I,
AD-A018 047
DEVELOPMENT OF A MINIATURE GAS-BEARING CRYOGENIC
TURBO REFRIGERATOR
R. B. Fleming, et al
General Electric Cor,.orate Research and Development
Prepared for:
Army Mobility Equipment Research and Development Center
October 1975
DISTRIBUTED BY:
National Technical Information ServiceU. S. DEPARTMENT OF COMMERCE
343134 I
Contract No. DAAKO2-71-C0026
DEVELOPMENT OF A MINIATURE GAS-BEARINGCRTOGýE TURBO REFEMIATO
R. B. Fleming. et al.Power Generation and Propulsion Laboratory
00) Corporate Research and DevelopmentGeneral Electric Company
Schenectady. New York 12301
0 .. D D C
19 November 1975 DC 81975
b. ZA ES PC TAU
Final Report for Period 15 Jan'iary 1971 - 1 October 1975
Approved for public release; distribution unlimited.
Prepared for
U.S. Army Mobility EquipmentResearch and Development Center
Fort Belvoir. Virginia REPRODUCED FROMh.r.-4 by BEST AVAILABLE COPY
NATIONAL TECHNICALINFORMATION SERVICE
SPmf-0.1S VA. 221$1
SRD- 75-097
Une lstisfic ________
SECUR~ITY (.LAS154ICATIOl4 ;)F TH~IS PAGE (W1hw. helm Engidtati)___________________
REPORT DOCUMENTATION PAGE HFFRI(A soiilr~.I IN(;
I REPORT Ni'MAE.4 ).OVT ACCESSION NO. 3 PR1-(IPIENT'% rAT AL(,i UIMPIIIf1
4. TIL (.n Sott0 I. TYPK Of' Kr~oRT & ~Plr~om i
Developmient of a 'Miniature GasB-earing ilReor1t5 J anua rv 107 1 -OC LLb uvIi
C ryogenic T urbo R efrigerator, 6 ~ F R I GO rR .0i i n
17~ ~ t B,1 OfV3lernirngDj 13. Colyer, * O1ATOONNJ~E&
R. K. Te rbu sh, B3.B. Gamblde. It.0. Oney DAA K02 -71-C- 002(.
S PIONMIO CNANI ZATIO4 NAMEAN tfP5 PROAMLMNPFOr ,C
Corporate Rtesuarch and Development A4Aa*P N1NMIGenernl Electric Com~p anySchenectady, New York 12301
I Ii. CONPe't.LINGOO'FIC9 NAME AND ADDAISS5 It, MI.IP'Ofl OATS
U. S. A imy Mobility Equipment Research and October~ 1075Ileveiopnient Center 13. NUMNEOF PAGES
F.ort Ielcvoir, Virginia 220610 Call4l l1)1. 1u1, c~~eTTIt, MbNlTcRnNG AGENCY' NARi4 I Abonetti eofoitt !ffl75 EURT LAS l hnt
Unclassified
Ua~ECLABSSPIATIO miýW [lAw"SCH EfU LF
7S. D1TI19UIS1ON STATEMENT (ol lhl. RePwOt
IApprc~vve kir~ puhlicý LcilIODSi.; diStvibIl ion 11[it'il nit(rl,
i7. DISTRIBUITION STATEME.NT (of IliC abstraef *iilprei nm Slack 20, it dIll.,eiil Veii Horptl)
ilSU51PPLEMENTARtY NOTE3j
IS. K I.Y WORO'.; (Corilinr. oto tov"ro , I it nt'w.Amar- and Idtintllv by blot It moiflbpi
C ryogenics, Rehfrigerator, Turboalternato r, Heat Exchange r
20. AIIISTRALC1 (f'omfinuo on ,eivepam sireif 11e e a ttl Idoi dehll by. hinch ,urnb,,
The work described in this report wakH undertaken 1o adlvance thu devvkop-mont of nminfiature cryogenic refrigeratorts that have high re~liability, long
DD I JW7 1473 EDITION OP I NOV 65 11 OftSOLETItncB(4 r
SecUNITY CLASS IFrATiýeij At 'I P A l,
Un clas ;Ifled59CURITY CLASSIFP CATION Of THIS PAOE(Men P18 fMntoredg)
Block 20
A distbiguishing feature of the equipment under development is the use ofhigh-speed lynamic, compressors and turboalternators, in which all rotatingparts are wispended on self-acting gas bearIngs. Such at mystem avoids slidingcontact and wear during operation and avoids the contaminatuion problemns
7 ~associatet' with oil lubricated compressiors used in conventicnal cryogenicrefrigerators.
The wook described in this report was directed finally to theý developmentof a system consisting of three cryogenic turboalterneitors, six helium con-trifugal com pressor stages, and a cryogenic heat exchanger with seven heatexchanger riwdules. Work was limited to the turboalternatovs and heat ex-changers. /' cryogenic turboalternator was developed and successfully testedto temperatutres In the range of 80' to 100'K. In a parallel contract, a similarturboalterneitor was tested, with no load, at a temperature of 0. 811K.
The conti-act effort fell short of Itp. goals because of difficulties encoun-tered by a vondor in the construction oi the heat exchanger assembly. Theheat exchanger was designed and constructed using a novel approach thatpromised a significant advance In the technology of cryogenic heat exchangersand a sizablv reduction In Mqize nnd weight. Itowever, it was found unexpectedlythat not all developnient. probleris had been solved before construction began,anti after a two year delay in delivery, the heat exchanger was found to be un-usable for its intended function.
The work re~x~rted on the turhoalternator represents an advance in thetechnology of cryogenic refrigeration. It is expected that heat exchanger de-velopment, problems can be solved with further effort.
Unclassified
SktCURITY CI. ASIFIIC ATION OF THIS5 PA13K(Whomnts Ffou dieredI
OSNIRAL@ILICTHIC
t
.... ... ..........
4 FOREWORD
This final technical report covers deveoppment work per-formed under Contract No. DAAK02-71-C-0026 with the U.S.Army Mobility Equipment Research and Development Center,Fort Belvoir, Virginia. The work was perlormed by the"Power Generation and Propulsion Laboratory in the Resnarc)hand Development Center of the General Flectric Company inSchenectady, New York. The work described herein coversthe period from January 1971 to October 1075.
The program was under the direction of Dr. L.I. Anistutz,of the Mobility Equipment Research and Development Center;this report was submitted in October 1975.
Tie General Electric Program nl1Mnager waa Dr. H.13.
Fleming.
The principal contributors LU this pvogrurn wee:
R. B. F'lmitmngD.1B. ColyerR.K. TerbushH.1B. GambleW. H. One,y
fii
SINEAL O'L'CTRIC
TABLE OF CONTENTS
Section
1 StMMA .. "................................. I
2 INTRODUCTION .................... 3Objec(tive . .......... 3tBackground. 4.. L'oznpli•hMcntaS . . . . . . . . . 5
3 1 IR [NOR I(GRATI ON SYSTI:M . . . . . . ......... 7systvin- IDescz'ption .. . . . . . . . . . . . . . . . . . . . . . 7
( .rl SvYtoni.............. . . . . . . . 7Sy tenl srIon4idf'red ................. ...................... 7
P-ositive ItPplavernent C'ornpjr'esaor. 7Illbrid ('!yvIt .. . . . . . . . . .. . . . . . . . . ... 7
• .... Sviten ('onci............................
Split Pvo'vh . . . . . . .. . . . . . . . . . . . . ..9¢._ ~~Finul Svstvlin Sthlvctiun .. . . . . . . . 10
Cy Le A 11;klý'Hi S .. .. .... . . . .. . ... 12("OrlpUte v Prot"I'alzn .. . .. . . . . ... 12System Design (:'riterit . . . . . . . . 12
Systenm Weight . 12Rtesults for 5-WLatt S,',Strni . ........ .. . 14
()'f-I)ue ign (Cahculatio( s. . . . . . . ... . 17RCM11ltS I'01 :'I-\V~ktt Systemn~ .. e.. ... ... 18
()ff-I)hign CaLculationn,. .. .. .. . . . .. 22Systvill Ou . ... ... ...
Guncra'd .. .. ... . . . . ... 23
C'ryogenic' Piping . . 25loulc-'rhonion Vulv . . . 25IHttuid NittroL'n C!ooler . . . . . . 2(
Cold Ilnd Cooler'. 4 . 0 . . 4 . & . .... . . . . . . . . 11)
4lU i t.AL wak I. . . . . . . 0 . 0 . . . . . . . . . . . . . . 213,nayout of "rgogtenic 14votion .T.u . .o.tno . . .. . . '2
4 TUit OA lTc NATOR )i. . . . . . . . . . . . . . @ . . . . . 31General ,. a.. .. ..... s ao* . )s . .. .. .. . .. *. . . 31"TurhonItt rnator l)esi ns . . . . . . . . ... ... .32
Pr'eliminary IDusigns ,1F'inal Design of the 141K Tutitboalte~rnator .. ... . 32
Alternator Design.k , .. .a ev . . . .. a. .. . .. . o o 36Gras Htearing• Anal ysis and IDesign ....... 3 61
Assornilflv Procedture, 3 6TIurhoaltr- rnator lPut-forrna.nec .... * . . I * 36;
Turbogltcrnatto *. , l . . . . . # ..... . . . . ..Týu rbtx lttrnattor t'I rfotrnuttnce IAi 't HVrdu t1to
lP 'ograni4o6. 37
Preceding pail blankV
SUNINAL ILICT1IC
TABLE OF CON11NTS (Cont'
Section EREe
4 Tt IRBOALTERNATOR (Cont'd)......... .Journal Bearing Material*..... .. 43
Very Low Ternperature Test............ 43MERDC Turboalternator Assembly....... . 53Preliminary Turboalternator Test....... .. . 55
MERDC Turboulternator Open-C(ycleTests . ... ..... so ........ o.. .. 56
Final Turboalt'ornator Open-Cycle Tests ..... 60Performance lest Results.... ......... 63
5 CR'OGENIC HEAT EX(HANGERS . . . .. . . . ..... 69General . ..... ... .. *fee of... ... .. 0a . .* 6* s
Design and Construction . ........... . . . . 69Test Results--Kinergetics..&.. a..... .... 69Test Results--General Electric ....... ,.. 69
General Electric Test Facility .......... 70Test Method . .. ...... ... . . .0. . .. . 70Test Results. ........ ... .......... 73Discussion of the TestResult ui........ 75
6 RE IOMMENDATIONS .......... . . . .. . .s. . .. 79Turboalternators ...... .. . . .. . .. . .. .. 79Cryogenic Heat Exan.angers . ............ 79Compressors . . . ... . . ., . . . .. 80Systems .o. . . . . .. . . . . .... . . ... . 80
Appendix
I PR 'LIMINARY TURBOALTERNATOR DESIGNS ...... 81Turboalternator DEosign Goals. . . . .a . . . .0 . . . ..a a 82Design for 140K Tiirboalternator . . . . . . . . . . . . 84Design for 55"K Tturboalternator, . . . . . . . . . . . .a 85Design for 170%K 1 urboalternator............. 93Mechanical Arrangement. . . ........ . ..... 100Turbine Wheel Struss and Deflection Analysis.. ... . 101
14"K Unit so #, . . a..e.. ..... o.... ao. 9 101551 KUnit . ........ . .. . .... . ...... 103
170"K Unit .. .. .. . . . ...a. .. .. . 103
11 GA -BEARING ANALY41S AND DESIGN.. ....... ... 105Bearing Design Requirements. .. .o. . . ..a . . . .o o 105Bearing Type Sele. tione . .. . ..... ...... . 105Journal Bearing D-sign . .. . . . .. . .. .. .. .... 109
Procedure.. .... . .. .. a a $. .... 109
vi
TABLE OF CONTENTS (Cont'd)
Appendix
11I AS-,IEAR]NG ANALYSIS AND) DESI(IN ((ont'd) ....Pu.rl'urranc . . . . . s . .* a a * a a a a a s * 111
Thrust 13earing D)esign ......... ... .. ... . 128Procedure . ..... ......... . . . . . . . 128
v•Performanceo** st o0 e sG e 129
III T1 IuMOALTEHINA'roli A.SHI;MI•IY PRO)C(EDURES . . . . . 147Preliminary Cleaning and Hlandling . . . . . . . . . . . 147Clean Room .A\strnblv. . • 147initial Whloe-to-Shuit \ssemblv . ........ . . . . . 149Hialanring the Rotating Assembly. ........... . . 151
F- Inner Thrust Hearing Assnihblv (CPart 22) ........ 151Outer Thrust IHearing Assemblv (Part 13). . . . . . . . 15JPrelimina rv J ournal Ihearint ,\ ,ssembly
and Stator Inst.illation . . . . . .............. 153Shaft Housing Assmlrlv. ................... 155Nuzzle A\ssomblic~s.. . . . . . . . ... 157
P'relinilnarv Thrust Ik tarin4 Shimning .. .. .. ... 17Installation of' Thrust. Proxirnitv Probe . .............. 160Final ,Ul rnal IHa ring I'ad .\ddjustments ......... 162l'urb alt riatol (•pe raion. .. . . . . .......... 164Mugnetizing Siaft hil•agnet. .. . . . . . . . . ....... . 1Shalft IkHUlIMICU I'(InjL lt . . . . . . . . . . . . . 169Parts list [or \ssoniblv I)rawingres . . . .. .. ..... 170
IV ,M;1I ,1 H('..\ 1'I( )NS ,'(M) A .A' S1 ()I,' ) V\'I ,N C RY(OGENIC
IltEATr1x iiC\N( 1::I . . . . . . . a # o a 173
iurn a II Pe to n in e and I're s sure Drop . . . . 1 7"S't~l-WAUV }p'lO l'l~tl *II( .'t, ' Dro .... 175Gorh' ll'ev.: ....... ...... ...... ...I%
Attaohnivnt to I1unlngr . . . . . , ....... . . 176Support ot Componnts ... . . ......... 176Stuctutl Ili it lak. . . . . . . . . . . . 177
Design Lifv. . a . 0 . . a . . 177Cleanliness.. . .. . . . . . . . .. . .. . . . . . 177Testing I'"'o.edures -- Thernal .......... 177
Cryoguni, Tusts. ......... .... . . . . 177Data Fxtrapolation . 178
Testing I'rocedures -- I•r, ssure I)rop. • . ... 178Flow Tosts . . . . a.... 0. ... . . .. 178
Data lxtrupolation . . . . . . . . . ... 178'rheLrmal C'vcling 0 . a . a * . . 0 b a . a . 178
vii
r SIN IAALI* ILICTA It
TABLE OF CONTENTS Ft"nncl
Appendix
IV SPjN•CIFlCATIONS FOR A SET OF SEVEN CRYOGENICHEAT EXCHANGERS (Cont'd) ...................
Testing Procedures -- Leakage .......... 179Stream-to-Stream Leakage . . ... .... . 179External Leakage. . .d.. t off. . a. . .... 179
Witness Testing .......... ......... 179Design Review ...... . . . ....... . . . . . 179Schedule . ... 179
Deliverable Itms.. ....... . .......... 180
V KINERGFETICS INCORPORATED FINAL REPORT .... 181
VI HEAT EXCHANGER ANALYSIS...........S.. . 20P
Relating Results to Design Conditions .......... 209Heat Transfer Data. .... a ... ofse .... 110a 20P
Pressure Drop Data ................... 210Data Reduction Computer Program . a . 0 . ..0. . . . * 2101-ffect of Helium Leakage, Thermal Radiation.
and Axial Conduction on Heat ExchangerPerformance .. ....................... 215
Stream-to-Stream Leakage . . . . . . . . . . . . .. 215Leakage to Casing .................... 215Thermal Radiation and Axial Conduction ...... 216
VII RE I"hRENCES . ..... .e44...4* . . . . . .. . . . .. . 21P
LIST OF ILLUSTRATIONS
Figure PaeI Hytirid-Cycle System tSchematic Diagram) .... . ..... 8
2 Split-Cycle System (Schematic Diagram) ........... 10
3 Claude-Cycle System (Schematic Diagram --
Corresponds to Computer Run 479) . . . . . 114 Tuboalternator Cryogenic Performance Power
Output Versus Speed.......... . .......... . *1 16
5 Design Point for Systenm with 3-Wati rapacity at 4. 40K(M ui1479) .. . . . . .. . . . ... pas.... 19
6 Inptit Power Versus Turbine Inlet Temperature ....... 21
7 Cryosection (Schemaitic Diagram) .. . . . . ....... 24
viii
LIST OF ILLUSTRATIONS IContd)q
Fi1gure ul
8 0 ~i I I gL Sv.Stter (St. I u~1 - I :1 ic ~ 'i I M ig v; I it 1). ... 26
9) T 1 ' 1 W t I LhLtIVC I )i Mt '1lI U! i(l It 2 7
110 ICf.%. -o -Nl i tzot .et I.r~L~ . . . . . . . . . . . . . 302
14 I'l 1huiAlte''1-11.1110 '1- . 6 * . * * . 33*
15 hIrj lw 'i 1111i, ltor it w 1 Is 1 1't I I t, iIl I .. . . . 0 . 3
III it, -a Litk. .41
17 I'tst lii.Itliti'in 11101'zijIu I . * 44
t)rjt't( 'Vk It? 'll I -I h II I ~~~I I-1'111 ' ' I I J~k tfl11 I r0 4 rar . . . . 45
'T., 100,IIh1 .0. .k.. .. \.. . . . . . . 54
20 S I{O I-f I i llIt Ill S \a I A ~ h 'J Il H0 01 1A I L.\- I L I I 5k
I'tz11~ ~ tu00 ' 000 I 'V . , . li . ~ ) .11 . . Al ...... . . . . 58
Ii k0,0 ) I:\ tk I U. .U ý A . . . . 57
ofun (1-1ML )pu'n-Cvilteru~twith k1k1 ih 1 watI ulancIItt ViLl,, N mlmd 100,ID 000 IUN1. .. . . . . . . . . . . 50
.111d tl 100 00 HI'M.~u~ . 24 .I. .... . . . . . . . . . . . 511
25 Vhl 1,M )v!~ 'urball: rmitr j'pt- ilitt igilltho S)oIame). w
1121K Os 610. 1 -Wott load, nild P18~, 9401 IWN'M withNov'.-M Prossui'u IRiuii ~ (it'.. 1* .. 6 2
*IIIIALI ILICTRIC
LIST OF iLLUSTRATIONS (Cont'd
Figure Plige
27 MELRDC Turboalternator Operating in the SameConditions as Figure 26. . . . . . . . . . ...... 62
28 Ovorall Efficiency Versus Velocity Ratio ........... 64
29 Wheel Efficiency Versus Velocity Ratio . . . . . . , . . . . . 64
30 Fluw Factor Versus Veiocity Ratio a . . . a. . ... .. .. .. 65
31 Corrected Torque Versus Velocity Ratio . . . . . . . . . . . 65
32 Flow Factor Versus Corrected Speed. . . . . . . . . . . . . . 66
33 Ehlvtrical Output Power Versus Speed. . . . . . . . . . . . . 66
34 MEHDC Run 103, Data Point 1, Data Obtainedwith Ucusing Bleed Valve Closed............ ..... 67
35 MF:RDC Run 103, Data Point 2, Data Obtained withHousing Bleed Valve Open, but not Corrected forHousing Leakage Flow. . a o a . o * s as a a s . a s 68
36 IeAt Exchanger Test (Schematic Diagram) . . . . . . . . .. 71
37 Heitt Exchanger During Assembly . ..... .. .. . .... 72
38 Design for 14°K Turboalternator ....... ....... 86
39 De.•ign for 55'K Turboalternator . .*. ... . . . D1
40 Des4ign for 170"K Turboalternator, 0.625-inch Wheel . . . 04
41 Design for 170"K Turboalternator, 1.00-inch Wheel. . . . 07
42 First-Stage W heel ................ . . , . .. . . 101
43 Cold-Condition Eftectivv Stress Contours. a .. .0. . . . ... 102
44 DL j lections for Wheel in Cold Condition a . . a a . .. a. . . . 103
45 Pivoted-Pad Journal Bearing (Schematic Diagram). . . . . 106
46 Cryogenic Turboalternator Tilting-Pad Journal Bearing. 107
47 Ginibal-Mounted Spiral-Groove Thrust Bearing. . . . . . . 108
48 Spiral-Groove Thrust Bearing (Schematic Diagram). . . . 108
49 Design for 2.0-g Gas-Lubricated Journal Bearings . . . . 114
50 Design for 1.0-g Gas-lubricated Journal Bearings .... 117
51 Design for 0. 0-g Gas-l ubricated Journal Bearings . . . . 121
52 Journal Bearing Performance as a Function ofG oal oadingo a . . . . . . . . . . . . . . . ..aa a a , a , a 124
x
V
GININALO*ILICTRIC
LI? f OF ILLUSTRATIONS (Cont'd)
Figure Pg
53 J ournal HI -iI, F'V 1104 I'llo t1 iie III CV 18 l N e F1 0ioOf 6-1.Lo diig . . . .............................................................. 125
54 Jouttiul licuring Decsigni (ihurawric and .1a~d Frequen cy
5(3 iiCH1iuni-f Ublie)LIted Sp~itl-AGrkOOVU TIhtuLIt btariig, IOhV.ignHuim No. 47100O3. RaLti)ot 01 L)utIiti toI Iriidt' Oiametur 2.0. 131
57 1L itn- bi itdSpi rat -( oovu Thrum Hem-IIng h , 1.)e si4,zlRun No. 471.1004. klW~t ioor OUtside to InSidC DIU]IHQVtei 2.0 * 13(3
581 Ie liutii-1,1,u1)1iaýt~Vd Sp~i r:I-( ;l'uoo' H~I' i-iot licavtII4, RaLtio)ol (.SW Wu et 1n1d k' Di .wtin t -' 2. 00, 1 i~ig' Run Nu,4 71)1003 .. .. .. .. .... .... .... .. .... .... .... ............ 139
50 I lI'iufl1-I.,UbIi-LCte, ThPUSl( HU(' 'Iiiu~ g,12 210- Mic routic (1'uov e Ihpti, z1 ''~l~( 1is 0
V1oa0led O S ide Ld Sid(tu 'e . *. . ., . .. . . . . . . . . . 141
1~ .0 I~t tci 14 'd Sod~ i d 0 ( 10, "ani I I 10 . . . . . ., . I -'1263 llintutael pa- 1toc Tlii'w'I H';uit'iw, 948-
62 I[V lX)II en 1l'il'I d U1-11hq P-\1, ll, 1015 Ii i\i i...........14i
(;. 4,u l o ib a tv ini t 111 a() I II-Il ~. . . . . . . . . . . . $ I . . 6 a . . # 6 1 48
01 5 W elc I "1l&'c AsI. lIIII\ .. .. .. ..... . ........... . . . . 1 . *0J
66 I1')LIVIM 'Itv ltt1ý1, hSIVH, \st-wiiihI., kilun, I jLi liltol 151
67? TIurIXMP11tilt1or' IrtI101 IhLtw-it PkMiIng .\M( . .ll . . . .. 152.
68 (zvboaterli ltolt wtor 'liltitugThrus Heaint -\s-.wf11IY. 1
F 70 l irtt- .g e S a't lieeL Therus (H;i . .. .......... * . I. . . $6. I 6 15
70 ivs -s~lg W ive 11111ýr W) . .. .. .. . . . . . 15 G~
71 '1urmxW alt'rnimtl.o' Ofni'lrl iin-io lt S*),1'litlmlhin ti 'IhilrustH~eariing I'Po41:iPtin.. .. .. .. .... .. .... .... .... .... ........ 15 8
MININALO 'tCTRIC
LIST OF ILLUSTRATIONS (onted)
Figure Pg72 Thrust Clearance Ring Positioned in Inner and Outer
Thrust Bearing Assemblies .... . . . .... .. .. .. .. . . 159
73 Open-Cycle Test Station for Turboalternator..o sso.... 161
74 Turboalternator Assembly on Open-Cycle Test Stand .... 164
75 Principal Turboalternator Instrumentation. 0........ 165
76 Open-Cycle Turboalternator Test Station.. a • •. .... 166
77 Open-Cycle Turboalternator Inlet Gas Instrumentation... 167
78 Modified Model MV-6 Balancer..... .............. 169
79 Balancing Cradle Assembly........ . . ........... 170
80 Claude Cycle System (Schematic Diagram) ... .. .... 174
81 Pressure and Temperatures Throughout System ...... . 175
82 Sample Data Run .......... . . ............ .... 212
83 Stream -to-Stream Leakage ............... . . . . 215
84 Leakage to Casing ........................... 216
85 Radiation and Conduction . . . . . . . . . . . . . 217
LIST OF TABLES
Table
1 Heat Loads for Hybrid-Cycle System . ........ ... 9
2 System Design Calculations -- 5 Watts at 4. 4°Kv. . . . . . . 15
3 Design-Point and Off-Design-Point Computer Results.. .. 22
4 Estimates of Heat Conduction Rates. ............... 29
5 Alternator Design and Calculated Performance .... . . . . 35
6 Comparison of Wheel and Nozzle Characteristics....... 36
7 MERDC Turboalternator Flow Test .......... ...... 55
8 Heat Exchanger Test Results ....... . .... ........ 73
9 Leakage Measurements.. ... ... . . . . . . . . ... .... 74
10 Turboalternator Designs . . . . .a. .. . . . . . . .o a a a .. ... 81
xii
0 INIRALo ILtTUIIC
LIST OF TABLES (Cont'd)
Table tae
11 Turboalternator Design Summary . . . . . . . ....... 83
12 Design ' ariations for 14"K Turboaitcrnator . . . . . . . . . . 85
13 Design Variations for 55WK Turboalternator . . . . .... . 89
14 Design Variations ior 1 70'K Turboalternator. . . . . . . . . . 90
15 Designs with a Maximunlm Operating Speed of 250, 000RPM ...... . . . ... I * a 6 ii
16 Design Speed Performance aL 2. 0 G... 112
17 Gas lHearing Performance Summary . .............. 113
18 Turbualternator Tilting-l'ad J1 ournal (Gu s- HearingDesign Summary ...... . . . . . . . . ........... . 127
19 Thrust Bearing Design Comparison . . . . . . . ...... 140
20 Turboalternator Spiral-Groove Thrust flearing DesignSummary . . . . . . . . . ........................ . . . . . ... 144
21 Axial Shimming Measurements for '.'urboaltnrnator . .t. . 158
22 Parts List for Single-Stage ludial ImpulseTurix.altvrnator. . .. .. .. ....... . so * . ... . . . 171
23 Parts List for Inner' Thrust c;lmbal Assembly. . . ...... 172
24 Parts List for Outer Ththust Cimbal Assembly .. .. . . .* 172
xiii
I INIHPALO@ILICTRlc
SUMMARY
The work described In this report was undvr-taketri to advanve the develop-ment of miniature cryogenic refrigerators that huve high reliability, longlif o, minimumI mlaint~enance requireinents, and no mechanical vibration.The type of system investigated utilized the Claude cycle for the productionof refrigeration at 4.,4 K. TempcaratureR at thim level are suit~able f'ornumerous appli cat.ions of superconductivity,
A distinguishing feature 01' the equipment uinder development is tile useof high-speed dynamic compressors and turboallernators, in which all ro-tating parts art, suspended on self-acting gas bearings, Such ui systernavoids sliding coni net and wvear during ope ration andi avoids tile cnntaminationproblems associated with oil lubricated compressors mused in conventionalcryogenic refrigerators.
The work described in this report. was directed finally to the developmentof a systemi consisting of three cryuog'nic turbonlturnatorsl, six helium cen-trifugal compressor stages, anti a cryogenic heat exchanger with geven heatexchanger modules. Work was limited to the, turhoalternat or and heat ex-changers. A cryogenic turbotiltoernwtor was developed and SUCCeSufuilytested to t eniperat~uros in the ran~ge Of SOO to 100111. In a parallel contract.a similar tUrboalt ernator was teast ed. with flo load, at a temperature of9. 80K.
The cort ract. effort fell short of Its gorils becatiscv of difcLi 'lult i e encountered*by a vendor in thev construct ion of t he heat. exchanger iissenibly. Theli hr-at ex-
changer was designed and constructed using a novel approaewh that promiseda significant advance in the technology of cryogenic hevat VxchangerS RndI asizable re I icc . on in size andI we i v t. H ow(VOT~, it WWI; foit111 n a lXPOetCludiythat not all developmt-ent. problems had heen solved ibcforu cmisiruction began,and after a two year delay in ci ci ivery, the heat exchanger was found lo beunusable for, Its intended function.
The work report~ed on the hi rhoal ternator represents an a' iva fle in thetechnology of' cryogenic ref'rigVratiotn. It IS expeCted th~ hat ltc I.eXChnger'development problems can be solved with further effort.
81N OIRA L 0IL|¢TRI C
Section 2
"INTRODUCTION
The application of superconductivity has been considered to reduce thesize and weight of several types of electrical equipment. Since supercon-ductivity requires temperatures in the range of 4" to 10'K, cryogenic refrig-eration must be provided. A reduction in the overall electrical system sizeand weight can be realized only if the uize and weight of the refrigerator issufficiently small. At present, there is a need to reduce the size, weight,and maintenance requirements of cryogenic refrigerat:ors, and to improvetheir reliability. With such advancements. the promise of superconductingapparatus and other cryogenic devices can be realized.
OBJECTIVE
The objective of this contract was to advance the development of high-reliability, maintenance-free, cryogenic refrigerators. This advancementwas to be accomplished by the design, construction, and expem imm-ntal evalua-tion of a 4. 4'K refrigerator, based upon the Claude cycle principle, utilizinghigh-speed compressors and turboalternators with all rotating parts suspendedon gas-lubricated bearings.
The tasks that were originally to be accomplilhed under this contract
are outlined in the following four line items:
Line Items 0001 and 0002
* Conduct refrigerator cycle studies.
* Ihusign, construct, and test 100 N turboalternator.
* Dlr',ign, construct, anti test tIronstfornier,-rectifier.
9 Ass(-1bl), and tcSt 801K ricfrigerator (two-stage reversed" Line Brayton cycle).
Line Itenms 0003 and 0004
* I)Dsign, constrLctl, and test. two re•elerative motor,-compressorsfor the first and second siages of a three-stage system.
* l)eslgn. construct, and test two power conditioners and transformer-rectil'iers for the two now motor cor•npressors.
o Modify 10"K and 80"'K turboalternators I.t.) satisfy both the interimand final 4.4'K refrigerator specifications.
9 Design, construct., and test a 4.4'K intir, im refrigerator usingpositive displacenmnt compressors.
Preceding page blank]-3
GIN IEAL O ILICTRIC
9 Construct a final 4. 40 K refrigerator by replacing positive displace-ment compressors with three single-stage dynamic compressors,
As the work progressed, these original tasks were modified, au will be
described in the following paragraphs.
BACKGROU
The present contract was preceded by an earlier Mobility EquipmentResearch and Development Center (MERDC) contract (No. DAAK02-68-C-0320),and was conducted in parallel with an Air Force contract (No. F33615-71-C-1003). The technology and some parts and equipment from both these con-tracts were utilized in the present contract. In this report, references willbe made to reports from these other contracts (Refs. 1 through 5).
As the presient contract was originally constituted, the system was tohave included three helium regenerative compressor stages, two turboal-ternators, and a set of five cryogenic heat exchangers. Of these components,one compressor stage and one turboalternator were to have been made avail-able as Government furnished equipment from previous MERDC contracts.The set of heat exchangers were to have been obtained from a previous GeneralElectric funded development program. The refrigeration capacity using thesecomponents had previously beon estimated to be between 1 and 2 watts at 4. 4K.
When detailed cycle studies were performed early in this program, it wasfound that the system then being considered would produce only a fraction ofone watt. In order to increase the probability of reaching the required tem-perature of 4. 40 K, it was conside red necessary to construct a system withhigher refrigeration capacity as its design goal.
A number of system changes were considered to improve capacity. suchas:
a additional regenerative compressor stages
* positive displacement compressors for higher pressure ratios
e heat exchangers with higher effectiveness
* different cycle arrangements
During the time these systems investigations were being carried out, allwork on the contract was stopped except for cycle studies. A stop work orderwas in effect from 31 March 1971 to I November 1971.
After consideration of a number of alternatives, the final system selectedhad a larger refrigeration capacity (3 watts at 4. 40K), six centrifugal compres-sor stages (instead of three regenerative compressor stages), and three turbo-alternators (instead of two).
"4
SINURALO*ILICTRIC
Ixte in 19'73. MEWi)(' dvci df~d to cut havi on the q.wopi of the contract,and to eliminate most of the work in Line Itenis 0003 and 0004. In December1973, the contraict was imodified to cover Just the t~oilowing taisks:
9 Conduct refrigerator cycle studlies.
0 Design, construct, and test ai 10'K turhualternatnr.
*Design and construct a cryosect ion consisting of a Net of heat ex-(,hangers, a .Jcule-T1honison loop), at vacuumi vessel fur thermal in-4ulati orl, and necessa ry pipifing and instrumnwtatn ion.
9 Test the turboalternator Lind thP Joule -Thomison loop tit final oper-ating temnperatur-CS, t.S[Ing Hliuild nitruigoil rot, precooling, and apositIve displacement ciimpre.ssol,
rhe heat exchanger described in Section I was de~igned and constructedby a vendior who tised a niovel approach ri.r design aind cinstruction. Unfor-tunately, not till the cieveloptneit problems had bet' n wo-rke-d utti prior to con -struction, and a numube r of tinetpeciecI di I'fk tilt. les were encountered by thevendor during constritc'tion. Tilt, exchanger w delivered almost two yearslate. and upon testing, it wais found that the thermal effectiveness was too lowto permit cainidown nnd test of thr tiirl.loalternntor near its design temperature.
ACCOMPLISHMENTS
Thu. w't rk de scribrlhc ini this i*e po i-t ril va ii ced the k developmient of c ryogenicrf'frige rat ion NV Mu~sing ni ii LIt ure hi glii-speýed I urbon Ite rntot'B with self -acting gos hen ritig:.
Designs we re de'rivvd tfor s ;'1 em itl that tcendvd towa rd nitiinlnum weight andin put po wer-, w ith in cc H a ii (on si rtli ts 11 comiponient sizes andlc ret'rige rat ioncapacity. System cit igns arv r dPest HiI.ed tinder Sec In, 1 trieattonSystem."
A I U rbouiterndiatir' wa c ~iiinstrictecl aLid was I e.ted tit I ~temperaitures in therange of 1100 to) 1001K. EIx oisvt nivi 'a u oviii nt s of' pe'rio rmnnete parameterswere mande it. thos, tenipprtuire,4.
Ti de itCcm ini' the hi-havior oft h urunliv hirorilentor ot even lower teniper-atu rex, Ui Ut it a ni W,(8 vl4tV'~tPI using he ickni ir0C LIIpatSx 01ta were tested underthe NI32Ri)( c)nt ract, wi-th the v~cepti on WPthe t urbine nozzle and the alter-nator. ThiiIs un it wan, t esi.ri unir'r ti p rog rnn spon sored by t he AdvancedResearchi P roj.ct s Agenicy tn a terniper0t nre or' 9.80K (poss illy the lowesttenipertu rt' at whichh a tuborh'aliernwtor w itli self-acting gra s bea rings hadbeen tested). Op)t atIxs or I uriwda Itcrutini con si rod inn and testing are givenin Sectljon 4' oitltern r
'Pllv cr vogeit'li Ieat &xli14i'd\ e0 pmiinit was, iml holl'y c'onpleled underihis cunt maci. I lowe\ cita tuiiwi r Id, dfeA igo and)( 'olsis ruet0 Jon tulvaiire's were
OINIRAL49 I LII!TIC
made toward the development of a cryogenic heat exchanger that has thepotential of sizable reductions In size, weight, and cost, compared withpresently available heat exchangers. Details of the heat exchanger develop-ment are given in Section 5, "Cryogenic Heat Exchangers.
6
ISINIRAL@ ILECTRIC
Section 3
* REFRKMERA71ON SYSTEM
SYSTEM DESRPN
ORIGINAL SYSTEM
As discuss.d in Section 2, "Introduction," the original system that wasto be constructed under the contract consisted of three regenerative corn-pressor stages, two turboalternators, and a set of five cryogenic heat ex-changers. The detailed cycle analysis conducted early in the program shuwedthat the refrigeration capacity at 4. 40K was only a fraction of one watt, rathertha*i the I to 2 watts that had been estimated in earlier, less precise caleula-"tions. With a predicted capacity of only a fraction of one watt, the risk of notbeing able to reach the desired temperature of 4. 40K was considered too high.Therefore, means to achieve higher capacities, using as many already avail-able components as possible, were sought, Some of the alternatives thatwerp considered are discussed in the following paragraphs.
SYSTEMS CONSII)ERED
Positive Displacement Compressor
One means of increasing the refrigeration capacity Is to increase thepressure ratio of the compressor. The addition of more regenerative ecor-pressor stages was considered to be too costly; therefore, positive displace-ment compressors were considered. The compressorrs had to be noncon-taminating, small, and lightweight. MERDC personnel carried out this in.vestigation. In their search, no positive displacement compressors werefound that were compstible with the system goals of small size and lightwe ight.
llybi'id Cycle
During the investigation of positive displacement compressors, a cyclethat takes advantage of higher pressure ratios was considered. The Claudecycle results showed that higher pressure ratios benefited the cycle perfor-mance because of higher Joule-Thomson coefficients, but at the expense ofdegraded turbine performance. A hybrid cycle (Figure 1) wris considered inan attempt to overcome this problem. Each of the compressors shown isassumed to be a single-stage, positive displacement compressor. This cycleis a combination of a two-stage reversed l'rayton cycle and a ,1oule-Thornmoncycle.
L;
SININAL ELISTRIIC
tippe
I~Met lywn
Th..mini hli'l
r r,
Hevat load ,~I~
Figure 1. Hybrid-Cycle System(Schematic Diagram)
In the analysis, the pressurp levels were first established arbitrarily:
9 Compressor suction -- 1. 2 atm
e Intermediate -- 3. 6 atm
* Compressor discharge -. 5. 0 atm
Heat-exchanger pressure drops were ignored.
Next, the flow rate through the -Toule-Trhombon loop was calculated to bir2. 14 g/sec, to produce the required 5-watt refrigeration capacity at 4. 41K.Next, the heat that must be remove(] from the Joule-Thomuion loop at the 16"Kcooling station was calculated to be 16.0 watts, The corresponding heat loadat the 75()K level was 38. 0 watts. A summary of heat load.-. is given in Table 1.
Next, a computer cycle calculation was carried out for a reversed P3ray-ton-cycle system with the total heat loads given in Table 2. The flow rate forthe reversed B~rayton cycle was 11. 7 g,'sec, giving a total flow rate in thefirst-stage compressor of 2. 14 plus 11. 7. or 13. 8 X/sec.
r'able I
HEAT LOADS FOR .IYBRI-CYCLE SYSTEM-"-'• =Total Load
_ •seul joule-Thomson an Reversed
rTemperature Capacity Load traytn Cycle
(01K0 (watta) (watts) (watts)
S4.4 5.0 2 .
16.0 0.0 26.0
75.0 56.0
The power inputs were:
staie kw
First 12.4
Second
Total 14.0
This input power is about *even percent lower than the power for tin
equivalent Claude cycle.
Another cycle variation that was considered, the "split cycle," is shown
in Figure 2. This stem is similar to the hybrid cycle above. flowever, in
the split cycle. the refrigerant fluids in the two sides of the system are en-
tirely separate; the only cummunication between the reversed Brayton and the
Joule-Thomson cycles to by means of the two counterflow heat exchangers,
which transfer hent from the high pressure stream of the ioule-,rhomson
system to the cooling stations of the reversed Brayton cycle.
There are several advantage- of this arrangtment, a'irst, the crm-,
pressor for the Joule-Thomson system provides the mjotive force for cir-
culating the helium through the cooling pasSages in the prilnary heat load.
This arrangemctt is less complicated and rnorc reliable than the use or a
pump for circulating a secondary fluid at the 4.4 0 K level.
Another adantage is that the helium gas in the reversed Brnyton cycle
system can be completely sealed and maintained at a high purity level. high
purity Is particularly needed in this part of the system berceuse of the close
"tolerances in the gas bearilng. Opening up and recharging helium in the
primary heat load would not affect the helium charge of the reversed liraYton
cycle.
SININAI* ILICTRIC
[P
Jowl# fllo'd 1
Figu|,o M~re 2~. S'ptt-ycl Sytem*. (SheatcDigrm
A•'M, further-)1 adanag is* tha th ubmcieyo tervre ryo
os I n mth r
Iwacocuethtnihr thew, spi cyl o h rdccewudb
I 5.1 55 5l*gIt l I DI
siuula of satisfac v p osi tive dis
tiona comreso stages. ps.
After inv ure 2. splt-Cycle System (wtchematic Diagram) --c
A further •dvuntage is that the turbomachinery of the reversed BraytonIcycle is compatible with low pressures, high volume flow rates, and low pres-
sure ratios. In contrast, the Joul -Thomson cycle requires high presy sure-ratios nd oenly small volume-flow rates g these requirements are more com-patible with positive displacern..nt compressors. The optimum system may.
pherefore, consist ae , turbncopressors in the .eversed hirayion cycle andpositive displ~wernent compressors in the ,loule-Thomso)n cycle.
It was concluded that neither the split cycle nor the hybrid cycle would besuitable because of the unavailability of satisfactory positive displacementcompressors iind beca~use o•f the extra weight and size associated with addi-tional compressor stages.
FINAL SYSTE•M SE1LECTI1)N
After investi~latton of systems with 5- to 10- watt capacity, it was con-cluded that contract goals could be met best with a Claude cycle system of3-watt capa~city at 4. 4K. Instead of the original three regenerative com-pressor stages, the compressor selected has six centrifugal compressor
stages. Two rentrifugal stages are mnunted on a single shAft, one on either
10
GIN IN At* SLICT RIC
s ide of the driv~e niot ov. The re will, Ilivrvh re, I4he tli-ve tonipre sso r mo dule s,A
with two com~pre ssor stages~ pet, ni o iii! Th ''l el rvilif'ugal coni prv~ssr SCSwere
chosn i pIcc'of the regeneerLiive c(Impressors LwCLIuse, in larger sizes~,centrifugal compress;ors are mnuch nimre efficient ihan regenerative compressors.
In place of the original two turbouilternators. there are three. The num-ber of turboalternators had to be increasod to three in ordler to increnfse sys-tern capac~ity without resorting to a larger turboulternator frame size.
The heat exchainger selected consists (if seven heait exchanger n-odules.
A schematic diagram of the final systeni 4elected, based on computerRun 479. is given in Figure 3. JDetnils (if Ihe system analysis, and the baqisfro the final systemn selection, are giveni below under CTycle Analysis."
II' .1
1 1,VrI
10F
0 a
-0'
I )iagru ii - - ('n-oarvst nrs to~ Ooni-
putt'r Huni 179)
SEWERALOILcTmIO
CYCLE ANALYSIS
COMPUTER PROGRAM
The computer program. "REFRIG," that was used for cycle analysis, isdescribed in detail by D. 13. Colyer et al. (Ref. 3, Vol. 1, pp. q-28).
o'•, SYSTEM DESIGN cRrPERIA,
The prin'ipal design criteria were reliability and total system weightand size. Other criteria were cost and input power.
The goals t'or total system weight (not including dewar or interface withthe system to be cooled) were:
9 rior n 5-watt capacity at 4. 40K, the goal was 400 pounds or less.
* For a 10-watt capacity at 4.4 0 K, the goal was 500 pounds or less.
For destin. the maximum heat-exchanger effectiveness was set at 08, 5percent, a conservative design goal giving reasonable assurance that the per-formance could be achieved.
The centrifugal compressor impeller tip speed was set at less than 520rn/sec (1700 t't/sec).
The compressor suction pressure was set at above one atmosphere.
At the time the cycle studies were being made, there were two turbo-alternator frame sizes being designed under contract with the Air Force(Ref. 3). (The "small" and "large" turboalternatnr frame sizes in this sec-tion refer to these Air Force sizes.) It was desired to design the MERDCturboalternator to the small frame size, which is the size previously designedand constructed for MFRDC (Ref. 2). It was believed the development riskwould be higher with the lnrge frame size.
SYSTEM WrINIT
A complete analysis of system weight was impossible without knowing theexact system design. However, an approximate weight calculation was de-veloped and applied to tho system calculation as a mneans of evaluating designs.
An appr oxmate weight was first developed for all comx)nePnts, based onelectrical power, input to the conditioner-controller. rhese weights wereestimated by (leneral Electric engineers, based upon compressor modules ofabout a two-kw power level, each module having two compressor stages andone motor:
12
rIAoI INI
Comnponent Weight
Compressor subqystem (two stages in each miodtile)
Power conditione r- controller 4 lb/ kw';
KMotor 5 lb/kw
Compressor (frame, impeller, etc.) 7 lb/kw
Aftercooler (for transfer of heat from heliumto cooling fluid) 2 Ib/kw
Hleat rejector- (for final rejection of heat fromcooling fluid to olt') 71lb/kw
TIotal compressor subsystem 25 lbfkw
Heat-exchanger core (Calculatedby computer)
fleat-exchanger headevs (00" of coreweight)
Turboalternator (including piping and filter) 10 ib/turboal-ternattor
Dewar (not included in system weight calculations) 70 pounds
It is apparent that basing the compressor subsystem weight purely on thebasis of input power is a crude appr'oxinintion. A report by P'. (;. Wapuitopresented much mor-e detailed comprr'ssor weight vstimates (Ilcf. 6). Insome cases, the Wapato repor+ estimates differed cons ide rably- from thosegiven abo)ve; In those instances averagevs wore generailly ttoken bvtween theestimates by Wapato and by the General El1ectric Company. In the sumnmavyof the resulting weight calculation, which appears below, the following symn-bols are used-
Symbol Meaning_
N Numiber of compr-essor modules, eaich having t wo cent rifuga~lcompressor stages aind otir motor.
P IE'lect rical inp~ut power to power conditioner-controlievz, forthe itli compressor module (kw)
W Weight of a single compressor module., with two conmpri.essionC ~stages, n~ot including moutor' (pounds)
W Total weight of compressor nfterc'oolers ~ind heat rejector,h fur one compressor module (pounds)
'Assumes conversion fromi 60 to 1500 hertz; also assumies water cooling,(Estimnated weight for air-cooled power vonditioner-controller is I I lb/kwv.
AIN IRAL ILICTRIC
Sy Meaning
W Weight of motor, in one compressor module (pounds)m
W Weight of power conditioner-controller for one compressorPmudule (pounds)
W Weight of total compressor subsystem (pounds)t
For the motot, the weight was calculated to be:
"W Wm = 4.0 P m/3 (1)
The compressor weight was:
W 20 (2)
The weight of the power conditioner-contruller was the same as above:
W 4 ;: (:4)p 1'
The aftercooler and heat-rejector weight was also the same as above:
Wh = 9 P1 (4)
Adding all these weights torether gives, for the total weight of the com-pressor subsystem:
W 20 N + (4.0 P2/ 3 13 Pi) (5)
This equation was used to ('stilmati: system weights in the analysis thatfollows.
RESULTS IV'O)R 5-WATT SYSTI:V
A numbr' of systems were Atudied, with refrigeration capacities of 5watts at 4.4-K and secondary loads totaling :30 watts at higher temperatures.A summary of the results is presented in Table 2.
The total system weight i- ihe sum of the compressor subsystem, the heatexchanger core weight with ten percent of the core weight added for headers,and 10. 0 pounds for each turbine (including piping and filters). Dewair andradiation shield weight are not included in the system weight because theseweights will be shared by the system to be cooled.
As meil toned previously, hliere are two turboalternator frame sizes,large and small. The frame size can be determined by Figure 4, which wastaken from ( olyer and Oney, Figure 4. 2. 3-1 (Ref. 1). A large-frame
1 4
G SIN RAL, ILLCTRIC
iI
E--
Ii }
IA~ IM t t
,• : " • .. .1 .! . .. .•. . • . ! . . - 1 .
. .. .w .q .r .w .s .e e .ur... w. ... .I
.
U NI~ ~ ~i
*• £15
OINERAL * ILICTRIC
Q If (S 5
a Run ul165I U Run, 10.
- - - 32.
509
, I
~-20--
* -- - - f- -T-
Tr i
50 too "D0 300
Speed ithouumrnda of rpim)
F'igure 4. Turboalternator Cr ' ogenic PerformancePower Output Versus Speed
turboalternator is considered to be a unit whose speed-power characteristicfalls abo%e the highest curve in the figure, while a !inall-framr turboalter-nator falls on or below that highest curve. l'or example, as unit with a spOE'dof 1 8 0 , 0 0 0 rpm and a power output of 50 watts would require a small frame.-while a unit of the some speed with a power output of 100 watts would requirea large frame.
Many of the coimputer runs were masde In attempts to produce a systrnldesign with all turbines of a small frarne size. This was foiund not be be pos~-sible with the refrigeration capacities hosvn. In all runs, [it least one of thethree turbines exceeded the small frame size.
16
-4.* 0.'.-
-INIIAL *IE@ CTIIC
It was concluded, on the basis of only a few runs, that there was no ad-vantage in four-turbine systems and that the additional complexity was notjustified. Two-turbine systems (Runs 414 through 432) showed larger systemweight than three-turbine systems. The choice was therefore a three-turbinesystem, with either four or six compressor stages (two- or three-compressormodules). Most of the runs were made on three-turbine systems.
For systems with four compressor stages and three turbines, the opti-mum system was represented by Run 404, with a system weight of 530 pounds.
For six compressor stages and three turbines, the optimum system wasrepresented by Run 433 (system weight 451 pounds). This system comes closeto the goal of 400 pounds for a system having a 5-watt capacity at 4. 40 K.
Off-Design Calculations
Present computer programs are 'decign" programs in that, generallyspeaking, system performance is specified as program input and componentgeometry is calculated and presented as output. A true off-design programwould do the opposite: It would accept geometry as input and would calculatesystem performance over a range of values for various parameters.
Such calculations were beyond the scope of the present contract; however,pseudo off-design calculations were performed using the present design pro-gram. Starting with the system of Run 433 (six compressor stages and threeturbines), the heat-exchanger effectiveness was lowered to 98. 0 percent todetermine the effect of missing the heat exchanger design goal of 98. 5 percent.At the same time, in Run 442 (not listed in Table 2), the two coldest heat loadswere lowered to a level that would maintain the same total helium mass flowrate (10. 4 g/sec) as in the original design (Run 433). The resulting heat loadsrequired to maintain this flow rate were 3 watts at the 4.40 and 120 K levels(compared to the original design of 5 watts at each temperature station). Theheat exchanger and turbine designs of Run 442 changed only slightly from theoriginal design of Run 433, and the compressor design was unaffected becausethe flow rate and pressure ratio were the same.
The same pseudo off-design procedure was applied to design Run 404 (fourcompressor stages and three turbines). In Run 445 (not given in Table 2), thetwo coldest heat loads were found to drop to 2. 8 watts each (from the originaldesign of 5 watts) as a result of the heat-exchanger effectiveness droppingfrom 98. 5 percent to 98. 0 percent.
In Run 451, the same off-design procedure was applied to design Run 404,but this time the turbine nozzle coefficients for all three turbines were loweredfrom the design value of 0. 90 to 0. 85, which dropped the turbine efficiencyabout five percent. The two coldest heat loads were found to drop from theoriginal 5 watts each to 3.8 watts.
Likewise, the same drop in nozzle crEffcients was used in Run 459, ap-plied to design Run 433. The two coldest loads dropped from the original 5watts each to 3. 8 watts.
17
GENERAL * ILICTRIC
Next, starting with design Hun 433. the heat-exchanger efficiencies were.lowered to 0J81. 0, and the nozzle coefficients were dropped to 0. 05 at the sametime, in off-design Run 466. The twu coldest loads dropped from the original5 watts each to 2. 2 watts.
Likewise, starting with design Run 404, the same parameters were changed
the same amounts in off-design Run 470, and the two coldest loads dropped to1. 98 watts.
It can therefore be seen that systems with four or six compi.essor stageshave about the same sensitivity to changes in heat exchanger and turbine per-formance and that the design for 5 watts at the coldest station appears ade-quate to maintain some capacity even in the event of depressed heat exchangerand turbine performance.
RESULTS Fel1 3-WATT SYSTEM
All attempts to design a system with 5 -watt capacity at 4. 40 K and withthree turboulternators of small frame size failed. It was found that thecapacity at 4. 40K had to be reduced to 3 watts in order to achieve the desiredsmall frame size for all three turboalternators,
Figure 5 is a printout of Run 471), which represents a system having 3-watt capacity at 4. 40 K. There are six compressor stages (three compressormodules) and three turbines. All three turboalternators are of the smallframe size. This run was selected as the system design basis. (See Vigure3 for a schematic ditagram of the system.
The Run 471) system has a compressor subsystem of thre'e modules, with
two centrifugal stages each (a total of six stages), and three turboalternators.
"The total system weight was calculated to be 386 pounds:
Component Weight (pounds)
Compressor subsystem: 262.7Hent-exchanger core B14. )Ileat-exchanger headers (10% of core) 8.5Turboalternators (including filters and
piping), at 10 pounds each 30.0Total 3116
The dewar and radiation shield weights are not included in the alove calculation.
.efore Run 479) was finally selected, a number of computer cycle calcu-lations were made in an attempt to optimize the system. lV'irst, the temper-ature into the coldest turbine was varied. The results are shown in FVigure 6,*Using E'quation 5
1 8
P. 1lh~t
OINIiAL@ EILEIC
RUN
REcaI S 1 I 8371ST 19S15171
RUN NUMBER 479IT IRAT I8 NS * * * *l 9t * ,i * 9 9 * * * * **,I* * * e
041 lee84
..IIENE SEQUENCE * 4 It 3 a 1
LSOATION LOAD(W) TCMPCK) FLOWiStgtC)S3.0000 4e400 0106S3.000 14.00 34360
3 5.000 5MO0 134664 M00 17000 0.7343
OOYLE DESIGN POINTS ? YES
Loco POAWN) TNPCC) L.C, PCATM) TEcPcK)26 I 1.00 5.747 I a 1.180 4m406
3 3 33174 4-400S3 11268910 13-84 l 2 8.814 31.57* 3 3:800 5,747 2 4 3.374 49400
.1 5 116 11.45 a 6 1,157 12,95a 7 a864 14003 1 1:856 54.35 3 It 2s642 491173 3 2.58 13.54 3 4 1.157 120953 5 i1145 48'73 3 6 1.139 51473S3 7 1.64* S5.004 1 21655 166.1 4 8 1.873 15S.3S4 3 11.856 54.35 4 4 1.139 51*734 5 Ao.86 353.6 4 6 16129 163.54 7 11871 370.05 1 1.914 335.0 5 2 16085 166.15 3 1.182 363.5 5 4 10100 330'4
TUR9IALTIRNATORS 7 Yts
0""......?UIRSSALTEatNATgIts-........ -SICe DIA(Ir)' PWRIW) IPrIC. RP41 0.5000 17,14 0.3730 318021.3 0.7000 46*47 0*3746 16169504 1,000 61*89 0*3061 1867650
CR-41 321
Figure 5, Design Point for System with 3-Watt Capacity at 4. 40 K (Run 47D)(Sheet 1 of 3)
In
, BINIRAL I*ILI CT IIC
LfC. AOMeFRACT. W*LOSSCW) FL@#rAc* SP*VtLoRAT*2 0.3056 0s,7241 lases 0.37733 0.1670 1,701 1.378 0636094 091074 4.772 tells 093340
1,609 ToIRGQ*N19 ToLD CO. TsM1I4 P'.(It) BRG.PWR*CW)9 09379OP-01 0.13682-0t 0.I1601-02 0.36473 0011169 0.14459-01 001160CE02 1.05584 063117 0 1681 Coal 0.11601-08 4.102
LOC. J.RG.NI. JsLD CO.s 9LHTbs1.C UT 9D IA.a 0.1663 0.41411-01 2.0653 I'.67 7 0 a49 59C40 1 61,2104 2s000 0,5561C-0 1.350
USE WROWIR PRO0GRAM ? YES
mo-m---HE--4AT 1EXCHAMGER DESIGr4-amoom---LOC.e WTCLBS) L ENO 'I4C CH) V@LCCU*.FT.) EFFECTa8a 1*243 11650 0.80351-01 0.9650a U 0.1509 1.395 002469C-00 0*91233 B 18.90 47.32 0.2604 0.98513 U 0,8474 4.017 0.1613t-01 0.90354 1S 99910 58.51 0.6707 0.96494 U 1.671 4.629 0*35866-01 0.3736S a 36.519 60.68 009353 006151
TOTALS 84.91 1.*944
CENTRIFUGAL COMPRESSOR? YES
owes m-----CENTRI FUGAL COMPRESSR-mom am so-
POWER IO 0040IT-CONTROL CKW) 13.12AMBIENT TEMPERATUREC CK) 328.0PRESSURE RATIO 29649MASS FLOW MOME) 7.686
MODULE IN* PGWECR(KW) RPM MST9SPDCFMS1 4.6819 0.193001.05 667.8a 4e391 049I00Ec*05 667.83 3.903 0*9100E.05 667.21
BEARINGS
F'igure 5. Design Point fo~r Systrim 1-ith :WEttCapavityat 4. 411K (Win 470) (Shoot 'I (f 3)
20(
BiI# UAL* ILI CYRIC
t4IDULt ToSR~oNfs. T*LD@CG. T.MIN fro(INI 9RQ*PWIR.W)160332 0.46161-0! 0*13231-02 453.7
2 4.435 0*33459-01 0*13231-02 474.23 3o274 0.25859-01 0*13231-OR 517&4
KGDULIE JeoRG*NS. JoL~eCts JeNIM IP.(IN) CRII.SPICIDit 0.000 0.*7 6051C-01 0.12231-iOg 0
a 12400 0.a55071-01 0.o121831-02 03 1.000 0.'42371-01 0 e11123102Ol 0
STAGEC SPoSP99D A IRO *IMrs TIP 3PD(F'PS DIA. (IN)1 0.43351-01 0.6367 1696. 4.-271
a 0.40601-01 0.6153 1696. 402713 0.39699-01 0.6077 1619. 4.3004 0.39011-01 0.*602111 1563i 3o9365 0'38639-01 G,5993 1501. 307796 0:38429-01 0.5964 1441. 3.698
STAGE PRISS.RATIO CSSLstIFPCTV91 1.214 0.7669a 191104 0.77653 10184 0.76204 1.166 0074695 Iets5 0.73116 1.136 0.7148 C R- 41 J2
Figure 5. Designi Point for System wvith 3-Watt Caparity at 4. 40 K (Run 479)(Sheet 3 of 3)
14.0
13.5
O 13.0CH- 4144
12 113 14 1Temperature, Inle~t to Cul.Wst. Turbfine (OK)
P igur~e' 6. inpul Power V \ersi~us l'urbine 11110 1Temnpertl tire
21
46N IRA L ILICTRIC
whipch indicate*s that the tvfrnriieratwrc of I 4ý'K is~ optimum from uri input power-Istandpoint.
Scvs~rul additionail cWi1.-uIlat ions wfrr* mrride w~ith v. r'ititintis in the warmert wo turbine terntwrrmtures. While slightly irjwterr Input powers~ (:o)Uld be achievvdby c~haninng those3 lemperaktures, chaingesx in tt'rnper~iturus in u direction touttain lower input, powers also ctii~i~ed vl. Irqitit one turbine to oxcoorlrr the maxi-mum power lcvr.ýl of a turboulterniutor with synall frumuo size. The require-mrent of iHmiiI Vrrunno size therefore impo~ii,- the t-emeli)ratures us4ed in JUun 479)for the %varrrflt: two turbinf.s.
Off-lfrsdgn CuLCulations
P.'-t-.Udo Ofl-4tPNlgri VIACIOU1Lt ons w'.rr, tippl k'd ito ritsign-v prtl. R~un 479, todetermine In , .rt'fec~t or (1 radv~ir Pc tforrmitice or cryoLit~nv leIit-o Icxc~hange rsa.nti t~urboulterna~t~ors Tublhe N1 s4ho WN i Ii vt tilts N f rus i ~gn-pt i t RIun 47) anmdthree off-ckes 1ri-pcolat rufns. The tot ul f'low raite thut wuis hrilrI tipprox iratelyconst unt toir u Il t heos run s was t I I celrit - po int v I uc of 7. V9 g /, ~c cl
Taliit. 3i
(iiPi,:ii ANDi, CO-
I'lvIIItt f ,' 1 7'L 1: ; 1I l I ,.w I I' h t IllI,111
4O - J1t14 114i t lilt 1 1, 1111 il w.l i I f..i clorIfitIt-fit
PilltiV-itE H
1, It I r i l 'I
'lihe (it-of) it) turtbinr -tt'~io./A tctlOic iinti From (1 "0 111 C. 117 t~iii vacl 'iqrox i-mnutel~y it five-ptcret-tt. drop In ovvrtilI tit'litie ellaitviri ' it. A cani It; siten 1ritnithe tablilr'1 cvcrii- in thf, tit ike ly c vittit I ha hoht u rbtiiri uirl hlmvi l.-(lc'xlarmjr 1pu r -rortinace weri ilegrarlec by thru 111tlnmtut Hhilwn. thic'r woitilr sl tl bit, wiort,t~han one watti ill r0'rgij.-ration cCaticlllK Lit. elitah or II( hewO ctiddest rel'rig.ertition
14t5.ii~fl22
*IlIRAOI IIlCTIICO'
(PER1MEAL SYSTEM DESIGN
GENERAL
The experimental system included only the coldest turboalternator and aheat exchanger that was to be cooled by a liquid nitrogen stream that take.the place of the warmer turboalternators, as shown in the schematic of Figure 7.
The design goals for the experimental system that were used as guidesduring the design phase of this program were:
s Refrigerator Maintenance-Free Life. This goal was to demonstratethe potential for long life (ultimately perhaps t0. 000 hours or more).The contract goal was given as 2500 hours,
* Shock and Vibration. The refrigerator was not to be subjected toshock or vibration loading, except during normal handling and ship-ping while the refrigerator was not operating.
* Acceleration Loading. The refrigerator was not to be subjected toacceleration loading beyond normal gravity loading; however, to beconservative, all structural design and turboalternator bearingdesign were based upon a steady acceleration loading, in any direc-tion, of twice the acceleration of gravity,T Temperature Capability. The unit was to be designed so there wouldbe at least a 0. 8 probability of reaching the design temperature of4. 40 K, with no useful refrigeration load at that temperature,
RADIATION SHIELI)
Calculations were made to determine whether the thermal radiation shieldshould be thermally connected to the 1700 K or the 550K refrigeration load sta-tion. It was found that if the shield is connected to the l170K station, as muchas 130 watts could be radiated inward to the lowe- temperature regions of thecryosection; this condition would hnve necessitated multilayer reflective in-sulation both inside and outside the radiation shield -- an undesirablecomplication.
With the shield connected to the 55OK station, the inward radiation wasfound to be less than 1. 4 watts, eliminating the need for insulation inside theshield.
With the shield at 550K, the external heat load on the shield was conserva-tively estimated to be less than 4 watts, if a 5-cm thickness of reflectivemultilayer insulation were used outside the shield (such a thickness wouldcontain 60 to 80 layers of insulation). This heat load would be the majorconstituent of the 5-watt load at 550 K, assumed In the cycle studies.
23
GININA L ILICTAIC
P4 4
IL~
2I-1I
01INlAtO IhlCTOIC
(•RYOCX1ENIC PIPING
1Y. Cryogenic piping is required to carry gas between turbinew and heat ex-changers with an acceptably low pressure drop. Calculations were made forpipe diameters, with the following assumptions:
F Ior each pipe carrying gas either to or from a turbine, the ratioof pressure drop to average pressure must not exceed 0.001.
- Pressure losses caused by wall friction are negligible, comparedwith losses caused by bends or elbows in the pipes.
o In each pipe to or from a turbine, there are four right-angle bendsand a total pressure loss corresponding to four times the velocity head.
These assumptions are somewhat conservative because in some pipesthere are fewer than four bends and there is probably less than one velocityhead loss in each bend.
Results of the piressure-drop calculations showed that a 1/2-inch nominalpipe size, schedule 5S, would te RA^qliately large for all cryogenic piping,from a pressure-drop standpoint. This pipe has an inside diameter of 18.0mm (0. 710 inch) and an outside diameter of 21. 3 mm (0. 11410 inch),
Because this pipe size is reasonaibly small, yet large enough for inech-anical support or the turboalternators, this sive was tentatively selected forall piptig to and from tile turbines and to and fromn the. loile-Thomson valve.
JOULE-THOMS)ON VALVE?
The simplest approach for the construction of a Joule-Tthornson valve isto use a standard needle valve and to modify it with the addition of a long stem
sealed within a vacuum-tight sheath. The valveO could thereby be maniallycontrolled nt the warm end. The stem screw threads wer(., inltiall.h locatedat the warm end, but in a later design change, the iliveauis were relocated tothe cold end.
Choked flow of an ideal gas through a sharp-edged orifice was assumedfor sizing the valve orifice. A Whitey C'ompany valve. Model 2H IP4-i316, withan orifice diameter of 2. •6 mmn (0, 093 Inch) wa. selecled. An orifive of ihissize, with an upstream temperature and pressure corresponding to cycle de-sign-point conditions, would result in a calculated helium flow rate of 6. 11 g/sec.
The cycle-design flow rate through the Joule-Thomnson valve is 2, 1 g/sec.It it believed that this ratio of flow rnaeý'- Is udequntely large to assure controlat the design point, despite uncertninties In the calculnilon (the grealesi un-certainty of course results from the assumption of an ide4l gaH). 'he vn•lvegives nearly linear' control from the closed to the fully opened stiem position.
2D
rGIN IRA L* ELECTR ICI
A - in d i ctI vd aibt ive .n iq 1tqiI i ( ) 11 rI) cuo I e I' was vv t~~~Ivli Ito I ak v I hi 1) 1i((c
(if the wan r-st'~ heai't txc'hm gp tr.
T1he nifethlicld HeeIU Ne( III -,wvuiilIHIIS I h1is Wh~ile~ attalininlg vi simvil I tl'Pfl-vC1r.0
is to tll bl bUI ll('t tile fil%% bvt)PI el~l I Ilk, I \%,I sti'vnin by\ providc ;,ll , II largi' l 11(1w Ilithit- (oldet'v 81'LI?11, This1j uhllulmlif. uml he2 liccl ImIpII ishiec Lisi 8.1)lmii I `igur'v 1).T1he rosult Iin tompert)'t~itri (li~t riIllit h ll j N SHlltWI ill PIlijLtre
AI lilt IItfphtll('
11 ii kimi
T he w vt~lt"194 ill Ihi orrili~l nli'lt 1.ý 1 it , I i cilli d %,1 I'm lit, iýt
l i ft vo III \li it l. 1s h 1trl . ctol h II i . hIIIi lvjI'li'lr t 11 . i (I tic 11w Ii ý I'mll i ll' o. 1,11.4i l'
' ll ItI ' r (111( k ', I% I I It'' ( It I -dSI 'l ll 1111 'It T i. l' w .t I I l l rit( - I 1~'W~ 11;. .11( 1i d
SINIRALO ILI CTI C
I
•rll~Sall Tim l~lpr rall uPi,
I Wtr-revi-L
Cold St riti -ltes Lai-gar
Plow litae
DMatai~i. AkoflM Warmeal Heat Exchansetr LE,-H
Figure 19. Temperature Distribution
Control Valve VacuuWarm Heliumfrom High-Pressure -
StreamC /d
_________._______ ____ ilIuLiquid LL) Het
N itrogenr -
-ii-.. . . . Wo L1 of Vacuunm
-- Veouel
Figure 10. Helium -to -Nit rogen Ileat E'xchanger
that a helium flow i ire through the control valve of O, 3 gfaer provides a suf-ficientlv large flow unbalance between the two strearns. The warm-strennmflow rate (Figure 8) to 5.47 g/sec (the sum of the flow rates throiugh the 1Olue-Thomson valve and the 140 K turboalternator). The c'old-si ream flow rate t,14therefore 5,47 + 0.3. or 5.77 g/sec, 'Tre capnctty-rate rotio, C, is calc'ulatedto be the ratio of' mass flows for uqual hent capacitic,:
C11 C i C 1 - fx.47/5.77 0. 15
27
GI IEN AL4 IELECTRIC
The heat tramsfer eft'ectliveness for a counterfluw arrangement is:
-NTIJ(1 - C)1 -eI: ' -ce - U( - C)
For number (t' transt'ver units (NT LI) of 6b (the design value oft the warmestheat exchingit.), the effectiven.uss is calculated from the above equation to be0. 998, andl tho coild-end st, ream-to)-stream temipirature difference Is cu1eulatedto be 0. 44"k. l'his value is susfiviently amall to match 'he streamr-too-utrevimtemperatur' dil'rtr'vi-ie tit the waril end of the adjacent colder heat exchanger.
COLD ENI) (' )()1 ,I,
A method I'mfr cooling the cold end of the systern was devised. As showntin Figure 7 a sm.ill diameter tube leads helium from just upstream of theJoule-Thomsm i valve, through a room temperature control valve, and finaillyback to the low Itressure side of the compressor system. The purpose ofthis gas olleri was to) remove rd. lenst. part (if rint all) of the warm heliumstream retuviiing thniough the c()unterflow heat exchangers during the coolhlownprocess. Itlli wed to pass thro)ugh the heat exchangers, this warm returnstreum wouldl 1.'ansfer heat, t the incoming cold strean, thereby lengtheningthe ,ooldowri process. Once the h oule-'Mhonn.on valve dropped to below theinve rsion t.-.ii)(1rature and produceod cooling, the hieed sitream valve would betclosed, and iill of the gas woul(I return through thie heat exchange r system.
iIIPA'L' IL4EA K
Tlhe c ryo;(c-Ct i on (lcsign concept was l.( mount ell v'v'ytm(cction compo)neI~tN
(tur bralternmat i-s ,1()ule-'l'h4)mson valve, filters, and radiation shield) dirvet lyon the heat exihlatnger or on the liquid nitrogen cooling• sHttge. Thus, no strtuc-tural supports between room temperature and cryo)eidh temperatures wouldahe needed, a.nd,1 all conducted heat leakage woiuld be thirough tnitrurnentatn imleads and ihr'(,ligh the ,l ouilv-Thomson valve stern.
A sunmalkl'y or calculated heal. lenks is gi.ven in Ta ble 4.
The ahov, heat leaks, do not include conduction thi (ough liquid iou.troge)iitubes, which will he installed to expedite rapid coolhown oftthe vold elnd ol'the heat exhl1ugV1er system.
Trhe buse; 1't )r the heat. leak talc.ulations were:
T The (cIoaxial (cable assumed is Microd(hA., lncirtl )rated, Nit. 250-4011:4.This calv c ('I• ntains sepveri strands if' 0. 004 1 -inch-dhia meter ' ,(lppe rwire ii, the inner conliactor, and 64 strands (of 0. 0033- in.h-diameterc(;)ppe. wire in the shied. A length of i. *2 mete'rs was assiunned.
211
M *eNAL ILIECTRIC
Table 4
ESTIMATES OF HEAT CONDUCTION RATES
Calculated Rate of Heat Conduction (w)Thermal Conductor .-
To 140K Level To 4. 4*K Level
Seven coaxial cables for proximity 0.275probes
Three alternator power Leads 0. 072(includes Joule heat)
Leads for three platinum resistance 0. 056 0. 028thermometers (four leads for eachthermometer)
Three pairs of thermocouple leads 0.020 0. 010
Two pressure lines 0. 007 --
Joule-Thorrson valve stem -- 0. 063
0,430 0. 101
The total of 0. 275 watt given In Table 4 is for all seven cables. Ifthe cables were thermally connected to the 80 0K level, with a 1-meterlength between that temperature and the 140K level, the heat leakcould be reduced to about 0. 13 watt.
* The turboalternator power leads were optimized by the method ofMIcFee (Ref. 8). The conductor diameter was calculated to be 0.41mm (0. 016 in. ) for an assumed length of 1. 0 meter.
It was assumed that two platinum resistance thermometers would beused at the 140K level, (one at the inlet and one at the outlet of theturboalternator) and that one thermometer would be used at the dis-charge of the Joule..Thomson valve.
Copper constantan thermocouples would be mo-inted at the spme posi-tions. These sensors would be used for monitoring cooldown only(their sensitivity would not be adequate for sensing the final steady-state tern pe rat ures),
* Pressure lines were nusumed to be stainless steel lines 1.6 mm.(0. 062 in.) outside diameter and 0. 81 mrn, (0, 032 in. ) inside diam-eter, 1. 2 meters long.
LAYOUT OF CRYOGENIC SECTION
A layout of the cryogenic section showing the location of all cryogenic com-ponents is shown in Figure 11.
29
rr
*EUIRAL* I �
I 2
� U
4 I.-. A
t I- I
4. -�
I ... ...
* -- -- * I. A
� L IL** - iF*�* * it,
" 4 hi
30
0 IN IN At ILICTRIC
Section 4
TURBOALTERNATOR
GENERAL
The turboalternator arrangement used in this program is shown in Figure12. It comprises a higkl speed, radial inflow, impulse turbine mounted on apermanent magnet alternator shaft. The shaft is supported and positioned onself-acting, gas-lubricated bearings. Operating tests at speeds up to 400, 000rpm have been conducted on this type of turboalternator. The bearing, windage,and electromagnetic parasitic losses are reasonably low. The high speed a!-ternator has proven to be of both sound design and high efficiency. Tests withelectrical loads up to 109 watts have been conducted with electromagnetic effi-ciencies of 98 percent. The bearing system allows operation in any orientation,free from gas bearing and rotor instabilities sometimes found in other bearingsystems.
Tih'ust Hearing |ii
aiaiil litt.
'I'il'l~ilA' ISpirai
h n d en arin •upporti t e it h
with an operating pas film thickness of about 300 micrcoinchE..'s. These journalbearings ai'e of the self-acting, tilting pad type. They were incorporated toensure stable operation throughout the operating range and at any attitude.
311
0 IN INAL * ILI CTNI C
Two iflwf ird p1ITniflh.Z spiral gr' ovr'd thrust beairings position the shaftaxially. Like the journal bearings, the thrust bearings are gas lubricated,typically using a 500) microinch filmn thickness.
The entire bearing systemn tosBelt-aligning bectiuse the thrust bearingsare gimbal mounted and the tilting-pad Journal bearings are self-aligning,Satisfactory operation of the cumplete be ii:ng system is independent of theaccuracy with which adjacent parts tire munufactured,
The turbine energy io absorbed by the perniancrnt magnet alternator. ThiscOmpact alternator is the most practical device for extracting energy at cryo-genic temperatures when thnt energy will be expended at amnbient temperatures.The two-pole magnet operates within the stator whichi is wound 3-phase withinlow-lose iron laminations. These laminations have beeri made by conventionaldie punching and by photoetching, a method which eliminates crimped edges.The alternator rotor consists (if a hitgh fieid-strength plat inui- cobalt magnet.The rotor Is constructed by brazing the magnet, to shaft ends. rhe groundmagnet surfaiCe has the same diameoter as the shaft.
TURBOALTERNATOR DESONS
PRELIMINARY l)l,'Sl(iNS
Pr'elin.itnary designs of the three turboulternaturs needed for the refrig-eration system were made (.Apperidix 1, "Preliminary TurbualternatOr IDe~ign.").The final dectigns oif the, two wtirnwr turboulternoltors were not completed be-eause it was decided to manufacture and test only the coldest turboalternator.Tlhe final design of the coldest turboatternator is discussed In the followingsection.
I"INA1, tESKICN OF THlE 1 40K TU RIKAL1TERNATOR
The overall mechalnival configuration design of the turboalternator isshown in Vigure 13 (excerpted from Ueneýral E'lectric I)rawing 58HE477). Theactive components, turbine, alternator, rind bearings are mounted on a base-plate (Part il8), which allowsi simple attachment of the assembled turboalter-nutor to the rematinder of' the refrigeration system.
All enclosure seals are indium -coated C rings. The base material or theC' rings is 30O4 stainless steel, to match the thermal ooe ffivient of expansion ofthe flanges. The low mass of the seals ecoupled with the excellent thermal con-tact between the sual and flange, established by the soft indiurm ooating that es-sentially becomes bonded to the flange suriutres, ensures geometric, tempera-ture uniformity in the seal area during system tvmiperaturC excursions. Rela-tive motion betwern the seal und the flange dwves is thus prevented. Althought ~the thermal coeffiieint of expansion o1' the indium couting is almost twice thatof the base metal, the spring loading of the C, ring coupled with the extremesoftness of the indium overcomes this mismatch.
32
MURALS ELECTRIC
'41
WTI
33
~I GENERAL * ELECTRIC
Axial posi itiftoing of the turN)oalt niator shaft is accomplished by spiral-groove gas bearings. rhe becirings tire self-aligning. because of their gimibalmounting. 'rhe bearing material is beryllium-copper alloy, which is comptit-ible with the nittrided thrust runner at the low uni! loading present (in contrastwith the journal bear'ing pads where the high loading at start and stop, due to
The gro~ves are hotoad.hprethedeang srfc iniie bernsajutline contact between the Journal Lind pd,prcuethue of this material).
meno poviionis adeforseven capacitance -type displacement probes withthe placements proceeding from the turbine end of the shaft:
Probes Function
I Axial shaft movement
2 Shaft orhital movement
2 Shaft orhital (periphiery of thrust collar)
I Jlournail bearing pad movement
I uter thrust bearing movement
Th turbine wheel is 1/2 itich in diameter and io made of aluminum. Thewheel is secured~ to the shaft by an interference fit. Proper selection of thefit dianieters can( wheel shape allows thle afisembled rotor to be operated atdesign tempevature, ut. safe stress levels. Experience has shown that turbinewheels can be removed from tho shaft and can be replaced without disturbingthe rotor balance.
AL4TERNATORi lElSIGN
Trhe alternatior design is simitlar to previous designs but, because of therequired design-point values of' speed and polwev. greater advantage can N,taken of the very high copper CondCIItivity ait tile I 411 inlet temperature thanIn earlier designs. Because of this high conductivity, the armature currentcan be high foi, this alternator, which allows the terminal voltage to b.e low-ered at the reqcuired power. Thus thle flux density can be lowered and thle airgap can be inc reased.
Dlecreased flux density reduces core losses. The large air gap coupledwith a tooth w~idth o)f 0. 050 inch reduces the slot harmonics and ussoeiatecllosses to a nfegligible level.
Table 5 Hiiows I he valrulated features and performance (if the alternutor.Electromagnetv Icffiviency is (estimated to be M). 4 percent, whichi is higherthan the effiriveny for lirfvirus designs.
SIONUAL* ELECTRIC
Table 5
ALTERNATOR I)ESIGN AND CALCULATED PERFORMANCE(140 K inlet temperature)
psr4fot'maUnceParameter (Computer lDesiin
D- 23000)
IStato. -
Outside diameter ipunehinI (in, 0, 1140
Inside diameter (In, 0, 291
Length (in,1 0,416 SMaterial Oin, ) 0,004 Hlymu 10
Number or seots 12
SpLral (dog) . 1
Tooth width (In. 0,050
Cirvuits I
Connection Wyet Pitch (0) 83, 3
Conductors per slot 10
Wire I - 0,0008 In. I I F
Turns in series per phase 100
Pitch factor 0. 9466
Di)itributiun factor 0, 961
Skew factor U. 9•3
S1tacking factor 0, 911
Resistance, 201C 0) 4. 4fl
heakqeg reactance 11) H, 54 et 1400 iit
Rotor I'leld
i)tameter OIn, 0, 261
I,ength (in O, :474
Material tC'. Pt
Air gap (in) 0, 0IN
Perfornmance (10" K)
Speed (rpm) 100, 000
iPower tactur 1, 0
Output (*W 29, 1i2
Iine-to-iine volts -- no load 30, 0
I~ine-to-kine volts -- full lacd 3, 4H
IRelsitanue at I•'N (6i) 0, 03•1
('urrent (anp) 0, 4100
Current density (anRpiin) 11. 100
Core Inam (w) 0, L:1
Copper luNI (W 0, 0.:0
Electromagnetic efficiency 1%) 49', 4
35
GE1NIRIA L 1EL1ECTRNI C
UAS i~A~IGANAL.YSIS AND 1)I1'-'lJN
Thu gat,~ he~aring analysisi anid de'sign is 1wesente d in Appendix 11, "Gas-Bea~xring AmialvsIs and fiesign.
ASSEMBLY PROCEDURE
*IITh pOWrIC-11.1T fo' t a~is c.'rnb m ilt- wccr~i tu-but urratov Ili prese~nted lit Ap -pi'ndix Ill. Fl1itsai ciP-1diX IS rupruitcLvdIA tvwn11 11 Ieparutit ducurnient titledInstructioni VIanuial: 'i'tuhL)Lulttlrtat o Aki.~mblyy Pruce'dures,
TURBOALTERNATOR PERPORtMNCE
'Il It BOA lTF R; lNATiOIR MVAN U VACTI jJjEJ
Di )ri mg Owe ni cll' of this prop rum 1 It. was dee idud to uttlz ii't' l uIt-boul~t(e -Iruitor made ;ilait'ul from a ie r'nitwivi: 11. S.T Air- I'o ee vontruct.;: The im-po)r'tatit dinli- 4 itsiti of the) wheel anrd ticizzlt a re almIiost Identia I'i t) the uni~tdesigiied Ifci Ni ii I ), alnd ctovu kit imis Iticit'at e thot the Air Force turbine wouldI
meet then r'e' lu Irr ncts () f the MEI~R IN( tests~.A vomlp~i rwmcc (it tii' whet'1 arid mozzle civicracterist les fur' the MEH l)C wcnd
Air I Ficr' mnill 14 gi veii fin Tublc HI. TPhe Mir Vý'ifvvc mitt was ojriginall 'Y de -mignied t~wic I\' luldIit, Mie% IW' i miei 14 ILLt. Ini theL [VI i'RC tes ts, only the
('OMPII 'kSi)N ()V ll \ %' Il' , AND)N( /ZI L- ('ILARACT'I'IR STICS
A~r hirco 11.50Cm
HN 'I 0. IM. 1*)I (1-IL1iNimlv Wc~ rc'ic 37i 23Nc'
An ~ 11-w l. auc wla d 0.rpilnc'c BObil 'r l Mil q)UW MMc'oile'mciwtII CI A tt c i' Ii Ir 811 . 41 4 1a thI A I Vor,1 c &l trit Ill)io ' 2 tc1 0.271c~i I i ptn'. i)I~cgf'Lull1 utnpr Ivc ti ItI~'i' -i V1 1.1c( ,u 1c1,~hit iii I ii
lilt' I''or'ir' I ''nxin~, it was used INi I he MII . I )(V chatrcumo'aszlk
(out~~( psirum if Nwe Wf~ i7 (-1
Air tit lu mwzi diHigivl m Im ltt'(r O N-11 C iivi~ ilerijim: it
rI INIRAL0@ ILEOTRIC
A new nozzle flange was designed and fabricated to permit disassemblyof tho turboalternator housing. (The Air Force unit was designed for welding,which made disassembly inconvenient.)
The parts used in the MERDC assembly are shown in Figure 14; with theexception of the alternator and the nozzle flange, essentially all of the otherparts were previously assembled and tested on Air Force contract, as in Build10. The operation of this first stage Air Force turboalternator, Build 10, itdescribed on pates 108-111 of the Air Force final Technical Report (Ref. 5).
The report also includes the following information pertinent to the MERDCturboalternator:
The turboalternator open-cycle tests and procedures are described onpages 83-86.
The results of open-cycle performance tests on the first stage, 0. 5-inch-diameter turbine wheel at room temperature and 177 0 K are described on pages111-155. In reviewing these data, the fact that the MERDC alternator has 1/3fewer conductors in the windings must be kept in mind1 This reduces the out-put voltages and affects the total power. The alternator phase winding resiai-tance it also reduced approximately 1/3, thus reducing the power loss pevphase and the total power loss, and affecting the electromagnetic efficiency.
TURBOALTERNATOR PERFORMANCE, DATA REDUCTION PROGRAM
A data reduction program was used to reduce the open-cycle data on boththe Air Force and MFRDC programs, This data reductionrrogramn was pre-pared to evaluate single-stage turboalternator performance. The alternatorvoltages and currents, thermocouple voltages, turbine pressures, and alter-nator frequency are the data recorded and used in the computer program.Perfect gas relationships are used throughout the program. Any perfect gascan be used in this pr')gram with suitable input oft
e Gas constants
e Rotometer constants
# Alternator housing leakage constants
The program is now set up only for, helium and nitrogen.
The open-cycle turboalternator temperature (OCTRMT) program wasprepared for opon-cycle tests conducted at room and cryogenic temperatures,using FORTRAN rV computer languagu. lee in used for the room-temperaturethermocouple reference junction and liquid nitrogen is used for the cryogenictemperature test reference junction. All principal turboalternator instrumen-tation is shown schematically in Figure 15.
37
1'.=
I SUNRAL *ILI CTIRICI
z 0
U)
.40
L4 H0
~104
38
r I *NEAL I LICTHIC
6AS S Pt4
'tVEAO
krc c DTHROU511S
14 lo 1-i4 k-1v I II IvkRldu r pst
1 11 No .Ik t1 1Ivii1t c t leIc y
VI \ ltag , Li , toNvutaj kOLH
II h n \o hiwe NoI -
TO \,lr Tuitorv '111111-v Lt l WIill H )
T INoz'ali Tvilpii 1tu'tuInt (il meVtUtv
ru I~)V 11tu Ii'pd Iz
VDI EI RALO ILECIC It
'I' ll" rlatug rprl- ii'1 ion ptlr1'rl iIt-iI 'fi hI(, rir I'fu i hif Air, ln r-e thu l till wli 1wit lti'if-l
t e All s4 wfi(t'e' (liitli'Pl \%i v w it 11 1 h I t, et It4-r-11,11 ()I i I1 I v( l l'd o)t a !.wlhI y'rl IaItI Its 1.I
TvI . fli n - llilt' p-4t 'i u 1.1 iIwor lii 'itll asi : I li, 111, C ~I l
'it'' 'll 4.I1l1 I' I 4l it i ii 1 I ll' 1 111. jl i 1 ~th t t\ i'dill'4- hl' Slf.4 I ll-v:t
1.11w I,~ i i I H''t it , I ill *.t 11 '' ii ii l' v 1\ 11hf~t l i:, -1i l l 't' I l lii
'111,1.111d J ; l ltiit t' d v*II
S ia I ll (I 'lb lm I t I I ' 1-v 1 , 4t, 11W4lt~ I I' 1t t vi i l~l II''Ii ll -if t ii v -I.
I I f 4 1 i 1 1 41 4''. If I ' A 11- l w f , .111 pi tt ill 1 1% 1 , 1- 41 4 lit '' I -w l I' 1 4 Il -I ll t 'A 1 I,
ill I ll t 1, iii ;4,11 1. i t ll, . I, I' l ItIt I :itjtIi t'l l l , Ii'l k li t , 1.1 'Il 1ii It) it lulv
? IR~fA Lo*LICTRIC
~e 1., tN M04TH. flAY
Ln******~J~L f~ LPH ppimOALFF 'P~ rl
sin..... .... s s..s. ss.. n...........~~
PW sWATT***********************0S2 .s05f12 a.. q...s.05 0 .9.1 .
*4'*****JOULF~ LOSS/?IHtFNWTT1I3******** rT, JOIF LOS~aWATT~).50 o 055n. p550 0aP. I 610 0 ,A40;w 9
*~****$LOT rFrmpsat***** AV() SLO!T ro ij '.If) lrS /t)I
r~Jn IN Tok ruR r) ro AC~T T nR# T~J.V5r(. PsiRe149.70 136#88l 1009 0'11
44 ý5014419 0.9043 ýs10
coi TRetG r.Da t~re TIP SPvFn V6"Le RAMTO,99b030.3708 4!)9& t ,0
JNL #PWA s ?eIVaILvCL* 'rH *Let W. T owq.,
F'igurp~ 16, Typtcal Stngle MEiD)C' l)ata Point 8t (.ryogellic, rtnipevattie
M OEIRALO I LICTRIC
* l..n, 7 -- Joule (12R) tosses are shown for the separate phases andthe total of these losses. These losses were determined frorn themeasured current and phase winding resistance, which is a functionof the average slot temperature,
* Line 8 - Two slot temperatures are tabulated along with the average
slot temperature. Thc temperature is measured from thermnocouplespositioned 120 degrees apart in the stator slots. The winding rests-tance is a function of the average slot temperatv-e.
* Line 0 -- Performance factors and characteristics of the turbine are4hown. The turbine inlet temperature is measured at the nozzle
inlet. The exit temperaturc is calculated from the calc lated tem-perature drop across the turbine wheel. The temperature drop iscalculated from +he iLrput power and avtual flow acrous the turbinewheel. The inlet/exhaust pressure ratio is the total pressure ratioacross the turbine.
* Line 10 -- The wheel power is the power inpitt of the turbine wheelto the shaft. It 4s determined by adding the total electrical power' tothe sum of the losses. The wheel efficiency is the wheel power overthe isentropic power across the turbine wheel. The windage poweris the sum of the shaft and wheel parasitic losses dedured frum thedesign computer program and is a function of speed, temperature,and ambient pressure. The bearing friction power is calculated fromrelationships generated from the design computer program and is afunction of speed, temperature, and pressure. 13olh journals andboth sides of the thrust bearing are included. The corrected torqueis the torque (in. -Rb) divided by the absolute inlet pressure in inches()f mercury.
* Line 11 -- The temperature drop efficiency is the calculated temper-atu,'•, drop over the isentropic temperature drop. It represents thepotential efficiency if there were no flow leakage loss. The tip speedis the turbine wheel tip speed. Shown next is the velocity ratio of thetip speed to the spouti[nK velocity.
* Line 12 -- Next listed is the total measured flow through the turbineno-,zle from one of the rotometers. Prior calibration of the roto-Metei- provided the proper constant for the flow equation, The inlet/nozzle pressure ratio is the inlet pressure to nozzle exit pressurerr'tio. The nozzle/exhaust pressure ralio is the nozzle exit pressureto turbine exit pressure ratio. The flow factor is a grouping of flow,temperature, and pressure, which is a turbine nozle performancechc -racte rist ic.
SLine 13 -- Next Jist e*a are bearing performance factors. The journalbearing power for one set of three tilting pads is determined fromfhe measured assernhly average pivot film clearance listed. Thefriction power for the loaded side of the thrust bearing is also deter-niinord from the thrust bearing loaded side clearance. IHoth the
42
$IN IAAL$ILICTI I C
assemnbled Journal pivot filmr thlic kness and fhe load sirde thrustbearing e-learance are 1listed in in ic roiric les.
A listing and flow chart of' the open-vycle data reduction computer programused to reduce the MERDC cryogenic turboalternator data are given in Figures17 and 18.
JOURNAL BEARING MATERIAL
As described in the Air P orce Final Report (Ref. 15), problems~ wereencountered in manufacturing and operation of turboalternutors using tiltingpad journal pads made of nitrided 304L stainless steel. Air Fiurce turbo-alternator assemblies, Builds 8, 9, and 10, were made with a new design oftilting pad Journal bearings, fabricated from a prorni':ing, new material,Kentanium, Grade 185. Since B~uild 10 is the one usu:d for MERDC tests(with the exception of the nozzle flange arid alternator), the MERDC unlit usedKentanium bearing material.
VERY LOW TEMIPERATrURE T'EST
Results of the preliminary open-cycle tc~sts made on the Air Force turbo-alternator assembly, Build 10, were very encouragling. However, the AirForce contract was terminated before further tests could he completed.Since similar tilting pad bearings wotilu be required in a turboalternatorassembly being designed on an ARPA contract (No. DAIIC- 15-?720C-0235), aspecial open-cycle experimental test was conducted on the ARPA contract,using the Air Forcei turboalternator assembly, Build 10, to operate at tem-peratures uppr"oaching liquid helium. The turboalternator was iristal1led intoa sealed housing and operated in the nornmal. manner on the open-cycle systemat approximately 100, 000 rpm, with a 9-walt load, until the assembly reachedU900 K. Then, with a valving arrangement, the turboalternator was graduallychanged to operate on crild heliumn gas supplied directly from a pressurizedliquid helium dewar. The turboalternator continued to operate satisfactorilywith a minimum turbine exhaust temperature of (). 80K obtained. Essentially,no performance data were takei (luring this test because of the limited amountof liquid helium in the prpssurized dowar. Thue mnain purpose of the test wasto observe the operation of the bearings, by nionitorin.0 the proximity probesignals, at the low temperatures. As mnietioned previously, the unit testedat 9. 80 K was identical to the MERAW unit, with the exception of the nozz~leflange and alternator.
This experiment was successful and a signifivant milsstone was achievedin the development of turboalternators for cryogenic refrigeration, This was,pe.rhaps the lowest temperature ever achieved with 1Iturboexpander operatingon self-acting gas bearings. It is the first step in establishinug the feas ibilityof turboalternator operation with gas-lubricated bearings, using very lowviscosity helium. rhe 9. 8)K( helium gats Is prtibably the lowetM viscosity fluidever used in a hydrodynan~to bearing system.'n
4 3
I H
... SA L .. ......
itI
Li-
~ ~4-
IiiLiL
I 44i I-. I-ll
A :: :I .I-
V ii'',t (
# Ii I ' _ _ _ _ _ __ _ _ _ _ _ _
44
it;tl.. ,' ji'tw ~r 01/0 /11
1001" Deq. (C0L1V S P90qRAM.*.**12JL: P(GN(A4 1 IE~VAUIJArE rFIS nArA FROM~ 0PFl-YlP# I'qrLlP-ST4(lPri 4U,ý FrJ,4qAL r[KVA rVR a (SM1* ?PNFiA TURR ANI (l tY04PrNI t, TPrMPlpATlJRPri Ai'ii ot0,et4AJcC ACJY PorotarC V3Ai145SC NHIS PRO0GrFM H4AS RREV~ M00!pICIE F' TiESt WI l NNO LFAKAIP FROMI 46Q HOUS1N0 re ArOPCT T~41S HAS MADE NNSOLPT1! (NOW 01r.16147t; L4NF31 V64#Vfiti967#9vM,)5RODI6Q0I iOC M0L)IFICD L1NE5 r(C.JUIRPOf re CHANnir Tl4, PWRAM !N(ALU0TI *-n
I til 0-NEW Lvi, r4e F0 mery Pt l~0l3AM IV-,LlJ0! Mw 4DI1 A q
200Q CO~MMON !I.Za 1I:laJI.sJ2m.I.J4pJ9*KI *L0.LI &L9
26U ,OMM04ON HOP).l*D. R~, Sr ICat vL96b COMMON U'JMmit480i FILENAME i~s#R!F.r4flFA4#wminf
jo IC ****FILF~qAMFMI r RILWF IN4 @Or) LOflATI04~ WITH Rl4 iiPrYT TO (*PMMP'I
J2U 1PIL[Llir *'00r14)Ar'3,4u. iu imrv
J40~ INflii O~Dl#Dfl).¶j4
Q J ý A L LO#L I sL 2 PL V s MO a MI #N 4 4544UJ 1EAL lot.M4aEae'JlRflNFJ#,4lTaJPW4$) L)ArA P1/3914139]/
%)4UU FLOWmircH C0N43TANTS
WJ S/ 61074469.RPI NFUR 304-9.7fI-I0i'0 FL, TA q4VT-Sd li~p. M11.1 sO61300 S/ 69014477?Y~ ru~pqI 110u07nIOiSOA FL* TA qVe44 1JPr l11m4eA"O~C SOON IIUIA19RWAý3 TRKq 13'd-eVO.0,1 P.v fINWIS9f1'9Af IR8196M64U0C */14 71017A12041 fUSE 14-27f1l.i0-S Pt.e (lJqVT49ATA() IRwA*Pl
610C LLA)4 CALmi~4r7I) CCIVSANT W9C0OJ APPIr! TO SPIP11IF~"411C lurl~iJff WHMFFL DIAMPrTFR*660 IF(UAi alt "HE1LIUM") 00 TO gflOIOU~ 4Mumtoi3ILwleII5 qC8'O646$ 014.6611 Ra3gAeII MP TO 950
7 0U d!)lt144 1 mAn 1vJpumm~A tS NFlTb4P.R 14LILIM NPIR NITRf1UP40'7lA) d loti6JO 9,ý3 1(tikFf aNr. 11ICP') 00 re 700dwO~ I$F.Fig 1 00 re 775d40 70U3 IPr(AFr 4#0L4210) 0O TO 735860 1KFPJol i i0 re 775dkiJ 125 i'14l~ts" BAD INPlifaSEFrRPFdr',* JUNCTI84 IN N!! T~4FS jf,,F 4J(P L494
Figure 18. Open-Cycle Turbo&aIternatc r Temnperature Programn (Sheet I of 9)
1ININAL@ ILIOTIIC
OC rrMim1 01/ OIAi7 S
o0i.c .,Tet$I 93u !ý04Ca~~A# *Ws "C"). uI89s4l
v60 MP.40C1*IO) LqjqtL*DJ#Drv04.OqDIR6964C 7(ClIAS .C1i "HFI.LUM") *AND. (DR ore) Oo5)i d9rviw.A&AnV65(" LF(((J'%M sCO. *E fROMPj*) sAN~fl (DR *rPjo nogis) wq(,pNv0,n.,
* ~~961C. 411(046 .CW. '"4U1'M") @4-4101 MM)F strl IsM) W9(FNPNdAo.
VH m; rAN I #W) L.N0PMP~oJPW~f.0)MmDA
1VVQ CAN I o 10 ) MD PI)*1 141u00 vCADC 1.10) L40#j)r.*R.L4.RA
1040 '4f IoL~- A* NTa 1 00 I *F- A*PL(e$ HP'xI F-A*'4P
114J ýKAUClaImI) L.1cPIM.jM1&Y*.NdJ5
I I 6UQ U~'~ I AN 1 bIidu k0NSIS4mu.0t*i'I
I i60Ci *****t'F%0(MA4t:1 DAMA ý1rUTIV*
IJ6J U u$@
I 31U SUMN0 CC1#4 II.1di4" A14.E.L eg P4FrL1FJ.is iJ
146U 'ALL rHK4~M(:PrL( 1NHFrJPMI v ri )I 4du -:ALL rHE,4MrL(I? 7PrJ .:J44a 4)
I h4Ut W~xW9U0N*bkPý (H9*R.TI14 45N.*"Sm -.491b0 1O:J ~S~6apb*M*7 47.
1640C i'um. ~t843S
Fitgure 18. Open-Cycle Turboalternatmr' TIemperature Program (She'it 2 of 9
40
HNINIAtI* ILICTRI C
WA[~IG Am1 0 J/03/75
116U qRNVI*12
ldl r~ori-U9l8aJ J`0*4104 1+03o mr*I'. PJJUQNb rouo oi8uo o
I$dd) j~wno*ld**haI VJO JJOU~J*130**
19 40 E 813A(G4+JA+JV')
ludj Zu bir 47fl6OJ*fl V C),V400 LIUU014'
o ' u KI a i* 6 J14 t( I' 1 1.4 M
dJLJ90W81 fJ4.J300..I R C 'L9*7 47 2*a.P I
.,!U99C-?IJOC V%.Ik P@0WFH
41 4u usma~d.*401jow-01I-4*rAJI 60 U RU'MU* 11** AF~V
iU 4U A t)DMo#UhCio(Ia FJ#4r~'RrG#DM)
,dl U C iWrAL WJ140FAGF LOSMfiE
d.JOUC J 0UH 4AL. 40440J RICT104 FRO t3@WF4 lq~lR (RCV!,% TAWATQ)2.J (CL.L 'OWFK(L9,eU#3sCMI~flJaJPWeH4aPD~)
ig-AUJ H I ?Ll @UiaF- *Mt* (L9**! )*c DT**A)*t I -R* -It41 01.L24.dU -3f~RP*lrRr
JI4790C
A49'M 5C IIIIEIL kFI'1CIIEJCY
Figurr' 1 8. C)i-(vl rurboulterntiatr TPemperature P'rogram, (Shr1i't 3 of 0I)
47
M INEAL0 LICTAI'UC
CCI'Km1 ~aI1/0375
JOUO U-0 I I 10 I 4F MC.L( 1PFJ*MV& VaI)CR)391~ JNEAL ý1Vji4 ANMVJ943~ fit) 1I0 0OUsiUO0#I10M
JV ý3I 0 PLr4JO,41AL ArPE'~r4TAT1*O PIP ~?p~'~I
41 1J U ULI 73
4h6V O-OMMuO JMi4"6 ýjCU4 O' 11d,1jeIIJIeJý$hJ3,J4.J9.h(IuL0.LlbL9.
4413 bU ImOM(W4~
'414bi) a o AI ~4IL5UU 3 P oM CIHA~bW. 4 I01AAAI ( I UA 5Jb4 .......... ... **............*.....
4bbi IK ~1 41 41pr'e~IN .101AM11'4 34bbji r'rtl'dI j4 . )4 1JI(I ,q 0,4b,4 FV--tId4A*4HjI'0 S t~AGF tRR PL T rR ATOR I4b ý6 0AL*d 1 N1 n 01ft 6wI opNMvort 1'44 a 46HVIJE4 , CfLr IPisV 4 OMANC1P~ W I T4 AM * 44 MIA4q)
WiV I~kfiom1l(.44Ao3I4AI #4iHge9ja rFrt.itr46613 1F00.0 f .l[1J* %42"I) PRNINT 1 14661 iIo04M4'1(ot4XMh3H4r #914CHY001Oiol'~ilm TFPMPP'RAIIlP)41~ s$0i141'( ai46003 latiqr It~
49W. I*e**
49503 P'INt I
Fitgure 18. ()pen-C'yvle Turbonite~rnotor 'rewnpe'ratuve Program (Sheet 4 (if 9)
411
M IuEAL@ ILIECYR IC
5-u 1M 1 IWIS)Q o Pf34A'JP!
5400~ PA',r41f 4
135 U OKI.,r 41)bbWW MM I QQ~raA sI I Hr TOT PW H 4X 7 4.SPD#RPM# 8 9 o 14 OVALL PP!..
57 5U 004Jt Va
bYbU OpliNr 4
6duo o"14 a
6300L JIJ FOOM~IN 1l(AaJAH.***S***** PHA~r nlRFTA~*****j63 S3UA I H (4,CA -i Loesa.wAtf~l
65003~ 0141 4T a
66JU 34100NAI I DAO0 .5N************,Q4A~iP loW~aWA TTS************ aAX s
6dwli ph141 4r A
AVQOU J 0U,%MAr( lUI36~4******JOU4F LOS5jPFHA~qrWATTRq**.****.4Xp
IvJ ti)40 004M T ( I~ JALL a * W M FTo'sS6
iiou opii qr aiibu Ior(ia ir 4lj.01) 4 ? F OMA r( I JA * S I H $1.s t TFP lP.4a * lb AVII SLOT 1*.R wrnn qsrlopv
1400 t 1',41 4 46
Figure 10. 0pon-C4 ,Ncle Turboolter'nator Tempetiituro P~rogram (Sheet 5 of 11)
40
0; 01/03/15
7 boo P)M41'~4 4
76WJe'~~ 4d~,40 Af(IO9#I0'41TJ? IN ToRs~eA1aOb4!U 1!X T&q.AX.I0HAf'T T F~
76 50 pgolit~d I
7950 SiOMATOlU 9t4WHL.6 0WI~..7Km94WI4L.. I-,,s7i.9HI4'4Pnf 0W*#
$000 IVC1d m0t. le) LluO.U
Q $S0 aillIJF
dR50 WSROIMM~ OA.111406144. tfRQg,5X#9HTvf0. PPP..PIC9WTP 19FPr),%qUJ300 atL)MvP~.6 R4ATIO)d350 PIaNP1 58. rW.L I aUp VSd4JO( btiI~FOHMAr(1A# 1 9 .4# AIG PF~s ls IOAOqa 09s4p1noFS . 4)6450 )#NINF I
d6UU SIOiIILOW FAc?.)d6bU 0`141,41 611.W#PPIINNP1VlI.#67JUJ GiFP0MATC IOA. FlO.iJ.a. o~l07.74oPI0.7.74s.pqa)A~7u? e8IIN I,
V716 A 9Am0HrN.L@*CLs1dil? Wk~f4tIJI1.*6*l4PJ W01406'I .A$PL.C
U720 i'AINr 66.iPD#W@HI#OITPL#WSRf12
81JU eINil~I 11014IT 410914~T I
$$OU LWl'JU33 4U6?(SUrINE Pi0WEfl(PFnvisnflA.(ht)IA.#lPaHP~of)IUUIQ DIMt~i4Sl9N I1M .,PChP(A) Y.' P A
I UJ N) £1 ICMC$a i) * ,53 573S63 F-O80Ia16o 3037 a 17I 199AIAt - 7e 10444 1As
10 100 42 01727 48# -6s57 21933,1IUIk0t DATA FAC #)FIC(s) l
Figure 18. Open-Cycle Turboulternator remperature Program (Sheet 0 of 11)
50
r SINIAL * LIGINmIC
7 10 18, J APIMGM hoho7.II40'74~3P["0te74oPS074R29.a.itO3A71AR*6.I17A7
I oil4Lu roilrd i~j'0GjAM egrA1NS T14V Ptr LqJS IN WATTS Or A TP J01IRMAL 0U1 UA soc 0NkiI~rI*1 OV i f i~r o 10 rci MI6FA PtVOTIPr AV *6i 144r) L1'1IPS10t~uJ WU HrLIOiM GAi
I JUVGC ie)m RorAT104AL 80~nRIQ3JUIC a~~ VOA; rimpomra!.(l.. (tPNP AT SorimP)lilic.) be'i)MAw MAAdIMUM 4i~rArtIAAL. 801.s~O
IUJu 'M 111C P u GAMI. IrM AT041 FOR ROTH IPfl I
10 . 4u M, 1) 1 - a J 0u q AAL. 1) 1A - FTM t Ra 4I j ibJ c uAiA br ~iC'jrb ('oIrW!. OF POLY)Iujoeut: i) PurcIM * V1 LAMADA AT rww5 His'
iU~fjQ, ) LV'4IN)A MI4 VS (LAMPI)A mA)*Lmn MI"4)1030c ) mil Vs %JflUMAL. DIA
1U4JU mH4mýa'I.1 U44u Va 11 SI'70~LAMIld( 1 I +#0S 611l*PLAMM)
IbJ 1 F( tU 1dIb H h) s)Ure'i
I U 5 1 I~F( I i-ATH 0 - 4( 1.1V) 0
1 J! ý')I P ( . IA Ia s 1 4 0 H ~ e #41V A.. F ( -1(~ 1 ov4 o mY P
I Igurtj ,0%f I Oýn-'y j Tubaltra 4rnprtr )r~r (Mhee 17 of!)
11) 3 & H rif as A , A I) ( 4 aL ý. a4) 111,ALL FP FI I ~m A * Aa *So s na Yp IVI-I~ ~ ~ ~ ~~~~~* u jjIFIt O1041
F *EOIIS11ALO ILIaTRIC
ccE4 r ~ T(c -ii t 001/
IUA6J PJ0 t~'1) a v AL * w Artt03.I U661) cio ro 6oo
1i3743 IOUUO14 101 0
Aov
10174J I AAR -K
iQ IF 60 10 1J11 14t
1019 5p .j A a .W( I -r(2))* I)- (V 1Y3,u% ý)yJ vaI C A# ( I *W ' T #
In Ii-) JC 1. ( ~ r' I )I K~A .JII IJ 34. ~~L 29 4
III U 1 Ali y a~ A I dl
II0? U ~ i
I I Ii ) I ) 4( 0( J*OMI k'M 44La * J J* )
I o I I "d J I CON r N
IL~r I I . 1ni f -(vI I or, I ) i(LIi1110 0''~Er~ woo' 1'u ri (rhe' I Of I
Iid 130 £ICNA ED *GE a 75000.) CMDus I 15*NR ED**(.-*Pf)5)iI 140C uPLDF 13 tURBINF DISC WINT)A(30 LOS,%
l12170 MVNAED 01 . 12 000s) t1UNm55e33Fa*.4R~rI**(.I11AA)l41d0c 00 IS WHEEL MAC1(SIDt OPTIMUIM GAP1419Uluuo 60 0 G0m5*GPUN'*D1Iaooc NAJ IS JOURNJAL DIAM. REY-40LDS NJIJMRF11aI UL JNk~ai *367*N*DJSQ/t 9, 551r6*CeNSr,
I 12.CcODS IS JOUR~NAL r)IAMFTFIR DRAMI C7TPIC I PN1,19JO c06u$,Ld89/cNRtEJ**0f25)182-40C ILSF IS ALTPMNATOR SH4AFT WINF)AME LSO.9.IAtbi) ,'LSFu:3.3*DSH*CDS*4l3*DJSQ*0J5Q,'(97OpFq*SV3)
I~~ t~ ~ E ;d .MLr~i 10000e) CG05AN7eTNNfIAJUU 1P'(MJi1EG e.tE. 10000s) COI~~I7r04T/'1lf*OlI431jo CMG IS ALIFNNATMt4 CIAI MOMFNT rTIOFFIPINThdldj0 PLOF' IS ALTERNATOR GAP0 WZIJDAt1IF LOSRLý330 ,L(WR~U0.0J32*CDG*)M*iN3*DGi**Ao'(F.7 Fg *SV3)I.2J4JC TOTAL WINDAnFl L05SS5S1 PLDF+IMJ~LIRFP~rlId-.Jbu 1t1rJHNI .)3 60 Ely0
F'igu t'c I t3. C1pii~- (yc 'Ic Trboalte riato' ~rem~pvratunrv Program(Shc'vt 9 Of 9)
MERI)C'' if(A FbNAPRA4 MIY
An us~somm 1.' (ii'wtitg or' tiw NIPM I)C t urbotilt ernato r Is s ho wn in~ Vigu re 19P.
Upon transfer ofillth Air V()ore turhociltevilatnr, luild 10, to thle M'EH)MCc oftrmvte the un it was~ comnplet elY ( disci ssemiblvd, All v r'it ic a pa rtsN to be re'-u sed( wvv r a refullY exam i ned ntrjjcj(, 11 ta m I)c cC'op. 'I'lhe 14ci ft (partt 7) jotirim 1be a ring suirfac'es a ad t he NoitAllikIn m011NMtt nI heaciring puds (pacrt 23) were lightlvre tinpperd t ) remove slight I ndivat ilm (inot' wunrv i'h'e h imitrig sect iol (pal~ rtIMwas d Iril1led and In ppecd to rviposilinthI Ile jou mal Ifi ring pr'ox imit y probe(pafrt I0) to 1)'Ilu~t. the probe anigic 'mo viewing the( larger Kentcntinm ,jour'nalben rinp padl. N( %v OWL110 herrring gimblal pivolt4 \%'. tv ins ta~~lled in the thrustben ring axsenfm hi es (pui rt 1 3 and 22). 'I'l.' nvw pivoit hav'e matinag ball andsocket Pita to redticp thle pin v In the thmi.st hen ring tissvimblies, tis noted inthe, Air F'orero turhoalterriator. butild 10.
'Ilre no./?VIe sval Immlwet 1cm tvst wos tmade with tlhe riovze hlock and newfil latigv' ito enrsm r'e a miifrwili met ail-to- 110 llo seal lit thlt noz"V.le diam~eter'.
1'his Ivs t' .-I 15 lt'ctrlwi'dI~ abov iite und I tu katlor r leitilm.' i oplpt-c%'(h' te~st wiad
..e t .n ....( ...
6':
4111
-L I I
Figr . TA A6li".
* I,," 54.
- ~,'
a Id
III
Figurea 19. TA Ass*,'ni1,y llrawin• 588E477
54
,o
S IIINIRAL O ItlOTlilC 'ii
All of the turboalternator parts were ultrasonically ch,,•z•ed and sealed inclean plastic prior to assembly in a clean room as outlined irl Appendix Ill,"Turboalternator Assembly Procedures."
The thrust bearings were shimmed to give a total axial shaft travel of 0.6-thousandth of an inch between thrust bearings and to provide a I. 5- to 2.0-thousandth of an inch turbine wheel to nozzle axial clearance. The journalbearings were adjusted for 360-miorotnch radial clearance,
PREIIMINARY TURBOALTERN.ATOR TF, STi
The turboalternator flow tests described in the Air' Force Report (Ref, 5)and results plotted in Figures 75 and 76 of that report were repeated on theMERI)C assembly. Results of the tests are shown in Table 7,
Table 7
ME RI)C T L IRH( )A(.,rrERNAT()R FIOW TEST
'l'e St Gas u,,, Agreement
Nozzle ctflibrat i•,n lie <3(n,) turbine, v.,iieet)
Noz •,1•' call brat I, ,n N,• < 3
(no turhLnP wh,,el)
Ze I'() '•peed lie .•3(with turbine wheel)
" Z,,r•, at•e¢l N• <O !S(wit:, turbine wi•eul)
i• i!
}Initial z'm)m ternpe ra•u,.es •H)et'ation•] tests (()ctober 1973) era the turbo-
altern,'.Jt•r' a:.•.•embly with the shaft magnet unmagnettzed and magnetized gavee×celhnt r'esulfs. 'lasts were n mde at speed• in the aea of 100,000 rpm.The proximity probes indicated very stable operation, and the shaft orbit probe•ignals pr,)duced e×ceptionally sn•all orbits of 30 to 40 mlcroinches at e•,chend" of the •haft, very similarto }luild 10. The thrust, journalbea,,ing, andouter gimbal ring pr.xin•ttv probes al,•o shc•wed very stable operation° wi•h
.,slight once per revolution oscillation noted. I •gures 20 and 21 show the
proxi•nity probe signals oblained. (The f•tint shadow or double images in theoscillos(,upe traces a |e caused by b•ckg•,ound noise in the pr,•ximlty probeinzt rumental icn. )
Alter'na•(),, Wlndin• 'l',,•t. ({,,)orn len•l)erature resistance rneasurernents weremade on th• MF, R1)('alter'•mtor '3-1-,hase windtnR•. The average winding rests-tam'e w},s !).135• •. O. •.5"',. 'rhe Ai•' I.'orce average wlndtn• resistance was12, 52• •d rr•,)m te•l:•,,rature and l. fi4.q al liquid nttr,,•rn tel'opec[duPe. Sinc.e
.55
IINIAL ILIITRIC
Figure 20. Shuirt Or-bitlt ol' NI iIII ' Turu boal-ternator Assemb], ()pepratlng atRoom~ Temnipe ralui~re, No I aod, and100, r)(0O l ) N. (()•cillriscope Sen-sitivity f', proximity I)j,'obe signal, -'200 .4n/(1,m.)
the same wire was asged in both winl(tlngS, v resistance ratio of 7. 65 wouldapply to both windings.
Figures 22 and 23 show the slnusoidtil, si-phase voltage outputs fromthe alternator. (The uscii11c.ope ,'eric~al scnsitivil.y is 10 volts per centi-meter and the sweep is 0. 1 millisr,cond pe r c~entimeter in both figures. )These photographs verify, to a degree, the low voltage regulation and thefreedom from harmonics one wOuld pridinIt fromi ptist similar experiences.The terminal voltage appears approxinmtoly Los it should for the conditionsof this test.
MERDC Turboalternator ()•pon (Cyle lesis
Since considerable 1oo-mn tenperatire. teat data lhd been obtained on theAir Force progranm, it wo,8 deciided to itstis ii :ssemnbly at temperature•approaching that of liquid nitrogen, with helium g.s, in the open-cycle teststation. In this test, the tur-hoatlternator is inistaalled into a sealed housingwith feedthroughs for the elec ri•,l (conections. The heliuni gas to the tur-bine is precooled bv flowing the pu .s thr mgl" h n coil immersed In liquid nitro-gen. Strips of plastic mateitil were( 1tW11ad t•, I 11h turboalternator supportstrand, to direct the cold noxzle hau~tsi pgai oaijund the lurboalternator sealedhousing, cooling the whole assemibly. A ticLI aur wos phaced over the turbo-alternator assembly, to isolate the ,4YsvinM I'vw noni• nent moisLure. The tur-bine exhaust gas still expunds to :Itm.ph)Mercr,
SENEIAL 1* LECTRIa
l'igu re 21 M\ I-L)C Turhrurilternl .or, A sseirdt ty,()pvr'al 111. lit l ito im-1T'ýll) )C pri tilv, NoI ('1 (I an ~1fI00, 000 fl M ('~L''I1lit t h rt.!sIt proLh(' Ix ia; :41lfa i ,nt i ou I
1,l .) t ,Iv i I i uI, ~ ,,i I, i r Ii 1j's ; T'1 -:' r t, 2 is4fl)w jeiir a I r 1"1 r1;(' rirj pI d I fl" I ( I t inn ; Trove'3 is tli !u mnlur Ilivust I ca rkivt- ginibhail ring
invlty proho Hliia,1 4100 lt'i/c'rn Vtr'titlcay;Iloizo)F'1'.lt 5Aý'vvp~ 11, 0 Ins /cvI.I
The perfornmance. nivistrvirenwo cul tao hebitaIitid undL i:1he effectiv) aver-age nozzl?.e preo snu ru can bv rncas cciisuc Ili cnjunction iiwithi othle r pe~rform'ancefactors. The (..fective uvertige noxv.1v prossuircr wYotld be th11 pressure in the,sealed housing. The thiree preSS1.1" cc ta its )b~laInr(I fromn combinations ofthe Inlet, nozzle, andI exit pressurcs will vary with 1.he volocity ratio In acharacteristic niannier,
The turboultevi'nalor %Was o)pe tetkd Po no' hotir waa' raoom temlperalture,100, 000 rpni, and no load, t~o purvi' 11i- systemi-i wilth cvi inclers o1' helium gas.The dlew point of the heliu turnws, -6 5V. TheIw gas press~ure ilap connocting thesealed housing vwas discotircn'ted I() pa r.pe I li turbi nriterinaor housing and the
F ~~~sealed hocusing. 'ihtit Inirdcv hittt tie t~ hirlwAlit tnililir, cssernibly at aslow rate, liquid nil rugnn wns -itmO~Y tiddvle h) ai dvw:ir' cn)iilining the coolingcoil.
After' 2-1 /4 hour's 0I' cmoflfng, t lie (- Iei pe:1 W1)1It~urt tir t Ilile niozz~'le Inletreached 109rK, rind thne tilt-inuf im, whitii11V iwli ivipt rtifE twe s 1 65"K. Thieturljoalternallor utqi ieiale slit. sfatnu0iI ItN 11- HpoodrIS ill 1lit o r'rIf t of 100, 000rprrn with an 1 1-\vfi loa o n. n thIlic i tt
I IN IIAL*I, II O11011
Figure 22, Alternator, 3-Phase, Line-to-Line Voltage Output of MERDCTurboalternator, Operating atRoom Temperature, No Load,and 100, 000 RPM
Figure 23. Alternator, :3-Phase, Line-toNeutral Voltage Output of' MERDC'rurboalternator, Operating atPHoor Teniperature, 8. 5-WattResistive l,oud, and 100, 000 RPM
$ IN IRA L LIECT I C
The only problem e~nc'ounte red during cooldown was that the leads fromthe orbit prox imity probes failed and the operni ion of the turboalternatur hadto be monitored by the remaining probes.
The test plan was to check the operation of the turboalternator and to oh-tain performance data by operating the turbouiiternator in a vertical position,thrust end up, with the nozzle- inlet -to -exhaust pressure ratios of 1. 49 1. 1.),and 2.4 (pressure ratio of 2.4 wasn the design condition).
Tests were conducted by adjusting the gas flow through the nozzle to obtaina given pMISSUre ratio, and the turhoalto rnator speed wns cotLrolled by varyv-irig thE resistive load onl the alturnlAtlr with speeds railging frurm low operatingspeeds up to 200, 000 rpm. Nine datak points were obtained on M\/Fl1WC Hun 100with a pre.44ure ratio of 1, -4 and speeds ranging from 68), 300 to 211, 320 rpmn.
The nov~zlc pressure rutio wits theni adjusted to 1. 1), and 5 data pointswere obtained on MlEIHIC Rtun 101 with the turboalternator speeds rangingfromi 118,320 to 196,* 240 rpmi, As thu tui'hoalternator speed Wats adjustedtowards 180,000 rpm slight roughness was noted on the thrust probe, thlejournal bearing probe, and the outer gimbal ring proximity probe signals..This indicated thrit the shaf't many have been Just touching thle thrust bearingsurface, but in general, this did not affect the operation. At this pressureratio and speed, the sealed housing pressure was 84 inches of watev aboveatmospheric,.
An thle turboalternutor flow wvas gradually inc reased for the, 2.4 pressureratio, the proxim ity probe signals bevecnie rough und the I urbitiv speed de -creased, Increasing the gas flow continued to decrease tile turbine speed.Trhen, the t urbinti stopped, Attem pts to restart the tu rhine failed.
The til rhoalte rnator was a llowed to wa rni up to room temipt ratu re. TIhereappeared to be c~onsiderable moisture present in the turboalternotor, asevidenced by tile fact that tapping the suipport stand did not move the shaft asit does norniall~. After purging the turboalternator with a low flow of he-liumgas for approximately one hour longer, the turbine started, Its operlition wasrough tit fit-At. but smloothed out. quickly. All of the proximity prohe signul Hwere identlicl to those obtained before starting he cold tests,
There are two possible explanatiu,.ns for unsatisfactory Ope ration Lit thlehigher pressuv e ratio, Pirst, the vessel in which thP unit wOs tested (lid notpermit u bleed of' gas from the alternator region to the dlischa rge side of theturbine. Wor this reason, Increused preswire on the back side of thle turbinewheel mlay have overloaded the inner thrust. hearing. (The final i nstallationof the turhonlturnator included Ll gCIs bleed to ecounteract this effect,
A second e%.planatton is that moisture wos deposliedi in 11w I urhoulternatorand ecaused thle shutdown. [here' wus soniv e vidence of -this; as thle Milit WEISabeing warmet] up and purged with dry gas. the bea rings LlppeLL red sticky onlthe osc illoscope trat~es, which would bv the case if mloist orc were presernt.This effect waH present until somep timei Lifter thev unit was fully warmied up.
511
IUIINALO ILITI IC
FINAL TURBOALTERNATOR OPI.N-CYCI.,E SESTS
To complete the open-cycle testing, a second series of tests was con-ducted in December 1974.
A number of gas supply systems were considered to eliminate the pos-sibility of moisture causing the above problem. It was decided to use heliumgas from a trailer containing 40, 000 standard cubi' feet of helium. Thepurity was guaranteed to be 99. 995 percent or better, with a dew point lessthan -1 10OF (less than 2 ppm water by volume).
A bleed line and valve were also connected to the turboalternator sealedhousing to reduce the housing pressure during the open-cycle tests if required,
The proximity probe feedthroughs were adjusted to improve the contactwith the mating lead connectors and prevent opening during cooldown.
The second open-cycle test was started using the same procedures forpurging and cooling as described above. The turboalternator operated verywell during cooldown, and this time the orbit proximity probe leads did notopen. Figures 24 and 25 show the proximity probe signals obtained with theturboalternator operating with the nozzle gas Inlet temperature of 89 0 K, I1-watt load and at 100, 000 rpm. rhe proximity probe signals are very similarto the room temperature, no load operation shown in Figures 19 and 20.
Figure 24. Shaft Orbits of MERI)C rurboal-ternator O)perating in the SecondOpen-Cycle rest with 89)0 K Gas,i -Watt Load, and 100,000 RPM.(Oscilloscope Sensittvity of Prox,-imity Probe Signals 200 pin/cm.)
60
SIIIRALO ULIChAiG
Figure 25. MNI2)Dc1 'Turboaituinututr operating inthe Same (.1onditin aS US i Iguru 24.('rrucc I is the thr-it probe, uL×iX lshaft notiorn between thrust betinrirngs'iTrare 2 Is the ,ournmi bciring. padmotion, Trace 3 1s the outer thrustbearing gimbol ring motion. O(scil-1lopu()pe sensitivity to proximity probesignal -- 400P iln/ m Vtrtivlivy; h =r-i7ulltsl swe tp - . 0 i•1s/c"n,)
As a comparative check, two data points were repeated in MERDC Run102 with the nozzle inlet to exhaust pressure ratio of 1, 4. There warn no no-ticeable change in the operation of the turboalternator, The nozzle pressuroratio was then increased to 1. 9, and one of the lower speed data point, warnrepeated in MERDC Run 103. with the nozzle inlet gas at 83°K, 29. 5-watt al-ternator load, and 117,000 rpm. This data point also repeated very well;however, the thrust proximity probe indicated that the shaft thrust was justbarely touching the thrust bearing. It was, therefore, derlided to reduce thehousing premoure from 59 inches of water to 27 inches by opening the houaingbleed valve, since this condition would become worse at higher speeds andhigher nozzle pressure ratio, The reduced housing prE.sure corrected thethrust operation and the data point was repeated with lower housing pressurebeing the only difference, The nozzle pressure ratio was then increased to2. 4 to be cjutaih thatl the huuahlOM bleed was sufficient. 'The turboalternatoroperated well; four data points were taken during this MTIUDC Run 104 withspeed. ranging from 166, 500 to 18H, 940 rpm und loads frotn 57. 7 to 60. 1 watts.The fixed three-phase, "Y" connected, resistive load points were the limitingfactor in varying the speed and load over a wider range, Figures 26 and 27show the proximity probe signats obtained with the turboalternator operatingat 188,940 rpm and 60, 1-watt load. "igure 27 shows a slight oscillation of theshaft between the thrust bearings, but there was no indication of touching or
0 1
09N IRALe ILI STUIC
Figure 2(3. Shaft Orb.its of ME'~JUX Turboal-ternator Ope~rating with 821K Cam,610. 1-Watt 1,oad, arnd 1811,940 R1PMwith NUZZIL' Pressur Itu'l~tio uf 1A(0scI1licnsorip -4ensl~U tIvily or pr'uxl'imIty probe signuia1- 200 u in/vcm,
2
Figure 27. MERDI.C 'To -boui1 rna~to- Ope ratitng inthe .LU11e ( 'mi cnit ns L is Iigtire 26. (r race1 iN t he Iimi't t pt-ohe. ,ifa HA~4~18haft motionhwel wi' thrust br'a.ingi4;g T1racie 2 fig thejJur?1id hei t ring pud mth)t Ionj 1Trace 3 to theotitvr thriuis he~itIivin grImilh ring motion.()scilloscope sriiiII vitY of p~roxinity probes* iglia h. .1 f()0 U Il ,n I vvi-y t ic. ,i li horizontal
6 2
*INNIAL@ ILIOTIIlC
rough operation. Since the housing bleed did not have to be readjusted at tilehigher pressure ratio, three additional data points were taken for MERDCRun 103, pressure ratio 1, 9, at higher speeds.
After the above data points were taken, the turboalternator was stoppedand restarted cold without any difficulty,
The operation of the turboalternator was very successful during thesecond open-cycle test, Results of this test with very dry gas gave evidencethat moisture was the cause of the turbine stopping during the first cold teot,and that the sealed housing bleed will be required in the cryosection to reducethe housing pressure when opernting the tu boolternator with nozzle pressureratios greater than 1. 8.
PERFORMANCE rEST RESULTS
The data obtained in the first and second cryogenic open-cycle tests werereduced and the performance test parameters were plotted in Figures 28through 33.
The data reduction program was not modified to reflect the operation withthe housing bleed line open, since only four new data points were obtained inthe second open-cycle test, with the nozzle pressure ratio of 2. 4. A ratherlengthy program would be required to establish the correct parameters overthis wide temperature range. Therefore, the points obtained with the bleedline open are identified in the curves,and their values should be consideredas trends, not actual values. An indication of the effect on the data, causedby the housi-ng bleed valve, can be seen by comparing the reduced data fromMERDC' Run 103, data points I and 2. Data pobit 1, shown in i,'igure 34, wasobtained with the bleed valve closed. Data point 2, shown in Iigure 35, wasobtained with the same resistive lond and with the turboalternatot' operatingat approximately the sane pressure ratio nnd speed, but with the bleed valveopen to reduce the housing premsure from 59 to 27 inchies of water.
Also. the eqlmatinn used in the dtitu reduction progi-am to cvalculate thealternator winding resistance per phase versus lemperature was derived for
the Air Force alternator acld was n(t niodifiecd fr tie lower resistance wind-ings of the MENU)(' alternutor; theret'ore, the winding resistance pr phase,
'he joule loss per, phase, tuld tittil tlule Iloss shown in FIgures 34 and :35 wouldbe approximately one third less. 'lhiu would iucrease the elect romnagneticefficiency slightly,
63
Atciul .l 4 Iit .SY I
IFigurrp T'. 0v'i ai .:trrtiochny vcrstis vvi'hiuty mat ic(1Iclitmi (h~is at O3H K)
I,64
0 1, IUI loUlM I41 4.uI'.I
0 1.4 11 * W 114
0 I. I104 idlluIll~
... i 1 ' Il i.' i 1 .
Filgure 30. Hlow Factor Versus Volucity flatio(Hfelium Gas at 830 K)
0
I ~ 44110 -1
11 1 AI "I 1-
Ofei ttum Uas ;a~ t HA3` K)
r SINUALO IELECTO It
0 1 41"' )DO IH I' om S'Ii
13 I'll 101 4 l~l S I,
0( IOAIlUm4id4 av.jkA
0 1 '11111 NIII'I ll44 iia fl lil ' I. -
0 2.' III11 With 11iitiI flI i v 1 1
Vi"gure, 32. Flow Fac:tor Vermum C2orrected Spt'c-d
(Heliumi Gas at BI3"K)
11 I l "
1%1 A I t , 11 l .,
4111 liumi ( ;as n~t 83' K)
61V,
AlA
******** 011V FCLL P Foi ilMAVJ F'***********
IU f EL. 0 1tIj P lM OVALL .FF~. flIN 51r~,
FLF ooo4 ir 13F4FO Nd **l *su41
4f 1.s30 41 0 W' AI 6 )-S 1)o
**~~***l )**R*''.4I aF ' ,AMPS** 0***,**** rlw i
*******JO(J'- LO ,'IPH. mwA r T l**.~** t*r? J0cI~LF c,'
I UH 1~qs ,qq74 r ,, o- "ks6e?
I iv .1 I' * 4 6 0) 'iý I # 6 9 ' Vi I o*q 7 '1ý
j ) .? , 1* A..
I'igurp 34, MEIU1) Hun 103~, hit h i 'sit I, I )ato Hbtaii iii' \kitll limiUintg
HIL-e Viive C~i 17
GENERAL* ELECTRIC C
. I V(LP IA 1 0-0 '"v.•l rioJ'n Tl
i~~~ ~ L• ~ i•• , F Pi Wit Pi Ch', ýý ý ',J I I H Hl Vl i. t / l~l I. A1
I r:~ey'riI~''~aq rpMr~ f''9
. . .-. .. " ..v . . .. . . .. I ' ., wb4 . f.AY '...,.
V A 4. !' * W)D VOA 1, 1., Uk'P'r'I + I I F I',,''-VI ')* f "4 G'-j fi Q +'s
"I * -' I * 2.1 *l o'>1 '.71o?1"ll>~
4 4 h6 w*$ i .'" A iiI ClJi.'.•iJI,'~ l' 494'-." -44***-4- rn!'•1 t..rI-',.!'•!-"
Ii. .,l l I' I, t. ',l l I l l ,I I ; I ' I' ''' l 'i '"i I 1o ltl t' •.1 1 4, k U s. \4I 1\ rI' J,-it ., 0 f- '! O T' Ii{• ,
..ii ,- A ,
+ i . ") if,1 1- .," 'T T' I ,'I4' ioL1 ..'
Ill I • lli 1 'v 'i '41'1 p- I, •'• "/•I:I - l'4 # *4-i4 ¥ 'th"4*#iit4" *i - h+i :"'i !
~~~~~~~~~~~~~~~. ... . . . . .. .. ........ ... t i I i i I a i i ! l I l l
I 1,• i I ..*"{ '.- ,qI ro I.'11o);
''IL.. * A ')IL. -4 *" *1' 'I - * -. A
it -' II , ' pt i , l• ' • lI ' 1 "
j,• ,/ ft I i4 3,. " .4 1 !*J I. r : I r,
J /I,' . -' i i* i * '. * - ' - I , .,
- II - , *" • " " iI• )I o
t[i 4 i :i'• ;r' Al 4Ii ]1 1 ' 1(1111 1 03, I)LPtl;i |I ' llnl 2, * I It ;i ( )hI lUird A• •ittl I (I•)ill !,i; I 41e,'I\ I-v •, t)tfi ,111l3 t1,41 ( oi-rpctr'(:! 34)1" Ihln iti~llt' I~tc;ikue-! i 1 "
68 f
SU'IIRi !AL * i LlClIRIC
Section 5
CRYOGENIC HEAT EXCHANGERS
GENERAL
Refrigeration cycles which employ dynamic rotating machinery operateat lower pressure ratios than cycles which employ positive displacementmachinery. For n given refrigeration load, this requires a higher mass flow,These conditions place a requirement on the cryogenic heat exchangers forhigh thermaleffectiveness and large heat transfer surface area. An ef-fectiveness of 98. 516 is required in the warmest heat exchanger of the select-ed cycle. An effectiveness this high implies an extremely large number oftransfer units, 66, which also implies an extremely high lateral thermal con-ductance. A conflicting requirement is that while the heat exchangers musthave high thermal conductances, they must also have small pressure dropsbecause of the low pressure ratio of the cycle,
A type of heat exchanger considered to be well suited for this applicationwas the stacked screen heat exchanger. The perforated plate heat exchangerwas considered, but the required plate porosity and hole size would have mademanufacturing extremely difficult, Since stacked screen heat exchangers hadbeen successfully constructed in small sizes by Kinergetics Incorporated,the decision was made to purchase the heat exchangers.
Specifications for the heat exchangers were selected based on analysisof the cycle, These are presented in Appendix IV, "Specifications for a Setof Seven Cryogenic Exchangers.
DESIGN AND CONSTRUCTlON
"I'he design and construction of the heat exchangers is discussed in thevendor report enclosed as Appendix V, "Kinergetics Incorporated -- FinalReport. " Because the vendor was unable to prevent stream-to-stream leak-age in the cold exchanger, It was decided to make this exchanger inactive.This was accomplished by blocking it off at the joint betwern the aluminumheader and the wire mesh header section at the warm end of this exchanger.
TEST RESULTS .- KINERGETICS
Th, poirtion idfthe t(Istinig which kas c(miplh-ted by Kin,'rWtivs is prt-S4edttt,(in Appendis V.
TEST RESULTS .. GENERAL ELECTRIC
"I'hiu dc ffi( , ',ie e il 'l'(]d at ki, elt!Eti t [il w )rpIi1rlt(ld r'esultedI if) ad I I , ' ;'th t, i + ,q . rT 1at t. SI '. a , 'niluitdd I oll ' li t ii In
GENINALO ELICTRIC
with MEJRDC, that it was in tile best interest of the Government to a('t':2pt
delivery Uf the huat exchangers and complete the testing at Gý'neral Elc:tric:.
GENERAL EL,EC'TRIC TEST FACILITY
The Resarch and Developtim.,,nt helium test facility consists of a positivedisplaceme.nt c ompressor, uftercoolers, and leak-tight plumbing necessaryfor the testing of high efficiency heat exchangers and turbomachinery. It wasdecided to use this facility to test the heat exchanger,
The compressor is an Ingersoll Rand Conmpany unit, Class ESI- I N 1I,- 2,which has a solid lu, ricated piston. The bore is 17 inches in diameter and thestroke is 9 inches long, nnd the unit can b, operated either in the single ordouble acting mode, The mnotor' is 60 horsepower, 1800 rpil.
With the suetton pressure aboive one atnmiospher'e, the unit is simigle a'ting.tinder these conditions, the rated performance for helium gas is:
( ha Cacteristic P.rforimnc'
Inl0,t pressure 0, 121 I MN/nmi (17. 6 psia)
Dis'harge pressure 0. :3,1 MN/n1 ;,(52 psia)
Inilt temperatur, 332:1,0 (1 036oF)
Vl)lillllO flow at inltt 0. 187 i'ili/sev (396 elmi)
IIelieirn density at inlet 0. 175r kg!n i'
1lteIim mass flow 32. 7 g/sec
This nmvss flow of 12. 7 grams per scoind, and the prn-,sin'i , r'tio ,ct':2. i!5, ,re n,•ije thani adtijunte, fit' testing thle heat ,x,-ha'geur.
'!" M I F; T ll:l I it ) I)
Tlhe phavsi ('ir U rrungte'ment or lie h,'at exchanige r test i• shi•\wnin in I"ligir:16. ihleliurm 1.1 Suppl (id t0'4m hy thaciliti co lmp S r'essor. It l Esis itiniv.i
thie hitfh p ,'-s~lie side lit' heat exc hLungrl i' 7 Ib't'<,re ftinwinl mut .uppl" 7 to( 11helit•itd nitrogw.rn dewar'. 'T'he lines to and from tihe liquidl nit edvevne dew ir' , iv'linsulated, A ,alvi is inu'ladled foir' eontr'oiling ti turii ng s- reim !•tu',,g ts'tr,
T'he hliuhm i,'lur'ns from tihe dewar, tlows In i''tUril 5, and flOws ri0i1li Cix-
'hnng|'•, s 6 :iuil 7, beforv rc'utur'iing~ig I othe' c liio'i45n1'. 1 l cl i a rl.ngi • , (I 1ii, I'lm (li\'v
lIiv ' ut., n. t' '-- l '" 'i ,.ti l'i ' ) It-l tE'in 11 h, ' !itW4.4. ,'rli vi('h1i•' h(i uii'- I'in:m rI t t , tu'. i n
* it' I I'-- l'l i 'f t c'l' T )1'
4l'O l il' d ' u ,t dl" l i i'' lt I 'W1 I i" 'lint I'l c'cIin. :' t cc'tc-" c
T'he ttini'i'atures c0' Ili, t Irw -divreams ,rvi t'e't a ski I I-' l rlr••hel'd'li . ;iL i 0
;iltt 'l ' I l itt. et U x' #;, l iu ge r', w ith1 C-i4 ip t'v I "i'll' s at•ll? i a lltin, 'I'i,"111' :4 lii
,i 't - if , i ito -:t t' r i s. 'l'h. rI II,-I s IS() I iIV r III'iAn it tl.' II II thi, I l1. , I li, ii-
DIN I RAL. ELICTRIC
Capped legend
N*,,o•I I Ou I indvo.1 ~ - Hand Volvo
NitrogenJuneonedt hanger I Flow Nozzle
'7Temperature (Copper-1 rj'-{-Ar• •"•lconstantan)or e ® PressureWart Exchanger 2 •
Capped r" sup ':8 > Pressure Diff
H HeoterHeat Exchanger i apeCapped Variable Area
ff U Flow Meter
Cooldown I Temnperature DiffBleed Heat Exchanger 4 (Copper- con etant n)
Heat Exchanger , 7 NitroIen
Heat Exchanger 6
_________________ iit4
"" M etl~et
Hoot E g I
I. '¾L. '
41
MIIN ALO ILICTAIC
return 1. The reference junction of the thermocouples u.4ed to measure thewarm end temperatures of the heat exchanger ;ire in an ice hath, and allothers are in a liquid nitrogen bath. Figure 36 shows temperature differencemeasurements, A Ti and A T2, but th'-se measurements were not taken be-cause grounding of the thermocouples during installation made them inoperablein the temperature-difference mode of operation.
Pressures are measured at points indicated in Figure 36 with pressuregauges. The a P measurement indicates the pressure drop past an A. S. M. E.nozzle of known diameter to provide a mass flow measurement.
Figure 37 depicts the heat exchanger during assembly. In this figure, thevacuum bell jar is not yet in place over thi exchanger on the base plate. Thetubing to and from the cold end of the heat exchanger can be seen to be curvedin two planes at the connections to the heat exchanger. This is to ensure goodthermal mixing of the gas in the tubing before the temperature measurementsare made. The nitrogen dewar is on the right and the cryogenic valve in thecold helium stream can be seen to the lower left of the nitrogen dewar. Thetubing leading to and from the nitrogen dewar was later insulated with 1-inchthick foam rubber.
72
.. ... .. ...
r .,r ii i.s t iI i r,,,
Aftev InHWl tpnt t ntt of the heat exchange r, it was dliscove red that tilt ten-SJItuI od ce~ignwd tW ninkiii~n tOw heat rvxvhaimilgo wundt' axial c'ompr).ummitmiwos not load ted to the proper tens niu by t~he vunfld( 11w'th tendS on in thu rmt )dwassubmequentl Incmvreased and addlitionial te sts WOV MenadC. T IMh 11VIeSLINe MAtSnucle ibthore the Adx tenvioun was increased arec i'eteriutt to as Test 1 in thisrtepovt, and t hE tet'St Rftelr the tenkSiOning LaVe I'flu'reLd to ws Test 2.
The test bega n with air stream -to- streami hoL um leak rate mie sa reumentThe st reani-iitl-st roun lea k was fouund to be 1. 3 7 x 10- g see /p uS t rooml
* ~~teni pe rantu on ad nea r atmosphe ric pressuare. This lviak was c'onside red 5I11LLienoughi to War rMit pirOkreding withi the ef feet iveness and pressure drop eniv-
'ivlie hot f'\(0hungt wus vLS 'oled by thiwtiwig hut luni thr-OUgh tile VXChlI~vlngand. he(L W adwr ttt Ik Iq nittI'guiII W ie all owing ai port ion of thle flo w to e sca iwOut tl he bleedt. T he? I' 1.)0 Cloi rate was suc'h that 'T2 andt 1 IT tiec ruced at. u
r 1at v the lA~ iono 1excSirg wasooldwi th: dat Z m taken uising tiehz floingJpt-oeedure. '11Pi premsures were mut cit IS cl es~rd vuilues, and the mass flow\YA~ irdjumtil Th ''e flow Msuili attI eICln Was a I towd in 11IO i UL11 tin ajitPt.~-
* fl~~int ely I htutur t'tfove lata, pohinslist Id Iin 1'utde HI "Y' vp lnii . hi 1Wi~ I, A.Itft-rAl t'tkv piiints wereo taken t.t or near' the desi4gin prosqurvi rat ini while Hi h
sM Rt wCIS Ilo ,k tl .I Ihr Ril ifiin0 Iiu prvwsLire I LOS0 t iilt t coIIIi IId thV ti' iii WWIti
wa'LLi taiken lto olis rye I tilt v'fe ttcl 'P the pressuu ' dill' v uinndto 1hw tiwWt lg,jIvul kge Nut wueu Htri 'as oa WIT ti.and v1Ysfeet i \envs-s .
HEAT~~l t-W IAN e: it1,
.OINI RAL E LIECTRIC
After the six points in Test 1 were taken, an attempt was made to measurethe leakage from stream-to-stream. It was found that the leakage had increasedso that it was above the range of the instrumentation.
Because of this high leakage rate, it was decided to attempt to increase theloading on the heat exchanger tension rod. It was discovered that the rod wasunder tension well under one-third of the design value. The rod tension wasthen increased to the design value (15, 800 lb warm).
During Test 2, additional flow meters were used to allow measurement ofInterstream leakage over a range of several orders of magnitude. The stream-to-stream leakage was then measured during the Test 2 cooldown. Table 9 pre-sents the leakage data as the heat exchanger was cooled.
Table 9
LEAKAGE MEASUREMENTS
Test No. T2 (OK) T5 (OK) DP (psi) MI, (g/sec) ML/DP (g/sec/psi)
Test 1 29)8 298 5.6 7.677 x 10-5 1.37 x 10-8
Test 2 300 300 5.1 3.3 1 10-a 6.48 x 10-15.6 :1.56 X 10"11 6. :6 X 10"G
9.4 6.44 x 10-5 6.86 y 10"9.8 7. 13 l0-5 7.27 x 10"
204 300 10.8 :3.03 X 10", 2.81 y 10"311.8 3.63 X 10"- :-3.08 y 10-3
154 2117.5 6.5 2.02 e 10-1 3. 12 w l0-37.2 2.71 y I0'' 3.76 y 10-"
120 221.5 4 8,,07 100" 2.01 x 10 -24.5 1.02 x 10"- 2.27 y 10"05.5 1.24 x 10-t 2.2 x 10-a
123 139OK 3 2,42 y 10-1 ,4. 07 x 10"3 2.42 10-1 11.07 10IO P
(O8 202nK: 5 1. 11' 10"1 1. 11 i5 j nn-•tO :(. 77 10• 1 77 ,Ill 0-P
1 1 2 . 5 5 . I I 10t O " 4 . 0 !' 1 0 "P
,io flow through the bleed
74
r. ..INI .AL ... LI.... IC.
After cooldown, the efficiency and pveessure drop were again measured.These are the two points listed under Test 2 in Tarbie 8. The first was takennear the design pressures, and the second was taken with the valve in the tub-ing leading to the nitrogen dewar wide open.
DISCUSSION OF TIHE TEST RESULTS
The data taken during this test were taken at temperatures lower than thedesign condition temnperaturcs. The section titled "Relating the Results toDesign Condition," in Appendix VI, lieut Exchanger Analysis," presents ananalysis of heat transfer and pressure drop in the heat exchanger. Equationsfor scaling the data to the design condition are developed.
rhe specified pressure drops through heat exchanger 7 are . 019 atm (. 271)psi) on the high pressure side and . 022 atm (. 3234 psi) on the low pressure sideat the rated flow of 7. 688 g/sec. The measured high pressure side drop can beseen in Table 8 to be in the range from 1. 7 to 1. () psi.
As a result, at least in part, of leakage, the heat exchanger will not meetthe 98. 50 speciZication for effectiveness, The measured effectiveness at de-sign pressures never exceeded 92, 6'%. When the stream-to-stream pressuredifference was reduced from the design condition, an improvwd effectivenessas high as 94. 90/o was observed. This indicutes that stream-toi-stre.am leakageis at least in part causnig the low heait exchalnger effectiveness.
The method usedl toi cralculate effectiveess anild NTU frorm the data isdescribed in the section titled, "l)ntn Reduction (oCrnputer Program," inAppendix VI. The effevliveno•ss calculotion Is bWsed on the tiveragr, (f thete mperature (lifferr'nc es between st rvan s at the Waicm ant1d cold' 1) nd iof thi,heat exchanger.
Lt'fi'c' ivetiv smI -'1'1 T 14- ' T 72
"T'he effect of leaikgi till till, tit tx a'lgailcl, 'ft't vetiulless Is tlnii•y d illthe section titled, ""ffert of Ilid tin 1etik :e, 'lerimil ItI diattwio, and Axiii("atiduct ion on [ leat lEx h-ntige o' rfiormnliiic.," lit Ap ipen dix \ 1. 'l'h ieriftect iostreamn-to-stream leakaigo in shown it) ile•siec tIII ctld i'ld !lmp ir i , L t t'ferenlp (QrT2) arnd to huve littlP effect ,n the warm end temnie rat tire difference( II l). A leal. to the caving of the lih it t x(hIang% ,r t.vIas.eS t0i0 11W !)VeHt kIirside (if the heiti Fx'chtiliger alnd iln,r','tis's tho, tlx t, iio ro, 'iffi', ,tll tinl 'm )t hends. hi ruo wring .pprotix in te' 4'1piat (lIo, a rv devr+l• t'p d ill th' t itrill, Hivi'ttionat Appienidix VI.
1 1'MI7 T -- 5 -tT_
rININAL@ELIChINIC&T I M 1,
6T- - -12- fuc stream to stream leakage
Applying these equations to the data in Table 8 yields somre informationabout the leaks in the heat exchanger. In Test 1, the large difference in thetemperature differences at the two ends of the heat exchanger indicates thatthere was a large stream-to-stream leak, This leak appears to be substan-tially less in Test 2 aftert tension on the bolt has been increased. This canbe seen from the fact that the temperature differences at the two ends of theheat exchanger are closer together, A comparison of the warm end temper-atuv'e difference in the two cases indicates that tht leakage to the casing wasnot arfected substantially by the tighltening of the bolt.
The two data points in Test 2 show the effect. of decreasing the stream-tn-strt'am leakage on the strearn-to-strean't temperature differences, Theratio 6Tl /6T2 is much larger in the second data point, where the pressuresin the low and high pressure sides of the heat exchanger were brought closertogether. This would be expected from the preceding equation.
The first 'Test 2 data |p)int was taken at the design pressures for theheat exchanger. When the temperature differences are applied to the pre-ceding two equatitons, a streim-to-stream leakage of approximately 1 /3 themnnss flow and a leakzue to the casing of 1/20 the miass flow are indicated.
The spec tifiatitn for the heat exchanger st ream-to-stream leukage was10'"1 atim cc/ s,,c at 20 psi pressure diif',rvntial, which cvrrSponds to approxi-,lately 1, 6 x 10"" , /se,, The leakage test results in Table P indic,,ate thatthe heal exchungeet' exceeds the ipecified nimxinium stream-to-stream leakageat t'o i too m lo pr• •ila 'e.
Se erali feat ares ItI' the leakage f'rom st reo'rn-to-stream Lire ini, lntied h ' t It,dauta in l'tible 9. Tl'he room temlperaltare htakagR. deeretised aftt r ilht r•.i'd• , .4ItInll'd(, :,8 ,,aNii. ili(,•I•td Ihy thi, thermral dah,, ' I'tk .lvl;,•yiii 'easvs drtanma:t .a Yly with dII r-vt'(! u |4111g t enlIpe rat ut, 'The , ekule w IS vXlerj ilivelitt I I h)InId ti Ihe lineairlv dependent om the strean-i-ta -strurani pressure, (iltl't'-I 'evf, 'llis III-,ticvates thath Ithe Ink is laminar. The d(lpenlde nch (if the teak on tvinlr:tur,ctan be derived trinm th,, Iamtna, equatin lI',' the trictinn facto r.
61,a.
(4 l'AMI
IIt•I) ti .:, ci t: ntrr
i i -'tl.... I "it
GIIIiAL@ 11!ISTIIC
Sviscc)stty
A cross section
L, length
The leakage increases faster with decreasing temperature thlan can bcexplained by the increase in density and Ole decrease in viscosity consideredin the above equation. This indicates that the hole through which the leakageis occurring .hanges geometry with decreasing temperature.
The last two data points in Table 9 indicate that the stream-to-streamleakage is more sensitive to the temperature of the cold heat exchangers(which are pressurized but through which there is no flow) than it is to, thecold end temperature of heat exchanger 7. T his indicates that the stream-to-stream leak is in one of the colder heat exchangers. Since thko other heatexchangers have lower design opurating temperatures, the stre, m-to-st reamleakage would probably be much worse when cooled down to design conditions.
The next to last data point indicates that the heal exchanger would leakapproximately 1/4 the mass flow from stream to ,treani when operating atthe design mass flow and pressures. This agrees fairly well with hOw pre-vious estirnate of the leokage Imsed oil the warin and coul Itemperaturedifferences.
The leakage to the (cas.ing culd t )' II? anv ()f till, hII t vxcv i iger'ý , andthere is nothing in the data to, iulicit, whert. it is, If tHl h i laik ix: also in coneof the colder heat ,xc'hnmg~ers, 0l will iticrens(c it the, actuatlI o)perntit.igtemperatures.
Section 6
RECOMMENDOA1O
TURBOALTERNATORSa
Miniature turhomilternators with self-aicting gLIB bearinigs~ )III%,- beden de-v'eloped under this contract and other cmit racts to the point that they appearwvell suited for those applic'ations where t'ivliabtlity and long life are (ot pri-mary importance. Provided that a fluncontnit at itiitg system mill be di evlo)ped,it appenrs thait the only drawback to thle t 'ype of i1urboaitertiator testedl under
* I this cont rac't is the relatively high ciost * It is, theref'ore. !'ectnimended that* ~further effort be devnted to COSt reduction In the aren of turboalte rnators.
* CRYOGENIC HEAT EXCHANGERS
The diffivult lea withI development of' a heat exc~hatige r of 11hi' puast ic undw I I,( me sh t 'ypor are demecribed In this report, Untortunat clv, thet cauues ofthle '' )n t ruct ion and lea kuk pojroblenms wvre not identified iii this pri jIe ci It
* is, therefol'e, difficult ito ma ke recomtnendat ions for furt her, work oil thistvpr' il exc'ha nge r except to prpose at monre fundam enta ui ppr-oach. with uit1' tailed Ina lyi icci I aind v'x pp rinivnit nI inl vesat Ikgat ion of pta St i(' -1 ()--met a1I hnding
and cliffe rent izd t hlernial c'ontract ion. tirthe r i-econimendatinns ran lit nitdeia It i' I tIPie' flx't (ICvefl ipnIelit phasHe. now being, vtioducteci unde n' ant it er c mli' rut,is comiple-ted.
()Ithoir t , Npes ()f plastic' and metal ecX(hunpe tol' UP thei p)(t-f~tratt'd-pla~t' tpeaI-Id tlil'tvxpanlded ini ('Ll ti \ s'. The I)V Ifilatt'it' pluti.t' tvpu %%us I-u'et''t-d fuI) ust.Ini this cimtint rt bera usc I if ts high 't isl c~impa rod with thle wi re me sh. Ih Il"v'vet, illie p 'IwI4 outed- p1 'iti' heat e' i'liiig'e'ha. IIS ntty-1111N beenl parlHIN' do-vchipdatid I'mutid it) he v a lvelvIckiight . I'e n'mps NuteMI n n rdctino;ttl-'1rrirao' ;'kits -with (-(%'ri's wmuildI hv uiflivd. I.:Xp.'mtdI.(I IlIvi'lal is a
stk\ (1 " Iutit I 1cmIt 41,. I 1' !, irl-'ac't I hat ha :I iv;, ( bil -iti I t vvi' i;.t:i ot vd fir its -I li'hevat v*X'Ichaiguyc r . I 1'o' limina rut dvv i' It pa rit I tostIlt.- illdI c~lIl 11. hat i'a Iri IH
nol r'a pi-ilfr wi',vh~i I 'thi s imit '1.ntnnfir w tl I, 'tI tt:IatdIii i
Ii.' amto'iIw pt(h \'40'? II I i (J t\u l t'ti ssin dfkto iii iilu P ' s no'i~ A n ' l.I tnti orh no' t'
0 ININAL) ILICTRIC
APPENDIX I
PRELIMINARY TURBOALTERNATOR DESIGNS
Cycle-study computer Run 479. Figure 5, established the tu rboalternatorrequirements showi) in Table 10. Also shown in this table tire principal turbo-alternator design parameters given by the same run, The three power outputsand overall efficiencies must match the cycle requireniviis. These turboal-ternators, of course, were not optimized in the cycle study; hence, the wheeltip diameter, admissiun, and speed tire shown only for reference purpoves.
Complete turboaiternator deignn situd hWs we I' ,0on1du1Cted for all three tifthe stage. for the required cycle c onditions. The best operating geovet rit'
Table 10
TI'R IHOALITEll N ATlO.) D)EMSOCNSIrnt ('--,y'Clc Study Min 4711
CLycle Study lutamtentrs Turl o-lt z'ia 1 uv lnlet T',t, pvi'atu ,'t14" '5K 170*K
Inlet tvr tperatur., C ) 125. 2 (U! 0 30k, 0
lnh't prit,• ouz, (ptiA) 2 81.4 22 11, 7 42, 1
O(Utlet pri,.4ure, 1p1 ia) 17, 22 1I". H2 I ii, F,7
Outlet pre.SSu rr, ( atril I 1 62 1,145 1, 1 2 H
M a mN flow (l1". 1I;, K .1 11. 910 HI
\lMa.4 Ilt (III , (1107, 1, 0 27 (1 0. 0011.14
M ass I1(m, (Lc , ' , 1.1 1. 3f; .-49, (0, 714l
J (,, I I fH! 1
I',• i &,',, t tll)L~l,'' ll, ti.' tit),(4 i) 1,I 1. O f M W
......in. pole blank
Preceding page Mank
I.
OUNEKAL * SLICTAIC
Iit. 'ui' lug LIvHil�t1�4 V.'i'I'C u�v�tiiiii'.'d lot' �Ll tlll'CU 5taL�L'�, �tid t'titiMiCJi' I'�ttiUflM lot'tht thrUidt - IK)i I ting di 143 griN v.' t' LV z'i'v It'WI' d. I .ini it iii u.� t'uuii.li tions on tin Luil tint
comply tc aliLti \5i M V. � inadc jot' tiit' 0. CO-inch - di anictc v t ur hi ut' w h'u 1 ot tiicI �K tuz'bin'. :\ t'uvli'w wns then niade of the two 1tt�rn&itv rc�at'heN lot'tlv signing and cun t*tt i'U,' I lug tin' 1 70�' K tw' I*iLI1 tt' i'tn t itt' 8t�igt, attil 'o�i�iUc r dl ion
was given to (hi' Usi' ol' in i.�istitn� Uovt'z'unnicnt -tUI'n�iMht'(l unit.
'I'ht,' thz't't,' tlc'u�ili.�ns that wt'z't' z'cc'oninwndL'LI alt' shown lii 1'ahh' 11. .*\
0. (12 � -inc il-Li) �in ii�ti' t w iicu I di.' sign I �c al Mo t.how ii In i' thi' 1 70"k I 'a ii hIlt.'. 'I'lu'12, S, Arn�v I u ulittie rei'�' ,'t'.'d lit it� th.' unit IWL'\'lllLI8ly c'iuixl ruc'tvcl InIdV u' I.'. �.
Ar'tit'.' ('oud ra.'t Nt.. I )AAI%(P2 -UH� ('-0320. Ibis wilt iiL&i� 8 wli.'eI tiittittut(i iii
IURSOALThRNATOR DESIGN GOALS
( 1rie goal fop t he ttirtitialtc� ruatti ?' designs ''. as In usc' Intent i�ii purl s whet-tNt'?' PoNlililt' in all tiit't'i' Iuut'bttalt.'ruiutt'r's. I 'iii'ltit'i', liii' clet.iigui itt' the t"�'dst -
iii lit' itt I' liii I * I'' I ii iii, -i'iil Ii' t,.. 'ut t �ii �. I I ' I I I ii I. .1
I it' :hlt't I.
IWEIAL9 LECTRIC
R I itI ) A~l E
I~ T I. H
4AI'1) %
~ ~ ~ I i I g I o .J u 1 1 1 .1 a4 I t I~ 1~ 1 4 h a.v t I~ "i ,, Is 14 - 4'ali
4446' A l k f 11 H U M ' IV I : Uii- I M:J~ 0 11V I
I'~~~~~' K)''4II~ i
III to So Il-.1 r lo) I'l -t '1 '
I, v
h,
•'•' • O NiNAL ti gl
The design approach anticipated for the turboalternators has been to usethe laminar-flow, radial-impulse turbine blades that have a con erging flowpassage. This flow passage is considerably different from that previouslyused on the early General Electric and U. S. Army turbine wheels. Theseearly turbine wheels used a constant cutter diameter to form the blade pro-files, and therefore there was a constant-flow, cross-sectional area betweenthe turbine blades. This cutter diameter enters into the design program cal-culations and also the performance of the overall turboalternator and is asignificant computer-program design variable. All the turboalternator de-signs were therefore considered with a nominal cutter diameter, so the trendin efficiencies could be seen alouts with the blade-height-to-cutter-diameterratio. This ratio is a significant consideration because it is an index to howtime-consuming and costly it is to out the particular turbine wheel with the
miniature-tracer milling system used.
DESIGN FOR 14KTURBOALTERNATOR
A number of runs were conducted using the design-point computer pro-gram that designs partial- admissIon radial-impulse turboalternators, Over-all results are shown in Table 12, where various nozzle angles were usedin conjunction with a 0. 50-inch-diameter turbine wheel. For the coldest tur-bine, it is desirable to maintain a small turbine-wheel diameter, to kecp thedesign operating speed up to a reasonable level; however, previous expert-ence in manufacturing and assembly has indicated that the turbine wheeldiameter should not be below 0. 50 inch. In this particular design, there isno appareni reason that the turbine-wheel diameter should be larger to de-crease the design speed further. This table shows the variety of changeswithin the design before the final design basis was adopted- 90, 000 rpm isthe best design speed, along with a nozzle angle of 80 degrees. The overallefficiency is a few percent higher than the design requirement, providing adesign contingency. A configuration of 23 blades is best for high efficiency,which also results in a blade height-diameter ratio of 2.0. The fact that the23-blade configuration is the most efficient one i.s contrary to what has beenseen on other designs, where a larger number of blades can improve efficien-cy. The last two columns show the change in axial clearance efficiency. Anefficiency gain of about two percent can be obtained by decrewising the axialclearance from 2. 0 to 1. 0 mitl.
Table 12 indicates that there is a trade-off between the nozzle angle,speed, and Ulade height-to-diameter ratio, The larger nozzle angle increasesthe blade height to accommodate the flow,
A complete printout of the adopted design point, computer run 4791007,is shown in Figure 38, which shows all of the significant input and outputInformation. The journal-bearing clearance and power loss compare rea-sonably well with that for a similar run with the tilting-pad gas-bearing me-lector program of the Franklin Institute Research Laboratories. Of course
84
Table 12
E. DESIGN VARIATIONS-FOR. 140K TURBOALTEI'NATOR
Run t, RM I R'UA,3 A'."n4 ROO .6 Run?7: ktun8
N91121 mile(dro$) 7at6 ¾: ..8 IF sFo . . ..
Nominal cutter diameter (in.) 0. 03061. 0.09081 0.60501 0. 03081 0. oil 1 0. 03051 0.03051Mlade height (0h. 0i5U O:o '0.0606 0.0895 Q0 0841 9,9701, 0.01-01 0. 06"U.
Sldýhegttoqto-dsntt aio170 1.70 1170 1.00' u. 59 3.00 1.0goAxical clearance (LA.) 0.003, 0.003 01003 01003 0,003 0.003 0.0019*Adoptod 4s pr~ltminaiy design point, C R211 Ji
ik there should be a good correlation, because the design-point computer pro-gramn for the turboalternator incorporates the tilting-pad gas journal-bearing
selector program. Howeverý, there is a slight difference in the value of theturboalternator design-point program, which shows a journal friction ofI0. 09007 watt, while the bear Ing- selector program shows a power loss of0. 0875 watt. The actual total power in, of course, the sum of both sides ofthe thrust bearing and both journal bearings. In the design-point computerprogram, only the one loaded side of the thrust bearing has been Included.Both the thrust- and journal-bearIng film thicknesses do show reasonablevalues: the thrust-hearing film thickness (loaded side) Is around 400 micro-
Inches, and the journal pivot-film thickness is shown to be about 140 micro-
DESIGN FOR S554 TURBOALTERNATOR
The first s eries of design studies was conducted with a 0. 70 -inch -diameterwheel, as was used in the cycle-design program. With an 80-degree nozzleangle, the beat speed was determined for a 43-blade turbine wheel (Table 13).A worthwhile increase in efficiency could be obtained by decreasing the speedfrom the 161,700 rpm calculated by the cycle computer program, because thecycle program optimizes the turbine on the basis of aerodynamic efficiency,and the parasitic losses are not included in this optimization. Inclusion ofthe parasitic losses lowers the predicted speed. It can be seen that with a0. 70- Inch -diameter turbine wheel, the efficiency can easily exceed the cycle-design requirement of 0. 374. At 130,000 rpm the predicted efficiency isU. 40-1.
A second wheel diameter was Investigated with the possible prospect ofadapting the existing U5. S. Army turboalternator with a wheel diamneter of
0. 625 inch. Table 13 shows the turboalternator design results representa-tive of different speeds with different numbers of blades, It is seen that thecomputer program predicts efficiencies greater than the cycle-design
835 .
GIN IRAL~ ILICimIC
'16~I4 CASK 41910 .7
EL9,1'rAC:IAL Rut.JkN our1'r cwArm) j 42E
0 VltvAiLL PFFPI( I EJC Y CFIMAC T I O) 1. 111 6,4t+.0J
£'JL~r r~iMP~t'~ Utl (K) 0.#1 40 UK+ 01
I.-JLEC tJRWSSUiAC (ArM)
'3JrL4-*A PýF.jA (AN)0 0* 1 172E+01
KLUOW(L~A33N) 0ee7)
FLOW1~G 0 *6U~I4
VUL4'A OF F' ILADE K0l 0 . 130 E+~U.3
1 LA 0 :~ H ll G WH rCW1AN 0 10 'U13 7 CE-21691
t iur 38. Deig Uor C4C Ttro1t4o EAcc t IN 3)
L 0 A0 C I S0 1 E-06
oyriLDEEtfa 0 s 10E 1
PHUMA.RY LOSSKISALTERNAFOR f0FAL (WATT5) oi12+0SUM OF F'RC101~% LOSSES (WArrs) 0s6J67kE+QQ
NOWSt~AUR P'hNF00VA.JCF 4O.OTHERC, SrAUc 14 TEMP() ,4uE0TorAL HW~ LEAK~ cWArTS) 3.EXIT TEMIPCitArumc cK) 0#I1218R+O2-SNAFr HEAT LEAK CWA~rri 0.HOUSO HEAT LRAI< CWA~rS) 0.
GAS BEA4t1'JGStie0rArINU AS~iEM4LY WE:IGHf CL8S) U $95ELI
AICELIERATION OF GRAVITY C"U") 0010u0E+0)1rer. 13KARI.'40 FRICTIWJN (WAI`C 3 ) 0a175SE+00
,P INPORMANCK TERAiISENrtNW'I HEt%() cmT 0.71 54E+U.4HYGNAULIC EFF..o FIRST I'Erii 0 e59 6:3 R+00HYDR~AULIC PEF~s. SECOND i) HIM U1l~
flIP OLFARAN~CE ~FT.F COR490r1I&I Js9:ý6~3C+00
RLAIUE Rte 40s. IF7P COAR, 0 0 1 uu5C.#jI
ALTEtANAt~ii 04IP FLUX CAGAUSS) 04335'/F.+01ALTtivorwi COR~E LOSS CIEFP7. - ~t ýALrmqjrwtidr 010POI~ LOSS COEFF. U.7k33.0.
[E.'IP!.ANrU* WH4EEL P)(iF GA3 (R) 002) +
'3LAOE 1*'l)KCF. A'ULL CUGHEKS) 01 5H.0:ISPEC0IFIC DIAMrER s204+0
HYORAULIC LFICIE'41Oy CF)4ACTI~JA a45E0eLwc'r,46MAG~p ,re m CtmuruIN) O.V93')EU0O
WHE~EL.lob $ WIA 'G)Liul (WA~rb) 0*3076E+-),)SHAP1 POWER minJFur (WATTA) 0**10i3R.+Oý jCH-216D-2
Figure 38. Design for 1411K Turboalternator (Sheet 2 of 3)
87
lINIUAL I* ILICTRIC
DESIGN CASE 4I1IlUU7
PAI4AS1TIC LOISSES1 (WA~rrSTUitSIJI D15C FIC~1TION 0*243C*OOOJOURi'4AL 'OIAM SHAFT NPC1C10i1 O49~A~pvrr GAP rK0T16N 0 OI'l 5.-0 I9cAtLIeU FAI(CT10i4 092756MU
WHIKEL FIP CLIEAiiAlCr (10skuuroBLAOLV lAS.SAUE CUfrhkr IiIAMEf~C. C(P0 J.3015 01rSLADE FHA1L1M3 EDGEC r($1UPF (10~ 0m4UOE0K-OJADMISSION AAC (i11ACT1~3:N) U.*d7149+00ADMIii1lJ.-J ARC CUEGRbki) U*9772[409
N~LE ANU4LE CDEGACEC) O00800c+08BLADE ANU'LE (DEGtCE18)Us6-0 +
WHI4CL 1,4ldDl DIAM4ET~i (IN) 06OJISICOUUSLADE.' PtLS.SLIIL SUuA7ACL ~AADIJU (114) 0 a 47 6.-J I
T~ALAD SUCTION £14AFAK LhADIIJ (IN) U. id 61 F-U I
JOUA'ýAL 111CC SH4AFT LF.40TH CIN) 0O.I366F2~o1ALM10rwii~ DIAMETER~ (11) 04..A011 0E0ALfv'JirIoi4 mpam.'p L6Nc'rH (IN) 0 8J.316E+00ALiEJ~rONI P[AI7. SPEF~.Web.~s) O.I~jL+4C03A~lC J R4ADIAL GAP (1.0~d 0.1 5011-01STArJRt OdVtHANO (10 a~4 04,)a0li[- UIsm~itt LAMINAriav D1'~. cv 0IN44)KWHEEL. 'ACK sioc urzommum GA13 cz11) U9161OKI-02o
VELCL.0 fAi~sisouriqu vELociry (FpOe6hds+MdeL~K uIsCHAi'Wb VLldIr (F~ PPS) O*l.6Jflt*03SL.AUEIJA LFr RELATIVE VJkLOC[TYCFS) ).IlE16+O,3ALADF.IqL~ H iAUIAL VFLWdU~rY CPI3) 0,1061'C+03BLA)F~ xi\4~r PtLvr1VC MACH l~l~p 0 s17 4U k1*O
TURBINE DISC fRKYNO1LDS 406 Ue6JS4ZA+U6JOURNAL DILAM~1rEA REYNOLDS N0.0l$6C0ALTR.AArmi GAP REYNOJkLDS 40a U~d7?76Z+US
DRiAG (ICVlM~N'i4
JOUR(NAL DIAM1RCm DRA~G COEFPe . 1. 4~- 0 P
ALritiiAFoi 13AP M01-MAT CJEFirs 0*1 396K.-032 IC1-2169-3]
Figure 38. Demign for 141K Turboa~lternator (Sheet 3 of 3)
88
F%
Table I 3
DESIGN VARIATIONS FOR 550K TURBOALTIOHNATORNominal Blad de-height- Azial
Nozze Speed Overal No, Cutter Bed to-cuttir- Cleariane
RnAngle SpEd fiOvenacyo i Mae[ Cleaanc(del) (rpm) (tration) Blades Diameter 1in, diamaterS i:• On.) Rqatio
WhiI amotort 0. 70 Inch1 80 100,000 0.3740 43 0,02151 0.0859 3.06 0.002
"2 80 151,000 0,3837 43 0.0315? 0, 065 3.05 0.002
8 s0 145,000 0.3906 43 0.02157 0.0657 3.05 0.002
4 ;o 145,000 0.39;5 43 0.02157 0.0057 3.05 0.002Sso0 140.000 0.4000 43 0.03157 0.0056 3.04 0.002ai so 0 135,000 0.4031 43 0.02157 0.06056 3.04 0.003
: • 8 30 O0 0.03 3 0.02157 010656 3.04 0.002
Wh iDiatneter: 0.625 Inch
8 80 130,000 0.3945 37 0.02253 0,o058 ,. s 0.o0029 80 135, 000 0.3970 37 0,09253 0, O2bi 1.99 0.002
10 s0 140,000 0,3983 37 0.02253 0,0657 2,i92 0.002
11 s0 145,000 0.3985 37 0,02353 0,0857 1.92 0.00212 80 145,000 0.3011 31 0.02767 0.0657 2.38 0,002
13 '10 145,000 0.3865 39 0.02985 0.065 3.2.0 0.002
Wheel Diatnoter,.. 0,,7?Lnoh14 80 135,000 0,3600 47 0.02107 0,0657 3, 12 0,002
15 S0 130,000 0.4024 47 0,02107 0,0656 3.12 0.002
16 60 125,000 0.4045 47 0.0210? 0.0656 3,12 0.002
17* 80 110,000 0,4038 47 0.02107 0.0654 3.12 0.00218 8o 130,000 0.4051 53 0.01823 0,0656 3.60 0.00219 so 120,000 0,4016 43 0,0234 0,0657 3.80 0,00230 80 130,000 0.4234 47 0.02107 0.,054 3.11 0.001
*Adopted am preliminary design point iCR-11741requirements, but not as great as in the larger 0. 70-inch turbine wheel design.Operatlng at a higher speed is also not as desirable.
A third wheel diameter was considered for this second-stage turboalter-nator, as shown at the bottom of Table 13. With a 0. 75-inch turbine wheeldiameter, a variety of computer runs was made, with a varying number ofblades and some changes in design speed, It is swen on the table that slightvariations will produce the expected trends of slightly increased efficiency,with a larger number of blades. There is no compolltng reason to chooseone of these designs over the other; however, it is desirable to maintain lowdesign speed, to minimize bearing losses aid provide an adequate speed mar-gin for much faster refrigerator-system cool-down, The r commended
$1INRALe ILICTRIC
(Ie Sign I• ,w I 1 ';'•1 T lh , 1:;, iiO , 7, W1 1re ',7 Ire ) ItA, ri' P0 q, •It ii'e rdh', a u. 75•-itlch-d IVt it--U, IttiI' H•, 1 ,11 t, Wh oj)4t'cL it 1i4 at 120, JOG rpm . A comlplete )p.;intoiatof the d t',si i 't'o fbis 55' K tillt( Iilltornahlor is .4hown in I"tigrt 'e 3f),
DESIGN FOR 170K TURBOALTERNATOR
'fabic 1 I shows unmfnlar t of the' I urhbo Ittrrnator compute r'- progril,designs coni•i•th d. 'rht, fivmt sOe, oi t'ornputer design re.itilts Is shown for a0. (125- inhh - iani :,tet, ' ut-hine wlvet'I, 'l'hi wleeul diamntivv, was selechted he.-cauiet of tht possibilily of uMIn I lthi, ,'•isting U, S. Army turboalternatou', Thedeign 14hov - tlint the riquired e,('kt'• ,ny (f' 0, 3062 apparrntly can he rnetwith a H4liall at' in hult Witli a v z'li .h tiesPign speed of 230, 000 rpni. Oftoll htta 1 'Io(sAig speed allows lilt'li, oljipotunitý for a wide, margin froom aninaitia qtt[ ,wll Spe ed. ol ll ho o rdtul. wH. 250,000 rplni. iigtutv 40 show s a coni- ,
pletV dea~tý!11- lt-t0t• 'a~l'l ('01111)It~ltt ' I)PHJllItl~l. '
l'uhl 14
I)151(',N VAIUAI'I.)ONS l.') 170' T TUIRHOAI.TIIINA'I)TOR
Nor-mle Overall Nominal Blade Slade-height- Axial1R'&| Angi f8peed No. Cutter to-cuitter(deAle (rptll) Ef'flc eiy Blndes tanister Helght te clearane
(dieg. )(n.) Ratio
WhMl2 Diameter' 0. (125 InchI 80 -00,00(l 0.313, :17 0.02253 0.o0 19 3,84 0.002
2 80 :,10,000 0. 31115 37 0. 02253 0. of;18 3.64 0.002
3 80 :!20,000 0,3184 :17 0. 02253 0.0618 3.84 0.0024 80 3 0, 00U 0,3188 37 0.02253 0.0618 3.84 0.002
Wtee Digaeter: 1. 00 inch5 80 180,000 0.2863 43 0.0325 0.0621 1.91 0 002
( 80 170.000 0.3036 43 0.03235 0.0619 1,91 0.002
10 80 130,000 0.3155 43 0,0325 0.0618 1.91 0.002
it 80 100,000 0.3121 43 0. 0325 0.0618 1,91 0.00213 80 1r0,000 0,3210 47 0.0294 0,0617 2,10 0.002
14 80 0O,000 0,3281 53 0.0256 0,0617 2,41 0,0021s 0 150,000 0.3282 51 0,0226 0.0617 2,73 0.0016 80 150,000 0.3277 67 0.0194 0,0617 3.18 0. 002
R C-217 5:
A 1. 0-idh -diameter turbine whti.,1 was then evaluated to satisfy the samerequiremenit,, but to operate at a much lower design operating speed. Thecomp'ter rwiu show that with so•ne varlation in the number of turbine bladesand opeeds, a best operating condition can be 150,000 rpm with a 57"bladeturbine whet I and with a reasonable bladle-heiglht-to-cutter-diamneter ratio.This conditlin is shown as fRun 1( in Table 14. Figure 41 Hhows a completeprintout of t,•, design-program Output.
00
I INESALI ELECTN It
D Li L G 4 CA i E7.O
F.pcr.ýICAL POR~ aurur CjA iTS) a .,J6I9+ud
O~VERL.L EFICIE'4'.Y CK4Acr10J '3.4'jlE
I'JLel rh;.lPEAr~tJi CA~) 090
OVRýALL rI3 l2Et: npoA r r ui iu,~Ir (ii) 3 I AN I '..ýj -.~00V'ML TEMPERAWRLJ '.Mtai (A 907llK
iautt i'p~miatJiAr4jqP)(rc
wii~ ~ýSSIJH4. (Afr1) ý) 1 1 41F+r'WI.;JiF 44A1 06 L. 47VX+*O I
FLO'~ tCIG/H4F CI'J)K+
FLOA )(Uli.). 0 sI .487)W*O I
WHE-(. ri? DIAMEE ,i(11) 0 1 WJ F+Q3
ACMLiAs AF CCLBADE5[)" .1100I'J)
LOA )(LIIS# )p J* 3I4071K-0
'31'r4I0F WM[i414 tiU~, .10K +7I)+
rigure 3fl Deig fo 050 aubatratr( I." oJ )
CLEie r DlA ti~lkl ,700 iE)J
PRt4AiY L033ES
Al K44ALNEA wwrL u.MN
I'Wk SiA~4'UE oA~kA
HFUA ULL rKPRAIC. (K)~I r*){ G 44 1;?E+ U
HOUrU AFArAA Cwl'S U#~~ ~~~.a
AUYCr CL C411 F(Hlv11 y3[~LC cF (~101 'J)
!;Fre uL-'IC Pitcr2- cw~r rrs 91,1~
PEFigure' 39 e infr55 ubalento Set f3
HYDRJI' F~o Kitjr mimJ@617921
IEN I AL@ ELECTRIIt
DESIGNJ CASE A9t
iPAriASL rIC LOS~SES CwA rrs)TUU31'JE ISC~ F~uc riaq U a 7213 t". )JOUR44AL DIAM~ iHAF1r FRICrIONa*, 300*J'ALTE~ANAUIR GAP~ CFlkkJ ).19Ao-
GCOMiETiA'
ALAUF l2AJALLW (~EDGE THCRA1EfE CIO e400 .~QI-0
'4Osi~laq li~c CoEutiCe) U 6 6 d C+J-.1
4LO ANiLE (Dl~u!EF.S) 1) 6 * +
WHEEL INSZ!ED I)IA.IEMI ( 1j4) 3 ý+
RL'ADC SUM0i4 :SU-iUE R4A')IUS CL10J U*d3J6bF'-J
r0fAL MA~dI'AE S-IAFr LEiGFH CIA 0O' s(1JOURN~AL JHAI~' ukiJL)rkii (IN') 0..6I04
MEýrsrTOrJ MU!KI L~EAU (IN) U.)O6lUIEU')ALTER44f3i 12 K R I Ir s~ GI Jt I'Ju KePssv L 1j . -+U
WHELr~r IMOK 4' j~rr4~ CA (11) Jel I l)KJI-I
VIrr0~ L;N~ IL r. I IE-SP U~ I'l: 4G~ VELO II r' Y A ( I FPS) Je U 4ý0
'1LAUE I NIL Ef EIL ri vE VELAC' I*rY ( FeS 0. 11 X4.fJ)
riY4iLU j)~Iei Z$
8'igUrK P39.AU [)euignLO for u5 0uatrao (See 1 of 3:)
RVYALOS 0* 63!"Ka
fjH0ALfEdtNAf+, P~Atif(AL A01LSS10I ,tADIAL IMPULSEDESIGNJ P011'T C0"1PUTZii aur1UT
I)CSIGM CASE 479.3004
RKAGERAIL0O4 PWA Oir cwArrs) a 06411 E+02EL&QfiNI(AL POWER~ OJrAUJ' cwArrs) 0 *63;241E+02$pF0 00301 '~~) 0 .2ji00E06OVViýiLi. EFF CIENJLOY CFAurio,4) 0339
viNLP Frim fivi~rit (it) 0*3060E*U.3'
OV~ii'ALL TCE1PE-4ATUJ0.' OHO~P 00t ij,j006F4.Oa
1,qLE1' Pti.iUR[ C0 a4 4P. 10 I.t+1*'l
Pasuot.: .,trio 0 .2 ESOI F+FL~)w (L-Vapl) 0) e 161 4[-J2
WI4FEL 1'1$ 'flA'4rER (11') u0b 3 ý0NUIJ4~E OF~ RLADF. 0@,37J0K*0P
CHttUSI' arAiL'J(
0IAM~rKA 4AA71 0 .buU,2 KJOCLýi ro ouke ttri~j 0110E
CL~ctL +3.AODSIIJUK( 1) Us 4ul tr-.3j
L.0AUC Li.S ) 'J89hd'.:-o ILOAD cJJ oiJ'i'3E~ J 52 410 ;. -,)
Figure 40. Design for 1709KTurboalternator, 0. 625-inch Wheel (Sheet Ilof 3)
9l4
BIllN I AL* RECTARIC
ALiJFI UrL C WAI~s 0 '4 7 A ý) E. +~ 0SU F irtRIClL(1 L-3S~t~ (w' r rs) 0 I-'M 3I F.+J I
tOrAL 'h4.Ar LF.A-ý CW'~t:i) 0,61F+.F. I' r rp c rji i. r, i ýk' 15 *
HO ~UjG 110tA1' L RAK ( W'frTS) 0 s 6 4-,S1 F.
IS r.' A ' " IUJ CW .1 U1d N61Cý s
t0 rA r:~~ I t.Jli t~qi yjr wL I t Ci r JL4. ') /j / ib IA'IL F~ i Vii oF u;&%v ry~ cvl3. Uj2I~ I* j[r'+ I
SISE ,VJ.jtr iJ A 0U' ' r); ij 9 1 4 1 E.+jI
HT 1)s i-( Ii IJ EFF 1 CIi'UW (TK& 2s :9 /1 ý-) Ir I r) CL v4 r< IL 11,1 1 F F. F F , ~ I'J , 0. 111 )C T3 I ON I Ai) t
81.9iU 11 L,*s A ,% *F~ CORd 1)4' s I17Ej
AHr~L r ENNA I'k ir I1"L J C K U' k 1.3) -)SJ. '1j I7 * u).I
44 Fi 4 f,'b V ri i IL03r1-)ij
ALi f±' V.IN.A 9A N~ i J t3 t) Fi *.J I:J
C rNjIAkI ' J-0,i~ KJit. A ''I S ri *, 'i.~*. C A9, F+O
N gur 40. iaqr orIA 7.11~ ')['urVOL s1t4aOW C, 0:;.)a 12-)uc 4-? i (Ehe - ot' I
Al I~ ibun..iw.~i
M*INmAL@ I UOTRIC ii clj~%3 CS 47930U4
h3ArA51i1C L~IjSkS cwArrs)
JOULR4A'L ULAM .SHAFT vii~criimJ 64:ALTENhAAF UI LiiAr' I1)4 J *6347v.-0 1
G F., '0 1 E N~ yWHE:EL L'1P CLEA~iANC( C14 0 st) vi (.)Uthk. -T t
6LAI)K TVI1LI.'Q E06LiI. NICt4ES C £.) 6 .10 1) U h:
AA)1IS:zU0 AR(~C IDl-'U IL j) U*661)bE+Q4
A04L.i KJV 143iF (IKUrFS '0)J d34000-j I
OLAý' NLF. COE(iid~h) U0 L'.J)0
13LAUE~ P.0:S5LJ1(E ShURFVIF RAD!Ui (1.4) 0s4J.43F -,)IBLA 9t%.4 r 10 .4 J U i iF1U K t% 0)1 U S 14 0J 234137K- I
IJJURAML SHAFf DIfAMlUIF (14 2 10E+JJJR9'JL K-iKE SHAo~r LEA.T8 (I .qJ) Oa I J66F4J I
AL.~ ~i~J~ I A4EK2'FCIý (1q 0i 0 .461 UE+JJ!%Lrh:N4Af)r *AAi'.Jvr LKNUNr C IN) ).04-ý3'I*j.
*ALr(,-1ATR0 R~ADIAL U A C 1') o 11CJJ r r..'i'4A"J(3 (1 4J)
SVAfui L1AivN~rLj4 IiL*m (oo'),)4;E)
S r'0j TI V U. VELU,'i ry ( o'ras) J 4 2 +J 4
RLA) IN0'1' r(EL.A 11 We ~L~r, I r YC(FP% 3 .3918E.*03RLA U L IL E r D 1A L V EL -3 , Ir t CFFJ 30 ).J)2E + J 3
9A0WI 4LI~ WrtiP7LA r IVL. I A1 H hJ4~FI UA El )I 799;c+Uj
rAEYALUDi '4UMHb)(S
NVRRIVý DISC RYAOLL~i *"J~ Jse4 A.4 6 E+ J
ALrR4ATO GAP riY')LO di. Q*344E0
JL Ut~iN~IA: 0-0-6 ;4) PjKP '0* J b i~J E-'j2~I~6-
Figure 40. Design for 1 700K Turboalternator, O.62b-inch Wheel (Sheet 3 or3
L 96
0GENER AAL E LECTRNI C
r.JJ L,) 14L I'E 4 k. l'J Ur R P Rr L Ai)OM1I 10 4J r4iA )I 1 AL I A JL S--)RSIL3N POI'JIr GH9UfoLr Juirtur
u)ESLGN CA:;E 479J301
:~IE~i1 w4i~W *jur cimrrs) O.46uU E+ci.
I .JL Ewr i~l ,i 1.,owJ~ ei ) o Lir * A T )L I 7'l:+P .ciSPE(AL Cim aJ, 1J~0 5, 1) K + 1
OVE14ALL ILPPICIENOV (FJRA r~CTIO) 0.l 7JJE+00
MAP LERAr~,' P P.uI E~.5~~ 0.1 KIS 0O+JOJIALEC'r MIPý.ATURE CO~i L). W60E*03
14LF'r FF.~ I. C)MT AI (K I 1't-
'.VIEAAL rimiL.AR,)rmK OtC p (K) 00 1~))~0iaur1e r~iL!ArwE~ 3x m)09! SO 0.'+ 0
uu~rL~ £i'F JLAfW E~ (K) J , 1i 5 ,2 E') 3~J
LI qLE ttFs~ .~4i uie iC * A a0 .-410 P,+jJ4
I~qLET~ PrESG N (Al'l)i 04.3 * *'J
J~l1ff( rI;MSUR ,~W'(41',A ) U 4 13 1 AKF.+) I
..IP A!)(L13/4R U a 58 1 ' E+ 1)I
91 V. &fl?~ 0141,10~ IPt (1 0) 11.) ) tý7+0 I
iigure 4. DesIUgn (o~170) 1uhaten r 6.0-u Wh1 See f3
I'+Wzi~ ~ C7E..4
SININALOMIETRIC
ALr~JI(,ATOR TgrAL CWArr3) 04640O3F+01
slim s Iw Ftlri
ro rAL -4PAT LEhAK( (WA NO0 O.37JIJu
AiI'fIrLNP. ASSVVFBLY wu(31r (LJ3ý) s,7K1
?4tNrAiLZC E~F.. Fi ijr rpiw(~ *17Aý0HYDR~AULIC~ WV.Fa.iIlJ *) #~ U*W41JF~I-UI
,'hIYO(i~.L1C F.F~ NiL;0 f~o s49Krip "LEAAP40 .EF ;J ,4rjME ii . U*945.3F+)O
NAL r FI) FFF 00+0C( 0A a"~ 9 140, F+0
ALrý:?JA hifj~ii LJ~ij CoiE~¼ Je2J56KC.3
'W Vzl4L.UA~IC PLO 00i')~a'- ,3 4 I)!D )t; I V J* F L OWJ~ h (LCi ) .0 ,, 7 0 2 3 - 1 1
NP F5FIraspuriuV.a0 I M-J 16 * ,
WH 04,L. E 011C I R~ 4C f ( FAVIC 1* 10 ) 36 5 6 K + u u
ai4'L P~1 ~' ~J I :~ [ .it r4 %; I' I) *J 774
j~AI PJ1K UfP'ur cw*Ai ci o.tA: i) C-26-
Figure 41. Design for 170"K Turboalternator, 1.OO-inch Wheel (Sheet 2 oft3)
9ININAL* ICTEIC
DESIGN' 17ASE 4i79 30 1
AAA~ASIrIC LOSSES (W~rrO
JOUR~NAL UIMAM 5rHAFr FIG 1CI'N 0 0 98f/a,+Uj
ALCUE1V4AFR UAP F1C~I'0,* 1 lE ,Iii .46 rac ho.4W.IKKEL rie ILAAAA~'CE (IN) 920%F13LADC I3AiSAGE uOrrEA DIAlA4fCN (IN) J 44 d641E-U IBLADE rRAIL1.4G ~LOUR THLOKNS (1~4) J.400UE-J2A41MSS10'4 AriC(U FAU1004) O16E'.AUMlSSIfON ARIC COI) RKES) U A IVU 0*J,.N44L ANGLE Ci)CRJ6iC) 03J&l3LA'E AN'GLE (UrCURECS) U6&E
WHFKL 1ASlDE DIAMEMN C 10 *4J)*JI9LADE I PES~JRE SUiRPAUK OALlUS C 14) U*4dh7FPU1
rIWJr'L MACH14E SI'4AFr LE4UINH CZ'4) 46P+J0U~'AL~HU~ ~lAEr~4C1') '026IUF.+J¶J
JZURAAL FREE~ SAAFr LE~'JiiH (14) J *I 366E+()I
ALrh~iC4ATOA ViAUN,.Jh LE4(3fH CIN) )eJ6UA.I.-3')
ALM44PAT0I4 RADIAI UAP C 10-j6 C )01 ~Ot. I~USrArJ~t O)VEd4A'I (14)J64!0F-IsrA rj L~m1IN'io, 1 1 atz~ 6 (LO' .) * ) /1 M "+,)WHEEL. BAC'( SIDE O0ATLMIMA (34P C1A) J * 1 173 E-J I
Vir.LOI~ r I S£"OIJ IL 4( VELLk',1 fY ( P?~) Je ~41? 0+ j4
N04L U)il'(:~k~oIi V~FJ-0C1 rY CFPeS) Ja1 430,'SLADE tm'.Er io.LArivE VELVV~lrY(FPS) )eJO'J~kLUJqLALM': IALicr RIO~IAL VELO(I ri CF?3.) .)3U*Apj;2+U.3FiLAiJK IALEI' dCL;IVU!~E M1ACH i J1i3l J. OhlOI lE+)C
iF.Y *JLO4 AJ"ISEti-400Ef ?PAISI3I *0;Y A0L.S 406J J *J .366iW.4J4
JJUiANA&. I)IAMWrEA I)V~Li 000'IP UIU 6 10tK-0 I
Figure -11. Design for 170"'X Turboalternator, 1.00-inch Wheel (Sheet 3 of 3)
WI hI ALO I LIGITRIC
Between the two different turbine-wheel diameters, the obvious prefer-ence would be the 1. 0-inch wheel diameter, where there is about a one-percentefficiency advantage; the largest overall advantage would be in operating at adesign-point speed of 150, 000 rpm, rather than 230, 000 rpm.
MECHANICAL ARRANOEMINT
A mechanical-arrangement layout was made for the 140K unit, based onthe design approach used in the small Air Force turboalternator frame else.
"The turboalternator was designed to incorporate elements that couldreadily be developed to provide suitable performance and that could be adapt-able to quantity production. A layout of this turboalternator is shown inFigure 13.
The turboalternator is mounted on gas-lubricated journal bearings.Three hardened pads at each journal bearing support the 0. 25-inch-diametershaft with an operating gas-film thickness on the order of 200 mieroinches,These journal bearings are of the self-acting tilting-pad type and are capa-ble of stable operation throughout the operating range and at any attitude.
Two inward-pumping, self-acting, spiral-grooved, thrust bearings posi-tion the shaft axially. Like the journal bearings, the thrust bearings are gas-lubricated and typically operate with a 500-microinch gas-film thickness.The entire bearing system is self-aligning because the thrust bearings aregimbal-mounted and the journal tilting pads are Individually self-aligning.Satisfactory operation of the complete bearing system can therefore be some-what independent of the accuracy with which adjacent parts are manufactured,
The radial-inflov, impulse turbine wheel is 0. 50 inch in diameter. Theradial-inflow wheel is convenient for close blade-tip axial clearances tominimize leakage. The turbine nozzle will be designed for partial admission.
The turbine energy is absorbed by a two-pole permanent-magnet alter-nator. This compact alternator is a very practical device for extractingenergy at cryogenic tamperatures when that energy will be dissipated at aremote location. ThL two-pole magnet operates within the stator, which itwound three-phase in a core of low-loos iron laminations.
A vacuum-type enclosure is welded for the final assembly. Proximityprobes are installed to monitor the position of the rotor and gas-bearingelements.
100
TURBINE WHEEL STRESS AND DEFLECTION ANALYSIS
140K UNIT
The 140 K unit turbine wheel is shown in Figure 42, The wheel is made of6061-TS aluminum and fits onto a stainless-steel shaft. The extreme operat-ing conditions for the wheel are:
Condition Temperature Speed (rpm)
Cold 140K 90,000
Warm 321i K 250,000
L . ... .. . o. ,on-in. D~tL,
0, ', 0:3-in, IdaI
Figure 42., First-Stage Wheel
The coefficient of thermal expansion of aluminum Is greater than that ofsteel. The wheel -shaft assembly is made at room temperature; therefore,an the warm condition is approached, the wheel expands more than the shaftand becomes loose nin the shaft unless the wheel shaft assembly is made witha slight interference fit at room temperature.
The stresses caused by the interference fit are increased as the tempera-ture is lowered, The thermal stresses are therefore very much higher forthe cold condition than for the warm condition. Because the wheel is small,the centrifugal stresses are very small, At 250, 000 rpm, the peak centrifugalstresses are only about one-third of the thermal stresses at 140K, From thestandpoint of stress, the cold condition is therefore the critical condition. Theprocess followed in the analysis of this wheel was to determine the room-tern-perature interference required to maintain wheel-to-shaft contact in the warm
101
1*
UNUAL* EUCTUIC
condition and then to perform a BtreSB analysis for the cold condition withthis interference fit.
The interference fit required is 0. 05 mu on the radius or 0. 1. mu on thediameter, This fit is sufficient to offset the differential thermal expansionand the sm�ill centrifuga.l force, tending to sepaz'ute the wheel and shaft in a.waz'rn condition. The wheel stress warn analyzed for the cold condition, usingthis value 01 interference fit. A plot �f the effective stress contours is shownin Figure 4�i. Note that the stress levels are low everywhere except in regionA, and even in region A, the stress level is acceptable. Actually there is no
- F contact with the shaft In region A, because the shaft is slightly relieved to aidin pilotirtg the wheel for assembly.
2 I 122 IIUS
1212 II U
III
12 *12
Hegtot� A
Jo I,,
I'igur� 43. Cold-Condition �ffe�tive Stress Contours
Bocauso centrifugal effocts are small for this wheel, there is almost nobendingU so the tip deflections are very low, Deflections for the wheel inthe cold condition are shown in t'igure 44. Deflections for the warm condi-tion were nut calculated but arc expected to also be small, because the con-trifugal s�rui�3ses are small.
l3ecausv the 0. 1-percent creep-stress limit �t 70F for 30, 000 hours is40, 000 psi, the wheel design is considered acceptable. Creep data F�12t cryo-
OM IRAL@ ILI CYR IC
N'
Figure 44. Deflections for Wheel in Cold Condition
The actual design range on this aluminum-wheel interference fit is 0. 2..to 0. 4-mil diarmetral interference. to assure that the wheel is attached at alltimes and can resist the operating torque.
550K UNIT
The 55'K wheel design is 0. 78 inch in diameter. The extreme operatingconditions for the wheel are expected to be:
Condition rernperature Speed (rpm)
Cold 85OK 120, 000
Warm 32211K 250,000
The wheel may be made of either 6J061-T6 aluminum or 6 Al-4V titanium,a trade-off between:
0 Ease of making tJhe aluminum wheel, but possibly limiting the miaxi-mumr speed below the target of 280, 000 rpm.
* tUsing titanium, but ensuring that the target upper ~ioped limit of2 50, 000 rpm can he obta ined,
170 0K UTN[T
The 170oIK wheel design in 1. 0 inuch or 0. (I25 ink~h in diumuter. The ex-
trerne operating conditions expec'ted for the two different wheels are:
103
I INIRAL ILIOTl 10
SpeedCondition Temperature 1. 0-inch Wheel 0. 625-inch Wheel
Cold 170K 150, 000 rpzn 230, 000 rpm
Warm 322 0K 250, 000 rpm 250, 000 rpm
From prior results, such as the work completed for the U. S. Air Force,it is expected that: the 1. 0-inch-diameter wheel must be made of 6 A1-4Vntitanium to achieve the target-warm speed of 250, 000 rpm, and the 0. 625-inch-diameter wheel probably car. be made of aluminum, with no problem.
Design considerations for a titanium wheel are somewhat different fromthose for an aluminum wheel. Taie coefficient of thermal expansion is lessthan that of stainless steel. The wheel-shaft assembly is made at room tem-perature; therefore, ai' the cold condition in approached, the wheel contractsless than the shaft nnd becomes loose on the shaft unless the wheel-shaft as-sembly is made with tn interference fit at room temperature. The stresses"caused by the interference fit are relieved as the temperature is lowerv-d andare increased as the temperature is raised. The thermal stresses are there-fore very much higher for the warm condition than for the cold condition. Be-cause the warm condition also has a higher speed, it is the critical conditionfrom the standpoint of stress.
The proccss to be followed in the analysis is to determine the room-tem-"peraturo interference fit required to maintain wheel-to-shaft contact in thecold condition and then to perform a stress analysis for the warm condition,"with this interference fit.
The design of a 1. 1-inch-diameter turbine wheel is described in the pro-gress report to Wright-Patterson Air Force Base (Ref. 4, Vol. I, Section 5).
104
aNINRAL ILICTRIC
Appendix 1i
GAS.SUARPNG ANALYSIS AND DSION
,EAINO WOSIG ,um'mNTS
ar. The overall design requirements for the journal and thrust bearings• are:
Lubricant -- Heliunm gas from cycle working fluid
Contract design life goal -- 2500 hours
Ultimate design life goal -- 10, 000 hours and more
Shock and vibration loads -- None while operating; normal handlingwhile not operating
Acceleration load -- None while operating, but design for 2. 0 g inany direction while operating
Design temperature -- 14'K (25. 2*R)
Maximum operating temperature -- 125PF (585 0 R)
Starting -- Many start-stop cycles
Orientation -- Both vertical and horizontal
Additional requirements, as a cutimequence of the operating environment,include a constant bearing ambient pressure 1, 19 x 10" newtons per squaremeter (17.22 psla) and a maximum speed of 200, 000 rpm. This maximumspeed will be experienced only at the maximum operating temperature at thestart of system cooldown. After cooldown has started, the speed will be grad-ually decreased until the design speed of 90, 000 rpm Is reached at the designtemperature of 14"K.
8EARING TYPE SMELCONS4Only self-acting gas bearings were considered for this design,
Externally pressurized bearings were not seriously considered for thefollowing reasons:
0 Refrigerator cycle efficiency requires a low anibient pressure,near atmospheric, In the rotor housings.
0 Refrigerator cycle efficiency would he lowered bwcnuse a portionof the cycle gas would have to be diverted through the bearings.
0 Ducting the bearing exhaust gas involves u mechanical und thermnlheat-leak complication that is considered impractical.
105
GIN I AL I LIMRI C
r he h earings must be isolated from thv rotor (,avity by noncon-ta( ting seals, rho design of the 9va's ':ould N, uis complicatedas the design of the bearing itself,
Tilting-pad journal bearings (Figure'. 45 and 46) were selected for thejournal@ be'oause of:
* Confidence in ultimate uccoss
0 Prior- manufacturing and .Iryogontv tvst E'xporit-ncc'
* Brond st~bility range
0 Inhoreat self-alignment
a Reasonable tolerance to dirt ingestion rand thermal distortion
NIo Q11,011.
Figure 45. Pivoted-Pad Journtial He'aring (Scilerutic D~iagram)~
!SIUNSAL. ILICTRIC
F igurye 483. C'ryoge'nic T urbon 1ter nutor TilItin g- tId Joturmi tuiv:r,ving
A\ doUbli-titeting hydrodynamnic 111rukt hunrirng with 11 gimbti11 Nsyi.Itni1 11I)11~was mtleettd he~cnuse (.f.
0 Ckwfiidence iN the Ul11UIL, Sin UCC41916
* Iv'ior n~inutneturtl ig 111d cr'yogenic tes t vxprePhit11c
* Isuilthiliity for rimy tittl tudi pl um g-lIoniidng
a Sui tabl~i ity for ctmiplct tti' t Mtifnimupiint.
A H piral -grooved in w~ rc pump ing g~o iietry i(shiiw n In I-I gI ivoH 47 tind 48)
could be tibt ineci Nthu opertited with n gm'inter Itnid viapmt I i.N wiu lcts iino, *loanB tha n '~nny iither c onfign ratitfi
107
GINIRA Le IELIOTU10
it I ~ I III I
'I.
it. Ill* 1 11, 1 II In
lit-i;
Ft gut.v 48B. Spi ral -Iroovir 'Ihl ru st I In~a vin)ý (Sel'oninti I I )tagrarn
OINIRAL@ ILECTRIC
JOURNAL BEARING DESIGN
V • L PROC EDUREt E
The procedure for journal bearing design .consists of the following steps:
0 Set up criteria for selection,
0 Select bearing parameters inid purformntnce characteristics forpreliminary design.
9 Dutermire first bending critical spevd of rotating assei bly,
0 Determine the stability nmid response of the rotor/journal bearingcombination.
Many of the selection criteria tre based upon engineering judgment,based upon past experience. F'or example, the principal Journal bearingdesign goal is to maintain a fluid film separation of the bearing surfacesin the cryogenic environment. All that is required is a nonvroro minimumfilm thickness, but there arc .wo other' film thickness(,s that are usuallyconsidered. One is the pivot film thickness choson as a dhsign goal inadvance of thermal distortion data. The sectmd film thicknes is chosenas the absolute minimum acceptable film thickness, whiich should take intoaccount basic equation accuracy, numerical solution :acuracy, and antici-pated manufacturing tolerances. Simillnr consido rations apply to criticalspeed@ and other aspects of the design.
The criteria used in desigtninig the jouIr nal hiaritigs itre:
9 Pivot film thickness it '2.-g btvady-stnto load without beturlng surfaucdistortion will be 100 miicroinches.
0 Absolute mitaitnuni fili thiL'kne, ss will be 10 ilL i'(Ji- 'ht:w e ,
a Power loss must be low.
a Shoe pitch, roll, amd radir•il translntiou:•i n1t1MLI 1 tI' qf1 tI(tieH, withundistorted bearing turfaves, must be 25 prcont above or 5 porcontbelow the oporating sped extremes.
* First oending critical speed must be 25 percnt above .ac operatingspeed rango.
* Whirl threshold speed intst e nbove the, operating s•p•e(d riinge.
* Naxinum nondiniuostont|l pivot filn thickties• (1I11) will bh 0, 75,for pod stability.
Mlininmum nondiniensi,,nal pivot film thiclmvnss will he 0. 20, to limitbearing frictirn,,.
*iVluximumi pivot point stress (hertz) will he 100, 000 psi.
u ~I O h
SOENEHAL* ULICTRIC
'The jutmrnaI hea rings w(erc decs•ignged with a stI(aJct or cor1putor program,JSE I,('TI (oi•r. 4, App. VI), that uo' tailned the f'(ollowing:
( Coefficient of nondimensional polynomiAls for single-pad load, powerloss, radial stiffness, and pitch axis stiffness versus bearing numer(A) at constant nondimunslional pivot film thicknesses (11P) of 0. 20,0. 25, 0. 30, 0. 40, 0. 50, 0. 60, and 0. 75.
SLo.,gic to compute internally coefficients of nondiniensional poly-noniials for single tpid load, power loss, radinl stiffness, andpitch axis stiffniess and load versus TIP at constant A.
* (oeffirients of nondimensioiial polyno•laJls for pad inertia versusshaft nmass for constant I', at the threshold of translatory whirlinstability.
a A routine based un betua theory foi- computing two rigid-bodynatural frequencies Atd the first bending critical speed of asystei cuonsisting of fo,,r hbars, three mnasses, and two bearings.
0 A routine for computing the increase in journal diameter due to
centrifugal force.
0 kogic for testing film .hickieHses rtelative to input criteria.
0 Iiugic for testing tile proxinility of the following frequencies relativeto the end points of an oprating speed raiige:
- Shaft rigid-body trarnslations and rotations
- Shoe radial tranOlation
- Shoe pitch axis rotation
T'rans latory self-,xcited wnli r1
0 I ogiv for varying the machined-in (lea rance, preload, and preloadspring -liffness if frequenY or film thi'knteivs tests are not passed,
@ 1,,gir for cotnputing perf'r niance chara•ctristirs if the niachined-in clearance, preload, antd preload spring stiffness are specified.
* Ikogiic to detvicnine the pivot ball radius so the hertz stress will be100, 000 psi.
The original pad data cont; tied in the selet tot' program were producedby a nunierical solution of the Iratsient Reynolds equation. Since the resultsare based onl a disturbance rrom equilibrium, both steady-state and stabilitydata were obtained simultaneously. A)! pod de•'ign data are hased on a pad
110
-,m I
r
ENERAL* ELECTRIC
Lrc length uf 100 degrees, with a pivot location of 65 percunt and a length-to-diameter ratio of one.
PERF ORMANCE
The .ISEi.,LT journal bearing selecto. program was used to design thetlting,-pad Journal bearings for the turboalternator. InitiAlly, a target max-imum speed of 250, 000 rpm for roomn-temperature operation was used asinput. With this high maximum speed, it was found that maxinium rigid-bodyeritical spretd for the 2-g operation was too close to the cold temperature do-sign o•perating speed of 90, 000 rpm. The reason for this is that the high,250, 000-.rpni speed required it stiffer gas bearing at the lower, 90, 000. rpmdesign speed, The variations of the maximum rigid-body critical speeds ureas shown in Table 15 for the three levels of g-loading of interest. Then,the stiffness of the shaft was reduced from 7, 000 pound-inches". With onlythis change, the maximum rigid-body critical speed wis reduced by onlyabout 1000 rpm.
Table 15
DESIGNS WIT-I A MAXIMUM OPERATING SP.EED OF 250, 000 RPM.I 7000 lb-to. 2
(;Lodindtn 0.0 110 2.0
1)c'M sr q it l wirln g n u mn b t r (A ) 0 .2 5 0 1 ." 10 ~ - . 51 5Maximum rigid-budy 751200 1 05,400 85, 00 2 " 96i, 2UO 127, HOO
crl ct l1 p,,d (rpm) ..'I -2280J
As shown on paige 144 of tleferencc 4, the spring rate of typical finishedturboalternator shafts of the subject design are 50, 000 and 38, 0001b.-in.Using the simpLe beam formula likened to conditions of the test, the shaft stiff-hess5 is: ElSL.
EI .48
whIere E -, Yuungs modulus (psi)
I Second nmomnet of area (inch 4 )
S Shaft spring rate (lb/In.)
1, Span between supports (inch
The correigponding shaft stiffness values, El, are 7410 1nd 5640 pound-inches 2 . lin'ce, the 6000 value appeared to be a reasonable a'verngv and wastisf.d for tilt subsequent design runs.
It was thvrerore decided to lower the maximum operating speed to 200,000rpm. Trh, 2-gr operating rangeI of hearing numbers wits first investigated, :1nrd
111
UNIlALO*, LICTRIC
the lowest bearing lose was determined. The results are shown in Table 16,where a bearing number of 0. 350 provides the loweat journal bearing powerlose of 0. 1542 watt. The maximum shaft rigid-body critical speed is com-fortably below the design-point of 00, 000 rpm.
Table 16
DESIGN SPEED PERFORMANCE AT 2. 0 0
B'xlearing MIAIXmLI Spring- MountedNumber .,l1gR(, S1nl't Ri gid-13dy Fixed III, ot-Prn d Pivot-Pnd Flhm, N mb r l,:'rillg IL ilor I"1[lM ThieknessLani A bda Losrýtw CH' tivs I S•L.L-d 'rhicktirms(nd) , (rpnii) ( ")
S0,,40J0 I, ' ) 8a . 0,.4. 16' 241
0, Iluo 0. 15 42 81, 5'!I 16H 245
0, W000 ,). 1151fir 80, 0,01 167
0, 2500 0. 16.97 80, ,i, 184 2!16
Complete design runs were then made for the three g-loadings of interest:
Run Number 0- Loading
4701001 2.0
4791002 1.0
4791003 0. 0
A summnary of these performance runs is given in Table 17. The threeruns are printed completely in Figures 49, 50, and 51.
Pertinent performance parameters from these three computer runs aregiven in Figures 52 and 53 as a function of the g-londing. Figure 53 shows thevariations of bearing power loss and pivot film thivknres for the two extremesin operating conditions of speed and temperature, F'!iure 52 shows that thehighest calculated rigid-body critical speed is safely under the design speed,and the spring-pad radial translational critical speed is well above the max-irnum operating speed. The pcd-pitch (,ritical frequencies are shown opera-ting reasonably above the respective operating mnd design speeds.
Next, tn analysis was conducted to evaluate the effects of the spring pad,spring rate, and film thicknesses. Figure 54 shows the results of computerruns of constant values of nondimensional spring-pad film thickness for vary-ing values of spring-pad spring rate. This nondimensional film thickness isthe ratio of the actual film thickness to the operating machined-in radialclearance. Also shown are the 1-gdcsign values. The results indicate that
112
kNINALIOILIOTNIC
Table 17
UAS BEARING PEHIVORMANCE SUMMAIHY
AiIis: 1.07 i 17. '
%%illmiiI
-11. 000 1.11Will
1-111 t joi Ilia wt ts,
Liv
flit,. 'i,, klil-KIl( ,l J ill
.1 I I1
ýA ~ ~ ~ 1 1- Mtli
J3[LCTU 14e42CST 1197
*****$AS-bLUBRICATID JOURNAL BEARINGS*S****
aeDESI SN SPICIFICATIGNS** .
DESIGN RUN NUMBER 4791001
DESIGN SPIED (RPM) 90000.00TEMP. AT DESIGN SPD (R) 25.10AMBICNT PRESS#DESIGN SPD(P&IA) 17.31VISCOSITY&DE&ION(LU*StCIIN**3)O.9 1003.09MAX. SPIED (RPM) 100000.0TEMP. AT MAX SPD (R) 565900AMBIENT PRESS&MAX SPD CPSIA) 170311VISOOSITY&MAX CL§*SECofIN**1)093IDOt-0S.JOURNAL DIA* (IN) 00961ROTOR WEIGHT (LU) 0*0596LEFT OVERHUNG WT. (LR) 0.0074WT* 1ETWIEN BRGS. (LB) 0.0396RIGHT OVERHUNG WI. (LB) 0.0051LEFT WT* TO LEFT ERG. (IN) 0.1450LEFT @NO* TI CS (IN) 1.34000CO TI RIGHT ORO* (IN) 0*5700RT BRG TO RT OVERHUNG WT (IN) 091650S LOADING 1.0000JOURNAL WALL T4ICK14ESS (IN) 0s1309YOUNO.INCRTIA. (LICINCCI) 006000C+04PERCENT OP MAD FOR PAD THICK 0.1530PAD LENGTH (MN 00.1610POISSINS RATIO (ND) 0.9600YOUNOS MODULUS (PSI) 0.30003.08WT DIN OP Je MAT. CLBIIN**3) 0.11500WT DIN OP PAD MAT CLBV*IN**3) 0626000ANGLE BETWEEN PIVOTS (DIG) 110000000MINIMUM PIVOT rILM TK. CI10 06000100
TFIX* T
SPRING STIFFNESS CL3IfN) 0.95561.03MACHINED IN CLEARANCE (IN 06000358STARTINO PIVOT FILM THICKoCND) 0.643000BALL RADIUS (IN) 00014000SOCKET RADIUS (IN) 0.014300
ONLY ONE PRELOAD IS BEING OONSIDERID
Figure 49. I)euign for 2.O0-g Gas-LubrIcated -Journal Bearings (Sheet 1 of 4)
114
~INIIAL@ ELECTS IC
LAMBDA LOPPLAMNDA. DES, SPDs 09350LAMBDA& MAX, SPD* 96694CLECARANCE. MeS SPO.s (IN) 0.00030193CLEARANCE. MAX. SPD* (IN) 0900037117WHIRL SPEED (RPM)&DE5ISN 1041163.33
WHIRL SPIED (RPM)&MAAIMUM 696350.04
THIS SECTION INCREASES MLUTTER rREO
ENTERING PRELOAD LOOP
$****GAS LUBRICATED JOURNAL 8EARINOS.****
*OCOINDITISNS AT DESIGN SPEED**
LAMBDA 00D 0.3501AMBIENT PRESSURE (PSIA) 17.33CLEARANCE (IN) 0.000381JOURNAL DIAMETER (IN4) 00216006ORO* TRANSVERSE STIrr. (LBIIN) 3926.06BAis VERTICAL STEPre (LOIfN) 979.48IRS. POWER LOSS (WATT$) 0.3543
SHIS WITH PINED PIVOTS
LOAD (LS) 0.1056PIVOT rILM THICKNESS (IN) 06000166PITCH STIFFNESS (IN"L9,*RAD) 3.1333PITCH CRITICAL TRIO. (RPM) 9130567.7
SHOES WITH SPRING MTD* PIVOTS
LOAD CLS) 0.05399PIVOT FPILM THICKNESS (IN) 0.000145PITCH STIFFNESS (IN-LBIRAD) 0.5110PITCH CRITICAL rRIO (RPM) 110476.5TRANS* CRITICAL rRIO (RPM) 360516.4STIrF Or PRELOAD SPRING (LB/IN) 955.6PIVOT SOCKET RADIUS (IN) 0.0143PIVOT BALL RADIUS (IN) 0.00140
Figure 49. Design for 2. 0-g Gas -Lubricatedt Journal Uetarlngs (Shoot 2 of 4)
IININAL IiDThIC A
*$GENERAL CONDI TIINS*s
WHIRL SPEED LIMIT (RPM) 6961t00,FIRST SENDING ORIT. SPD, (RPM) 359873.,SHAFT RIGID BODY GRIT SPD (RPM) 013S0o1SHAFT RIGID BODY GRIT SPO (RPM) 54231s4SHAFT RIGID BODY CRIT SPD (RPM) 47410M6SHAFT RIGID BODY CRIT SPD (RPM) 34075.0SHOE PITCH INERTIA (IN-LIo.ECI) Oo$34@E-OBWEIGHT or SHOE (LI) 0000359THICKNESS OF SHOE (IN) 0.0800MACHINEDoIN CLEARANCE (IN) 0000381START-UP CLAMPING FORCE (IN) 0.START-UP CLEAR ON TOP SH4E (IN) 0.00053
*****GAS LUBRICATED JOURNAL *EARINGS*****
**C$NDITIONS AT 4AXIMUM SPEED*S
LAMBDA (ND) 3 *694AMBIENT PRESSURE (PSIA) 17.10CLEARANCE (IN) 0.000378JOURNAL DIAMETER (IN) 0.141008BEARING X-STIFFNESS (LBIPN) 3173*931"BEARING Y-STIFrFNISs (L/IN) 159.3518$EARING POWER LOSS (WATTS) 1.60
SHOES WITH FIXED PIVOTS
LOAD (LB) 009954PIVOT FILM THICKNESS (IN) 0000131PITCH STIrFNESS (IN LBIRAD) 5.4367PITCH CRITICAL (RPM) 304696.1
SHOES WITH SPRING MTDV PIVOTS
LOAD (LB) 0o1018PIVOT FILM THICKNESS (IN) OOCOT70PITCH STIFFNESS (IN LBRAD) 3.4173PITCH CRITICAL (RPM) 055316.0TRANSLATION CRITICAL (RPM) 50276800
Fifure 40. I)i•ign for 2.O-g GaR-Lubricated Joi rnal Bearings (Sheet 3 of 4)
116
9ONINAL ILIOTIIC
LIST
I IMPTS 1414lEST 12,019,17i
t00 II SO 479100100O0 90000. 15.3 I9o.33 9.i1-tO 10O0000. 5B5.
300 i7.13 3.1*t-0 4461 .0074 .0396 40057400 .945 1.38 *t570 ,165 005965 800500 e1305 .31 ,gO e26 30.16 6000.600 *153 .161 *0001700 0 I$0g0 T
)00 ,0003811 99556 . 643 .0140 00143
Figure 49. Design for 2. 0- Gam-Lubricated Journal Bearings (Sheet 4 of 4)
JSLOTU O6IRI2 T 1190,0172
*e***GAS LUBRICATED JIURNAL .,EAANO3e*e**
**DESIGN SPECIFICATIONS**
DESI3N RUN NUMACK 4791003
DESIGN SPEED CtPM) 9000000TEMP. A1 DESIGN SPO (4) 250,0AMBIENT PRESS#DEGIGN SPD(PSIA) 17.31VISCOIITY.DE SIONELBSOECC INS*91009100E-09MAX. SPEED (RPM) O000000TEMP, AT MAX SPD (R) 51,000AMBIENT PRESS#MAX SPD (PSIA) 17.21VISCiSITYPMAX (LB$SEC/I'Ns9) 0.3100E-0SJOURNAL DIA. (IN) 0.961IOTSN WOIGNT (LB) 0.0056LEFT IVERMUNG WTe (LB) 0,0074WT. BETWEEN BROS* C(B) 0.0396MIGHT OVERHUNG WTo (CL) 0.0057
Figure 50. Design for 1. O-g Gam -Lubricated Journal Berarings (Sheet 1 of 4)
117
3 SNESAL ILIOTRIC
LCIFT WTs TO LEFT RAG* (IN) 0.2450LIFT IRG. TO CO (1IN) 138300CC TO RIGHIT BORG (114 00$700A~I T BAG TO At OVIRIIUNS WT (IN$ 0.16500 LOADING 3.00000JOURNAL WALL THICXNtSS, (IN) 0.1305Y@UN0*IN9RTIA& (LI*IN**1) 0660001#04PERCENT OF RAA rOA PAO TMICX 0.3530-PAD LI-40TH (IN) 0.0610P01550145 RATIO (NO) 0.1600YO'JN03 MODULUS (PSI) 0*30001*0IWT DEN or j# MAT* (LBE'IN**3) 0911000WT DIN OF PAD MAT (L9##IN**3) 0412000ANGLE BIT WEEN PIVOTS (DIG) 120.00000MINIMUM PIVOT FILM TKs (IN) 0.000100
TFIXe
SPHING STIFFN~SS (Ls.I'I) 0695569+03MACMINED IN CLEARANCEt (IN) 0.000352STARTING PIVOT FILM THICM.(ND) 0.590000BALL AADILIS (IN) 0.034000SOCKir RADIUS (IN) 0.014300
ONLY 0141 PRELOAD IS BECING C9NSIDERID
LAMBDA LOOP
LAMBDA# DES. SPD. 0.350LAMBDA* MAX. SPO. 3.694CLEARANCE. DES. SPDs (IN) 0000034123OLIAitANCI. MAX. MPe (114) 0.00037817WHIRL SPEED (ItPM)sDISIGN 1041163oI1WHIML SPEED (APN)oMAXIMUM 6961800.04I
THIS SECTION INCRESEs FLUTTER r410
ENTERING PARIAD LOOP
***$*GAS LUBRICATED JOURNAL BEARINGS*****
**CONDITIONS, AT DESIGN4 SPEID**
LAMBDA (ND) 0.3501AMBIENT7 PRES3URE (PSIA) 17.31CLEARANCE (IN) 0.000303
Figure 50. 1eomign fo~r 1. 0-g (.ras-Lubr~cat(-d Journal .13naring~s (Sheet 2 of 4)
IIAI(AL ILI0T.i0
PJOUNAl DIAMETER H 10ESS 0IN) OBPIm TCHANSVER STIFFNSS (iLR N) 1563.65PITCH CRTICAL STIFF* (000 910076.0
SHG9S WITH PIXND PIVOTSo-a
LOAD (WL) 0067:PIVOT FILM THICKNESS (IN) 000O01lPITCH STIiiNESS CIN-MLItRAO) 12149-PITCH CRITICAL FRICE (RPM0) 1100750
TRAS.RT WITH CL PRING ( TPM PIVOT3
LSIAD PLA PL9)55.6PIVOT CILM THICSKNES (IN) 0.00OOISPITCH STL rrNDUS (IN-L)RA) I0.1459PITCSH CRINTICAL CRT P C RPM) [email protected] C ODY CRITE CPRPM) 7398321.9SlTIFFe fir PAHII.AD iP/lIN6,CL9,,IIIM 95546
PIVOT SIGCID RADIUS C IN) 0R0143• PIVOT NAtLL RIADIUSI €IN) 00140
**GENERAl. CONDI TIONS**
WHIRL. SPECED LIMIT (RPM) 69)61804O+* FIRST SENDING CGRIT* SIPD9 (RPM) 3591699o6'.MIArT RIGID BODY CGRIT SIPD CAPM) 73935*4
SHMAFT 14101D BODY CRlT 3PD (RPM$ 56399*3
SHArT RIGID BODY CRIT SPD (RPM) 43091.0SHAFT RIGID BODY CHIT 0PC (RPM) 33018s4SHOE PITCH INERTIA (IN-LBSE(C) 0.5340€-081IINGHT IF SHOE (LO) 00000359
THIC:KNESSI OF IIHiC (IN) (:01000MAI:HINICD-I N C:LEARANCE ( IN) 09 0003821START-UP C:LAMPING FORCE(IN€l) 0,START-UP C:LEAR IN TIP SHOE (IN) 0100053
*****OAS LUBRICATED JOURNAL OEARINGS*o***
**CONDITIONS AT MAXIMUM $PEttD*
LAMBDA (ND) A's 694AMBICNT PRESSURt (PSIA) 17,119CLEARANCE (IN) 0.000378
Figure 50. I)emign for 1. 0-g Gas-Lubricated .luurmial l3earlngs (Shoet 3 of 4)
11.. .
IINIIAL ILIOTRIC C
JOLUMNAL DIAMETER (IN) 0,R61006$EARINO X-STIP'PNISS (LS9,IN) 2699.0o6SCEAING YSTIFPrNESS (LS'IN) tST71789SEARING PoWER LOiS (IATTl) 1.77
SHMfS WITH FIXED PIVOTS
LOAD (CL) 0 * 3407PIVlT FILM THICKNESS (IN) 0000i42PITCH STIFFNESSIN L( tAD) 409269PITCH CRITICAL (RPM) 190047,1
SHIES WITH SPRING MTD, PIVOTS
LOAD (LB) 001144PIVOT FILM THICKNESS (IN) 0.000016PITCH STIFFNESS (IN LS'RAD) 4,0771PITCH CRITICAL (RPM) 163660.0TRANSLATION CRITICAL (RPM) 50/19.6
LIST
I2NPTS O08135CST 1t,020.079,
100 I150 4791009100 90000a l5.1 17. i 9,11-t0 *00000, 565.300 17.16 3141-10 .161 .0074 e0396 *0057400 4845 1483 o570 .165 .O0165 1.000 w1305 s26 .16 *.6 30416 6000.
600 ,153 .l61 .0001700 0 1100 T900 e000360 95Si6 e590 90140 o0143
Figure 50. Design for 1. 0-g Gas-Lubricated Journal Bearin•es (Sheet 4 of 4)
120
JSELCTU 170OOE3T 12,0k9 /72
*****GAS LUBRICATED JIURNAL BEARINGS*****
OMDSI GN spEciriOATIONSs*
If
DESIGN RUN WPMSE 4791003
DESIGN SPEED (RPM$ 90000.0TEMP. AT DESIGN SPD (A) 95.30
i AMBIEINT PftISS#DtSIGN SPDCP$IA) 17.13
MAX* SPEED (RPM) 300000.0AtMPIEAT MAXESS.M (R) (PSA) 001TEMP. AT MAXSMA SPD CR) IA 517.0VISC431TY # AX (LB*SEO/fIN**2)O.3I0OE.OS
IJOURNAL DIA. (IN)041NOTIR WEIGHT (LB) 0.0526WFT OVERHUNG WT, (LB) 0.0074WTe SETEE IAGS. CLI) 0.0396RIGHT OVERHUNG We. (LI) 0.0057WET WT* TI LIFT $RS* (IN) 0.214s0
LEFT SRI. To CG (IN) 1*380000 TI RIGHT ERG@ (IN) 0.5700AT MG4 ?01 AT OVERHUNG WT (IN) 091650IU .LOADING 0.JOURNAL WALL THICKNESS (IN) 0*1305YOUN@*INERTIA* (LBIOIN*011) 006000t#04PERCENT or NAD Or4 PAD THICK 0*1530PAD LENGTH (IN 0*2610PIISSONS RATIO (NO) 062600YOUNG$ MIDULUS (PSI) 0.30001408WT DIN Or Js MAT. CLB#'IN**3) 0.16800W? DIN or PAD MAT CLIVIN**3) 0.9600ANGL9 BETWEEN PIVOTS (DIG) 120.0000MINIMUM PIVST FILM TK& (IN) 0.000100
TFixe T
SPRING STIFFNESS (LOOPIN) 0.95561403MACHINED IN CLEARANCE (IN) 0.000389STA,'-TING PIVOT FILM THICX*(ND) 0.530000NALL RADIUS (IN) 0.014000SOCKET RADIUS (IN) 00014300
ONLY ONE PRLIAD IS BEING CONSIDERED
F'igure 51. D~esign for 0. 0-g Gas-L~abricatcd .Jou~rnal Beanrings (Slit'tt 1 of 3)
121
$IN EHAL@ ILICTIIC
LAMBDA LOIP
LAMBDA* DES* SPD. 0.350LAMBDA* MAX* SPD* 20694CLEARANCZ, DES. SPO. (IN) 0,00038113CLEARAN~4CE# MAX. SPDo (IN) 0.00037117WHIRL SPEED CRPM).DISIGN 1041163,1l
WHIRL SPEED (RPM)eNAXIMUM 696150.04
THIS SECTION INCREASES IPLUTTER rREO
ENTERING PRELIAD LISP
*****GAS LUBRICATED JOURNAL BEARINGS*****
**CONDITIONS AT DESIGN SPEED**
LAMBDA (NO) 0D,03S0AMBIENT PRESSURE (PSIA) 17.38CLEARANCE (IN) 0,000351JOURNAL DIAMETER (IN) 0S 1001eRG. TRANSVERSE STIrre (L.BIN) 1313077
BRO, VERTICAL STIMPT eLIlIN) 891.76BRO, POWER LOSS (WATTS) 01442
SHIES t WITH rIrXED PIVOTS
LIOAD (L) 0.0743PIVOT rILM THICKNESS (IN) 00o00090PITCH STIriNESS (IN-0L/RAD) 167,140PITCH CRITICAL RIrAg. (RPM) 174044.2
SHOES WITH SPRING MTD. PIVOTS
LOAD CLS) 0,0743PIVOT rILM THICKNESS (IN) 0.000030PITCH STIFFNESS (IN-wLSVRAD) 107491PITCH CRITICAL MIEG (RPM) 17961683TRANS. CRITICAL PrREQ (RPM) 431994.4STIrr Or PRELOAD SPRING CLB'IN) 95.96PIVOT SICKET RADIUS (IN) 0,L143PIVOT BALL RADIUS (IN) 0,0140
SeGINERAL CINDI TI ONSO*
WHIRL SPEED LIMIT (RPM) 696160.0iIRST BENDING CRIT. SPO. (RPM) 3396317.
',HArT RIGID BODY CRIT SPD (RPM) 67372.7
Figure B1, 1 ).Desgn for 0. 0-g (Cas-Lubricanted ,1ournal TBvariiwns (Shwtvt 2 of 3)
2I .
010INRAL, ILICTIIO
SHAFT RIGID BODY CHIT SPD (RPM) 5b595.1SHAFT RIGID BODY ORIT SPD (RPM) 39341.5SHAFT RIGID BODY CRIT SP0 (RPM) 38569.5SHOE PITCH INERTIA (INeLS-.SEC) 0.53401-08WEIGHT OF SHOE (LI) 0,000359THICKNESS or SHOE (I0 00000MACHINED-IN CLKARANOF (IN) 0.000381START-UP CLAMPING FORCE (IN) 0.STARTeUP CLEAR ON TOP SHOE (IN) 0.00053
*****OAS LUBRICATED JOURNAL BEARINGS*****
**CONDITIONS AT MAXIMUM SPEED*.
LAMBDA (ND) 20694AMBIENT PRESSURE (PSIA) 1742oCLEARANCE (IN) 0.000378JOURNAL DIAMETER (INI 0,361008BEARING X-STIFFNESS (LBIIN) 9696.998BEARING Y-$TIFNESS (LB1IN) I520.435REARING POWER LOSS (WATTS) 1.75
SH1ES WITH FIxED PIVOTS
LOAD (L) OP' 70PIVOT FILM THICKNESS (IN) 0000051!ITCH STIFFNESS (IN LBSRAD) 4.5259PITCH CRITICAL (RPM) 277993.6
SHOES WITH SPRING MTDV PIVOTS
LOAD (LB) 0.*270PIVOT FILM THICKNESS CIN) 0.000253PITCH STIFFNESS (IN LBOoRAD) 4,4550PITCH CRITICAL (RPM) 275819s6TRANSLATION CRITICAL (RPM) 51761109
LIST
IRINPTS 171lIEST Ilk19.72
100 1150 4791003200 90000. S*.t 17*22 9.E1-tO 200000a 585s300 17.11 31.1-10 ,061 .0074 .0396 .0057400 .145 1*38 .570 .165 ,05065 00500 @1305 ,88 ,18 .26 30.16 6000.600 s153 .361 .0001700 0 1800 T900 .00038B 955.6 .530 .0140 s0143
Flgurn 51. * )nsiagti r¢or 0. 0-• (hCos -tm, lir to Jot urilLil I(• Ii'I1kYs (shlrt, 3 of :.J)
I2 M
GEN ERALO& ELECTRIC
90, 000 rpm,--- ýIBu rlng Nii. 0. 3502200, 000 rprri - -13varing No. 2. 694
90
U)
50
0
5 00
4~ 00 __
4A I)V
2002
!-`OUNIAALOILICTIIC
90, 000 rpm H-Iearinig No. -0. 3502200, 000 rpm R earing No. -2.8694
280 Pivots
0 also Spring
_______Spring
P4 Lq Fixed
200
1.1.0
0 .02.0 Ldu-22871G-)nmdinlg
Figure 53. Iomi'a.al Bearing I1vi-formnace nH at Vunction of CI-Luading
125
*ININAL0 ILICTR IC
90. 000 r pi400, 0000rl• . .
•; o, e0. 43_ !
200, ON I~f
*, : 400,000 ......___ _._
•' O~~, A 3.
•:. , .'~~~~100, 000 --- " 'OBI _t s
• )'i .... ~ ~ ... ...-. "--•- - • - ---
I moo*n I a
4000
00, 000 .... . .
0 .... . . ..I .... .. .. . . . . .
0 0 . . - -. I
nL lJ ... V]...
F/gurc 54. Journal Roaring Desiga Cl.arance and Pad FrequoticyVersus Srijug-PNd Spring Mitt'
12,6
[ ......... .. aw
OsINulALO ILIoTIgC
the spring-pad spring rate is not too sensitive above a value of 600 pounds perinch. There is a decrease in the spring-pad pitch critical frequency, butthis is of no significant consequence.
Probably the most critical problem with varying the spring rate to asofter spring is the practical aspect of setting the initial desired bearingclearances,
A complete summary of the selected design is shown in Table 18.
A comparison with a similar previous design shows the Wright-PattersonAir Force Base contract design uses a 584-microinch ground in clearance, as
Table 18
TURBOA LTE RNATOR TI LTING- PAD JOU HNA 1,GAS-BEARING DESIGN SUMMARY
Characteristic Design Parameter
Type 3-shoe tilting pad
Pad wrap angle (deg) 100
Pad pivot location (% from leading edge) 65
Pad material and weight density (lb/in.*) 304, 0. 28
Pad coating and surface finish (rms) Nitride, 4
Diameter (inch) 0. 261
Pad length (inch) 0.261
Angle between pivots (deg) 120
Journal material and weight density (lb/in.3 ) 304, 0.28
Journal wall thickness (inch) 0. 1305
Cold machined-in clearance (inch) 0. 000382
Ball material, coating, and surface finish (rme) 304, Nitride*:, 4
Socket material, coating, and surface finish (rms) 304, Nitride*, 4
Preload spring stiffness (lb/in.) 956,0
Shoe pitch inertia (in.-lb-sec') 0. 534 x 10-"
Nominal weight of shoe (pound) 3. 59 x 10"
orominal thickness of shoe (inch) 0. 020*Subject to change IR279
127
compared to the 382-microinch clearance for the present design. This wouldlead to lower bearing number and higher bearing power loss, as was seen fromthe trend of Table 16.
THRUST NAMNO DEIlGN-!i• PBOCEDU RE
The process of establishing the selection criteria for the thrust bearingwas essentially the same as for the journal bearing; the design-goal filmthickness corresponding to the maximum load was assigned in advance ofthermal distortion data.
The design criteria used in selecting the thrust bearing are:0 Maximum design load of 2. 0 g
0 Minimum film thickness of 100 microinches with parallel,undistorted bearing surfaces at maximum load
* Stable operation with minimum possible gimbal pivot frictionand damping from zero net load to maximum load
* Low power loss
' Minimum outside diameter
e Maximum gas film moment
0 No rigid-body natural frequencies in speed or load range
* Maximum stiffness over operating range
The general thrust-bearing design procedure consists of the following
principal steps:
0 Set up criteria for selection,
* Optimize load capacity with respect to film thickness for axiallystable face geometries.
e Determine bearing and gimbal-ring moments of inertia and pivotfriction characteristics necessary to provide stability with a mis-aligned thrust runner for both axial translation of the rotor andangular rotation of the gimbal ring,
The spiral-groove thrust bearing selector program for uniform clearanue,STBSUC (Ref. 4, App. IX), was used to obtain design and performance param-eters, This sarne program was used to compute the film righting momentfor the spiral-groove bearings, at a load of approximately 2 g.
The moment was then checked, using the nonuniform clearance program,SPORTH (Ref. 4, App. VIII). Pivot characteristics were computed so that thepivot friction monment equaled the film moment with collar misalignment atmaximum load.
128
OIUIRAL@ULIlOTlRiC
Program STBSUC, the spiral-groove thrust bearing selector program foruniform clearance, was used first to evaluate the thrust bearing design. Theprogram is based on analytical expressions for computing the load capacity ofa spiral-groove bearing with a uniform film thickness. In addition, the powerloos, axial stiffness, and tilt stiffness are computed; several other parameters,such as the Reynolds number, are computed and listed as an aid in the designselection. When the bearing design parameters have been selected, a plottingprogram is used to compute and plot performance characteristics for a double-acting bearing.
The principal parameters for the bearing design study are:
0 Ratio of outside diameter to inside diameter
0 Design film clearance at which the load it to be optimized
For a small value of optimized film clearance, the load-versus-clear-ance curve will have a steep slope, maximizing the film thickness at highload at the expense of the film thickness at a lower loading. This film thick-ness is normally selected as a compromise between the film-thickness safetymargin at maximum load and stiffness at normal operating conditions.
"The design parameters are usually selected so that:
e The Mach number is less than 0. 8 at the outside diameter.
e Minimum number of grooves is less than the maximum numberof grooves. The minimum number is based on minimizinggroove entrance effects; the maximum number is based on manu-facturing considerations.
e Relative @wash amplitude is less than 0. 15 X film thickness,
0 Reynolds number is less than 1500.
* The bearing number lambda is less than 50.
e Convective and transient inertias are much less than one.
o Steady-state and dynamic local compressibility are much lessthan one.
• Thrust-runner tip speed is less than 800 feet per second.
Based on the comparison of the load capacity predicted by this programwith available experimental data, the theoretical load capacity and stiffnessare multiplied by 0. 75 as a safety factor.
PERFOiaMANCE
The principal dimensions of the thrust face were established. In orderto evaluate the sensitivity of the cdepth of groove on this design, a met ofthree different groove-depth designs were considered. A range of the three
129
---- ---- -_w
*IN11IRAL*ELICTRIC
domignt; for- wh~ichI UKIQ groov depthi was vaI~ed and the collrrcpornbig optmuc~luirantco of 'i-imil groove depth in shown oil Figeure 55 Alimo shown iii tht-k'oreronL?(' d (Inwhio tol'rianu.' range Pu~ItubishrI.
400- H .imu No.
AA
A I IDouigiir Itun Nis.3o 00 4 701003i
2 00 IItuim No.1'7111002 II6(100 110(0 1000 12~00 1400
clvoov u Dei' ) 6AIII ( . )
Figi l*ro 1), "'P1 tiMIs BeItling Cl roove I Ii pkh Versus C It'll I'fl l nvp
j'he Hv'Jucivd tiv'H~gi r-un nlumber Im 4791lf003, shown on Fi'gir'v nil. Thotitpul. of 1,11H rmi putuhr p,' )grl'i vun ll'u~ proes(I s the' de'i gn Nmtived Imt-wm, 0imcit
rexpc'ctcdt Ir'tiii imv( tli-tisl. 1c'uitig riwi(' cnity. rIitiiv I~imwmfg inimithI Is irii, u-ti
0 Ti Lit IhidicIatilig thet gtis limed iiM the' hiil ivi~oni
0 Wtillis ol' nkiiido ditiniviti In Wailde dlimm-CurII
0S uill II inttionul opt-ad (r pi).
STUR 1 N 15 s26EST 12/21/72
200 4791003300 .00031 90000. o287 25.2 17.22 .003400 9209 973 1.93 71o2 .010 9.4E-6500 9.4E'.6 &05265 2.0 0.0600.1 25 25
RE~AD YOLD SKhSUCU
READYRUN
sFrRsucu 15127E57 1P/21/72
HS7.IUMLURiRI0ATEfl SPIRAL 0F~uOVE THRUST REAR [NO
RATIO nF OD TO ID a 2.0000
*** INPUT***
OFISION RUN NUMRER 4791003
SpunD, RPM 0. 90OOE405 C3RO(1VF Lt/flriO N 1TH 0 * 7300E+00TMPWRATUR1ý mn~G. 0,252OP*02 MROOVF/RIDOE WIDTH 3*1930E+01RP'ARINO 10 IN. 02.2170FE*00 iiROOVFE ANGLh DHOC fl.71205*02AM~!2rJT Pis~sS PSIA 00t7229+02 FXP COEFP.RWAININ F 0.94001-O5.1ASH ANC1L.FDO~ 0#30O00E-02 KXP COEFIF.HUN, VIN F 0094006-05nt pEN 8140 MATgLV/IN3 0,2890E+00 MIN ALLOO, RUID0I NIN 0. 10006-01SHAFT N9I(1NTqLA 0.52656-01 CLEAR AT OPT t1RVDIN 0,31001-03
***OUTPUT***'iACH NU,. AT U0.02920 MIN* NO, ORDOVES 2UJIMBAL 'ATD RRO TVK.qIN 0,62b05-01 VI3Q0S!TYLR-$FC/IN2 0.40205-09MAX NLJ. OF OROOV~b 39 REAkINO or), IN. 0,5'740FE+OUMIOLECULAR MFI~, IN. 0,21630-06 WeO INER. IN-LR-5FC2 0*21678-06ORrOVE Dl--PTrHv IN. 0,94558-03 TIP AP~FFl FT/SSC (),2254E+03
Figure 56. Helium -Lubricated Spira1-Groove Thrust B~earing, IOesign RunNo. 4701003, Ratio of Outside to Insidc' liametc'r 2. 0 (Shrc't 1 of 3)
1 31
*UNUALILSCTRI C
CLEARA14CIP, MICRO1INCHES 25 50 7LOAD LB 0.1339E+00 0.12615+00 0.Il55+00POW6A WATTS 0.73255+00 O.3Af4054'0O 0.26359400REL.ATIV E SvoASH AMPLITUDnE 0019p8B+0I 0.30068+00 0*21919+00MAX prIn3RfTION TO CLEAR RAT. 0.50025+00 0912825+00 095747E-01AXIAL FRiI0URNCY,CPM 0.97251+04 0*1548E'+05 0.16662+05TILT FREQUEN~CY, CPW 0.3553E+05 0,56552+05 0.6086E.05AXIAL .STIFFNFSS WtIN 0*141539+03 0*3584R*01 0,41518+03TILT S3TIFFNESS IN.LIJHAr) 0* 1277F+0.1 Oo$3039*01 0,96151#01ESPrFc.-* VIS. L8SESC/11l4**2 0.30221-09 00g910-09 0,30529-09RlI-YNUl0Ds NO, 0.644/9.02 001329F+03 0.1'7941+03LAmnlT)A 00~7008+02 0a217SE+02 0.96671+01CONVECTIVE INERTIA RATIO 0014808-03 0.59208-03 0,1l32E-02rWANSIENC INERTIA RATIO 0.9428WO0 005979R+01 0.17561+01LOCAL C0I4IR!SSo FtATIO,58 0,67513+00 0*164AH*00 0.750I1-01LOCAL COMPRESS. RATIO DYN 0,1631R-01 0.6513S-02 0*3115F-02
1,IAt., Ltl 0.a 1 1.71T + 00 fl.C204rE,-oI O.)035E-oIP0L~i~TC .ý02.1t"+O 0.1652P.*00 0014026+00
W7,A CSNiS A-LITUDE 0.91 w I I. +on o,7.1 '' o0 0. 1n688400MAXJ~.~C~ Týj rLf''AP RAT. o 0.P46P.-0 0,218R3R..O 004499-01
A~ IL U:~Y,~0. 1 71W'+05 0.,1720P+03 0. I 6*0
A.~IA0 ,SI 11.19 ILI/I -1-0 1A 0, 4424P~+03 0.4256E+03r 1 1 .-, ~3II i-v-ir. s, 1,1-L I/ RA ti r),I.)IiOr!+0 P 0.10)2564*0 0.9m61E+0l
!!Cie ý .. ,- TC:IN 0., WoFE-oo n. 1970R.-09 0. 3056E-09H.':y"'1jr'1. rql). 0) .659P+ol 0.321+0 0. 3QHE*03LAN'IP.)A 0 o 1)43 th+(U oo.1490F+ol 0#2417E+01
QllH 11 lVL-' llJij- 11I A RATI 10o P364R-02 0, 17POF.-O2 Oo53286-02
LWCAL 0'3M~Ih'IýSS ?AI'I(J,3S U,4219E-l -0I O.7006-OI 0,1875E-0ILfCUAL C1.11101481, ATIOIDvN 0. * I04E-07? 0. I 193-02 0.7487E-03
CI.R6AtIAhCR, "WROV'I 'CHII 1 75 W225
PU,4iLR W'A'IT I~ 22PF+(Yr0. 0,083- ý00 ) o97l99-0I
\kx l)19TIIioin'6l -rj c"'lAR RAT. 0.10606-01 O.Rl6lr-0>.~ 006458E-02A, I AL PlC)Jr4.1(jrVi': 0 * S23E+0li 0. I 30'-')5 0 *1I4366+05I IU L ii k"~ f, 'A o*.929E+o5 0.o6id?H +015 0,1246E+05iAXIALI. IW+1iI-.3 LR/IU U60R0 09351SCo~+ 0. 0105+03
T1F~~.v LT. Ll41.1/P'I*n 0..e9I25F0 *00977F+01 0. ,714E+01
;q 'YN(11LI11 J, 0*46bl5-'+l0 ,51 R,~ +().I~ 0.1982F+03LAM"P'IA 0.Ia 1i61-#0 1 0.11 50R+01 0.00706+01
l¼CI i- HdRTIA 14ATIII 0ol?0ri- 0,04717-0P.019E0
LoW.~t 61i.AP'llts RATIO,. 0~ ob 11791.'0 1 1. 05V~-0I 0,14134Ein02U)WAL 011PHI1.56. IhAri, DY! 095~514q-03 vi.4U40ý-cl 0.2QO4-01
Figure 56. f~leluni- ,ubricvated Spiral- Cvoovc, rhmist B~earing, D~esign H1111No. 4701003. Ratio of Outside totinuide Dinmete'r 2.0 (Sheet 2 f 3)
132
...... ............
IEN IRALO LI CYRI C
C A f IJI250Y) 300LMOJ, LH 0.4P'11E-o1 I s a..Iao&W-0I 0,100C4H-01
iWA~L 118 us 10AtI11LIi~I ~ U*I ) I,+(Y) 0 a I I 07f.400 0, 1 (AE+noC
t1~F~R'IUIPICY4 ia't4 0.4859E*01) 0, 447 -+o~i 0 .4099F40ýA\IW. ~I Fq 1ri, -L/ I 1. 0,26O40~~~1 U 0. 24 1 ::*0 01 'll2a+401Vt .' 14" ~I 7!- !'IN-URA") 0,61101301 0 0 1) 1 1011 00.4161E~401
~~~~ Vt4 ,~3CI*2400013~-09 0 ,4001 - -U 0, 400 1r-09~0~;1: *i 664 17+01 0, 7, 111(+01 0, Thr7dE+O,1
C -iiV' :1.'f1VH P191 A RAT IO u, I 4f1); -() 1 0.1 /9 i -01 0, $ 2 'irOW11~.Y ~~ 11ý,ri 1,A f OfA ?Iý1 10. 1 1, f , 0 I v1 ',1 714+0 1 0.l 1701 r. +0
L,,iCtL. COWW A sI, Uilos 0,67h1M-(l 0. 93 11:)10-0; .) 48OIýE-0LJi .CLfHPRS, 1AT0l,r~~ n~pqlo 0 1 02Y-- a .1 '10 1-f;,, 0 .1 F Ir-01
U1Lr.- LA 0. 7t- OC :, 1)1 1Mb-0I 0.l1f10.29-0 I
itII.A CIVr." 3aA!-1i A,,%LICU:,k 0. 1 11W 0. , 4A 4F4.O10.1 ,tiOR.F00
'Iv LT FW:l1U1-',y, t ý, W, 0 , 4 7,11n RW , 14 ~7 1r1+01, 0 4P1LHOUJY Lil/iA C00I 1Y4P+).l 0,0 4 :4- Vý0'3 u. 1 o" I
T1 U -STWFFW9q c),4~A 10 W40'+f I 0, 04pI,+0I 0, V) 399I:+oItfl v%, VIS 1Aq. L V~I~/ I Nfr 0 .4004E~-90) U a40Uor 09 (, .4107F-09
W.YA11- 141.16 OIA64I15*03 0 a. 9.1051,40+0 0, V~. .1
OUIV ~IV I1A r11A HAT 1( 0. 25~0 1 H.-U 1 00.~9011n-01 0.U* 1()10ý-FTiqA:':;I -NT TNkiTt A kATI 00 I704F+01 O .61911+U 0., I 106E+~0 IL;ICAI. C(TP-1LJ HM. -..;.Tlrl1 1) 1 A 391-4H, -u ? , 3444i-02 U . :100 1;-0 2
ALM', 1", 04 10? . ilI
AX~ I ý. S II I- .- i~C I /I 1 0.91 400 IU5?L~~~:y:1~fl; 1. .0 ')TOV1A 114' flAP 0. at0 V:~~i1I'I\'C~~~f OA VP I*r? t +j I 0£I~A'~*~J I ~~IA Aior I, '1 CI
IFigurc 56, I ItiiUiu- 1,uh1zh~ted SpI mrl -Ciome T:'hritt ibvoILilg, Iusign lim,No. 470)100%9 Hut~i o of ( u t 8dt, to In Mid t, I )iu iitC i 2. 0 (Shvv't S9 (it 3)
M INIAL ILI OTWI C
a Gas temperature (OR) used in computing the viscosity.
4 Bearing inside diameter (inch).
0 Ambient pressure (pmia).
* Thrust-runner awash angle, peak-to-peak (dvg).
0 Weight density of the bearing material (lb/in.') used in calculatingthe moment of inertia for tilt frequency output.
*Shaft weight (pound) used in computing axial frequency output.
*Groove -length/ bearing -width ratio (nondimenuional) measuredradially.
*Groove/ ridge -width ratio (nondimenslonal) measured at constantradius.
0 Groove snnle (deg) measured between tangent-to-groove and radialcoordi nates.
* Coefficient of thermal expansion for bearing material (in. /in.OF)used to coniputv dishing due to bearing film shear he-atihg.
* Coeffticient of thormal expansion for thrust-runner material (in./In. "F) used to compute dishing due to bearing film smhear heating.
0 Minimum allowable ridge width (inch) and minimum thicknues ofridge used to compute maximumn bearing number.
0 Clearance at which load capacity is to be optlimized (inuh).
The output consists of it group ofi overall performanve favtorw whkivhInclude the following,-
*Mach number (nondimenuional),based on helium and the thrustbearing outside diameter, used to indicate the potential of comn-prehsibility effects at the tip of the thrust hearing.
*Viscosity (Ih-sc/ in.5), based on tippropriate Irnu tit hinput tvmperv'turic.
*Moleculur' nman-free path (inch), bnsed on the appropriate gas ueto compute the offective viscosity.
*Minimum number of grooves, 'OMIpLItod on the hnaill that theC p~reh -sure ripple across the groovea in 10 percent of thev prosii~tir rislealong the grooves and related to the n~issumption that thc edgecorrmwtion is negligible.
*Maximum nuan'er of groovevi*, which 11na1UfVc turin1g 1and stru( turalconviderations have based on tin input quantity that 11mits the ridgewall thickness,
* a~tnimated moment of inertiti of the bviaring, with respect to n diam-eter, through the mldplane (in. -lb-sve),
1 34
IONIRAL t ILICTRI C
0 Assumed thickness of the thrust disk, based on dividing the annularwidth of the thrust disk by 2. 5.
0 Bearing outside diameter obtained from the inside diameter and the
ratio of outside diameter to inside diameter.
0 Groove depth (inch).
0 Thrust-runner tip speed (ft/sec).
Individual prformance Items are then tabulated in 16 columns, for one
thrust face only, with a heading for the land film clearance in microinches:
* Load capacity (lb), incl4ding the 0. 75 sanfety factor referred to above.
* Power lons (watts).
* Relative swamh amplitude (nondirnensional). This is the ratio ofcalculated awash amplitude to the maxdmum swash amplitude at"contract, The calculated awash amplitude is based on the swashangle, the film tilt stiffness and the moment of inertia of thebearing, calculated on the basis of an assumed thickness-to-diameter ratio.
Maximum distortion- to- clearance ratio (nondimensionai). Distor-tion is due to bearing viscous shear heating and is assumed to flowaxially, causing the thrust disk to take the shape of a cup of asphere.
* Axial frequency (cycles/minute), which is natural frequency bamedon the axi1l film stiffness and weight of the rotor.
*' Tilt frequency (cycles/ ninute), or natural frequency based on the
film tilt stiffness and bearing moment of inertia,
* Axial stiffness (Oh/in.), again Including the 0. 75 safety factor.
* Tilt stiffness of the gas film (Ib/rcid).
* Effective viscosity (lb-Aev/in.'), This quantity Is used in cal-culating bearing performance and is based orn the film thickness,viscosity, and molecular menn-free path. The mean-free patheffect reduces We viscosity a•rd therefore the load apacnity,becoming more pronounced as the film thickness demreuxes.
4 Reynolds number (nondirnensionnl), Thia is the ratio of inertia toviscous forces, Indicating whether the assumption of laminnrflow used in the present analysis is valid, Tho value should beless than 1500.
* Convective inertia ratio (nondimensionul), This is bnsed on thegroove width and tanRentinl velocity, indicatting whether the as-sumption of riefrligiblo inertia ffeCt is vvlid. The r itio shouldbe very much less than one.
I ;;,
$IN INAL@ILIOTRI C
0 Transient inertia ratio (nondimensional), based on the nominalbearing gap and axial oscillation, indicating whether the assump-tion of negligible inertia effect is valid. The ratio should bevery much l.ss than one,
0 Local compressibility ratio, steady-state .(nondimensional)based on the tangential velocity, indicating whether the as-sumption of quasi incompressibility is valid. In this analysis,a sectionally linear pressure profile is assumtod. To be valid,the magnitude of the circumferential pressure fluctuation mustbe small compared to the local pressure level. This ratioshould be very much less than one,
0 Local compressibility ratio, dynamic (nondimrnesionul), basedon axial oscillation and referring to the ilssumption defined above.This ratio should be very muc'h less than ono,
Figure 57 shows the corrosponding dosign conmputer run for a singlethrust bearing fa'e at the maxinum speed conditions of 200, 000 rprn and 58 5'R.
, SrUH 1K 09s4SRS'r 02/27/13
20U 4191004,1Ot 000311 PUtX00, (o,287 I 1,2? Qo.QQ400 0.•89 0.1, 1l.93 '11,2 ,'.•)0 9,0-6'00 9,4E-6 ,0.,.P6S P.u 0,06UO I 100 Pti
kfAnYOLr) sarsucu
READYRUN
SST"FSUCU W 1 P l',!"r O)/P 71/73
HFLTIUM LWJiRCA'rP) 3tPIRAl OP20UVR rHRU,'r qPAlPJ0
RAt!(l uF rj TO I') * 2.0000
*** I N RJ'I***
lo It I N iUWN NUMAPR 4791004
(10000P+06 OROOvL! 1.1w/oRn ,ir III. 1o00k+O0CEMPhRA Ili I,r)1'i, O "1,''11P,(I 0 0I #4/rI'D,1F -1.)'I ., 19.10 ,*0,1II•AI+NII l'•lN,'•,•qlF*,(10 0140110. ANMI.H4rI~ 0) 411 2UF:,w2AqI'ITI" .fl~. ,,'A fit I NP?40, SXP CU1',4rO', VVIN F' 0.,Q00F•O-0
SWASH Ql'.irlW ,, I's 1OF-O2 E•Px0hUN,1N/N F 0.940rM-O5Tr 6Lwi r l MAr Lit,,/IN• 0,• M40O9O0 MIN ALLflN, i o: I %, TN 0, I 000F.-l
SHAqi .iou ru; 0O,,,,,U-0, CLEW AT OPl' I'RV.r), IN 0.,1 OF.-O1
ligure 5'7, Illoium-Lubrieatnd Spiral-Groove Thrust Honring, lhosign HtunNo. 4791004. Ratio of Outside to Insido Dirtmotvr 2.0 (Shoot 1 or 3)
13 U
i!**oujrpu r***
MACH NO,. AT 1`10 0,11I 36E4(0O MIN. T-). O',iQUVhS R0IMHAL mmO PH13 TK., IN .)6P'50F-01 V I SCr1,TY , LR-SiC/~ 1111 )*066F-08MAX( NOL, LJ1 flWflCV'b 19 BIEAR I N, s m, IN. M.74014+00miLELICULA4 MFPP, IN, 0.,~C0l00110 W),l UINFR INI.rRSIFCP 11 P.6 1E- U6
CLVAHA'ICI%, \ICRUTNCHEPS 100 12 150 *
LCAtn. L9 0*1P09EF0) 0.? I42FP4O 0,1041E+01
NI-.A rV9 SmASH kqPE.Ir TUF a P 42 1 P+00 0,.4 1t4AIT-0t WJj E-0 It MAX DISTOWTQ4Tif rl CLVAP HAJI. OII770F-01I 0). A0 I F-0 1O.4166E-01
AYIAL FHHUUPN0-Y4CPI ,14 034<~) 1 .1,413P 14+05 0,1' J10'Fl+()TILT FHrC)UV4CY, CPM Da. I 114 4P+06 (). I 61~a0 0.1 'i67FeO6AXIAL STIFF~NESS LR/IN .1190[ie0 ,3eIio 0.V?0bF+U4f I I. ST1II4'NE';4 INq-LPI?Ar) O. 1,)P'W+0ý u, I IOPQ+0P~ 0. 90461:+02
HFYHO(LD'5 NU. 0). 1.11.+ 1 0 .1117PF--01 0. r)00+0OILAMPIDA i9 2 1 4*n F.+,>0.I 1 1P 0. 4r)Obp+eCONVELT.1V INEWTIA PATIO 0.PQ 1,?E-(4 ().A644F-(4 O.669IF-04
LOlCAL COMPRESSe RAM0,18~ ~ I'~f 005i~.0 0,1178F+00LOCUAL COMPI4ESSI, WATIt3, YN 0,;)44 iý H -f 1 ()ý1414-01 0,1.1I".29-01
LJA,., LQ 0.Q3UQ9 ý (') Q*'1 14E4,0O 0v 1 09 Fe+CO.41,I 1 .S 1.8 V. +0 1'O 0. 111155+01 o. 10719+0
A'iIAL FI?F JfllJY,('ij, 0,.7?114w#0 0924E0 0.bf`47I405
11 LT. F~i .JF I cy,:1" ,t1 1) 1. 1: +0A 11 1 01 .?206 0.1 j4PF+0OA
II 1.1, sr [IiFi I FNF LI/IeA 1. +0 *0 q .4 b 0 p r). 94k 11 F+.O?
, YJL~)~ , 'i 4u4ý o..dI e~c, 16.+)F I 0, ,PtflQF+0 IU.~ ~ 0 .ýi.O 10' 4~ , dF*12 0. 1
rtC N 'I Ii, I I'Ui X11kA1 AlF[.) 0. Q10~' 1 9-0' 4 . 1 1u PO-( 0,1,)05-01
I 'iCAL C'A-U,1i1,5 01 ~ 0I7F7F*O0o U. .141 ;)+00U1CALNe C)il 1 I0.ii'V' 1 (,ro- ~ l 00 I 0, 7MOCIP-O1
6LA1 kAJ , '. I6 1 i+ 1 6.H~ .''399 a 7
i-6LAI IV ý h~ A 'Iid[1* ) Pj l 0 i~' -0 1 0.* 4110 0ý-( I (I . 4 V1.3 E-I) I#A:, MS))TýT IP 1.1 CLP I 11, 0.? lil .- AI 0. 1 4j-0l C. I 1F-
I~ 'AL si:il: I- ! /1 1 I* 1.,(~~.);) 0. 1 IT*) 04 /iF +02
LA~ VR~ Ii 14 ,4b. '+(
H I r' I'7 I oI IT .1 r 1 dr 4 ae Ap ~1- uv 'w.N vii gI'& g I o.
No. 47 D IOU4, Hintio of hit WeIII I IIs¶4h(00 I1 V teZ V 2, n (S h If :2 If N)
13 7
01INIA1 ILECTRIC
CLFA4ANJCý, 'A; >'L5IC)h175LOAD, L" O.?9`t11F+OJ 1.)41125~F+0.) 0.)94rP+Ofl
RFNLATIVS ý,sJA' A.qYUThjrW' 0,47.11-01 le¾iME7P.01 O.5I09ý-OIK1A X r) ,I m? T IO a i 'ri, ýA R u . J .106 M -6 1 MMSO-OP 0. AI4AF-()AXIAL FRQP'Cq.~ '), IR7P4i5 0.? O Os.1243E+05T ILT PPR `JFNCY, CP'A 0,130O4+06 0.I20MR+06 0O1105F+06
AX IAL SI' Hj+P: - i 1.11 /1 N 0,2191 E+04 0. 1 Ab5)F~+0 0.1 573F-604Iru I 5I .TI~r'' ri-L.;At 0,507';H+02 0~.4$106E+07 U.1~640+02?
LA iF4OA 0.8774FW+0 0.710)F+0 I O.eSci5PE+0l
COA sIV ECr I vh i: .?I IA 1 hA I'l01',1 Q.! C'-0I 0.1541E'i-010 41to-LOCAL 0,07603.1A~f (.)1Q-~ O.IjA36'"-0 d*'j1A$4E-0JLOCAL Cfl.iYNL: '; 6 r ii n, 2)6-4)') (J*1l';o' .IthOE-)P
CL'AiA I 4CTrli FIO -?I ) : I A -IA1I. .1 * 11 I 4-u P 5 O 0. 47 50~:3AX IA.' ~ILAJ u.'CI4 0. P IO' */d 4fl V 00 . I ;4FO
*rILT itý :'jJANYC . 109 F.O11r. L. I O.F 0.74163-0OI
AXIAL FSI1'II-tNP5SLP/1'J 0.I33I;+4 ().IPY4404 -010 14I*1
TI LT 'UVIFFH!l19S 1.1-LV/0?) 0.ch~ 7*~~t 0 :)1 1F+0?
tEYN0LJ. I,-''. 9 1 111'i F+(); U. I .* I 9 4.C'P 0. I Opp.4O?
CUMNVECHb"; RJWTATIOAi 0,476i6F-01 0.53~69F-01 0.6019F-01
LOCAL C0Al4F; R. ATI1,TYiI (1, 14Q'W!:4ý 0. 121 O-O2 0. 1 CI F-07
CLEAkANCP, MICHI1ONCHES 475LOAD, L9 0.1627E+0nPOVI'ER b~AMr 0,1778E+01RELAT1VE SviASH AMPLITUr)P 0.B7.AIR2-OMAX flISTIRTIO'J f17 CLEA4i PAT, ().5200E-OP~AXIAL FRFUE!JCY'CPV. o.2310E+05TILT FnTiQUENCY, OPM O.qHF.4OI+9AXIAL ;TIFFNrAS LV/fl 0.8121E+01TILT srIFFNF1S, IN-LP/RAr) O.IHHIE*OýEFFECT, VIS. LP-SiiC/IlJ**2 0,270IF-OAkFWYIOLDS NO. 1.) 8 5 L-402
LAMRDA 0.40144E+01CONV~c'rivE INEPTIA RATIOr 0,6706F-01TRANSIRNT IN174TIA RAVO( 0.2745E-01LOiCAL COMPRESS. RATIO,SS 0.1I69IF-OILOCAL COMPRFSS. t4ATI0f,'1YN .P4-'
Figure 57. Heliurn-1Lubricated Spiral- Groove Thrust Bearing, Design RunNo, 4791004. Ratio of Outside to Inside Diameter 2. 0 (Sheet 3 of 3)
With an 800-rnicroinch total cleurance for the two thrust facein, a cmbined thrust bearing perfoxrmance run is shown in Figure 58 for the 948-nileroinch groove depti.. The various performance factors, such as load,power, axial stiffness, and tilt stiftnf-ss, are shown mib a function of the
- -~~:2- --- =... ___1- 8
'LONIRALI* ILICIRIC
f-s0 55I
- IW ,6 AI. O 4' do '@53 e "3355
41% ~ 5 1.
M W4 t.'W 0 u s a a~N4 0 J a a
C a 04ow 0 w o
- '4%1% N44 N P%4
a c s
ff rg as aYN a
- _j
A* 4C~~ v V5~ 4 -*'1 owsu_ m *sm ~s~as-IsOU
A. Ad *W p p W~
III
Ud~ U5 5j
~ * ::-E .~ ~m~ ~m 3m
ILI~E
1 ~ ~ ~ ~ . 4.Ca~ U 1~-
.~ ~-~ .. 'Ne
loaded side clearance of the bearing. Thi information for the computer pro-grams is shown graphically in Figures 59, 60, and 61 for the three differentgroove-depth designs that were considered.
The designs from these bearing performance runs are compared inTable 19 for the three groove-depth designs. Shown in this table are the signif-.leant performanco factors for 1-g, 2-g, and 0-g operation. It can be seen that.the largest groove depth, of 1220 microinohes, is not satisfactory to meet the criteria:
Table 19
THRUST BEARING DESIGN COMPARISON
Speqd (rpm) 90, 000
Temperature f R/0 K) 22. 5/14. 0Viscomity of helium (lb-sec/in.d) 9. 1 X 10"16
Ambient pressure (paln) 17. 22Rotor weight (pound) 0, 0507
Computer Run Numbers4791001 4701002 4791003
Desig grouvo depth (ln.) 1220 610 948
1-g Operation
Load (pound) 0. 0507 0. 0607 0. 0507Power loss, both faces (watt) 0. 170 0. 145 0. 150/.And clearance ýin,) 170 240 210Axial stiffness (Ib/in.) 230.0 45.0 36.0Tilt ttiffnede (in. -lb/rad) 5.6 10B0 0 8.4
2-a Operatiort
Lund (pound) 0. 1014 0, 1014 0. 1014Power loss, both tauns (watt) - 0, 200 0. 235Land clearance Oin.) below 100 140 100
Axial itiffneua(lb/tin,) -- 118 45,0
Tilt atiffnes, (in.- Ib/ rd) 34.0 10, 5
0-i Oparation
Load (pound) 0. 0 0. 0 0.0Poaer loss, both farom (watt) 0. 1153 0, 1282 0. 1196iand cleatance O.in.) 400 400 400
Axial mtifftems (lb/in.) 163, 4 1M!. i) 181. 2Tilt stiffneau (in. -lh/rerd) 3. 7n 4, 77 4. 24
140
OI GNEALO * ECTRIC
SpeuL' 11O, 00 LI)IlI Ambienft pi-v4:url, 1 0? (I)Tt.mpera~turv 2.5. 2' It ( )utsidut eliiiriiiter 0 ýy i> mi.DwA, Nt, 4' . 314 i Im.1i n dind. ~nicu r , 1. '287 it,Design Itun N~i. 47191001 T tod"ul~I uivuaIancl. I' . U 11111
CC
00 0.20 0,40 060 1) 0, 0 0 0, UO 040 (1.6t0Cli Iuu.'' (m N) Cu&~ltirme (1110dI)
u 6 1 1)(
Figre50 Hlim-ImriatdSpralGrov Truilinrng 120
MiChc Grcv et, Crori-lfeUa
Luade Sid C~vim~w
' SMINAL * ILICTRIC
Sp,,! '0,OUO vpi')l A11itiiirit pzvtlutle 17. 220 pMulhI1Ipueitur , I '215. 2"11 )Ltnitidi dilume ter 0. 5174 in.
DWM, No, 423D435 lInsdt, dianvteut 27B in.Design Run No, 479f1U02 TWtOl nxit )erwanct , 0,8 mil
N?
PIi
SII ,
A l 60 1)o ' 2 ' " U, . .. 4 0 a o , UCleav'ailvu (niL*) 11¢tLi~e(..1141l
SIR
Figure 60. Helium -1,ubriteted, ,Spiral-Groove Thrust Bearing., 810- iMicroinch Groove Depth, Performance as a Function ofLoaded Side Clearance
142
IINIRA Le ILI CRImI
Sp, ed •0oD000 rpm Ambient prui'.iur, 17. 220 ~l)iuTenmpterature 5. 2,•i ()ttiddr dirmhtil r 0 . 574 in
Dwg, Nu' 423D435 ftn•Jid diurnwtv' U. 287
Deviiln Hun No. 4791003 'rOtil Axial oltii'unutIt U. 8 molu
ka
IIL
U ,00 o. 20 040 oý OU ýW, 0O 20 0 10 0. bio,-
Clearancelf (mils) u* a u
4.00 U. 2U
Figure 61. H-eliurn-Lubrtoated, Spiral-Groove Thrust Bearing, 948-Microinch Groove Depth, Low-Temperature Performanceas a F~unction of Loaded Side Clearance
143
IINI RAL@ ILICTRI C
of a minimum thickness of 100 at the 2-g operating condition. The other twodesigns are satisfactory, and there is little difference between them as faras the performance features are conuerned. However, the design with agroove depth of 610 microinches should be considered a very limited designsince the tolerance on maintaining this type of groove depth is more seriousthan maintaining a groove-depth tolerance around the larger, 048-microinch,clearance. It can be concluded that the existing thrust bearing design draw-ing, 423D435, will adequately suit the turboalternator thrust bearing design.Table 20 shows the complote geometry of the thrust bearing design.
Table 20
TURBOALTERNATOR SPIRAL-GROOVE THRUSTBEARING DESIGN SUMMARY
Charatcteristic Design Parameter
Outside diameter (inch) 0, 574
Inside diamcter (inch) 0. 287
Inside diameter, grooved region (inch) 0. 364
Number of grooves 15
Groove angle (uug) 71. 2
Groove width to ridge width at constant radius (nd) 1. 03
Groove width (deg) 15.809
Ridge width (deg) 8. 191
Depth of groove (inch) 0. 000048
Collar awash angle (deg) 0, 003
Total axial clearance (inch) 0. 0008
Bearing material, coating, and surface finish (rms) Beryllium copper,none, 8
Collar material, coating, and surface finish (rms) 304L, NIrrIDE, 4
Figure 62 shows the performance to be expected for the high-speedoperation of 200, 000 rpm and 585"K under the maximum-speed and rnaximur-temperature operating conditions.
144
. .--, - . "-....1--00 "- - i i- -....
NINALOILITUIC
Speed 200,U00 rpm AImibtUIet p3rUIHUlr 11. 220 pi• Tvmpurutur'e 585.0011 )utdu dtinriter' U. 57•4 In.
S~DwMI, No. 421)'1436 Inallde diarmtur' & 0, 207 mh5•; Dedign I1un No. 47111004 rowtu axial dhwirolnev 0, 8 wil
. 4
C;i
Go QG O, 1o 0.40 0.60 00,00 0. 2U 0,40 0,60Clutw'aflnce Wllf) Cluar&nue (m•l )
to
rell
OU 0 0, 20 , 40 U. so1 q;•. {.1o 20 11 (, 10,(H
(Avh aniv•l lt (mlsJA) '!u|Zll,,l~t' l ,• (110.
E~ -- -1=
Miarotnch Gr'oove Depth, Room -Temperature; Pprf'ormance,as a F~unction of Loaded Side Clearance
5 ...... . .. .... ...
AL ILICI1I8
Appendix III
TURBOALTERNATOR ASSEMBLY PROCEDURES
This manual pruvides detailed assembly and disassembly proceduresfor the turboalternator development by General Electric Corporate Researchand Development in Schenectady, New York, under Contract No, DAAK02-71-C-0026. All instructions and procedures described are based on techniquesthat were successfully employed by Corporate Research and Developmentduring the course of th development program.
Special assembly tools are shown in Figure 63. All part numbers referto turboalternator assembly drawing 588E477 as shown in Figure 19 and tothe photograph of parts in Figure 64, unlees otherwise specified.
PRELIMINARY CLEANINO AND HANDLING
1. After manufacturing, inspect all parts under a microscope to ensurecomplete deburring and satisfactory surface finishes on all criticalareas,
2. Ultrasonically clean the electric stator in a solution of Freon* andvacuum-bake at 120VF for 8 hours. In'tall a liquid nitrogen coldtrap in the vacuum line between the vavuum pump and the furnace,to prevent oil migration. Ultrasonically clean all other partsseveral times in clean solutions cf chlorothene to ensure removalof oil and grease films and any particles clinging to the surfaces.
CLEAN ROOM ASSEMILY
1. Perform all final cleaning operations, assembly procedures, andinitial testing in a clean room to ensure imaximum cleanliness of thecomplete assembly.
2. Prior to entering the clean room. ultrasonically clean all parts andseal them in clean plastic. This step includes cleaning and sealingof tools, fixtures, and so forth used in the assembly procedures.
3. Final-clean the turboalternator parts in a clean room area, in asolution of Freon.
4. Assemble the turboalternator inside pressurlzed clean bencheswithin the clean room area. Air through the clean beticheai isfiltered to 99. 96 percent of 0. 3-micron particles.
!Freon, a-precision cleaning agent manufactured by E. I. DuPont de Nernoursand Company. Inrn.
Preceding page blank 147
I IN I AALO ILICT8IO
(f) tlllgl' Shnlft
-Opp I let tiTut-bi4lfItl ilW anI -hl.0 ll ilyf
.inio
Figure 63. Turboalternator Assembly Tools and Fixtures
fth-As bumnug Probe Hojlder I4wUvId1M abatax' MotionNosaPinopart 14 PatINosals Asse %blyt
Nossi, lr Putt Is
Journal Hearn Sapo temsni , JB~ i
Proximlty ProbePart III
Outa li ~t?~axtal Imuanbi J a rbaxll4ian u W I ad.IIMr ax hu k W
Intiar Thrustl Iiaaa tng lAsovii*4~I
F~igure 64. Turboulternator Parts
148
CAUTIOHandle all parts carefully during assembly to
prevent damage to critical surfaces and dimensions.
aportion of the assemblybecause most lubricants are very volatile at roomtemperature and will cause serious contaminationin the refrigerator system.
INITIAL WHIL.TIO.SHAP, ASSEMILY
Assemble the turbine wheel (Part 19) onto the unmagnetised shaft (Part1) as follows:
NOTI: Inspect finishes of turbine wheel bores priorto assembly to ensure that the chemical conversioncoating and lubricative plating were not damaged inprevious assemblies.
1. Scribe lines at each end of the shaft on the outer diameter, alongthe axis of the shaft and along a line on the back of the turbine wheeladjacent to the shaft, for realignment after the balancing operation.
2. Install the thrust end of the shaft in the brass shaftholder (a) andtighten the clamping screw. Place the turbine wheel (Part 19) onthe opposite end of the shaft, aligning the scribed lines. Thread thewheel pusher assembly (a) into the end of the shaft, using the "ireadsin the inside diameter of the shaft grinding center. (For details ofthe wheel pusher assembly, see Figure 65. ) Then. after recheckingthe aligmnent of the scribed lines, tighten the wheel pusher assemblyby holding the end of the threaded rod (Part 8, Figure 65) with awrench on the flats and tighten the nut (Part 11, Figure 65), forcing1the tarbine wheel onto the shaft.Normally, a torque on the order of 1. 0 in. -lb is required to push
the 0. 8-inch-diameter aluminum turbine wheel onto the shaft with0. 0004-inch interference, wheel-to-shaft fit. A sudden increasein the torque value indicates the turbine wheel has been pushed com-pletely onto the shaft and the stop on the inside diameter of the tur-bine wheel is touchitig the end of the shafto
3. Remove the wheel pusher assembly from the shaft and insert thespecial hexhead cap screw (Part 32) in the threaded shrift center.This hexhead cap screw, with the top ground flat, is used for theviewing surtfce of the thrust proximity probe during final assemblyand testing. Using the socket wrench (b), torque the hexhead capscrew to 5. 5 in. -lb by holding the shaft holder at the opposite endof the shaft.
140
A'OIN4 IRAl~ to, MGMk,
.,N.RAI.,MawI
415
IRAL O ILICT!IIa
4. Remove the shaft holder (a) from the shaft (Part 7).
WALANCN THE ROTATING ASSEMBLY*
After the turbine wheel has been installed onto the unmagnetized shaft andthe hexhead cap screw is torqued in place, the assembly is ready for the bal-ancing operation.
1. Place the assembly in a balance machine and balance it at 6000 rpm,to the maximum unbalance displacement in each plane of 2. 0 4in.For balance correction, remove mawrial from the turbine wheeland shaft, as outlined in Figure 66.
I * .lt*,Ie, lbt l lS'
I)•| hlll111lllli~rQ •tt, I l• 1.. l 4,
-'1 'm
- - - -I F "-
SMI, to ,
I' I I f. 15111411,11
Figure 66. Turboalternator Shaft Assembly Balance Specifications
INNER THRUST BERN SEBL PR 2
The part numbers in this assembly appear in Figure 87. Preassemblethis assembly on the bench.
1. Install the pivot screws (Parts 4 and 5) in the mounting flange (Part1) and in the gimbal ring (Part 3),
2. Assemble the thrust bearing (Part 2) in the gimbal ring by adjustingthe pivot screws to approximately center the thrust bearing,
3. Repeat the above procedure to install the assembled gibal andthrust bearing into the mounting flange. After the above parts are
TOTaiunclg equipment to discussed in detail in "Shaft Balance Equipment,above.
IN___UR•RIST IARNO $$|DLY(PAT151
GIN I lAL ILI CTRIO
/]
i -
- ,t
Figure 87. Turboalternator Inner Thruat Bearing Assembly
assembled, acourstely center che gimbal ring %nd %hrust bearingand adjust the tipping force of the girnbais.
4. Center the gLmbklby uding a .No. A8 drill, 0.07 6 incn Ln diameter, tomeasure the spacing between the gimbal ring and the mounting •.ange.Adjust the pivot screws accordingly.
8. Use a No. 54 drilt, 0. 055 Inch in diameter, as a guide to adjust thespacing for centering the thrust bearing in the gimbal.
8. Position the inner thrust bearing assembly on its test fixture (j).When the thrust bearing ix properly centered, the assembly willslip easily onto the fixture.
7. Adjust the thrust bearing pivot screws and the gimbal pivot screwsuntUL a 1. 5- to 2,. 0-g breakaway force is measured on a force gagewhen pressed against the outer edge of the gimbal ring and thethrust bearing at the midpoint between pivot.. Then lock the pivotscrews in position by tightening the locking set screws (Parts 8 and 9).
8. Recheck the centering of the thrust bearing on the test fixture (Q)after adjusting the correct tipping force.
NOTE: The gimbal rings and thrust bearings are de-signed with a spring at the pivot sections to maintainthe tipping force over a wide range of temperatures.
Gi IIRALO ILIN Im C
OUTER THRUST BEARING ASSEMBLY IpART 13)
The same procedures are followed in the assembly of the outer thrustbearing, as outlined above, for the j4nner thrust bearing. Test fixture jaused to center the thrust bearing, is designed to ar.cept both thrust bea:-ligs. For details of this assembly see Figure 68.
6 Q 71
K. " '
, '\ J4I
A~i•A
TYPICAL THIVRUT BEARING ASSEMBLY
Figure 68. Turboalternatcr Outer Thrust Bearing Assembly
PRELIMINARY O.OURNAL BEARING ASSEMBLYAND STATgR INSTALLATION
1. Assemble the inurnal bearing support clamps aind stems (Parts 26and 27) onto ýhe housing sections (Parts 8 and 13). The spring stemsin bearing stem s.,pport assemblies (Part 27) must be inserted fromthe inside diameter of the housing before the stem cl~arps are at-tached to the housing, The other journal bearing support stems canbe inserted through the stem clamps from the outside of the housingsections.
2. Position the journal bearing stems off center to allow clearance forthe installt•!on of the gage shaft (f) into the stator section of thehousing (Fart 8), from the nozzle end. The nozzle-to-housing
153
S, ... ., . . • ....-- - -__... . . - , -, - =-i -=. . ... ........ .. . .. :.:-- : ::..--. . .
GEINIL ELECTRIC
rabhi t fit dlrnionsions oni Jih housing 11'L' U!"d to position the gageshaft vrrtically and to o'cntev the gage shult into the housing. An-Hernble the gage shaf't holder (g) to thie housing nozzle flange withthree screws (Part 49). Then gradually tighten the centering screwin the gage shaft holder wo positiun the gigui shaft securely Into thehousing.
3. Assemble the three~ journal bearing pods (Part 23) Into the housing.Cheek the pad orientation before tussumbly (Figure dO). Hold the twosolid steirts in thu assemblies (Part 268) ftr-mly, rso the pads areagainst the gage shalt und tighten the stem ulamps se'Ž,urely.i Posi-tion the s~pring stem ahsembly (Part 27) so the pad is hold gentlyagainst the gage shaft, without applying pressure from the spring.Tighten the spring stem clamp securely.
T eilting PadJei 'u~K~'nao itn-la ora
th'ilz!, jwin Shft 'in Th1"s 1,10l-1n
Ons ~ ~ ~ 1- tIf
INL itLITRIC
4. Remove the gage shaft holder and gently tap on the end of the gageshaft ,ith the back end of the brass shaft holder (_.) to loosen thegage shaft from the housing. Do not allow the gage shaft to dropout of the housing. Insert a dummy shaft (h) through tie housing andgently push the gage shaft out as the dummy shaft is inserted. Theends of the shafts must be held together during this operation to pre-vent the pads from dropping off. Tape the dummy shaft to the hous-ing so it will not drop out.
P 5. To install the stator (Part 10), the stator leads must first be fedthrough the housing and out through the hole in the housing wall.Orient the stator above the housing so the leads can be drawnthrough as the stator is inserted into the housing. Press on thestator shell to be certain that the stator in bottomed in the housing.Then tighten the set screw (Part 46), to secure the stator in position.
6. Assomble the journal bearing padr, in the outer thrust sections of the
housing (Part 13) in a manner similar to that described above; how-ever, it is necessary to temporarily remove the thrust bearing andgimbal assembly from this section of the housing by loosening oneof the pivot screws in the housing sectiot,. Insert the swnre gageshaft (f) from the thrust bearing probe holder end (Part 14) usingthe housing rabbet fit dimension again to center the gage shaft.Mount the gage shaft holder to the outer thrust bearing flange andtighten the centerinig screw in the holder, to position the gage shaft.Install the pads and stems as outlined previously, using the sameprocedures to remove the gage shaft and install the dummy shaft (h).Then reinstall the thrust bearing and gimbal assembly over the dummyshaft. Readjust the pivot screw previously loosened, to give the
1. 5- to 2. 0-g tipping force and the pivot locking screw reset. Par-tially remove the dummy shaft from the assembly, but position itto hold the journal bearing pads in place and allow retesting of thecentering of the thrust bearing with the bearing test fixture (j).
SHAFT HOUSING ASSEMBLY
1. After completion of the rotating assembly balancing operation, installthe brass shaft holder (a) onto the thrust end of the shaft and tightenthe clamping screw. Hold the shaft securely by the shaft holder andremove the probe viewing screw (Part 32) from the shaft, using thesorket wrench (b), Remove the shaft holding fixture from the shaft.
3. Remove the 0. 5-inch-diameter turbine wheel from the shaft, usingwheel puller (d), Disassemble the halves of the wheel puller assembly(Parts I and 2 of Figure 70) and reassemble them around the turbinewheel, as shown in Figure 70, Hold assembled Parts 1 and 2 of thewheel puller (d) with a wrench on the, flats, as the pulling force isapplied by tightening the screw, Part '7 of Figure 70. Normally, atorque of 3 to 4 in. -lb is required to remove the wheel.
awnmnn¶¶ew�wmrEa'R7 �'¶YDY fl�TV1WWflflIW Y�Irrt
SIN IRAL * ELECTRIC
0---- Li
a
IIII -A
-2- N I . I
/ r"" I /t I
I k -3
I '7.I 1<> >'-I
c"I I- I I -/ I 1)
I I I/' �"I I I�\ I
/ I
K" II - \. ""�
-, £ "/ j N�', a''< I
/ A#M1 Is/ ' -" / N
/ I
156
rINIRAL 0 @ Y!I-
3. Place the preassembled, inner thrust bearing assembly (Part 22)onto the previously assembled housing section (Part 8) with the
journal bearings and stator installed.
4. Thread the tapered shaft extension (W) into the end of the shaft at theturbine end. Insert the tapered extension and shaft through the innerthrust bearing assembly. Position the end of the shaft extensionagainst the end of the dummy shaft and gently remove the dummyshaft as the shaft in installed. Keep the two shaft ends together sothe journal bearing pads do not drop off the stems.
5. Remove the tapered shaft extension and, using wheel pusher c, in-* •stall the 0. 5-inch-diameter turbine wheel and probe viewing screw
(Part 32). * After completing the above assembly, remove the shaftholding fixture (a).
NOZZLE ASSEMOLIUL
Preassemble the nozzle assembly (Part 18) on the bench. Do not assem-ble the gas inlet and discharge cylinder (Part 15) at this time. Ensure that
r the c-seals are centered in the seal grooves before the inner and outer nozzlesections are assembled. Orient the offset mounting holes in the nozzle sec-tions for correct alignment during assembly. Torque the six No. 4-400. 5-inch-long assembly screws to 5. 5 in. -lb, to ensure a metal-to-metalseal at the inner section adjacent to the nozzle channels.
PRELIMINARY THRUST BEARIN2 SHIMMING
Depth micrometers are used to measure the shaft position to obtain theaxial shim requirements (Figure 71). Table 21 lists the clearance specifioa-tions and the required depth micrometer measurements and calculations todetermine the thickness of the shims. The following procedures are usert:
1. Place the partially assembled turboalternator in a vertical position,thrust end up, on n stend with a cutout to clear the turbine wheel.Press down on the inner thrust bearing mounting flange to be certainit Is fully seated onto the housing.
NOTE: Threaded holes have been provided in theouter flanges of the thrust bearing assemblies(Parts 13 and 22) for jack-out screws, if required.
2. Position the thrust clearance ring (k) onto the Inner thrust bearingmounting flange (Figure 72). Measure and record Distance A fromthe top of the thrust clearance ring to the top of the shaft thrustrunner.
*Described above under "Initial Wheel-to-Shaft-Assembly."
157
SIN, L I_ i, 731'.
i•~~~0 140'. -- •"
..
M in, In.
+ -- - . -,- ----
Figure 71. Turboalternator DirniusiorLa for Shimming Thru'stBearing Posltion
Table 21
AXIAL SHIMMING MEASUREMENTS FOR TURBOALTERNATOR
Wheel-to-nozzle clearances 0.0015-0.0020Shaft total axial iravel between thrust 0.0004-0.00ve
bearings OPtv,Inner Thrust Bearing glum 444,ob~t mecment AA. Distance from top of thrust clearance
ring Wk) to top of shaft thrust runner
B. Repeat Measurement 1 with nozzleinstalled
C. Inner shim (X) required: X.(A - B)+.0.0015
Outer Thrust Bearins ShimD. Repeat Measurement A with nozzle as-
sembled and inner thrust shims (X)installed.
E. Distance from top of thrust clearancering (W) to outer thrust bearing surface
P. Outer shim (Y) requireds Y • (0. 732 - E)- (D - 0. 6505) + 0. 005
158
Thrlut
• Fi•,•r T•.Thru• C~arace gnl[ oeitonedtn ner arancter
pint
s nA Inwhe e. iten
/ / A
____
/)el
i i
Figure 72. Thrust Clearance Ring Pohitioned in Inner and OutegrThrust Bearing Assemblies
3. Measurement of Line B requires installation of nhe nozzle, Posutionthe nozrle assembly on the housing, aligning the mounting holes inthe nozzle with the holes in the housing flange. T hen evenly tightenthe assembly screws (Part 49) to bring the nozzle fully into the*thousing. Measure and record Distance B.
4. Calculate the inner thrust bearing shim requirements, as instructedon Line C.
5. Install three shims of thickness x, calculated above, between thehousing and inner thrust bearing mounting flange, 120 degrees a~part.Be certain that the inner thrust bearing shims are positioned so theturbine wheel Is not touching the nozzle and the shaft thrust runneris resting on the inner thrust bearing surface, Then measure andrecord Distance D from the top of the thrust clearance ring to thetop of the shaft thrust runner (the value obtained should be the sameas that for Measurement A above).
6. Place the thrust clearance ring Wk onto the outer thrust bearing &a-sembly (Figure 72), Then, with the thrust bearing parallel, measureand record Distance E from the outer thrust bearing surface to thetop of the thrust clearance ring.
7. Calculate the outer thrust bearing shim requirements, aq instructedon Line F.
NOTE: Five mils are added to the above distancemeasurements. The outer thrust bearing cannot bemeasured accurately. Final shim adjusti oento aremade with proximity probes for greater accuracy.
159
$ININEA L9 ELECTR IC
B. Thread the tapered shaft extensioin (G) into the thrust und of the shaft.Install the outer thrust bearing assembly by positioning thle dummyshaft against the end of the shaft extension and slide the outer thrustbearing assemnbl~y onto the shaft,
9. Remove the tapered shaft extension and inuert the three outer bearingL shims, of thickness y, calculated above, between the inner and outer
thrust bearing mounting flangei;, 120 degrees apart.
10, With the outer thrust bearing in position, turn the lihaft fromn thethrust end to be cortain the turbine wheel and shuft are free to rotate,Then gradually tighten the thr'uut-end assembly sc'rews (part 40).Do not overtighten the screws; they should be just snug enough tohold the shims in place,
INSTALLATION OF THRUST PROXIMITY PRCBE
1. Mount th~e turboalteri-intor nssembly onto R support stand, turbineend down. Three lengths of 4-40 threadod rod can be used in placeof three alternate nozvzle assembly iscrews to attach thle turboalter-nator to the support, The turboalte rnator support mhould be mountedon twn indexing tables ki0. thv turboalternator can be rotated froml- avertical tu a hurizontal position and can be rotated radially nbouLt theaxis of the shaft. Figure 73 is n typical test station showing oscillo-scopes, the Wayne- Kerr Instruments, and n turbualte rnator mountedonl two indexing '.able s.
2. AdJv~st the Indexing tables to position the turboaiturnntor verti cally,with the tu-rbine ond downi. __
3. Inutall. the tl "lst proximnity probe (Part I1I) into the nozzle bl~c~kby carefully turning thu probe in until it ttouc hvs the, und of tilt Fihe rt;then back it out rslightl-, and tighten the jam) nut.
fAU'rO
Crtre MuIst bV ex.'rcimod in cnrs'ying out thim pro-
istrumnent (oel.101)
n~terntor huItig
I INEAL@ ELICTRIG
Figure 73, Open-Cycle Test Station for Turboalternator
NOTE: All of the individual oscillators in eachWayne-Kerr instrument MUst be disconnected andrewired to provide operation frorn a separate,single, 50 kIIz asupply; otherwise, the individUtAloscillaturs k. Ill ope rate out of phase and willCaUSC huating nnd erratic signals,
7. Set the oscillosc ope Y-axis sensitivity to 0. 2 d- c V/cm miii d theX-axis to a sweep of 2 nis/cmy, Adjust Y-nxis gnin control to4. 9 cm for 1 volt of d-c inpuLt.
B. Switch the Wayne- Kerr instrument to the ('11EC'K position, Shortthe osciloscope \'-nxis input nid ndjust tht! zero position to R pointthat is I cm down from the top of the sc reeni scalE',
F). Switch the Y-axis inpuft on the! oci111ot'vol t() th, d -c position anid nd-just the Wnyne -Kerr set ndju tmtnint to give n 4. 9 em de flection onn
the omcillomcfqpe.
161
rBININAL*ILET
10. Switch the Wayne-Kerr iWrtrument back to the READ position, Thethrust proximity probe is then ready to be used, and after Lhe ad-Justments outlined above, the sensitivity of the oscilloscope in 40014in/cm.
11. Readjust the thrust probe position so the output signal froan theWayne-Kerr instrurrent appears at the middle of the oscilloscopescreen,
12, Measure the total thrust bearing clearance by noting the thrust probesignal level on the oscillcscope tiler the above setting. Then rotatethe turboalternator 180 degrees vertically and note the thrust probeoutput signal on the oscilloscope with the shail resting on the outerthrust bearing. Change the outer thrust bearing shim height (Y) asrequired to meet the spucified total axial shaft t-avel of 0. 0004 to0. 0000 inch, and record the actual shim thickness on Line F ofTable 2i, for future reference.
13, The thrust proximity probe Js also used to measure the clearancebetween the turbine wheel and the nozle. Reposition the turbo-alternator vertically, Ath the tWrbine end down. Temporarily in-crease the thickness of the three shims (Y) by 0. 001 inch, anoremeasure the total travel %f the shaft between thrust bearings withthe thrust probe. Then gradually decrease the thickness of the threeinner thrust bearing shims (x) in steps of 0. 001 inch. As the valueof the inner thrust shims denreases, the thrust probe may have to bebacked out so it does not touch the shaft.
A thickness of the shim. (N) is obtained where the turbine wheel justtouches the nozeliel ting the shaft off the inner thrust surface, re-ducing tho total travel of the shaft. Once the above condition is ob-tained the thi.'knews of .hims (x) can be calculated to provide a 0. 0015-0. 0020-inch clearance between the turbirne wheel and nozzle, beyondthe inner th.rust bearing surface, Record the actual shim thicknesson Lino C', for future reference.
14, Install shims x and y of the corrected values to meet the axial clear-anvv specificntions, Align the scribe, marks on the outside of thehousing at the, thrust end to align the journal bearing support stemsat each end tif the housing assembly and tighten the thrust assemblyscrewm (Pnrt 40),
FINAL JOURNAL BEARING PAD ADJUSTMENTS
1. Adjust the indexing tnble to position the turbonlternator horipuntallyand ootate the turbonlterfmator until the pad stems between the shaftorbit prominity pr'obe mounting holes are pointing directly up. Inthis positi,1,1, the turboalternntor mhnft im rusting between the hottumpaldls
162
IAt IL6GTSIC
2. Install the four proximity probes (Part 11) adjacent to the shaft, twoin the housing at the turbine end and two in the inner thrust-bearingmounting flange. Tighten the probes with the jam nuts.
FCAT UION7
Care must be exercised in carrying out this pro-cedure, to avoid damaging the probe Textolitethreads,
3. Connect the 10-foot leads between the probes and the Wayne-Kerr in-
struments. The probes and Wayne-Kerr instrumients are identifiedby looking down on the turboalternator from the turbine end. Thefirst probe on the left is X%, and the corresponding probe on the right
is Y1. At the thrust end, the probe on the left is XI and the probeon the right is Ye.
4. Check to ensure that the ground lead is connected to the turboalter-nator housing,
5. Each of the probe outpo't signals from the foui Wayne-Kerr instru-ments must be filtered and connected to an oscilloscope with dual-trace amplifiers for both th,7 X axis and the Y axis, Connect the out-put from Probe Xt to the X axis and the output from Probe Y, to theY axis of one trace, maud Xj and Ve to the X axis and Y axis of theother trace.
6. Set the scope sensitivity on alI four channels to 0. 2 d-c V/ca ondadjust the four channe1i to, 4,f 9 cm for I volt of d-c input.
7, Switch all four Wayne-Kciýt" ihktrunit'tts to the C(HEC'K position, Ad-just the four channels to givo q dpf e~ction of 4. 6 cm by adjustingthe Wayne-Kerr set adjustnt't•and switching the scope betweenthe shorted Input and the d.-c inout,
H, Switch the Wnyne-Kerr in;tri~nnents hack to the READ position.
in. Short the scope inputs,, and adjurt the two zero poostions to pointsthat are I cm down from the top and 2 cm in from the right. Switchthe scope inputs to the DC position. The orbit probes are then readyto be used, and after the above adjustments, the scope senuitivityis 400 6in/(nm,
10. Readjust tlh, four probhe positions so the scope signals are just on thvbottom left corner of the screen. This adjustment is made with theturbonutt.rnatnr in the position dewcribed in Step 1.
11. The shaft cLearanvoes are measured by rotating the tirhboalterinatorassnibhly 120 degi'ees to the rii'ht and left of the position de sc 'ibhrdab•,v for setting the probes and plotting the probe readings in thethrve positiouns. An act urlate method (if roc'ordingl the three ptisitionsis ti take a triply expomure of thu turbine vnd only, unid then of the
163
IIIN- IRIeILIT
t.111ust tnd only, with, I. jeopt, cnitvu c'u t a ~sectionl of the scope gr'idfroin a photo and use this~ svct iOn tO nwIaSIT U xVth longt hs of the trianglelegs formedby the three pointm, in centimeters, The legs of the tri-tingles should be approximately equal, Skibstitute the lAvOerage Valueot' the triangle lugs in the foilowing equation to deter-mine the mleanbeatring pivot -point radiul clear'ances In rnlcroinches:
(.t- 115 G., x (the length of the averuge leg in centimeters)
'rho radial clearance, Cl, of the shaft should be 24012G inicrolnch'os.Greater differences III thle rudial clearancus will requi1re adjuertment,of the spring atems and pads W~art 27) monitored by the proximnityprobes and the oscilloscope to obt,..in the proper clearance'. Duringthe adjustment of the Jouraul boaring puda, tile ,upport. stemsa canl onlybe moved minute di~tunces or the journtil bearing pads m~ay drop offthe sterns, In this cusv, the Journal beurings must be reset to the
-r gage shaft, as outlined ubovv under "f'rel Iminary Journal Bear~ingAssembly and Stator Instullation."
TURIOALIERNATOR 21PERATIONS
T'rel~itniary, turboulternutor performunce tests are made in an open-cycleteut station (Figures 73~ and 74). F'igure 75 is a cutaway view of the turboalter-natc~r &toexnbly for open-vyvle testing; a schematic diagram of the station ioshown ih'lrt*re 76. hligh-preNSsure gatS (holiumi or nitrogen) 1. supplied fromgas cylindoer banks Rnd im expanided acrops the turbinp to atmospheric pressure,Liquid nit~rogen temperature tvwts arre pvt-formed by precooling helium gaswith a eooIt~kg coil Immersed in liquid nitrogen.
Figurt, 74. TLuO)Miltlt 11INIM \MStlI uhl~v [In Opull-( ,c It Tust Stand
ININL ILCTRI
Altvi-nator~~V p*tuny hrz
Il N.9 -. e 111 P M S I U1.IJIU Y
HD 11 L 1ti N 4SA ll 1 V
H 1 ****li VjItag ,(% o , l.
H3 44utV lug v ls
S4 M I..,1O . 4. li
I V,
.J,-44
*Ul R•ALO ILI ,omo
;,mows aIl
• ', I• 7 ... . .. .
- 1 &P . ...
I ,,_,_____o_____
Figure 76. Open-Cyele Turboalternator Test Station
Figures T3 and 74 show a two-stage turboalteriator mounted on twoindexing tables in the opesn-cycle teot station.
The inlet gas flow instrumentation (or testing twn-stage turboalternatorsto shown In Figuire 77 along with a moisture analyzer and temnperhture -mea -suring instrumentation, The moisture analyzer is used to monitor the mols-ture, co ntent of the helium gas supply prior to low temperature tests, to re-duce tho poss ibility ot ice crystals hind~ering the operatiou of the turboaltcr-nator,
Watur and mercury manometers, not shown in the above figures, areused to measure the housing and nozzle pressures.
A temporary nozzle gam inlvt fixture (a) is ~attached to the nozzle atisembly,Trhe same size-c seals used in the nozzle assembly and the gas connectio~nh arealso used to assemblP the gas inlet fixtures onto the nozzles, Attach the gasinlet fixtures to the nozzles by tising three 4 -40 threadud rods to replaca al-ternator screws in the nozzle assemblies. Tight'.m three 4-40 nuts againstthe gas inlet fixture flanges to comprers the cseals. The threaded rods atthe turbine riad should be of sufficient length to attach the turboalturncitor tothf, support stand on the indexing tables, Tlwo gas connections are provided it)the gas .nlet fixture. The larger connect ion is for the gas inlet flow, andl theorraller connection im used to meabure the nozzle inlet preadure on a mercur-ymaromete r.
166 lf
- -tf
OERI. LOILE. R
ItI
. ... ... -...
IFigur((1 77 . (pen-(~l 'y Ti 'i> bov ltv vnator I nlet Gas~ Insi~rurnintatictn
ThIS Itli J'''42 Il I4 1m;1 II14XI 1;414 IlIIm: ig c(. ' risi ( I. O()l osu I4 It) d duO Ii'lig tis!1~ h
t n 111 1 Iiv Ii cIý ii II 4I tl titi J iI 1iI t~ . ~~ ~ L- It
li, iim III ,1 4t444413)4l 'I Ii Il I 'lUgjJ-)wI't1- 1 A
ju tt1 'K! 11 (11-; ; 4.1.ur(f0- 03 t 003:1 !4'1111'h4I()II I the1
W I .( - I0!1 1( 1 l t I (.I ~
DIN IRA LILICTRIC o h n~
onertheshf or-bit proximit~y probes can be used to monitor thespeed by coninecting the probe output signal to another scope and cal-culuting the speed from the oscilloscope sweep rate.
.3. Check to ensure the proper operation of the turboaiternatorL by moni-toring the proximity probe signals on the oscilloscopes. Thes orbitsshould be quite small in diameter, (approximately 1/3 cm. or smalleron the oscilloscope screen) and uniformn. If the shaft is hitting, theorbits and 'he thrust signals will be quite distorted and will contin-ually var-y in size and amplitude. In this case, the thrust and orbitmeasuremet)ts made with the proximity probes will have to be re-peated in ordert Lo correc.t a shim adjustment or a shift in a padposition. The pad probe a.nd thrust gimbal signals should be uniformlyonce per oscillation.
MAGNETIZING SHAFT MAGNET
Trhe measurements and adjustments outlined above tinder "Installationot Thi-ust Proximity Probe" and "Final Jiournal Bearing~ Pad Adjustment"cannot be made with the shaft magnet magnetized, because the force of themagnet's fliuid would nut allow the shaft to float freely when mneasuring theclearances. Therefore, the turboaltez'nator mlUst be disassembled to removeand magnet ize the shaft, The assembly procedure must therefore be reversedto r-emove the shaft for, magnetizing. Keep the shims identified while the parts
ave disassembxled, and use the dunimy shafts to hold the pads in position asI ~the shart is x'enioved.NOTEI: If i1 is necessavy to c:ompletely disassemblethe turboalternator, use a hole gage to assist in ther-vmoval of thu stator , Adjust r. 0. 3- to 0. 4 - inchjutre to thle smalIcst dhiameter and insert it throughthe stator'. Increasv the diameter- of the hole gageand gently lift the stator out of the housing.
1 Plac' the shaft, with the turbine wheel removed, in a plastic con-tainuir (c) to pr~oteut the c ziticai shaft sur-faces from the metal fieldy(1ki' on the magnet xzixng (,oil,
2 Subjccit the shaft nmagtivt to a unifouni field of 25 WU, in a plant- per-p'd 10ii toC thilngthl of the shaft, to full 'v magnetize the magnet.
3 R, .d-ixil iv Ili i brialie i-nituln using~ txi ýýlAiie proccduties Nowit tn ~~iiv l .ll'~atlur- lavi bei'vi ad~j~xe hall the ori'
,;1 11)., 'I roig iiisq. till) 1'Y
rG INUIRAL *I ELCTRI C
SHAFT BALANCE EQUIPMENT
The tvrboalternator shafts (Part 7) are balanced on a modified Model MV-6balancer (F'igure 78), manufactUrod by Micr'o IBalancing, Inc., in F'armingdale,New Yorik,
F'igure 78, Modified Mdudl MV-6 Bulancer'
Afte'r, review ing the part icula r btilaticlng requiremnwfts Of 111C.'It tLJobU lt(I--natntr with the~ i'pVVvslt'Itn VL'4 (tf M~icrzo Bailanicing, Inc ., thtŽ tolIlowinug modifi uat iofl were. maud to )rn)v ide iimax rn urn1 Li nba laiicv di spluc umuft inl t'atIi
plalit of. LtVox imatc ly 2. pin :
* ilalancing speed of' Oil, skaff was change'd to 6000 rpmn, which re-qu~ire aIn Uf ddi tomitl hlltr In the Niulet kT)1icti of htic, Mode'l NIV -I
s Iltiitv ditty, 1/8i -inch -wide(.. Mittr i'ecordifg tape was Lised fCm thrdrtivil hielt butlwi''il III ldithalt anid Owi liUmtcr'.
il It'l ph c(t 1 ;1-i 11I ',-, I tll- 1 1, Ill IS1, ib I , anid II\ I'm I mI i I I / I li t Il
GI IENRAL ELECTRIC
Figure 79. lBalancing Cradle Assembly
PARTS LIST FOR ASSEMBLY DRAWINGS
'T'able 22 thrcugh 24 list putris for thu 14"'K ruditil impultu turlb illveilaltot' tidll 'tol' iflflvi' 1ii1d DUtt tillfS1 giltltgimbUl US MtbliC .
GENEIPAL* I LECTRIC
"Table 22
PARTS LIST FOR SINGLE-STAGERADIAL IMPULSE TUIRBOALTERNATOR
+ i,;.li,'t*.ti 0,0. lbl•lesnjm,
N: I lt-.r.ll
A ,,40l....
I.ti,.4 ".l*.,I.-l . 411l'le Ile, IW eI ' W-n h 4141".1,I
1 I I IIi...u g .151 w 11-d- Nll -041-1.,1I I silel"t eiti" ielge kh. hellik , 4.00 Iil 0.401 1 1 , I.elele.l 'lret Illillm1,•
Ill %Ip''I 4ild~lli 3I
0 i I I',1,,h 1 14140l K1 .II I ll..,,.e 41luJl ii4fl.j
I I lir si i eit l'l,411144il I- IP1 011!11111
* I I '. i .terl 1i1i ..I N.- 643ll0 11(-1
,I.hil,,Ill.~r~re Oiieu
S" IIII.,I. .il eih~hl h I l g411411l. I
I e,. i ,11,1 I, +l ilI I 1 1114011I
h' t~,'n wIe MiiOBl I 1
I I.' 11'. 1,¥'1 "silol 4l I 'g1110411. I
1V Inter 1.11101 enlist i'i.04011 lest 1.
I 5 * ln l.ll Itin~d ll I ll i .
I. I "lleet,l.h ml~t hqilll4llll i
.14 IIn04 1 1 V.1 1 1 ....19ll-, -CI '
.e l 4 4 ... "I . 'l1 ll 1il,1., ... ....u . . . -
, il! i4 .eIIA
I p. Ihe ,Iii l illr I'1
N, ll, ll. ij~t 141, le.r ~le, I,. tleleii~nAll,.h,
I .- i..i ii1114 hOniene'ree..,lIelr ee, Iollle leLii., hil
'I lIletl ie e• h el l IllA4001.1 III I.h llle hlhle I W.01teel.I4
IIr ,Itl
hI Ca clit .~ I. 034 lI,*~e, il ll Iiilel eI el.. I; lll eel ,Il I, ,Il'i 11,
• I r..l- ei'e, le1 ,.,5 hIll~iA.I 100, .Illl~ lll l i.,Iplll
%111.1 P,.*i'llll '*1 ii! 14 ' 'lll Iceltl Ic.,/.' *'51,In e 1ftll' l Il
I. . . . . ..l , N.-lhill I0 01 i ,'.l l,l. . . .lillllIlt. l A ll lt
I'KA i i 'lpilt ili .ll~lt',::i'. il::I:......ltl
All I- i hrI !i~l tI~l ill• l 1n~ i
0ENERAL0 I LECTRIC
Table 23
PA% RTS LIST FOR INNER THRUST GIMBAL ASSEMBLY(Group No. GI)
r'PartQty Name Drawing No. /IMoscription;i No.
I I tiouming -- Inner 423D438P-1
gimbal
L 2 Thruot bearing 4231)435tP- I
1 3 Inner ginmbal ring 543C174P-1
2 4 Pivot lsrew 6641131 Il- I
2 5 Pivot screw 664H31 1 P-2
4 a Pivot stem $641131 7P- I
2 8 Set screw. hux No. 0-80 x3/32 long, cup point,socket stainliess steel
2 ) Set screw, hux No. 2-56 x I/ISlong, cuppoitnt,socket stainless steel
2 10 Coppur slug I '16 dim x 1/32 long, capper, HlI B34C
2 II Copper slug 1/32 dis x 1/32 long, cuppuvt, B111434C.
Table 24
PARTS LIST FOR OUTER THRUST GIMBAL ASSEMBLY(Group No. G1)
•t' P N, Nonle D)awing Nit. /h')acription
I I HIiuding -- outer 588kE4461- Igi m~biil
1 2 lhtumt II'MA1n4 4231)4.11511-2
I :1 Outer ginibal ring 4',43CM1l "'- I
2 4 i'Ivut icrltw 684113111'- I
2 5 I'$vot virew 864H311 I1-2
4 If PIvot Olt-Inl 6641431 171'- I
-7
.4 8 :A't-w t~1w, No. U-8U g:11:32 long, ('U 111111t.sucki l typte 304 •4ttt, ilo ta wtttIl
SOt Hut.1 W, 11'X No: 2- *• fi /x long., t'up Imint.
mio kitt 1,pl, :1114 t4tlniti ,4 .4tvuIl
'1p-wi - - _ - -•L A2 /h-i, 7TTTpll 1I:1'|
;NIEAL* ILtCTRIC
r Appendix IV
SPECIFICATIONS FOR A SIT OF SEVEN CRYOGENIC HEAT EXCHANGERS
This appendix includes a listing of the specifications that accompaniedthe purchase order tu Kinergetics Incorporated on 20 November 1972. Therewere two sets of' modifications of these specifications. The first set was senton 17 July 1974. These modifications instructed the ,endor to conduct thermaland pressure drop teats on only the warmer two balanced flow heat exchangers,instead of on all four, as had been previously stipulated, The vendor was alsoinstructed to conduct a test that would determine whether plugging of the colderheat exchangers would occur as a result of condensation of impurities pickedup by the helium gas in the warmer heat exchangers. The second set of mod-ifications was sent on 23 Dmcember 1974 and instructed the vendor to block offthe coldest heat exchanger and to cut a hole that would vent this exchanger to+he jacket.
A met of stven cryogenic heat exchangers is required for a three-stage,Claude cycle, cryogenic refrigeration system (Figure 80). The refrigerantis helium gas.
The vendor is required to design, construct, and pz'easure-drop test theheat exchanger system. Thermal performance testing by the vendor iW to bequoted as a separate option. While carrying out this work, the vendor shallbe called upon to coordinate with General 1lýectriv engineers to assuru com-patibility ot the heat exchanger Hystem with the rest nf the cryogenic r'efrig-eration system.
The heat v.xclianger system shallbe providud by the vtendor under a sub-contract with the General Electric Company, The prime. contract is botweenthe General ELectric Company and the tI. S. Army. AS L subcontractur, thehoat exchanger vendor shall 1w subject t 0 the terms aFud cnidition t)of the prim'contract.
The subcontract shall be on a fixed cost, guaranteed parformance basis.The pe;rformanvu guarantee is required for the perforoianc'e, pa tametc.rstested by the, vendor.
In the heat exchanger system, %eight is Important;t the vendor shall desipns for minimum weight. In replying to this request for proposatl the %'en-.dor shall pro\vide his e tait nate of the we•,tht and s l,,i tit' the heat e'xchiinger,based upon an initial design. It is rt,•.gnizod thai tho weight and site maychantig soinv% hat ns miti,'r dhsigln int)dfticat-.ons :art, i adf.fo' rot iti•ii atitllityWih tilt, i'4,L id tilth re ft'ilgirntihn syvst(e, 0
$IIIRA Le ILICTUI C
I ,, • Compressors mad'
Leat Rejection
iFlow .5 68 Orem/ See
Location 1
4
STurbine 3
Flow 3. 734 Gram/Sto
S6 Turbine I
.---- Flow a 1,.0 48 ram/See
Location 3
TS
St Joule-Thom on ValveFlow a 2, 10, Gram/Sec
"--IH-" Dewaowitabl, Seal
3 Hoa t Eachanger 3
,2
4 3
GINIRAL IELICTRIC
T-.
Speciflcations for the heat exchanger system and testing procedures are
given below.
THERMAL PERFORMANCE AND PRESSURE DROP
Specifications for thermal performance and pressure drop are expressedin terms of temperatures, pressures, and flow rates throughout the heat ex-changer system, as indicated in Figures 80 and 81. Note: The required ther-mal effectiveness of all balanced flow heat exchangerr is 98. 5 percent; however,the vendor shall design for 99. 0 percent, to be conservative, and the heatexchanger vendor shall be responsible for pressure drops only within theheat exchanger cores and headers. The General Electric Company shall beresponsible for pressure drops in the connecting tubing.
SPCAT4Mi Ti Loll PCMAIM)I I 11SO0 S.747 3 101110 4.406I 3 1 174 4.400a 2 6ol'6 13.84 a 201184 123.5713 3.600 .1074? A 4 1274 40400a s 10160 1,045 3 1o17 3.9109a 7 3104114 14.003 1 3.56 54.35 3 1 1.413 49.373 3 mor11n 13.84 3 4 l.157 to.9s3 5 1.245 46.73 3 6 1.139 io.733 7 .6842 55.004 1 3.885 16601 4 3 21671 155434 3 3.656 54.35 4 4 1.139 51.734 5 1.128 $53.08 4 6 1.313 163.54 7 3.671 1700.S I 8.914 335.0 S 3 I13 5 166.15 3 1.112 163.5 S 4 1.100 338.4
Note a:
1. Pressures are given in atmospheres (absolute),2. Temperatures are given in degrees Kelvin.3. Locationb I through 5 are identified in Flgurc 80.
"Location," above, is the cycle station. For ex-ample, Lucation 5-4 is the inlet to the compressor,where the pressure is 1. 100 atmospheres and thetemper•.ture is 332. 4'C.
Flirure 81. Pressure and 'T'rmperatures Throughout Systtem
STRUCT UiRE
In addition to transferring heat, the- hpet exchanger system is thiprincipal structural member of the cryogenic refr 4gerator. To minimrnivheat leaks and facilltite system r-onstruction, it is desirable not to relyMnt nxterlz l I ui)Io:'t 1 .e ruber: .', but i'L. the' to lu .pun•U the cr' .ýgt ,fiic tkmplt|Ontlriltd,
suclh as turb''iwlteriitttrit, on thu healt x ihanger4. 'riw rinal 'eig..'lno' id.i
EININALO ILICTRIC
and support of the cryogenic components shall he worked out in coordinationmeetings with engineers from the vendor's firm, the General Electric Com-pany. and the U.S. Army.
Gene ral Confighurat ion
The heat exchanger system shall be constructed either an a single unitor as two units. For example, Exchangers 1, 2, 3, and 4 could be con-structed an a single unit, with a uniform cross section, and Exchangers 5,6, and 7 could be constructed am another unit, with a different cross section.A rigid connection between the two units shall be provided, with the colderunit supported entirely by the warmer unit.
Attachment to Flange
The warm end of Heat Exchanger 7shallhe attached to a room temperatureflange, The vendor shall provide a headeL' capable of being rigidly attachedto the flange in a manner to be worked out with General Electric and U. S.Army engineers. This header is preferably the only means of support forthe entire set of heat exchangers and other cryogenic components,
Support of Components
If possible, the turboalternaturs and a radiatiorn shield shill N. supportedby the heat exchangers. The vendor shall provide rigid piping connectionsto the heat exchangers, and tabs or protrusions on the headers, where needed,to provide this support,
The total weight of each rf the three turboaltmrnators is estimated to be10 pounds. The center of gravity of the turboalternators is estimated to be5 inches from the edge of the heat txchangrrs. The weight or the radiationshield, to be supported at the warm end of Location 3 (lFigure 80), is estimatedto be between 4 and 6 pounds.
The General Electric CompLiny will select the piping wizes. with regardto pressure drop and strength, and will be responsible for miunrting the coni-ponents on the heat exchangers, The ýenorshlall hv, responsible for th, strengthof the hest exchanger and the strength of the mounting pretruthons and pipingconnections to the uxchangvrs,
Turbines and Joule-Thumson valves shall he mounted to t.i,. heat exchangerby means of demountable vacuum-type, seals. Theme seals shall permit rr-moval uf the turbine packages from the heat exhunng r during savleni develtip-merit. lalf of each seal (one flan!ge) shall be mounted direct lv on the heat ex-changer headers. The flanges mounted oni the heat extchanger heiderH shallbe the flat flange,4 (the grooved flanges will he provided It the (G;npct-al ,letric(Ceinponv anti will 1e attarhed ti thef turhine pa(ckages).
I 7fl
N I AL@ UCIH1111C
Structural Heat Leak
If the vendor' req(uiret2 external structural memb'tri for support of theheat exchanger, the turboalternatorm, or tht radilitioti shield, the heat leeksthrough theise members mliall he used to calculate a corrected thermal per-formunce of the heat exchangers. The corrected thernial performance muststill meet the thermal performance specifications described above under"Thermal P~erforma~nce and Pressure Drop."
LEA KAOE
The heut exiclitingrs dhiill In' bubble tight, from str'eLam to stream.Bubble tightness is defined on page 181 under "External L~eakage."
The hivit oxchanger muwt be luak tight to the outside, as measured by ahelium mams spvt-ronheter leek detector. Leakage ttest procedures art.specified b.'uw.
DESIGIN LlIVE
The dt-Aign life of the heat exchangerd shiill hI' 5000 liOUrs, with 100vouldown Mind w~itmup vycles reqt-,irud during refrigeratki'n system testing.T~he vendor tihall glioe ('onsiderrition to any dv~radtuinio of miaterials that mightutffvt perf ,rnvin,n during that period.
CLEAN1IANE14S~
Cleanlinoss procLdureis ihaul be observed during adsembly. Hea~t exchangerparts shaill hit d(,ro'nsed and ultirnsuniCilly UvntPfl'd before assembly.
\s selibii dhallitii hV pet' furnt md ill ai C. Itl ;k . of, Lactic, (Ieall roouii Tht,c i:Iti~d ~lsiiii hIl' Packaiged rorti hipping in it cleani heat- sealed plastic bag,
w Ith tixys nilr ogun pat pka ged inmide.
'I .'r 111(11 ;a lit-rfr ianunc te stinig OK~~I 1 Wu 11to1a 01MMe hVal to bU ltluvted de-pxiratelY bY t he vendur. If the (;t'neral Electric Compiany eloctei ill 1iiove thivrnailjiert. wiminre it-tit in' lierrorned by the . i.ndoll, thien the 0-rierx. Elu-tr ic (-'oi)-pany and the 11 S. Army will rvview tewt plans, lUet lig joel hoildS, -Alipiaritus,and instruni'ntation with the vendor and will gi%:t appruo al befori" testing bugins,
('ryogerti ItTsts
The fflur hulaiul('d fCinw,' XullIiLlors Wihlti hit, le,'tvc ild.i dIIilo fl\h the01
ht~tumi p~us, 'n'iv rl(,A rtled t,hii 111. ýI. ,rid, jid( thw~iii'jil vt'ro, iIr,.
IINIAAL* ILICTtlC
surementa shall be obtained at two Reynolds numbers above anid two Rteynolds
numbers below the design Reynolds number.
Data Extraeplation
Data for balanced flow heat exchange'ru shall be extrapolated to the a etualoperating temperature, accounting for average fin thermal conductivity, Longi-tudinal heat conduction, and average helium density, thermal conductivity,and viacosity. Heat transfer data obtained in balanced flow exchanger ttutsshall be uaedto predict unbalanced flow exchanger performance at operatingconditions.
Errors in both the experiments and the data extrapolationj must be suf-ficiently small to assure aciieving the thermal performance goals. A sim-plified analysis of errors shall be conducted by the vendor,
TESTING PROCEDURES -- PRESSURE DROP
Pressure drop testing shalI be performed by the vendor even thoughthermal tests may nut be performed by the vendor.
F'low Tests
The vendor shull test both streams of each of seven exchangers with ni-trogen gas flowing at room temperature and near atmospheric pressure.Pressure droposhrl I be measured with Reynolds numbers at two points belowLind two points abovv the detign Reynolds number.
Similar testai@ll he rmade with one stream of one heat exchanger usinghelium gam tit room tritperaiur., and neast atmosplihric pressure, t'4 n tedtuf the e-xtrulml~tjii method.
Data E x-rapnation
' rhe vvndoh)v sha I I vxtrapolute prowsurt, drlop re-ulItLt to) a tual opertttingconditions ror all exchangers, taking into accouunt average gas density darlvi a'usity.
'11INIRMAl CYCLINM;
Iefore leakage testing, all extthaiige a t hall he cyt led tit',i''e t'voomtemperature anrd liquid nitrogen temperature at Irast three' tinw t,,'. Ir' ith
:L2'lt' l111UDt blV OLCvOMnlii'lheu Ill IIc MM than i)nt, tholu'. he Yihling vinaý h :be -C41i'•l•l~lShtd I).ý 4U111.'t'igkg the exchangerst • inl liquid mll gtl l,!
NIRAL *ILICTRICt ~TE~STING P~ROC'EDURES - LHAKAUIE
Stream-to-ltruam hc-akagte shal hiii IlivasuMLId in all l.ivat ex hngerm atolc.Wihile xhlgr tatup--ueb-o I OW .0 itli 11i portsif
IPtilt, umIlluliguiI'i bhiwt ~k d , e~xce~pt fro 011V puil (111 I.- H l i4idt', tilwt.- VXCan~ile' 311011 Ihv~ p'ividurleiA .'vu tt one' 141dil w itli imliu ii gas at 20 psig. With thv other didt
-Ab~iout Ltio4d' iLprosure, tilt- leak-age gas sliha be c'ollected aind n1 iva-* AuLI .\cI " MOIdA hi 1ellti age k' \ I skll~ bu ve mas riioky rte atut im than I
10 O01t m-101111/HOV', whit-i to rr'4pttntlm tpproxi mat oly to 1iubblt ti ghtnesis,
I CU k 1 I Ithim i tt . IL It. ' llit ou ter iv inls sinli b1 e tnt'a sa:e~d bY mound3 i i fa hcoJlutn iisus mpect romilt'te hty 1nw, a sonsitivity or ait kaust 10" Lurr -lite rm/ HII.With t~lw eowhwmitirs at 1,0111 tornpt'rolurv, tilt' inwitld mhhll ho i' enotlctod
it) thte Wil'k doot 'tor and t'V aunteut * i eliuni shall lt- itt roduct d in ai ph s -
tit! hag out. s i i, Olt' t'xe~hanger to Iindicate tho t' xt it-nco or leaks, and uimy vi r'kmhlnl be lot'atod wlaLpt'n. liv 10C Ti ttst shalil ho repontoid with thv ox -Ohallgor 10l0"O~K. Aniy l'akago kirficltiml Ill till Itoak deti'etor Khntll Illrepa1 red,
A\t IteaH1t tttuk' SL1114e Ot li et-A tv?11isfuc Ivtiti (if :Wl'forit uiid b I lit tit- i
anid p romigu ii drop tt-st u shiall he \vit vsmut hy it Gonvi'flt £ EI l tric a ndI/ ora
v iff iUt' 'I itl il ' I lrtsl 'l
I.1-1A 1 1
f Prepare initial deign and review I F1 March 1972
with General Electric.
Prepare final design aind review with 14 April 1972General Electric and U. S. Army (onetrip by vendor to Ft. Belvoir, Virginia).
Fabricate heat exchnrger system. 15 June 1972
Conmplete pressure drop t&id heat 14 July 1 172transfer teats ({General Electric visitto vtandur).
Ship deliverable item,, 17 ,uly 11972
D)ELIVERAIBI.E I'TEMS
The following tt, it at" to be delivered unoler thie contracet.•'etn l)esv•,rlptio~t
I 011V (1) Set Of AM.Ven cl y CrYOt.lC htwat VXcjhutnger'S
nidvet. t3 copies of ii lettier repo~rt, mncludinig at briefdvUiic t'iJA n Of test llpt l indl w n cedu r•L e dl,,e , r'esu~lts
of heat t at iaflut o tedti n• (if p|rt•Vortll d) and pr l•' al Z'-
dlop te st IIng ( ibioth a ic t" s 't 'enlt3 :,tad extrapl -I.Ati 1a to at ua l U4t t ad4II| Is), tlo. a Iuma:1r)
(if satl'ucta a 1 c-1• ,'limtttoIal
3 T'thrue (3) k.-tM of hut.at exchattlr .Ivat.m drawings(outlilne alld dettliltld (1r'a1winI s are r'quiredI ,
I i. - ] 1 [. I I. -- I-i-- -
I
KI Ref FP-2025
: FINAL TELtINICAL REPORT
GENERAL ELECTRIC PURCHASE ORDER NO. 002-207165
November 1972 thr'ough April 1975
Prepared By
M. A. Elkan
Kinergetics Incorporated6029 Resed BoulevardTarzana, CA 91356
ItllI
g
TABLE OF CONTENTS
Section
Introductlon 1
II Exchaoner System Configuraton 1
III Stress Analysts of Port Tubes 4
IV Test Procedure and Results 9
V Summary and Conclusions 24
182
LIST OF ILLUSTRATIONS
Fi ur, page
I Heat Exchanger Section 2
2 Basic Exchanger Construction 3
3 Layout - Cryogenic RPX ,KI Drawing 3700) 5
4 Photograph of Heat Exchanger 6
5 Schematic of Heat Exchanger System 11
6 Schematic of Pressure Drop Measurement Test System 12
7 Test System Prousure Drop 13
8 Thermal Effectiveneu Test System Schematic 17
9 Helium Flow vs. Nitrogen Boil-Off 22
10 Helium Flow vs. N2 Boil-Off For G.E. and Air Force 23
LIST OF TABLES
Table Page
I Pressure Drop Measurements (Nitrogen at 700 F) 15
2 Thermal Effectiveness Test Data at Five Flow Points 16
3 Masa Flow Rate, Nitrogen BoIl-Off, Thermal 21Effectiveness
1 83
I. I NTRODUCTION
Th's report describes the test apparatus, procedures, and results of heat transfer testingand pressure drop testing of the heat exchanger fabricated to General Electric
Specification No. GL- 12913. The exehanger -ristem Is part of a three-stage Claudecycle 4ryogenic refrigeration system. The refrigerant Is helium gas.Due to problems In the thermal performance test system, the 98.5 percent designeffectiveness of the warm end, balanced section of the heat exchanger was not direct-
Sly c. monstrated In testing. The degree of vacuum required in the vacuum test chamber,to ninlmize external heat leaks during testing, could not be attained. A helium leakI. In the helium flow circuit In the chamber is suspected as the cause. The Inadequatevacuum and possible helium leak combined with an extrapolation of the cool-down flowvs. liquid nitrogen boll-off data point to a large heat leak In the exchanger test system.The pressing nature of the schedule did not leave time to pursue a cure of the problem.
The cold end balanced flow section In the exchanger was bonded In place but Is mnop-e.oive. It is sealed off from the remaining operative six sections of the heat1 exchanger. Difficulties encountered with leakage in this section during assembly ofthe heat exchanger and time limitations were factors In the decision to Install the sectionas a "dummy" in the system.
II. EXCHANGER SYSTEM CONFIGURATION
The heat exchanger system consists of a set of seven heat exchanger sections, fourbalanced-flow and three unbalanced-flow, alternately mounted in series. Each sectionIs composed of a sandwich of alternating .0075 dia. mesh copper screen parts andthermoplastic sheet parts which have been heated and presed to force the thermoplasticthrough the screen voids. The flow passages are formed by precutting holes of thedesired size, shape, and location in the thermoplastic sheets. Figure 1 is a photographof a completed section, and Figure 2 indicates the construction concept. The fusedthermoplastic forms solid walls which separate the gas flows from each other and fromthe outside atmosphere. Heat is transferred to and from the gas by passage over the wiremesh. Heat transfer between passages Is accomplished by conduction through the screenwires, which run unbroken through the plastic separating walls. At the same time, theplastic forms an effective Insulator to conduction In the longitudinal direction. FlowImbalance compensation is obtalnqd by varying the flow area In the center portion of theexchanger as discussed by Cowans I].
The exchanger system Is effectively clamped together In the longitudinal direction byan Inconel bolt which runs down the open center. This clamp greatly strengthens the
T1.Z.W.owans, ADVANCES IN CRYOGENIC ENGINEERING, VOL. 19, PlenumPress, New York (1974), p 437.
1f84
FIGURE 1 - HEAT EXCHANGER SECTION
185
COPPER SCREENS
S(PLASTI SEPARATORS
EXPLODED VIRW
SCEN STAC
FUSEDPLASIC-,FLOW PASSAGE$
PLASPASEPAATOR
CROSS STCCTION
FIGURE 2 - BASIC EXCHANGER CONSTRUCTION
I Ilt
exchanger and helps to prevent leakage at all bonded joints. The bolt is a Q, inchdiameter Inconel 718 rod, threaded or. the warm end. The flanged cold end Q.r edin a stainleu steel cap structure which transfers the load to the exchanger body. Thethreaded end carries a nut and washer which transfer loads to a set of three Bellevtlesprings. It was originally intended to use five Belleville springs, but it was not pos-sIble to start the nut on the bolt threads with the five In place because the exchanger wasslightly longer than anticipated. The next lowest odd number of springs had to be used.Them. springs are carried In the warm end manifold structure, and the loads are trans.ferred to the exchanger body through this structure.
The heat exchanger Is enclosed in a stainless steel cylindrical shell of .050 inchwall thic!4nens with a flat .25 inch thick end plate welded on one end and a flange,
with a bolt hole pattern, welded on the other end. The can is bolted to the warm endmanifold of the exchanger and uses a neoprene O-ring between the flanges for sealing.There Is a nominal .050 Inch gap between the exchanger outer surface and the insideshell wall which has been partially filled with seven one-inch wide strips, spaced alongthe length of the exchanger and bonded around .he circumference of the exchanger, ofa furry, compliant cotton material which retains its compliance Ut liquid nitrogentemperatures. This material serves to impede thermal conduction currents between the coldand warm end of the exchanger. Holes are cut In the shell to provide for the supply andreturn port flanges, which emerge from the manifold and connect to the cryogenic systemturLines and other auxiliaries. Special laterul movement bellows have been designed notonly to provide for longitudinal movement of the heat exchanger body with respect to theouter shell, but also to retain any leaking helium within the shell. Each port flange hasa bellows with one end welded to the outer shell and the other end to the flange on theend of the port tube (see detail in Figure 3). Therefore, when the heat exchanger bodyexpands or contracts longitudinally due to temperutuLe changes, the bellows de.flectlaterally.
The Belleville springs are sized und deflected to provide a .larnping load varying from15, 800 lb warm to 12,000 lb cold. The load varlation occurs because of approximate0. 126 Inch differential between the exchanger body and the bolt during cool-down,
The exchanger weighs approximately 250 lb and can be seen as a layout in Figure 3 andIna photograph In Figure 4.
I1l. STRESS ANALYSIS OF PORT TUBES
The design criterion relative to port tube strength in the heat exchanger states thateach port tube shall be able to withstand a bending torque of 170 newton-meters (125ft-lb) at the exchange, core wall, and a force In line with the pipe of 1110 newtons(250 Ib) In either direction. The port tube is Illustrated In the KI drawing package Indrawing 03711. The length of the tube from the flange end to the exchanger core Is3.965 Inches, the outer diameter Is .875 Inch, and the Inner diameter is .635 Inch.The following Is a calculation of the stress In the tube due to the bending torque applied(assuming elastic bending theory is applicable).
=- • • . . ... i • ll • - m illi.'
0406/ QIAA.~wv *PA&0~ -I >Al-a , 'I- l
t L ý%
7-i 4.IO0MU* AP6$bi
C)~ V !0& "-..
IVr
Ap .0.0 C.So qbwa
"Os' -k :.:
'ANO~ LO Z--N
'fl . --. tCis 2W 0
lBd wSC.
___ICA LA~.jGE DTAL
LIZZ
A~R' 4Aee As a- o
F -A- I
6. 011W~"M14NAT
IIre, A -I p
I0pI -. bn06 wS -0 .18 .ICAi~
- ASIA - *.. --- 401PP.* Ae.4 It; -14-1614 .t M.jw4 'All PIN UAtyp .
OW" AsS Sibif...ai
-IfAN I~ Am-0 M 1, llD"*'p
0- WI4-O UN.e n L "oto
11 "T --- !
fo i Sia 10 JaM =INA6 -,
VI -- .M 11 11 - - - - - -t 1--1& - - - - - - ;- - -
mum 0111Hz ., -rv -7
qIII1ýq_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __1_ _ _ _ _ _ _ _ _ _ _ _0
- -... 7771--
1 ti l
FIGURE 4 - PHOTOGRAPH UF HEAT EXCHANGER
I 8~'
My
where:
c = stress in tube due to bending (lb/in2
M bending torque (ft-lb)
y - distance From neutral axis to outermost wall (in)
I moment of Inertia (in4)
For this particular caoe:
M - 125 ft-lb = 1500 in-lb
y = .875/2 in
7? 4 4 1.875 4 .635 4 4I 2 r (R - R 7 " "-"= .0208 in
7 (1500 in-lb) (.4375 In) 231,561.1 lb/in2
.0208 in4
This bending stress is acceptable for the material being used, stainless steel 304, whichhas a yield stress of approximately 35,000 lb/in2 . Should the bending stress exceed35,000 lh/In0, plastic elongation may occur. Only gross plastic elongation would becritical to the design.
JI
The stress in the header pipe due to a 1110 newton (250 Ib) Force in line with the pipeis analyzed as follows.
For the case of tension or compression:
F
where:
s a stress in the pipe due to tension or compression (lb/In 2)
F - tensile forc, (Ib)
A - cross-sectlinal area of pipe (in2)
1110
for this case:
F = 250 ibR2)2 .675 2 .635 22"
A M It(R2 .R - 2 .
= 250 lb 878,34 lb/In2
The stress due to compression or tension is well below the yield stress,
For the case of buckling due to compression'2
it ElP' - f2EIELe2
where:
P E elastic buckling load (Ib)
E modulus of elasticity (lb/n 2)
I moment of Inertia (in4)
Le =effective length of the tube (in)
for this cas5:
E = 29 x 16 lb/in2
1 .0208 In4
Le*, 0.7L = 0.7(3.965 in) = 2.7755 in
PE -" 2, (29 x 106 1/in2 )_(.0208 In4) 4 7.7282 x 1o5 lb
(2.7755 In)
The force required to buckle the pipe is far in excess of 250 lb.
* The header tube Is best represented as a pinnea-fixed column In which cose the
effective length Is given as .7 L.
t11
~'Ij
IV. TEST PROCEDURE AND RESULTS
The test specifications called for four separate performance tests:
1) External leakage
2) Stream-to-stream leakage
3) Pressure drop
4) Thermal effectiveness
The following sections describe the procedures used and the results obtained.
1. External Leakage
Leaks through the exchanger shell were.measured by means of a helium mass spectro-meter having a sensitivity of 1.5 x 10 Torr liters/sec. With the exchanger at roomtemperature, the Inside was connected to a diffusion pump and leak detector andevacuated. The system was evacuated for a period of three days with the best heliumbackground reading, on the leak detector, being 4400 divisions. Bagging theexchanger assembly and filling the bag with helium gas produced no significantchange In the background reading. The calibrated helium leak of 3.7 x 10 " tdcc/sec corresponded to 80 divisions on the leak detector. Thereofre, the 4400 divi-sions would correspond to a helium leak no larger than 2.04 x 10-6 std cc/sec. It ispostulated that the high helium background reading Is due to outpassing of theexchanger, since It had been stream-to-stream leak tested with helium upon completionof assembly.
A second external leakage test was to be pertormed upon completion of the thermalperformance tests, by which time the exchanger would have experienced several cool-down cycles. The test was to involve pressurizing the exchanger with helium gas andusing a helium sniffer, connected to the leak detector, around the outside to detectleakage through the shell. This test was not performed due to the request to ship theexchanger.
2. Stream-to-Stream Leakage
Stream-to-stream leakage was measured in all heat exchangers at once, with the temp-erature below 1000 K and also at room temperature. Directly upon completion of thethermal performance test, while the exchanger system was still cold, the vacuum chamberwas opened. With positive helium pressure on the system, one of the bellows hoses wasdetached from the exchanger port and both openings were capped. Thlb physically sepa-rated the high pressure side of the exchanger from the low pressure side. The high pressureside was then pressurized to 20 psig. The low pressure side of the exchanger rose at a
192
rate of 10 psi /21 sec until equilibrium with the high p'essure side was reached. Thisloak corresponds to less than 1.8 percent of the operating helium flow rate.The same test was made, without changing the test arrangement, after the exc'langerhad reached room temperature. The high pressure side was pressurized to 20 psig andIr • valved off. The low pressure side of the exchanger showed no Increase In pressure,
although the high pressure .ide dechyed at a rate of approximately 8 p 2i/minuti.
It appears that a relatively small stream-re-stream leak, at room temperature, opensup at operating temperatur, ,. The pressure decay on the high pressure side was dueto a leak in the adjoining helium circuit outside of the exchanger. The exchangeritself was then separated from the adjoining hoses and LN2 dewar and both high andlow pressure streams were pressuie decay tested at room temperature. In each case,the stream to be tested was pressurized to 20 psig of helium gas and valved off fromthe source. Decay on both sidws was less than .2 psi/minute.
Preliminary room temperature stream-to-stream pressure decay leak tests had been per-formed on both the high and low pressure streams at the exchanger upon completion ofassembly. The low pressure side decayed at a rate of approximately .2 psi/minute andthe high pressure side decayed at a rate of approximately .12 psi/minute. Both leakrates were considered acceptable.
3. Pressure Drop
Pressure drop measurements were made on both streams on various combinations of two(balanced and unbalanced) of the six operatino exchanger sections with nitrogen gas atroom temperature. Each exchanger section could not be tested individually, exceptfor #7, due to the design of the exchanger system, which does not provide an Inlet andoutlet for each stream of each section (Figure 5). For each section combination, pressuredrops were measured at Reynolds numbers two points below and two points above thedesign Reynolds number.
The procedur, was to flow gas from a compressed nitrogen bottle into the exchangersection(s) with a pressure gauge and a laminar flow meter in the system (see Figure 6).The pressure drop through the section(s) and dnwnstream of the test system was indicatedby the inlet pressure gauge. The pressure drop through the section(s) alone was foundby subtracting the downstream test system pressure drop, at the given flow rate, from theoverall pressure drop read on the Inlet pressure gauge. Data was taken separately forpressure drop vb. flow in the downstream test system alone (Figure 7).
Pressw.-e drop measurements were made prior to the external leakage test, in which thecan was evacuated. It was decided not to make the one required pressure drop test,using helium gas, until later in the program because the Introduction of helium gasinto the exchanger would reduce sensitivity of the leak detector to be used in theexternal leakage test.
iNLET OUTLET
BALANCED SECTION #7
UNBALANCED SECTION #6
BALANCED SECTION #5
2S -- UNBALANCED SEI'NTION #4
BALANCED SECTION #3
UNBALANCED SECTION #2
3RBALANCED SECTION #11(INOPERATIVE)
S - SUPPLY PORTR - RETURN PORT
FIGURE 5 - SCHEMATIC OF HEAT EXCHANGER SYSTEM
194
PRESSUREFLOW GAUGE
REGULATOR
LAMINARFLOW METER
REULf~IATOR
* ~HEX C
FIGURE 6 - SCHEMATIC OF PRESSURE DROPmEASUREMENT TEST SYSTEM
,,....~ a b J .............
u1j
1916
N '~v w alr"*1 O.M w, ~
rhe set of pressure drop measurements performed wasn determined to be Invalid becausethe pressure drop through the testing system was much greater than that through theexchanger section(s), After It was determined that the data was Invalid, It was decidedto continue on to the external leakage test and come back to the pressure drop measure-ments after completion of all other tests. The order to stop work and ship the exchanger
* was received before the pressure drop measurements could 6o remade, Table I presentsthe data taken on the first attempt.
The following equation, taken from the Hagen-Poiseville equation for laminar flow,would be used to extrapolate pressure drop test data (N2 subscript) to actual operatingconditions ( He subscript).
where: His. ~He 4N2
-pressure drop through section(s) at operating conditions
A PN pressure drop through secticn(s) at test conditions
A P pressure drop through the test system alone
QN 2 - volumetric flow through test systm
ll He aoperating mass flow rate
0 He - average cnsilty of helium gas In section(s) under operatingconditions
~He - overage viscobity of helium gas In section(s) under operatingconditions
ON2 - average viscosity of nitrogen gas In section(s) under testconditions
*4. Thermal Performance Test
The raw test data for tle wairmest balanced section Is presented In Table 2, IncludeidIs the system flow rate, pressure, and nitrogen boil-off flo~w rate, This datciwas taken
* with a vacuum, In the test chamber, varying from 8Sx 10O' Tore to 25 x 10 ' Tarr,The cool-down was started with a vacuum of 3 x 10-4 Torr, but this vacuum rose tothe above range after 1-1/2 hours Into the cool-own, Apparently, a leak In the systemopened up at this point. A schematic of the testing system Is shown In Figure 8.
197
ro4P
198
Helium Flow System (1) NitrogenReading Preutres Boll-Off (2)
(In. H 2 0) (psi) (% of 4)
2.50 14.6/11.2 43
3.20 14.6/10.2 50
3.53 14.3/ 9.7 48
3.9 14.4/ 9.2 52
4.35 14.1/ 8.7 52
1090 K at #1 supply port, 20 x I0"3 Tarr vacuum.
Equipment.
Flow Meters - F4 Fisher and Porter Flowrator Model 10A27350,tube No. FP-3/4-27-G-10i80
PressureGouges - P29, U. S. Gauge OU-2579-A
P47, U. S. Gauge AW-2926AB01
VacuumGauges - VAC3, CVC Type 0 PH-100C
VAC4, Thermocouple vacuum gauge NRC 802-A
Thermocouple
Bridge - Ti, Leeds and Northrup, Model 8693
(1) See Figure 8 for location of presure gauges.(2) Flowmeter R4, maximum flow 11.0 scfm air @ STP.
Table 2 Thermal EFfetlvenoess Test Dataat Five Flow Points
199I
N2 BOIL-OFF
N2 BOIL-OFF
VACUUMPUMP IIRLS
I NSULATI ON
TC
TC TC2 DLOWART:3LTCR
FLOW FLESfREGULATOR HPUPPUMP BrY-PASSVALVE SUE MK-P PVALVE
BY-PASS VALVE TC.- THERMOCOUPLE LOCATION
FIGURE e THERMAL EFFECTIVENESS TEST SYSTEM SCHEMYATIC
200
To obtain exchanger thermal effectiveness, the data is reduced as follows:
1) The Moriam flowmeter curves are entered at the appropriatedifferential pressure to obtain the corresponding scfm of air.This value Is then corrected for helium viscosity (x .912) andmultiplied by the density of helium, calculated at 100 I andthe system outlet pressure, to obt,,in the helium mass flow rate.
2) The N2 flowmeter readings are multiplied by 11 scfm (the maxi-mum flow rate) and the density of nitrogen at room condtlonsto obtain the N2 boil-off mass flow rate.
3) The temperature difference across the gas flows (ST) is thenfound by comparing the heat capacity of the helium flow tothe N2 mass flow times the latent heat of LN2, I .e r
1 N2 X N2 fHe CPHe 6T
orh •N 2 XN2
6T7 1lHe CpHe
4) The effectiveness is calculated by the equation:
6 T
AT + 67T
where 4 T is the temperature difference between the gas enteringthe warm end and the gas leaving the cold end.
A sample calculation of the data reduction procedure is shown on the following page.Table 3 shows the reduced data for the five flow points observed during the test.
The calculated effectiveness of the exchanger, extrapolated to operation conditions, isalmost 93 percent. The design effectiveness was 99 percent and the required effective-mess Is 98.5 percent. A linear extrapolation of the five helium flow vs. nitrogen boil-off points shows a 32 pergerit boll-off at zero flow (Figure 9). This, along with thepoor vacuum of 25 x 10"4 Tnar under which the test was run, is Indicative of a largeheat leak in the testing system. Correction of the boll-off data for the maximum flowpoint recorded (4.35 Inches H20 @ 52 percent boll-off) gives an efficiency of 97.11percent. It is believed that the linear extrapolation is conseevatlve and that a con-cave curve, which opens upward, would be a closer approximation. This belief Isgiven substance upon making reference to a graph of helium flow vs. nitrogen boll-offfor a similar high efficiency heat exchanger constructed several years ago by KI forGeneral Electric (G.E. P.O. 002-206289) (Figure 10).
•U1
Effectiveness Data Reduction(typical)
Data Point 1:
He flow reoding 2.5 Inches H20, p - 14.6 pIig
N2 Flow reading 43 pervcent
Helium Mau Flow
Qafr(scfm) " 2.5 10 " 31.25
viscosity correctlon factor Is .912
for o " 1."x10-4poai
Io.r , 1.80 x 10-4 polso
QOme(scfm) = .912 x 31.25 a 28.5
"H28.5 ft (1. 11 869 xo10 -5 Ib4
m__ _ _ _ _ _ _ _ _ _ _ _ _ __in 7 .In3
H OHO •.ON. ilx 7., 1728 1440.__a ft3 \60 se 7m rn
29.3'bHe - 4.168 ram/see 'He " g38.3 x 3633"t x-
Nitrogen Mass Flow
QN2 (scfm) a .43 (11.0 scfm) - 4.7360 454
A% N2 QN2 IN2 4 r .0 f tbj We
14.7 x144i N2 - 2.695m/Wic ON2 3,
202
EFfectiveness Data Reduction
(continued)
ST and A T+Ho CpST thN 2 AhN 2
1 'N2 Lh N26T T
rh He CPcal watt sec
AhN2 , 1310 Ca 195.83 w- sacAN =M4l gm
C -1.25 BTU - 2.9042 watt sec
6T a 2.695 x 195.83 . 43.59 F
ST " 11S F - (-2600 F) a 378 F
Effectiveness:
:• ~AT +g
378 ,8966s + 435
t He rhN2
Helium Maus Flow Nitrogen Mass Flow Effectiveness
(gm/sea) (gin/,,c) (%)
4.1687 2.69" 89.66
5.3360 3.134 90.82
5.6753 3.008 91.80
6.2300 3.259 92.09
6.7700 3.259 92.826
Table 3 - Mass Flow Rate, Nitrogen Soll-Off, Thermal Effectiveness
204
9,9 K-I
al.
205
C
LA
I2
C!
z 66
w -@
66
(im {f1Ij JO %) ONIOVIM •3IL-3W M~li •
206
V. SUMMARY AND CONCLUSIONS
Problemn with the thermal testing system and time limitations were the main causes ofthe apparently Inconclusive dita on the performance of the heat exchanger. It Isbelieved that the 93 percent effectiveness oa ihe exchanger, which was determinedfrom the one thermal .erformance test made, is due to leakage within the vacuumchamber, which can introduce severe heat looks, and Is not Inherent In the head
exchanger itself. Extrapolation of the Helium flow vs. Nitrogen boil-off data pointslend credance to this view. External leakage tests revealed no gross vacuum leaks,but the tests were made at room temperature and after the system had been exposedto helium gas. The background Indicated by the helium look detector was relatively
Stigh, due to heNum and outgassing In the heat exchanger; this prohibited detectionSof medium and smelt looks. Stream-to-,trom lockage was small enough to have only.*minimal effects on the performance.
Future testing of the refrigeration syStem by General Electric will hopefully revealthe effectiveness which the heat exchanger Is capable of attaining. Much knowledge"has been gained In the design and construction of compact, high effectiveness,cryogenic heat exchangers through this program. It is believed that the state-of-the-art has been advanced by this heat exchanger.
207
JENIRALO ILIOTIHIC
Appenodix V1
HEAT EXCHANOER ANALYSIS
RELAIINO RESLTS TO RESION COMMONTo simplify testing, the data presented in this report were taken at tem-
peratures slightly lower than the design condition temperatures. The datawere taken at Reynolds numbers above and below the design conditionReynolds numbers.
HEAT TRANSFER DATA
The flow in the heat exchanger is in the laminar regime, but becauseeach layer of screen must form a new boundary layer, the heat exchanger hasdeveloping laminar flow. The following correlation approximates the curvesfor stacked screen@ at low Reynolds numbers (Ref. 7, pp. 129-130):
St a Prr"O" Re"le * NTU As/Af
where:
St h/Oc
Pr uc/K
Re x GD/KK - Thermal conductivity
u - Viscosity a 5. 023 x 10"0 x TI-11 7 g/cm/sec
c a Specific heat
h - Convective heat transfer coefficient
G a Mass flux
As a Heat transfer surface area
Af = Flow cross section
D - Hydraulic diameter
NTU - Number of Transfer Units
The Prandtl number is a weak 1,inction of the absolute temperature, sothe temperature dependence of the convective heat transfer coefficient isdetermined primarily by the viscosity. The following proportionality canthen be used to scale heat transfer data taken at conditions other than thedesign Monditionsu
NTJ a TO-msa/O."a
Precedlng page blank 209
I ONEINAL* ILITCYNIC
PRESSURE DROP DATA
Because the flow is laminar, the friction factor is proportional to the in-verse of the Reynolds number:
fa I/ReAP a Ou/ p
AP a OT ''G47
where:
Re m GD/u
f - &P/4(L/D) (G*/2p Af')
AP - Pressure drop
L a Length
P Density
The pressure drop is therefure generated at the warm end of the heatexchanger. The pressure drop can be scaled to design conditions by theratio of the warm end temperatures raised to the 1. 647th power.
DATA WDUCfON COMPURTR PROORAM
ThL thermocouple measurements in millivolts, the pressure drop acrossthe nozzle flowmeter, the pressure at the entrance of the nozzle, and thebarometric pressure are inputs to a computer program that calculates themass flow, temperatures, effectiveness, and NTU for each data point. Thefollowing paragraphs list the program and explain the variables. A samplerun is also presented, which represents the initial data point in Table 8,
The program was originally designed to include the thermocouple meas-urements 6T1 and AT2 in Figure 35. Because of the shorting of themethermocouples, the program was altered to ignore these readings, In thesample run, these thermocouples read zero.
Following is the nomenclature for Program HEDRI:
I Symbol and Explanation
170 TEST - test number; MO, DAY, YR n month, day, and year ofof the test.
180 MV1 .... MV6 n thermocouple millivolt readings (numbers 1, 2,3, and 4 correspond to the thermocouples in Figure 36. Number5 is the difference couple labeled A TI; number 6 is the differencecouple labeled AT2.
210
#NIIALO ILItTRII
Line Symbol and Explanation
190 BAROM - barometric pressure in mm of mercury; PI1N u pressureupstream of flow nozzle (paig): DPIN u Pressure drop across flownozzle (In. water).
197 MVNOZ • thermocouple millivolt reading at the flow nozzle.
200 DI n diameter of pipe upstream of flow nozzle (cm);D2 - diameter of flow nozzle throat (cm).
210 M - molecular weight of gas (g/mole).
220 CP - specific heat of gas (J/gK).
230 MU - gas viscosity (poise).
240 TTAB - number of entries in temperature-millivolt table.
250 DTTAB - number of entries in temperature difference-difference-couple millivolt table.
550 DTLAV a lengthwise average temperature difference between heatexchanger. (This average is based upon four differences: thedifferences between absolute thermocouple readings at the two ends,
and the difference couple readings at the two ends).
590 P1 absolute pressure upstream of flow nozzle (bar).
600 DP - flow nozzle pressure drop (dynes/cm').
610 RHO w gas density upstream flow nozzle (g/cmr).
620 AHSQ - area ratio squared.
630 A2 a area of flow nozzle throat.
640 MI)OT - mass flow rate through nozzle, assuming nozzle coefficient
650 NHE x Reynolds number, based on the mass flow rate.
660 CV = nozzle coofficient, based upon curve fit of American Societyof Mechanical Engineers data.
670 MIw'P - new mass flow rate, based upon nozzle coefficient.
700 EF1 i thermal effectiveness.
710 NTUA - apparent NTU.
720 11T r heat transferred from stream to stream.
Figure 82 is a sample data run, which represents the initial data point.
211
M*IN AL ILtETRI C
100C -----HtAT EXCH4 DATA IRE0IJCU9N----101110OC ----- N'3ZZLt FLOW MCTEtq-----
1900 -.---- HtF*JUNC1grN3 IA4 AT 273.15 I( AND M4 AT 77.35 k-mem-130 INTEGE TtSlDM~etDAYsY~sIsTTA9#DITA9135 REAL MVIMV2.MV3hMV4,MV5,MV6,ýMVNSZP11N4aMUsNRtaMDOTNIUAeMVA140 DIMENMISN MVA(AO),,TAce0),DTA(10).DMWVA(I0)150 DATA (MVA(1). u.8"0049.I9408EO.I..5..91514 1 .S4lDI.SO4Dl.3 65DI.9 73s3 .6 gla4B3IOpS.O4DS.SID6.62I.o7.d78B154 DAIA (TACI). IuItlS)ý77.,35.SO*e90.al0O..I0.lI0..O~l30.,Id0..IS%# ISO.. lAO.. t8O..200..2R0..1A0..260..980..30O.,3R0#.*~157 DATA (DTA(I),IutS)V77.. t0oo.. 10..lSO..2O.90260..300..340.,158 DATA C0D1VA(l).1mI.S)/60..Sl .S.40,5e34.,29.U,26b7,14,.Uogeti159160C ---- NPUTS...170 RCADo 1EST#M1#D0AY.Yvt1SO HEAD# MVI#MVR#MV3#MV4#MV5#MYA190 I4EAD* BAN6MPPIIN.DPIN195 MVIwMVI*5.54631 MV4vMV4*5*5463196 PItNNmtPJI*2*306197 MwVNOZmMYA200 DivAs351 D~w2*0282RIO M44eto cpassoe30 MUw20OPK-6240 TTAR.ISl250 DTTAP~aR2511i0('C ----- PRINT INPUTS--fte-
330 PRtINT,*DAT~i ,sMils"t/'DAY #'PsYA340 PKINTs"MILLI VOLT AtA[31146, 1#90#0#.5#61'50o P~i4ITsmvi.mve*MV3sMV4,%MV~sMV6360 PRNINT,'NOULE TtMnP (94V~w"sMVNQZ170 PR1NT#'flAROM PHES5 (MM~rP)w"s.RAKOM300 PR14Ts"P)ýSS UJPSTHKAM NOZ?,Llr (IN tah1J390 PRIN1D"NOZZLE. DXLTA P (14.1 HRO)mu.OPIN400 PI4INeTa"NOZZLE DIA (CMi (Ie"v~is" (2)nu.D2410 PRINTs'MOL WT (AO'MSLt)s"vM4B(0 PHINTs"Sp'C 14T (J'0 R)w'.CP430 PHI4N3.VISC (PSISKInm'MUA31d~ir . ..... IKMPERATURES .....m46nl CALL I-NitRP(MVNeEDTNSý,MVATAD11IAR)A70 CALL IN1tRPCMVlTl.sMVAs1A*,1A8)49O CALL I1C14ItH(MV9i.TBDMVAsTAs1'TAPý'9fl CALL t'I.41KP(MV3s13sIMVA*lAs1TAM)SOO CALL IN~tHP(MV4PTA.MVA#TA*TTAP)SIO CALL I~ltPc.5ecT9.13.#DTflM.D1A.DMVA.D11AP)
F1gure 82. Sample Data Run (Sheet 1 of 3)
212
~IN IA LILI CT I C
904r~1)1 D]Uim) I 11rilioMVis
s A
'190 Plw(Pvq*9'1.4+MAmmmM17%i.
if APO flPrTS 'io
A10 APs o7f,*ADa*L0'0440 MDAsJ'mCI'~,)* (9 4 A ,I
A SO wCamp i o (. 7PQ09~ij
A700 mr,, or('1 weil i~~
7p1
710 14i1 1Jj." Flo # I . l'Pr r"
790 I NCI*. 1 I i *I I 1",ILLS P- a' -
770 tPCM ,-A I) 0p 1 rD90
990 Po9 "1 P( A!'"sA 1) )9.9.
1010 9R NY1 F1- KA~ IAI)F VALF-~( -:W41*V Y~I1~( -l)'l'1~~
I n Pi-s Ta \ii1 , ".NJ ' I , II -toorAfF P~ I P AJ "
1% n~rr 82.11m lr Iat Nu 1N r' 2 ol 3)16v10s
M ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ 2 4K1PmINIa":)!, kFl
M IAEAL@0ILESTAIO
HEDRI 08118/75
i0A(0 STOPI OHO F.N 0
HEDRI 15116 08118t75
* I85lPsO.75,67.0.243S.O.6l 73.0O.OOO7750.07#1 .5#9.7
DAMF 7/ 2. 75MILLI VOLT RKADINOSU I,2&3#4s5#A
A,7315000E+O0 7.56700001-01 e*4380000t101 6.1A360001*000. 0
NIZZLC TEMP CMV)s 0;.11636000t.600NAROM PRESS CMMHG)a 7*5007000[*02P~tSS UPSTREAM NOZZLE (IN 403. 3o45900009+00NOZZLE DELTA P (IN. HRO)u 9e7000000E+00NOY/.XLK DIA (CM): M@l 60350000OR400 ME~) 240282000E.0OMOL WT COMSOLE)aSPEC NT (JIG X0. 562000000C4O0VISC (POISUE) 20O0000000t-04111TPUTT1. 3.0259695[+OP. DEG x(T~w 1.1715610E+02T38 9.81319 169t*0lTim 9.8U7200t*01?0T!.MPERA1II)4E DIF!H!A.C!PSrIP~r (C'PLt So 0. TPuT3o P95SA392Bkt+0IDIFF ClULJL 6. 0. TI-T4w 1&3870~5O+01PP~m 9*0600l103o-olAPPAN. NTIJs 9 04A35012E# 00'jHU.0 ~3*05711999+04 *41ZZLR CtO3Tu 9,AOF44a47r.-O1 04 FRsu 20004N4CE46Q000MASS F~LOW (0.ISKC)w 9.3i8296AE400HKAT TN 'JS (WA71) 9,31531e0t+03
PROGRAM STOP AT 930
Figure 82. Sample Data Run (Sheet 3 of 3)
214
NINIALO ILIOTRIC
"FFECT OF HEUUM LAKAGE, THERMAL RADIATION, AND AXIAL CONDUCTION
STREAM-TO -STREAM LEAKAGE
If radiation and conduction are ignored and if there is no leakage to thecaning, the control system shown in Figure 83 can reprepent heat exchanger7.
- - - - M
MM~14 'rjMT I ,M-M
-- ' •-.. ... - L-.
Figure 83. Stream-to-Stream Leakage
r In Figre 83:
Mi s Strearn-to-stream leakage
M a Heat exchanger mass flow
T, w Temperature of the gas from the compreshor
T2 a Temperature ot the high pressure gas leaving the heatexchanger
T., - Temperature of the gas returning rrom the nitrogen dewar
T,6 Temperature or the low pressure gas leaving the heatexchanger
C Specific heat of the gas
(M-ML)C(Ta-T:,) + MC(TI-Ts) U 0
ML -
LEAKAGE TO CASING
The casing of tho heat exchanger- 1I connected to the warm end of thelow pressure stream. A leak to the casing therefore bypasses thr low pres-sure stream'. Assuming the radiation and conduction are negligible andassuming there is no stream-to-stream leakage, the control system shownin Figure 84 represents the heat exchanger.
215
,,. SIilIAL*ItL6TlEI¢
T1 T aIIlll -
MMTo
I MLT#b. T41 To
Figure 84. Leakage to Casing
In Figure 84:
T.' - Low pressure stream temperature before mixing
ML a Leakage to the casing
Ts. To, T3, T6 - Temperatures as defined in Figure 83
An energy balance for the above control system yields the followingrelationship between the mass flows and temperatures:
MLCTg + (M-ML)CT4u • MCT 4
M L /M • (T6' - TI)/(T4 ' -To)
The following approximation can be made to simplify the above equation.
,T6' < T%
This approximation approaches an equality as the effectiveness of theheat exchanger (without a leak) increases and as the size of the leak increases.This equation ezqentially notes the fact that a heat exchanger with unbalancedflow will have a pinch at one end. This approximation can be combined withthe previous equation to yield .ha following equation for calculation of theleakage to the casing of the heat exchanger:
MLL < (TL - T4)/(T 1 - To)
THERMAL 1ADIATION AND AXIAL CONDUCTION
The control system in Figure 65 shows the effect of thermal radiationand axial conduction on the performance of the heat exchanger. It is assumedthat the leakage in the heat exchanger can be neglected.
216
IN I A ELOSICTRIIC
M IM
M -
SFigure 85. Radiation and Conduction
where:
.QR a Hadiation to the heat exchangerQc z Conduction down the heat exchanger
Tt, T•i, T3 , T 4 x Temperatures as defined in Figure 83
The following energy balance can then be written to show the effect orthermal radiation and axial conduction on the temperatures in the heate'zchange .:
M(U(T-T 4 ) + MC'('3'1-2) + Q4 Q+ W 0
S... .. Q ('r -' 3 ) "MC
Thermnl raclintlon and axial Londuction aipponr as n ciir, renc, in the torn-perature dirr, once: at the warm and old ends of the heat exchanger. If' Itwere not rot, the leakage in the hoPt oxchnngvr, this would irovide a methodror exact calculation of the thermnal rndiation and axial conduction. the lastdata point in i'anle 8 has the least aniount of lakag, b'catist, the axial comi-pression of thr heat exchanger had hen incr•nsed and the pressure differencebetween Ftr,.nms is small. The diff rence in the stream-to-stroam tompe,.atures at eithutr end or the heat ('xchanger for this vase is I. 4"K, rhis isin part cauFed by leakage, but It does provide an upper bound on tht! f•ect ofradiation and i-onduction. This would decreasr the effectivenoss or the heat
exchanger by 0. 35 peru:ent.
217
... ... ..
jNIIAL@ItIETRIO
Appendix V11
1. D. B. Colyer and W. H. Oney, High-Speed Cryogenic Alternator Develop-mnPhase I F~inal Report, U.S. Army MoJ'ilty Ecauirment Research
aid Development Center Contract No. DAAKO2-88-C-0320. Fort Belvoir,Va.,* July 1970.
2. R. B. Fleming, RK. Terbusb, and D.E. Colyei, -- Ipment of a SingloStage Cryogenic Turborefrigerator, Phase 11 Final-Report, U.S. ArmyMobility Equipment Research and Development Center Contract No.DAAKO2-88-C-0320, Fort Belvoir, Va., March 1972.
* ~ 3. 1). B. Colyer et al. . Design and Develo'pment of Cryogenic Turbo Refriger-ator Systems, Phase A Final Report, Report No. AF1FDL-TH-71-117,Air Force Flight Dynamics Laboratory Contract No. F33615-71-C-1003,Air ForieO Flight Dynamics L'iboratory, Wright -Patterson Air Force
* Base, Ohio, December 1971.
4.~ ~ ~ ~~1 1)-.9le cta. an and Development of Cryogenic Turbo Refriger-sktor Systems, Phase 1M inal Report, Report No. AlIFI)L-'rR-72-154,
Ai orce Flight Dynamics Laborutory Contract No. F93815-71-C-1003,Air Force Flight flynamicir Laboratory, Wright -Patterson Air Force Base,Ohio, April 1973.
5. D. B. Colyr'r (it al. , Design and Developmenit of Cryogenic rurbo Refriger-ator Sstems Phase C Final Report, Report No, AFFDL-74-93, AirForce Plight Dynamics Laboratory Contract No. F33815-71-C-1003, AirForce Flight DynamnicR Laboratory, Wright -Patterson Air Force Base,Ohiu. April 1974,
6.P.C0. Wapato, Investigatiorn of Thermodynamic Cycles and Components forTuro-ryp Rfrigerators, Air~cosearch Manufacturing Company Draft,
U3.S. Air Foret Contrnct No. h'336l5-60-C-l540, Wright-Pntterson AirF ~Force BflBL% Ohio, August 1970.
7. W. M. Kays nd A. L. London, Compact Beoat Exchangers, McGraw-HillBook Company,, Now York, N, Y., 1D64.
8. R. Mcl-ce(, "Optimum Input Leadm for Cryogentc Apparatus," The ReviewF of Scientific Instruments, Vol. 3O, No. 2, February 10BO, pp. 98-102.
Preceding Pill blink
211)
