A 008995

AD-AO08 995

FABRICATED HELICOPTER TRANSMISSION HOUSING ANALYSIS

Alexander Korzun, et al

United Aircraft Corporation

I

Prepared for:

Army Air Mobility Research and Development'Laboratory

Jarnuary 1975

II

DISTRIBUTED BY:

National Technical Information ServiceU. S. )EPARTMENT OF COMMERCE

EUSTIS DIRECTORATE POSITION STATEMENT

The information contained in this report is a result of research

conducted to investigate advanced structural concepts for a helicop-

ter main transmission housing.

The objective of the fabricated helicopter transmission housing study

was to provide a housing design that was structurally superior to the

conventional magnesium cast housing, with additional attributes of

reliability, maintainability, and cost.

The report has been reviewed by the Eustis Directorate, U.S. Army

Air Mobility Research and Development Laboratory and is considered

to be technically sound.

This program was conducted for the Military Operations Technology

Division under the technical management of Hr. L. Thomas Mazza,

Technology Applications Division.

...,. ........ ....."" ....'" ....."

plslUl1hh/AVAILA6hl:I1 CODES

DISCLAIMERS

The findings In this report are not to be construed as an official Department of the Army position unless sodesignated by other authorized documents.

When Government drawings, specifications, or other dats are used for any purpose other than in connectionwith a definitly related Government procurement operation, the United States Government thereby incurs noresponsibility nor any obligation whatsoever; and the fact that the Government may have formulated, furnished,

or in any way supplied the said drawings, specifications, or other data is not to be regarded by implication orotherwise as in any manner licensing the holder or any other person or corporation, or conveying any rights orpermission, to manufacture, use, or sell any patented invention that may in any way be related thereto.

Trade names cited In this report do not constitute an official endorsement or approval of the use of suchcommercial hardware or software.

DISPOSITION INSTRUCTIONS

Destroy this report when no longer needed. Do not return it to the originator.

UnclassifiedSECURITY CLASSIFICATION OF THIS PAGE ( ohen Date Enlered)

REPORT DOCUMENTATION PAGE READ INSTRUCTIONSBEFORE COMPLETING FORM

I. REPORT NUMBER J2. GOVT ACCESSION NO. 3. RECIPIENT'S CATALOG NUMBER

USAAIMDL-TR-74-14 AtPA 00 9?r4. TITLE (and Subtitle) S. TYPE OF REPORT & PERIOD COVERED

Final Report - December 18,FABRICATED HELICOPTER TRANSMISSION HOUSING 1972 to October 18, 1973ANALYSIS 6. PERFORMING ORG. REPORT NUMBER

7. AUTHOR(#) 11. CONTRACT OR GRANT NUMBER(@)

Alexander KorzunStephen Schuman DAAJ02-73-C-0022

9. PERFORMING ORGANIZATION NAME AND ADDRESS 10. PROGRAM ELEMENT. PROJECT. TASK

AREA A WORK UNIT NUMBERSSikorsky Aircraft Division of United AircraftCorporationStratford, Connecticut 06602 TASK 1F162203Al1902

II. CONTROLLING OFFICE NAME AND ADDRESS 12. REPORT DATE

Eustis Directorate, U.S. Aror Air Mobility January 1975Research and Development Laboratory 13. NUMBER OF PAGES

Fort Eusts. Virginia ;604 30014. MONITORING AGENCY NAME 6 ADDRESS(iI different from Controlling Office) IS. SECURITY CLASS. (of this report)

Unclassified

IS. DECL ASSI FICATION/ DOWNGRADINGSCHEDULE

16. DISTRIBUTION STATEMENT (of il Report)

Approved for public release; distribution unlimited,

17. DISTRIBUTION STATEMENT (of ihe abstracl entered in Block 20, IH different from Report)

IS. SUPPLEMENTARY NOTES Repioduced by

NATIONAL TECHNICALINFORMATION SERVICE p[JIW .S SUBJECT TO (HANUS Department ol Commece O M

Springield, VA. 22151

IS. KEY WORDS (Continue on reverse side If nece.eary and Identity by block number)

Fabricated Helicopter Transmission Housing, Magnesium Casting, Corrosionresistance, Custom 455 Stas.nless Steel, Ti-6A1-4V Titanium, CRT, NASTRAN,Welded steel truss, Titanium truss

20. ABSTRACT (Conlinue on revereD side It necessary and Identify by block number)

Tne objective of the fabricated helicopter transmission housing study was toprovide'a transmission housing design superior to the conventional magnesiumcast housing. The new design is a welded steel fabricated truss-like structureapproximately 15 percent lighter than the magnesium casting, and in small quantities offers a 30-percent reduction in cost. The truss structure is superiorto the casting in reliability and maintainability. It is corrosion resistantand not susceptible to creep. The fabricated housing is feasible using today's

DD , AN 73 1473 EDITION OF I NOV65 ISSOSOLET I UnclassifiedSIN 0102-014-6601 1I SECURITY CI.ASSIFICATION OF THIS PAGE (norn Data Xritred)

Uncl ssified.UIJRITY -'.ASSIFICA '.Wi OF THIS PAGE(Wen Date Entered)

materials, analytical methods, and fabrication tecniques.

The aircraft selected as the vehicle for this study was the U.S. Army CH-54Bhelicopter. The new fabricated truss-like housing was designed to be inter-changeable with the present CH-54B magnesium main transmission housing. It meetall the interface and functional requirements of the cast housing design. Loadsfor flight and crash conditions, as well as stiffness criteria, were developedto permit structural analysis and comparison with the existing casting design.

During preliminary design, seven candidate truss arrangements were consideredin conjunction with material and joining trade-off studies. Of the sixteenmaterials considered, only titanium Ti-6Al-hV and stainless stell Custom 455had an adequate combination of ultimate strength, fatigue joint strength, andhigh-temperature creep properties. Of the five joining methods investigated,only welding had the static and fatigne joint efficiencies essential for afabricated truss-like housing made of titanium or steel.

The structural analysis of the seven candidate truss arrangements with titaniumand steel was simplified by grouping the arrangements into three fundamentallydistinct configurations. The geometry of each arrangement was coded and thencorrected through use of a cathode-ray tube (CRT) display program. The analysiswas then performed by a redundant NASA structures analysis program (NASTRAN).During the analysis, the results were iterated and compared with a c'onventionalcasting to satisfy stiffness and support conditions.

From an anlysis of the preliminary results, a single detail configuration evolvea truncated conical structure with rings at each bearing or interface connectedby many structural members forming a truss. Further analysis yielded the finaldetail design configuration: a welded steel truss arrangement weighing 305 lbs.An alternate titanium truss arrangement weighs 313 pounds. The present magnesiutransmission housing weighs 362 pounds.

To determine the effects on the housing design of a ±25 percent change in inputtorque, the CH-54B aircraft was scaled parametrically, and new applied housingloads were developed. Included in the load development were crash loads andflight loads as well as servo, scissor, accessory, and internal gear and bearingloads. For a 25-percent increase in input torque, the scaled-up welded steeltruss fabricated housing weighs 362 pounds. For a 25 percent decrease, thehousing weighs 254 pounds. The alternate welded titanium truss fabricatedhousing weighs 386 pounds, for a 25-percent increase, and 262 pounds, for a25-percent decrease in input torque.

Areas requiring additional development effort before the fabiicated housingconcept can become operational include a welding development program andimproved analytical methods. The welding development program is necessary toimprove joint integrity through better quality control of welded joints. Improv danalytical methods should in..'ude determination of skin covering strength effect ,a method for plastic and thermai stress analysis, development of a preprocessorto reduce coding of input data, and development of a specific program to handlethe required structural analysis.

A fabricated housing should be built and tested as outlined in the proposedfollow-on program to verify cost, analysis, fabrication techniques, and weightsavings.

UnclassifiedSECURITY CLASSIFICATION OF THIS P .GE("4hon Data Enter d)

--

PREFACE

The program reported herein was conducted during a nine-monthperiod from December 18, 1972, to October 18, 1973, for theEustis Directorate, U. S. Army Air Mobility Research and Devel-opment Laboratory (USAAMRDL), Fort Eustis, Virginia, underContract DAAJ02-73-C-0022, Task 1F162203A11902.

USAAMRDL technical direction was provided by Mr. L. ThomasMazza, of the Eustis Directorate, Technology ApplicationsDivision.

The program was conducted at Sikorsky Aircraft, Stratford,Connecticut, under the technical supervision of Mr. Harold K.Frint, Sikorsky Aircraft Transmission Design and DevelopmentSection. Principal investigators for the program wereMr. A. Korzun and Mr. S. Schuman of the Transmission Designand Development Section, and Mr. W. Degnan of the Material and-Processes Section.

I '

TABLE OF CONTENTS

Page

PREFACE 1.. ....... ....... . i

LIST OF ILLUSTRATIONS .............. 6

LIST OF TABLES ................. 16

INTRODUCTION . . . , . . . . . . . . . . . .. 19

DESIGN APPROACH ................. 26

Aircraft Selection ............. 29

Description of Aircraft . . . . . . . . . . 29

Description of Drive Train ......... 30

Description of Main Gearbox . . . . . . . . 33

Description of Maina Transmission CLsting . . 43

Design Criteria for Main Transmission • • • 47

Design Criteria for the FabricatedHousing . .. .. . .. .. .. . . . . . . 48

Interface Requirements for FabricatedHousing .................. 49

PRELIMINARY DESIGN . . ....... . . . .. . 50

Load Development .............. 50

Stiffness Criteria . ........... 74

Selection of Materials and JoiningMethods . .. . . .. .. . .. .. .. . . 78

Candidate Materials ........... . 79

Rationale for Joining Method Selection . . . 90

Candidate Materials for Skin Covering . . . 98

Candidate Truss Geometries ...... . • • 100

Preceding page blank

3

Modeling of the Truss Geometry . . . . . . . . 114

Method of Analysis . . ........ . . . 119

Discussion of Results of Analysis . . . . . . 125

DETAIL DESIGN OF THE TRUSS FABRICATED HOUSING . . . 130

Design Criteria . .. .. .. . .. . . . . 130

Description of Truss Geometry . . . . . . .. 130

Analysis . . . . . . . . . . . . . . . . 139

COMPARISON OF THE FABRICATED TRUSS-LIKE HOUSINGWITH THE CH-54B CASTING . .. . . . . . . . . . . . 141

Cost Comparison ....... .... . . . . 141

Reliability and Maintainability Comparison . . 142

Corrosion Properties Comparison . . . . . . . 145

Crashworthiness . . . . . . . .0. .. . . . . 145

Heat Dissipation ............... 146

Survivability and Vulnerability .. . . . . . 146

Manufacturing . . . . . . . . 0 . 147

Summary . . .. . . . . . . . . . . . . 148

EFFECT OF SCALING A FABRICATED HOUSING DESIGN . . . 149

Parametric Scaling of Loads ......... 149

Analysis . . . . . . . . . . . . . . . . . . 158

DEVELOPMENTAL AREA DEFINITION . . . . . . . . . . . 162

Analysis . . . . . . . . . . . . . . . . . . 162

Experimental Verification .. .. ...... 163

Test Plan . . . . . . . . . . . . . . . . . . 163

Design Guidelines . *. .. ..* . • 168

AREAS FOR FUTURE STUDY . .............. 170

4

Page

SUIMARY * . . * * . * * * * * * * * * * * * * * 172

CONCLUSIONS .... ************** * 173

LITERATURE CITED . . . . . .. . 174

APPENDIXES I - CII-54B DIMENSIONS AND PERFORMANCEDATA . . . . . . . . . . . . . 175

II - MATERIAL CHARACTERISTICS . . . . 183

III - DISCUSSION OF SYNERGISTICINTERACTION BETWEEN CREEP ANDFATIGUE PROPERTIES . .. . 280

LIST OF SY14BOLS. . .. . . . . . 281

LIST OF ILLUSTRATIONS

Figure Page

1 11-5 Main Housing ........... 20

2 H-19 Main ousing .i ........... 20

3 11-34 Main Housing e........... 20

4 H-37 Main Housing i. ........... 20

5 Three-View of 11-3 Main Housing .. .. el

6 11-54 Main Housing, Composite Views . . . Z2

7 H-53 Main Housing, Side View . e . e .. 24

8 11-53 Main Housing, Front View . e . . . . 24

9 XH-59A Main Housing, Three-Quarter View . 24

10 Three-View, CII-53E Main Housing ..... 25

11 Flow Chart of Approach to FabricatedHelicopter Transmission HousingAnalysis . . . . ,e . . . . . . . . . . . 27

12 U. S. Army CH-54B Aircraft . .. . . . . 30

13 U. S. Army CH-54B Aircraft, Three-ViewDrawing . . . . . . . . . . . . . . . . . 31

14 Isometric of Drive Train Components of theCH-54B Aircraft .............. 33

15 Intermediate Gearbox, CH-54B Aircraft . . 34

16 Tail Gearbox, CH-54B .......... 35

17 Isometric of CH-54B Main TransmissionDrive Train Scheme ........... 37

18 Rear Cover and Accessory Location of theCH-54B Main Transmission . • • ..... 39

19 Cross Section of the CH-54B MainTransmission .............. 41

20 Isometric of CII-54B Main TransmissionHousing. ......... • • • • • . . 44

6

Figure Page

21 Cross Section of CH-54B Main Transmission,H~ousing . . . . .o . . .. 45

22 Forward Crash Loads on Housing Due toRotor Head and Accessory Inertia • • • • 51

23 Initial Loarting - 20g Forward CrashCondition . . . . . . . . . . . 54

24 Initial Loading - 18g Side Crash

Condition . . . . . . . .. .. ..... 55

25 Initial Loading - 20g Down CrashCondition . . . . . .. .. .. .. .. 56

26 Sign Convention for Applied Main RotorLoads . . . . . . . . . . . . . . . . . . 57

27 Bearing Reaction Geometry . .. . . . . . 60

28 Outer Shaft Load Geometry . .. . . . . • 63

29 Loading for Accessory Drive Train . . .. 66

30 Summary of Bearing and Gear Loads . . . . 71

31 Applied Housing Loads for FlightConditions (Hover) . . . . . . . . . . . 72

32 Applied Housing Loads for FlightConditions (Symmetrical Dive andPullout) . . . . . . . .. . . .. . ..* 9 * 73

33 Shell Model for Casting StiffnessAnalysis . . . .. .. . . . . . . .. . 77

34 Specific 107 Cycle Fatigue StrengthVersus Temperature ....... 80

35 Specific Fracture Toughness VersusSpecific Yield Strength . . .. . . . .. 83

36 Specific Stiffness Versus Specific YieldStrength . . . . . . . . 84

37 Specific 107 Fatigue Strength VersusSpecific Yield Strength . .. . . . . . . 85

38 Specific Ultimate Tensile Strength VersusTemperature .............. . 86

7

Figure Page

39 Specific Ultimate Shear StrengthVersus Temperature ........... 87

40 Various Joining Methods . . ...... 91

41 Friction Trapped Joint Used for CompositeMaterial .. * * * # * * 0 0 . . * * 92

42 Typical Welded Joint .......... 94

43 Typical Welded Joint .......... 95

44 Sketch of Typical Skin Truss Connection . 99

45 Truss Geometry Constraints . . . . . . . 101

46 Plan View of Truss Geometry . . . . . . 102

47 Half-Size Model of Typical TrussHousing . . . . . . . . . . . . . . . . . 103

48 Sketch of "A!' Frame Truss Design .... 104

49 Sketch of Pyramid Truss Design ..... 105

50 Sketch of Pure Truss Design . ...... 106

51 Sketch of Pure Truss Design . ...... 107

52 Sketch of "A" Frame Truss Design WithComposite Material ........... 108

53 Sketch of Pyramid Truss Design WithComposite Material ........... 109

54 Sketch of Pure Truss Design WithComposite Material ..... 9.9.. 110

55 Three-View Preliminary Layout of "A"Frame Truss Housing .......... . 1i

56 Three-View Preliminary Layout ofPyramid Truss Housing.. ....... 115

57 Three-View Preliminary Layout of PureTruss Housing .............. 117

58 Typical Truss-Like Structure in DisplayProgram ................. 121

8

FiurePage

59 NASTRAN Card Deck Setup...... 122

60 Truss-Like Structure with Some InitialError . . . . . . . . . . . . . . . . . . 124

61 Photograph of a .Designer Using CRT ToChange a Structure Element . . o . . .. 124

62 CRT Display of "A" Frame Truss Design . 126

63 CRT Display of Pyramid Truss Design . . . 127

64 CRT Display of Pure Truss Design .. . 128

65 Detail Design Layout of the FabricatedTransmission Housing (View Looking Aft) . 131

Detail Design Layout of the Fabricated66 Transmission Housing (View Looking

Forward) . .0 * 0 0 0 0 0 0 0 . .. . . 133

67 Detail Design Layout of the FabricatedTransmission Housing (Plan View) .. .. 135

Detail Design Layout of the Fabricated68 Transmission Housing (Auxiliary Plan68 View) . . . . . . . . . . . . . . . . . 137

69 CRT Display of the Final FabricatedHousing Transmission Design .. .. .. 138

70 Detail of Typical Casting MountingArrangement ........... .o . . 143

71 Detail of Typical Fabricated HousingMounting Arrangement . .. . . . . . 144

72 Historical Data - Hub Moment Versus DesignGross Weight .I..... ...... . 150

73 Historical Data - Hub Moment ConstantVersus Design Gross Weight . .. . . . . 151

74 Design Gross Weight Ratio Versus DesignTorque Ratio ... . 152

75 Head Moment Constant Ratio Versus DesignTorque Ratio .. e. e. . . . 153 4

9

Figure Page

76 Main Transmission Rotor Group Weight

Versus Design Torque Ratio ....... 154

77 Main Rotor Group Weight Ratio VersusDesign Torque Ratio . . . * . . . .. . 155

78 Flight Control Group Weight Ratio VersusDesign Torque Ratio . .. ..... *0 156

79 Head Moment Constant Ratio Versus DesignGross Weight Ratio .. .. . . . 157

80 CHI-54A/B Regenerative Test Stand - Viewof Test Stand. ........... . 164

81 CH-54A/B Regenerative Test Stand ControlRoom . * * * * * . . . . . .e ] 9 * * * a 164

82 CH-54A/B Regenerative Test Stand - Viewof Drive ................ 16&

83 Test Rig To Simulate Main Rotor ShaftLoads . . . . . . . . . . . . . . • • • • 165

84 Isometric of a Tail Gearbox Housing as aTruss-Like Structure . .. . . . . . . . 171

85 Forward-Flight Performance, Sea LevelStandard . . . . . . . . . • . .... 176

86 Hover Ceiling, Two-Engine CH-54B . . . . 3.80

87 Productivity Versus Mission Radius,CH-54B . . ... . . . .. 181.

88 Forward Climb Performance, CH-54B . . . . 182

89 Effect of Temperature on the Tensile andCompressive Moduli (E and Ec) of CastAZ9lC-T6 ............ . . 188

90 Effect of Temperature on tile PhysicalProperties of Cast AZ91C-T6 . . . . . o 188

91 Dimensional Changes of AZ9IC-T6 SandCastings at 2000 to 400OF ° ° • ° . • ° • 189

92 Effect of Cycles on KT, SM = O + 10KSI, KT = 3.3, AZ91C-T6 MagnesiumAlloy °190

10

Figure Page

93 Constant-Life Fatigue Diagram, AZ91C-T6Magnesium Alloy, KT = 1.0 . .. . . . . . 191

94 Crack Growth Rate Versus Stro:ess IntensityFactor, AZ91C-T6 With As-Cast Surfacein Laboratory Environment . .. . . . . . 192

95 Constant-Life Fatigue Diagram For Weldedand Heat-Treated 6061-T6 .. . . . . . . 194

96 Constant-Life Fatigue Diagram For 6061-T6Wrought Products at 200OF .. . . . . . . 195

97 Bending Modulus of Rupture For 6061-T6Round Tubing .............. 98

98 Constant-Life Fatigue Diagram For 6061-T6Wrought Products, K = 1.0 . .. . . . 198t

99 Constant-Life Fatigu, Diagram for 6061-T6Extrusions • •.. • •........ 199

100 Crack Growth Rate Versus Stress IntensityFactor, 6061-T6 Extrusions ....... 200

101 Constant-Life Fatigue Diagram 7175-T736Aluminum Alloy, KT = 1.0 ........ 204

102 Effect of Cycles on KTj R = 0.00, 7175-T736 Aluminum Alloy .... ...... 205

103 Effect of Temperature on the UltimateTensile Strength (Ft) of 7075-T651Aluminum Alloy ..... 206

104 Effect of Temperature on the Tensile YieldStrength (Fty) of 7075-T651 AluminumAlloy . . . . . . . . . . . . . . . . 207

105 Constant-Life Fatigue Diagram for 7075-T736

Wrought Products at 200OF ........ 208

106 Notch Sensitivity of Ti-6AI-4V Weld Metal(Kf Versus K t ) at Two Steady Stresses . . 214 4

107 Kt as a Function of Defect Geometry . . . 215 I

11

Figure Page

108 Constant-Life Diagram for Shot-PeenedTi-6A1-4V . . . . . . . . .. . . . . . . 217

109 Fatigue Crack Propagation Rate inTi-6AI-4V, BETA-STOA . . . . . . . . . . 218

110 Fracture Toughness of Beta-ProcessedTitanium Ti-6A1-4V . . .. . . . . .. . 218

il Constant-Life Diagram, Ti-6A1-4V AnnealedSheet, Room Temperature, Kt = 1.0 . . . & 219

112 Crack Growth Rate Versus Stress IntensityFactor, Mill Annealed Ti-6A1-4V TitaniumSheet . . . . . . . . . . . . . . . . . 220

113 Plane Strain Fracture Toughness ofTi-6A1-4V . . . . . . .. * . . . . .. 223

114 Constant-Life Diagram of TitaniumTi-6AI-4V BETA-STOA 0.5 Inch From As-Quenched Surface, Polished Specimens . . 224

115 Effect of Temperature on the UltimateTensile Strength of Annealed Ti-6Al-4VSheet and Bar . ....... . . . . . 225

116 Effect of Temperature on the TensileYield Strength of Annealed Ti-6A1-4V Sheetand Bar . . . . . . . . . . . . . . . . . 226

117 Effect of Temperature on the CompressiveYield Strength of Annealed Ti-6A1-4VSheet and Bar .............. 227

118 Effect of Temperature on the UltimateTensile Shear Strength of Annealed Ti-6A1-4VSheet and Bar ............. 227

119 Effect of Temperature on the UltimateBearing Strength and Bearing Yield Strengthof Annealed Ti-6A1-4V Sheet and Bar . . . 228

120 Effect of Temperature on the Tensile andCompressive Moduli of Annealed Ti-6A1-4VSheet and Bar ....... ...... 229

121 Effect of Temperature on the ThermalConductivity of Ti-6A1-4V (Annealed orSolution Heat-Treated and Aged) ..... 230

12

Figure Page

122 Effect of Temperature on the SpecificHeat of Ti-6AI-4V (Annealed or SolutionHeat-Treated and Aged) . .. . . . . . . 231

123 Effect of Temperature on the (Mean) LinearCoefficient of Thermal Expansion ofTi-6Al-4V . . . . . . . . . . . . . . . . 232

124 Creep Properties of Solution Heat-Treatedand Aged Ti-6AI-4V Sheet . .. . . . . . 234

125 Effect of Temperature on the PhysicalProperties of 17-4PH Stainless Steel . . 237

126 Effect of Temperature on the UltimateTensile Strength (Ftu) of 17-4PHStainless Steel (Bar and Forgings) . . . . 238

127 Effect of Temperature on the Tensile Yield

Strength (Fty) of 17-4PH StainlessSteel (Bar and Forgings) . . . . . . . . 238

128 Effect of Temperature on the CompressiveYield Strength (Fc ) of 17-4PH StainlessSteel (Bar and For ings) ........ 239

129 Effect of Temperature on the UltimateShear Strength (Fsu) of 17-4PH StainlessSteel (Bar and Forgings) . .. .. . .. 239

130 Effect of Temperature on the Bearing UltimateStrength (Fbru) of 17-4PH StainlessSteel (Bar and Forgings) . . . . . . . . 240

131 Effect of Temperature on the Eearing YieldStrength (Fbry) of 17-4PH StainlessSteel (Bar and Forgings) ........ 240

132 Effect of Temperature on the Tensile andCompressive Modulus (E and Ec) of 17-4PHStainless Steel (Bar and Forgings) . . . 241

133 Constant-Life Diagram, 17-4PH Bar andWire . . . . . . . . . . . . . . . . 242

134 Plane Strain Fracture Toughness of TenHeats of Annealed and Aged Plate as aFunction of Yield Strength ....... 244

13

Figure Page

135 Effect of Temperature on CompressiveYield St ength of 280-KSI Bar .. ..... 244

136 Effect of Test Temp .:ature on TensileProperties of Annealed and Aged Bar . . . 245

137 Effect of Test Temperature on ShearUltimate Strength of 250- an6 280-KSIBar . . . . . . . . . . . . . . . . . . . 245

138 Effect of Test Temperature on BearingStrength of 280-KSI Bar . . . . . . . .. 246

139 Effect of Test Temperature on the TensileProperties of Two Heats of 250-KSI CVMSheet . . . . . . . . . . .. . . . . 246

140 Constant-Life Fatigue Diagram for 18 Ni(250 Grade) Maraging Steel; RoomTemperature, Kt = 1.0, Smooth, Polished . 247

141 Kf Versus Cycles, Welded 250 GradeMe.raging Steel, R = 0. . . . . *..* 248

142 Effect of Temperature on the PhysicalProperties of AISI 4130 . . . . .. . . . 251

143 Effect of Temperature on the UltimateStrength (Ftu) of AISI Alloy Steels . . . 252

144 Effect of Temperature on the TensileYield Strength (Fty) of AISI AlloySteels . . . . . . . . . . . . . . . . . 253

145 Effect of Temperature on the CompressiveYield Strength (Fcy) of Heat-Treated AISIAlloy Steels . .......... . . . 254

146 Effect of Temperature on the Shear UltimateStrength (Fsu) of Heat-Treated AISI AlloySteel .. . . . . . . . 254

147 Effect of Temperature on UltimateBearing Strength (Fbru) of Heat-TreatedAISI Alloy Steels . .. 0 .. . • 255

148 Effect of Temperature on the Bearing YieldStrength (Fbry) of Heat-Treated AISI AlloySteels . . . . . . . . . . . . . . . . . 255

14

Figure Page

149 Effect of Temperature on Tensile and Compres-sive Modulus (E and Ec )of AISI Alloy Steels * 256

150 Typical Tensile Tangent-Modulus Curves atRoom Temperature for Heat-Treated AISI 4340Alloy Steel . . . . . . . .. . . . . . . . 257

151 Constnnt-Life Fatigue Diagram for 4130/4340Steel Weldments, Heat-Treated to 150-KSIUltimate Tensile Strength . . . . . . . . . . 258

152 Constant-Life Fatigue Diagram for 4340 SteelWeldments, Heat-Treated to 180-KSI. . . . . . 259

153 Constant-Life Fatigue Diagram for Carpenter

Custom 455 at Room Temperature and 200°F • • 262

154 Baron Epoxy Characterization - Goodman Diagrun 264

155 Goodman Strain Diagram at 107 Cycles for EIand S Glass Epoxy Unidirectional Laminates 268

156 Goodman Strain Diagram at 107 Cycles for Eand S Glass Epoxy + 450 Laminates . . . . . . 275

157 Goodman Strain Diagram at 107 Cycles for Eand S Glass Epoxy 50%-0*/50%-90* Laminates 276

15

15!

LIST OF TABLES

Table Page

I Gear Mesh Speed and Reduction Ratios -

Main Transmission . ............. 33

II CH-54B Acessories, Design Powers, Speedsand Unit Weights .............. 38

III CH-54B Component Weight and Center of GravityLocations . . . . . . . 0 . .. .. . . . .. 53

IV Summary Crash Load Design Criteria . . . . . 57

V CII-54B Planetary Gear Data . . . . . . . .. 70

VI Strength Criteria for Final Selection ofMaterials for Fabricated Housing Analysis . . 78

VII Material Cost Criteria for Final Selectionof Materials for Fabricated Housing Analysis. 79

VIII Material Trade-Off Selection Chart . . . . . 81

IX Thermal Conductivities of Material Candidates 88

X Static Joint Efficiencies of Weldments forMaterial Candidates ............. 97

XI Summary of Weight and Housing Stiffness forVarious Candidate Materials . . . . . . . . . 129

XII Cost Comparison, Conventional Casting VersusFabricated Housing Design . . . . . . . . . . 141

XIII Summary Table of a Cast Magnesium Housing withthe Fabricated Truss lousing Design . . . . . 148

XIV Changes in Applied Housinq Loads for a + 25Percent Change in Input Torque ....... 159

XV 10-Hour Dual-Engine Endurance Test Cycle • . 166

XVI R/H Input Accessory Pad Loading . . . . . . . 167

XVII Dual-Engine Overtorque Test Cycle . . . . . . 169

XVIII General Dimensions, CII-54B ......... 175

XIX Rotor Dimensions, CH-54B .......... 177

16

Table Page

XX Fuselage/Landing Gear Data, CH-54B . . . . . 178

XXI Mechanical Properties of AZ91C-T6 . . . . . 185

XXII Mechanical Properties of ZE41A-T5 . . . . . 187

XXIII Mechanical Properties of 6061-T6 . . . . . . 197

XXIV Mechanical Properties of 7075-T736 . . . . . 203

XXV Fracture Toughness Properties of 7175-T736Aluminum Alloy Forging . .......... 209

XXVI Typical Room Temperature Fracture ToughnessProperties for Several High-StrengthAluminum Alloys .............. 210

XXVII Mechanical Properties of Titanium Beta ForgingSTOA MIL-T-9097 Type III, Composition A • • 216

XXVIII Mechanical Properties of Titanium 6AI-4VExtrusions, BETA extruded MIL-T-81556, TypeIII, Composition A. ............ 221

XXIX Typical Elevated Temperature Deformationand Rupture Properties of Mill AnnealedTi-6AI-4V Sheet .............. 233

XXX Mechanical Properties of 17-4PH StainlessSteel H1025 . . . . . . . . . . . . . . . 236

XXXI Mechanical Properties of 4130/4340 Steel . . 250

XXXII Mechanical Properties of Carpenter Custom455 . . . . . . . . . . . 261

XXXIII Physical Properties of Boron Epoxy, Boron

Filament and Resin Matrices . . . .. . . . 265

XXXIV Elastic Constants of Boron Epoxy Composites 267

XXXV Strength and Proportional Limit of BoronEpoxy Composites . . .9........ 269

XXXVW Physical Properties of Glass Epoxy Composites,Glass Filament and Epoxy Resin Matrices • • 271

XXXVII Ultimate Strength, Proportional Limit andElastic Constants for Glass Epoxy . . . . . 273

1?(

Table Page

XXXVIII Elastic Constants for Graphite EpoxyComposites .. . ..... .. .. ..... 277

XXXIX Physical Properties of Graphite Epoxy GraphiteFilaments and Resin Matrices . . . . . . . . . 278

XXXX Strength and Elastic Constants for GraphiteEpoxy Composites ............... 279

I1

18

INTRODUCTION

Since 1946, helicopter transmission housings have been made ofcast magnesium alloys, generally AZ9lC and ZE41A. Such housingsare extremely light in weight and can be cast in complex shapesfor gear and bearing supports. Their disadvantage is that thealloys possess relatively low strength and are subject to prob-lems of porosity, segregation, and other defects as a result ofthe casting process. AZ91C magnesium casting alloy has anallowable fatigue strength of only 3000 psi, compared withhigh-strength steel, which has fatigue strengths approximatelyten times this value. In addition, magnesium alone lackscorrosion resistance. The corrosion resistance of magnesiumhousings is achieved through use of protective coatings, such asbaked resins and acrylic lacquers. Low fracture toughness andlow fatigue propagation resistance also make cast magnesiumhousings susceptible to ballistic damage. Another disadvantageof magnesium castings is that tiey require excessive wall thick-ness to compensate for core shift during pouring and to assu.eadequate fill in large sections. Weld repair of castings wasdeveloped to solve this lack-of-fill problem and is widely usedin the aircraft industry. The problems with present-daycastings have increased as cast housings have become increa- Jsingly large and complex, and there is an increasing need forstronger and lighter-weight designs.

One of the first uses of cast magnesium for a helicoptertransmission housing was on the H-5 aircraft. This casting wasused primarily for oil retention and for gear and bearingsupport structure. A tube-trus3 arrangement was designed, inconjunction with the casting, to transfer the large main rotorloads to the airframe. Figures 1 through 4 show the casthousing designs with tube truss structures for the H-5 in 1946,H-19 in 1949, 11-34 in 1953, and H-37 in 1954. In 1958, with thedevelopment of the H-3 transmission, knowledge of magnesiumalloy properties and casting technology had improved suffi-ciently so that a three-piece cast main housing could bedesigned to react main rotor loads in addition to internal gearand bearing loads. The H-3 housing, shown in Figure 5, isshaped as a box and is an assembly of three separate castings.In this arrangement, the main rotor loads are reacted in the

* center casting and then transferred through flexibly connectedend housings to the mounting feet.

Housing deflections associated with this design caused problemsin developing proper bevel gear meshing. In addition, low-stress/high-time creep problems occurred in service, requiringcorrective overhaul procedures. With the H-54 transmission,designed in 1961, the problems of lack of stiffness and creepwere eliminated by a design in which main rotor load path tothe airframe was minimized. Figure 6 shows the two-piececonical H-54 main housing, which consists of an upper cast

19

FReproduced from-best available copy.

Figure j. 1- ~i~~H~ig Pi gUre 2. 11-19 MIain Housing.

Figure 3. 11-34 M~ain Housing. Figure 4 . 11-37 Iain Housing.

20

Figure 5. Three-View of 11-3 M1ain Housing.

21

- I!

Figure 6. 11-54 Main lHousing, Composite Views.

22

magnesium housing and a lower forged magnesium support housingwith four mounting feet. At the time, the H-54 upper casthousing represented one of the world's largest magnesiumcastings. Casting technology continued to improve, however,and the H-53 main housing was designed in 1963 as a one-piececonical housing have cast mounting feet as shown in Figures 7and 8. The XR-59A main housing, shown in Figure 9, designed in1972, is very similar to the H-53 but less complex. The CH-53Ecast housing designed in 1973, is similar to the H-53, but ismore complex because of the third engine input. This housing,shown in Figure 10, makes extensive use of the weld repairprocedure to obtain a structurally adequate casting.

Due to the size and complexity of a cast housing, it isdifficult, if not impossible, to analyze it to the degreenecessary to minimize its weight. The result is often an over-designed and overweight housing.

Little improvement is seen in the future of magnesium castingtechnology, and changing design requirements will demandincreased size and complexity for transmission castings. Staticand fatigue strength of cast magnesium housings remains rela-tively low. At transmission overhaul, the rework frequency ishigh as a result of corrosion and creep. These problems arestill unresolved.

As a result, a new design is being sought that would eliminatecasting and material problems, enable accurate structuralanalysis, and provide a structure with improved stiffness andsupport characteristics. Improvements are also sought inweight, reliability, crashworthiness, corrosion resistance,inspectability, and maintainability.

One approach that is being considered is a skin-covered truss-like fabricated structure that performs the functions of a castmagnesium transmission housing. By exploiting the latestadvances in material technology, joining methods, and structuralanalysis techniques, such' a structure promises low weight, highreliability, improved crahworthiness, increased corrosionresistance, increased access for inspection, and elimination ofcreep.

% . .. . . . .

NRpoducedl from

t s avalable copy.

Figure 7. 11-53 Main llousing,

Side View.

Figure 9. XHI-59A Main Housing,

Figure 8. 11-53 Main Housing, Three-Quarter View.Front View.

24

le

II

Figure 10. Three-View, CII-53E Main Housing.

25

DESIGN APPROACH

Figure 11 is a flow chart of the approach taken in designingthe truss-like fabricated transmission housing.

The cast housing was examined to ascertain interface require-ments between the fabricated housing and its mating housingsand accessories. The fabricated truss-like housing will replacethe upper main casting and the lower support forging and meetall the functional and structural requirements of the CH-54Bh )using.

The design criteria were then established. These includedapplied loads for flight and crash conditions, and housingstiffness requirements. To determine crash loads for 18g side,20g forward, and 20g vertical given in Reference 13, weights weredetermined for all accessories and modules suspended from themain housing, and for all dynamic components. Items includedin the component weights were the main rotor head, rotor blades,swashplate assembly, main rotor controls, and internal gearing.Flight loads on the fabricated housing were determined for themost severe loading condition, i.e.,the symmetric dive andpullout maneuver. Stiffness and deflection criteria werederived from a simplified membrane analysis of the cast housing.

Truss geometries, materials and joining methods were theninvestigated. Design layouts of possible truss configurationswere produced for preliminary analysis. Concurrently, materialsand joining methods studies were conducted. The materialsinvestigated included steel, aluminum, titanium, magnesium andcomposites. The joining methods reviewed for possible use withthese materials included bonding, brazing, bolting, rivetingand welding.. This investigation resulted in the selection ofthree possible truss configurations and the deci ,ion to usewelded steel, welded titanium or brazed Borsical. Designplan views of the three configurations chosen are shown inFigure 11.

Through an analysis in a redundant structures program (NASTRAN),and a comparison with the cast housing design, the fabricatedstructures were modified to satisfy the design criteria. Forthe analysis, the structure geometry and the applied loads forvarious maneuver conditions were coded for NASTRAN. Thisredundant structures program was used in conjunction withSikorsky's Data Editor Program, which has a cathode-ray tubedisplay capability.

At the conclusion of these analyses, a final design config-uration was selected. A plan view of the final design config-uration is shown in the flow chart. The final design wassubjected to structural analysis and to assessment of reli-ability, maintainability, inspectability, cost, weight, crash-

26

1E

ESTABLISH DESIGN CRITERIAFLIGHT LOAD

CRASH LOADSTIFFNESS REQUIREMENT

LAY OUT VARIOUS CADESIGNS UTILIZING

TRADE-OFFLZIN

MATERIALS AN~D JOI

METHODS

CONDUCTJOINING METHOD

TRADE-OFF

Figure 11. Flow Chart of Approach to FabricatedHelicopter Transmission HousingAnalysis.

27

LAY OUT VARIOUS CANDIDATEDESIGNS UTILIZING VARIOUSMATERIALS AND JOININGMETHODS

COMBIh'E VARIOUSCONFIGURATIONS FOR 3FUNDAMENTALLY DIFFERENTDESIGNS WITH VARIOUSMATERIALS

ANALYZE REDUNDANT STRUCTUREUSING -NASTRAN

CON

"A" FRAME TRUSS PYRAMID

COMPARE FABRICATEDHOUSING WITHCASTING DESIGN

DUN'T STRUCTURETRAN

DETAIL DESIGN'CONFI GURATI ON O

SENSITIVITY STUDY± 25% IN LOAD

FINAL DESIGN IV

CONCLUSIO

worthiness, corrosion resistance, and heat dissipation charac-teristics.

To acquire trending data, a sensitivity study was conducted fora +25% change in input torque. Payload to gross weight fractionana rotor tip speed were held constant. Rotor size, head momentconstant, and group weight fractions were determined. The fab-ricated housings were then resized, with the newly developedflight and crash applied loads and the scaled internal gear andbearing loads.

During the course of the study, problem areas were uncoveredthat would require additional development before a fabricated

housing design could be incorporated in flight hardware. As aresult, development areas were defined, and a test plan waswritten to demonstrate functional and structural aspects of thedesign.

On the basis of the analysis, the characteristics of the final

design were assessed, and overall conclusions drawn.

Aircraft Selection

The U. S. Army CH-54B is the largest operational helicopterin the free world. It also has the best payload/gross weightratio. With expanding heavy lift helicopter uses, weight andcost might be reduced significantly through application of thefabricated truss-like housing to the CH-54B aircraft.

The CH-54B aircraft was selected for this study because of thelarge size of its components, the large number of aircraft inthe field, and the existence of extensive reliability andmaintainability data.

Appendix I presents the dimensional and performance character-

istics of the CH-54B aircraft.

Description of Aircraft

The U. S. Army CH-54B helicopter (Figure 12) chosen for thisstudy is a heavy-duty vertical lift aircraft. It has theability to pick up, transport, and place heavy external loadswith precision.

The CH-54B has twin engines rated at 4,800 horsepower each andtransmits 7,800 horsepowerto the main rotor. The six main rotorblades form a diameter of 72.2 feet. The CH-54B helicopter hasa lifting ability exceeding 12 tons and an overall length of 88feet 6 inches. Figure 13 is a three-view drawing of the CH-54Baircraft.

Preceding page blank29

Figure 12. U. S. Army CII-54B Aircraft.

Description of Drive Train

The drive train for the CH-54B helicopter (Figure 14) consistsof two JFTD12A-5A engines rated at 4,800 horsepower each, amain transmission, intermediate gearbox, a tail gearbox andinterconnecting shafting. Power is transmissted by each enginedirectly to a four-stage reduction main transmission, whichreduces the engine speed of 9,000 rpm to 185 rpm at the mainrotor, an overall reduction ratio of 48.54 to 1. Power is alsotransmitted from the tail/accessory drive section of the maintransmission through the tail drive shaft at 3016 rpm to theintermediate gearbox. The intermediate gearbox in turn drivesthrough the tail gearbox to the tail rotor, which turns at 850rpm.

The intermediate gearbox (Figure 15) consists of an input bevelgear, idler bevel gear, and output bevel gear. The designincorporates a self-contained lubrication system. The sump-located oil pump is driven from the back of the idler bevelgear.

The tail gearbox assembly (Figure 16) consists of a right-anglebevel mesh, with the output bevel attached through a series ofbolts to the tail rotor shaft. The tail rotor servo, which isfixed to the back of the cast magnesium gearbox housing, isattached through thrust bearings to the pitch control shaft.The pitch control shaft is concentric and supported in abushing by the tail rotor output shaft. The pitch controlshaft, which is capable of linear motion within the tail rotor

30

Figure 13. U. S. Army CII-54B3 Aircraft, Three-ViewDrawing.

31 (

0 8SCALE (f'eet)

t, Three-View

0 4~ 1 .2 16

SC ALE (feet)

A

Figure 14. Isometric of Drive Train Components of theCH-54B Aircraft.

output shaft, supports the pitch beam. Tail rotor blade pitchis controlled through a series of pitch links. The lubricationsystem of the tail rotor gearbox is of simple splash design.

Description of Main Gearbox

An isometric of the CH-54B main transmission drive train schemeis presented in Figure 17. Table I contains a list of reductionratios and speeds for each gear mesh.

TABLE I. GEAR MESH SPEED AND REDUCTION RATIOS -MAIN TRANSMISSION

Input Output

Gear Mesh Reduction (rpm) (rpm)'

Input Bevel Set 27/53 9000 4585.9

Main Bevel Set 25/76 4585.9 1508.2

First-Stage Planetary 78/55/188 1508.2 442.3

Second-Stage Planetary 166/32/230 442.3 185.4

Tail-Takeoff Bevel Set 76/38 1508.2 3016.4I


(c

0 2SCALE (in.

Figure 15. Intermediate Gearbox, Cli-54B Aircraft.

34

AL~ ROO AI OO

SEV IC HF

PojM.A.

Figure 16. Tail Gearbox, CU1-54B.

35

rPITCH LINK

TA"IJi ROTORPITCH SHAFT

4B

Main Rotor

Tail-TakeoffCBevel Set

Main Bevel Set

Input Bevel Set

V Firs - Stage Planetai7

Second-6tagePlanetary

Figure 17. Isometric of CII-5414 ain Transmission4Drive Train Scheme.

Preceding page blank 37

The acqessories driven by the main transmission are located onthe rig,.t-hand input bevel housing and on the main transmissionrear cover. Figure 18 shows the location of the various rearcover mounted accessories. Table II lists the CH-54B access-ories, design powers, speeds, and accessory weights.

TABLE II. CH-54B ACCESSORIES, DESIGN POWERS, SPEEDS, ANDUNIT WEIGHTS

Design Speed Weight

Accessory Name Horsepower (rpm) (ib)

#1 Generator 49.2 8025.7 64

#2 Generator 49.2 8025.7 64

Tachometer Generator 1.0 4200. .2

Utility Pump 53.0 4321 22.16

Winch Pump 100.0 3797. 20.

Second-Stage Servo Pump 24.0 4585. 14.1

First-Stage Servo Pump 19.0 4321. 14.7

Tail-Takeoff Drive 1220.0 3016

Figure 19 is a cross section of the CH-54B main transmissionassembly. Power is transmitted from the engines through adrive shaft, with a torquemeter mounted concentric to the shaft,to a first-stage spiral bevel gear mesh. This mesh has a 27-to-53 ratio. It is designed to transmit 4,800 horsepowercontinuously at an engine input speed of 9,000 rpm. Thespiral bevel gears used in the first stage have a 3.313-dia-metral pitch, a 2.875-inch face width, a 20-degree pressureangle, and a 25-degree spiral angle. The first-stage spiralbevel gear mesh is identical on right and left inputs to themain housing. This permits both inputs and housings to beinterchangeable.

Located between the first-stage bevel mesh and second-stagebevel mesh on both the right and left sides are ramp rolleroverrunning clutches. These clutches enable disengagement ofthe engine power turbine and first-stage spiral bevel mesh fromthe rest of the drive train during autorotation and under

38

Oil Cooler

#1 Generator

r f irst-Stage

enerator

Figure 18. Rear Cover and Accessory Location of the4CH-54B3 Main Transmission.

39

single-engine operation. The clutch is arranged with the outerhousing driven by the first-stage bevel gear and the camdriving through to the second-stage bevel gear.

The cam output of the ramp roller overrunning clutch drivesthe second-stage bevel pinion of the second-stage bevel mesh.This 25-to-76 ratio bevel mesh has an output bevel gearcommon to both inputs. This common bevel gear combines powerfrom both the right and left engines. The spiral bevel gearsused in this second-stage bevel mesh have a 3.125-diametralpitch, a 4.0-inch face width, a 20-degree pressure angle, anda 25-degree spiral angle. Lubrication and cooling oil is fedcentrifugally from an oil distribution tube inside the second-stage pinion to the free-wheel unit and the support bearings ofthe pinion.

Mating with the 76-tooth common bevel gear is a 38-tooth bevelpinion that provides the drive for the accessories and tailrotor. Included in the accessory packages are an oil cooler,APU, generator, hydraulic pumps, and tachometer generator. Aramp-roller overrunning clutch mounted on the tail-takeoff bevelgear shaft permits the APU drive to be operated without drivingboth the main and tail rotors. This feature permits groundoperation of the accessories through the auxiliary power unitwith the main engines off.

The common bevel gear of the second-stage bevel set, which isconcentric to the main rotor shaft, drives the input sun gearof a two-stage 8.14-to-i reduction ratio simple planetary gearset. A simple planetary is defined as a driving sun gear,planetary carrier output and a fixed ring gear. The first-stage planetary set has a 3.41 reduction ratio and containsseven pinions. The first-stage planetary pinion is mounted ona two-row roller bearing whose inner race is supported by twoplanetary plates. The ring gear is rigidly bolted to the maintransmission lower housing. This ring gear contains bothfirst- and second-stage planetary internal gear members.

The carrier of the first-stage planetary is connected throughdowel pins to the sun gear of the second-stage planetary. Thesecond-stage planetary contains eighteen pinions, has a reduc-tion ratio of 2.385-to-i, and is similar in construction tothe first-stage planetary. The lower plate of the second-stageplanetary is splined to and drives the main rotor shaft.

40

INPUT SECTION SWAGUIDE T

ROTATED 900 FORGUDCLARITY,

SCALE (in.)/

Figure 19. Cr055 Section of the C11-541 ain Transmission.

41

R

-INPUT MODULE TAIL-TAKEOFFACCESSORY COVER

I TAIL-TAKEOFFBEVEL PINIQN

SECOND-STAGE

ORT

4 8

Cr055 Section of the C11-54B3 Main Transmnission,

41

Descriptior of Main Transmission Casting

The main transmission ho -ing consists of an upper casthousing of AZ9lC magnesioi., a lower housing forged from ZK60A;and lower cover, accessory cover, left and right inputhousings, and sump, all cast in AZ91C. Figure 20 is anisometric of the CH-54B main transmission housing, showing uppertransmission housing. All seven housings, in addition to pro-viding reaction and support for main transmission components,provide a containment for transmission lubricant. They serve asa support for servos, oil cooler, scissor bracket, rotor brake,and accessories. They also contain interface pilot diametersand attachments, and serve as a conduit for distributing oil tovarious components.

Three servo support brackets are mounted on the main trans-mission casting at increments of 450, 2250, and 3150. Theservo support brackets react induced loads from the swashplateon the servos. A swashplate guide is mounted to the top of theupper housing and serves to locate the swashplate and reactany induced side loads. The scissor bracket connects the uppermain housing and the stationary swashplate. This bracketpermits the swashplate assembly to translate up and down as wellas be tipped out of a horizontal plane, but it prevents rotationof the stationary swashplate assembly.

Mounted on the main transmission casting are brackets thatpermit connection of a series of structural tubes used tosupport the main transmission oil cooler and blower assembly.

The main transmission housing reacts first-stage input bevelassembly loads, second-stage bevel gear and pinion loads, first-and second-stage planetary reaction loads, and main rotor shaftloads. The tail-takeoff and accessory drive bevel pinion,driven by the second-stage input bevel gear, is also supportedin the main transmission housing.

The main transmission housing is attached to the main trans-mission support housing through a series of studs. The supporthousing has four mounting feet with three bolts in each mountingfoot. All net rotor and transmission loads are reacted by themounting feet and distributed to the airframe.

The main rotor shaft is supported by a roller bearing at the topof the upper main housing and by a set of tandem-mounted splitinner race thrust bearings in the bottom of the lower cover.The lower support housing, which transfers loads to the air-frame, is located between the lower cover and the upper mainhousing. The lower cover transfers the side and thrust loadsto the lower support housing. The upper housing transfers sideloads from the upper roller bearing to the lower support housing.


IN,

Figure 20. Isometric of CII-54B3 Nain Transmission Hiousing' 4

44

0'0

C 0

Figure 21. Cross Section of C11-54B3 Main Transmission Housing.

45

ansmission Housing.

Attached to the lower cover is the oil sump casting, which hascored lubrication lines and a pad for an oil pump drive. Theaccessory cover is attached to the rear cover flanges of themain transmission assembly. All of the accessories, with theexception of the hoist and utility pumps, are attached to, andsupported by, the rear cover.

Attached to the upper main housing are the left and right sideinput housings. The input housings support and contain thefirst-stage bevel gears as well as the aft engine support. Arotor brake disc is attached to the left side inputbevel gearshaft, and a brake caliper assembly is mounted to the left-handinput housing. The brake reaction torque is carried throughthe left-hand input housing to the main housing. Drives forthe hydraulic hoist pump and the utility system hydraulic pumphave been provided on the right side input housing.

Design Criteria for Main Transmission

The CH-54B main transmission design is typical of modernhelicopter transmission practice.

Castings and housings of the main transmission are designed forcrash conditions of 20g vertical, 20g forward, and 18g lateral.The worst loading condition for the structural design of themain transmission is the 20g forward crash.

Gears and shafting are designed for unlimited fatigue life atmaximum torque conditions. For example, since the CII-54B cancontinue flight with one engine inoperative, the first- andsecond-stage bevel meshes are designed for maximum single-enginepower of 4,800 horsepower, whereas the first- and second-stageplanetaries are designed for 6,800 horsepower.

Bearings are designed for a minimum Bl0 life of 3,000 hours ata prorated load. The prorated loads are determined from theexpected operating spectrum and represent a weighted averageload. The shafting is also designed for maximum slope under thebearing of .0007 in./in. at maximum operating torque.

Steel liners are provided at the interface of bearing outerraces and the magnesium castings. These liners serve both asstress buffer and wear member. The bearing liners are designedto have a positive interference fit with the magnesium housingat 250 0F, while maintaining a low enough interference fit at-650F to keep the magnesium from yielding.


47

All threaded fasteners on the main transmission are providedwith locking features to prevent loosening of nuts. Nuts andbolts, cotter pins, clevis pins, or Shurlocks are used toprovide positive retention of larger nuts. On smaller threads,using standard nuts, a deformed thread type locking device isused.

The lubrication system for the CH-54B main transmission istypical of current modern aircraft pressure oil systems. oilfrom the gearbox sump is pumped through an oil filter to an air/oil heat exchanger. The blower fan of the heat exchanger isdriven from the accessory case of the main transmission. Theoil is then forced from the heat exchanger into the main trans-mission manifold. Oil from this manifold is distributedthrough a series of jets to bearings, gear meshes, and oildistribution tubes within the transmission.

Design Criteria for the Fabricated Housing

The truss-like fabricated housing, in order to be an acceptablealternative to the conventional magnesium casting, must meetthe CH-54B functional requirements and in addition possess thefollowing attributes:

The fabricated housing must be at least as strong asa magnesium casting.

Stiffness of the fabricated housing should be similarto that of a casting. As a minimum, the stiffnessshould be sufficient for proper operation of thedynamic components.

0 The fabricated housing should weigh less than aconventional casting.

Maintainability and reliability should be at leastequal to those of a casting design.

Inspectability of a fabricated housing should besuperior to that of a cast housing.

The fabricated housing should be easier and simpler tomanufacture than a magnesium casting.

Total cost of a fabricated housing should be lowerthan that of a casting.

Overall heat transfer should be greater than that ofa magnesium cast housing.

0 The fabricated housing should exhibit characteristics

of improved survivability and reduced vulnerability.

48

In assessing these desired attributes, a quantitative approachwas used where possible. In certain areas, such as thermalcharacteristics, a true assessment will be possible only afterexperimental data are developed.

Interface Requirements for Fabricated Housing

The fabricated housing must be interchangeable with the presentCH-54B casting. A cross section of the CH-54B transmission isshown in Figure 19. The items and modules that must be inter-changeable are labeled. Included in the interchangeabilitylist are input section assembly (first-stage bevel mesh andhousing), tail-takeoff/accessory cover, lower cover assembly,swashplate guide, ring gear support, second-stage bevel gearhousing, tail-takeoff bevel pinion housing, and the scissorbracket. The upper housing and lower support housing for whichthe truss structure is modeled have been joined together, asshown in the isometric of Figure 20. This permits a lightertruss structure and eliminates the need for a mounting flangeand row of studs. The servo bracket fittings have also beeneliminated, since it is possible to have a ligher truss struc-ture through use of a more direct load path. All oil transfertubes between the housings and all fittings, manifolds, andjets have been kept in the same location. In brief, thefabricated housing is directly interchangeable with the upperhousing casting and the lower housing support considered as asingle unit.

All internal transmission gears, bearings, and seals areidentical with those used in the current CH-54B transmission.Bearing fits and backup are adjusted for thermal expansion, sobearing clearances are the same under operating load as in thecurrent CH-54B casting. Seal fits are maintained as in thecurrent CH-54B transmission at normal operating temperature.The fabricated housing is designed to maintain the location ofthe jets and oil flows that are used in the present CH-54Btransmission design. Bearing supports, in particular bevelpinion bearing supports, are located in the plane of the presentcasting walls. Truss members are so oriented that loads canbe distributed easily. The transmission mounting arrangement isthe same as for the present CH-54B, with four mounting pads andthree mounting bolts in each pad.

49

PRELIMINARY DESIGN

In the preliminary design phase, the objective was to select amaterial, fabrication methods, and structural arrangement suitedfor a detail design analysis. The approach was to conduct aNASTRAN analysis of various combinations of materials, joiningmethods, and structural geometry under crash loading conditionsand to use the results of those analyses as a basis for deter-mination of the most promising housing configurations.

Load Development

Table III is a list of component weights and center of gravitylocations taken from Reference 1, CH-54B Weight Control StatusReport. This data was used in developing crash loads for theanalysis. The loading conditions were 20g forward, 18g side,and 20g down. Figure 22 shows the load orientation for forwardcrash conditions due to rotor head and accessory inertia. Theapplied loads for these conditions are shown in Figures 23, 24,and 25, respectively.

The applied loads on the upper and lower main rotor bearingsupport rings result from the masses of the main rotor head,rotor blade, control rods, scissor, and swashplate assembly.These loads are determined by multiplying the masses by a loadfactor and then calculating the reactions at the main rotorshaft bearings.

Loads transmitted to the main housing through the servos aredue only to the weight of the servos themselves for bothforward and side crash conditions. For the vertical crashcondition, additional loads due to the rotating and stationaryswashplates are also transmitted to the main housing. Loadson both the right and left side input housing consist of themass effect of the aft end of each engine plus the weight ofthe first-stage bevel input housing assembly. The loads on therear cover flange consist of the mass effect of the oil coolerand supports, rear cover assembly, first-stage servo pump,utility pump, generator, and tachometer generator. The totalapplied load on the rear cover was distributed on the rear coverflange by a series of point loads. The internal gear drivingloads were not added to the crash load conditions, since theyare small in magnitude. The applied crash loads are summarizedin Table IV. Figure 26 is a sketch of the sign convention usedwith the applied main rotor loads.

To develop the applied transmission loads for flight conditions,the critical maneuver condition was used. The aircraft loadsreport (Reference 2) and the structural design criteria report(Reference 3) indicate that the critical loading conditionoccurs for a symmetric dive and pullout maneuver.

50

30.344 Tn.

Servo Inertia Loads --

Inertia Housing Inertia

Load Ta

Figure 22. Forward Crash Loads onflHousing Due to

Riotor Read and Aiccessory Inertia.

-mm* - Rotor Applied Load

30.3k In.0.4 Upper Bearing Raaction- Ring

rtia Loads Distributed Rear Cover Inertia

62.426 In.- -" " _ _i

H Inputusing Inertiaad

Main Rotor Thrust Reaction Ring

using Due to

Inertia.

Rotor Applied Load

Upper Bearing Raation- Ring

Distributed Rear Cover Inertia

Main Rotor Thrust Reaction Ring

fn) H- 00 000000 0 o~orCC 0;0004J1 V4 VJHN H

4IHE-4

E-40H *i OI- Ir- iM D ON0 VNOO00000 NO01A

H 0 . 6 * 4 * 0 0 * 0 * * 0 0 0 * 0 0 0

W) I ULfl 10 ) w V MN -4M Hr -)C'H HCNH H r-4 Nr4I-) CNCN CN N NC N 'N NN

0

W* 094 00 00 0- H IV N 0 OO O 0 0 0 0D 0

U A a* g 0 0 & 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

V W9 % -4C'4w r- r- - n e 44 44 ; r: a%* am- fn- H m m4 H Hn Hn fn m NHH% omm n%mm m mMmm H

0q

4. C~ HH

00U

H4 4J jai E)~

H :4 0 U)

U) 13 P4m) >-P U0-4.CJ 0 m H

H J(d 0 1 r 0 00 H .4

H P JU00r U 4 VH~) fI 0 4W~ ) V 0 - 4)*0Id( 4.l.- Cn &4d ~ ltI Im4JC1)) n :3 0 0~t

0) r40 U) toH 04 m 0U)U) H 9 )V I Mdr H00r4W) CO H rd00)04 0 *)rH 90)4

V rz U r4c E 0 4) t)E-I WPrecedin pag blan 531 4 ?. ,A

00 0 04 d 0 f- a) (d S 4 (1 N U E W

S18 loadsk 200 Lb

ETEEach

69o Lbb

-81,100 L

X X I

Mounting Feet Constrained in All DirectionsExcept for Twist About of Each Bolt

Figure 23. Initial Loading - 20g Forward Crash Condition.

54

, =- -- ll ill -li ... ." i .. ..

18 loads

200 Lb Each

5 2 0 0 Lb 11 0

'~165,300 Lb,

690.Lb---

Lb5200 L

-81,100 Lbervo Loadc

Mounting Feet Constrained in All DirectionsExcept for Twist About of Each Bolt

Figure 24. Initial Loading - 18g Side Crash Condition.

55

i1

18 loads

200 LbEach

5200 LbS Sationary Scissor

i 1730: i

1730 L

1730 Lb

84,950 Lb

Figure 25. Initial Loading -209 Down Crash Condition.

56

I , , , ° , i II "-

TABLE IV. SUMMARY CRASH LOAD DESIGN CRITERIA

ConditionLocation Forward 20G Side 18G Down 20G

Upper Roller Bearing 165,300 165,300 0

Servo Pads 690/pad 690/pad 1730/pad

Input Section 5,200 5,200 5,200

Rear Cover 3,600 3,600 3,600

Lower Housing Support -81,100 -81,100 -84,950

+Z

V VERTICAL SHEAR

Mz

- +X AFT

11x D

+YLMY

S/

+ Y LEFT

Figure 26. Sign Convention for Applied Main Rotor Loads.

57

For the critical flight conditions, the loads at the aerodynamic

center of gravity of the main rotor are:

VR - 94,263

SR - 5,220

DR = 14,685

M = -15,426

MyR = 1,322,271

MzR = -70,106

These loads are resolved into reactions at the main rotor shaftupper roller bearing and the lower stack bearing set. Figure 22shows the geometry used in determining the reactions, which arelisted below:

RAx - 12,641

RAy = 9,676

RBx = 27,326

RBy - -9,676

The thrust associated with this condition is 94,263 pounds.

Internal gear and bearing loads were developed for each primarymesh of the CH-54B drive train.

The first-stage input housing is supported by the fabricatedhousing structure, which must react a moment due to the weightof the cantilevered input section. The loads due to this momentare much smaller than those due to the second-stage bevel meshand were ignored.

The second-stage input pinion is driven by the first-stage inputbevel gear and in turn drives the second-stage bevel gear. Theinput pinion is a right-hand spiral bevel gear rotating counter-clockwise as viewed from the shaft end. Gear data for thesecond-stage input pinion are as follows:

Pitch angle Y = 160 51

Pressure angle c = 200

Spiral angle 1 = 250

Pinion diameter dp = 8.0

58

I

Face width F = 4.0

Pinion rpm n = 4585

The axial component of the gear tooth load is given bytan V sinY + sin W cosY

Wx - Wt cosW

Wx W t (.560)

The radial component of the gear tooth load is given by

Wr tan0D cosY - sinW sinY (2)t cosW

Wr Wt (.257)

The bevel gear mean pitch diameter is given by

dm = dp - F sin Y (3)

dm = 6.892 inch

For the maximum fatigue design power condition of 4800 horse-power,

HP 63025 (4)T = rpm

T - 65,980 in.-lb

The tangential tooth load is given by

t 2T (5)

Wt = 19,150 lb

59

Substituting the max.mum tangential tooth load in Equations 1 and

2, the axial and radzal components of the gear tooth loads are

wx - -'560) (Wt)

Wx 10,724 lb

Wr = (.257) (Wt)

Wr = 4,920 lb

The bearing reactions are calculated from the geometry shown inFigure 27.

i0,724 - 19,150 lb 3,4146in.

3.92in

7.899 in.

Figure 27. Bearing Reaction Geometry.

At the tapered roller bearings, or end A (Figure 27), thevertical and horizontal reactions are given by

RA (Wr) (3.92) + (Wx ) (3.446)V -7.899

60

RAv = 7,120 lb

RAT = (Wt) (3.92)7.899

RAT = 9,500 lb

Thrust = 10,724 lb

At the roller bearing, or end B (Figure 27), the vertical andhorizontal reactions are given by

R (Wx ) (3.446) - (Wr) (3.979)= 7.899

RBV = 2,200 lb

(Wt ) (3.979)RBt = 7.899

RBt = 9,650 lb

The second-stage bevel gear is driven by the second-stage bevelpinion. In turn, the second-stage bevel pinion transmits powerto the first-stage planetary sun gear and to the tail-takeoffbevel pinion. The second-stage input gear is a left-hand spiralbevel gear rotating clockwise as viewed from the shaft end. Thegear data for the second-stage input spiral bevel gear are givenbelow:

Pitch angle Y = 570 21'

Pressure angle 4) = 200

Spiral angle W = 25*

Gear diameter dp = 24.320

Tail-takeoff face width F = 1.375

Pinion and gear face width F = 4.0

61

EL---.

Gear rpm n = 1508

The axial component of the gear tooth load is given by

tan(O sinY- sinticosY (6)Wxin = Wtin cosqj

Wx.= .0866 Wt.in in


tan(DcosY + sinWsinY (7)Wrin = Wtin Cos

Wr in = Wtin (.609)

Using the geometry of Figure 28 and summing the moments about

each reaction point yields the following results:

MB = 0 in butt line plane

R.355 - .0205 Wt (8)1 35Wout in

IMA = 0 in butt line plane

RBb I = .531 Wtout - .00152 Wt in (9)

IMB = 0 in station plane

RAsta = .930 Wtout - .0296 Wt in (10)

1MA = 0 in station plane

R = .0700 W + .00152 (11)RBsta Wtout Win

The critical design condition is 5400 horsepower to both inputs,with 1220 horsepower taken off at the tail-takeoff bevel gear andthe remainder delivered to the main rotor. The input and outputtangential tooth loads are

62

Forward

RB .9

0 5 tin R sta Wx t

Fiurt2. Ouehf od Gemer

W I 63

W = 10,560 lbin

Wtot 4,960 lbout

Substituting the tangential tooth loads in Equations 8, 9, 10,and 11 leads to the reactions at A, the tapered roller bearingloads, and B, the roller bearing loads (Figure 28):

R = 1544 lbA bl

RA - 4300lbsta

RB = 2617 lb

RB sta= 331 lb

The axial load on the thrust bearings is found from

RA = (.0866) (Wt. ) (2) + (.0866) (Wt )x in t

RA = 2258 lbx

The tail-takeoff bevel pinion is a right-hand spiral bevelpinion driven counterclockwise by the main bevel ge'ar. Thegear data for the tail-takeoff bevel pinion are as follows:

Pitch angle Y = 240 54"0

Pressure angle ) = 200

Spiral angle o = 220 52'

Face width F = 1.375

Tail-takeoff speed n = 3016

The axial component of the gear tooth load is given by Equation(6).

Wt can sinY - sin t cos YCos t

64

I

WX = Wt (-.2162)

The radial component of the gear tooth load is given byEquation (7).

tan4 cosY + sinti sinYWr Wt Cosqi

Wr = Wt (.5359)

The bevel pinion mean pitch diameter is found from Equation (3).

dm = dp - F sinY

dm = 10.297

The tangential component of the gear tooth load is given byEquation (12).

W 126050 HP (12)Wt dm n

Wt - 4952 lb

The accessory drive loads acting on the tail-takeoff gear shaftare sh<vn in Figure 29. The loads are based on the maximumaccessory power requirements shown in Table II.

For the utility pump, the tangential component of the geartooth load is computed from Equation (12).

Wt 126050 HP

Wt = 237 lb

The radial comnonent of the gear tooth load is given by

Wr = Wt tan'I= 237 lb

The horizontal and vertical gear load components are foundthrough use of the geometry shown in Figure 29.

65

Tail-Takceoff (Freewheel)

2 Utility WinchPump 53 HP

NN 235Tail-TakceoffSpur

90 ,V/Generator 3

0

260 30'

Figure 29. Loading for Accessory Drive Train.

W W =-98.5 lb

W = -2371lb

For the first-stage servo, the tangential component of the geartooth load is computed from Equation (12).

W = 85.37 lb

66


Wr = Wt tan 22-1/2*

Wr = 35.3 lb

The horizontal and vertical components of the tangential andradial gear loads are calculated with the aid of Figure 29.

WH = -Wr sin 300 + Wt cos 300

WH = -91.5

Wv = Wr cos 300 - Wt sin 300

Wv = 12.1 lb

For the generator, the tangential component of the gear toothload is given by Equation 12, in which

Wt = 442 lb


Wr = Wt tan 22-1/20

Wr = 183 lb

The radial and tangential gear tooth loads are resolved into

vertical and horizontal components with the aid of Figure 29.

WH = Wr Cos 90 15' - Wt sin 90 15'

wH = 110 lb

Wv Wr sin 90 15' + Wt cos 90 15'

WV 466 lb

For the tachometer generator, the tangential component of thegear tooth i - is given by Equation 12, in which

= 5 Ib

Wr Wt tan 22-1/20

Wr = 21b

67

The radial and tangential gear tooth loads are resolved intovertical and horizontal components with the aid of Figure 29.

WH = Wr cos 260 30' - Wt sin 260 30'

WH = 1 lb

Wv = Wr sin 260 30' + Wt cos 260 30'

Wv = 5 lb

The resulting loads on the tail-takeoff gear that drives theutility winch pump, servo pump, and generator are given by

W = - 98.5 - 91.5 + 110

WH - 80 b

Wv - 237 + 12.1 + 466

W = + 241 lb

The resulting loads on the aft tail-takeoff spur gear thatdrives the tachometer generator are

H 2

Wv2 = 5 lb

The bearing reactions at the tapered roller bearing, A, and theroller bearing, B, are given by

RA = (.5359) (Wt ) (10.91)

v 9.22

+ (.2162) (Wt) (5.15 - I.v2 (1.94) -Wv(2.37)

9.22

RA = .7548 Wt - 63v

Wt(10.92) + WH 2 (1.9 4 ) - WH(2.37 )

RAH = 9.22

68

RAH = 1.1832 Wt + 20.77

RB = (.5359)(Wt)(1.69v 9.22

+ (.2162) (Wt) (5.5) + Wv2 (7.28) + Wv(6.85)

9.22

RBv = .2271 Wt + 182.9

RBH = Wt (1.69) - WH2 (7.26) + WH (6.85)

RBH. = .1832 Wt - 60.19

For the maximum tail-takeoff bevel pinion design power conditionof 1220 horsepower, the bearing reactions are found to be

RAV (.7548) (4952) - 63 = 3674 lb

RA = (1.1832) (4952) + 20.77 = 5880 lb

RBv = (.2271) (4952) + 182.9 = 1307 lb

R = (.1832) (4952) - 60.19 = 847 lb

The two-stage planetary unit, which is driven by the outputbevel gear, drives the main rotor shaft. The two-stageplanetary unit consists of two simple planetary units, withsun gear input, planetary plate output, and ring gears. Thering gears are attached to the upper main housing through aseries of studs. This housing counteracts the reaction torque.The planetary gear data are given in Table V.

The loads on the upper main casting from the planetary unitsresult from the planetary weight and reaction ring gear torque.The torque on the first-stage ring gear is given by

Wt DNT = 2

TR = 753,217 in.-lb

The reaction torque on the second-stage ring gear is given by

T - Wt DN2

69

TABLE V. CH-54B PLANETARY GEAR DATA

Diame- Number Numbertral Pitch of Pressure of

Reduction Pitch Diameter Teeth Angle PinionsRatio Pd D t 0 n

First Stage 3.41025 - - 7

Sun - 7 11.142 78 22.5 -

Planet - 7 7.8571 55 22.5

Ring - 7 26.857 188 22.5 -

Second Stage 2.38554 - - - 18

Sun - 8 20.750 166 22.5 -

Planet - 8 4.000 32 22.5

Ring - 8 28.750 230 22.5

T = 1,342,912

Figure 30 is a summary of the internal bearing and gear loadsthat must be reacted by the truss fabricated housing. Figures31 and 32 show the applied loads on the fabricated housing.

The loadings are identical on the CH-54B servo bracket and thefabricated housing servo support bracket. For the fabricatedhousing design, the servo brackets have been eliminated, and atruss arrangement is designed to keep the servo attachment lugin the same location. The load on the attachment lug consistsof a steady load plus vibratory load of 2,500 + 6,000 pounds,from Reference 2.

The loading on the stationary scissor bracket, located on thetop of the main gearbox housing, is a result of reacting thedrag torque and pitching loads of the stationary swashplate.The load that will be used to design the stationary scissorconsists of a steady load plus vibratory load of -250 + 1,100pounds, from Reference 2.

70

I

A B

t.0

w A

F igr 30. Summary of Beain an Gea oadsou

n7

w*

z13 Loads 12 Lb

x y

12~1 Looda6 L894 Lb 550 Lb -96T76 Lb

12 Loads 188 9500 Lb 12 Loads[100 Lb

Figure 31. Applied Housing Loads for Flight

Conditions (H1over).

72

z

13 Loads 12 Lb

X Y

161L12 Loads 4275 Lh 16

264 3 '674 Lb1

\2200 -, 5880 Lb 12 Loads- ,. 1070 Lb

12 Loads 188 Lb

. 331 Lb.' %12 Loads

, 12 Loads;

7855 Lb 27326 Lb 66L

• 2 Load 7123 Lb

Figure 32. Applied Housing Loads for FlightConditions (Symmetrical Dive andPullout).

73

Stiffness Criteria

With the results calculated for the cast housing, only lateraland vertical spring rates can be computed at the top of theconical shell.

The stiffness requirements for the fabricated housing weredetermined from a simplified analysis of the conventionalcasting. The conventional casting was modeled as a thin-wallconical shell with boundary conditions shown in Figure 33.With the use of Reference 4, the vertical and lateral springrates were calculated as follows

wctno Calw + V c til~ (3

ctn2 [0 Cvu [ cp + 6u V ctna (14)

E 7r r 3.3 6.6

where:

P is the horizontal applied load

V is the vertical applied load

r = 7.5 inch - upper cone radius

r 2 19.9 inch - lower cone radius

a = 230141 - half cone pitch angle

h = .63 inch - shell wall thickness

= .35 - poisson' s ratio

r _ 7.5 - .3678

a = ctn C h[12(1 -ui2)] 1/2

Zl 0

74

The coefficients for use in Equations (13) and (14) aredetermined from the curves in Reference 4.

C, = -10.1

CV = -15.0

w

C = 17.0

1Cv = 0u

Substituting the coefficients:

w ctn2 a [-10.1 p + -15.0 (2.3294) VE rI L 3.3 6.6

u ctn [ 17.0 p]u = E r I l 3.3

Since there is no vertical load applied to the casting V isequal to zero.

w ctn2a [ -. 3E rl 3.3 P

U = ctna[ 17.0 p1

E rl 3

The verticF'l deflection is

w - .1084 x 106 P

The vertical spring rate due to a horizontal applied load is

KVertical = 9.22 x 106 lb/in.

The lateral deflection is

u = .08019 x 10-6 p

75

The lateral spring rate due to a hoizontal applied load is

KLateral = 12.47 x 106 lb/in.

To compare the fabricated truss-like housing with a convent-ional housing, the methods must be comparable. With the resultscalculated in the study it was possible to determine the springrate of any member of the truss structure. For purposes ofcomparison, the spring rate of the truss structure (both verticaland horizontal due to a horizontal load) at the center of thntop of the truncated cone was used.

76

wi V

-7.5

p

Figure 33. Shell Model for Casting Stiffness Analysis.

77

Selection of Materials and Joining Methods

Of the sixteen materials considered for the fabricated housingdesign, only three appeared to be usable in the nearguture:Ti-6AI-4V, Custom 455 stainless steel, a Borsical Whencompared with the titanium and BorsicalM , the stainless ste .1alloy, Custom 455, has the highest allowable critical bucklingstrength, highest fracture toughness, and best combination ofspecific ultimate shear strength, specific ultimate tensilestrength, and specific fatigue strength with temperature.

Welding was chosen as the joining method for steel and titanium.This choice was based on thc fact that welding has the higheststrength-to-weight ratio for steel and titanium of all thejoining methods considered. razing-bonding was chosen as thejoining method for Borsical. Once again the decision wasbased on the strength-to-weight ratio as compared to otherjoining methods.

As discussed previously, the crash condition, 20g forward, 139side, and 20g down, is the critical design condition. As aresult, specific ultimate strength was the primary ranking item.Table VI is a list of the final material candidates and theirultimate and specific ultimate tensile strengths.

TABLE VI. STRENGTH CRITERIA FOR FINAL SELECTION OFMATERIALS FOR FABRICATED HOUSING ANALYSIS

SpecificUltimate UltimateTensile TensileStrength Strength*

Material (psi) ((psi) /('.b/in. 3 ))

Borsical® 95,000 1$-00,000Titanium, Ti-6AI-4V 140,000 875,000Stainless Steel, Custom 455 220,000 785,000

* Specific ultimate strength is equal to ultimate strength,divided by density.

The final material candidates are ranked by ?ost in Table VII.

78

.2I

TABLE VII. NATERIAL COST CRITERIA FOR FINAL SELECTION OFMATERIALS FOR FABRICATED HOUSING ANALYSIS

Approximate c't

Material Rank Dollars/lb

Stainless Steel, Customer 455 1 1.40Titanium i-6A1-4V 2 9.00Borsical Q2 3 20.00

Candidate Materials

The sixteen materials considered were composites and alloys ofmagnesium, aluminum, titanium, and steel. Important propertiesof each material are summarized in Table VIII. The table showsabsolute and specific properties, at room temperature, 200 0 F,and 600*F for tensile strength, shear strength, fatiguestrength, and fracture toughness. A compilation of the mater-ials data collected during the study is included as Appendix II.

Additional data on the characteristics of the materials isincluded in Figures 34 through 39. Figure 34 shows therelationship between fatigue strength and temperature. Figure35 is a graph of the specific static and fatigue properties ofthe materials under consideration. Figure 36 contains data onthe specific tensile moduli and yield strength, while Figure37 is a plot of specific fracture toughness versus specificyiele strength. Figure 38 shows the relationship betweenultimate shear strength and temperature. Figure 39 is a plot ofthe ultimate tensile strengths of the materials.

Whereas magnesium and aluminum alloys have relatively lowstrength, fatigue, and creep properties, they have excellentthermal characteristics. Table IX lists thermal conductivitiesof the material candidates. In the worst case, magnesium hasabout twelve times the thermal conductivity of steel alloys.

The thermal characteristics of a fabricated housing can beassessed realistically only through testing. The material(steel) finally selected for the fabricated housing was chosenindependently of its thermal characteristics, since only 12 to15 percent of the heat generated in the drive train is trans-ferred through the housing. The convective heat transferooefficient, the. most significant factor in determining overallgearbox housing heat rejection rate, can be determined onlythrough testing, and no analytical estimates were made,

79

3 0Alloy System Le ed Line

Aluminum-----------

Titanium -

Magnesium250 s- -- -- Steel -

SBorsical

S20 0-- -osi

x-1 Ti-6Al-4v (a)

0w Ti-8mo- -M- 3AJ.

V

O Ti-6Al-4v (oc -4)150 Custom 4i55

S100. 15T3

250 Grade Maraging

AC91-T

00 10 20 0 6 50 6.

HPRTRE(F

*NT:60 aiu Srnt si.%e i ubro ae

Fiue3.SeiiH0 CceFtgeSrnt

Vess epraue

*8

DENSITY ULTIMATE TENSILE STRENGTW (ksi)

MATERIAL/CONDITION (lb/in. 3 )

R.T. (Ab. R.T) 200OF (Ab 2000 F) 600°F Ab 600°F

AZ91C-T6 (CLASS 3) 0.0652 27 (414) 23.5 (360) 6.8 (104)

ZE41A-T5 (OVER 50 LB) 0.066 26 (395) 24 (364) 10 (150)

6061-T6(bar) 0.098 42 (429) 38 (388) (YAhr) (86)8.8

7175-T736 0.101 76 (750) 10,000 hr (720) (Y2hr) (415)73 42

Ti-6AI-4V(oc -(3) 0.160 130 (812) 118 j/37) 97.5 (610)

Ti-6AI.4V(P) 0.160 140 (875) 127 (795) 105 (650)

Ti-8Mo-8V-2Fe-3AI 0.175 180 (1030) 160 (915) 145 (828)

17-4PH(1025) 0.282 155 (550) 149 (529) 130 (460)

250 GRADE MARAGING 0.289 295 (848) 244 (844) 240 (830)

4130/4340 0.283 180 (636) 175 (618) 160 (565)

CUSTOM 455 0.280 220 (785) 214 (765) 187 (668)

"S" GLASS/EPOXY 195 (2900) 195 (2900) -175 (2600)"E" GLASS/EPOXY 0.067 160 (2390) 160 (2390) ,145 (2170)

HIGH STRENGTH GRAPHITE 0.057 134 (2350) 134 (2350) ,120 (,- 21004

(E=20 x 106)

HIGH MODULUS GRAPHITE 0.053 100 (1880) 100 (1880) ,,90 (,o,1690)

(E=26 x 106 )

BORSIC ALUMINUM 0.098 95 (930) 95 (930) 95 (930)

81

TABLE YTMATERIAL TRADE-OFF SELECTION CHART

STRENGTH (ksi) ULTIMATE SHEAR STRENGTH (ksi) (Reduced op____ ___ ___ __ ___ ______ ___ __ ___ _ _ ___ __ ___ ___FATIGUE STRE

2000 F) 600OF (Ab 6000 F) R.T. (Ab R.T.) 200OF (Ab 2000 F) 600OF (Ab 6000 F) R.T. (Ab R.T.) 2

160) 6.8 (104) 15.0 (230) 16 (245) 4.5 (68) 4.8 (73)(CLASS 3)

364) 10 (150) 20.0 (300) 18.5 (280) 7.7 (115) 6.2 (94)

188) (YAhr) (86) 27 (275) 24.5 (250) 5.6 (57) 6.5 (66)8.8

720) (2hr) (415) 42 (415) 40 (395) 23 (228) 10 (99)42

037) 97.5 (610) 76 (475) 71 (444) 57 (356) 25 (156)I (Plate)

F95) 10i (650) 86 (537) 80 (500) 64 (400) 42 (262)

15) 145 (828) 105 (600) 97 (555) 78 (445) 28 (160)

129) 130 (460) 100 (354) 95 (337) 71 (252) 37 (131)

44) 240 (830) 150 (518) 145 (502) 120 (415) 20 (69)

f

18) 160 (565) 108 (382) 105 (371) 99 (350) 35 (124)

65) 187 (6e0 135 (482) 130 (465) 113 (403) 39 (139)

900) - 175 (2600) 5.4 (80) 5.4 (80) - 2.7 (,'- 40) 18 (270)390) - 145 (2170) 5.2 (77) 5.2 (77) -2.6 (,- 38) 12 (180)

350) "-120 (-N2100) 16 (280) 16 ,280) ~ 8 (,-140) 30 (525)

380) -'90 (-'1690) 1 (150) 8 (150) ,v 4 ('' 750i 22 (415)

30) 95 (930) 17.5 (180) 17.5 (180) ,4 (,40) 17.5(T) (178)

-LE TEKMATERIAL TRADE-OFF SELECTION CHART

(Reduced approx. for Scatter & Size Effect)ENGTH (ksi) FATIGUE STRENGTH 107 CYCLES(2Oi F CRITICAL BUCKLIN

00°F)j 600°F (Ab 6000 F) R.T. lAb R.T.) 200OF Ab 20,0F 600°F (Ab 6000 F) R.T. (A b R.T.) 201

5) 4.5 (68) 4.8 (73) 4.2 (64) ,_ 1.2 (18) 16ksi @ (245)R/t=17

0) 7.7 (115) 6.2 (94) 5.3 (80) 2.4 (36) 22ksi @ (333) 1R/t=17

0) 5.6 (57) 6.5 (66) 6.0 (61) unkown (tan. mod)25 (255) 2

5) 23 (228) 10 (99) 7.5 (74) 2.5 (84) (ton. mod) (495)50

57 (356) 25 (156) 22 (138) 15.5 (97) 107 (670)(Plate)

0) 64 (400) 42 (262) 37 (231) 26 (163) 117 (730)

78 (445) 28 (160) 27 (154) ,v 145 (, 825) '1

71 (252) 37 (131) 35 (124) 30 (106) 126 (446)

120 (415) 20 (69) 19 (66) 16 (55) 200 (692)

I

1) 99 (350) 35 (124) 34 (120) 31 (109) 135 (477)

113 (403) 39 (139) 38 (135) 33 (118) ,v 156 j-, 557)- 2.7 (,, 40) 18 (270) 18 (270) -16 (,, 240) v 60 (, 895),-2.6 (,u 38) 12 (180) 12 (180) '10 (,v 150) ,v 70 (,v 1040)

-, 8 ('-140) 30 (525) 30 (525) - 25 (,v 440) 182 (3190)

,v 4 ("-750) 22 (415) 22 (415) $,1 18 (,, 340) 100 (1880)

v 4 (,.V 40) 17.5(T) (178) 17.5(T) (178) ov 4(1) ,v 40 200 (3750)

FRACTURE TOUGHNESSCRITICAL BUCKLING 'TRENGTH (Fcc OR SECANT) (ksi (in ) Y _ _

R.T. (Ab R.T.) 200OF lAb 2000 F) 600°F (Ab 60OF) R.T. 200OF 600OF16ksi @ 1245) 13 (199) 3.2 (40) 11.7 ,50R/t=17

22ks* @ (3331 19.5 (295) 9.8 (148) 14.1 -,35 -55R/t::17

(fan. md) (255) 23.8 1243) 3.9 (40) very high higher highest25

(tan. mod) (495) 46 (455) 4.4 (43) 23.8 34 >> 3430

107 (670) 100 (625) 75 (469) 40 44 >, 44

117 (730) ,110 (687) 82 (512) 74 81 81

S145 (, 825) N-139 ,.,795 ,W100 (,-570) 52 60 >> 60

126 (446) 114 (405) 77 (273) -55 > 60 > 60

200 (692) ',' 190 (657) 175 (605) 67 90 90

135 (477) 120 (425) 97 (343) 110 124 >>124

, 156 (, .5571 ,-50 (,535) ,'120 (,429) 90 -,100 100

N 60 ('89!) - 60 (-895) %, 45 (.o 670)70 (iv 1040) - 70 r, 10401 .% 55 ( 815) N/A

182 (3190) 182 (3190) ,, 145 (25401 N/A

100 (1880) 100 (1880) , 80 ( 1500) N/A

200 (3750) 203 (3750) N160 (,,3000) N/A20.5)(

0

0I I H

I-'m

'-(4

0 0 4

Co)

- q.4 -H

U u

00

HM

'0

C>0 0C) .p 0 0

Lr.\ *dr I. ) .2 C

Precdingpageblan

83o

0

4)

P4

U)

C,

W 440

- - 0

140 MH

r-i NCd V2

U] N

ct2.r -tIba L'

co p

-,d d

E-1 *,-i

00UU

L C)0

~~~~ 0 f 0 L 0 LH*r CJ H H

P84

0_____ ___ ____ ____ 0

.0

cdy

0_________

*0 0C~j H

U) 4

H 0 >

Hr 0 CICI

I b o

0jE1 l V44 4

0 0~ 00

0 04 04

C')UH

85q

100

900 -Ti-8Mo-8V-2Fe-3A1

800-"250 Grade Maraging

700,T-A-~(~

Ho -Ti-6Al-4V(o( -8)

~ 600 4130/13140

500-....

4oo-

3 00

100- 1 6061-T6

010 100 200 300 4600 500 6;0

TEMPERATURE (OF)

Legend

Alloy System Line

Aluminum--------

Titanium -

Magnesium

NI OTE: Strength at 600OF is After 1/2 Hour Exposure Steel - -

Borslcal

Figure 38. Specific Ultimate Tensile Strength VersusTemperature.

86

600- - - Legend

Alloy System Line

Aluminum-----------

Titanium - -

500 - - Magnesium

Steel

-Borsical9

40-0 Ti-8?40-8V-2Fe-3A1

1 0 IN250 Grade Mara~ing

~1) Custom 4~55

H300-

U) 17-11PH

200. 71T5-T736

ZEl41A-T5

AZglc-T6

C)100H 606l-T6

Borsice.,

0 100 200 36o 4OG 500 600

T&W4ERATrORE (OF)

N OTE: Strength at 600 0F is After 1/2 Hour Exposure

Figure 39. 11?ecific Ultimate Shear StrengthVersus Temperature.

87

The magnesium alloys were eliminated from consideration on thebasis of low yield, ultimate and fatigue strength at room andelevated temperatures. An additional problem is that thinwall castings are difficult to achieve. Casting technologyhas remained static, so magnesium casting alloys were con-sidered to hold no future promise.

Aluminum alloys 6061-T6 and 7175-T736 in wrought form wereeliminated from consideration due to poor properties at hightemperatures. In the 600*F temperature range (emergencyoperaging condition), all aluminum alloys are significantlyabove their aging temperature. Interaction of fatigue andcreep is not defined in this region, so the material isconsidered to be an unnecessarily high risk.

Titanium alloys, Ti-6AI-4V in either the alpha or alpha-betaform and Ti-84o-8V-2Fe-3A1, show promise for application in atruss fabricated housing design and were retained in the study.

TABLE IX. THERMAL CONDUCTIVITIES OF MATERIAL CANDIDATES

Thermal ConductivityMaterial (Btu/ft/hr ft2 OF)

AZ91C-T6 32.4ZE41A-T5 61.8

6061-T6 967175-T736 70

Ti-6AI-4V 4.7Ti-8Mo-8V-2Fe-3AI 4.8

17-4P11 8.5250 Grade Maraging 14.54130/4340 21.7Custom 455 10.4

"S" Glass .12"E" Glass .12

11S Graphite .8IIM Graphite .8

Borsical 56

The data on the material selection chart shown in Table VIIIand the data in Appendix II are displayed graphically inFigures 34 through 39.

88

Steel alloys 17-4PH, 250 Grade Maraging, 4xxx series, and Custom455 were considered. 17-4PH was eliminated from consideration,since an improved version with higher strength is available inCustom 455. Production experience with 17-4PH is directlyapplicable to Custom 455. The 250 Grade Maraging steel hasexcellent properties, but it was eliminated because of produc-tion and processing problems. One problem with 250 GradeMaraging processing is that a stratified composition (banding)often results which is associated with austenitic segregationand alloy (chemical) segregation. Property variations indifferent lots and this stratified composition problem haveoccurred in Air Force production. As a result, 250 GradeMaraging steel was not considered for fabricated housings.The 4xxx series alloys, in particular 4130 and 4340, have thenecessary specific properties, but require heat treatment toobtain these properties. The 4xxx series alloys were eliminatedfrom consideration, since the quenc'. rig process during heattreatment will result in some warpage even with a quench fixture.

Glass, graphite, boron and PRD-49 composites were consideredin the study. The high-strength, highly oriented filaments inthese materials make them possible choices for use in fabricatedhousings. However, when joining methods are investigated, itis found that effective shear transfer methods for these mater-ials are undeveloped. Thus, these materials were dropped fromconsideration.

From an investigation of all the figures and materialcharacteristics, the most promising materials for design of afabricated truss-like hou ng are Ti-6Al-4V, Custom 455 stain-less steel, and Borsical (9.

The material study did not include an evaluation of fatiguecrack propagation. Although damage tolerance analyses,including ballLstic lateral damage, fracture toughness, andfatigue crack propagation exist and are in routine use, amore conservative approach was applied in this sutdy. Forthe study, initiation of a crack in any member or joint wasconsidered to result in immediate loss of load-carryingcapability of the entire member. This approach simplified theanalytical work in the NASTRAN analysis and added conservatismto the study approach.

89

Rationale for Joining Method Selection

Six methods were considered for joining members of the fabri-cated housing: mechanical fastening, riveting, bolting,welding, bonding, and brazing. Figure 40 shows the differenttypes of joints applied to a fabricated housing design.

For a mechanical, friction, riveted or bolted joint, a fittingor overlap area is required between the two joined materials.This results in increased weight and a greater cost per unitvolume of joined material than for any other fastening method.

Figure 41 is a typical friction trapped joint that would be usedin a composite material design. Increased axial load in thefibers results in axial load in the fitting plus radial tensionor compression. The mechanical joint is an unnecessarily heavyway of making a connection, so it is unsatisfactory for afabricated housing design.

Riveting was eliminated from consideration for the followingreasons:

i. For the large loads in the structural members, manyrivets are required to transfer the load. The resultis a heavy, bulky joint. The typical fabricatedhousing has more than one hundred joints.

2. The clamping force in the rivet, which permits loadtransfer through friction, cannot be controlledadequately.

3. Expansion of the rivet in the hole is not alwaysuniform and can cause stress concentrations andnonuniform load distribution in the rivets.

4. Vibration in the structure can cause loss of clampingforce in the rivet, which would require excessiveinspection and re-riveting.

Riveting is one possible method of attaching the skin coveringfor retention of oil in the truss housing structure. Becausethe loads are not large, this is an ideal application forriveting.

Parts that aze riveted can also be bolted. If a joint isdesigned for rivets, a bolt of the same size can be substitutedfor each rivet. The resulting joint is stronger by approxi-mately fifty percent. The best loading design occurs when thefasteners are loaded in shear. When it is impossible to avoidtension loading, as would occur in at least some of the maintruss joints, the bolt should be used over a rivet. Animportant advantage of bolts over rivets is the ability to

90

wb

43-

4r.

*rl

.rd

0

.4-I

0

4-) 00 N'

PCI -4

*rJ.14

43rS

91

II

Figure 41. Friction Trapped Joint Used for Composite

Material.

92

control the clamping force accurately. The clamping, orpreload, force tends to distribute the load over a larger areaand give the bolted joint greater fatigue strength than theriveted joint. The bolted joint has the added advantage ofeasy assembly and disassembly with standard field equipment.

With each joint, bolted or riveted, some form of fitting isrequired. The design of the fitting is heavy, due to the manyunknowns in the analysis of fittings. Typical practice in thedesign of aircraft is to require a fitting factor of +.25 asthe minimum margin of safety instead of a 0.0 margin of safetybased on yield or ultimate criteria. The analysis of joints isquite complicated, due to the combined stresses in the joint andfastener, tolerance buildup, and the values for allowablechafing stress and stress concentration factors for fatigueapplications.

The relatively high weight of the bolted or riveted joint andthe relatively high cost of fabrication make the bolted orriveted connection unsatisfactory for the fabricated housingdesign.

Bonded joints for a fabricated housing design would requirelarge, heavy lap joints to have a high joint fatigue strength.The bonding agent serves a multiple purpose, since it also actsas insulator to prevent galvanic corrosion between dissimilarmaterials.

To use bonding properly in a fabricated housing design, it isnecessary to protect the adhesive from- the internal gearboxenvironment. The joint would be exposed to approximately 240 0Fsynthetic diester base oil. This oil has a severe degradingeffect on most bonding agents. At present, only two possibleadhesives, American Cyanamid HT 424 and Hysol EA 912, appearable to survive in the internal gearbox environment. These twoadhesives may allow the placing of structural joints internal tothe oil-holding skin covering. Ilowever, these adhesives requireverification, so bonding will be considered only for joining theskin covering external to the load-carrying truss structure.

Brazing exhibits relatively low joint strength and is thus notconsidered for attaching structural truss members. 1},ever,brazing is the only feasible i of joining Borsical , sinceit is not weldable. Borsical , appears to have littlepotential in a fabricated housing application due to the highrisk of brazing.

The most practical way of fabricating a truss-like housingappears to be weldment construction; welding offers a simplejoint design and high static joint efficiency. Figures 42 and43 are typical welded details considered acceptable for afabricated housing design.

93

Figure 42. Typical Welded Joint.

94

Figure 43. Typical Welded Joint.

95

The titanium alloys considered in this study may be joined byelectron-beam welding, plasma-arc welding, inertia bonding, ordiffusion bonding.

Electron-beam welding is generally considered more reliable(and more expensive) than the other methods as long as thejoint can be designed to allow machining off the weld protru-sion which occurs on the side of the weld away from theincoming beam. This protrusion is the region where weld voidsoccur,

Plasma-arc welding is less expensive than electron-beam weldingbut requires close quality control to limit hydrogen gas voidsin the weldment due to reduction of water vapor by the moltentitanium. The state of welding art has been advanced tothe point where the hydrogen void is only a minimal problemin fatigue. Stress concentration factors from hydrogen-gasvoids are negligible for static (crash loads) analysis, and thestatic joint efficiency is very close to unity.

Inertia bonding (also known as friction welding) requires asymmetrical structure since no control exists of the angularposition of one portion of the strucLure relative to the other.Present structural designs are asymmetrical, effectivelyeliminating this process from consideration.

Diffusion bonding, either low-pressure long-time or forge-diffusion bonding, may be effectively applied to housingfabrication. Properties have been degraded by long times athigh temperatures (press-diffusion bonding), and costs ofpress-diffusion bonding do not appear to hold much promise forreductions in the future. Forge diffusion bonding may beapplied to attachments where maximum strength is required,applying some other fastening method to a lower-stressed area.Results of Contract Number DAAG46-73-C-0126 (USAAMMRC) indicatethat static and fatigue properties across the forge-diffusionbond line are the same as the parent metal properties. Datafrom that contract will be available for follow-on design in ashort time.

Static joint efficiencies of weldments for various materialcandidates are summarized in Table X. While magnesium andaluminum alloys exhibit high static joint efficiencies in theweld, they rank low in the product of ultimate tensile strengthand static joint efficiencies. The only materials beingconsidered for a welded fabricated housing design, based onmaterial study, are Custom 455 stainless steel and Ti-6A1-4V.

96

TABLE X. STATIC JOINT EFFICIENCIES OF WELDMENTS FORMATERIAL CANDIDATES.

Product of Specific UltimateStatic Joint Tensile Strength and Static

Material Efficiency Joint Efficiency (psi x 103)

AZ9lC-T6 .95 394ZE41A-T5 .95 375

6061-T6 .95 4087175-T736 N/A Not Weldable

Ti-GAl-4V .95 790Ti-8Mo-8V-2Fe-3AI 1.00 1030

17-4PH .95 522250 Maraging .90 7624340 .90 572Custom 455 .95 742

Evaluation of joint fatigue strength is extremely difficult inlight of the number of variables involved. A qualitativeapproach requires that design requirments for fatigue strengthbe met through quality control.

It remains to be seen whether a rigorous fatigue analysis ofwelded joints is required for transmission housings. The highstatic crash load conditions normally size the structure, andthe fatigue loads remain relatively low under lg forward flight.However, some qualitative discussion of welding quality controland fatigue strength interactions appears required to outlinethe state of the art, the trends of the state of the art, andthe method of analysis which can be applied to fatigue analysisof welded joints.

At this time, hydrogen voids in titanium weldments are beingfound in the thicknesses which would be used in a built-uptransmission. These voids are of the order of 0.006 inch indiameter (maximum). With improved techniques, it is expectedto find voids whose diameters are as small as 0.004 inch with80 percent probability. Voids as small as 0.002 inch in dia-meter have been the origins of fatigue failures in laboratoryspecimens. Theoretical and experimental analyses agree thatthe controlling variable in determining the stress concentra-tion of a void in a weldment is not the diameter of the voidbut the ratio of the void diameter to the distance of the void'scenter from the free (stressed) surface. With the residualstresses from shotpeening, the effective depth of any size void

97

may be limited to the depth of the residual compressive stressesfrom peening. This limits the effective stress concentrationfactor of the worst possible void (described above) to a theo-retical stress concentration factor of 2.6, the same as a filledrivet hole.

The effects of voids in titanium weldments are negligible instatic conditions. Sound welds in titanium approach unity forstatic joint efficiency. Joint efficiencies in fatigue may befurther improved by applying a static load slightly in excess of50 percent of the yield load of the structure. This would yieldout the local areas around any voids and induce residual com-pressive stresses on springback of the surrounding structure.To apply this technique, the fatigue stress vector must be known,must never change direction, and must always be low enough sothat cyclic-stress induced fading is not a problem. Althoughpromising, this concept would require further research beforeit could be applied in practice.

Custom 455 is not prone to formation of voids in welding. Itis possible, should the heat-input to the weld go out of control,to overheat the material and boil off copper from the weld.However, this is not a normal problem with plasma-arc welds anddoes not require serious consideration. Welds in Custom 455have been shown to have nearly unity joint efficiency. Furtherimprovements in welding techniques will probably improve thewelded strength of Custom 455 more than titanium (i.e., heatinput control, cleanliness, radiographic *nspection).

Candidate Materials for Skin Covering

A skin covering material is required to contain transmission oilwithin the truss members of the fabricated housing. A number ofmaterials were considered, including stainless steel sheet,titanium sheet, and a composite. Various joining methods werestudied, including mechanical fastening, brazing, welding, andbonding.

The combinations of skin-truss investigated include Carpenter455 stainless steel truss with a stainless steel sheet covering,Ti-6AI-4V truss with a titanium sheet covering, and a stainlesssteel or titanium truss with a composite covering. Figure 44 isa sketch of the detail of typical skin-truss member joints. Theconnections include brazing or welding the sheet to the tube(a), a bonded joint to tube structure (b), and a mechanicaltrapped fastening system (c).

Sealing provisions require that the assembled joint be oil tightfor at least 5,000 hours at 200 0F. The skin is nonstructural.Thickness is stainless and titanium is .020 inch. The relativelylarge panel size results in oil canning if any appreciable loadis applied. The joining method of attaching skin to truss is,

98

(a)

(b) 4

Figure 44. Sketch of Typical Skin Truss Connection.

99

therefore, expected to be lightly loaded. At 600OF forapproximately one-half hour, skin-panel truss connection neednot be oil tight, but the skin panels must not fall out of placeand cause meshing problems with the gears. The skin panel trussconnection is exposed to hot, synthetic, diester base oil withthermal cycling for long periods.

Access panels are provided to permit inspection of the inputbevel gear mesh, tail-takeoff gear mesh, and planetary gearunit. The access panel is provided with quick-disconnectfasteners used for joining the access panel to the skin panel.A simple gasket or "0" ring is used for sealing the oil in thetruss structure.

For joining the skin covering to the truss, mechanical fasteningwith screws and a simple gasket are practical, but the heaviestsolution. A welded or brazed joint in fastening the skin tothe truss is a simpler solution, but requires equal compati-bility of truss member and skin material to welding or brazing.The lightest solution is bonding fiberglass/polyimide to trussmembers. Two products that would be acceptable are AmericanCyanamid HT424 adhesive film and Hysol EA 929 adhesive film.

There remains the problem of verification of the performance ofthe adhesive or, at worst, definition of the shear strengthdegradation that would allow accurate calculation of a reliablebonded skin on the built-up transmission.

The approach taken for the detail design configuration is theone with the least amount of risk: a metallic skin coveringwith gasket screwed to the truss structure. Even though thisis slightly heavier than other covering methods, the skincovering weight is only 6 percent of the total weight.

Candidate Truss Geometries

The truss-like housing shapes considered reflect the presentcasting design. This is a result of interface and inter-changeability requirements with the CH-54B upper and lowerhousing casting and forging. Figure 45 is an isometric of thegeometry.

Since there are infinite possible truss geometries that couldhave been chosen for a fabricated housing design, the problemis how to minimize the possibilities. Analysis of the variousloading conditions shows that the largest loads act on theuppermost ring. All the net housing loads must be distributedto the lower mounting ring to transfer load to the airframe.Included are the main rotor load, net bevel gear reactions,tail takeoff gear and accessory weights, ring gear reactiontorque, and many minor loads. Three geometry arrangements wereconsidered: an "A" frame design, a pyramid design, and a pure

100

Figure 45. Truss Geometry Constraints.

101

truss arrangement. A plan view of each design is shown inFigure 46.

CONE I cow. 2 CONE 3

Figure 46. Plan View of Truss Geometry.

A half-size model was constructed as illustrated in Figure 47.The model has all the required interface, support, and mountingrings. With the model it was easier to determine load and pathand to identify such requirements as servo pad and bevel gearroller bearing supports.

With these data a number of isometric sketches were made toshow the geometry in more detail. Figures 48 through 54 aresketches that correspond to the "A" frame design, pyramid design,and pure truss arrangement of Figure 46.

Three-view preliminary layouts were rade of the candidate trussarrangements.

The "A" frame truss arrangement, Figure 55, consists of a seriesof rings and eight main load-carrying members arranged ortho-gonally to each other in pairs in the plan view. They form an"A" shape when viewed from the front or side. This arrangementtransfers main rotor shaft horizontal reaction loads at theuppermost ring, through the mounting ring, to the airframe. Thehorizontal rings which the "A" frame members support providestability and transfer loads from side to side. The base ringis tipped five degrees in a side plane relative to the mainrotor shaft and is constructed as a circle instead of thepresent casting ellipse. A vertical ring welded to the "A"frame member and stabilized through supports provides thereaction for the tail-takeoff bevel pinion tapered rollerbearing set. Parallel to the tail-takeoff bevel support ringis the rear cover support flange. This flange, which is

102

iP P

Figure 47. Half-Size Model of Typical TrussHousing.

103

Figure 48. Sketch of "A" Frame Truss Design.

104

F'igure 49. Sketch of Pyramid Truss Design,

105

It

Figure 50. Sketch of Pure Truss Design.

106

Figure 51. Sketch of Pure Truss Design.

107

Figure 52, Sketch of "A" Frame Truss Design WithComposite Material.

108

Figure 53. Sketch of Pyramid Truss Design WithComposite Material.

109

Figure 54. Skctch of Pure Truss Design With

Composite Material.

110

1+

Figure 55. Three-View Preliminary Layout of "A"

Frame Truss Housing.

-I-

'N'N

'N'3 II.

I III J., - ~ --- - - - IIf.. - ii I. .t.1 1.-i- '3 .3 33I .*Ij I £..~ - -3- -.4- -3.I, I I ,1 ;, 33

-3-33 -3'

'3 ~ , -I ,. 3321 33 13.:~ I'

- / -I, 33 '3 I.z.33 / 31 3333 / 3~ 33 . ~ I.*,, * 33 33

.*III 333 / , 3 33 A

/ 3I.1 / 331 I,/ III 3I '1ii / 33k' ~ L./ &. -.. :: 1 -. ~ -~

33 -3 3

'3 '333 3, '3-~

3,- I, 333 I 33

33 3333 3333 33

3 -3 3333 33

.33. 33 3333

33 33

331'3..

33 3333 3333 33

3333 3* 33 3'..

Q246

SCALE(ifl.)

iminary Layout of "A"

ing.

1

--- - - - -.

*3 3 I

SCALE (in.)

supported from the main structure, contains an "10 " ring grooveand bolt holes for mounting the present rear cover assembly.The second-stage inputs located on the right and left sides aremounted in rings supported by the sheet metal structure of themain housing. The first-stage input casting bolts directly tothe support ring mounted to the sheet metal structure. Thesecond-stage input bevel pinion roller bearing is supported bythe sheet metal structure.

The pyramid truss arrangement, Figure 56, consists of a seriesof rings and eight main load-carrying merabers arranged to havea theoretical intersection point to form a cone. This arrange-ment transfets main rotor shaft horizontal reaction loadsapplied in any direction at the uppermost ring through themounting ring to the airframe. The horizontal ring, in additionto supporting bevel gear housings and providing ring gearreaction torque, provides stability to the pyramid members andtransfers loads side to side. The base ring containing themounting is tipped five degrees to the side relative to themain rotor shaft and is constructed as a circle instead of anellipse. A vertical ring joined to the middle horizontal ringand stabilized through supports to the top horizontal ring andpyramid members provides the reaction for the tail-takeoff bevelpinion tapered roller bearings. Parallel to the tail-takeoffbevel support ring is the rear cover support flange. Thisflange is supported from the pyramid member by a series oftubes. It contains an "0" ring gcoove and bolt holes forattaching the present rear cover assembly. Each second-stagebevel pinion thrust bearing housing located on the right andleft sides, is mounted in a ring supported by a series of tubesfrom the pyramid main members. The first-stage input castingbolts directly to this tube-supported ring. The second-stageinput bevel pinion roller bearing ring is supported by a seriesof tubes from the pyramid structural members.

The pure truss arrangement, Figure 57, consists of a series ofrings forming elements of the cone and a series of diagonalmembers to transfer loads between the rings. This arrangementtransfers the main rotor shaft horizontal reaction loads,applied in any direction it the uppermost ring, and transfersthem indirectly through txie mounting ring to the airframe. Thehorizontal rings, in addition to supporting bevel gear housingsand providing ring gear reaction torque, provide stability tothe truss members and transfers loads from side to side, Thebase plane ring containing the mounting feet is tipped fivedegrees to the side relative to the main shaft and is con-structed as a circle instead of an ellipse. The tail-takeoffbeve.l pinion tapered roller bearing set is supported by avertical ring joined to the middle horizontal ring and isstabilized by supports attached to the other truss members.Parallel to the tail-takeoff bevel support ring is the rearcover support flange. This flange, whicl, is supported from


the truss by a series of tubes, contains an "0" ring and boltholes for attaching the present rear cover assembly. Thesecond-stage input bevel pinion roller bearing ring issupported by a member from the top ring, which is an oiltransfer tube, and by a series of tubes to the truss members.

Modeling of the Truss Geometry

A model of the fabricated housing truss geometry was developed.

Nine assumptions were used to simplify the analysis.

1. The skiai covering does not contribute to structurestiffness and carries no load.

2. Any connection to another member has completerigidity.

3. The bearing loads are sinusoidal and lie in thesame plane.

4. Any structure element is a beam in which the shearstresses are ignored.

5. Each connection has the same properties as thestructural element it is connecting.

6. The mounting ring used to transfer loads from thetruss housing to the airframe is mounted rigidly.The mounting pads are constrained in five directions,but allowed to twist about the axis of each bolt. Thebase ring in areas other than the mounting pads isconstrained in the vertical direction only.

7. The truss structure members join other members on theline of axes of their common centroid.

8. All applied loads are concentrated forces applied asdiscrete point loads in the structure. The aggregateof these applied loads has the correct magnitude. Thesinusoidal distribution for bearing loads has thecorrect magnitude even though it consists of a finitenumber of points.

9. All cross sections use structural elements that haveno holes, and bolt loads are concentrated on thebolt centerline.

114

Figure~~~~~~1 56 he-iwPelmnr aoto

PyaidTus ouig

Ik~115

'17

ij

/of

f==It /

ii i"" -

--- -, .- .... 1-.

~~02 k 6 5

SCALE (in.)

of

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

-44

Figure 57. Three-View Preliminary Layout of PureTruss Housing.

Preceding page blank L17

I I I \ \;-* -. . . . .---

i/ ' //

I ~ -

SCALE (in. )

Method of Analysis

Extensive use was made of the NASTRAN finite element computeranalysis coupled with a CRT graphic display.

The finite element method was used as a part of NASTRAN. Thisprogram has a library of eighteen finite elements that can beused to perform a variety of analyses.

Although specifically designed to solve large structuralproblems, NASTRAN has an efficient executive subprogram thatselects only those parts of the program needed to perform therequired calculations. With the statics option of NASTRAN,three-dimensional frameworks can be analyzed by modeling thestructure with bar elements connected at grid points. Thebar element of NASTRAN is a generalized beam capable of trans-mitting axial load and torque as well as moments in two planes.It enables the user to model the stiffness of beams havingasymmetric cross sections by defining offsets from grid pointsto locate the neutral axes. The program calculates displace-ments of the structure, and internal loads are referred to theneutral axis of the beam. By inputting fiber distances from theneutral axis and specifying the orientation of the cross section,the user can define four points of each beam element crosssection, where the stresses are calculated. When the materialdensity and allowable material stresses have been inputted, the

structure weights and margins of safety are calculated.

Capability exists for conwlraining any of the six degrees offreedom of a point. This enables fixing rotations and dis-placements at a point. In addition, the capability exists forspecifying the spring rate at any particular point in thestructure.

Loads are applied to a structure at a grid point by specifyingthe magnitude and the direction of the load. The capabilityexists for handling uniform loads anO sinusoidally distributedloads.

As part of the input data for a NASTRAN analysis, it isnecessary to specify grid points relative to some arbitrary setof axes and the connection of these parts. The elements arethus defined. With a large, complicated, truss-like fabricatedstructure of an asymmetrical shape in two coordinate directions,the voluminous input data required make it next to impossibleto check the data without display capability.

Display capability is provided by a Sikorsky Aircraft developedprogram employing an IBM 2250 CRT console. This device consistsof a cathode-ray tube (CRT), typewriter keyboard, function box,and light pen. The CRT console is linked with SikorskyAircraft's system 370 Model 145 computer.

119 Preceding page blank

The display capability makes use of the Sikorsky developed DataEditor Program. The designer views the geometric structurecreated by his connectivity cards, checks previously inputteddata, and interacts with the computer to create design changes.

The Data Editor Program has the following capabilities.

1. The Batch Card Processor section runs in the batchmode to read in the NASTRAN deck. It establishes thedefined connectivity relations and stores the resultsin the permanent files storage. A printout isgenerated with minimal diagnostics for the NASTRAUdeck.

2. The graphics Files Storage section permits the user tostore and retrieve up to twenty individual NASTPANdecks and their associated display data in a permanentstorage area.

3. The graphics Display program permits the user todisplay any selected portions of the connectivitygeometry of a NASTRAN deck retrieved from permanentstorage. With the display program, errors inconnectivity can be detected and corrected.

4. The graphics Card Edit program permits the user todisplay and alter the card images of a NASTRAIl deckretrieved from permanent storage.

All three graphics programs are interconnected by the NASTRANSystem Monitor. This assembler language program permits theuser to enter any of these programs, operate on his data, andreturn to the monitor repetitively. The Batch Card Processormust be scheduled as a separate batch run to be completed priorto the scheduled graphics run.

As an example of the method of analysis, consider the modelingof a typical truss-like fabricated housing structure. Thefabricated housing design, Figure 58, is a simplified model ofa truss-like structure to replace the present casting. Barelements are used for the connecting members, because they arecapable of having specified section properties in the planeof the neutral axis and orthogonal to it. The bar elementsare so joined that there is rigidity at joints, and the weightof the bar element is considered to be uniformly distributedover its length. The structure has a conical shape, with a ringat the upper main shaft roller bearing, two intermediate rings,and a large-diameter base or mounting ring. On each side of thestructure are input support rings and an input roller bearingreaction ring. In the rear of the transmission is the accessorycover support ring and a tail-takeoff support ring. The servosare modeled as a series of pyramids at the servo locations. All

120

support rings are connnected to the main structure by barelements. The rings, as modeled, consist of a number of pointsdefined on a circle with straight line segments connecting them.These straight line segments are the centerline of a barelement, each of which has specified section and materialproperties. Each defined point, or grid point, is located inthe global coordinate system or in a coordinate system definedrelative to the global system.

I!

Figure 58. Typical Truss-Like Structure in DisplayProgram.

Figure 59 illustrates the cards used in the NASTRAN deck setup.For the display capability, coordinate system definition cards(C0RD2c), mesh point description cards (GRID), and connectivity

A cards (CBAR) are necessary to execute the batch card processor.The results are stored in a file, and error messages are printed.If many errors are found during a run, a considerable savina ingraphics time can be achieved by correcting the deck and re-submitting it for another run through the batch card processor.If there are minor errors, they can be corrected at the CRTusing either the Display Program or the Card Edit program.Figure 60 shows typical errors in the NASTRAN deck setup.Figure 61 shows a designer changing an element in the DisplayProgram with the light pen. When the structure shown on theDisplay Program has been corrected and modified to the satis-faction of the designer, a deck of the updated and correctedversion is punched through use of the Files program.

121

1 4

H 4)

Pii

PL44

0U

00

Ltn

0 $4

P- C

02

If they have not already been added, it is necessary to addthe rest of the NASTRAN deck setup at this point, as shown inFigure 59. These cards include a series of control cards tospecify the type of output required, internal moments, stressesand displacement for each end of all elements, and descriptiveinformation for each loading condition. A series of connectivitycards (CBAR) specifies each bar element and the grid points itinterconnects. Also specified on connectivity cards are theorientation of the neutral axis of the section and the applicableproperty card.

The next series of cards consists of coordinate system identi-ification cards, which specify the orientation of a particularcoordinate system relative to some arbitrarily defined globalcoordinate system. Specified for each new coordinate systemrelative to the global system are the location of the origin,direction of the plus "Z" axis, and direction of the azimuthplane. The capability exists for definition of coordinatesin rectangular, cylindrical, and spherical coordinate systems.For the truss-like fabricated housing, coordinate systems aredefined in rectangular and cylindrical syst-ms. Various loadingconditions are added through a series of force cards, whichspecify the loading case, grid point of the applied load, andmagnitude and direction vector.

The grid cards define the coordinates in any specified axesand the degrees of freedom of point. It is possible to definereference points, interfaces, and boundaries.

Through a series of material property cards, the allowablestresses, moduli, and densities are specified.

The section property cards are referenced by the connectivitycards and determine the applicable material, as well as giveareas, inertias, and torsional constraints of each bar element.

The sequence cards (SEQGP) in the deck setup are generated bya program called BANDIT. The purpose is to reduce the use ofcomputer time for stiffness matrix inversion by minimizingmatrix size. The sequence cards renumber all grid pointsinternally within the computer for a minimum matrix size.

After deck setup, the program control cards are altered toprevent execution of the program, and the program is run throughthe data check. The data check helps locate errors that mayremain after checkout by the display program.

123

Figure 60. Truss-Like Structure with Some InitialError.

Figure 61. Photograph of a Designer Using CRT ToChange a Structure Element.

124

When the NASTRAN deck is executed, the outputs include a dataecho check, grid point weight generator. table, estimates ofdecomposition time, and the accuracy of the solution received.In addition, a displacement vector table, load vector table,table of forces in bar elements, and table of stresses in barelements are printed out. From analysis of the various tables,section properties, element number, and location cards arealtered. The deck is then rerun to obtain a solution.

Discussion of Results of Analysis

The three housing configurations initially considered in thepreliminary analysis are shown in Figures 62 through 64 as theyappeared on the graphic display terminal. Views (a) and (b)include the loading spokes which were used to model sinusoidallydistributed loads at the bearing reaction rings.

Of the three basic designs, the pure truss was by far thestiffest, followed by the pyramid and the "A" frame in thatorder. The analysis shows that the pure truss was most effec-tive in distributing loads over the greatest number of members.During forward crash loading, for example, a substantial portionof the load is carried by the side support members for the puretruss configuration. In the case of the "A" frame and pyramiddesigns configuration, nearly the entire load is carried by theforward and aft support members. These load distribution char-acteristics of the fabricated housing designs have tremendousimpact on the vulnerability of the transmission. Should a keyload-carrying member be lost in either the "A" frame or pyramiddesigns, the housing would collapse since no other load pathto the airframe would be available. The truss design, however,would simply redistribute the load over other members, therebydecreasing substantially the possibility of housing failure.

125

U)a)

r:4

I4-0

1261

4

.4o

0.9

E O,

0

.0

0)

04-

127.

U)

44

0

>1

.1-I

E-4

U)

D-4

128

All three designs were analyzed in stainlpts steel, titanium,and Borsical® configurations. Borsical proved to be notonly the strongest, but also the lightest of the three materialswhen incorporated in the housing analysis. Stainless steel wassecond in both categories, with titanium ranking last. Table XIsummarizes the results of the analysis. The stiffnesses shownin this table are for a load applied in the center of the uppermain rotor bearing ring.

On the basis of these results, a pure truss configuration waschosen fo~rdetail design. Although the analysis showed thatBorsical was the superior material, Custom 455 stainlesssteel was chosen as the mat ial for the pure truss design.The brazed--bonded Borsical was eliminated due to itsrelatively low joint strength.

TABLE XI. SUMMARY OF WEIGHT AND HOUSING STIFFNESS FORVARIOUS CANDIDATE MATERIALS

Weights, lb"A" Frame Pyramid Pure Truss

Material Design Design Design

Stainless Steel 327 309 278Custom 455

Titanium 330 318 290Ti-6A1-4V

Borsical 290 265 230

Stiffness, 106 pounds per inch

Stainless Steel 11.2 12.0 16.2Custom 455

Titanium 10.0 10.4 14.6Ti-6Al-4V

Borsical® 12.0 14.0 18.9

129

DETAIL DESIGN OF TIlE TRUSS FABRICATED HOUSING

Design Criteri.

The final loading diagrams for the fabricated truss-like housingwere shown in Figures 23, 24, 31, and 32. Conditions studiedwere forward crash, side crash, hover, and symmetric dive andpullout. The loads for these four conditions were derived fromthe weights and CG locations of the components listed in TableIII from the aircraft loads report (Reference 2), and from thestructural design criteria report (Reference 3).

Description of Truss Geometry

The detail design of the fabricated housing evolved to a sym-metrical true truss design, as shown in Figures 65 through 68.The truss-like fabricated housing is shown with skin and accesspanel details in Figure 67. The housing consists of foursupport rings, each with constant cross sections. The topring is used to react loads from the main rotor shaft rollerbearing and the bevel gear outer shaft roller bearing. Thesecond ring reacts the loads from the second-stage bevel gearouter shaft tapered roller bearings, and the third ring is usedfor reacting the ring gear torque and lower housing loads. Thelubrication oil transfer glands are located on this ring. Thelower cover, which contains the main rotor shaft split-inner-rape ball bearings, is the same as the present CH-54B casthousing design. The bottom ring contains the mounting pads tofasten the transmission housing to the airframe support beams.

A left and right side input ring is used to support the second-stage input bevel pinion assembly and is also used to cvpportthe first-stage bevel gear input housing assembly and the rearhalf of the JFT4D engine. These rings are supported by a seriesof tubes from the main truss structure. An input ring on theleft- and right-hand side, also joined through a series of tubesto the main support structure, is used to support the rollerbearing of the second-stage input bevel pinion assembly. Aring located in the tail-takeoff section supports the tail-takeoff bevel gear tapered roller bearing attached to the tail-takeoff accessory flange ring. This ring is supported by themain truss structure through a series of tubes. The tail-take-off accessory flange supports the rear cover, cover-mountedaccessories, and oil cooler support structure.

130

SEC A-A-

ROTATED 90' C.C.W.

Transmission Housing (View LoigAft).

Fiur 6. etilDeig ayutofth Fbrcae

D

41 - IIF

// iOJ

SEC B- 02SCAL (in

I0

0~9

)0 2 14 6* ~~VIEW D-DSC E ( n )

Figure 66.* Detail Design Layout of the Pab icatedTransmission Housing (View LOokingForward).


0

0

0

0 g

VIEW E-E

S AE (in.)

gn Layout Of the

Plabricated

33

A

B

G '

ROTATED

SCALE (in.)

PLAN VIEWFOR MAIN HOUSING STRUCTURESEE AUXILIARY PLAN VIEW C

Figure 67. Detail Design Layout of the FabricatedTransmission Housing (Plan View).


G G

F ASTNERS

ot f thariaesin (Pa Viw)

1X

02 1.6

AUXILIARY PLAN VIEW C SCALE(in.)

Figure 68. Detail Design Layout of the FabricatedTransmission Housing (Auxiliary Planview). Preceding page blank

137

4-)

rl4)

0

4J )

4.4

r4E-4

P14

0.,

0*0

.0 t

-.0 004

138

Analysis

The final truss-like housing design geometry was modeled as aseries of thin rings with connecting truss elements. Figure 69is a series of photographs of the CRT display of the geometrycoded for analysis in the NASTRAN program. The inside ringdiameter contour is determined by bearing or mounting interfacerequirements, and the appropriate properties were coded for theanalysis. The section properties of each of the rings werekept constant to simplify manufacturing. As a result, thering is not fully stressed for most of its circumference.

The analytical output from the program gives the internal forcesand moments at either end of each structural member and theresulting maximum axial and bending stresses. In addition,displacements at each end of the element are given. The cross-sectional properties and/or the location of the members weremodified as necessary to satisfy the overall stiffness require-ments and to obtain a minimum margin of safety.

The result of the analysis for the crash design condition is astructure weighing 305 pounds and having a minimum margin ofsafety for the ultimate strength of Custom 455 stainless steel.Maximum deflection of the upper ring for the crash loadingcondition is .480 inch.

To determine structural adequacy for normal flight requires thedesign to be reanalyzed for both the hover and symmetric divean- nullout conditions. Adequacy for these conditions is basedon minimum yield margin of safety.

The material properties of Custom 455 are: ultimate allowabletensile strength, 220,000 psi; yield allowable tensile strength,135,000 psi. The minimum margins of safety for the yield andultimate allowables were positive for all conditions.

Analysis of buckling for the truss structure members indicatedthat the critical column length is always greater than thetruss member column length for normal flight conditions.

For an analysis of fatigue loading conditions, the ground-air-ground (GAG) cycle column was considered to be critical. Themost highly loaded member has the following bending and axialloads:

Mxx = -146

Myy = 398

Fa = -4456

139

The section properties for this structural member are

Ixx = .0675

1yy = .0265

A = .4687

The offset to the critical point on this section is x = .4583,y = .6041. The maximum stresses were found to be

fxx= 2530

f = 3560

fa = 9570

The maximum vibratory stress for the GAG cycle was found to be

fvib = 7830 + 7830

, The margin of safety is

A .S. = 1 - 1 = .16

7830 + Kt 7830

Fty (.05 En )

where Fty = Tensile yield strength

En Endurance limit taken from Table VIII

1t = Concentration factor

140

COIPARISON 01' TIE FABRICATED TRUSS-LIXE HOUSING WITHTHE CH-54B CASTING

For a comparison of the fabricated truss-like housing with theconventional casting design, the following were considered:cost, reliability and raintainability, corrosion characteris-tics, survivability, vulnerability, and manufacturina ease. Thecast housing and the fabricated housing were not compared on astructural basis since the castinq is less amenable to preciseanalysis. Characteristics such as weight, cost, stiffness, andcorrosion resistance were comparable, however.

The fabricatgd housing weighed 305 pounds and had a stiffnessof 19.2 x 10 pounds per inch. The cast housing weiahed 363pounds and had a stiffness of 12.47 X 106 pounds per inch.

Cost Comparison

Table XII shovs comparable costs of the fabricated housing andthe cast housing. Materials and labor cost data for orderquantities of 5, 10, 25, 50 and up were Oetermined. Toolingestimates of $211,900 for the fabricated nousing and $283,500for the cast housing have been made. The fabricated housingon a unit basis is approximately $11,000 less than the presentcast housing.

TABLE XII. COST COMPARISOt, CONVENTIONAL CASTING

VERSUS FABRICATED HOUSING DESIGN

Unit Cost

Quantity Conventional Housing Fabricated Housing

5-9 $38,500 $26,600

10-24 $36,280 $24,450

25-49 $32,100 $22,050

50 and up $28,880 $20,750

141

The savings reduce to S7,000 with 50 or more units. In additionto being less expensive, the fabricated housing offers a savingsin lead time of three months.

Reliability and Maintainability Comparison

The most significant problems common to all CII-54B cast housingsin the field are nicks, dents, scattered corrosion, and stresscorrosion. Figure 70 is a sketch of the conventional castingarrangement. With the typical casting mounting arrangement,corrosion takes place in fifty percent of the qcarbo::es at areas

D, E, F' and on the housing and airframe interface surface.Area C corrosion results from the bolt being installe(I to the air-frame without dichromate primer. Area F corrosion is a result ofthe lack of sealer o in, the presence of moisture at the mountinginterface. The mounting interface, which has the ma(gnesiumcasting on one side and the aluminum forging on the other,corrodes due to the difference in electrolytic chemical activity.This interface is caulked on the extremity, but there is separa-tion or cracking due to high maneuver vibratory loads. Leakinamay also occur near the barrel nut by capillary action, sincethe barrel nut cuoes not completely fill the hole.

On the fabricated housing, the material and joininq methodsselected pi'eclude the possibility of these problems. As anexample, almost all housings coming to Sikorshy Overhaul andRepair facility exhibit corrosion inside the mounting feetbushings. These bushings are cadmium plated and are used withcorro:ion-resistant bolts. The corrosion that appears is theresult of wearing of the cadmium plating as the bolts aretightened.

Bushings are used on fabricated housings to enable reuse of thecomponent in the event of wear or damage. IT:he bushinq flangealso provides a large enough area to keep unit stresses low,while the press-fitted bushing body provides a large shear areato react shear loads. Figure 71 is a sketch of the detail ofthe mounting foot with bolt and bushing installed. Shims areemployed as sacrificial elements to prevent water from seepinabetween the bolt and housing. These shims are approximately3/16 of an inch thick to prevent splitting under high bearingpressures. The steel bushings are installed in the, fabricatedhousing with wet zinc dichromate primer. The bushings arefrozen and then pressed into the housing with an 10 " ringunder the bushing flange. when the bolt is installed in thebushing to mount the housing to the airframe, the bolt isinstalled with wet zinc dichromate primer on the bolt with an"0" ring under the bolt head. The "0" ring fits between thechamfer of the bushing and the root radius of the bolt head.

142

TWELVE POINT EXTERNAL

WRENCHING BOLT

E

STEEL BUSHING

CASTING

D

C

ALUMINUM BEAMOF AIRFRAMESUPPORT STRUCTURE

BARREL NUT

Figure 70. Detail of Typical Casting MountingArrangement,

143

/-t1011 RING

"O"0 RING

STEEL BUSHING

SACRIFICALHOUSING -SHIMS BONDED

TOGETHER

HOUSING MATERIALAIRFRAME MATERIAL

II

BARRE NUT?

AIRFRAT!--MATERIAL

Figure 71. Detail of Typical Fabricated Housing

Mounting Arrangement.

144

The failure modes of a cast magnesium housing in decreasingorder of importance include corrosion, creep, fatigue due toGAG cycle, static yield, and stress corrosion cracking dueto pressed in bushings and pins. On the fabricated housingthe material and joining method selected preclude the possibilityof these failure modes with the exception of stress corrosion.The problem of stress corrosion is reduced through the use ofCustom 455 which has higher fracture thoughness properties.

The problem of field inspectability of the cast housing has beeneliminated with replacement by a fabricated housing. Presentcastings to be fluorescent inspected require the removal ofthe protective finish, paint and resin coatings. The fabricatedtruss housing made of Custom 455 stainless steel has no pro-tective finish on it, since it is self-passivating, and may beinspected on the aircraft through magnetic inspection techniques.

The detectability of cracks on the fabricated housing is simplerthrough on the aircraft inspection techniques and is not criticalto structural integrity of the housing. The fabricated housingis a redundant structure and the loss of any one truss memberis easily tolerated.

It is expected that a reduction in housing repair cost will beachieved through the use of a fabricated housing since problemsassociated with corrosion and creep have been eliminated. Over-all reduction of housing failure rates can be expected but aredifficult to quantify without test data.

Corrosion Properties Comparison

Custom 455 stainless steel is far superior to cast magnesiumwitl respect to corrosion. The stainless steel, including thewelded areas, requires no surface protective film because itis self-passivating. A cast magnesium housing requires a seriesof protective surface finishes to maintain its corrosionresistance. Even though much work has been done on surfacefinishes for magnesium and the results are good, self-protectingmaterial* was considered a good goal for a fabricated housingdesign. The scattered corrosion problem would thus be elimi-nated. Stress corrosion does not exist, since in the weld areasof fabricated housing, the structure is normalized afterwelding.

Crashworthiness

The work done on crashworthiness indicates that the fabricatedhousing design would survive for the indicated loading of 20gforward, 18g side and 20g down even though there would beconsiderable yielding.

145

Heat Dissipation

A typical main transmission rejects approximately 12-15 percentof the total heat generated through its housing. The rest isrejected through the oil heat exchanger. The actual heattransfer coefficient of a cast housing is unknown, since thecast housing includes dichromate primer and acrylic lacquer onthe exterior and a baked resin coating on the interior. Thefabricated housing rejects approximately one-third of the heatthat a cast housing rejects. This results in a 5-percentdegradation in overall heat rejection performance and is aconservative estimate.

Survivability and Vulnerability

The basic study requirements for the fabricated transmissionhousing omitted a survivability/vulnerability analysis. Theredundancy of the ultiple-structural-clement design makes theimproved survivability aspects obvious when compared to thesingle redundancy of the present monocoque casting construction.With multiple element design, resistance through 23mm IIEI isinherent with no weight penalty. The survivability/vulner-ability improvements from this design method are several and arelisted below:

- The design is multiply redundant compared to thesingly redundant approach of castings.

- Ballistic damage or destruction of one or morestructural elements removes only those particular loadpaths and may lead to eventual fatigue crack initia-tion in other paths as opposed to immediate fatigueinitiation and propagation in monocoque structures.

- Gears are already considered survivable againstalmost all hits from higher threats because of theirbulk and hardened steel construction.

- Bearings have remained a problem: additional designfreedoms allow alternative solutions to bearingsurvivability that did not exist before.

- Truss structures may be left open during assembly(and skinned over later), allowing additional stopsto be installed on gear shafts to prevent lateralmotion of bearinqs after ballistic damage and toprevent separation of gear mesh.

- Weight savings inherent in multiple element designmay be applied to additional passive bearing armoror to multiple redundant bearing installations oremergency lubrication systems, heat pipes, etc.

146

Fabricated transmission housing materials have beenchosen to operate under emergency conditions (6000 Ffor one-half hour) with minimum deterioration of thoseportions of the transmission not directly damaged bythe projectile(s). Repair of this battle damage hasbeen made far simpler and, with further development,may be reduced to the field or depot level, ratherthan the present approach of replacing entire trans-missions.

It is these gains, more than any others, which make the fabrica-ted transmission housing concept so promising for future.incorporation in Army helicopters. At a time when the impor-tance of survivability/vulnerability is becoming recognized instringent design requirements for future helicopter production,a design technique exists which removes the transmissionhousing from attrition, forced landing or mission abort andsubstantially reduces the problem of repair. The technique alsosaves weight which can be applied to improved survivability ofcritical high-speed bearings with sufficient amounts of weightreduction left over to improve performance of the vehicle. Itis strongly recommended that future fabricated transmissionhousing requirements include in-depth studies of the surviv-ability/vulnerability aspects outlined above and that design,fabrication and test include definitions of trade-offs (weight,cost, survivability/vulnerability improvements, performanceimprovements) which should be investigated to meet future Armybattle-damage needs.

Manufacturing

The fabricated housing requires considerable labor for jiggingand hand welding. The casting requires a lead time of 9 - 10months for pattern equipment, tooling, and machining. Thefabricated housing requires less machining and offers theadvantage of easier machining due to increased access (no skinon) areas. The fabricated housing permits repair duringmanufacture much the same as weld repair procedures are used forcast magnesium. The fabricated housing also has the advantageof permitting manufacture of small subassemblies before joiningthe main structure.

147

Summary

A list of parameters for comparison of the cast housing with thefabricated truss housing is shown in Table XIII.

TABLE XIII. SUMIARY OF A CAST MAGNESIUM HOUSING WITHTHE FABRICATED TRUSS HOUSING DESIGN

Item Weight Lead Time(lb) (mo)

Fabricated Truss Housing 305 6

Cast Housing 363 9

The steel fabricated truss housing has considerable advantageas a replacement for the cast magnesium housing. The fabricatedhousing offers a saving in weight and lead time and is lessexpensive.

148

EFFECT OF SCALING A FABRICATED HOUSING DESIGN

The fabricated housing was additionally analyzed to study theeffect on housing size and weight for a +25 percent change ininput torque.

The approach taken in this study consisted of scaling theaircraft size up and down and then identifying the significantparameters that correspond to a +25 percent change in torque.Since a mathematical model does Not exist for the CH-54Baircraft, a utility transport aircraft was modeled and the loadswere scaled to the CH-54B. This comparison has inconsequentialerrors in the estimation of fuselage inertias and weights, sincea crane helicopter has almost no cabin area. Figure 72 is aplot of hub moment constant versus design gross weight. Figure73 gives historical data of hub moment versus design grossweight with aircraft built to date. These figures illustratethe validity of the scaling procedure.

Parametric Scaling of Loads

For the scaling procedure, disc loading and tip speed were keptconstant along with number of blades and payload/gross weightratio. The rotor diameter, gross weight, power required,component weights, and rotor shaft moments were so determined.

Dimensionless plots were developed to provide a basis fordetermining flight loads and crash loads for +25 percent changein design input torque. To determine new deslgn gross weights,Figure 74 was used. The design gross weight ratios are 1.15 and.76 for +25 percent change in design torque ratio. Thus, thegross weights of the sensitivity study aircraft are 54,100 and37,700 pounds. Figure 75 is a plot of head moment constantratio versus design torque ratio. For +25 percent change indesign torque ratio, the head moment constant ratios are 1.375and .875. These ratios correspond to head moment constants forthe sensitivity study aircraft of 1020 and 6540 ft-lb/deg. Itis assumed that the flight loads vary in direct proportion tothe head moment constant. To determine the crash loads, anumber of plots of weight group areas were generated (Figures76 through 79). Plots are included for main rotor group weightratio versus rotor design torque ratio, and servo group weightratio versus rotor design torque ratio. Under crash loadingconditions, the rotor shaft reactions at the upper rollerbearing for +25 percent change in input torque are 201,660 and133,450 pounds. For the side load on the lower thrust bearing,the reactions are 98,100 and 64,870 pounds. The flight controlweight group ratios indicated that servo crash loads changedwith a +25-percent change in rotor design torque ratio to 793and 558-pounds.

149

10~ __ _ -

0

0

Q 10 20 30 4i0 50 6Q 70 80

DESIGN GRO0SS W~EGHT (X 10 00 lb)

Figure 72. Historical Data - Hub Moment Versus DesignGross Weight.

150

I8 55

Is64

--

S56

4 - --

34- - -

C162

K ].-- 561 -

I 562 61

0 5 10 15 20 25 30 35 4o 45

DEsIGN GROSS WEIGHT (X 1000 lb)

Figure 73. Historical Data - Hub Moment ConstantVersus Design Gross Weight.

151

2.0-

1.5.

1.0

11.00-5

0

rm 0.5

I

I Ii I

0

0 1.0 2.0 3.0 4,o

DESIGN TORQUE RATIO

Figure 74. Design Gross Weight Ratio Versus DesignTorque Ratio.

152

4.0

3.0O

0HI

E-4

2. G

0 1.0 2.0 3.0 4.oDESIGN TORQUE RATIO

Figure 75. Head Momnent Constant Ratio Versus DesignToqeRatio.

153

3-0-

0

o2.0-

1.0

S .0 ~ 1. 20 .0 4,

DES-IGNTRU AI

VesDeGN TOrQUE RATIO.

154

3.0-

0H

M 2.0-

0

E41.0"0

PrIi

0

0 1.0 2.0 3.0 4.

DESIGN TORQUE RATIO

Figure 77. Main Rotor Group Weight Ratio VersusDesign Torque Ratio.

155

3.0Cr

o 1.0

0.

0 1. I. .

DES-INT~ AI

Figre78 F igh oto ru egtRtoVruDe ignTrueRto

156

3.0

0

E-4

E.0

00

14 0

0 1.0 2.0

DESIGN GROSS WEIGHT RATIO

Figure 79. Head Moment Constant Ratio Versus DesignGross Weight Ratio. j

157

For example, Figure 75 shows lead moment constant ratio vcrsus-design torque ratio. The changes in head moment constant were1.375 and .875, respectively, for +25 percent change in inputtorcue. p'he loads on the upper roller bearing were scaled

proportionally with the change in head moment constant. For+25 percent change in input torque, the upper roller ))earingreactions were:

+25 percent - 25 percent

Rax = (1..375)(12640) = 17380 Pay = (.875) (12640) 11060

Ray = (1.375) (9670) = 13300 Ray = (.875)(9670) 8460

For the lower thrust bearing horizontal reaction, the scaledloads of +25 percent in input torque were:

+25 percent -25 percent

1Rbx = (1.375)(-27326) =-37573 Rbx = (.875)(-27326) = -23910

rby = (1.375)(-967G) -13304 Rby = (.875) (-9676) = -8466

Table XIV summarizes the changes in applied loads used in thestructural analysis, compared with the fabricated housingdesign.

Analysis

For a +25 percent change in input torque, the weights for thefinal loub.e.ag design in stainless steel were 362 and 254 pounds.

For the craslt.conditions, the minimum margin off"safety was zero.For flight cAditions, the minimum margin of safe'ty that isacceptable was the same as the baseline configuration. For bothdesign criteria, the fabricated housing was adequate.

For a buckling analysis, with the allowable critical bucklingstress taken from Table VIII, Fsecant = 156,000 psi at roomtemperature for Custom 455 stainless steel. For the fabricatedhousing design with the truss length members selected, the trussmember column stress was always less than the critical columnstress.

For the low cycle fatigue condition (ground-air-ground cycle),'he highest loaded truss member for the -25 percent change ininput torque has the following bending and axial stresses:

158

TABLE XIV. C1C11

-25% Input Tor !

CrashLocation x y

Rotor Shaft Roller Bearing Reaction 133,444 133,440 -1i

Rotor Shaft Lower Thrust Bearing -64,864 -64,864 23,;

Left Side Input Ring 3,860 3,860 51

Right Side Input Ring 3,860 3,860 51

Tail-Takeoff Accessory Flange 19

Servo Supports 558

Scissor Bracket

Combining Bevel Gear ReactionI1

Rotor Shaft Lower Thrust Bearing

Left Side Input Roller Bearing

Right Side Input Roller Bearing -1

Tail-Takeoff Bevel Pinion -2

1

159

TABLE XIV. CHANGES IN APPLIED HOUSING LOADS FOR A + 25 PERCENTCHANGE IN INPUT TORQUE

-25% Input Torque Baseline ConfigurationCrash Flight

Crash Flicht Forward Sidex y x y z x y x y z

133,444 133,440 -11,060 8,467 165,297 165,297 -2, 4J. 9,676

-64,864 -64,864 23,910 8,466 -81,103 -81,108 27,326 9,676

3,860 3,860 5,390 7,125 8,040 5,147 5,147 7,120 9,500 10,720

3,860 3,860 5,390 7,125 8,040 5,147 5,147 7,120 9,500 10,728

19 loacis 165 each -9 19 load 200# each -12

558 each 558 each 7,481 690 each 690 each 8,550

-762 -762 -13 -953 -953 -16

1,962 -248 -1,692 2,617 -331 -2,256

37,128 15,408 Thrust & Ring Gear 49,476 94,260

1,650 2,200

-1,650 -2,200

-2,755 4,410 9,636 -3,674 5,830 12,840

S FOR A + 25 PERCENT

Baseline Configuration +25! Input TorqueCrash light Crash Flight

)rd Side Forward Side

y x y z x y x v.165,27 -12, 4 9,676 201,860 201,860 -17,381 -13,304

8 -81,108 27,326 9,676 -98,120 -98,120 37,573 13,304

7 5,147 7,120 9,500 10,728 6,433 6,433 3,900 11,375 13,404

7 5,147 7,120 9,500 10,728 6,433 6,433 8,900 11,875 13,404

19 load 200# each -12 19 load 255 each -15

690 each 690 each 8,550 793 each 793 each 11,756

-953 -953 -16 -1,191 -1,191 20

2,617 -331 -2,256 3,271 -413 -2,20

& Ring Gear 49,476 94,260 -61,848 101,328

2,200 2,750

-2,200 -2,750

-3,674 5,880 12,840 -4,592 7,350 15,056

fb = 5959

fa = 12925

The maximum vibratory stress for the GAG cycle was found to be

fvib = 9442 + 9442

The highest loaded truss member for the +25 percent change ininput torque has the following bending and axial stresses:

fb = 7135

fa = 11,427

The maximum vibration stress for the GAG cycle was found to be

fvib = 9281 + 9281

The allo7able material endurance limit for Custom 455 stainlesssteel to 37,050 psi. The margin of safety for the -25 percentchange in input torque is M.S. = 0.72, and the margin of safetyfor a +25 percent change in input torque is .'1.S. = 0.75.


161

DEVELOPMENTAL AREA DEFINITION

Before the fabricated housing concept can be incorporated intoactual flight hardware, certain areas will require furtherdevelopment. During this study, these potential problem areaswere identified. These areas are associated with the analysisas well as the fabrication and qualification of the housingdesign.

Analysis

Because of the extreme general nature of the N,1ASTRAN program andits capability of handling very large matrices (orders of 10,000have already been solved), it is not particularly efficient onsmall programs, and large expenditures of computer time result.The amount of computer time used increases geometrically withthe number of truss members in the structure. For the trusshousing of the preliminary design configuration, with approxi-mately 300 elements, about 20 minutes of actual computer programutilization time was required to calculate the stresses, dis-placements, and margins of safety. For the final housing designconfiguration, about 60 to 90 minutes of computer time wasrequired for approximately 700 elements in the structure.

The CRT capability permits display of the truss geometry in anyorientation. Through use of the Data Editor Program, connecti-vity changes are permitted, but new element and geoimetry cannotbe added at the tube without going back to the Batch CardProcessor (see page 102). Use of this program at the CRT consumesabout 20 minutes, during which nothing else can be done at thetube. The handling of more than one deck simultaneously withthe Data Editor Program would be desirable. To checlk loaddirection, a display capability for the applied loads is needed.To help in understanding the IASTZATI output, a display ofnormalized deflections relative to the housing geometry displaywould be extremely useful and permit quick assessment of loadpath and imiprovement in structure weight. The analysis of thetruss housing with NASTRAN! to obtain a minimum weijht l.siqn wasaccomplished vith a trial-and-error approach. Minimum weightdesign programs, in conjunction with NASTRAN, would permit atheoretical solution for a minimum weight truss structure.These results would then be changed to limit the number ofmembers of different sizes.

A plastic analysis capability will make possible a lighterstructure design for ultimate criteria under crash loadingconditions. The bending analysis that was completed assumes alinear, elastic relationship between stress and strain. Aplastic analysis program would account for the redistributionin load capability o9 the structure members as the yield pointis exceeded.

162

A thermal analysis capability will permit the calibration ofinduced internal stresses as a result of the nonuniform transferof heat from localized heat sources within the gearbox. Thermaleffects at the truss housing interface with the airframe supportstructure would also be evaluated.

Experimental Verification

The fabricated housing can be substantiated and verified byan endurance tesL program run at operational loads. Asseri:)lyof the gearbox would verify that all of the interface require-ments associated with the dynamic components (gears and.arings) as well as peripheral equipLient (accessories, inputsection, etc.) are satisfied. Structural substantiation of thehousing structure itself should be shown by a 350-hour endurancetest program as outlined below. Also included in the test planshould be a no-load lubrication test and an overtorque testsimilar or equal to the following:

Test Plan (Suggested)

A proposed test program should consist of a no-load lu'-ricationtest and 350 hours of bench testing with simulated loading torepresent hub moment and thrust on the housing. The programshould be much the same as the program designed to qualify the

CII-54B for FAA certification.

Dench testing should be conducted in Sikorsky's CII-54,/S-64 Hain Transmission Test Laboratory. The regenerative benchtest facility to consist of three closed mechancial loops (twoinput and one tail-takeoff) in which the required torques arelocked in. The required shaft speeds to be provided by a sinaledrive motor and eddy current clutch. The test comnonentexperiences the torque equivalent of the rated test power,independently of actual power supplied. The test facility isshown in Figures 80 through 83. A hydraulic load-absorbingsystem loads the accessory pads of the test. transmission rightside input. To simulate main rotor shaft moments and thrust,the housing dolly and the fabricated housing upper rollerbearing support ring should be arranged to permit loadingthrough hydraulic cylinders, as shown in Figure 83. This wouldpermit the testing of the fabricated housing design under allpossible loading conditions prior to a flight test.

Before the 350-hour endurance test, the transmission should besubjected to a production transmission acceptance test.

Thirty-five 10-hour endurance tests should be conducted inaccordance with the spectrum presented in Table XV.

163

Repodce from.....best availablle copy.

Figure 81. CII-54A/13 Regen-Figure 80. CH1-54A/B Regenerative erative TestTest Stand -View of Stand ControlTest Stand. Room.

Figure 82. C1H-54A/3 Regenerative Test Stand -View

of Drive.

164

Figure 83. Test Rig To Simulate Main Rotor ShaftLoads.

165

0 0)raH co

a) 00~ 0 0 -30 000 0 0 C) C0in 0 C) H1 -104r-4090 0C0 000 0 0C) C cl, 00 0)

r-I H H-IH H HqH H H- HHI H- r- 14J

4) 00 00 C 0 00r.a 00 0)0 0 H>1

ro W r- .* % % . %. I )U -1 00C 00 0 4.)

in :

4-) 00 0)0 0

ul 3 -I 00 00 0 HC

P (4 cNN NE-4d II)in Lin inV)a

U445

r- Ha) a). 00C 00 000 0 0 0 00 0 04.14 :'4 0 in Otin 000 0 0 in i LIn! in c c

Z 11( t 0 -HC H N v 1.4 -l r I0N A -

111 0

0 (d:IQ 0 0 00 000 0 C 0 0 00 C4Jt 4w 00 4 ON 0) 000 kD %D ~ ON 0 m 0 o o

I E-4H I f- N t-r m oDk in in m Tw

4J0

:D in LO i co 000 1- -r co 0 0 4.)

(n MH f IH MH mN (7 N m c N Hm

00 HI

4-1 $4 >1:10 0 0 0)0 000 0 0 0 0 a 0

* 4 04 :0 'F i o in0 0 00 r- *' CO 0 0 H-> , 1 O04 o ~ m m mCo co m 00 m ()

x 4HPW, M r-I n H MMCN N N H >i4

(I) 0)0 0)r

0 0 0

NN 0

C'0 £0dH

0 0) 4-10 0 P*dOJ C4.) 0 'X P X P

O1 0. w0 .0 . 0. P. HC

5.j 4 -* > ' r-IH (nHd ar 0)) 4) 4.) pe Q 0

0 Co (a fu0 0 0 0 000 0. > 0 0u M 4 ; o'. coU 'fl tDu o 0 Z P

166

During at least 10 test cycles, the main transmission right-sideinput hydraulic pumps should be driven and loaded to the maximumrate specified in Table XVI.

TABLE IXVI. R/11 INPUT ACCESSORY PAD LOADING

OperatingAccessory Pressure Flow Speed Forse-

(psig) (gpm) (rpm) power

Second-Stage Servo Pump 3,000 14 4584 30

Hoist Pump 4,000 30 3796 80

To demonstrate main transmission housing overtorque capability,the main transmission should be subjected to twenty-two 10-second cycles at 8700 total input shaft horsepower and 1100horsepower at the tail-takeoff. A complete dual-engine over-torque cycle is described in Taole XVII.

The following data should be obtained during the tests:

1. strain gage measurement of the fabricated housingfor all loading conditions,

2. spectogp DPhic analysis of lubricating oil samples at25-hour Latervals during the bench test,

3. transmis. ion oil pressures, temperatures, and flowat all test conditions during the bench test, and

4. housing surface temperatures.

At the end of 200 hours of endurance testing, the fabricatedhousing transmission should be condition inspected. A spectro-graphic oil analysis, chip detector inspection, and strainerinspection should also be conducted. The fabricated housingtransmission used for the bench test should be completelydisassembled on completion of the 350-hour test. The trans-mission components, truss housing, and skin covering should besubjected to a condition inspection including zyglo/magnafluxand a conformity inspection. The test report should includethe following items.

1. 200-hour and 350-hour inspections performed duringprogram.

167

2. Spectographic oil analysis performed during test

program.

3. Applied housing loads and strain gage measurements.

4. Conclusion and recommendations.

Design Guidelines

1. Use a structural design of many lightweight truss membersfor redundancy, rather than a fewer number of large-sizemembers.

2. Assume that the skin covering does not contribute tostructural stiffness and carries no load.

3. Assume the bearing loads to be sinusoidal and to lie in thesame plane.

4. Make use of a rigid, weightless loading spohe to inputsinusoidal bearing loads.

5. Assume each connection to have the sa , e propcrties as thestructure it connects.

6. Use welding for joint fabrication. Welding offers simplejoint design, lighter weight joints, and high static jointefficiency.

7. Provide access for machining joints after welding toeliminate stress risers.

8. Locate truss members so that their line of action passesthrough the centroid of the ring cross section or trussmember it joins.

9. Initially size structure for crash loads.

10. Use syimnetrical cross sections for truss members. Theyare easier to code and use in the analysis.

11. Minimize the number of different size sections for thetruss members to simplify manufacture and analysis.

12. Use constant cross sections for reaction rings to simplifymanufacture and analysis.

13. Use of a redundant structure computer program, such asNASTRAN, is mandatory.

168

4J dP

00

E-H 00 04

4.3 U 0 000 0: C

o 0H0 0

I4

14 0 A

0 It0

000H9 0Nr

C4.

000 0 N

H H

>1

4J1. WI 0 0 0)0 I 4

H

4

0x 44J

0d 0 0

*4 P-4 0 -

169

AREAS FOR FUTURE STUDY

Areas requiring additional development effort before thefabricated housing concept can become operational include awelding development program and improved analytical methods.The welding development program is necessary to improve jointintegrity through better quality control of welded joints.Improved analytical methods should include determination of skincovering strength effects, a method for plastic and thermalstress analysis, development of a preprocessor to reduce codingof input data, and development of a specific program to handlethe required structural analysis.

A fabricated housing should be built and tested as outlined inthe program to verify cost analysis, fabrication techniques, andweight savings. In conjunction with an endurance test program,a thermal mapping study should be made to determine the heatdissipation characteristics of the fabricated housing design.

To reduce the risk and cost of building a full-scale CII-54Bfabricated main transmission housing, a CII-54B tail gearboxshould be considered. The CH-54B tail gearbox was previouslydescribed in Description of Drive Train.

This housing considered for a truss-like fabricated housingreplacenment has the advantages of simulating all the loadingsexperienced in a main transmission housing and is smaller insize, easier to test, and less costly to fabricate.

The loading on the tail gearbox housing consists of the internalbevel gear reaction load supported in two different housings,tail rotor servo support mounting load to simulate variousaccessories, and tail rotor shaft loading to simulate the mainrotor loading.

Figure 84 is an isometric of a tail gearbox housing in atruss-like configuration. Tht- gearbox consists of threehousings jointed at the ring interfaces for assembly an-disassembly of the gearbox. The bottom housing contains theairframe interface mounting ring and a series of rings tosupport the bevel pinion. The center housing is a truncatedconical structure that supports the bevel gear and tail rotorshaft and reacts tail rotor thrust loads. The third housing isa cylindrical structure that supports the outboard end of thetail rotor shaft.

A tail gearbox fabricated truss-like housing offers potentiallya lighter weight design, simulates the loading of a main trans-mission, is smaller in size, can be tested in the tail-takeofftest loop shown in Figure 80, and is less costly to fabricate.

170

Figure 84. Isometric of a Tail Gearbox Housing as aTruss-Like Structure.

171

SU4HARY

A fabricated truss-like transmission housing was designed whichcould replace the conventional cast magnesium main housing onthe U, S. Army CII-54B helicopter.

The fabricated housing replaces the upper main casting and lowersupport forged housing. This design enables use of the presentinput housing assemblies, rear cover assembly, all internalgearing and bearings, and the swashplate guide. Design criteriafor flight and crash conditions, including loads and stiffnessrequirements, were developed to allow comparison with theexisting casting design.

A material and joining method study, which included sixteenmaterials with various joining processes, indicated thatwelded Custom 455 stainless steel should be used for thefabricated truss-like housing. Thin sheet covers of thesame material, mechanically fastened to the truss, with gasketmaterial for sealing, are used to retain oil and have only anegligible weight penalty.

A number of fundamentally different truss housing geometrieswith different materials were coded, checked through the useof a CRT, and analyzed in NASTRAN. A final design configurationevolved, which consisted of many structural members joining therings of a welded stainless steel truss. The weight of thefinal design configuration was 305 pounds as compared to 363pounds for magnesium casting.

A +25 percent change in input torque results in a trussstFucture weight in stainless steel of 362 and 254 pounds.

Before a fabricated housing can become operational, theanalysis must be correlated with experimental test data toverify the analytical model. A test plan is included fortesting a CH-54B housing design.

To reduce the risk and cost of building a fabricated truss-likemain transmission housing, preliminary design and testing shouldbe conducted on a CII-54B tail gearbox. This gearbox has loadingcharacteristics similar to the main housing but on a muchsmaller scale.

172

CONCLUS IO1S

The fabricated truss-like housing is feasible using today'smaterials, analytical methods, and fabrication techniques andrepresents low technical risk. A fabricated housing should bebuilt to verify cost, analysis, fabrication method, and weightsavings.

The selected material for the fabricated housing is Custom 455stainless steel for both the structural members and the non-structural skin covering.

The fabricated truss-like housing is 15 percent lighter than thepresent cast magnesium housing.

In quantities up to ten housings, the fabricated housing unitcost is approximately 30 percent less than the cast magnesium

Design and manufacture of a fabricated housing is expected tooffer a shorter lead time of approximately thre- months.

'['he fabricated housing was designed to operate at gearboxtemperatures up to 600 0 F, and its fatigue, ultimate, yield andcreep properties are superior to those of the present castmagnesium housing.

A large number of redundant structural truss members improvesfail-safe characteristics. A simplified, conservative analysisindicates irmproved survivability with ballistic damage andgreater credibility for survival against ballistic threats suchas 23 mm API and JIEI hits.

Verification of the fabricated housing design is necessarybefore it can be considered as flight hardware. Preliminarytesting should be conducted on a fabricated housing designed fora production gearbox such as the CII-54B tail gearbox. Thisgearbox, although smaller than the main gearbox, has similarloading characteristics and complexity. Test results would beindicative of the practicality of a fabricated housing design.

173

LITERATURE CITED

1. Caseria, P. T., ACorn.. WEIGHT AND BALAIICE REPORT, SikorskyAircraft; SER-64316, Sikorsky Aircraft, November 1970.

2. 14ongillo, A. L., FLIGHT LOADS AND CONDITIONS, CII-54B,Sikorsky Aircraft; SER-64442, Sikorsky Aircraft, August1969.

3. Frye, R., STRUCTURAL DESIGN CRITERIA REPORT, MODFL CH-54BHELICOPTER, Sikorsky Aircraft; SER-6238G, SikorskyAircraft, August 1969.

4. Stiegler, D., THE SIALLOW CONICAL SHELL, SikorskyAircraft; SER-50361, Sikorsky Aircraft, November 1963.

5. Sanders, W. W., FATIGUE BEHAVIOR OF ALUMINUM ALLOY WELD-kZNTS, Engineering Research Institute, Iowa State Univer-sity, Ames Iowa, No. ISU-AMES-7.031, October 1971.

6. Jones R. E., FRACTURE TOUGHNESS AND FATIGUE CRACK GROWTHPROPERTIES OF 7175-T736 ALUMINUM ALLOY FORGING AT SEVERALTEMPERATURES, AFML-TR-72-1, February 1972.

7. Peshak, C. M., THE INFLUENCE OF WELD DEFECTS ON PERFOR-AINCE, Welding Research Supplement (to the Welding

Journal), February 1969, pp 45S-56S.

8. Lobb, A. E., et al., TITANIUM ALLOY WELDING, SST TechnologyFollow-on Program - Phase I, The Boeing Company, CommercialAirplane Group, P. 0. Box 3707, Seattle, Washington,D6-6029, July 1972.

9. Anon., ARMCO 17-4P1 STAINLESS STEEL BAR AND WIRE, ArmcoSteel Corp., Middletown, Ohio. 1968.

10. Dunsley, J. A., and Walker, A. C., THE FATIGUE STRENGTH OFBUTT-WELDED AND PLAIN SPECIMENS OF 18 (250) MARAGING STEELPLATE, National Research Council of Canada, 1S-122, May1969.

11. Anon., CARPENTER CUSTOM 455 HIGH-STRENGTH STAINLESS STEEL,Carpenter Technology, 1971.

12. Degnan, W. G., Sikorsky Aircraft Engineering MaterialsManual, Sikorsky Aircraft.

13. SD-24 Vol. II General Spec for Design and Construction ofAircraft Weapon Systems (Rotor Aircraft).

NOTE: If references are not expressly given, the data sourcewas one of the standard handbooks, such as MIL-HDBK-5b orAerospace Structural Metals Handbook,

174

APPENDIX I

CH-54B DI.4EtSIONS AND PERFOIANCE DATA

The aircraft chosen for study in this program is the U. S. ArmyCII-54B TARHE helicopter. Overall dimensions are contained inTable XVIII.

TABLE XVIII. GENERAL DIMENSIONS, CH-54B

Overall Length 88 ft 6.0 in.

Overall Width 21 ft 1G.0 in.

Overall Height 25 ft 4.0 in.

Weight (Empty) 19,700 lb

The CH-54 is powered by twin JFTD 12A-5A Pratt & Whitneyengines with a takeoff rating of 4,800 horsepower. Theseengines feature isochronous steady state gove "ning, transientrotor droop of less than 4%, and rapid recovery to low power/high-rate descents (see Figure 85). Engine output power ismonitored within 2% accuracy by phase displacement torque-meters. With their low vibration environment, these engineshave a TBO of approximately 1,000 hours. The main transmissionof the CH-54B is rated at 7,900 horsepower for 30 minutes.

The main and tail rotors dimensions of the CII-54B arecontained in Table XIX.

The tail rotor incorporates a viscous damped tail drive shaftsupport which reduces high frequency vibrations, increases mis-alignment allowances and reduces stress on contingent supportstructures. It also uses self-centering pitch control linkswhich contain low-friction Teflon bearings. In addition, bothtail and main rotors have completely independent, self-con-tained, continuous lubrication systems.

The fuselage and landing gear (see Table XX) have been specifi-cally designed for the CH-54B's cargo handling duties.

Minimum fuselage bulk provides maximum visibility of cargo andcargo handling plus a large open space for the cargo itself.The wide-stance gantry type landing gear enables the aircraftto straddle close-in loads. The cargo handling system in singleand four-point configurations features a 25,000-pound poweredwinch with 100 feet of usable cable, a 25,000-pound four-pointsuspension system for pods or pallets, normal operation control

175

EQUIVALENT MAX POWER0KNT

6.ENGINE INOPERATIVE

GROSS~10 KNOTS 100 b

Figr 5 owr-lgtPromne e ee

Standard. M-OWE

'176E

a.) 4) 01 a) Q)

4-4 4 44-4 44.) 4J m1 4.)

W )0l04)0) U)U) U) 0 r (n)U ) w MU) (n r.M M W 1 $4 w ) 0W ) Q)0 4) 0)Q U 0) a)

w go 000w9 wozz~40 0U0.0 w4M) In UW M U) 4-4 *ri *ri44 .4. r4 134- 0)J4P I 10 -rl -rl~ - C.,1 04

Hn %o vA. NH H- 1- H0 0 4 4-) V 0

ON '00 0 0)0 Q 4.) 0*NH 0 C) 00 ) La) Q) cu Q) C

r- . *o 00* ** U -cUH 4-4 *U t-OND N N- 0 C4 0 C 0 W o L tA I-, N o *e4 o.0qrm')N " '4 CNI

m we r)N m v r- N N rH -H r- Z -iH H Z0 t-!,'H-4'D. aL ZH r-

LA -

0C.. 0

U)

O 4-) rH 4-4 4)

U) U0

H- u (n rd

o) 40 4E4 - j ON) Ho0rl 4o 4.) 0 43 -H)

04 idU) M 4J rd

so r. 0.U 0* U) Vo 0~ 4J A 4)

k 0 4 4 U > - 0 0 4)H OO 54 0 0 *-H *- $4 ly, U) 0 0

~~ 0 ..4 4 4) (v w 0644J 00C. U E-i Q0 aV 54k

N 40) 41 Q)~ 0 .4( 4 0

() 44H CA 4J'

4)U)m 00) o 0 04i OH. rl.-40r -W 5.1 0 (*r4 U) H. 04 4J-M5.0 ''d0 r

0 - 00 X 0 ~ 4J M .).rH -s0 54J ci 4J4 4.. 5.. 4) U0 0 Q 0 0000-. 04 00 .rig 0 : 04.49ZMr-r4

W0 0Q- *- Z 44 0%a r- 0~ P ' 0H M:3P00 C r 00 E- 4E4 'Wi 'E-4 4 a C4 co d LI))=

4410 4)4) 0 4J - V 5-A W 0~4.) 4J OP 0

4) r - ow r4) 4- ) 0jrI d 4 4 EN-

M 0 (0 p) r4* 4)d M r (dU)m U) rIc ) 0 i050.4 (

0 0 00 0 0 C 0 4 0%H4 0 0 M X M 0 P4

4JV ' ' ) 1 Jtr 4Jk V4J0 M43(V0177ri0

0

00

4. 03 tv

cn 0

4.)~4C X: 4 >I4J Hj 000

nr-4 H LnU n '.0 '.0 H ( 40 4 14 HnWN r-C .4-4

N1 Nn teI1

43l 43 >03 00 44

H~ 0LL (dL 0( .

N MH H) 9H 00, )r 11( - 43 .

: k 0u 4--.

>1~~t 414p4 a3r41 $ r- 0 4 -

03 0 :j )

C) 4.3 Nnr4 4 rA r -

-rf "T '-H >.r *r( '4r 4 Hto %D >i 4J M-i - $40 -A r4 0

14 0P-4 440 03 l Q -0'0 0 4J (1)0 0 0)3 A Q

*4a) H 4 U4 4 >*dj 4) H 41444. :. 4-:0 * 4) 9: 0H 0r

4J 4 -i 0) d -I 4) ( r-4 -

OH4M * 44 . 4ia)0 (d 03rU)' l I g03 0 4) r 03 q $4 0.. i( 0

r:: 43-I HS1r- 04 (d V r .-i rq 03ri 0 -4 i- (0( r)r 4 44 St. 09 Hl >U

>lr 1 td 434 t 0 W 0033 . 4 V4 a03-r ( 04 r. 0 0 04E3-4 u: (a 43I i >

0. - r' HH N N P 0 $40 U (D rIp kQ 0 HQWn ud Iv -A0 03(Ui pr- 4-p 4 Q) Q)a) a)19 It r

0)3)a 400 4 1 0d443 0 0 lo 0)> )0 r 4 0 ro 1 C r 0 £a ) H(n Q)i. H *d ()

P *dri 4 u( -d -l u 0 :j 00)E- H4 Q) 0 4043 4 -H 003(d 0- 4J 0

(n ) 44U r -A. ri - i 4-) 0 4-H 4J CU) r-i W

(-I ra0 0.1.%- ri 4.id0O 000 0 303 *r44 043 N00 X0U .74N NN0 >~0 >N >W > u.40

in W . a)U4r a) ~ Cr-N3 33 03j ~HH413 N H 4-I 4J03 r 4303330D 0.0 '0) 0...4.- 0 0i VU

>033.4 ~ ~ ~ 0' fu $4 3 0 0 .>U~~nU30 4N .3Nl

I~~.r N ~ ino d 178~

from all pilot stations, and multiple hard points with universalattachments for any type of load. Added precision and safety incargo handling are insured by the rear-facing operator position,which has full authority in all control axes and positivecontrol priority over the forward-facing operator position.

These design features all contribute to the overall performanceof the aircraft. With a large useful load to gross weightratio, the CH-54B has the ability to pickup, transport, andplace heavy external loads with precision. Although the emptyweight of this aircraft is 19,700 lb, it is certified to operateat sea level with a gross weight of 47,000 lb to a temperatureof 118F. On a standard day (temperature 58.7 0F), this grossweight drops only slightly to 43,000 lb (see Figure 86).Mission productivities, up to 480 Ton-Knots at 40 nauticalmiles radius (see Figure 87), are calculated assiming:

(1) Warm-up and takeoff - 5 min at transmission rating

(2) Cruise outbound with max paylad at 80 knots

(3) Hover out of ground effect and drop payload - 2 min

(4) Cruise inbound at 130 knots

(5) Land with 10% reserve fuel

For missions that include a ferry leg, climb capability exceeds3,500 ft/min at sea level standard (see Figure 88).

179

12 STD___ TEMP__10-

00

E-4 5

4-

2'

024 2323 O48

GROSS WEIGHT (1000 ib)

Figure 86. Hover Ceiling, Two-Engine CH1-543.

180

500-SELEL

00

S300

00

0 10 20 3 050

MISSION RADIUS (nrni)

Figure 87. Productivity Versus Ilission Radius,CII-5 4B,

181

12STD TEe !95°F

10 . _

8 _ _

47,000 Lb

4, Io V

214,000 Lb

2 I!

I

00 1000 2000 3000 4000 5000

RATE OF CLIMB (ft/min)

'Figure 88. Forward Climb Performance, C1I-54B.

182

APPENDIX II

MATERIAL CHARACTERISTICS

The material properties for the sixteen candidate materialsconsidered in the material and joining method study are includedin the following pages. Included are static, fatigue, and creepcharacteristics of the following materials:

Magnesium AlloysAZ9lCZE41A

Aluminum Alloys6061-T67175-T736

Titanium AlloysTi-6A1-4V Alpha-Beta ProcessedTi-6A1-4V Beta Processed(Beta Forging, Beta-STOA, and Beta Extrusion)Ti-8Mo-8V-2Fe-3A1

Steel Alloys17-4Ph16% Ni 250 Grade Maraging4130/4340Carpenter Custom 455

CompositeGlass "E" TypeGlass "S" TypeGraphite High ModulusGraphite High StrengthBorsical

183

ALLOY: AZ91C

General Description

AZ91C is a heat-treatable magnesium-aluminum-zinc alloy mostoften used for sand and permanent mold castings. The alloycombines good ductility with good strength up to 3000F and goodpressure tightness with good welding properties (important inthe weld repair of castings).

Physical Properties

Density .0652 lb/in. 3

Melting Temperature Range 7850 - 11050F

Coefficient of Thermal Expansion 14.5 x 10-6 in./in./OF

Specific Heat .25 Btu/lb°F

Thermal Conductivity 32.4 Btu/ft/hr ft20F

Fabricability

Castings

Castings require generous fillets and uniformity of wallthickness. Minimum wall thickness on sand and permanentmold castings is .125 inch. Casting temperature range is11500 to 12500F.

Welding

Welding of AZ91C is done in accordance with an in-housestandard. Weld properties after heat treatment to -T6 areas good as the parent material. Castings are welded in the-T4 temper and then re-solution heat treated and aged tothe -T6 temper. Buildup of AZ91C components by weldingis acceptable as long as heat treatment after welding canbe accomplished.

Brazing and Soldering

Brazing and soldering of magnesium alloys is not recom-mended.

Adhesive Bonding

Adhesive bonding of structural components is not usuallydone, but can be carried out in accordance with an in-house standard.

184

Mechanical Fastening

Care must be taken in mechanical fastening to avoidbridging and galvanic corrosion.

Corrosion Resistance

This alloy is slightly inferior in general corrosionresistance to wrought alloys, and to casting alloys notalloyed with zinc or aluminum. Care should be taken toavoid galvanic corrosion. Anodic Dow 17 or equivalenttreatment is strongly recommended.

TABLE XXI. MECHANICAL PROPERTIES OF AZ91C-T6

Specified LocationsClass Class Class Unspecified

Property 1 2 3 Locations

Ftu, ksi 35 29 27 17

Fty, ksi 18 16 14 12

Fcy, ksi 18 16 14 12

Fsu, ksi - -

Fbru, ksi ...

e percent 4 3 2 .75in 2 inches

E, 106 psi - 6.5 -

6i

Ec, 106 psi - 6.5 -

G, 106 psi 2.4 -

Poisson's Ratio - .35 -

185

ALLOY: ZE41A

General Description

ZE41A is a heat-treatable magnesium-zinc-zirconium rare earth(cerium) alloy used for sand castings. This alloy combines goodcastability and weldability, high strength at ambient andmderate elevated temperatures, and pressure tightness. Thealloy is considered to be as castable as AZ91C and, combinedwith higher mechanical properties, better creep resistance, andgood weldability, may be considered as a replacement for AZ91C.

Physical Properties

Density .066 lb/in.3

Melting Temperature Range 9900 - 11800F

Coefficient of Thermal Expansion 15.1 x 10-6 in./in./°F

Specific Heat .230 Btu/lb/°F

Thermal Conductivity 61.8 Btu ft/hr ft2*F

Fabricability

Castings

Castings are designed much the same as for AZ91C. Tendencyto microporosity is low, and the alloy solidifies in fineequiaxed grains. Tendency toward alloy constituentsegregation is low.

Welding

Welding is currently done in the as-cast condition, only inareas demonstrated free of microporosity, and the weldedassembly is aged to the -T5 temper (no intermediate stressrelief required). Welded ZE41A-T5 promises to have thesame properties as the base metal both statically and infatigue. It is easier to weld than AZ9lC-T6 and shows nodecrease in properties when the ground weld is compared tothe as-cast surface.


Brazing and soldering of magnesium alloys is not recom-mended.

186

Adhesive Bonding

Ahesive bonding of structural components is not usuallydone, but can be carried out in accordance with an in-house standard.


Mechanical fastening requires the usual care with magnesiumalloys to avoid bridging and galvanic corrosion.


No specific data is available in house, but reports from

England and the Continent indicate that ZE41A has the sameresistance to corrosion as AZ91C Anodic treatment with Dow17, and a good paint system is strongly recommended.

TABLE XXII. MECHANICAL PROPERTIES OF ZE41A-T5

Specified Locations Unspecified Locations

50-Lb 50-Lb 50-Lb 50-LbCasting Casting Casting Casting

Property Weight Weight Weight Weight

Ftu, ksi 26 26 25.5 23.5

Fty, ksi 17.5 17 17 16.5

Fcy, ksi 17.5 17 17 16.5

Fsuksi 17.5 17.5 17 15.5

Fbru, ksi 56 56 55 50

Fbry , ksi 41 40 40 39

e percent 2.0 2.0 1.0 1.0in 2 inches

E, 106 psi - 6.5 - -

Ec, 106 psi - 6.5 - -

G, 106 psi - 2.4 - -

Poisson's Ratio - .35 - -

187

10

0

P4 4C-

0 100 200 300 400 500 600 700 800

TEMPERATURE O0F)

Figure 89. Effect of Temperature on the Tensile andCompressive Moduli (B~ and Ec) of CastAZ9lC-T6.

100 - - - - - - - - --

0 15

0.0

0 100 200 300 400 500 600 700 800TEPRTR (0 F)A

Figure 90. Effect of Temperature on the PhysicalProperties of Cast AZ91C-T6.

188

10 H

H 10 Hr

4-

4.)

0 100 200 300 4oo 500

TEWEaRATUE (*F)

Figure 91. Dimensional Changes of AZ91C-T4 SandCastings at 200* to 400*F.

189

K f .

3.0' 1

2.5-- - -

1.

102 10 10 106 1.01N A - Cast Surface

s

Figure 92. Effect of Cycles on_ K, S14 = ,+ 10KSI, KT = 3.3, AZ91C-Tw magnesiumAlloy.

190

*12 0 - - 1 0 4 -

15Q'

0 5 10 1 5 20 25 30 35

STEADY STRESS (KSI)

IA

Figure 93. Constant-Life Fatigue Diagram, AZ9lC-T6Magnesium Alloy, KT =1.0.

191

10,000- Gross Stress A (PSI)11 I I54OO*±2030

III 5100*1,000- ±01

8100*±2030

100.0.-

0

10.0.

0

-1.0

0.1

.01 -

0 12 3 5 7 89,10 1112STRESS INTENSITY FACTOR

DELTA K (KSI i. )

Figure 94. Crack Growth Rate Versus Stress IntensityFactor, AZ91C-T6 With As-Cast Surfacein Laboratory Environment.

192

ALLOY: 6061-T6

General Description

6061 is the most versatile of the wrought, heat-treatablealuminum alloys. It is available in all wrought forms. How-ever, this aluminum-magnesium-silicon alloy is not as strong asmost other heat-treatable aluminum alloys. Good corrosionresistance is present and not significantly affected byvariations in heat treatment. This alloy possesses excellentformability in the annealed and solution treated conditions, andis readily welded by all methods. It is the only heat-treatablealloy commonly fusion welded.

Physical Properties


Melting Temperature Range 1080° - 1200°F

Coefficient of Thermal Expansion 13 x 10-6 in./in./OF

Specific Heat .23 Btu/lb/°F

Thermal Conductivity 96 Btu ft/hr ft2 °F

Fabricability

Forming

This alloy has excellent formability in the annealedcondition. It can also be formed in the T4 condition andaged to the T6 condition.

Welding

Welding of 6061 may be accomplished by TIG, MIG and spotwelding with two limitations: In T6, heavy sections havea tendency to crack, and in the annealed condition, spotwelds present some production problems. If the configura-tion allows, reheat treating restores parent metal staticproperties; if not reheat treated, 6061-T6 welded hasapproximately the strength of the T4 temper. Fusionwelding is covered in MIL-W-8604 and spot welds perMIL-W-6858.

Fatigue strength of welds has been well documented and aconstant-life fatigue diagram is included as Figure 95.Figure 96 is a constant-life fatigue diagram at 200 0F. Asynergistic interaction between creep and fatigue thatoccurs near aging temperatures is discussed in Appendix III.

193

R 0.5/

/R0.

R 0.50

10 /

/10 6 R. -0

00 10 20 30 4o Ft

STEADY STRESS (KSI.)

Figure 95. Constant-Life FatigueDiagram For Weldedand Heat-Treated 6061-T6 (Reference 5) .

194

Synergistic Interaction ofCreep Fatigue Not Defined

M 0

0 10 20 30 4ortu

STEADY STRESS (KsI)

Figure 96. Constant-Life Fatigue Diagram For 6061-T6Wrought Products at 2000F.

i95

Brazing

The alloy 6061 has excellent brazing characteristics withall methods, resulting in annealed properties of theassemblies.


This alloy is readily joined with most conventionalfasteners. Attention should be directed to prevention ofgalvanic corrosion.


The corrosion and stress corrosion resistance of the -T6 temperis the best of the heat-treatable alloys. In general, resis-tance to corrosion is not significantly affected by variationsin heat treatment.

1

196 :

TABLE XXIII. 14ECHANICAL PROPER.TIES OF 6061-T6

Extruded Welded JointsBar & Rod Shapes T6 As Welded

Property QQ-A-225/6 QQ-A-200/8 Welded HT to T6

Ftu, ksi 42 36 22 40

Fty, ksi 35 33 15 34

Fcy, ksi 34 34 - -

Fsu, ksi 27 24 19 26

Fbru, ksi

e/D = 1.5 67 61 - -

e/D = 2.0 88 80 - -

Fbry, ksi

e/D = 1.5 49 49 - -

e/D = 2.0 56 56 - -

e percent 10 - - -

E, 106 psi - 9.9 - -

Ec, 106 psi - 10.1 - -

G, 106 psi - 3.8 - -

Poisson's Ratio - .32 - -

197

I

1.7 ---

1.6 -

1.5 - -\

1.3. --

1.2

1.0. - - -

0.9. - - - -

o.8. - - - - -

0.7 - - - - -

0.6-

k0 10 20 30 4o 50 60 70 80 90

Figure 97. Bending Modulus of Rupture For 6061-T6Round Tubing.

H4

20,

0o1

0 10 20 30 4o 50STEADY STRESS (KSI)

Figure 98. Constant-Life Fatigue Diagram For 6061-T6Wrought Products, Kt = 1.0.

198

4o --

30

11910 10

10,00a

10. -r

'1 0.0k

.04

0 2 4 6 8 10 12 1)4 16 18 20 22

STRESS INTENSITY AK (KSI in.

Figure 100. Crack Growth Rate Versus Stress IntensityFactor, 6061-T6 Extrusions.

200

ALLOY: 7175-T736

General Description

7175, a heat-treatable aluminum alloy, is a special, high-purity, premium-strength forging grade of 7075. 7175 has achemistry within the specification limits of 7075 and has manyof the same characteristics. 7175 has been applied only toforgings. Although elevated temperature properties have notbeen defined, there is no reason to expect different percentagereductions than 7075.

The elevated temperature strength, however, is inferior to thatof 2024. It is available in all wrought forms and is readilyformable in the annealed and unstable solution treated tempers.This alloy can be resistance welded, while fusion welding ofthis alloy is avoided. Its corrosion resistance is slightlysuperior to 2024 and may be further improved by cladding. 7075is subject to stress corrosion cracking in the -T6 temper,particularly in the short transverse direction in thick sections(forgings), and has largely been replaced by the -T73 temperfor forgings.

Physical Properties


Melting Temperature Range 8900 - 1130OF

Coefficient of Thermal Expansion (Reference 12)

Specific Heat (Reference 12)

Thermal Conductivity 70 Btu ft/hr ft2 oF

Fabricability

Forming

The annealed condition has forming properties inferiorto those of 2024-0. In the unstable solution treatedcondition, 7075-W possesses the same formability as 2024-W.For maximum formability, the total time at room temperaturebetween quenching and forming should not exceed 30 minutes.The T6 condition has limited formability at room tempera-ture,

Welding

This alloy is not normally fusion welded. In resistancewelding, the performance of this alloy is inferior to 2024.For the transmission structure this alloy is not weldable,

201

and hence mechanical fastening must be considered.


Brazing and soldering of this alloy is not recommended.


This alloy is readily joined with most conventionalfasteners. Attention should be directed to preventinggalvanic corrosion.


General corrosion resistance is similar to that of 2024.Resistance to stress corrosion cracking of 7075-T6 type tempersis only moderate, as with the 2024-T4 type tempers. 7075-T73has a high resistance to stress corrosion cracking.

202

TABLE XXIV. ECHANICAL PROPERTIES

OF 7 0 75-T77366Property

Vlue

Fty? ksi

L 76

ST 71

Fty j ksi

L 66

ST 62

L 66

ST 75

Fsu, ksi 42

Fbru, ksi

e/D = 1. 5 110

e/D = 2. 0 141

Fbry, ksi

e/D = 1. 5 89e/D =2. 0 99

e.. percent

L 7

ST 4

E, 106 psi 10.3

Ec, 106 psi 10.5

G, 106 psi 3.9

Poso~sRatio .33

203

0

A

______ 0

/r /1

Lrv%

iOi

H XU

C') 4 0

4 4 H

'.r4 H

-0c'0

0/ usa

0 0 0

204

4.o Kt 3.0

3.0-

Kf

2.0'

1.01_ _ _ _4.i~ 05 l6 7~ 8~

CYCLES

Figure 102. Effect of Cycles on KT, R = 0.00, 7175-T736 Aluminum Alloy.

20

205

(sanoH) aKII anSoJlXa

o o 00 0 0o H H H H 0

0 0 0

0 r0

0C41

4.

0 1:4 1nH>

E-1 .4 H4

4-4 UH

No,,

20

(sxnoH) aMLL amfsodxa

0 0o 0 0 0 Hi

C 0 0 i '-0o H Hi H H4 0;

\00004 H

4J~f

(04

140E-4 r&i

0.0

'4414

440

* 0

.r4

207

60

50 - _ _ _ _ _ _

CREP/FATIGUE

_____ ___ SYNERGISTIC INTERACTION

H0 UNKNOWN IN THIS REGION

30~

10

~20

10

0~0 10 20 30 4o 5o 6o 70

STEADY STRESS (KsI)

Figure 105. Constant-Life Fatigue Diagram for 7075-T736Wrought Products at 200 0F.

208

TABLE XXV. FRACTURE TOUGHNESS PROPERTIES OF7175-T736 ALUMINUM ALLOY FORGING(REFERENCE 6)

Temperature

Room Transverse (WL) 22.723.4*23.8*

Room Short Transverse (TW) 33.129.133.5

31.9 Avg.

2000F Short Transverse 33.035.034.4

34.1 Avg.

0F Short Transverse 26.427.126.3

26.6 Avg.

-650F Short Transverse (TW) 26.026.726.1

26.3 Avg.

* Invalid test; crack length does not meet ASTMstandards.

209

$4HNe

> . N

02P4U

W) 0 ~ N)o, 0

>40

U)

0 r2N.4 0 ~C4 a $4H0 CO N Nn C N 0 t: k

04.- E- 4 on ko %

$4U V LA ) %a IA 1H~> r- CO f- MA 0. '.0;

E-

0111~r *0 * 0 0 f- k9 HH02 I o, L 0 C I; 0 I

E-4)

0r4 02-. LA4 0H

0 0

1 '49 0' 0

$0 9- N0 C N N H- H-x x x I I

N1 r-4 H m

$44 0R

$4 $4.$4I t))

wH 0

N N

EU N N r-4 P-4o

0E4 >1 4)L A A LH0 0n 0n 0 0 0 H)

'~4 N N- N- t- r- -

210

ALLOY: Ti-6A1-4V Beta Processed (Beta-Forging, Beta-STOA &

Beta Extrusion)

General Description

Ti-6A1-4V is a heat-treatable sheet and bar alloy possessingexcellent elev ted temperature strength up to 750*F, good notchtoughness, and fair to good weldability. This is normally usedin the annealed condition.

Physical Properties


Melting Temperature Range 30000 - 3040 0F

Coefficient of Thermal Expansion (Reference 12)

Specific Heat (Reference 12)

Thermal Conductivity (Reference 12)

Fabricability

Forming

Limited forming is possible at room temperature, withheating to 4000 - 120 0°F required for many formingoperations. It is readily creep formed.

Machining

Machining characteristics are similar to those foraustenitic stainless steels. Required are sharp tools,rigid setups, heavy feeds, slow speeds, no dwell incutting, and soluble oil coolant in quantity. Sawingis best performed with high-speed friction saws in therange of 4000 to 4500 ft/min. Grinding can be performed asa finishing operation in accordance with an in-housedeveloped procedure. Unlike heat-treated steels, there isno warpage associated with deep machining cuts.

Chem Milling

Chem milling has been applied to titanium on a limitedbasis to remove alpha case and to achieve specificdimensions.

211

Welding

Titanium may be fusion welded under an inert gas atmos-phere and then must be treated with proper post-weldannealing to obtain high static joint efficiencies. Spotwelding is done in accordance with MIL-W-6858. Whileelectron-beam welding can be done, it is usually notapplied to the generic structures made from the alloy.


Brazing and soldering are not recommended, but it is

possible to braze with proper protection. Latest develop-ments in low-temperature brazes appear to give adequatestrengths with minimum erosion.

Adhesive Bonding

Good bonding characteristics are obtained from all titaniumbonds when attention is paid to surface preparation anddesign technique.


Titanium alloys are readily joined with conventionalfasteners, including both aluminum and Monel.


Corrosion resistance of titanium alloys is excellent. Thehigh resistance of this material to attack by harsh chemicalshas led to broad applications in the chemical industry. Mostother metals except Monel and stainless stools will sacrificeto titanium under galvanic attack. The only basic problemwhich can be generated by a titanium application is the galvanicattack of adjacent, less corrosion resistant materials.

Note on Fatigue of Titanium Welds

As a direct result of Sikorsky research on welded metaltitanium main rotor blades, welding technology has and is undercontinuing development to achieve the highest fatigue crackinitiation resistance in welded titanium structures. DothSikorsky research and other research have determined that thesource of fatigue strength reduction in welds is small,

- including the hydrogen-filled voids distributed throughout theweld structure. The source of these voids is water vapor, andthe solution is to keep the water vapor out of the weld area.Water vapor will probably never be. eliminated but can be reducedby 'the methods currently in use and outlined below:

212

1. Avoid the use of filler metal in all welds. Forplasma arc welding, this limits the weld thickness to.50 inch.

2. If filler metal is used, only well finished, vacuum-annealed (degassed) ELI grade rod, protected againstmoisture pickup in the welding process, should beused.

3. Use low-moisture argon blankets, taking care thatmoisture pickup does not occur in transit fromstorage tanks to weld chamber.

4. Use automatic welding equipment to avoid the humanvariable.

5. Establish and control precleaning procedures.

6. Use good machined surfaces on faces to be welded andadjacent areas. Never weld sheared surfaces orsurfaces with flowed metal.

The notch sensitivity of Ti-6Al-4V weldments has been documentedin References 7 and 8. Figure 106 is a plot of Kf vs. Kt forwelded titanium. The notch sensitivity data shown in based onroom temperature data. It can be expected that at 600OF thereduction in notch sensitivity will be 73% at 104 cycles and

53% at 107 cycles. For material in the 200*F operatingenyironment, the reductiqn in notch sensitivity will be 9% at10 cycles and 12% at 10' cycles.

For the stress concentration factor of voids in weld, Figure 107from Reference 7 gives the theoretical relationship betweenvoid parameters and stress concentration factor. Only the ratioof void radius to depth below the surface controls the magnitudeof the stress concentration factor. For the problems of voidsclose to the surface, the solution is to shot peen the weld.With this approach, a stress concentration of 2.6 would allowanything up to a .008-inch-diameter void .015 inch below thesurface.

213

Legend

F = 0 KSI _ _ _

S

F S= 20 KSI__ _

4.0-

6,7$ 6

2.05

1.00

2214

5 -A

Kt

0 .1 23 .54

r/h

Figure 107. Kt as a Function of Defect Geometry.

(

215

N %0O0 0 v H m~ H

0o Ln tn V 0 %D 0 0 0 0

0 0C IV fiA 0 H D H~ H .r4~ ~ ~ ~ - Nf H~ N ~ f-

'H~~~I HH N H.

E-4 C . . . I IU3) LI IA It 0 N 1AU)n 0~ 0w Ch w.0 %0'. 0 0

m N M) 00 0u LH ODHI HP4~ H Hl N ( H N

2 I

0

0 0 10 H M' H0 0 to tor 0C, , 0 0

N 0 m m. '0 Hr tD 00 N. '.D '.0 0HN H1 NN HN

E4 H

£H

P4 N '.00 40 0 lw m ~ H

At0 0 '.0 %D0 Hr- 0N t-. %D0 w. %D 0 0

r- r-H.

04 E-

ON~~~ a '0-(h ~ ~~ ~ ~ 0v Ln H v' H n L o

1a Ln wC 0n a% 00 I OLA HA H. . .E4H H ;;4 C4 N 2IA oia H'*C'

HJ H3 C14 HA L,4 40$4 0d N.4) H N H N 4.) 11 r

H4 94 0 H 0 03 *03 fri H H -HQQM 4 03w H p -P r4 (4 N (a N 4 - %P3*4 w. 4. ) u l 2 0) 9 %0

PZ4 W 4 4 P4 414 04

216Al

00

4.,4

o0 t

o 04

/0

-1-0 0

4

0., 0*44l

'0q

(IS31 SMU XHOVc'I

217J

o053

0

NEAR SURFACE

0.1

0 5 10 15 20k ~STRESS INTENSITY RANGE, AK (KSI in. )

Figure 109. Fatigue Crack Propagation Superior inTi-6Al-4V, BETA-STOA •

170

m 150

013 10

4o 50 6o 70 80 goK I (KSI in. )

Figure 10. Fracture Toughness of Beta-Processed

u 0 Titanium Ti-6AS-4V.

218 '

~ iTO

120- -__ _ _

100

100~ ___10

60.

10

0 20 40 6o 80 100 120 1140

STEADY STRESS (1(SI)

Figure 111. Constant-Life Diagram, Ti-6A1-4V AnnealedISheet,. Roam Temperature, .1' = 1. 0 (Ref erence 8).

219

10,00

8,000 t 6,000300 RPM

100> - 20,000 ± 25,000 PSI,

ri/ 20,000 ± 15,000 PSI

20,000 155000 PSI00 M iooRP

2 ,0 00 ±RPM0 ,00

10 0 - (PAIRED IN) 1,800 RPM -

ROOM TEMAPERATURE

CD,

0 10 20 30 40o 50 60 70

AK (KS jn~k

Figure 112. Crack Growth Rate Versus Stress IntensityFactor, Mill Annealed Ti-6Al-4V TitaniumSheet.

220

H HO

ulN 0 0- -

D ~ ~ H P

HOW 00

0.

HZ Id r r- I I III I I I

~H

0E4 k%0 %

H r-4-

0 0 '.

1- 4 ' rIHi r- H N

u U HHL

00

H4U ) H- H -

'.0 0 4

-rlL -r ~ A00

OU4 00 W'.00. U*H 9 I4

0I IC-4 H Ho

4> 0 . 0 A.D U 0h 0

-A4 4)G to -oI

00 000r4(

W4 U -40221i ia

.r4 4J

0 ow

Wri V *.W4)r r r

ONQ~rItrd

lJ 4J 43

rd 0 rO 4J

Gi i U:W 0

a4) E-4wr4

0. OD 0 0

4-) E-4 *44J1

H 4 -W-

HW L0 00E

$4 -I *P4 4j0 r4 A

w >1

WN 4

(44 0 r4(Ua

M~W 4J 0 r

i4-) 0 4J01.

NW 4 44

WN -.-4c

*2r4 9: 4

44 W W4

0UW H0 44Mnwu NOo

222

10o TI-6AIA4V1/2 IN. PLATEANNEALED

8(

.r4 60PLANE TRAIN FCTURE )UGHNESSFATIGU] CRACKEI ROUND S ?EC

600

2C-T0.5

0.354

-600 -400 -200 0 1.00 400

TEMERATURE (0F)

Figure 113. Plane Strain Fracture Toughness ofTi-6A1-4V.

Note: Considerable scatter exists in this parameter both as a functionof testing methods and between-lot variation. Additional data onTi-6A1-4V Forgings (typical data) are as follows:

Conventionally forged and annealed 40 - 50 ksi (in. )Conventionally oiged and heat treated 30 - 45 ksi (in.)Beta forged and annealed 5T - 70 ksi (in.)

223

110-

04

8108 6 - - -

~70-

~6o

t 50

30 - - - - -

20 - - - -

* ~10 -- - . - - -

0 10 20 30 40 50 60 70 80 90 1 00110 120 130 140

STEADY STRESS (KSI)

Figure 114. Constant-Life Diagram of TitaniumTi-6A1-4V BETA-STOA 0.5 Inch From As-Quenched Surface, Polished Specimens.

224

20CISTRENGTH AT TEMPERATUREEXPOSURE UP TO 100 HOURS

S14-- -- - - - - -

S12 0-0-

0

S80-

40

200

0--600 -400 -200 0 200 400 600 800 1000 1200

TEMPERATURE (F)

Figure 115. Effect of Temperature on the UltimateTensile Strength of Annealed Ti-6A1-4VSheet and Bar.

225

200' I I

STRENGTH AT TEMPERATUREEXPOSURE UP TO 100 HOURS

18o- - - - -

160-

S120 --

o 100

- , 6 0 .-

40.

20

0--

-600 -400 -200 0 200 400 600 800 1000 1200

TEMPERATURE (OF)

Figure 116. Effect of Temperature on the CompressiveYield Strength of Annealed Ti-6A1-4V Sheetand Bar.

226

120 - STRENGTH' AT TEMPERATURE

1EXPOSURE UP TO 1/2 HOUR

LONGITUDINAL

S80-

8 60-

S20 - - - - - -

" 4o

S 00 200 4O0 600 800 1000 1200

TEMPERATURE (0F)

Figure 117. Effect of Temperature on the CompressiveYield Strength of Annealed Ti-6A1-4VSheet and Bar.

120 -STRENGTH AT TEMPERATUJRE

1 EXPOSURE UP TO 1/2 HOUR

LONGITUDINAL

S80--

0£ 0,

0O.

S20,

S 00 200 400 600 800 i000 1200

TEMPER;TURE ( F)

Figure 118. Effect of Temperature on the UltimateTensile Shear Strength of Annealed Ti-6Al-4VSheet and Bar.

227

1004

! -

0

E'20 -- - - -

02- 00 140 60 0 800 100

TEMPERATURE ('F)

Figure 119. Effect of Temperature on the UltimateBearing Strength and Bearing 'Yield Strengthof Annealed Ti-6A1-4V Sheet and Bar.

228

10

80-

C, 20

S-600 -4oo -200 0 200 4~0o 600 800 1000

TEMPERATURE (OF)

Figure 120. Effect of Temperature on the Tensile and* Compressive Moduli of Annealed Ti-6A1-4V

Sheet and Bar.

229

12.0 - - - - -

S 10.0-

2 .. - -

-200 0 200 4o0 600 800 1000 1200 11400TEMPERATURE ('F)

Figure 121. Effect of Temperature on the ThermalConductivity of Ti-6A1-4V (Annealed orSolution Ileat-Treated and Aged).

230

0.18 -

o.16 ---

0.12 -

r.)

r. 0.10 -

.o8

-200 0 200 400 600 800 1000 1200 1400

TEMERATURE (OF)

Figure 122. Effect of Temperature on the SpecificHeat of Ti-6A1-4V (Annealed or SolutionIleat-Treated and Aged).

231

6.oNOTE: EXPANSION COEFFICIENT SHOULD BE READ FRO/

ROOM TEMPERATURE TO THE TEMPERATURE INDICATED.

r5.8

0S 5.6-

C..)

5.2./

-200 0 200 oo 6oo 800 1000 1200 I4oo0TEMPERATURE,( F)

Figure 123. Effect of Temperature on the (Mean) LinearCoefficient of Thermal Expansion ofTi-6A1-4V.

232

$4

H 1 lii Io~ Iru r- i

0

U lII II I chco - %4DI ~ -

L1 iflO co II I I

Id E-4I

o r H4 c n(H 4

aN E-4 I qw C Orv)14 a'E4 l i j~~~I N

o 0 0 tr- m vm m 4 # o N( U ~OOM~4EA $4 H- NO~~ 0 ~ - r'H IH

* 00 -ri

rE,) E4H 4.)rd If * YOO~ r-r 0% % CI O-i 0r-4 LL r-N

:D> 4 C4 - m NO~-f m DNH r i Hr4$10 i U -4

P4 ko on4 1 0 0

E-4E-4 H- b n DI q I C)H r-4 ~ 0 (h mI 4.)0 dP EAv ~N C NH 0-

* -4- W)0 ocanu creI ~ 44~H H

* $4 $4W ~Z 4. Q

dpmr-I H0

H 0u H 0 mma'o mmv U) I Pfl' tcoH IE-

VU v WVN 1 -0~ ~N I H-

E-4 0IEE-1

a$o4 HOOO HOOO HOOO 0HiOOO - 00 H O -

E- 0$4

W) 4.)

04 0

a) -

233

TEMPERATUJRE (T) =FAHRENHEIT + 46oEXAMPLE 8000 STRAINI AGED TI-6AL-4v SHEETTEMP = 65 0 F 70QQF PERCENT 800f4STRESS = 70 KSI 675 0 1. oFTIMET~ 60 HOURS 600 FCREEP STRAIN =0.037 62 zUF 0 ' 750 F

0.o 0F 00.65601 *0F

I 0 .

72FI 0.002

~ '1* 0. 001- 650 uF

2 3 5 10 20 50 100 200 500 5 10 20 50 100 200TIME (HR) STRESS (icsi)

Figure 124. Creep Properties of Solution Heat-Treatedand Aged Ti-6A1-4V Sheet.

234

ALLOY: 17-4PH Stainless Steel

General Description

The 17-4PH alloy steel is a martensitic, precipitation hardeningalloy which exhibits good corrosion resistance and ductility andstrength to 600 0F.

Physical Properties


Melting Temperature Range 25600 - 2626°F

Coefficient of Thermal Expansion Figure 125

Specific Heat Figure 125

Thermal Conductivity Figure 125

Fabricability

The alloy 17-4P11 is readily forged, investment cast welded, andbrazed, Machining requires the same precautions as with theaustenitic stainless steels except that work-hardening is not aprobler When heavy sections are to be welded under conditionsof high restraint, prior stress relief obtain&A by four hoursheating at 11000 or 1150°F is advisable to maximiza fracturetoughness properties and to preclude cracking in planes normalto the thickness direction.


The resistance of 17-4PH to stress corrosion cracking inchloride environments has been found to be superior to the alloysteels and the hardenable chromium stainless steels.

2

235 1

TABLE XXX. M~ECHANICAL PROPERTIES OF 17-0P11

STAINLESS STEEL 111025*

Prope.rty Value

Ftu, ksi 155

Fty, ksi 145

Fcyi ksi (145)

Psksi (100)

Fbrur ksi(e/D =1.5) (245)(e/D =2.) (300)

Fbryp ksi(e/D =1.5) (216)(e/D =2.0) (238)

e, in two inches 12

Et 106 psi 28.5

Ec, 106 psi 30.0

G, 106 psi 11.2

Poisson's Ratio .272

H 1900 temper used as basis of properties inparentheses ()but avoided due to stress corrosionproblems with this temper.

236

0 o

0

0

f H 1

02

0

44

'4110 0Q

cli ri H

t UNcn r t I_ _ _ r- _ _ __.-

_am/% ___ *

o237

100STRENGTH AT 'TEMPERATURE

EXPOSURE UP TO 1/2 HR

00

0

S40 -

S20.

04

0 200 40o 600 800 10100 1200 1400 1600

4Figure 12.6. Effect of Temperature an the UltimateTensile Strength (Ftu) of 17-4PHStainless Steel (Dar and Forgings).

100-

STRENGTH AT TEMPERATURE

80.- EXPOSURE UP TO 1/2 HR

60.

P4

4 0-

0'-

0 200 400 600 800 1000 1200 1400 1600

TI"MPERATURE(O')

Figure 127. Effect of Temperature on the Tensile YieldStrength (Ft ) of 17-0P11 StainlessSteel (Bar ,ifd Forgings).

238

100STRENGTH AT TEMPERA-

80 "TURE EXPOSURE UP TO

0 6o0

20

0 200 +00 600 800 1ooo 1200 1OO 1600

TEMPERATURE (OF)

Figure 128. Effect of Temperature on the CompressiveYield Strength (Fcy) of 17-4P11 StainlessSteel (Bar and Forgings).

100100 STRENGTH AT TEMPERA-

TURE EXPOSURE UP TO

80

E*4

. 40 ",

20E-1

S 0, 0 200 4,00 600 800 1000 1200 1400 0

TEMPERATURE (OF)

Figure 129. Effect of Temperature on the UltimateShear Strength (Fsu) of 17-4PH StainlessSteel (Bar and Forgings).

239

STRENGTH AT TEMPERA-80. TURE EXPOSURE UP TO

k0

20,

20 460n 600o 800 1000 1200 1)400 1600

TEMPERATURE (OF)

Figure 130. Effect of Temperature on the Ultimate

Strength (Fbru) of 17-4PH StainlessSteel (Bar and Forgings).

100- STRENGTH AT TEMPERA-rzi TURE EXPOSURE UP TO

80 HR

60-

4o-

20-1

rx 00 200 4o 600 800 1000 1200 14oo 16oo

TEMPERATURE (OF)

Figure 131. Effect of Temperature on the Bearing YieldStrength (Fbry) of 17-4PH StainlessSteel (Bar and Forgings).

240

100

MODULUS AT TEMERATURE

80

0- -

0- 60 o 6o 8o 100 10 4o 1o

TEPRTUE(F

Fzigr 13.Efc ofTmeaueo th Tesl an

Coprssv Mouls(Ead___o 7- 1

24

0

H

'1 -0

0 C)

HH

-P N-

0 to0

H/ U

*r44

I 4 A

Ei H2 c4_

0 0

24.2

ALLOY: 18% Ni 250 Grade Maraging Steel

General Description

The 18% Ni 250 Grade 250 alloy is one of a class of maragingtypes which develops its strength primarily as a result of acomplex precipitation reaction in a low carbon Fe-Ni martensite.Fracture properties of heavy section can be directional and arelowest in the short transverse direction. Corrosion andoxidation resistance are somewhat better than 4340.

Physical Properties


Melting Temperature Range 26000 - 2650°F

Coefficient of Thermal Expansion 5.6 x 10-6 in./in./0 F(700 - 9000F)

Specific Heat .11 Btu/lb/°F (200°F)

Thermal Conductivity 17 Btu ft/hr ft2 OF(200OF)19 Btu ft/hr ft2 OF(600 OF):

Fabricability

Forming

Formability is excellent in the annealed condition, andmachining offers no special problems.

Brazing and SolderingI

Brazing and soldering of 250 Grade Maraging steel is notusually done.

Welding

Welding requires special precautions; the toughness of theweld is generally about one-half of that of the parentmaterial. Welding is normally done in the annealedcondition and is subjected to a full re-annealing afid agingprior to final machining.

Corrosion Resistance 4Corrosion resistance of 250 Grade Maraging steel is superior tothat of 4340. Banding has been experienced in and adjacent towelds as a result of alloy segregation in the welding process.

243

160 F -18 N.-8.5 Co-Mo-Ti-Ale

:-o11/2 To 1 IN.PLATE15000F, 1/2 HR, AC + AGE

Figure 134.PPl NE STRAIN FRACTUR.E TOUGHNESS of Ten

IIeats of Annelaed aid Aged Plate as aFunction of Yield ctrength.

32F -18 N -8.5 Co- b-Ti-A-le BARI 28) KSI

280

H F

~24o c

20 --A0 200 4oo 61)o 800

TEMERATURE ('OF)

Figure 135. Effect of Temperature on CompressiveYield Strength of 280-KSI Bar.

244

34o0- F -18 N -85 Co-Mo-Ti-Al BAR

300-M 1500YF, 30.MIN, AC + 90001', 3 HR

Ftu26o.- - - 300

18& 2200

~~80.

TEMPERAOMR (0 F)

Figure 136. Effect of Test Temperature on TensileProperties of Annealed and Aged Bar.

220-F -18 N -8.5 Co-Mo-Ti-Al

e i180.- - BAR, CVM

280 KSI250 KSI

100 A0 200 W0o 600 800

TEMIPERATURE (OF)

Figure 137. Effect of Test Temperature on She~rUltimate Strength of 250- and 28U-KSIBar.

245

'520'. e 18 N1-8.5 Co-MO-Ti-AlBAR, 280 KSI

r.4o

0o 20 F.o 60 0

240F 18 -8d=.51 C-oiA0.200 460OT INo 80EE

p300EMPATUE (OFF3HR

' EAT AGE 9000FB3 HR.~ HEA B250.3 i, 150 , R

20002002080 20

10 F

"0Cu

00

cm

H 1 104 cm 0

0 004

H4I /~ 14 00000O

___ J44

11n /

4J UN 1

$4 4J

T/1 0 c$4 0)

.11. i $4 $4

MA 0044

H *0 0H

000 A

00

0 00 %0 _-r0HH H

247

3.0,

Kf

2.0._ _ __ _ _

1031 0 ~ 1, 10~ 10

Figure 141. Kf Versus Cycles, Welded 250 GradeMaraging Steel, R = 0.1 (Reference 10).

248

ALLOY: 4130/4340 Steel

General Description

The 4130/4340 alloy steels contain carbon by the last two digitsin hundredths and various alloying elements to improve theirstrength, depth of hardening, toughness or other properties ofinterest. These steels are available in a variety of finishesranging from hot or cold rolled to quenched and tempered. Theyare generally heat treated before use to develop the desiredproperties. They may be carburized and then heat treated toproduce a combination of high surface hardness and good cbretoughness.

Physical Properties


Melting Temperature approximately 2740*F

Coefficient of Thermal Expansion Figure 142

Specific Heat Figure 142

Thermal Conductivity Figure 142

Fabricability

The alloy steels are only slightly more difficult to forge thancarbon steels. For cold forming, alloy steels are usuallyformed in the annealed condition. The machining of alloysteels is generally more difficult than unalloyed steels of thesame carbon content. The low carbon grades are readily weldedor brazed by all techniques. Alloy steels can be welded withoutloss of strength in the heat-affected zone provided that thewelding heat input is carefully controlled. 4340 steel isnormally welded to an in-house standard.


The corrosion properties of AISI alloy steels are comparable tothose of plain carbon steels. Alloy steels containing chromiumor high percentages of silicon have somewhat better oxidationresistance than carbon or other alloy steels.

249

o t.o H- 0 N In LA oo f% CD N f- Ln tAn coI Iro N H- Hq H (N(m N N

04)

0 40

o 0c 00 o r nN mL

%0 C'1 H 'IV N . >

IV 0 9A0 ON 0 0~ 0' H 0I" P1 0LA m' qwJ a% HCO ?H r N N1 H-

Hv r- H- H- H N HNfr-

H"E

H H H HN HH

0 0 mNa

1:4 o t- r- LA mo I N N N HHH H-

43LA E Ll U) Ir' 0 3

41) N Hq0)

HN rHN 4.)

*..4 d-tl -A U)q U)V)1

ob 0 H 0 w 4

o >1 ps > >1 0 % rl.,I 4J 4J 4J) U U) .Q .0 0.4 a) 14~ 4 ~ L rCr

*'- 0

0 0 wru I U) f4

250

0 c

H

0

4H

$4H0 0 4

00 0

0000

40 0

44 W

C))

0)

00 -t 6 \D O00)Oj CM j 0)i

2514

160-STRENGTH AT TE4PERATUREEXPOSURE UP TO 1/2 HR

14o- UNNOTCHEDNOTCHED, K 8.360' V NOTCH

- 0.505 - IN. HOOT DIA.120" 0.027 - IN. NOTCH RADIUS

120 0.714 - IN. SPECIMEN DIA.

.100-I

_80-

60-

40-

20-

-400 0 400 800 1200 1600TEMPERATURE (OF)

Figure 143. Effect of Temperature on the UltimateStrength (Ftu) of AISI Alloy Steels.

252(

3.6oSTRENGTH AT TEMERATUREEXPOSURE UP TO 1/2 HR

1240

0-

S 100--

8 o-

E--

-E-044o 80 120 io

T PERATURE (OF)

Figure 144. Effect of Temperature on the TensileYield Strength (Fty) of AISI AlloySteels.

253

10

E4

600

0

E-1 40. _ _ _ _

0-

20 A 40d 60o 800 100 100 1400 100o

TEMERATURE ('F)

Figure 145. Effect of Temperature on the CompressiveYield Strength (Fcy) of Ileat-Treated AISI

100 AlloySteels.1

8o.

4 o.

S20 -- -- - - -

0.'

0 200 400 600 800 1000 1.200 14oo 16o0TEMPERATURE ('F)

Figure 146. Effect of Temperature on Shear UltimateStrength (Fsu) of Hecat-Treated AISI AlloySteels.

254

100.

S 8o*

60- __ - _

0

~20.

0.40 200 400 600 800 1000 1200 14oo 16oo

TEMPERATURBE (OF)Figure 147. Effect of Temperature on Ultimate

k Bearing Strength (Fbru) of Heat-TreatedAISI Alloy Steels.

Pq~ 100.

* ~ 8o.

4 40

0 0 - -

~ 0

0 200 00 600 800 1000 1200 1400 i6ooTEMiPERATURE (OF)

Figure 148. Effect of Temperature on the Bearing YieldStrength (Pbry) of Heat-Treated AISIAlloy Steels.

255

0 160-,

MODULUS AT TEMPERATURE

120-

100 -

0

80,

60o-

z

40o

20 -

0--4oo 0 hoo 800 1200 i60 o

TEMPERATURE (°F)

Figure 149. Effect of Temperature on Tensile andCompressive Modulus (E and Ec) ofAISI Alloy Steel.

256

200- 0-S EE

150-

C2100-

50 - - - ___ ___

5 1 1 20 2 30

TANGENT MODULUS (103 KSI)

Figure 150. Typical Tensile Tangent-Modulus Curves Aat Room Temperature for Hfeat-TreatedAISI 4340 Alloy Steel.

257

2 t; Mai

0

H

0C',H-

0

Hm 00 10rr',

- 04

0 4- 4

N4.)W 4JC/M

HH

CO 4 4W

r.4J Hto U) /

Lr\ O r -4

0 CY

Lfn01 H

HH

1/ 0m C~j H

(IS31) SSaUHLS XHOILVUqIA

258

50 - -

~J10

01

10 20 30140 50 60 70 0O 90 100 110 120 130 140150

STEADY STRESS (KSI)

r Figure 152. Constant-Life Fatigue Diagram for 4340Steel Weidments, Heat Treated tol80-l1I.i

259

ALLOY: Carpenter Custom 455 (Stainless Steel)

General Description

Carpenter Custom 455 is an improved chemistry 17-4PH which is aprecipitation-hardening, martensitic stainless steel used forparts requiring high strength and good corrosion and oxidationresistance up to 600°F'.

Physical Properties

Density .283 lb/in?

Melting Temperature Approximately 2640°F

Coefficient of Thermal Expansion 5.90 x 10-6 in./in./°F(720 - 200OF)

Specific Hfeat Approximately 0.11Btu/lb/°F

Thermal Conductivity 10.4 Btu ft/hr ft2 OF(to 2120F)

Fabricability

The alloy Carpenter Custom 455 can be readily forged, welded andbrazed. Machining requires the same precautions as the austen-itic stainless steels except that work hardening is not aproblem. The alloy can be worked in the annealed condition,welded, re-annealed and aged to high strengths. Weldedstrengths are good. Static joint efficiency with the jointwelded, annealed and aged is in excess of 95%. Welded fatigueproperties show a 75% to 80% joint efficiency at 107 cycleswith bead ground off. The following heat treatment is recommendedfor maximum toughness in the weld and adjacent area: weld inthe annealed condition, re-anneal at 1525°F for 1 hour, andfan-air cool; then age at 975°F for 4 hours and air cool (coolingrate not sensitive).

20

260

TABLE XXXII. MECHANICAL PROPERTIES OF CARPENTER CUSTOM4 455

Aged Aged and Aged andProperty Annealed 9000F/4 hr 200OF at 600OF

Fty, ksi 140 220 214 187

Fby, .2%,, ksi 115 205 200 172

e%12 10 10 12

R.A. % 50 40 40 50

KIC, ksi - 73 125(in) 1/2

1KISCCp ksi -58 (117)(in)1/2

26

0 4

4

1 4-)0 0 a

I Hnt

/LO 4J 0~-

8r 0)4.)

U j~-

P4.or

r , J

S~~aHIS XUIVo1

2062

COMPOSITES: Glass, Graphite, PRD

The tables and graphs included in the following pages charac-terize boron filament, glass filament and graphite filamentreinforced polymer matrix composites. Included are elasticand strength properties of unidirectional, cross-ply and angle-ply composite, and physical properties of composites, filamentsand resin matrices.

Orthotropic elastic properties for unidirectional compositesfabricated from SP272-3M and Narmco 5505 prepreg boron-epoxytapes are given in Table XXXV.

263

0 i

0 Nrz I CO*\(YHHO 00

Q E-f0op-- 0w 0r

i~ 0 CJ-E- 0 0

0) E-0+

0 1-1 LH\ 04 4.)o' 0 U\

____ o 0

__ __ 0_ _ 0 0 4

Ii41Lrid

c\4 00~ -

H H

2644

TABLE XXXIII. PHYSICAL PROPERTIES OF BORO1N

SP272-

PROPERTIES Units Composite Ref.* Resin'Matrix Ref.* Cornpos

~. Fiber Volume Fraction % 54 Nom. 11 -53 Nor

coefficient of Thermal in/in/0 F - 2l.8xl0-5 M 2.5x1l

Expansion 13.lxlC

Density lb/in3 .075 M .0443 H.074

Specific Gravity -2.07 M 2.03

(water disp)

'Resin Content % by wt. 28.4 C 28.9

Resin Flow 10 Norn. M 14 N or

V olatile Ccte nt %2 M 10 ',aN

BarL;ol Hardness -85 -95 11 85 -9

* Thickness per Ply in. .0051 M .0052

*= experimentally measured values

C = calculated values

265

~XXXIII. PHYSICAL PROPERTIES OF BORON EPOXY, BORON FILAMENT AND RESIN MATRICES

[P272- lNaco 5505 Fib Eef.* Resin M4atrix Ref.* Composite Ref.* Resin Matrix Ref..* Boron

ii-53 liom. M

21.8x10-6 m 2.5x10-6 G 00 IM45~ 6 .x0

13.1x10-6 09010 2.xo627,

.0443 H.074 m .043 H.095

M2.03

C 28.9 C N/A

14 Nom. IM N/A

.110 Max. m N/A

H85 -95 141 N/A

.0052 M4 .004 dia.

'EPOXY, DOR0ON FILAMENT AND RESIN MATRICES

14armco 5505 Fiber

osite Ref.* Resin Matrix Ref. B3oron Ref.*

106 4.5QO 2.7xl10 6 M

m o043 H.095 M

it

C N/A

~orn M N/A

lax, M NI/A

95 FlN/A

052 H.004 dia. Nomn. m

03 .

04. to 4).A 0 ) U

~~4 4) C)C) C)

$4 0 p4'-,4.,4 0. 4.)

.r4 r4 -H P14 -14 A~-4 ($U) U) U) U) U) w) y) H o1

r 1,04 04 m4 C4 4-) 'E- ) U)

H to 0 %0 '.0 %0 to '003 Uto 03L 0 0a0 00 ar 0 '0 :1 Hn H- H H H- 4.) 54 0-r-%

r4 Ha%Id (0 xX X X x x 0S3a

o >0 H H- Do *,.AC) C - 0 H4 0 m' $4 ~4j>4 (a( H 1 (CH H ) 0

4 ) 4. 41

x .0 4.0S0 4 03 r-4O

m 4-4 u 41 -P40 003D 0 (d H-

145.4 $4 w 1 $4 w04o 0 0 1H 4.)3

U)4J $4 4

.EA .,4 .14 .i .V4 M4 41M) U) (A (n U) U 144I 03w0404 04 04 04 4) .14 .0 U)

E- t 4J ) 00z~~.0 4)4)0a:34 4) 1 t

0 r- H H H H- H- HU4 4) V AV4 0)3C) M r. >1 it .14 0

1 xX X XX 0 x>0 'dH 9 ' r 'C) N 0 He 0304 9H H-r- H H- C% H4 N 0 m('1rC Z .,4 rE-4 * * * * . * *00 0 S -4.4 .1(1) N1 ('1 H.1 03 4.44-) 0

IU) 1 ' - U4 140 3'M 0)4U) 0U) 4-4 : r 1-H -A *4J

16( 03'0 4 th

p4 0.0 0(4 0 U04 H-4V 0.0

H>4-4 0 0 -144)1Hd r- 0UH H4 Z r. 4)

X U) 0 0 (d U) o(144x 4-) .4 -14 O4 r 0 $4 %

9 4.) 41 030r,0 ~4) $-vi U0 VAs 4 4) -U4 0 ?.. -14(U (a c4 41 u) (U 0 -toU) H*-4 -44 -14 0) .00 0 4)V4 M) 0 V 4 tnr UH)w

E4 0 ) 00-14 U)4J (dU t )) f- H0444 (U 0$4 V14 k

0 0 (X 0 HO4 0.4) 0-1j0U (D 00 )~ 04) 4JU E 4"N

to r-4 HO M'0 N It if, I II f(U H ('10 H04 m'

11 It 0 t 11 1

H- N M' N m' e-I lu N-14 m' 1 ' 'H- N M'1 H N N4 H N A'.1 .14


------ ---- -

00

____ _ __ ___ ___ _______ __

00

V- H0 ~ u 41

>0

0 t- rI1

41~N >1

0d_ _ _ _ _ p_ I. 4 .

0 4J.W C;

0

0 400 (n ra U

f41

00 0 04 0A

9--

2684

$4I

TABLE XXXV. STRENGTIL AND. PROPO

Material Boron-EpoxyFiber Orientation

__00 90SP272-3M Narmco 5505 SP272-3M Narmco

Fiber Volume Fraction (%*54 Ilor. 53 Norn. 54 Nom. 53 'NomnDensity (lb/in3) .075 .074 .075 .074

Ultimate Strength (psi) 212,000 208,300 11,700 9,590"A" Allowables (psi) (18,400) (159,400)---"B"' Allowables (psi) (204,300) (177,300) --

Tensile - Proportional Limit (p:si) 136,000 120,000 6,700 5,200

Dat o1dulus (psi) 33 1x106 31.7:-.10 63. 1x 10 6 3. OxlO("A" Allowables (psi) (30.96x106) - --

"B" Allowables (psi) (32.065.-10 6 ) - --

Ultimate Strength (psi) 250,000 262,000 38,000 41,.000Compression-Proportional Lirnit(psil) 190,000 196,000 16,600 16,000Data

Modulus (psi) 33.1x106 31.7x106 3.6x106

Shear Data -Ultimate Strength (psi) 18,700 17,300 18,700 17,300Inplane

Proportional. Limit (psi) 7,000 6,500 7 0006 6,500Modulus (pi1.x0 .x0 .2.X10 1.lxlO

*Poisson's Ratio P214131= 1121=MJ31 PJ12= P13 1.12=1113.22 .-2-6 .0198

"B'o Allowables - 90% Probability with 95% Confidence"A" Allowables - 99% Probability with 95t Confidence

()-Statistical values obtained from experimental data.

* These are experimentally determined values and not to be used in computer pri

**These data are derived from narrow specimens and are probably. not representalPreliminary wide specimen data are almost twice as high.

269

TABLE XXXV. STRE14GTIL AND PROPORTION4AL LIMIT OF BORON EPOXY COMPOSITES

B~oron-Epoxy

Narmco 5505 5P272-3M Narmco 5505 SP272-311 iarmco 5505 SP272-314 Narmco 5505 SP272

53 lior. 54 Nom. 53 'Norn. 54 Norn 53 Nomn 54 llor. 53 klorn. 54 No.074 .075 .074 .075 .074 .075 .074 .075203,300 11,700 9,590 22,200** 171800** 108,000 103,400 --(159,400) -- -- -- -- -- (65,600) -

(177,300) -- -- -- -- -- (73,000) -

k120,000 6,700 5,200 ll,900** 10,700 33,600 35,000 -

6 .7l03 x1 .x0 .x0 3.6xl10 6 17.2x10 6 18.0x10 6-

262,000 38,000 41,000 30,000 29r000 225,000 245,000196,000 16,600 16,000 12,600 12,000 147,000 -- -

1.7x106 3.6x10 6 3.7x10 6 3.5x,06 17.8x106 17.6x10 67,300 18,700 17,300 65,700 63,200 29,000 27,700 -

6,5100 7,000 6,500 37,900 3,00380-.ll61.2x10 6 1.1x106 8.75x106 87,61.08x,0 6 1.0X106 -

,21=PJ1 P12= 1*13 P12=13= P21=1112= P1=Pll2= P=.065 pJ=.073 -

.26 .0198 .78 .80

lencelenceital data.

nd not to be used in computer programs*

and are Probably, not representative.ice as highl.

lIT OF BORON EPOXY COMPOSITES_

S450 00/50. h4b.- -5i.. 50% 5F-11 Narmco 5505 SP272-31.1 Narmco 5505 SP272-3M NarmcoL 0

m 5 3 Zion 5 4 llo. 53 Norn. 54 Norn. 53 Nomn..074 .075 .074 .075 .074

0* 17,800** 108,000 103,400 -- 112,300---- (65,600) -- (89,900)

-- -- (73,000) -- (100,000)Q** 10,700 33,600 35,000 -- 38,000

06 3.6x10 6 17.2x10 18.OxlO 6- 17.2x10 6

.0 29,000 225,000 245,000 -- 208,0000 12,000 147,000 -- ----

06 3.5x106 17.8x106 17.6x10 6 -- 13.0x1060 6-3,200 29,000 27,700 -- 50,000

0 37,0006 3,800 -- ----

106 8.7X106 1.08x,06 1.OXlo6 -- 5.75x10 6

2= PJ21=12= PJ=.065 5=073 -- J=. 743.80

bI

TABLE XXXVI. PHYSICAL PROPERTIES OF qGLASS FILAMENT AND EPOX9

Composites1002 1002 XP 251

Physical Properties Units E-Glass S-Glass S-Glass

Filament Volume 48 48 59Fraction

Coefficient of in./in. 4.8 x 10 - 6 3.5 x 10 - 6 -

Thermal Expansion OF uni at 0. uni at 00

Density lb/in 3 0.067 0.067 0.071

Specific Gravity 1.90 1.90 2.0 i

Resin Content % by 34 norm. 34 nom. 22 nom.

wgt.

Resin Flow 10 10 10

Volatile Content 2 max. 2 max. 2 max.

Barcol Hardness 70 70 75

Thickness per Ply in. .010 .010 .0075

Filament Diameter in. -


271

I. PHYSICAL PROPERTIES OF GLASS EPOXY CO)MPOSITES,'I GLASS FILAMENT AN4D EPOXY RESIN MA L. -.CES

comp osites Filaments matrices1I0 XP 251 Scotchply ScotchPlys S-Glass S-Glass E-Glass S-Glass X1002 XP 251

48 59 -

1cr6 3.5 x 10-6 - 2.8 x 10-6 1.6 x 10-6 - 3.78 x 10-500 uni at 00

0.067 0.071 0.092 0.089 0.043 0.0439

1.90 2.0 2.54 2.49 1.19 1.216

34 non. 22 nlom. - -- -

10 10 - ---

2 max. 2 max. - -

70 75 - ---

.010 .0075 - ---

-- o00039 .00039-

ding page blank271

TABLE XXXVII. ULTIHATE STRENGTII, PROPORTIONAL LIf-

Uni - 00

Scotchply Scotchply Scotchply Scotchply

1002E-Glass 1002S-Glass XP251S-Glass 1002E-Glas

Fiber Volume Fraction (%) 48 Nom. 48 Nor. 59 NoM. 48 Nom.3)

Density (lb/in. 3 .067 .067 .071 .067

Ultimate Strength (psi) 160,000 195,000 270,000 4,800

Tensile Data Proportional Limit(psi) 100,600 129,000C 188,000 2,400 C

modulus (psi) 5.5x10 6 6.4xi0 6 8.4,106 1.5xi0

Ultimate Strength 90,000 110,000 120,000 20,000

CompressiveData Proportional Limit 45,200 67,300 105,200 -

Modulus 4.70xi06 5.8x506 7.3x1 0 1"2x106

Shear Data Ultimate Strength 5,200 5,400 9,940 5,200

Modulus(In-Plane Shear) .47 .90x106 1.0x106 "47xi06

Poisson's Ratio P21=0.26 P21=0.26 p21 =0.26 j12=0.0 64


273

TABLE XXXVII. ULTI IATE STRENGTH, PROPORTIONAL LIMIT ANJD ELASTIC CONTENTS

Uni -0~ Uni -

Sctcpy S thyScotchply Scotchply Scotchply Scotchl1OO2 E-Glass lOO 2 S-Glass XP251S-Glass 1002E-Gl1ass 1002S-Glass XP215S*4 8 Nomn. 4 8 iloiil. 59 lorn. 4 8 Morn. 48 Lion. 59 Moa.067 .0G7 .071 .067 .067 .071

1) 160,900 195,000 270,000 4,800 6,700 11,900i) 1~ o 129,000c 188,000 2140c3,900

4,500) 5.5-r106 6.4%c10 6 8.4x10 6 1.5x10 6 2.1x10 6 2.4oxli

9o.0,00 110,0000 120,000 20,00026r045,200 67,300 10%)5, 200

43004.70x106 5.8.xio 6 7.3x10 1.2x106 .xO5,200 5,400 9,940 5,200 5,.400 9,940

r) .47 .90.V1o6 1.0x10 6 .47x10 6 .90x10 6 1.0x10'~I21=o.26 P210.26 P21=0.26 pl2=0.064 11 12=0. 089 120

*NTS FOR GLASS EPOXY

I~j.JJ...LU'50%/50%i - 900 ±45' 00/900

tchply Scotchply Scotchply Scotchply Scotchply Scotchply Scotchply15S-Glass 1002E-Glass 1002S-Glass XP251S-Glass 1002E-Glass 1002E-Glass XP251--Gla

1o. 48 Nom. 48 Nom. 59 Nom. 48 Nom. 48 Nom. 59 Nom.

1 .067 .067 .071 .067 .067 .071

,900 22,200 28,900 30,300 75,000 105,000 130,000

,500 - - - - 25,600 24,000666

4x10 6 1.50xl06 2.2xO6C 2.8x10 3.5x106 4.OxlO6 5.2x106

500 23,000 76,400 88,000 100,000

,300 - 43,000 43,000 43,700

3x,0 6 . 2.80x10 6 3.60x10 6 4.40x10 6

940 42,000 43,600 5,200 5,400 9,940

0 1.5x106 1.9x106C 2 5 x06C .47xi06C .90x106C 1.0xl06C

2=0.076 P=0.63 P=0. 46C p=0.51c P-0.11 c P=0 .12C P-.13

0\0

0V

H

1 10

V)$2M 0- 0 I

00 4

00~

0

tr4

04

0 tn

0 U)I

'4

Cdj CM Cdj H H H H

s a'rio 0v 7M0T (9 -OT *UT/UT; HIV1I1 a2IMN RM~ XHOJNU~IA


275f

0

0

000C'

U)lN 0

0 10

uo

0 1

0 4 i

0b0 $0

I-- LrZ

00

4L

0l0.Id 4

00044

UN)

0 0 0 0 0 0 0 r

C ~ ~ A C4% 0 0 0P

276

The orthotropic elastic constants for Scotchply SP286 andFiberite E1305 graphite epoxy unidirectional laminates aregiven in Table XXXVIII.

TABLE XXXVIII. ELASTIC CONSTANTS FOR GRAPHITE EPOXY COMPOSITES

Elastic Constants

Scotchlx SP286 & Fiberite E1305

Unidirectional Configuration High modulus High Strength

Ell Modulus-Longitudinal psi 26 x 106 20 x 106

E22 Modulus-Transverse psi 1.6 x 106 1.5 x 106

G12 Shear Modulus psi 0.57 x 106 0.61 x 106

P12 Poisson's Ratio 0.300 0.300

Vf Filament Volume Fraction % 55 55

27

277

LA0

,41

Cl04

H 01 0C

9WH 4 "1 0)1 L) %

U)

P4 U) 14HE1 0 4J0

04 'n tH2 0) C44 a In NI

k 0 Hn

E- .14 :3 I I CHl k) O.8.3 0 t.

:9 : Ln IV 0 H 00 0

0- (1 H

44A Z~ LA 0; 14

% o 4

I n 0 0

oto.

" 143 N 0 n LA N N0* OP A

I-~ ).~ LA 0 A H A

> H

14 4.)

1.43 L 0 N1

0I 0 0 0- W0 4

LA LA I 0 -- H A U - 4 A4

4J' LA1 l t lAMu to 0 -4 :t 0 I

LA $4 LA 0, & o

278

o 0A1 a t .0 0D e-i 0 H-

I E4 N 2z) 0 0 0 Xcmw 0 H 0D IV -i LA

0 4b x % b %.00n '0 '0 '0 L

U)~ 0k %0 0o

H 0 x4 xXEn d N LA Nq H- q

0 01 v

%.0 H- U) H

'.00 %.0A. 0 0 0 0 0_ E-1 0 0 H Oi X O H -4 c

LAx x

T1 r-4 N NA N N to

On 0 0 0 0E- A En Co a N ) x C

o nL %.01I.0 a k0

o- 0o 0I 0- r- H

0 %DOH

0 Ln W L o % 'E- T 06 *b o'.0

E) r N H H 0

o0 0 0

E- t 0 C) 0 a 0 tC.) 0n 0 H 6 -

m0 H 0 H D 0

r- H N H C)

E-4'.0 o. 10

0 0 0 HD 0now 0 A 0 H 0 X- 0In rA a H~ C) x 0a

T 0 %.0 Ln '.0 %b L

H N 00 N co * 0

I .*

N 0 H4UP- 4JJ (n 41

43 U)r ) .Wr

V U) 4M1 U -4.) Hj rj H W Ea ) :J

$40 N4 Ur r. QJ4JQ LfW U $4 4J04 a) 1 ) C) En~ 0)U 44 40 r 0

D4 0 -4 H w H4 $1 9 4 u $S4 U

279

APPENDIX III

DISCUSSION OF SYNERGISTIC INTERACTION BETWEEN CREEP AND.FATIGUE PROPERTIES

Synergistic interaction of two variables is the nonlinearaddition of the effects of two variables in which the sum maybe significantly greater than the individual parts. Applyinga synergism to creep and fatigue is difficult without test dataapplicable to the specific design stresses.

Most of the research to date has been in characterizing thesynergistic interaction of creep and fatigue on high-temperaturesteels. In the case of aluminum, 6061-T6, the interactionproblem is compounded by the fact that cyclic stresses at temp-erature will affect the overaging characteristics of the aluminum.This overaging will occur faster or at a lower temperature thanthe static test data reported in MIL-HDBK-5A.

The nature of synergism with respect to creep is that lowcyclic stresses have a tendency to cause creep rates to increasemarkedly. Under certain circumstances, materials can changecreep rates by an order of magnitude. Total creep elongationat failure can also be reduced.

With respect to fatigue, creep stresses and creep (permanentstrains) have a tendency to break up dislocation networks whichare a contributing cause of fatigue crack initiation. The netresult is an apparent increase in fatigue strength particularlyat appreciable steady stresses. Some of the steel constant lifefatigue diagrams for elevated temperatures in MIL-HDBK-5 show amarked improvement in fatigue strength at combined elevatedtemperatures and steady stresses.

In some cases, the fatigue strength of the material at steadystress and elevated temperature is higher than the fatiguestrength at zero steady stress and the same temperature. Thetest conditions (material, temperature, cyclic stress ratio,steady and vibratory stress) must duplicate the design applica-tion. Significant errors can result without well-matched testcondtions.

Test data to assess the synergistic creep/fatigue effects onnotch sensitivity has resulted in mixed and conflicting inter-pretations. Apparently, if compressive residual stresses areinherent in the machining and the machined stress concentrationrelieved by synergism, then the notch effect becomes more severe.If the stress concentration becomes blunted by creep, then thenotch effect becomes less severe.

280

LIST OF SYMBOLS

A area, square inche.,

a characteristic radial dimension

pC stiffness coefficient

u

v

C stiffness coefficientA u

pC stiffness coefficientw

vC stiffness coefficientwD pitch diameter, inches

DR rotor force - aft, pounds

dm bevel gear mean pitch diameter, inches

d p pinion pitch diameter, inches

E modulus of elasticity, pounds per square inch

e elongation, percent

Ec modulus of elasticity in compression, pounds persquare inch

F face width, inches

Fa force at a, pounds

Fbru ultimate bearing strength, pounds per square inch

Fbry yield bearing strength, pounds per square inch

Frib force on rib, poundsFsu ultimate strength in shear, pounds per square inch

Ftu ultimate strength in tension, pounds per square inch

Fty yield stress, pounds per square inch

281

fa axial stress, pounds per square inch

fb bending stress, pounds per square inch

frib stress in rib, pounds per square inch

fvib maximum vibratory stress, pounds per square inch

fxx maximum stress in x direction, pounds per square inch

fyy maximum stress in y direction, pounds per square inch

G modulus of elasticity in shear, pounds per square inch

g acceleration of gravity, feet per second 2

h thickness, inches

Ixx moment of inertia about x axis, inches4

I yy moment of inertia about y axis, inches4

K geometry factor, inches 4

Kf notch sensitivity factor

KIC fracture toughness, pounds per square inch (inches1 /2 )

Kt stress concentration factor

KLateral lateral spring rate, pounds per inch

KVertical vertical spring rate, pounds per inch

MA moment about bearing A, inch-pounds

MB moment about bearing B, inch-pounds

MxR rotor moment about x axis, inch-pounds

Mxx moment about x axis, inch-pounds

MyR rotor moment about y axis, inch-pounds

Myy moment about y axis, inch-pounds

MZR rotor moment about x axis, inch-pounds

N speed, revolutions per minute

n pinion speed, revolutions per minute

282

P radial shear, pounds

RA bl bearing reaction in butt line plane, pounds

RA horizontal bearing reaction at bearing A, poundsh

RA bearing reaction in station plane, poundsSta

RA horizontal bearing reaction at A, poundst

RA vertical bearing reaction at A, poundsv

RA bearing reaction - main rotor, poundsx

RA bearing reaction - main rotor, poundsy

RB bearing reaction in butt line plane, pounds

RB horizontal bearing reaction at B, poundsH

RB bearing reaction in station plane, poundsSta

RB horizontal bearing reaction at B, poundst

RB vertical bearing reaction at B, poundsv

RB bearing reaction - main rotor, poundsx

RB bearing reaction - main rotor, poundsy

Ra bearing reaction at a, poundsx

Ra bearing reaction at a, poundsy

Rb bearing reaction at b, poundsx

283

Rb bearing reaction at b, poundsy

RA axial load on thrust bearing, poundsx

rI radius at top of conical section, inches

r2 radius at bottom of conical section, inches

SR rotor force, pounds

T gear torque, inch-pounds

TR torque on first-stage ring gear, inch-pounds1

TR torque on second-stage ring gear, inch-pounds2

u lateral deflection, inches

V vertical shear, pounds

VR vertical rotor force, pounds

Vf filament volume fraction, percent

WH horizontal gear load component, pounds

Wr radial component of gear tooth load, pounds

Wr. radial gear tooth load in, poundsin

Wt tangential tooth load, pounds

Wt tangential tooth load in, poundsin

Wt tangential tooth load out, poundsout

Wv vertical gear load, pounds

Wx axial- component of gear tooth load, pounds

Wx axial gear tooth load in, poundsin

w vertical deflection, inches

284

YS yield strength, pounds per square inch

Z elevation of shell above base plane, inches

a cone angle, degrees

4 spiral angle, degrees

V pitch angle, degrees

cpressure angle, degrees

V Poisson's ratio

1z

1270-75

285

A 008995

Documents

Transcript of A 008995