub is fabricated from aluminum with a beam-wise flexure, giving a coning virtual hinge offset of 1.35inches. Feathering is accomplished through a spindle arrange-ment on the grip. The blade is constructed of a twisted ta-pering S-glass spar with the airfoil shape formed by balsaglued to the spar. A fiberglass skin has been applied to theleading edge and segments of music wire are embedded in theleading edge to provide mass balance.

The test stand consists of a focused, sprung pylon with aneight horsepower hydraulic motor and rise-and-fall swashplate.The mast was made from a section of drill rod.

Rotor structural properties were determined either experi-mentally or analytically. The rotor component masses weredetermined by weighing a blade and the entire rotor.

21

I

. Pylon -Sws l T M

AM.' Swash t

Hydraulici

~motor

'IlI

Figure 2. Rotor model on test stand.

TABLE 1. MODEL CHARACTERISTICS

Radius, inches 36.0

Rotational speed, rpm 800

Undersling, inches 0.65

Precone, degrees 2.5

Pitch-flap coupling, degrees 0.0

Blade chord, inches 2.62

Twist, degrees -8, linear

Airfoil NACA 0012

Coning frequency, per rev 1.06

First inplane frequency, cantilevered,per rev 1.39

Pylon longitudinal frequency, per rev 0.668

Pylon lateral frequency, per rev 0.679

Distance between teetering axis andflapping stop contact point, inches 1.6

Flapping stop angle, degrees 4.5

23

The mass of the hub plus pitch horn was derived from thesemeasurements and assumed to be uniformly distributed. The hubwas subjected to a load deflection test and a beam and chordstiffness were computed from the tests. The blade stiffnessdistributions were calculated using the dimensions and mate-rial properties of the spar and fiberglass nose cap. Due toits very low modulus, the contribution of the balsa to theblade stiffness was neglected. The Rotor Frequency ProgramDN9100 was used to find the rotor natural frequencies and modeshapes, with the latter used to locate the rotor inplane andvirtual hinge location. The out-of-plane virtual hinge loca- rtion was assumed to be at the center of the flexure, a suppo-sition borne out by load-deflection data.

The pylon natural frequencies were determined by placing alongitudinal accelerometer at the top of the mast, pluckingthe mast longitudinally, and observing the resulting frequencyin the accelerometer trace. An identical test was conductedfor the lateral frequency. As only one component of the modeshape was measured, and six were needed for the mathematicalmodel, a NASTRAN model of the pylon was constructed and aneigenvalue analysis performed. The computed frequencies werein good agreement with those measured. The computed modeshapes and measured frequencies were used in the simulation.

A flapping stop was constructed by attaching a machined alumi-num block to the mast and attaching a block with setscrews tothe rotor hub. The setscrews were adjusted to give a flappingstop angle of 4.5 degrees, and the flapping stop contact pointwas 1.6 inches below the teetering axis (see Figure 3).

The flapping stop spring rate was determined analyticallyusing the previously constructed NASTRAN model of the pylon.Referring to Figure 4, two opposing shears (S) were applied,one at the teetering axis and one at the flapping stop loca-tion, in order to model the couple due to flapping stop im-pact. The spring rate was computed from the angular deflec-tion (0) of the shaft at the teetering axis. The resultingspring rate was found to be 71.68 feet-pound/radian.

The rotor hub was instrumented for critical beam and chordbending moments, which were monitored during the test todetermine if stress allowables were being exceeded, and a po-tentiometer was installed to measure hub flapping. Tip-path-plane flapping was recorded on videotape. Strain gages wereapplied to the mast 2.8 and 11.8 inches below the teeteringaxis to measure mast bending moments. Pylon longitudinal and

24

dMan

Flapping Stop

ii

ii.

Figure 3. Rotor model and flapping stops.

25

lateral rocking angles were also measured. All eight quan-tities were recorded on an oscillograph for runs made at 800rpm and e75 = 0.0 and 3.0 degrees and values of B, (at the

hub) between 0 and 8.25 degrees. Several cases of severeflapping stop impact were encountered.

'~4LLS

S XLFS

M S * XLFS

K FS M/0

Figure 4. Analytical model- for determination offlapping stop spring rate.

Hybrid computer simulations were run for the same cases inorder that bending moment and shear time histories could becompared. Figures 5 and 6 present a comparison of the uppermast bending moment time histories for two cases with flappingstop contact. The measured and computed time histories arein good qualitative agreement, with almost identical oscilla-tory (one-half peak-to-peak) amplitudes. Adjusting the testdata for the obvious zero shift (the steady moment should bezero), the measured and computed time histories then agreealmost exactly, except for a slight phase difference. (Thisphase difference may be due to a slight misplacement of theazimuth event instrumentation in the test.) The lack ofhigh frequency content in the computed time history is dueto the program's allowance for only the flapping, coning, andfirst inplane modes, in conjunction with the three-quarterradius aerodynamics.

The predicted and measured overall oscillatory (one-half peak-to-peak) moments are plotted versus rotor cyclic in Figures7 and 8, again with favorable agreement. The computer pro-gram underpredicts the moments at control angles, which result

26

60

i 500--

Test Data

40-

]1 II30 I I

"QI I I' I I I

* 20• I a

- I

-10 I Ia' I

I I

\I II \I / I \

-20 P."--Computed Data

-30 I~I I ~ II

-40- V

h Rotor

-50 Revolution I

Figure 5. Comparison of measured and computed uppermast bending moment, model rotor, e75 0o,B 1 7.50.

27

60-

50 TestData

40 ii .'40I

30 1"$4 I I I

I I I I I I

0 I I I I D

.6I I I I I- o I I I

SI I

I I I I I-10 I I I DataI I I Ij I

-40I II II I I

t Rotor-JRevolution

Figure 6. Comparison of measured and computed uppermast bending moment, model rotor, 875=3 ,Bl-7.5.

28

Adim"

540 Test Data

-- Computed Data

.4--.64 /

3

4 .1

6440. 2 --

04

0 - -

0 2 4

Rotor Flapping, deg

i Figure 7. Comparison of Measured and computed upper~~mast bending moment, model rotor, e75-0'

29

,. -,9m~:, +.+ '+. _ '- -. . .. - . . . .

5 0 Test Data-- Computed Data -

4.r.4

w 3i 0II

S 2 >11~

I /

r

0 0

O00(2 - I0 2 4 6 8

Rotor Flapping, deg

Figure 8. Comparison of measured and computed uppermast bending moment, model rotor, e75=3 .

30

A". ....... .. .... .. ...~ .. .... .. .. .....I1 l".. .l n + + ... ... .." +. ....]. .+" + " " .....

- . .. . . .. . . . ... . . . .-- - -

in hub flapping angles less than the flapping stop angle. Aninspection of Equation (ib) indicates that at these low flap-ping angles, the only contribution to mast bending is theinplane hub shear, which is not a directly measureable quan-tity. Apparently, the ARHF01 program is underpredicting theseshears. Since the interest in this study is in mast loadsduring flapping stop impact, it was thought that the data pre-sented in Figures 5 through 8 demonstrate good agreementbetween theory and test.

2.3 UH-lH SIMULATION

Program verification was followed by a simulation of flappingstop impact for a full-scale UH-lH helicopter. Inputs for theUH-lH were developed and used first to conduct a preflightanalytical investigation of mast loads due to flapping stopcontact with and without the hub spring. The same inputs werelater used for postflight correlation with data recordedduring the flight test program.

2.3.1 UH-lH Model

The UH-lH inputs are listed in Table 2, and a detailed inputdata set listing is shown in Appendix A. The majority of theinputs were derived from the geometric description of thehelicopter. A complete twenty-segment structural model of theUH-lH rotor blade was analyzed, using the Rotor FrequencyProgram DNAM05 (Reference 4, Section 10), to identify virtualhinge locations and natural frequencies. A NASTRAN model of aUH-lN helicopter was modified to yield a model of the UH-lHpylon, which was used to compute pylon frequencies and modeshapes and the flapping stop spring rate (130,000 feet-pound/radian). However, five items in the input data require fur-ther explanation.

First, the hub mass is input as zero because the NASTRAN modelused to generate the fuselage elastic modes had a point massat the center of rotation equal to the total rotor mass minusthe mass of the blades outboard of the zi plane virtual hinge.This ensures that the total rotor mass ip properly accountedfor in the pylon dynamic response.

Secondly, the coupled inplane natural frequency was checkedbefore data runs were conducted. This test is performed byincluding only the inplane rotor mode and the pylon modes inthe analysis and integrating the equation of motion. Theinplane mode is excited impulsively and the coupled naturalfrequency determined from the transient response. The South-well coefficient input for the inplane mode (KZEDAP) was setto yield the correct coupled inplane frequency.

31

TABLE 2. UH-1H CHARACTERISTICS

Radius, feet 24

Rotational speed, rpm 324

Undersling, inches 5.2

Precone, degrees 2.75

Pitch-flap coupling, degrees 0.0

Blade chord, inches 21.0

Twist, degrees -10, linear

Airfoil NACA 0012

Coning frequency, per rev 1.17

First inplane frequency, cantilevered,per rev 1.13

Pylon Longitudinal Frequency, per rev 0.594

Pylon lateral frequency, per rev 0.546

Distance between teetering axis andflapping stop contact point, inches 7.3

Flapping stop angle, degrees 11.5

32

7 . -. - . .. .. .

Figure 9 gives the spring rate versus flapping angle for thenonlinear spring, as well as the best curve fit for Equations(4). In the nomenclature of that equation,

KL = 300 (ft-lb)/deg/rotor

KNL = 0.100 (ft-lb)/deg)4 71/rotor

OB = 4 deg

r = 4.71

Fourth, since the landing gear model was not required for thisanalysis, no data were input for this part of the model. Thelanding gear data given in the appendix are the default valuesset by the program. Finally, fuselage aerodynamic data werealso unnecessary for the analysis and were not input. Thevalues printed are the internal defaults.

2.3.2 Preflight Simulation

The simulation performed prior to the flight test used thisUH-lH model in the form of a rotor mounted on a pylon, withoutan elastic fuselage. The rotor longitudinal cyclic pitch wasvaried to compute mast loads for steady conditions at variousflapping angles. The rotor was simulated at both 2500 and8500 pounds thrust, with three levels of flapping restraint -no hub spring, a linear hub spring, and a full nonlinear hubspring. (The linear-spring cases were run with a spring rateof 330 feet-pound/degree/rotor, i.e., the spring rate in thelinear range of the nonlinear spring.)

The oscillatory mast bending moment and shear are plotted inFigures 10 through 13 for the six different cases. Comparisonof the 2500- and 8500-pound thrust cases show qualitativeagreement, but lower mast loads were experienced for the 2500-pound cases because the inplane shear portion of the moment issignificantly lower.

As to be expected, increasing the flapping restaint moment, byadding springs, increases the mast loads for a given flappinglevel. However, the additional control power provided by thespring will tend to reduce the flapping required for a givenmaneuver, thus alleviating some of this increased load. Inaddition, if the hub restaint is designed to spread theseloads over a much larger bearing surface than that encounteredduring flapping stop impact with the standard rotor, the shearstresses in the mast will actually be lower.

33

12 -- - - -

10 -- /

8-

x

4-1

zDesign Spring Rate-

04 Linear Region--/.IJ

Curve Fit forAnalysis, Equation (4)

0 1

Flapping Angle, deg

Figure 9. Total hub restoring moment due to hub spring.

34

- - ' -o , . -. ." , , F 1•-, *. " - -

35 a s No f II - -

Rotor with Linear Hub Spring -

en_ Rotor with Full Nonlinearo 30 Hub Spring -I

xr-I

!j 25 1-1lapping stop -

0

'15

0 1

//

'4

44to 50 i

/# '

o /

/.s

0 00 10

Main Rotor Flapping, deg

Figure 10. Computed oscillatory mast bending moment atflapping stop contact point, UH-lH main rotor,2500-pound thrust, hover.

35

70 ___ -- Basic Rotor, No Hub Spring------Rotor with Linear Hub Spring

Rotor with Full Nonlinear

60 Hub Spring

0

/

x 50-

- - -- -Flapping st

0 400

34 -2-0- - -to

0 1

00

• 36

'U /H -

0 5 i015Main Rotor Flapping, deg

Figure 11. Computed oscillatory mast shear at flapping~stop contact point, UH-lH main rotor,i. 2500-pound thrust, hover.

~36

35 --- Basic Rotor, No Hub Spring

--Rotor with Linear Hub Spring

I Rotor with Full Nonlinear0 301 Hub Spring

x4 25

4

4) ~PpingStop20/,

//a' /

O-

15

/ .0

-r4

0 5 10 15


Figure 12. Computed oscillatory mast bending moment atflapping stop contact point, UH-1H main rotor,8500-pound thrust, hover.

37

el

60 - , .... asic' Rt-or,N-o-H-ub Spring -

-- - oo with Linear Hub spring- - -

_______Rotor with Full Nonlinear50- Hub Srin~L

xT

40- Flapping stop

---------------0

04J

10/

0

00 1


Figure 13. Computed oscillatory mast shear at flapping stopcontact point, UH-lH main rotor, 8500-poundthrust, hover.

38

WI

2.4 UH-lH HUB SPRING POSTFLIGHT ANALYSIS

The ARHF01 program was further verified through a postflightanalysis of the full-scale UH-lH helicopter with the hubspring installed. Flight test data for two forward flighttest points and one tie-down condition were simulated usingthe same inputs as discussed in Paragraph 2.3, with the addi-tion of a 5-degree-of-freedom rigid-body fuselage.

The three test conditions that were simulated are listed inTable 3. The hub spring was installed for all three cases.The amount of thrust generated during the tie-down test(Counter 616 of Ground Run 17) is unknown. The hub flappingangles are based on a harmonic analysis of the hub flappingtime history and are the coefficients in the positive Fourierseries for flapping, i.e.,

P(t) = 00 + Pis sin Qt + A, cos Qt + ... (6)

The tip-path-plane flapping assumes a different sign conven-tion; e.g., positive longitudinal TPP flapping is indicatedby a negative coefficient for the cosine term.

2.4.1 Definitions and Sign Conventions

Mast parallel and perpendicular bending moments were measuredat points 16.0 and 46.75 inches below the teetering axis. Themast parallel bending moment is measured by strain gagesmounted to the mast in a vertical plane parallel to the bladefeathering axes, while mast perpendicular bending moments aremeasured in a vertical plane perpendicular to the blade fea-thering axes. These moments and their sign conventions areshown in Figure 14.

The equations for these two moments are

M1 = MOT + (H coso - Y sinqi)x (7a)

M = -(H sin + Y cosf)x (7b)

where

MTOT is the moment applied at the top of the mast dueto flapping restraint and flapping stop contact

39

TABLE 3. PARAMETERS OF TEST CONDITIONS SIMULATED

Flight GR17 197A 197ACounter 616 746 748Flight Condition Tie-Down Level Flight Level Flight

Gross weight, lb * 8000 8000

Center of gravity, in. -- 144.5 144.5

Airspeed, KTAS 0 66.2 118.7

Euler pitch attitude,deg 0.3 4.73 1.46

s 1.136 -0.511 -0.746

0i -0.392 -2.625 -2.510

Density Ratio 0.9925 0.9391 0.9391

*The amount of thrust generated during the tie-down run isunknown.

40

4)0

o

0

t-o

0o >Al 4) 04cJ a) t

H-

4 4)

4-- 0

ro

414

4q 0 .I

to H I

00

00

00

5.4 4

00

441

-------------

H,Y are the longitudinal and lateral hub shears inthe fixed system, positive forward and to star-board, respectively

x is the distance from the teetering axis to thepoint at which the moment is measured

In the rotating system, in the absence of a hub spring, apositive parallel bending moment at a point in the mast arisesfrom a shear that is positive outward along the noninstru-mented blade. The primary aerodynamic portion of this shearis Lsinp (assuming that the lift is perpendicular to the tip-path plane). The remainder of this shear arises from theinertia loads in the blade spanwise direction.

In like manner, a positive mast perpendicular bending momentis generated in the mast due to a hub shear (in the rotatingsystem) that is positive toward the trailing edge of theinstrumented blade. The aerodynamic portion of this shear isthe difference in the drag between the two blades (D1 - D2).The remainder of the shear is due to inertia loads in theblade chordwise direction.

Simulations of each of the three test conditions were per-formed by setting the airspeed and fuselage Euler pitch angleto the values measured during the test, setting collective andcyclic pitch angles, and recording data after the transientresponse damped out. The recorded thrust and flapping angletime histories were then harmonically analyzed to determine ifthe desired mean thrust and one-per-rev flapping had beenachieved. If the difference between the desired and computedthrust was more than 200 pounds, or if the computed values of

and differed individually from the measured values bys c

more than 0.1 degree, the control position inputs were changedand the case was rerun.

2.4.2 Measured versus Computed Mast Bending Moments

Computed mast parallel and perpendicular bending momentshave been compared with those measured in flight, for bothmast Station 16.0 and mast Station 46.75. The computed mastparallel bending moments are in good agreement with the testdata; the computed mast perpendicular bending moments arenot in good agreement with the test data.

As was described in Paragraph 2.4.1, the mast parallel bendingmoment consists of a moment due to the hub spring and a momentdue to the thrust vector and rotor flapping. Since the hubflapping angle and the thrust were being matched for the

42

forward flight cases, the moment due to the hub spring and themoment due to shear resulting from the thrust vector tiltshould be an accurate representation of those portions of themeasured mast parallel bending moment. For example, thecosine one-per-rev component of the Station 16 mast parallelbending moment measured at 118 KTAS was -12643 inch-pounds.The same term computed using just the hub spring moment andthe thrust vector tilt is -15623 inch-pounds. The full ARHF01analysis computes a cosine one-per-rev term of -22200 inch-pounds, indicating that the spanwise inplane inertial termsare probably overstated in the program, and of the oppositealgebraic sign with respect to the inertia loads experienced

during the test.

Inspection of the sources of the mast perpendicular bendingmoment has shown that it arises from an inplane aerodynamicshear, equal to the difference in the total drag of the twoblades, and a shear due to the chordwise inertia loads. Thiscan be restated in symbolic form as

Shear = D1 - D2 + IL1 - IL2 (8)

Where the shear is positive toward the trailing edge of blade1, D1 and D2 are the drag forces of the two blades and IL1and IL2 are the inplane shears at the root of blades 1 and 2

due to inplane inertia response. The aerodynamic representa-tion of the rotor blade element drag is quite simple (Figure1), so the (D1 - D2 ) term will only be an approximation of

the correct value. In like manner, the blade inplane motionis represented by a simple hinge-spring-rigid blade mode, sothe (IL1 - IL2 ) term is also only an approximation to the

spanwise hub shear. The simplicity of the model results inincorrect or overestimated mast perpendicular bending moments.Therefore, there is a significant discrepancy between com-puted and measured mast perpendicular bending moments; theremainder of the discussion will consider only mast parallelbending moments.

Time histories of the computed and measured mast parallelbending moments are compared in Figures 15 and 16 for the tie-down test condition of Counter 616. Since the thrust actuallydeveloped by the rotor during the test is unknown, cases wererun at thrusts from 4000 to 10,000 pounds with both one-per-rev components of the flapping angle being matched to within0.1 degree. The computed mast parallel bending moment timehistories plotted are for the two extreme cases.

43

v-4

$4414

4'.9 E4

E4.

4.1

0 0440 P)4.

00W

Sn v- 4)4)4J

ju0g04J 4J14 to Id

tu -4 441 (d)IV 6i T

Q 4J4JC4J 0 15

IllWW 4WOO Isj 44JV

4) 00 2:3 :OE-w)0 M4

.4

444

In

V .m

'44

0 00)

040

0 - A4

0

U.1 I--*1

41 i4 *-* 1.4)Vwes 0

0100 4J4E~tJO -

45P

The computed parallel bending moment time histories for mastStation 16 (Figure 15) are almost exactly in phase with thetest data. The overall amplitude of the measured bendingmoment exceeds that computed at 4000 pounds thrust and is lessthan the amplitude predicted at 10,000 pounds thrust. (Notethat there is a subharmonic response present in the computedresults. This is due to a portion of the transient responsehaving very low damping. The investigators were unable toentirely eliminate this response.)

The measured and computed mast parallel bending moment timehistories for Station 46.75 are presented in Figure 16.Again, computed time histories are plotted for the two ex-treme thrust conditions simulated due to the uncertainty inthe thrust generated during the test. The computed timehistories exhibit the same phase, with respect to the azi-muth event indicator, as the measured time history. The am-plitude computed in the low thrust case is approximately thesame as that measured, while the amplitude of the time historygenerated in the high thrust simulation is approximatelydouble the measured quantity.

Figures 17 through 20 contain comparisons of the measuredand computed mast parallel bending moment time histories atStations 16.0 and 46.75 for the 66 and 118 KTAS test condi-tions (Counters 746 and 748). The computer-generated timehistories are in excellent agreement with the test data withregard to phase angle. The overall amplitude of the com-puted Station 16 bending moment is approximately 30 to 50percent greater than that measured, while the overall ampli-tude of the computed Station 46.75 bending moment is approxi-mately double that recorded in flight. Since the contribu-tion of the hub spring moment is the same at both stationsand very close to that obtained in flight, these discrepanciesmust be due to differences between the inplane shear computedin ARHFO1 and that actually experienced in flight. Thatportion of the shear due to the tilt of the thrust vector hasbeen well-matched, so the spanwise inplane inertia loads mustbe suspected as the source of the difference between the mea-sured and computed mast parallel bending moments.

Considering the simplicity of the mathematical model containedin hybrid computer program ARHFO, the results presented inthis section of the report demonstrate acceptable correlationwith flight test.

46

L ..... , ....I " ...- 1 I ... ' ..A1 1 i = ...

'-4

41

0 00

4) op

54

'U 47

In

I..-

48)

04

1

w= - =

4) -4 -1 4 - -

E-.

484

--7 t -

'4J

-4J

04J4

0) 0

0

449

Lin

- - -4I

0 0

1-

4- V 0'

41 W. 4-

0 -0001~r. - Tf4v

IV N

OT -q-U -~~c - B- - -

TOT -u -L9 - - -9 08W

so

... . .. . ..

3. HUB SPRING DESIGN

The addition of a nonlinear hub spring to an existing helicop-ter design requires a compromise between the level of springrates desired for handling qualities and the allowable loadsthat the new spring will add to the existing structure. Andfurther, the craftsman-like nature rather than pure scientificcontrol exercised in the composition of elastomeric propertiesnecessitates that some variation in desirable characteristicsbe tolerated.

3.1 DESIGN CHARACTERISTICS

The nonlinear hub spring was designed for the UH-lD/H helicop-ter to provide increased control power for reduction of mainrotor flapping in steady flight and maneuvers, and to elimi-nate metal to metal contact of the flapping stop and mast.Additional design objectives were a design that could be ret-rofitted with few modifications, a design that did not reducethe fatigue life of the UH-I hub and mast assemblies, and adesign that did not reduce the flapping limit.

The design of the hub spring is a trade-off between the con-trol power obtained with high spring rates and the additionalstructural loads applied to the hub and mast by the hub springthat were not present in the current design of the UH-lD/Hhelicopter. The method used to achieve this compromise is toemploy a flapping spring that provides a two-stage, linear/nonlinear, variation in hub moment with rotor flapping. Flap-ping in this case refers to the once per revolution teeteringaction of the main rotor hub.

A flapping spectrum was defined based on flapping data ob-tained during the UH-lD/H Load Level Survey tests and modifiedfor the flapping reduction in steady flight provided by thehub spring. This spectrum is presented as Table 4. Thisspectrum was the basis for calculations of the fatigue loadingof the rotor and spring components. The linear rate portionof the hub spring was chosen such that for over 90 percent ofthe service life of the helicopter, flapping would be lessthan 4 degrees which results in unlimited fatigue life of allexisting UH-ID/H parts. The maximum load resulting from thehub spring moment at 11 degrees of flapping was chosen suchthat no parts of the mast, hub, and blade assemblies wouldhave a negative margin of safety.

As previously stated, the constraints were that there be noreduction of fatigue life at low flapping angles and thatmaximum control power (i.e., hub moment) be applied to the

51

TABLE 4. DESIGN FLAPPING SPECTRUM

OscillatoryAngle Cumulative

Condition (deg) % of Time % Time

1 ±0.8 1.48 1.48

2 ±1.6 16.68 18.16

3 ±2.4 21.57 39.73

4 ±3.2 29.69 69.42

5 ±4.0 21.05 90.47

6 ±4.8 6.72 97.19

7 ±5.6 1.57 98.76

8 ±6.4 0.70 99.46

9 ±8.0 0.44 99.90

10 ±8.8 0.09 99.99

5

52

helicopter at high angles of flapping. A purely linear hubspring can only meet one, but not both, of these constraints;however, the nonlinear design was able to satisfy both con-straints. A linear hub moment is provided throughout theflapping spectrum by four elastomeric shear pads. Hub momentsat the rate of 330 ft-lb per degree of flapping are generatedby the shear pads. At 4 degrees of flapping, an elastomericsnubber (compression spring) contacts a mast sleeve providingadditional, nonlinear hub moment. The total, maximum hubmoment at 11 degrees limit flapping is 13,000 ft-lb. The equa-tion defining hub moment is given as Equation (4), Section 2.1.The additional control power provided by hub moment produced bymain rotor flapping is characterized by Figure 21.

14-

o12-

10- Nonlinear

x

8-

- 6 Transition- Point

4- Flapping~stop

S2 Linea

0~6 8 10 12

Flapping Angle, deg

Figure 21. Hub moment provided by nonlinearhub spring.

53

.! ___

The hazards associated with the metal-to-metal contact betweenthe mast and the flapping stop at extreme flapping (greaterthan 11.5 degrees) are greatly reduced by the UH-I hub spring.The contact between mast and hub is via the snubber whichgreatly increases the area of load application during contact.In addition, the mode of load application is changed from arapidly applied dynamic load to a more gradually applied load.

The UH-I hub spring can be easily retrofitted to the UH-lD/H.In addition to the hub spring assembly, a sleeve set, a clamp,and stronger pillow block attachment bolts are required.Prior to installation, the thick wall mast (P/N 204-011-450-7)must be installed because of the higher loads produced by thehub spring. Hub spring installation requires removal of themain rotor and attachment of the hub spring to the yoke andthe mast.

The hub spring bracket was designed to have an infinite life.The hub spring elastomeric components have minimum servicelife of 1500 hours. The Miner's damage coefficient has beencalculated for the required minimum service life and was foundto be 0.4832.

A postflight analysis to determine the effect of the hubspring on mast bearing lives also resulted in a determinationthat a positive margin of safety could not be calculated forthe transmission case. This analysis used lower stress allow-ables and new casting factors for magnesium castings. Also,the maximum moment produced by the hub spring (13,000 ft-lb)was used in addition to the maximum loads measured during thestructural demonstration program. This finding will requireadditional testing or an alternate material (aluminum) to beused for the transmission case (See Section 7.6.5).

A drawing of the UH-lH nonlinear hub spring is presented asFigure 22. Figure 23 pictures the hub spring itself. Notshown is this picture are the clamps that attach the hubspring to the main rotor yoke or the mast sleeve. Figure 24pictures the hub spring attached to the main rotor yoke. Al-though the hub assembly is not attached to the mast in thepicture, the mast sleeve and yoke clamp are clearly visible.The hub spring, as installed on the test aircraft, is picturedas Figure 25. Wiring bundles, which appear in Figures 23through 25, are associated with the instrumentation used forthis program.

54

Flapping Axis

I Hub clamp

Compression__ "Spring

Shear Pad

Figure 22. Drawing of UH-lH nonlinear hub spring.

55

NONLINEAR HUB SPRING

IUB BER BLOCK

Figure 23. UH-lH nonlinear hub spring.

56

NONLINEAR HUB SPRING

.SNUBBER BLOCKS

SHEAR PAD

Figure 23. Concluded.

57

41

: '.4

C)C

*4

0a4LaO 0

u*0

1>ui

0 0)

I- * 4

I5

F-.

UU

C/)

59

I R Mr - WT-IA

Figure 25. UH-lH nonlinear hub spring installation.

60

3.2 FAILURE MODES

The failure modes of the hub spring include the failure of theshear pads and/or failure of the snubbers. The failure of ashear pad or a snubber could result in two hazardous conditions.First, material from the failed item could fall from the huband be struck by or lodged in rotating components. Second,the hub moment contribution of the failed item could be lostat a critical flight condition.

Foreign object damage by material from a failed item isthought to be a negligible hazard. This is a result of thesize of the components and the failure modes necessary tofragment a shear pad or a snubber.

The second hazard would be the effect on the aircraft due toloss of hub moment as a result of a partial or a sudden, totalfailure of the hub spring. The loss of a single shear padwould result in a 25-percent reduction of the hub moment up to4 degrees flapping and 8.25 percent at 11 degrees flapping.Because of the design there would be no vibrations introducedas a result of a shear pad failure. The complete failure (de-bonding) of a single snubber would have no effect at flappingangles below 4 degrees. Above this, a one-per-rev vibrationof increasing (with flapping) severity would be experienced.The vibration would be the result of the unbalanced hub moment(maximum = 7370 ft-lb at 11 degrees) produced by the unfailedsnubber.

Hybrid simulation of a complete hub spring failure indicatesthat no unusual or uncontrollable aircraft motions would fol-low the hub spring failure. In the unlikely event that thereis either a partial or complete failure of the hub spring, theanalysis shows that there would be little if any adverse ef-fect to the aircraft or the mission.

62I

4. DESIGN LOAD CRITERIA

In order to ensure that the helicopter can be able to with-stand flapping stop impact, a conservative estimate of theloads generated during such an impact must be available. Toprovide a data base to compare estimation methods, a para-metric analysis of stiffness parameters in the UH-lH rotor wasconducted.

4.1 PARAMETRIC STUDY

Hybrid computer program ARHF01 has been used to study the ef-fect of several variables on the value of mast loads experi-enced during flapping stop impact. A description of thecomputer analysis, and substantiating data, are provided inSection 2. The parameters to be varied were linear and non-linear hub restraint, pylon stiffness, and mast stiffness.The simulations were conducted by placing the mast perpen-dicular to the ground with zero wind speed and trimming therotor to 2500 pounds thrust or 9300 pounds thrust with thecyclic set to zero. The cyclic was then increased to intro-duce flapping and data recorded in the new steady-state con-dition. The mast shear and bending moment at the point atwhich the hub would contact the mast are plotted for a pro-duction UH-lH in Figure 26. (The production UH-lH rotor canundergo ±11.5 degrees of flapping before striking the flappingstop.)

4.1.1 Variation of Linear and Nonlinear Hub Restraint

The hub restraint as modeled in ARHF01 permits both linear andnonlinear components to be represented simultaneously. Theequations for the moment about the teetering axis were givenin Section 2.1, Equations (2) and (4).

For the production UH-lH, the parameters are KL = KNL = 0

KFS = 130,000 foot-pounds/radian/rotor = 2269 foot-pounds/

degree/rotor. KFS is obtained from a NASTRAN analysis of themast and pylon as described in Section 2.3. Note that sincethe hub spring characteristics are inputs to the ARHFO1 pro-gram, changes in the hub spring will not require modificationof KFS.

The initial parameters for the proposed UH-lH hub restraintare:

KL = 330 foot-pounds/degree/rotor

63

0 -00 0

o 00

C) A.4 04 UJ

00-)

S E) E) 4

4.1 41 00

4J : 9I 00

0

E-4U

.e44

0

00441

4.1 EU

4.1

.~0 5

.0 0 4

rd 4 4

w 0 0 1 t00 0 r4

00 q a

4J 44A3 _ _ _ 4

K NL = 0.1 foot-pound/(degree)4"71/rotor

r = 4.71

= 4 degree

The shear and moment for this condition are shown in Figure27.

Figures 28 and 29 give the mast loads for cases run with one-

half the spring rates (KL = 165, KNL = 0.05) and twice the

spring rates (KL = 660, KNL = 0.20), respectively.

The analysis shows that for the simulated tied-down configura-tion any hub restraint increases both the moment and the shearin the mast. In flight, the hub moment would reduce the flap-ping angle due to increased control power. The loads areintroduced over a large surface area, though, which is animprovement over the essentially point-loading of hub-mastimpact on the production UH-lH helicopter.

It must be noted that the mast loads for the nonlinear springbear further investigation. The behavior of the spring willdepend on the details of design in terms of how the compres-sive material is constrained under loads. This constraintwill govern the compressive spring rate and thus the behaviorof the spring at high flapping angles. The mast loads com-puted in this analysis are based on the assumption that thenonlinear spring force increases in the same manner (i.e.,

proportional to Pr) even after the conventional flapping stopangle is exceeded. A static test to the structural limit ofmast, spring and hub components would be required to determinethe actual spring characteristics at excessive flapping angles.The analysis of the high flapping loads by ARHF01 is thus re-stricted to the assumption of the high flapping spring char-acteristics.

4.1.2 Variation of Pylon Mount Spring Rate

Simulations were conducted for aircraft configurations whichhad half or double the pylon mount spring rate. For thesecases, the NASTRAN model of the UH-lH was modified and newmode shapes were generated to represent the airframe. Thepylon mount spring rate also determines the root boundarycondition of the mast and, therefore, affects the flappingstop spring rate. Accordingly, new values of KFS were compu-ted for the modified pylon spring rate configurations, yield-ing

65

*. . . . 7. 4J- a C4

.m* 0

to x'a4-

m~ Ut

_ _ _4- -4-

E-4 0 u

U5.4

4J 1

0 la)

o 0

0) 43 r

00L .l

uH 0

(a~ 0

m. _ __ 00

0. U')

E-

UN W

oo -4 I0

0 to

44 W

U,4 0-. to410

00m

[:41

~ 04

4' S

00 Id 0 0

LLA

0 0 0

E-4-

qT-4; '4u~UX0W BUTPUOS

67

4J M

VN

- ----- 040

10 4 1 -04

00

cc w __T I) __)

a: cc 0.

4 J 0 0 0 0

H )K 4J -Ii

0 0

mm 00 4

*i U

4* 0

0 0l

Nz 00

L 0 0 N%

$4 ) El 4__ _

E-_ _ _ _D

qT xo xSqT-q; ';uaurow buypuag

68

Half Pylon Mount Spring Rate KFS = 92,400 ft-lb/rad

Double Pylon Mount Spring Rate KFS = 157,000 ft-lb/rad

The results of these runs are shown in Figures 30 and 31.Softening the pylon mounts by a factor of 2 reduces the mastloads markedly at low thrust, and a lesser reduction is ob-served at high thrust. This result is not surprising, asthe moment due to flapping stop impact has been reduced dueto the large reduction in KFS. Doubling the pylon mount

spring rate increases the mast loads a small amount due to theincrease in KFS.

4.1.3 Variation of Mast Stiffness

Calculations were made using aircraft configurations in whichthat portion of the mast above the transmission had half anddouble its normal stiffnesses. Again, the UH-IH NASTRANmodel was modified and new aircraft mode shapes and flappingstop spring rates were computed. The resulting values ofKFS were:

Half Mast Stiffness KFS = 83,750 ft-lb/rad

Double Mast Stiffness KFS = 169,300 ft-lb/rad

The simulation results are given in Figures 32 and 33. Asexpected, softening the mast lowered the loads, while stiffen-ing increased it.

4.1.4 Summary of Variation of Parameters

The only aircraft modifications that lowered the mast loadsdue to flapping stop impact were softening the pylon mountsor softening the mast. Both pylon mount stiffness and maststiffness are determined by other criteria, thoughneither of these is a practical remedy. Installation ofthe linear and nonlinear flapping springs increases the mastloads but decreases the amount of flapping required for agiven flight condition due to the significant increase incontrol power.

4.2 STOP CONTACT LOAD PREDICTION

The prediction of mast loads due to flapping stop contact bythe ARHF01 program has been shown to be accurate for the

69

0El 4

0 0 00 0-41~ ~ 0 E .r

04 40- 1 1-3 ..A5. g (

U U- 0

It 0f 0 J 94.) 4J H,

Ul0 0 4

GE))

C))

.r40

04

I944

0) ) 044m

4J

044

-. 4 C 0 4

C ) 0)~ >

N urn 0 N

4 4J 4J

4, 0

0 EE-4-

p .0f

m N ~ f-H

LOT x qT '9.Oa avals

O T x qT-4; 'WueMow BUTPU9g

70

0 El No El 01 4.

430 El C : 9

00g El to4.D 0 ElI

H 00 4

0 Urz 0(I 411 414

0I 00 H

4.1El1c0 0E1 4 0

'14J '44

.. 01

0444

0

0 0

o0

0 04 roo0 tv 0

H0 Pk)

o u *0 H

41 004.3

41 C: aa

Il 00

HH

£OT xC qT-; ';uehoW 6uTpugg

71

v-40 0

8 8 ~0 4

0 N ~00 E) 4 t

03 0 H0 X0 0 0

H 04000 .0 04

0 o (d 0

0~ 00 t4J 4J ) 0

0A Elr4 0

4)

410 a

a4)0 9

4)0

0 014Io f2 w

0 4J

44H- -

Ln J 4J 1

to4 00 00 4J 4J1 (

4) 0 0Uu

0L 0i0 n 0(n4 04 HH

.OT x qT 'ODXOA XV074S

ox qT-j; 4;ueIoN Bu'Tpuag

72

'-4

$64.0 Uw :4

0o 0 0

E-444

.8.1 000.04

00 G ~ 0

4.1

r- 04

00-

00 ~ WEl 41

0 El r

4.14

o- 0

o 04In go " 0

00- 0 00'4.144

od 0o IA01A

E- 0* I

£OT x qT-4; 'lusOuzo buTPusG

73

model rotor as discussed in Paragraph 2.2. For the full-scaleUH-IH, however, the problem encountered in accounting for theinplane shear loads with the simple blade model used madecorrelation with test data more difficult. This is discussedin Paragraph 2.3.2.

The prediction of mast loads due to rotor flapping beyond theflapping stop has been discussed and presented graphically inParagraph 2.2 for a model rotor and in Paragraph 2.3.2 for thefull-scale UH-lH. While the model test agrees well with theprediction, no data are available for comparison with thefull-scale prediction. However, both predictions follow thesame trend of almost linear build-up in moment with commandedflapping. The proposed design criteria for flapping stopcontact loads are based on a comparison with these predictionsfrom the hybrid computer program ARHF01. The general trend ofmast loads after flapping stop contact for the basic teeteringrotor system (no hub restraint) is a linear variation at aconstant slope as indicated in the sketch below.

4 J

Flapping Stop

Flapping Angle

74

Prior to flapping stop contact, the moments are primarilygoverned by the level of rotor thrust and thrust vector in-clination, while after stop contact, the flapping stop springrate is the major source of moment. Table 5 gives the slopeof moment versus flapping angle beyond the stop as measuredfrom the ARHF01 simulation presented in Figures 30 through 33.These slopes may be considered as the effective flapping stopspring rate for the conditions listed.

The simple theoretical prediction of hub moment given in Table5 uses an effective theoretical spring rate (KEFFT ) given by

KEFFT = KFS + KT

The derivation of the flapping stop spring rate (KFS) was pre-sented in Section 2.2. The other term in the above equationis the effective spring rate due to rotor thrust (KT). Thisterm is derived using the moment on the mast at the flappingstop due to rotor thrust, and the flapping angle (p) that therotor thrust is tilted. While this theoretical model does notinclude inplane inertial shears, it does provide a reasonableprediction of mast moment. As can be seen from Table 5, thesimple prediction was within 25.2 percent of the simulationresult in the worst case.

4.3 MAST LOAD CRITERIA

For teetering rotors, the ARHF01 simulation has predicted alinear variation of mast loads with commanded flapping beyondthe flapping stop. A simple analysis based on the static cal-culation of a flapping stop spring rate using conventionalstructural analysis techniques is able to predict the charac-ter of these loads such that a safety factor may be applied tothis simple method for design load calculations. With thisprediction, the load due to flapping stop contact may be con-sidered in the structural analysis of the rotor system.

Present design criteria consider the mast and hub loads forconditions where the flapping stop does not contact the mast.It is proposed that the following flapping stop shear load(and its associated reaction at the teetering pin) be applied,in addition to the loads considered by existing criteria, toensure adequate load-carrying capability of the mast, hub,pylon, and fuselage in the event of excessive flapping.

This flapping stop shear load will be calculated as:

75

SI1o px .s

FS KFS(PX - S) x Px

where

S = flapping stop shear load

FS = factor of safety

KFS = flapping stop spring rate (ib/deg)

pX = commanded flapping beyond the stop (deg)

PS = flapping stop angle (deg)

The flapping stop spring rate is calculated statistically asthe force at the flapping stop (balanced by a reaction at theteetering pin) which produces a degree of mast deflection atthe flapping stop considering the entire mast, pylon suspen-sion, and fuselage structure.

The commanded flapping angle, PX, and the factor of safety,

FS, must be established by further tests and analysis to en-sure that variations in pylon configuration, rotor thrustlevels, and other design variables do not affect this formula-tion.

76

i' -Dq ~ p'-: ' -

o , •. •L 0 -p $ ,-w . - -, ...

TABLE 5. EFFECTIVE FLAPPING STOP SPRING RATE

Af0fol Simple Theory SpringVariable Thrust Spring Rate Spring Rate Rate*Parameter Stiffness (ib) (ft-lb/deg) (ft-lb/deg) Ratio

Mast Stiffness Baseline 2500 2269 2295 0.989

Mast Stiffness 1/2 2500 1450 1488 0.974Baseline

Mast Stiffness 2 Baseline 2500 2850 2981 0.956

Pylon Spring Baseline 2500 2269 2295 0.989Rate

Pylon Spring 1/2 2500 1560 1639 0.952Rate Baseline

Pylon Spring 2 Baseline 2500 2900 2766 1.048Rate

Mast Stiffness Baseline 9300 2900 2365 1.226

Mast Stiffness 1/2 9300 1950 1558 1.252Baseline

Mast Stiffness 2 Baseline 9300 3450 3051 1.131

Pylon Spring Baseline 9300 2900 2365 1.226Rate

Pylon Spring 1/2 9300 2050 1709 1.200Rate Baseline

Pylon Spring 2 Baseline 9300 3100 2836 1.093Rate

*Spring Rate Ratio = ARHF01 Spring RateSimple Theory Spring Rate

77

L ,A

5. ELASTIC ROTOR EFFECTS

The severity of the maneuvers required for high flapping issuch that the aeroelastic effects may significantly affectthe predicted values of hub flapping. In order to determinethe influence of elastic blades on predicted hub flapping,main rotor blade mode shapes were included in the AH-IG Rotor-craft Flight Simulation Computer Program C81 (hybrid version)data deck. The mode type and its natural frequency are shownin Table 6.

TABLE 6. MODE SHAPES

Natural FrequencyMode Type (/Rev)

1 Rigid Body 0.999

2 Cyclic 1.678

3 Cyclic 2.522

4 Cyclic 2.903

5 Collective 1.0436 Collective 2.438

7 Collective 3.044

The modified AH-lG data deck was used as input to the digitalversion of C81, which allows elastic blades; the predictedvalues of the hub flapping were compared to the hub flappingpredicted by the hybrid version of C81, which allows only rigidblades. This comparison was made for steady flight conditionsand transient flight conditions.

5.1 STEADY FLIGHT CONDITIONS

The steady flight conditions selected for comparison werelevel flight, airspeed sweeps. The airspeed was varied from40 to 140 knots in increments of 20 knots. The gross weightand center-of-gravity combinations are shown in Table 7.

In general, the parameters exhibit good correlation. Thedifferences in predicted main rotor flapping are less than1.5 degrees in all cases.

78

TABLE 7. GROSS WEIGHT AND CENTER-OF-GRAVITYCONDITIONS

Gross Weight Center-of-gravityCondition (Ib) (in.)

1 10,000 200.0

2 10,000 142.5

* 3 7,000 190.0

4 6,000 201.0

(Note: Main rotor hub centerline is located at sta 133 in.)

The values of main rotor flapping predicted by the hybridversion of C81, and the digital version using the quasistatic-time variant trim routines are shown in Figure 34.

5.2 TRANSIENT CONDITIONS

The transient maneuvers selected for comparison were a cyclicpullup to 1.85g and a cyclic pushover to 0.25g. Peak pre-dicted flapping during both of these maneuvers was 6.5 de-grees.

The hybrid version of C81 utilizes both stability augmentationand an autopilot as an aid to the user during steady state andtransient flight conditions. These models are not availableto the user of the digital version of C81, and the user must"fly" the model. Thus, the control motions for a simulatedmaneuver using the hybrid version of C81 will change dramati-cally when the same maneuver is simulated using the digitalversion of C81. The reverse is also true. Therefore, ratherthan duplicate control motions, an attempt was made to dupli-cate aircraft attitude rates of change.

Illustrated in Figure 35 is the comparison between the hybrid(rigid blades) and the digital (elastic blades) of the cyclicpull-up maneuver. Good correlation was obtained in the pitchaxis, while only fair correlation was achieved in the rollaxis. Both main rotor flapping and load factor do, however,show good correlation. In both cases, main rotor flappingpeaks with maximum rate.

79

AH-lG - Hybrid C81 (AGHC74)GW = 6000 lb --- Digital C81 (AGAJ7625)CG = 201.0 in.

i 1054100% Flapping = 112.5 deg

0 I II

O 0--------- - --

4)

V

4J~ t7

0 ,.- --

i0

.4 100

44

0 0 604J .-4i-H4 4)

0 0 400 0 40 80 120 160

True Airspeed, kn

Figure 34. Effects of elastic main rotor blades onhelicopter parameter predictions duringsimulated level flight.

80

*MM-

'a - - -

AH-1G -Hybrid C81 (AGHC74)

GW = 7000 lb --- Digital C81 (AGAJ7625)CG =190 in.

100% Flapping i 12.5 deg4J M

4 '

r-I 0:r4

:24JW

-4 -10-

80

Va r 40- ____

20____ ____

00 0 40 80 120 160

True Airspeed, kn

Figure 34. Continued.

81

AH-1G - Hybrid C81 (AGHC74)GW = 10000 lb --- Digital C81 (AGAJ7625)CG = 192.5 in.

w 10 , I I I

$4100% Flapping = ±12.5 degos I

0o 0-

W Q) -5 -_ --

-80

60_q, 80

44 40

:30

r 2000 0 40 80 120 160

True Airspeed, kn

Figure 34. Continued.

82

AH-IG - Hybrid C81 (AGHC74)

GW = 10000 lb --- Digital C81 (AGAJ7625)CG = 200 in.

i0

a) 10100% Flapping = 12.5 deg

4)0 0

0-r. 5

4J-- I

I100r -I 8

>4

•,4. 4-4

80

*d

80 12 160

True Airspeed, kn


83

__1;

-Hybrid C81 (AGHC74)Digital C81 (AGAJ7625)

10

o -10

i-l10

-41o

U0

*rqP4 -103

Time, sec

Figure 35. Effects of elastic main rotor blades onhelicopter parameter predictions duringa simulated pull-up.

84

00

4 f'rI

Ck ~ -101

'-4

01

04.)0

0

Pz4

IV

Time, sec


85

__ i ~ m~ . .r ., ;, - ,- - . r . - ,, *,. ,, . ,, .... -r ' -.

In the case of the pushover, shown in Figure 36, good corre-lation is obtained in both the pitch and roll axis untilapproximately 8 seconds into the maneuver. Main rotor flap-ping and load factor exhibit good correlation until approxi-mately 6.5 seconds. Loss of main rotor flapping correlationand the large flapping amplitude predicted by digital C81 isnot the result of the addition of elastic blades, but ratherthe increased sensitivity and lack of an autopilot in thedigital model. These factors cause larger cyclic commandsthan were required for hybrid C81 to match fuselage motions.These larger cyclic commands can result in as much as 4 de-grees more swashplate motion and, therefore, higher main rotorflapping. The poor correlation of main rotor flapping in thepushover maneuver is caused by the larger cyclic motions.

86

Hybrid C81 (AGHC74)Digital C81 (AGAJ7625)

10 1-

4I MI )

'-I I lb K - _ _ __

104)

-10-

4.)

4J~

*120

10-

040

04

0 2 4 6 8 10Time, sec

Figure 36. Effects of elastic main rotor blades onhelicopter parameter predictions duringa simulated pushover.

87

10

)

,-10-

S-20

S 20

0 10 - --

400 I11

2

I I I I I I__

'.4

0 1

0 2 4 6 8 10Time, sec


88

6. TACTICS EVALUATION

A review and evaluation of U.S. Army Aviation tactical litera-ture and an observation flight at the Army's UH-1 nap-of-the-earth (NOE) course at Fort Rucker, Alabama, was made in orderto evaluate the current U.S. Army helicopter tactics. Thepurpose of this evaluation was to assess the possibility ofcurrent tactics leading to high flapping situations. Anymaneuvers so identified would be included in the flight testprogram.

6.1 EVALUATION OF TACTICAL LITERATURE

The U.S. Army Field Manuals (FMs) and Training Circulars (TCs)provided by the Army for analysis are listed in Table 8. Thetactics and the associated maneuvers described in the litera-ture were judged as potentially high flapping producersagainst the critical flight conditions identified in Reference1 as distinct contributors to large flapping amplitudes. Ac-cording to Reference 1, large flapping amplitudes could beexpected when simulated helicopters were operated under any ofthe following conditions:

- at center-of-gravity extremes

- under low or negative g conditions

- with large, abrupt control inputs

- in conditions involving significant retreating bladestall

6.2 UH-l NOE COURSE

An observation flight over the UH-1 NOE course at Fort Rucker,Alabama, was made with a Government/Industry team consistingof a Bell Helicopter experimental pilot and an engineer andthe Army project engineer. The Army instructor pilot (IP) dem-onstrated the various techniques of low-level, contour andNOE flight over one of several NOE courses at Fort Rucker.The IP pointed out that these three flight categories differin speed and altitude variations exercised by the pilot. Theprimary variable that governs which flight mode is to be em-ployed is the type of enemy threat known or considered to bein the operating area. Low-level flight consists of constantairspeed and constant altitude, usually approximately 50 feetabove the highest obstacle in the area. Contour flying demon-strates constant velocity with varying altitude to remain a

89

minimum safe height above major terrain features. Nap-of-the-earth flight requires the varying of both airspeed and alti-tude to maintain the least possible exposure.

TABLE 8. U.S. ARMY TACTICAL MANUALS

Date ofNumber Title Publication

FM 1-105 Aviator's Handbook August 1974

FM 1-1 Terrain Flying October 1975

TC 1-28 Rotary Wing Night Flight February 1976

FM 1-51 Rotary Wing Flight May 1974

TC 1-4 Helicopter Gunnery September 1976

TC 17-17 Gunnery Training forAttack Helicopters December 1975

FM 1-5 Instrument Flying andNavigation for ArmyAviators March 1976

Most of the demonstration flying was done at tree-top level inthe 40- to 60-knot range, with some flight at several hundredfeet and about 80 knots. Seldom did the aircraft venture be-low tree level except for a demonstration of the NOE "quick-stop" and the masking/unmasking or bob-up maneuvers. Themaneuvers determined to be most critical in requiring abruptcyclic inputs and thus large flapping values, although noneseemed to task the flapping limit of the vehicle, includedevasive turn reversals and pull-up/push-over combinations.The turn or roll reversals were characterized by a levelflight zigzag maneuver. During the maneuver, the aircraftwas sharply banked in one direction and then the other inorder to return to the original course. Maneuvers that com-menced from IGE hover were the NOE quickstop, the bob-up, anddashes; all of which were demonstrated. The IP indicated thatlarge, abrupt cyclic inputs were uncommon. Such control in-puts were encountered rarely and usually only when a studentpilot responded incorrectly to a simulated or actual emergencycondition. Following the demonstration flight, the Army/BHT

90

' " *i -' " 4, . " - " - • -

team met with several other UH-l NOE IP's in order to deter-mine if any other operational NOE maneuvers were susceptibleto high flapping.

The BHT team interviewed the IP's at the Scout/ Attack Schoolthe following day. These conversations revealed that in addi-tion to the NOE course, there was an agility course. Thiscourse is limited in its use to Scout and attack-type helicop-ters, and a more aggressive flight profile is flown in orderto avoid simulated threats. Although no demonstration flightswere given, the agility course should be evaluated in order toassess the possibility of NOE flight and NOE tactics, asassociated with the Scout/attack environment, leading to highflapping situations.

6.3 TACTICS EVALUATION SUMMARY

Applying the criteria from Reference I for judging large flap-ping amplitudes, it was determined that flying at the forwardcenter-of-gravity extremes, maneuvers requiring abrupt, largeamplitude cyclic inputs such as evasive maneuvers, and thecombination of these could result in large flapping ampli-tudes. In order to determine the order of magnitude of mainrotor flapping during these NOE maneuvers, the sideward dash,the NOE quickstop, and the level flight roll reversal wereadded to the flight test program. As a result of the tacticsevaluation, three NOE maneuvers were added to the flight testplan in order that main rotor flapping amplitudes could bedetermined. The level flight roll reversal and NOE quick-stopwere determined to be of interest from the NOE demonstrationflight, and the lateral, or sideward, dash was selected follow-ing evaluation of the tactical literature. The flappingamplitudes measured while performing these maneuvers are pre-sented in section 7.2.2.

91

7. FLIGHT TEST PROGRAM

The helicopter used for this program was a U.S. Army Model UH-IH helicopter, pictured in Figure 37, equipped with a thick-wall mast (part number 204-011-450-7). The helicopter wasinstrumented as described in Reference 5. The contractedflight test program consisted of evaluating a baseline, orstandard, helicopter against the same helicopter with a non-linear hub spring installed. The conditions that were to beflown are described in Reference 6. Cost and schedule con-straints dictated that the hub spring be evaluated in lessthan ideal weather for handling qualities tests. In thefollowing subsections, the objective and results of the flighttest program are presented.

7.1 PROGRAM OBJECTIVE

The objective of this flight test program was to determine ifa nonlinear hub spring would improve the mast bumping safetymargin and, at the same time, not degrade the handling quali-ties, any component life, or the vibration levels of the UH-lHhelicopter. Improved safety margin is demonstrated by reducedmain rotor flapping in all flight conditions and increasedcontrol power at low-g flight conditions. In order to meetthis objective, a flight test program that included steadyflight conditions, handling qualities tests, nap-of-the-earth(NOE) maneuvers, low-g flight conditions, and steady highflapping flight conditions was defined and conducted.

7.2 EFFECT OF THE HUB SPRING ON MAST BUMPING SAFETY MARGIN

For the UH-I, an improved mast bumping safety margin can beshown if main rotor flapping is reduced for all flight condi-tions. This is achieved by increasing the flapping-stop-to-mast clearance. Additionally, positive control of the heli-copter during low-g maneuvers would indicate an improved mastbumping safety margin, since mast bumping has been associatedwith loss of lateral control power at low g. In order todetermine the effect of the hub spring on main rotor flappingand low-g control power, a series of tests representing

5Bonham, D. G., and Hollifield, P. J., UH-l HUB SPRING INSTRU-MENTATION TEST PLAN, BHT Report Number 699-099-095, BellHelicopter Textron, Fort Worth, Texas, June 1978.

6Hollifield, P. J., UH-l HUB SPRING TEST PLAN, Bell Helicop-ter Textron Rpport Number 699-099-099, Bell Helicopter Tex-tron, Fort Worth, Texas, August 1978.

92

Figure 37. Bell Helicopter Textron UH-lH

93

steady, maneuvering, and steady high flapping flight condi-

tions as well as flight at low-g were conducted. These con-ditions were first flown with the basic UH-1 helicopter andthen repeated following installation of the hub spring. Inorder to minimize differences due to pilot technique, the sameproject pilot flew baseline and hub spring flights.

7.2.1 Main Rotor Flapping Comparisons During Steady FlightConditions

As expected, the effect of the nonlinear hub spring was to de-crease flapping in steady flight conditions. Flapping reduc-tions are greatest when baseline flapping is greater than 4degrees, due to the additional moment produced by the non-linear compression spring.

7.2.1.1 Forward and Rearward Level Flight. Figure 38 com-pares main rotor flapping for worst-case conditions of base-line and hub-spring-installed flights. The greatest reductionis in rearward flight where baseline steady flapping is great-est. The flapping differential ranged from 1 to 3 degreesover the airspeed range tested. While overall flapping dif-ferentials were less in forward flight, the hub spring reducedsteady flapping to levels below 4 degrees for the entireforward flight speed range.

Due to wind conditions at the time the hub spring tests wereconducted, the hover point with hub spring installed is notavailable.

7.2.1.2 Sideward Flight. The comparison of main rotor flap-ping at the forward (worst-case) center-of-gravity loading isshown as Figure 39. Airspeeds from 30 knots left to 30 knotsright sideward were tested. The increased flapping in rightsideward flight is primarily due to the rotor inflow distribu-tion at low advance ratios (Reference 7) causing down aftflapping which adds to the down aft flapping due to the for-ward center-of-gravity loading.

7.2.1.3 Autorotation and Maximum Continuous Power (MCP)Climb. The reduction in main rotor flapping during autorota-tlon and MCP climb at both extremes of the center-of-gravityenvelope is shown in Figures 40 and 41. Although the increase

7Harris, F. D., ARTICULATED ROTOR BLADE FLAPPING MOTION ATLOW ADVANCE RATIO, Journal of American Helicopter Society,January 1972.

94

r-- *

- to

0 4C) 0H

H Mrx-4 a)

0

04 2

4-4

00U

0 4)

I 04

-~~ -- -4

00E

.4 -

.04-)

4404-It

0 co4

00

b9P 'fBUTddeia a0O4OX utPW

95

AD-AO98 ?"9 BELL HELICOPTER TEXTRON FORT WORTH TX F/G 1/3FLIGHT TEST EVALUATION OF A NONLINEAR HUB SPRING ON A LA-IN MEL-CETCfU)APR 81 P J1 HO4LLIFIELD, L W DOOLEY DAAJ02-7-C-OGW4

UNCLASSIFIED USAAVRADCO4-TR-800-27 NL

lflfl lfl lflfI

iimniimnmnnmmIEEEmhmhmhmmmmmmmnmmmmmmmmmfllflflfnmmmEEmhinEm

11&25

ICRI RSO TES ICAR

L ._

MICROCOPY REOUINTSCHR

SYM CONFIG FLT

0 Hub Spring 198AEl Baseline 192A

GW = 8600 lbCG = 130 in.

10100% Flapping ±11.5 deg

ro 80 -

6 .6

-4

.4.

0

-30 -20 -10 0 10 20 30Left Sideward Hover Right Sideward

Airspeed, kn

Figure 39. Effect of the hub spring on main rotorflapping in sideward flight.

96

SYM CONFIG FLT0 Hub Spring 195, 197AEJ Baseline 183A, 186A

GW = 8600 lb

8 CG = 130 in. __

0)100% Flapping ±11.5 degw

-H ra

r. 04

GW =8000 lb

4 100% Flapping =±11.5 deg0-

04 4

Calibrated Airspeed, kn

Figure 40. Effect of the hub spring on main rotorflapping during autorotation.

97

SYM CONFIG FLT0 Hub Spring 195, 197A'0 Baseline 183A, 186A

GW = 8600 lb6CG = 130 in. __

100% Flapping =±11.5 deg

0-

2 -

~r. 0

GW = 8000 lb

100% Flapping =±11.5 deg

0


Figure 41. Effect of the hub spring on main rotorflapping during MCP climbs.

98

. .. .. . . . - ,. - .S.. . . .. ". .. - " • - - , ... - , . . ' "" - Z .

in mast bumping safety margin is not as great as in the pre-vious conditions, main rotor flapping has been reduced by atleast 1 degree in all cases. For these cases, the baselineflapping is less than 4 degrees, thus the nonlinear compres-sion spring is not exercised.

7.2.1.4 Wind-up Turns. Main rotor flapping during coordi-nated wind-up turns to over 1.7g is shown in Figure 42. Thetest procedure used was to stabilize in level flight at thedesired airspeed and, maintaining fixed collective, turn whiledescending in order to remain at the same airspeed. Steady-state data were recorded at several g-levels during the turn.

For wind-up turns at 72 knots (0.6VNE), a near constant 1

degree of difference in main rotor flapping is held for all g-levels. At 115 knots (VNE), an even greater flapping reduc-tion is observed.

7.2.2 NOE Maneuver Flapping Comparison

The NOE maneuvers flown were the NOE quickstop, forward andlateral dashes, the pop-up and the bob-up, and the levelflight roll reversal. Each of these maneuvers and the ap-parent effect of the hub spring are presented in the followingsubsections. Table 9 summarizes the results of the NOE maneu-ver comparisons. With the hub spring installed, the NOEmaneuvers were characterized by either reduced flapping forthe same level of maneuver performance, or increased maneuverperformance with the same flapping.

7.2.2.1 NOE Quickstop. From 60 knots, the helicopter wasdecelerated to a hover by increasing collective pitch whilesimultaneously applying aft cyclic control in order to main-tain the constant tail rotor altitude for obstacle clearance.With the hub spring installed, main rotor flapping increasedslightly; however, the maximum pitch rate increased more than50 percent over that of the baseline from 8.0 to 12.5 degrees/second indicating the capability to decelerate faster with thespring installed. As expected, the installation of the hubspring resulted in increased control power.

7.2.2.2 Forward Dash. For the forward dash, the helicopteraccelerated from hover to 60 knots and then decelerated to ahover at the maximum rate. The dash, however, was not termi-nated by the NOE quickstop. In the case of the dash, mainrotor flapping was decreased by 0.5 degree with the hub spring,while maximum pitch rate during the maneuver was approximatelythe same.

99

GW= 8000 lb SYM CONFIG FLTCG = 144 in.

0 Hub Spring 198F

Airspeed V Baseline 187A6p VNE6100% Flapping ±11.5 deg

0 r - --- '- - - -

0-

0404 2

r2z

Airspeed = 0.6 VNE

MI I I

4) 100% Flapping = ±11.5 deg

0-4 J0

4-4

Cv44

1.0 1.2 1.4 1.6 1.8 2.0

Load Factor, g

Figure 42. Effect of the hub spring on main rotorflapping during maneuvering stabilitytesting.

100

~" 0

P44Z

Inn

0 0 LA NE-1 ~ I I.lL

0 m (dOIx- E-4 z -4

0-

EO~0 OA 44 0 N

w 4- 4 ) A A L

0U0

tnn

E0 H' 04 0K-4' W

E-4)

~ r-4

4J~ -A r- 0

004 ~~0 U4L A~- A L

10

7.2.2.3 Lateral Dash. The lateral dash was performed in thesame manner as the forward dash; however, the direction offlight was sideward and the maximum airspeed was 30 knots. Ofall the NOE maneuvers, the right sideward dash, at the heavy/forward center-of-gravity loading configuration, resulted inthe highest main rotor flapping. Figure 43 presents a timehistory of the lateral dash as flown by the standard UH-lHhelicopter and with the hub spring installed. At the time ofthe hub spring flight, surface winds were approximately 10knots so that the maneuver was started and terminated atapproximately 10 knots right sideward flight. Figure 43 showsthat main rotor flapping was at the target limit (10 degrees)for an extended period in the baseline configuration. Peakmain rotor flapping was reduced 2 degrees.

7.2.2.4 Level Flight Roll Reversal. As stated previously(Section 6), this was the only maneuver added to the test pro-gram as a result of the tactics evaluation. As flown duringthis program, the roll reversal was an evasive, zigzag maneu-ver entered at a 60-knot, trimmed, level flight condition.The helicopter was flown rapidly into a 45-degree banked turn,followed immediately by a banked turn of 45 degrees in theopposite direction, and then returned to level flight. Figure44 presents a time history of a roll reversal as flown withthe standard helicopter and by the hub spring-equipped heli-copter.

Although there was a slight increase in maximum main rotorflapping with the hub spring installed, the maximum roll rateincreased 25 percent from 38.8 to 49.0 degrees/second, a char-acteristic that was seen in other maneuvers. That is, if mainrotor flapping increased during a maneuver flown with the hub-spring installed, there would be an attendant increase infuselage angular rate of change.

7.2.2.5 Pop-up and Bob-up. In the case of the vertical bob-up, the helicopter started from an IGE hover, climbed verti-cally into an OGE hover, and returned to IGE hover. For thepop-up, the helicopter was accelerated into translational liftwhile climbing. Installation of the hub spring reduced maxi-mum main rotor flapping 2 degrees during vertical bob-up and1.1 degrees during the pop-up (see Table 9).

7.2.3 Steady High Flapping Conditions

In order to evaluate the hub spring during conditions of sus-tained high flapping, the test helicopter was loaded to themaximum allowable weight (8600 pounds) at the most extremeforward center of gravity (FS 130 inches). Data were thencollected during 30-knot rearward and right sideward flight,

102

GW = 8600 lb Flight 188A ConfigurationCG = 130 in. Baseline

- -- !

6- -W

T ~

-Hlot.Iv

8 12 16 20Time, sec

Figure 43. Effect of the hub spring on main rotorflapping and helicopter response duringa right sideward dash.

103

GW = 8600 lb Flight 198D ConfigurationC G = 130 in. Hub Spring

116 -~t '~0

ks~

XI

: H-I-'age

1041

-dib- -. .-

GW = 8600 lb Flight 183B ConfigurationCG = 130 in. Baseline

IdI t

-a.-

... .v...

Time,se

Figure 44. Effect of the hub spring on main rotor

flapping and helicopter response duringalevel flight roll reversal entered at80 knots.

105

GW = 8600 lb Flight l98D Configuration

CG = 130 in. Hub Spring

.

ILI

%-go ----------- -

1106

which are conditions of known high flapping (Reference 1). Asnecessary, the center of gravity would be moved forward inorder to obtain desired main rotor flapping values of approxi-mately 9 degrees. In the baseline configuration, tests werediscontinued at FS 129.5 due to lack of sufficient aft cycliccontrol margin for gust correlation and the 9-degree mainrotor flapping target being met. With the hub spring in-stalled, testing continued to FS 128.88 and discontinued dueto high loads in the hub spring hardware. At this extremecenter of gravity, sufficient control margin remained for gustcorrection, and main rotor flapping was 2.8 degrees less thanthat measured during baseline flights at a more aft center ofgravity. Table 10 shows the results of these flights.

7.2.4 Low-g Flight Control Power

A low-g flight investigation was conducted in order to deter-mine the effect of the hub spring on control power at low g.The low-g condition was achieved at the termination of aroller coaster maneuver, which was entered at an airspeed of70 knots. The helicopter was placed in a dive until the air-speed built to 100 knots, a climb was initiated and continueduntil the airspeed decreased to 80 knots; at this time, apushover to the desired a-level was executed (Figure 45). Loadfactor during the pushover maneuvers was incrementally de-creased toward a target value of 0.2g.

During the period of low g, left lateral steps up to 0.75 inchwere made in order to determine low-g control power. Thismaneuver as done with the baseline configuration and with thehub spring installed is shown in Figure 46. Figure 47 showsthat the hub spring increased the low-g control power for allg-levels tested. This figure also shows that at 0.2g, 42 per-cent of the lg lateral control power is available with the hubspring configuration; however, via extrapolation only 16 per-cent of the lg lateral control power is present at a 0.2g forthe standard UH-lH helicopter. The pilot was never able toactually reach 0.2 g in the baseline configuration. The maxi-mum main rotor flapping measured during all low-g, and low-gwith control power steps, was only 4.5 degrees.

7.2.5 Summary of Improved Mast Bumping Safety Margin Tests

As shown, the mast bumping safety margin of the UH-l helicop-ter has been improved by installation of the hub spring. Mainrotor flapping was reduced in all conditions tested and lat-eral control power was increased. Although a numerical valuecannot be assigned to quantify the improvement of safety mar-gin, reduced flapping and increased control power verify thatan increased flapping stop-to-mast clearance has been achieved.

107

TABLE 10. STEADY HIGH FLAPPING CONDITIONS

Center-of-Gravity

Flight Location Maximum Flapping*Condition (in.) Hubspring Baseline

30-knot rearward 130.0 5.50 8.20

129.5 5.80 9.30

128.88 6.50 -

30-knot right 130.0 6.20 8.80sideward

129.5 6.70 -

128.88 7.20

*Maximum flapping target value was 9 deg. (80% of total

flapping limit).

108

t _ _ _ _ _

GW = 8600 lb Flight 191 ConfigurationCG = 137 in. Baseline

-------- ~---. - -4 - -

.M jug * t%. ---

Time, sec

Figure 45. Effects of the hub spring on main rotorflapping and helicopter response during aroller coaster maneuver to low-g.

109

GW = 8600 lb Flight 199D ConfigurationCG= 137 in. I Hub Spin

I i+fr.~

.II~ --1 - --- -- -----.......

I1

04 2 62

Time, sec


11~0

AL.-

GW = 8600 lb Flight 191 ConfigurationCG = 137 in. Baseline

--- -- .~-. -..........

'-'

Time, sec

Figure 46. Effect of the hub spring on main rotorflapping and helicopter response followinga left control step at low-g.

G= 8600 lb Fih 9D ConfigurationCG = 137 in. Hub Spring

.-- --- ---

uio..

Time, sec


112

SYM CONFIG FLT

0 Hub Spring 199Do Baseline 191

GW = 8600 lbCG = 137 in.

-4

1.0-----------------

08

0.6L0 44 5-.. -

En 0.2 --4

w 0.0

-4

10 08 0.6 0.4 0.2 0.0

Load Factor, g

Figure 47. Effect of the hub spring on lateralsensitivity at g-levels less than 1.0.

113

H

7.3 EFFECT OF HUB SPRING ON HANDLING QUALITIES CHARACTERIS-TICS

In order to determine the effect of the hub spring on thehandling qualities characteristics of the model UH-lH, aseries of test points were flown in the baseline configurationand repeated with hub spring installed. The following subsec-tions present effects on control margin, static stability,maneuvering stability, dynamic stability, and controllability.The baseline flights were conducted with atmospheric condi-tions consistent with those acceptable during handling quali-ties testing; however, the hub spring flights were conductedat levels of atmospheric turbulence greater than those nor-mally considered acceptable for reliable handling qualitiestesting.

7.3.1 Control Margins during Steady Flight

Figure 48 shows longitudinal control positions during levelflight from hover to 30 knots rearward and from 45 knots for-ward to VNE at critical loading conditions. The critical

weight/cg loading condition for rearward flight is heavy/forwardand for forward flight it is light/aft. At the heavy forwardcenter-of-gravity extreme, an increase of 5 percent longitudinalcontrol margin at -30 knots is provided by hub spring installa-tion. Control positions during sideward flight are shown inFigure 49.

7.3.2 Static Longitudinal Stability

The apparent static longitudinal stability as indicated by thepositive gradient of the longitudinal control position withrespect to airspeed about a trim condition of 0.08 VNE is pre-

sented as Figure 50. As shown, the collective fixed staticlongitudinal stability at aft center of gravity of the UH-lHtest vehicle increased from a 0.07 percent per knot baseline to0.12 percent per knot with hub spring, as shown in Figure 50.There was no significant change in the static longitudinalstability at the forward center-of-gravity loading condition.

7.3.3 Maneuvering Stability

In order to determine the effect of the hub spring on maneu-vering stability, flight tests were conducted to determine thelongitudinal control position displacement required to developa steady-state acceleration during steady turning flight.Longitudinal cyclic position during wind-up turns to 1.7g wererecorded at 72 and 115 knots and presented in Figure 51. Asshown, the hub spring does not significantly change the maneu-vering stability of the UH-lH helicopter.

114

C)l

4-I-~ _ 4-1 04

H n 0f4 0

U )n0, H

_ _ _ __ _ 4 -i J0 4

4-I

$,W 0

to 0 0

$4 *,-04) : r

___~~~C -( _ _ - -4~U 4-0

44 40 0 0

440

o 0Co0 0 0 0 0 0 1000 r, %DO Lfn qr ' CN4 H

H 44i

44 PmJ % 'UOjTrrS~c OTTO3AD TVUTpnqTbuorI r4

P44

115

GW = 8600 lb SYM CONFIG FLTCG = 130 in. E Hub Spring 198ABaseline 192A

80

H 60

4J0 dP__ __ __ _ _

0 40C.iU-4 0__

w 4 204J (n

0U

Full left

r- 800o4-,

oT 60

4-4

M ~dP40

. ( -- [ - _

o 0 0 -

0 0

Full aft 0 40 30 20 10 0 10 20 30 40Left Sideward Hover Right Sideward

Airspeed, kn

Figure 49. Effect of the hub spring on control positionsduring level sideward flight.

116

SYM CONFIG FLT

Q Hub Spring l97A, 198DSBaseline l83B, 186A

GW = 8600 lbCG = 130 in.

H 70-

U'

H 4dP 60

ra 50 __ ____ _ _

S-r4w 40 __ ____ __ __ __ _ _

NOTE: Shaded symbol is trim point.

H GW = 8000 lbU ro

>1 t CG = 144 in.U 4 90 _ __

__

rd g 80 -4p

.4J -4I-4 .-P

g n 70-0 0 ~~60 8 0 2


Figure 50. Effect of the hub spring on staticlongitudinal stability in level flight.

117

GW = 8000 lb SYM CONFIG FLTCG = 144 in.

C Hub Spring 198FO Baseline 187A

Airspeed V N VNE-4 100

r-

U'

dP

o0 60

Airspeed = 0.6 VNEU

• 80

60 -0. -

4J -H4r0 401

0 0 .01 .1 .1 .1 .2 0

Load Factor, g

Figure 51. Effect of the hub spring on maneuveringstability at O.6VNE and VNE.

118

.4 so,

HZ

With hub spring installed, wind-up turns to load factorsgreater than 1.7g's at a main rotor speed of 314 rpm were con-ducted. During these tests, there was increased vertical vibra-tion and detectable cyclic feedback. Although no conclusiveexplanation for the increased vertical vibration was found,retreating blade stall was suspected at the 2g load factor con-dition reached during hub spring tests which resulted in cyclicfeedback. During baseline flights, load factors of only 1.6gwere reached and no increased vertical vibration or cyclicfeedback was noted.

7.3.4 Controllability

In order to determine the effect of the hub spring on con-trollability characteristics, aircraft motions following rapidcontrol displacement were recorded and analyzed. Althoughcontrollability is usually conducted in still air, tests withthe hub spring were conducted during a period when surfacewinds were approximately 10 knots.

Controllability testing was conducted at a hover and in levelflight at 70 knots with the test aircraft at both the heavy/forward and the light/aft gross weight and center-of-gravityloading configurations. Figures 52 through 54 are time his-tories of aircraft motions following a cyclic control step andare considered to be typical of these tests. Figure 55 pre-sents a summary of the hovering control characteristics asmeasured during these tests. Control sensitivity, the maximumvehicle angular acceleration per inch of control input, wpsincreased by the hub spring by approximately 15 percent inboth the pitch and the roll axis. Angular velocity dampingalso appears to be increased as a result of the hub spring.Control response, the maximum vehicle angular rate per inch ofcontrol input, in the pitch axis was increased 36 percent,from 6.7 to 9.1 deg/sec by installation of the hub spring.This increase in control response is not reflected in the rollaxis. This is due, in part, to greater rate damping beingprovided by the stabilizer bar, which is indicated by themaximum roll rate of the basic aircraft being 40 percentgreater than the maximum pitch rate.

7.3.5 Dynamic Stability

Acceptable handling qualities require that forces are set upto damp oscillations resulting from disturbances from trimconditions and that the vehicle tends to return to the trimcondition. In order to evaluate the effect of the hub springon this dynamic stability requirement, external disturbanceswere simulated by applying a rapid pulse control input about asingle axis. The aircraft controls are then held at the trim

119 I

GW = 8100 lb Flight 188B Configuration

CG = 144 in. Baseline

6-9 - 2YJ.- I.. .I ... .... . --t Ii .... .... ..----

-4----'

,---------------------------+...., ------ .. - --...--...--,--- - -

I it

• , , -r .i m e s e c' ,

F ig. 52. . i E of t h s i o m r

£- wI I I

s p a 'v r

. 4i2 0

-i . .. ..... . I ' : I

-- -S -4.- -l.. L--!-_ _,---

- . -- .... ... ::T 2 _,I: -

'i + - : 1 2 34

Time, sec

Figure 52. Effects of the hub spring on main rotor flapping

and helicopter response to an aft cyclic control

step at OGE hover.

120

GW = 8100 lb Flight 199C ConfigurationCG = 144 in. Hub Spring

fa AI I " --

- 0- -[4 ... .-'

" .

X L~f'I

Tie sec

- , -p qpp - .- ' - - " --'-

- ' - I .__. .. . _.. . ...

01 234

Time, sec


121

GW = 8100 lb Flight 188B ConfigurationCG = 144 in. Baseline

I ih-a I*

. .. I ' 1 "

1:I1 t7 -1-!--

4 J.-

- iI t

- .. , __ 1_.

...., -< H ..... _ .... __

--_ _ 1 .7-- -

.~ ...... I

0 2 3 4

Time, sec

Figure 53. Effects of the hub spring on main rotor flappingand helicopter response to a right lateral cycliccontrol step at OGE hover.

122

GW = 8100 lb Flight 199C Configuration

CG = 144 in.-: : [ ,-i-, T ,l ! .. I ' ,I: : I" :I Hub Sprng

ER A

7, "

U AL 4..

-- J --. ... 1 . . -----:_444.4

_1 ___

.... .. -- .. .__.._ . . -- - _ _

AGo - -M

- 1j-* 1 -: I I-- ---

I __1.

K j to

01 2 3 4

Time, sec


123

Ix

GW = 8600 lb Flight 183B ConfigurationCG = 130 in.S -Baseline

......... .-.. . .... .

- -_ - --- ---

i ,- . --•4

I - h i ' I: : . . . . . . . .. - ..

- . . . .. , I '

............

m_ -r-- ----+---I -• , -_t. .O..O1

- i. Li-.v-. I I

• " : t .. . -," ' OMM i

0 1 2 3 4

Time, sec

Figure 54. Effects of the hub spring on main rotor flappingand helicopter response to a left lateral cycliccontrol step at 0.6 VNE.

124

9I

GW = 8600 lb Flight 198E Configuration- - C G = 1 3 0 i n . - . H u b S p r i n g

- I ! -I o.- - -.-

- T t

)-co . . . ..

e I :-4 - -..finI -A

C -j

' c c" .......4 - --r "z I.

.. .... --- €I ag- - • _. . -' .. . a .. w-="

o I - IIB°-5 '---"4 -

-I , )a0 1 2 3 4

Time, sec


125

SYM CONFIG FLT

0 Hub Spring 199CO Baseline 188A, 188B

2.0 Roll Axis

2.0

1.5

El G,4 1.0 _

_.' 0.5

,-0

ow 0.0

2 Pitch AxisW2.0 -

I

t; 1.5 '

-40 1.0 77R-_

9 4 0.5

U0

- w 0.0 1 -

m>

0.0 0.1 0.2 0.3 0.4 0.5

Control Sensitivity, rad/sec /in.

Figure 55. Effect of hub spring on control sensi-tivity and damping.

126

condition and aircraft motions recorded. The dynamic stabil-ity tests were conducted at the light/aft loading configura-tion at airspeeds of 70 and 110 knots. As stated before,atmospheric turbulence levels during hub spring testing wereabove normally accepted values.

In comparing the time histories of the lateral control pulsespresented in Figure 56, it can be seen that the lightly damped,small amplitude, roll oscillation found in the baseline con-figuration is apparently eliminated by the hub spring.Additionally, coupled oscillations in the pitch and yaw axisare not present in the hub spring configuration.

If the time histories of the longitudinal control pulses(Figure 57) are compared, little difference in damping in thepitch axis is noted. As in the case of the lateral controlpulse, however, cross-coupled roll oscillations are apparentlyeliminated by the hub spring.

7.3.6 General Handling Qualities Summary

The principal objective of the nonlinear hub spring was tominimize or eliminate mast failure due to severe flapping stopcontact, not to improve rotorcraft handling qualities. How-ever, test results indicate a general improvement of thehandling qualities of the UH-lH helicopter following installa-tion of a hub spring.

The following improvements were noted:

- Increased control margins in forward and rearward flight.

- Improved static longitudinal stability at light grossweight and aft center of gravity.

- Eliminated persistent roll oscillations following cycliccontrol pulses.

- Increased control sensitivity and angular rate damping.

Although the presence of the hub spring was not readily appar-ent to the pilot in all flight conditions, the data show im-provement in most areas tested.

Another improvement is the absence of large main rotor flap-ping excursion during star*-up in high winds. Although thiscondition was not tested, it was apparent that main rotorproblems associated with starting in high winds were improvedby the installation of the hub spring.

127

GW = 8100 lb Flight 186C ConfigurationCG = 144 in. Baseline

It ; : . .. ... ...... ....... . -4

_I

-• I i

j

5 . z

0 44-i

f , " .. .. ........-..... ....... . .. ... .

Time, sec

andl heiope repos toalf atrlcci

control .. .. .

1 I I I

-I....-- -

- - . T I: . . . - - --

Time, sec

Figure 56. Effects of the hub spring on main rotor flappingand helicopter response to a left lateral cycliccontrol pulse at 0.6 VNE.

128

GW = 8100 lb Flightl99B Configuration

CG = 144 in. Hub Spring

.-- ... ....... ... i ------- .................

• . ""1.° -

- -,.,..

....... __.

i 1.

- 4 • 1 I

-t *-,-- - --. t- - J

--- J--

8- - 0 8 12 16

Time, sec


129

GW = 8100 lb Flight 186C ConfigurationCG = 144 in. Baseline

7 - L-.,/I +

--W -4-: --+,,

II } I ' ____ _,

0 Ii I 2 25

_t-,- I

-H" , -Tim, s ,ec'

ar he c te repos to anatcci

• ! I I - i- " !--'

130

.- .I-. .r --. ...-. ...

I I __

__" I ! ;

4 : I0 5 10 15 2 0 2:5

Time, sec

Figure 57. Effects of the hub spring on main rotor flappingand helicopter response to an aft cycliccontrol pulse at 0.6 VNE.

130

GW =8100 lb Flight 199B ConfigurationCG = 144 in. Hub Spring

- - - --w ------- .. *

- - - - - - - - -- - - - --...

-- -- - -

4 o -4-

n 00

____ __ __ _

0 015202

1131

7.4 EFFECT OF HUB SPRING ON VIBRATION CHARACTERISTICS

In order to determine the effect of the nonlinear hub springon vibration characteristics of the UH-lH helicopter, accel-erometers were located in the nose, at the crew stations, atthe aircraft center of gravity, and at the top of the verticalfin. Since the main area of concern is the vibration acceler-ation at the crew and passenger stations, only the vibrationacceleration data at the frequency of 10.8 cps (2-per-rev)during steady level flight are presented. Passenger stationvibrations were measured in three axes (vertical, lateral, andlongitudinal) by accelerometers at the center-of-gravity andcrew station by measuring (with accelerometers under the crewseats) copilot (left-side) vertical vibrations and pilot(right side) vertical and lateral vibrations. There should belittle, if any, difference in longitudinal vibrations betweenthe center of gravity and the crew stations.

7.4.1 Center-of-Gravity Vibration Characteristics

The vibration accelerations at the 2-per-rev frequency forthree axes (vertical, lateral, and longitudinal) at the centerof gravity are shown in Figure 58. This figure presents thedata for both loading configurations investigated. With oneexception, the data indicates that the vibration accelerationsare increased by installation of the hub spring at both load-ing configurations. The increased vibration accelerationsare, however, within the acceptable limit of 0.15g up tocruise airspeed, and 0.20g between cruise airspeed and limitairspeed as specified by MIL-STD-8501.

7.4.2 Crew Station Vibration Characteristics

Vibration accelerations measured at the pilot's station (ver-tical and lateral) and the copilot's station (vertical) areshown at 2-per-rev frequency in Figure 59. In the case of thelight/aft loading configuration, the vibration accelerationswere decreased at the crew stations. In the heavy/forwardconfiguration, however, the vibration accelerations increasedfollowing installation of the hub spring. In addition to thedata presented in Figures 58 and 59, the program test pilotstated that he felt that there was increased 2-per-rev vibra-tion in all flight conditions (Appendix B).

7.4.3 Summary of Vibration Characteristics

The variations of the data presented in Figures 58 and 59 canbe attributed to normal data scatter; therefore, from the datapresented, no definite conclusion concerning the effect of thehub spring on vibration levels can be made.

132

GW = 8000 lb SYM CONFIG FLT

CG = 144 in. 0 Hub Spring 195, 197ABaseline 183A, 185

0.08

4 0.06

~0.04- _ -/

-,4

0 0.02

0.00

0.20

0.15

0.10

0.05

0.00_____

0.06

10 0.04 - -

0.02-4

U 0.000 40 80 120


Figure 58. Effect of the hub spring on center-of-gravity2-per-rev vibration acceleration in level flight.

133

GW = 8600 lb SYM CONFIGCG = 130 in. 0 Hub Spring 195, 197AC Baseline 183A, 185

0.08

0.06

0.04a)1 Nk iN

0 0.02

0.00

0.06

$4 0.04

4o 0.0

0.00

0.12

-4

ro0 0.08

413

0 0.04 1 E-

C) 0.00 _ _ _ _ _

0 40 80 120



134

GW = 8000 lb SYM CONFIG FLTCG = 144 in. E Hub Spring 195, 197A

E Baseline 183A, 185

4 0.30-4

4J 0 2 0

4J 0.100 0

C 0.00

0.08

" 0.06

0.040

P 0.02

0.00

S 0.30

a 0.20

0 .I0

>P 000480 12

0 M- 101r 0.10

04

U 0.000040 80 120


Figure 59. Effect of the hub spring on crew station2-per-rev vibration acceleration in level flight.

135

SYM CONFIG FLTGW = 8600 lbCG = 130 in. 1 Hub Spring 195, 197A

13 Baseline 183A, 185

- 0.15Hrd0 .I0

4 0.10

4J 0.050

H 0.00

- 0.12

o' 0.08.4(U

40 004

C4 0.001- __

4i 0.30-(d

$4 0.20 -

42)



~136

As stated by the test pilot, the helicopter exhibited sometendencies to wallow around during start-up when the mainrotor was below 130 rpm. In any case, the increased vibrationwas not significant or objectionable to the pilot.

As expected, the stiffness of the hub spring used during thistest program is not great enough to introduce significantchanges to the vibration levels of the UH-lH helicopter.

7.5 EFFECT OF HUB SPRING ON MAIN ROTOR COMPONENT FATIGUELIVES

During the flight test program of the nonlinear hub spring ona UH-lH aircraft, several loading parameters in the main rotorsystem were measured and recorded. These data are presentedas fatigue damage summary tables for conditions where flightloads exceeded their respective endurance limits during thetest program.

For the purpose of vibration evaluation, wind-up turns wereflown during this investigation. These maneuvers are con-sidered aerobatic in nature and would therefore not be withinthe operating manual limits prescribed for the UH-lH aircraft.Also, to generate maximum flapping, certain flight conditionswere flown outside the gross weight/center-of-gravity envelopeand are not representative of normal UH-IH flight conditions.

Considering the nature of these "extreme" conditions, they arenot considered in the evaluation of the effect of the hubspring on component retirement intervals; however, they areused to assess the damage accumulated by each component duringthe flight test program. Paragraph 7.5.1 presents the fatigueevaluation for normal operation. Paragraph 7.5.2 gives thefatigue evaluation for the complete program, including retire-ment time adjustments as required to account for accelerateddamage accrual.

7.5.1 Fatigue Damage During Normal Operation

Fatigue damage calculations were performed for the followingcomponents: main rotor, yoke, mast, and pitch horn. The re-sults are presented in Tables 11 through 13. No summary wasneeded for the main rotor yoke, since no damage was incurredon this part during the normal flight conditions. In thesetables, the cycles per hundred hours are the total per flightcondition regardless of gross weight, center of gravity, oraltitude from the appropriate fatigue life determination tablein Reference 1.

137

0 4-) r-4 0 -4 C-E0 0) 0 0

L4

0

'.40

00

U

r 4 -. Ii14 E-4

4n) 0c

0 4-) (14 En

'0 z '%0L 1E1"-H +4 U

W -4 m Ur-4O

1-' 4

C, cj4-4

04 aU>40 (a 0

o4- C,)L1.0o)L E-i V

'-4

cju U) (a U-4

0 0 0

0~ '4-4

4-) UU- 0 J 0 - ) r4

Ca.0 0) U(a ~

1384

0) 0 k.o U) r-4t~ir4 M' r4 LA 0(d 4-) (y 44~ rN i4

uL r.4 0 ,.4 u-I

to- 0 0) 0NZ4 C) LA r

0E-4 4) 0 0

0 04 4 r4 0 0rU 0 0 t C4 C

0-44-

4E-N

LT4 r-4 u

E- 4-) 4)

U) -I ULA 0 -r-

0 z~ q* - w

E- "-H -H

4-)

E-4-

4)

d)2 0

U 0 v 1r-4~ fa (~I

E-4 4-)

lz 4J

'-4r-

rL44-4

0 V

.g 4-) 4)t.4.)

r-4 ~ 4 *-

V ,-~ 139

4) 0D U' 0 ('U0 -- H

a- u 00 -R$ 0d 0

0 0 0 0 00

E- (1 0 0>HQ) 0 00

P0 0) 0

r- LA) kf

CU')

4J -Q-H r-4f

Q) LO C') LA)H-4Z C4 M Itp -

0 00 r- ON

Hx -H-H -

ixU)o 004-

0

En 00 r

-4

00H- U

.:0 MH- .4-) -4 0

o 4-) Cin 0 a.4-) t3 a

-404J I 4J 4 4

0' 4) 1- r -4-I.rq '04) U '.0 W' S 4-)-4 -H~>1 0 m~C:4 Z)- 440

140

7.5.1.1 Main Rotor Blade. As shown in Table 11, there wereonly two flight conditions within the normal flight envelopethat produced stresses that exceed the endurance limit. Thesummation of damage for these two conditions during operationwith the hub spring is 0.001721. Since this is less than thetotal for these two conditions in Table 12 of Reference 8 forthe basic aircraft, no reduction in the main rotor bladeretirement interval should be required for aircraft operatingwith the nonlinear hub spring.

7.5.1.2 Main Rotor Mast. During the flight test evaluationof the hub spring within the normal envelope, only two flightconditions produced stresses greater than the mast endurancelimit. The damage fraction total for these two conditions is0.001751. In Reference 1, the mast is substantiated for anunlimited fatigue life with no stress greater than the en-durance limit. Therefore, the only damage expected for themast is that shown in Table 12. From this damage fraction, afatigue life of 57,110 hours was calculated. Since the lifeis greater than the recognized 25,000-hour service life, anunlimited fatigue life would be predicted for the main rotormast during normal operation with the nonlinear hub spring.

7.5.1.3 Main Rotor Pitch Horn. An evaluation of the mainrotor pitch horn was conducted, since its endurance limit isthe lowest in the rotating control system in terms of pitchlink axial load. The damage summary in Table 13 shows a totalof 0.000464 for the conditions where measured loads exceed theendurance limit. Since this total is less than the 0.001721total for the basic aircraft (Table VII of Reference 8), noreduction in the retirement intervals for the main rotorcontrols is anticipated.

7.5.2 Fatigue Evaluation of Complete Test Program

Tables 14 through 17 present fatigue damage summaries for themain rotor blade, yoke, mast, and pitch horn. The flightloads used are the top of scatter loads where maneuvers wereflown more than one time. The fatigue damage is the real-timecycle-counted total using each measured cycle as it occurredwith no projected spectrum. Where conditions were repeated,

8Orr, P., McLeod, G., Goodell, W., FATIGUE LIFE SUBSTANTIA-TION OF DYNAMIC COMPONENTS FOR THE UH-ID HELICOPTER EQUIPPEDWITH THE 48-FOOT DIAMETER ROTOR, BHT Report Number 299-099-135, Bell Helicopter Textron, Fort Worth, Texas, June 1964.

141

tUl 0 0 0 -40 )0 0 0C Y)00 C 0 00 it4-) C>00 00 0 00 00 mfl0 00a00 Cl)EO 0 00 00 00 0 00 00 00 0 0 0

0 0 00 00 000 00 00 00 0 00 0

4

W4 0

OD0

C.)00

E-4 0

I-I 0) -q,0

0 Wo-)U-)+

E4P U r - r 1 Cl) Cl) - (N N~ Cl) 1ir4 0 C-4 -4 H W I

0 ~ 0 01 a UU Gi a)-0 a 0 -4 4-)

Z" 0 00 'U 'U 'U 0'r-40 0 0r-4-4 0 00 a) s -

CH

1142

co 0 04-) 00aO 00aE-4 0 0a

0

U) '4-40

4)4C.)0-4 LA

4) 0

0. 0 0004C

E-z1

E-1 E-4 ) w

"(I H4))

P4 4JL W 4-)04 ur LO4- 4)

>4 En (ts 0 4 -)P 04

4Z4 44) 0

0 04): 0 ::

W)1) 0 0 0

0 000t 1

0 tp 4)

'44 -,

4 00 a

NOD

0E-4

0 010 4-)

0 04-)l

-H .4-)

0) Q)o0 0 4-)

4-)

4) 0 -P-g 4-4 4J

4--4

H

143

N r- - )N 0 -0

0 C ~~ .0O0 a0 0 0 u. ..4. . .0.00.0.0.

E-4 0 0 0 0 0 0 0 0 0 0

4-

u 4-)

0

E-tH

rX4)

E-4 0 4

0

E-4 04 Ulo0

0 v ) r--%40

24 Wz .4 r-I 0 m0 00 0 r-4z 24 r- 0 00 0 00(a0

C,, '0 .rz 0

0 0r-H >1

4 u

244-

a)0 0

9: 4)-6 14( 00

0 Ad VOl.rJ 4-)4- 0 0.4. 0 4-iC - r- N 4J 4J.H C) -1

V I 4J) Q 01 10 00

0 m - 4-J4Jr-4 'N .4-)Ul Vo Co Co .4.- 4 r-4 (a4)

w- 1 0 0 - Z 4 I I 44 4..4J t 0 Q 04(

, t ro 0I 0 VV0 0)34 -j e-4 4 W- 4

-eH VO d (a fu0 w0 0)(a t (a vi-I r- W 4-~b 0 : 3: 3: ' )Ota V) 3- ) 0) ) N 4) 0) 4 ju o

(04-) 4-) H MV '0 ( V4) (3 00 O H,-rq - 0)0()

U)V

144

0- m 0 - N 0 a a -4- 0c O' 4C C'0 0 0 0 m 0 qdq

000 0000000 0 0 0 00 0k-4 4-)E-4 000 000 0 00 00 0 0 010 E

'4-)

00£1144

0a)

u 0

~0 0~ OD)

0~4 01

E-4fl44 (134E- 00 00 00 00 0

'-4 00- 0

0 4- U 4

ix 00 -L rI r D N u -

0 Ix to ) 0 0 -I qr L) 0 r- r-4C14 m 001).E-4 4 - 0 0 0 0 0 r4 0 0 0 M 0

0 Z4 - 00 00 00oo00 0-r-

04-4 IIco

P4

*'-gr 000

£44.

>4145

the damage fraction is raised in multiples of occurrences.The damage fractions shown in the tables are expressed as anaccrual rate per flight hour during the 20-hour flight pro-gram. These rates were then compared with the hourly ratespredicted by the retirement intervals presented in Reference8. The main rotor pitch horn showed a damage rate that wouldindicate an equivalent flight time of 104 hours during the 22-hour test program.

Since the main rotor mast had no damaging conditions duringnormal operation without the nonlinear hub spring, it isconsidered that if returned to normal use in the standard UH-lH configuration, it would accumulate no further fatiguedamage. Therefore, the .008857 damage fraction in Table 16 isconsidered to be the total to date and, as stipulated above,the total for future operation unless new test programs areaccomplished. Since this total is far less than the 0.5 totalat which retirement would be recommended, it is consideredthat the main rotor mast on this aircraft would retain itsunlimited retirement interval.

7.5.3 Fatigue Evaluation Summary

There is no noticeable trend for variations in loads duringforward level flight with the nonlinear hub spring versusbaseline loads for the same conditions. Also, as stated inParagraphs 7.5.1 through 7.5.1.3, the fatigue damage on thecritical components during operation with the nonlinear hubspring is within the values established in Reference 7 for thebasic aircraft. Under these considerations, it is thereforeconcluded that the nonlinear hub spring can be installed onUH-lH aircraft without lowering the present retirement inter-vals, as long as the operational flight limits specified inthe operating manual are observed.

7.5.4 Recommendation

As stated in Paragraph 7.5.2, certain dynamic components didaccumulate fatigue damage at an accelerated rate per flighthour compared to normal operation. It is therefore recom-mended that the retirement intervals for all time-life com-ponents on the test aircraft be lowered by 100 hours as cal-culated for the main rotor pitch horn in Table 17. Thisadjustment is considered necessary on all components due tothe lack of measured flight loads on certain aircraft systemsduring the test programs. It is emphasized that this actionis applicable to UH-lH test aircraft, U.S. Army S/N 73-21805(BHT S/N 13493) only, and does not apply to all such aircraft.

146

1. F

7.6 PROBLEMS ENCOUNTERED AND SOLUTIONS

During the course of the flight test program and as a resultof inspection of the hub spring after removal from the testhelicopter, several design deficiencies and problem areas werefound. Further analysis also yielded a problem concerning thetransmission main case. The transmission case was not, how-ever, instrumented during this flight test program. For alldesign deficiencies or problem areas, possible solutions havebeen identified below.

7.6.1 Structural Bond Failure

During the flight test program, the structural bond betweenone of the elastomeric shear pads and the side frame of thehub spring failed. This particular failure would not occur ona production hub spring. In the case of the experimental hubsprings, the elastomeric shear pads were bonded to metal shimsduring the molding process. These shims were then cemented tothe side frames, after the parts had cooled, following mold-ing. It was this cemented bond that failed. In the produc-tion hub spring, the side member is incorporated into the moldand the elastomer bonded during the molding process. This isa much stronger bond and will not fail during operation.

7.6.2 Abrasion and Tearing of Elastomerics

Inspection of the elastomerics during the flight test programand following removal from the test aircraft showed abrasionloss of material on the snubber face and a tear of the elasto-mer of the shear pads. Abrasion of the snubber was caused byrelative motion of the snubber following contact with themast. It was determined that the snubber face and mast con-tact point were not parallel when snubber-mast contact wasmade and was the cause of this motion.

This problem will be eliminated by:

- Modification of the snubber mold to insure that mastcontact point and the snubber face are parallel;

- Increased width of the mast contact point to insure thatthe compressed snubber does not wrap around the mast; and

- Using an abrasion-resistant elastomer for the snubberface.

147

Tears in the elastomer at the upper inside corners of allshear pads were found to have started at an unrelieved stressconcentration. The stress at this location can be relieved bymodification of the mold to extend a fillet and by changingthe mold parting line.

7.6.3 Excessive Stress in Side Member

The stresses in the side member of the hub spring were suchthat fatigue damage was occurring at a rate that would makethe part unacceptable as production hardware. These damagingstress levels were present for all flight conditions whenflapping was greater than 4.5 degrees. A modification thatincreases the material cross area of the side member in orderto reduce stresses below the endurance level will be requiredfor this part. This modification can be accomplished withoutintroducing interference problems. Additionally, correctingthe angle of the snubber face will reduce the side memberstress.

7.6.4 Fretting of Mast and Yoke

Inspection of the mast and yoke following flight test showedthat there might be a problem with fretting of the mast andyoke under the clamps. The possibility of fretting can beeliminated by use of bluecoat or scrim buffers, or redesign ofhardware to eliminate motion of the clamps.

7.6.5 Transmission Structural Analysis

During the post-flight analysis of the loads introduced by thehub spring, it was determined, by analysis, that the main caseof the UH-lH could no longer be shown to have a positive mar-gin of safety with hub spring installed.

This is a result of a combination of several factors. First,the revision downward of stress allowables for magnesium, fromwhich the transmission main case is cast, found in MIL-HDBK-5,"Metallic Materials and Elements for Aerospace Vehicle Struc-tures." Secondly, the new casting factors to be used forClass 1A castings that are found in AMCP 706-203, "HelicopterEngineering, PartThree, Qualification Testing" (Section 7-4paragraph 7.4.2.2.1.6), and the additional stresses resultingfrom hub moments produced by installation of the hub spring.When all factors are considered, the transmission is calcu-lated to be static strength limited. A fatigue analysis ofthe transmission under hub spring loading has not been per-formed. A solution to this problem would be to replace themagnesium main case with an aluminum one similar to that foundon the Bell Model 212 helicopter. Another approach would be

148

to attempt to qualify the transmission by a static proof testthat subjects the transmission to the maximum stresses intro-duced by the hub spring in addition to the maximum stressesencountered during the structural demonstration flight program.

7.7 SUMMARY OF FLIGHT TEST RESULTS

The flight test program demonstrated that the nonlinear hubspring did, in fact, reduce main rotor flapping and thereforeimproved the UH-lH mast bumping margin of safety. As shown,main rotor flapping was reduced at all conditions tested, andlateral control power was increased at low-g flight conditionsby installation of a nonlinear hub spring in the test air-craft.

As expected, handling quality characteristics of the UH-lHhelicopter were either improved or unchanged by installationof the hub spring. As an example, control power was increasedapproximately 15 percent in the pitch and roll axis, andstatic longitudinal stability increased at the aft center-of-gravity configuration. Also, improved control margins weredemonstrated.

Although higher stresses were measured in the main rotor mastand hub at main rotor flapping angles greater than 4 degreesduring flights with the hub spring installed, there would notbe a requirement to decrease mast, hub, or blade retirement oroverhaul intervals.

149

8. CONCLUSIONS AND RECOMMENDATIONS

8.1 CONCLUSIONS

A nonlinear hub spring was successfully designed and flighttested for the UH-lH helicopter. This design met the retrofitcriteria as well as the additional load constraint that nomain rotor component time-between-removal interval would bechanged. Flight testing of the nonlinear hub-spring-equippedUH-lH helicopter proved the concept of the nonlinear hubspring as a method for improving the mast bumping safetymargin. Flapping was reduced in all flight conditions withthe greatest reduction being recorded when baseline flappingwas above 5 degrees. The nonlinear hub spring also greatlyimproved the low-g control power of the UH-lH helicopter.

Following an analysis of helicopter tactical instructionmanuals and observation flights over the UH-IH NOE instructioncourse at Fort Rucker, Alabama, it is concluded that currenttactics do not cause excessive flapping if normal operatingprocedures are followed. There were flight conditions, how-ever, during which main rotor flapping exceeded 8 degrees. Itis concluded that if the loading condition is heavy and for-ward, 30-knot rearward and right sideward flight and rightlateral dashes will result in main rotor flapping exceeding 8degrees.

The comparison of main rotor flapping as predicted by thehybrid version of C81 with rigid rotor blades and digital C81with elastic rotor blades shows that the main rotor flappingpredictions are not greatly altered by elastic blades.

Simulation of mast loads during conditions of in-flight flap-ping stop contact can be satisfactorily accomplished utilizinga modified version of the rotor pylon hybrid computer program(ARHF01). An aeroelastic model rotor was satisfactorilytested on a whirl stand providing additional flapping and mastcontact loads data. Comparison of data obtained from theaeroelastic model rotor and test helicopter measured data withthat predicted by ARHF01 demonstrate excellent correlation.

A main rotor design criterion is suggested for teetering rotorhelicopters. Using ARHF01 results, mast loads were predictedwith a simplified method that will allow contact loads to beincluded in the structural design of rotor concepts.

150

8.2 RECOMMENDATIONS

The nonlinear hub spring has been demonstrated to be a feasi-ble concept for increasing the mast bumping safety margin forthe UH-lH. The problems identified during these tests can beeliminated while designing for production articles to retrofitthe UH-IH fleet. The nonlinear hub spring should be subjectedto qualification testing. The shortcomings of the nonlinearhub spring identified during this test should be corrected andthe additional flight testing and fatigue testing conducted.Also, user tests of the nonlinear hub spring equipped UH-IHhelicopter should be conducted. The application of the non-linear hub spring as a hub flapping restraint mechanism onother helicopters with teetering main rotor systems should bestudied.

A proof test of the current transmission should be conductedin order to determine if adequate structural safety marginsexist with the additional loads introduced by the hub spring.If positive margin cannot be demonstrated, further effort isneeded to qualify a stronger transmission assembly.

Additional testing and analysis is required to fully establishthe design criteria for various pylon configurations, rotorthrust levels, and rotor blade designs.

151

9. REFERENCES

1. Dooley, L. W., ROTOR BLADE FLAPPING CRITERIA INVESTIGA-TION, Bell Helicopter Textron, USAAMRDL Technical Report76-33, Eustis Directorate, U.S. Army Air Mobility Re-search and Development Laboratory, Fort Eustis, Virginia,December 1976, AD A034459.

2. Dooley, L. W., and Ferguson, S. W. III, EFFECT OF OPERA-TIONAL ENVELOPE LIMITS ON TEETERING ROTOR FLAPPING, BellHelicopter Textron, USARTL Technical Report 78-9, AppliedTechnology Laboratory, U.S. Army Research and TechnologyLaboratories, Fort Eustis, Virginia, July 1978.

3. VanGaasbeek, James R. and Landry, L. Matthew, Jr., HYBRIDCOMPUTER PROGRAM ARHF02, Technical Data Report 699-099-015 to be published by Bell Helicopter Textron, FortWorth, Texas.

4. Van Gaasbeek, James R., ROTORCRAFT FLIGHT SIMULATION COM-PUTER PROGRAM C81, WITH DATAMAP INTERFACE, Volume I -User's Manual, to be published by the United States ArmyResearch and Technology Laboratories, Applied TechnologyLaboratory, Fort Eustis, Virginia.

5. Bonham, D. G., and Hollifield, P. J., UH-l HUB SPRINGINSTRUMENTATION TEST PLAN, BHT Report Number 699-099-095,Bell Helicopter Textron, Fort Worth, Texas, June 1978.

6. Hollifield, P. J., UH-I HUB SPRING TEST PLAN, BHT ReportNumber 699-099-099, Bell Helicopter Textron, Fort Worth,Texas, August 1978.

7. Harris, F. D., ARTICULATED ROTOR BLADE FLAPPING MOTION ATLOW ADVANCE RATIO, Journal of American Helicopter Society,January 1972.

8. Orr, P., McLeod, G., Goodell, W., FATIGUE LIFE SUBSTAN-TIATION OF DYNAMIC COMPONENTS FOR THE UH-lD HELICOPTEREQUIPPED WITH THE 48-FOOT DIAMETER ROTOR, BHT Report 299-099-135, Bell Helicopter Textron, Fort Worth, Texas, June1964.

152

10. LIST OF ABBREVIATIONS AND SYMBOLS

B0 Coning angle, deg

B1 Longitudinal cyclic feathering, deg

B Lateral cyclic feathering, deg2

D Drag force, lb

F Factor of Safety

H Hub longitudinal inplane shear force, lb

IL Inplane shear force, lb

KEFFT Theoretical effective spring rate, ft-lb/deg

KFS Flapping stop (mast) spring rate, ft-lb/deg

KL Hub restraint linear spring rate, ft-lb/deg

KNL Hub restraint nonlinear spring rate, ft-lb/deg

KT Effective spring rate due to rotor thrust, ft-lb/deg

KTAS Knots True Airspeed, knots

L Lift, lb

M Mast bending moment, ft-lb

M1 Mast perpendicular bending moment, ft-lb

M 1 Mast parallel bending moment, ft-lb

MFS Bending moment due to flapping stop contact, ft-lb

MTOT Total bending moment due to flapping restraint, ft-lb

r Order of nonlinearity

S Mast shear force, lb

T Thrust, lb

T Thrust coefficient

c

153

VH Level flight airspeed requiring maximum continuous

power, knots

VNE Never exceed airspeed, knots

WUT Wind-up turn

X Distance from teetering axis to a point at which a mast

moment is computed, ft

XLFS Distance from teetering axis to the flapping stop con-tact point on the mast, ft

Y Hub lateral inplane shear force, lb

P Hub flapping angle, deg

00 Coefficient representing flapping angle independent ofblade azimuth angle, deg

Coefficient representing amplitude of pure sine flap-s ping motion, deg

Coefficient representing amplitude of pure cosine flap-c ping motion, deg

PB Flapping angle at which nonlinear hub spring is engaged,deg

Ps Hub flapping at flapping stop contact, deg

Ox Commanded hub flapping beyond flapping stop, deg

0* Value of hub flapping beyond flapping stop

63 Pitch-flap coupling angle, deg

0 Main rotor rotational speed, rad/sec

Main rotor azimuth angle, rad

o Angular deflection, deg

075 Blade collective pitch angle at 3/4 radius, deg

154

APPENDIX A

TABULATION OF ARHF01 INPUTS FOR THE UH-lHWITH NONLINEAR HUB SPRING

Description Name Value Units

LIFT CURVE SLOPE A 5.70000 NOROTOR PRECONE ANGLE AO 0.4dO0E-01 RAOSAME AS BBtTA 3 10.72 FTVH TO SLADE MASS(CUN UbETA 10.72 FTFLAPPING STOP fFS 0.2U07 HADVH TO OLAGE MASS (EP OZETA 9.973 FT.SKID GEAR JAMPING CG 500.0 ND

PYLON CONTROL COUPLING TERMSLAT TO PYLON ROLL DAIOPP 0.0 NDLAT TO PYLON PITCH DAIOPT 0.0 NDF/A TO PYLON ROLL UBIDPP 0.0 NOF/A TO PYLON PITCH 001OPT 0.0 NO

CONING VH OFFSET E 3.240 FTCONING VH OFFSET EtETA $.240 FTIP VH OFFSET LZETA 6.000 FTFLAT PLATE AREA F 18.35 FT**2GRAV. ACCELERATION G 32.17 I-T/StC2LOCK NUMIER GA 4MA 6.293 NDHUB JNDERSLING HU 0.4333 FTFLAPPING INERTIA I Ft: TA 1i11. SLUG-FT2FUSELAGE PITCH INERT IF O.1108E 05 SLUG-FT2FUSELAGE ROLL INERT I( 3162. SL UG-F TZIR VH BLAJE INLRTIA IZ1:TA Sb.3 SLUG-FT2HUR SPRING KBETA 0.18',iJE 05 FTLa/NADSOUTH. COEFF (CON) KBETAP 1.229 NOFLAP STOP SPRING KFS 0.1300E 06 FTLS/RADSKID VENT SPRING KG 35O0. FTLB/RADPYLON STOP SPRING KP 3700. FTLu/RADCONT40L SYS SPRING KTHETA O.JOOE 05 FTLB/RADIP VH SPRING KZETA 0.5963E Ob FTLB/RAOSOUTH. COEFF (IP1 KZETAP 0.736 NO)8LAOe MASS MR 4.010 SLUGSENGINE TORJUE ME 0.0 1- TLtFUSELAGE MASS MF 248.7 SLUGSLANDING GEAR FRICTN MU 0edOOO NOPYLON ROLL STOP PPO 0.3500 RAOPYLON PITCH STOP PTO 0.3500 HAD3/4 OLADE RADIUS R 18.00 FTBLADE HAD GYR.(IP.CG RG 5.391 FTAIR DENSITY RHO 0.2233E-02 SLUG/FT3CONING FRED SQUARED WBETA2 14.4 Ik/iSC)2

IP FRED SQUARED WZETA2 fib?. (R/SECJ2FUS CG UNDER HUB ZCG 7.312 FTHUB MASS MH 0.0 SLUGSBLADE CONING INERTIA IOETAO 170.1 SLUG-FTZ

DELTA3 0.0 HADINDUCLO VEL TIME CON TAUI 0.2500 SECXDIST CG TO PILOT XPT 0.0 FTZDIST CG TO PILOT LPT 0.0 FTROTOR RPM RPM 424.0 PMPITCH-LAG COUPLING ALPHAJ 0.0 OEGLAT TO PYLON VERT D$3IOPL 0.0 NOF/A TO PYLON VERT DAIOPZ 0.0 NDPITCH TO PYLUN VERT OTHDPZ 0.0 NOPITCH TO PYLON PITCH TTI DPT 0.0 NOINPLANE MOUE UAMPING ZZETA 0.4000E--l PCTMAST TILT ANGLE THTI -O.2TzyE-01 RAOPIN TO FLAP STOP ARM XLFS 0.6063 FT

XILFS 1.333 FTX2LFS 3.8116 FT

SPRING CONTACT ANGLL RBRK 0.tv96IE-OI RAOSPRING CONSTANT KUBRK 4.J63 F TLU/OEGSPRING EXPONENT EXPURK 3.000 NO

155

APPEflDIX A (Concluded)

LANDING GEAR DATADE SCRIPTI ON NAME VALUE UN I T S

FWO GEAk DIbT TO C4. X(.F I0.)o FTAFT GEAR DIST TO CG. XGA 1"1.00 FTFWD GEAR SPRING KGF O000. Ld/FTAFT GEAR SPRING KGA 1000. LB/FTFUO GEAR DAMPER CGF 10.00 LOSEC2/FTAFT GEAR )A 4PER CGA 1U.00 LUhEC2/FTNO. FWD GEARS NFG 2NO. AFT GtARS NAG 2Z DIST. CG TO GEAR ZLG 3.000 FTGEAR TILT ANGLE IHFG -0.0 kAJ

GEOMETRYDESCRIPTION NAME VALUE UNITS

X DIST CG TO ROTOR XCG -0. 95b0 FT

AERORIG|OOUOY FUSELAGE AE.DYNAMIC COEFFICIENTSi

FLATPLATE LIFT CLOF 1.00000 FT**2LIFT(ALPHAI CLAF 10.0000 FT2/DEGLIFTIDALPHA) CLOAF 0.O0000E 00 FT2/DE.GSFLATPLATE DRAG COOF .. 00000 F T**2DA AG(ALPHA) COAF 10.0000 FT2/ULGDRAO(DALP'-A) CODAF 0.I0000F 00 FT2/DEGSORAG(ALPHA2) LDA2F .0IOOOOOE-01 F T2/DEG2PITCHING r4UMLNT CMOF 0.200000 FT**3PITCH( ALPHA) CMAF 2.00U00 F T3/DEGPITCH|(DALPHA) CMDAF J.200040 FTJ/DrGSPITCH(ALPHA**2) CMA2F 0.13)0000t-0I FT3/DEG2X-CENTtR OF PRESS. XAC 0.b00000 F rZ-CENTER OF PRESS. ZAC 1.00000 FTX-DOWNmASH FACTOR DWNSX 0.100OOUE U0Z-DUWNWASH FACTUR OWNSZ 0.100000E 00X-FREE-STREAM-VCL EFSO 116.300Z-FREE-STREAM-VEL ZFSO -9.6eJ80

NORMAL MODE DATA

MOE UHIH NASTRAN MODES WITH PARTIAL HUB MASS SET MH TO 0.0 (01-22-80)

MUOL

1 2 34 5 6

FREiUENCY WI 20.3-23 2J3.744 - 43.910? -0.0 -J.0 -0.0M4ASS MA 26.3b?9 18.1293 9.2954 -0.0 -0.0 -0.0JAMPING I 0.0500 0.0500 0.0500 -0.0 -0.0 -0.0HUB x UE3LXI 0.0114 -1.0000 0.0362 -0.0 -0.0 -0.0HUs Y DELYI 1.0000 0.00b4 -0.0047 -0.0 -0.0 -0.0HUs z DELZI -0.0439 0.0722 - 0v. 33 -0.0 -0.0 -0.0MAST PITCH DELTI -0.0022 0.217t -0.0039 -0.0 -0.0 -0.0MAST ROLL DELPI 0.2342 0.00t5 -0.0024 -0.0 -0.0 -0.0:V.ON PITCH DELPTI 0.0 -0.0 -0.0 -0.0 -0.0 -0.0PYLON ROLL OELPPI 0.0 -0.0 -0.0 -0.0 -0.0 -0.0PYLON Z DELZTI 0.0 -0.0 -0.0 -0.0 -0.0 -0.0CG x DXCGI 0.0 -0.0 -0.0 -0.0 -0.0 -0-0CG z DZCGI 0.0 -0.0 -0.3 -0.0 -0.0 -0.0PILOT STA X DXPT1 0.0 -0.0 -0.0 -0.0 -0.0 -0.0PILOT STA Z OZPTI 0.0 -0.0 -0.0 -0.0 -0.0 -0.0

LOAS SAMPLE MODULE

STATION NO. 123MODE 1 2 3 4 5 6

x 1.20000 2.20000 0.32000 0.44000 O.8000 0.01000Y 1.20000 2.20000 0.42000 0.44000 0.68000 0.01000

Z 1.20000 2.20000 0.32000 0.44000 0.68000 0.01000

156

--- - - ' . -..

C 0

APPENDIX B

QUALITATIVE PILOT FLIGHT REPORT ON THEUH-l NONLINEAR HUB SPRING

GENERAL

During our flight testing, two helicopter configurations weretested: (1) the baseline, standard UH-lH helicopter, and(2) the UH-lH helicopter with hub spring installed:

This qualitative report will addresss only flight profilesthat appear to me to be signficantly different from standardconfiguration UH-lH and UH-lH with hub spring installed.

FLIGHT TEST BASELINE

The flight tests to obtain the baseline data were completedwithout any major problems; however, the following should bementioned.

(1) During rearward flight at 30 knots with center of gravityadjusted to 129.5, a 9-degree (80 percent) main rotorflapping was demonstrated. It should be noted that thecyclic was on the aft cyclic stop to achieve the 9-degreeflapping.

FLIGHT TEST HUB SPRING INSTALLED

(1) During start up when main rotor speed was below 130 rpm,the helicopter exhibited some tendencies to wallowaround. This condition was not serious but differentfrom the standard UH-lH configuration.

(2) An increase of two-per-rev vibration in all flight con-ditions was noted.

(3) During high Tc turn maneuvers at VNE, collective fixed,when 1.7g's or above was obtained, vertical vibrationlevel increased with noted cyclic feedback. This in-formation was validated by our telemetry which showedhigh pitch change link loads.

(4) In rearward flight at 30 knots with center of gravityadjusted to 128.88, the maximum main rotor flapping neverexceeded 7 degrees. The cyclic at no time made positiveaft cyclic stop contact. Telemetry indicated high loadsin hub spring parts and main rotor mast during thismaneuver.

157

~i

(5) Low-g investigation did not exhibit any abnormal quali-ties. Control response appeared the same as baselineflight test. During baseline, .4g was obtained andduring flight test with hub spring installed, .25gwas obtained.

SUMMARY

The UH-I nonlinear hub spring appears to do the job for whichit was designed. It improves the UH-I by giving the helicop-ter a new safety margin to keep it out of mast bumping. Ifeel that more testing is necessary to determine flight limitsand serviceability.

/S/ J. HonakerExperimental Test Pilot22 March 1979

158

IIIIIIIIIIIIIIfllfllfl IIIIIIIIIII - dtic. · PDF fileas a means to reduce the levels of...

Documents

Transcript of IIIIIIIIIIIIIIfllfllfl IIIIIIIIIII - dtic. · PDF fileas a means to reduce the levels of...