Studies of aerothermal loads generated in regions of shock/shock interaction in hypersonic flow

f

,i ,qt._

4,-NASA

Contractor

Report

181893

STUDIES

OF AEROTHERMAL

LOADS

GENERATED

IN REGIONS

OF SHOCK/SHOCK

INTERACTION

IN HYPERSONIC

FLOW

(NASA-CR-181893) STUDIES OF AEROTHERMAL

LOADS GENERATED IN REGIONS OF SHOCK/SHOCK

INTERACTION IN HYPERSONIC FLOW Fina| Report

(Calspan-3uffalo Univ. Research Center)

339 p CSCL 20D G3/36

N92-I1319

Unc] as

0051147

Michael

S. Holden,

John

R. Moselle,

and

Jinho

Lee

CALSPAN-UB

RESEARCH

CENTER

Buffalo,

New York

Contract

NAS1-17721

October

1991

IW A

National Aeronautics and Space Administration

Langley Research Center

Hampton, Virginia 23665-5225

ABSTRACT

Experimental studies have been conducted to examine the aerothermal characteristics of

shock/shock/boundary layer interaction regions generated by single and multiple incident shocks. An

extensive review made of the literature on this subject showed that there was significant lack of detailed

high-quality experimental data in the high Mach number and Reynolds number flow regime. The

experimental studies presented here were conducted over a Mach number range from 6 to 19 for a range

of Reynolds numbers to obtain both laminar and turbulent interaction regions. Detailed heat transfer and

pressure measurements were made for a range of interaction types and incident shock strengths over a

transverse cylinder, with emphasis on the types Ill and IV interaction regions. These measurements

indicated that the peak heat transfer and pressure increased with the occurrence of transition in the shear

layer generated in both type 111 and type 1V interactions, and with increasing Math number. For some

type 1V interactions with flowfield configurations close to those for maximum heating, a flow instability

was observed which caused large temporal variations in tide peak heating. For completely laminar

interactions in high Mach number, low Reynolds number flows, the structure of the type IV flowfield and

the resulting heat transfer and pressure levels appear to be strongly influenced by viscous effects in the

shear layers. The measurements made in this study were compared with the simple Edney, Keyes and

Hains models for a range of interaction configurations and freestream conditions. For interactions where

transition occurred in the shear layer, the peak pressures were in general agreement with the predictions

for types III and IV interactions; however, the predictions employing laminar stagnation-point heating

rates for type IV interactions must be enhanced to account for radiated noise generated in the transitional

shear layer. For fully laminar flow conditions in a type 11I interaction, experimental results were in good

agreement with the simple predictions; however, the pressures and heating rates in a type IV interaction

were significantly overpredicted if multiple compressions were assumed in the inviscid jet. The effect of

sweeping the interaction is to lower the heating and pressure levels in roughly the same proportions to the

reductions observed when sweeping a cylinder in the absence of the interaction. The studies of

multiple-shock interaction demonstrated that the largest heat loads are generated on tide cylinder if the

shocks coalesce before they are incident on the cylinder. The complex flowfields and aerothermal loads

generated by multiple-shock impingement, while not generating as large peak loads, provide important

test cases for code prediction. It will be difficult to accurately predict the maximum heating in

shock/shock-interaction regions over a large and important part of the flight regime, because

free-shear-layer transition can take place at low Reynolds numbers, in either single- or

multiple-shock/shock interactions, and because of the occurrence of flow instabilities for type IV

interactions. The detailed heat transfer and pressure measurements presented in this report provide a

good basis for evaluating the accuracy of simple prediction methods and detailed numerical solutions for

NOMENC i ,ATURE A, H,N (l C+ Ch, St Cp CI, C c(7') D d Fshape

f

II h k(7),h', K M ,_.I_ ,,_.l: 01. l' I'r I rallo P pD r

(2, ,1(q)

q/'h'. QPR R R

Constant Used itt Pcak-Ilcating CorrclatiCm

Speed of Sotllld

O,.//,

.)'r,../'r.

: q/(o.,.U,,.(ll,,- llw)), Stanton Number

Pressure Coclficient, i_.qualion 4.9

Specific ih:at at (?t+)rlsli|llt Pressure

Specific Ilcat; ('hord I.englh, Figure 42

Specific Ileal, Appendix A

Cylinder 13iametcr

Ilcal-Penctratiotl l)cplh of Stfl_strate, Appendix A

Shape Factor 1/2 [N _)/Z, where A]': 1,2

Charactcrislic I:requency (llz) Total I;]olh;!ll'>y

= q/(7'+;w - 7"w), I lcat Transfer ('oelficienl

Thermal Conduclivily

/,eng(h of S/it_';+lr 1.ayer, /7i_,jrre 2

Macll Ntlmber

ltlcident Math Ntttnbe+

Con,,ective Nlach Number -- (U+_- U:,)/a:, l:it_,ure 4

S|lock-(.}t'lqct;l/Of Angle, Table I

PIessllrc; Separation l'(fint, lq/?.ure 10

Pr:lndtl NItml_cr

Non-l)imensional Pressure Ratio

Sl,llic Pressure

[]t'Klisttnbcd Sla[;naliorl Ptcssllre

Iicat Transfer (Rate)

Dytmmic Pressure

Fay-I_iddt, II llcalTtansler, Rulcrcnce +1¢)

Reattachmctd I_oint, I:igtHe 19

G,IS (_+OIISlalll (+ 1717.91 l-t-lb/shLt,,/°l_)

iii

NOMENCI,ATURE (conl.) Re ReD Rc,I Red. R I, IT r SL StA s T TS / _/tn_ix U,V U I¢ V X. x Y.y

= Ut/v, Reynolds Nund_er

= O,,,tLD/p.,,,

Attachment -I.i,m l,k.ynolds NI,mhcr, I'klualion 5.7a

Rererence Allacluncnl-lJne Reyn()l(Is Number, Equation 5.101)

Leading-l';dge Body Radius, Figure 11

hnplicit Marching-StCl) Size, Appendix A

Shear l.ayCr, Fit,,ure I

Altachmcnl--l.irle ,gtanton Nulnber I.al)lacc Variable, Appendix A

Temperature (°R)

Transmitled Shock, Figure I

Time; "l'hickncss of lllunt I.eading Edge or Nose

I.on!,,ih of Tinlc, for which ,qemi-infinilc <qolid Assuml)lion is Valid, Appendix A

Velocity

Widlh of Ru'allachlilelll Region

Cli;ir,lclerislic Vch_cily, Vitture 42

Nornlal I-)i.<,lance Mc;istucd in to Surface, Body Surface Coordin;ite

Normal I)i,',l;ilice Mc,;isured Away Ilonl SJurface

GREEK SYMBOI.S ii 1' A Ag 3 E : k('l)l(ec(7)) , Ai)pcndix A I_catlachnlcnl l\ilgle

Shock and lifilial <qllear--I.,iyor Angles, I;igilrc 15

Ih-es_,urc--(-irac]iOlil Pafanleter (= 0 lo 2)

Specific ileal Ratio (_1.4, Ideal Gas)

Bow-Shock Standoff Distance> [:iglll-O 0b

Width of Supersonic Jet, Figure 3 Bound,Jry l.ayer Thickness

Width of Shear I.ayer, I;igul-e 2

0'-I)10'+ i)

I_;ll-anit, lt, r I!socl hi l:.qilalion 5.5

NOMENCI,ATURE (cont.) /] 0 )-K 2 2, A #+ v 0 T

4,

to S U BSCRI PTS + 0 I 1,2,3,4,5, 6,7,8 A aw, AW I_,II D v,' FR, F-R IDEA I, I,A k,l I,E

Ntm-Dimensi_mal I)i._lanCe, Appendix A

Noli-Dinlensi_mal I.elw, Ih Scale, EclualioH 5.7b

Alll,>ul;ir Po._iliorl tin Itle ('ylillder Mo(lcl

Anl_i)li;.ir Po,_ilioii (ill Ih¢ (]yliiltlor Mllclt!l, [gp,urc 12

= /),l/(O,,, {/21)

l)ummy Variable, Al)l'_eIMix A

Swept Angle

MocIU Sweep Angle, Tilblc I

Yiscosiiy

Kinematic Viscosily

KircholT Tlanslormalit)li V;iriablc, Apl_Cildi× A

Densiiy

Non-Dimcnsi{mal G'ime

Orbilal Angle of Allack, Fit,.ure 13

Radial Frequency

Re ft21'orlce V;ihle

Undislcirbed V;i h IC"

lilili:ll Valtl_2

Ri2giori,_ ill 'l'ypo III ;iiltl "l+ype IV illlc'l';It:liC)ll._ _lS ilhisiratod by Figures 2 and 12a(o) AI Atlachillt, iil I.hlt,

Aciiab;ilic \V_lll Vlilllt"

At Body Diameter

Edge Condil ion

Fay-Riddcll Values, Reference 49

I¢lc;il V;ihlc

lllcidcill

I,a milla r- Fhlw V;ihlo

Nor_:II{NCI,ATU RE (cont.) max peak R Real ref, REF SL S TA G, stag s w, W _, INF

Maximtnn Peak Value

Peak Value

At Rcattachn_cnt Region, Figure 19

Real V;|lue Reference Value Shcar-l,ayer Value At Slagr_alion Region Shear-l.aycr Value Wall Value Freestrcam V,,luc vi

Table of Contents

Section Page

ABSTRACT ... i

NOMENCI.ATURE ... iii

I INTRODUCTION ... 1

2 REVIEW OF EARI.IER STUDIES ... 6

3 EXPERIMENTAL PROGRAM ... 29

3.1 PROGRAM OBJECTIVE AND DESIGN ... 29

3.2 EXPERIMENTAl. FACILITIES, MODELS, INSTRUMENTATION, AND FLOW VISUAI,IZATION ... 30

3.2. I Experimental Facilities ... 30

3.2.2 Models ... 32

3.2.3 Iieat Trans[er Instrumentation ... 37

3.2.4 Pressure Instrumentation ... 37

3.2.5 Flow Visualization ... 37

4 T U_'ST PROCEDURES ... 38

4.1 INTRODUCTION ... 38

4.2 EVALUATION OF STAGNATION AND FREESTREAM TEST CONDITIONS ... 38

4.3 REDUCTION OF MEASUREMENTS OF PRESSURE AND HEAT TRANSFER ... 42

4.3. I Measurement-Time Selection ... 42

4.3.2 Pressure Measurements ... 43

4.3.3 Ileal Transfer Measurements ... 43

4.3.4 Measurement Recording System ... 44

5 RESULTS AND DISCUSSION ... 45

5. I 1NTROI)UCTION ... ' ... ... 45

5.2 MEASItREMENT OF tlEAT TRANSFER AND PRESSURE DISTRIBUTIONS ON A CYIANI)ER WITIIOUT INCIDENT SIIOCK ... 45

5.3 STUDIES OF MACII NUMBER AND REYNOLDS NUMBER EFFECTS ON SINGLE-S! IOCK/BOW-SI lOCK INTERACTION ... 48

5.3. l Introduction ... 48

5.3.2 Discussion of Measurements ... 49

5.3.3 Comparison with Simple Prediction Techniques ... 75

5.4 STUDIES OF MULTIPLE-S[1OCK/SIIOCK INTERACTION ... 81

5.5 PRELIMINARY STUDIES OF SWEEP EFFECTS ON REGIONS OF SItOCK/SI lOCK INTERACTION ... 89

5.5. I Introduction ... 89

5.5.2 Experimental Studies ... 89

5.5.3 Discussion o[ Measurements ... 89

Table of Conlenls (cont.) Section Page 107 109 6 CONCI.USIONS ... 7 REFERENCES ... APPENDICES

A TECIINIOUES I:()R NUMERICAl, EVAI,UATION OF UNSTEADY HEAT FLUX

FROM TtlIN-FII,M GA(;ES ... A-1 A. I REVIEW ... A-2 A.2 SOI,UTION FOR CONS'FANT TIIERMAi, PROPERTIES ... A-2

A.3 VARIABI.E TI1ERMAI. Iq_OPERTtES ... A-5 A.4 TIlE RAE-TAUI.BEE AI.GORITIIM ... A-5 B SIIOCK/SilOCK-INTERAC!TION STUI-)Y DATA ... C-I C ML!I.TIPI.E-SIIO(_K/SIIOCK-INTEI,_A('TION STUDY DATA ... C-1 D ,qWEPT-StlOCK/.qlIO(,K-INTERA(71'If)N STIJDY DATA ... D-1

last of Figures Figure I 2 3 4 5a 5b 6 7 9[i 9b 10 II I 2a 12b 13 14 15 16 17 19 20 21 22 23 24a 24b 25 Page

Six Types of Shock-Wave lnlerference Pallerns ... ... 2

Schematic Diagr,_m of Type 111 Inlcrfcrence ... 3

Schematic Diagram of a Type IV Interference I'altcrn Impinging on a Cylinder ... 3

Variation n[ Transition Reynolds Ntmlbcr wilh ('onvective Mach Number ... 4

Sled in Motion al 7,000 fps on Test Track al l lolloman AFB ... 7

Sled on Test Track at Ihflloman AFB and Resulting Intcraction-Iteating Damage . 7 Damage to X-15 Pylon Resulting from Shock/Shock Interactions ... 8

Shock Impingcment tlcating on a Right Circular Cylinder at Math Numbers Between 2.65 and 4.44 in Investigation of Newlander ... 10

Free-Flight Shock Interference Measuremenls of Heat Transfer on Unswept Cylinder at Math Numbers up to 5.5 ... 11

I,;ffects of Shock hnpingement on Peak-floating Investigation of Bushnell ... 12

Illustration of Imcrfcrence Pattern Generated by a l lypersonic Wing/Fuselage ,hmction and a Typical Wind 'Funnel M,_dcl Used ... 13

Pressure and Ilcat Transfer Distributicm Measurements of Silcr and Deskin ... 14

Spanwise Ileal Transfer Distribulion ol_ ,qla_?nation Line of the Blunt Leading-Edge Investigation of llcirs and I,oubsky ... 16

Type IV Interaction(s) and Associated Ilcat Transfer Distributions on a Ilcmisphcre at Math 5.94. Investigation of Keyes and Ilains ... 17

Types Ili and IV Shock Interlerence Paltcrns Gcncrated During Ascent of a Mated Shutlle Con figuralion ... ... 18

Inlerfcrcnce Ilcaling I)islribution on OJbilcr Caused by Booster ... 20

Shock Inlcrfcrcncc I lcaling on Bhmt ('ylinder at Mach 3. Investigation of Kaufman ... 21

Interference l leafing _m P,igl_t Circular (iylinder at Math 2.25. Investigation of Ginoux ... 22

Typical Schlicfert I'hol_graphs of Inlcrfcrencc-Ileating Flow Pa|terns Generated by a Spiked Body During Studies C_mducled at Calspan ... 23

Shock/Shock-lnteracti_m llealing on Indcnled Nose Tip ... 24

Shock/Shock-lnter,lcli_m llcating on "'l'cnsi_m Shells." Investigation of Jones, I_ttshnell and Ihtnl ... 25

Analogy Between Type 111 and Rcattaching Shear I,aycrs ... 27

Comparisrm Between Nunmrical Prcclicli_ms of T,mnchill c! ai. Using Navier-Stokes Equations and Measurements of Keyes and tlains ... 28

Performance Clmractcristics of Calspat_'s Shock 'Funnel ... 31

Shock Interference Model Mounted in (ialspan's 48-inch Hypersonic Shock Tunncl ... 33

Inslrumcnlalion _qchenmlic I)iagram for Shock Interference Model ... 33

Schem,ltic lfiagram of Mtdtifflc-Shock Interaction Model ... 35

Model for Studies _1" ,qwel_l-Sh_ck/Sh_ck Inlcractions ... 36

Comparisons Ilclwecn McasUl-ed, Prcdiclcd lie,t! Transfer and Pressure I)istribulions Al_mnd ,i Circul:lr (,ylindcr al Math 8 ... 46

I,isI oI"Figures (coul.) Fign,'e 26 27:1 2717 27c 27d 27e 27f 29b 29c 3(la 3(>17 30c 3(kl 30e

Comparisons Between Measured, Prediclcd I Icat Transfer and Pressure

Distributions Around a Circular (:ylinde,-at Mach 16 ...

I lear and Pressure I)istributions in Shncl</Shock-lnleraction Regions Induced

by a 12.5 ° Shock {]eneralor Over a 3-incl,-Dianacter Cylinder at Mach 8 for

Run 59 ...

I lear and Pressure I Iislributions in Shock/Sl_{_ck Interaction Regions Induced

by a 12.5 ° Sh(_ck (;eneral<w Over a 3-inch Diameter Cylinder at Mach 8 for

Run 61 ...

I leat and Pressure Distributions in Shock/Shock Interaction Regions Induced

by a 12.5 ° Shock (iencralor Over a 3-inch-I)iameter Cylinder at Math 8 for

Run 60 ...

Ileal and Pressure I)istribulions in Shock/Shock-Interaction Regions Induced

by a 12.5 ° Shock (;cnera[or Over a 3-;itch-Diameter Cylinder at Mach 8 for

Rut) t ()0 ...

I lear and Presstlru Distributions in Shriek/Shock-Interaction Regions Induced

by a 12.5 ° Shock Generator Ow'r a 3-inch I)iameier Cylinder at Math 8 for

Run lO I ...

Ileal and Pressure I)islribulions in ghocl</.qhock-lnterac(ic, n Regions Induced

by a 12.5 ° Shock (;eneralor Over a 3. ir_clP I)iameter Cylinder at Math 8 for

l{un 99 ...

q'ypical Time I lisi(nies for Ileal 'l'ranslcr and Pressure in Type IV Interaction

Region at Math 8 ...

Ileal and Pressure l)islributions in Sllock/gh(lck qnteraction Regions Induced

by a 10.0 ° Shock (;eneralor Over a 3 inch-Diameter Cylinder at Math 11.6 for

Run 103 ...

and PresSille I)islribulio,ls ili Sll_)cl</SI)ocl, Interaction Regions htduced I(I.0 ° Sll_lc,li (;cneratol ()ver a 3-inch I)i;uneler Cylinder ,it Math 11.6 1or 1(1-1 ...

,;lid PlcSsiire l)islribiltions in gli(iclt/<qliocl<-Interaction Regions Induced

10.0 ° Shock (leneralor (-)ver a 3-tilth--Diameter Cylinder at Mach 11.6 for

113 ...

and Pressure Ilistribulions ill glioc[q./,qhock-lnteraction Regions Induced

I(t.(1 ° Shuck (;eneralor Over a 3-tilth-Diameter Cylinder at Mach 15.6 for

107 ...

anti l'ressure I)JsltibutJons in Sliock/ghock--lilteraclion Regions Induced 12,5 ° ,qlilit'k (;eneralor Over :l 3-ilicll [)i:uneler Cylinder fit Mach 15.7 for 114 ...

,iild Pressllre I)islribullons in Shoekl,ghtlck-lnteracl, ion Regions Induced

12.5 `> Shock (-]eneralor Over a 3-inch-l)iameter Cylinder ,it Math 15.7 for

i J5 ...

and Pressure Distributions in Sh_wl,/Sllock-lnteracllon Regions Induced 10.() ° Sliock Gene,ator Ow,, a 3 inch Diameter Cylinder at Mach 15.6 for 106 ...

and Pressure I)istributious ill Sllllcl<lShock Interaction Regions Induced

by a I1).(1° ghllck (;enerator Over a 3-tivoli I)iameter Cylinder at Math 16.3 for

Rut) 43 ... ;'; ... Ileal t_)' a Riiii I It,it 177 a Rut) Ileat by a Run I leal I_ya Run I lelil h 7 a Rtlil I leaI :'by a Run 1leal Page 47 5O 51 52 53 54 55 56 57 58 59 60 61 62 63 64

l,ist of Figures (cont.) Figure 3la 31b 31c 31d 32a 3212 32c 32d 33a 33b 34a 34b 35 36 37a 37b 38a 38b 38c I leal by a Run I lear by a Rtln I leat by a Run Ileat by a

and Pressure I Iistribulions in Shock/Sllock-lnteraction Regions Induced 10.0 ° ,qhock (;cnerator Over a 3 incl_-Diamclcr Cylinder at Math 18.7 for II0 ...

and Pressure Distributions in Shock/Shock-Interaction Regions Induced 12.5 ° Shock (lcncralor Over a 3-inch-Diameter Cylinder at Mach 18.9 for 116 ...

and Pressure I)islributions in Shock/Shock-Interaction Regions Induced 10.0 ° Shock (;cnerator Over a 3-inch-l)ian_eter Cylinder al Mach 18.8 for 112 ...

and Pressure Dislributions in Shock/Shock-Interaction Regions Induced 10.0 ° Shock Generator Over a 3-inch-Diameter Cylinder at Mach 18.9 for

Variations Region for Variations Region for Variations Region for Run 109 ...

Variations ol Peak I leafing and Pressure wilh Angular Position of the Interaction Region for Mach 6.4

of Peak llcating and Pressure with Angular Position of the Interaction Math II.0 and 11.6 ...

o1 Peak I lcating and Pressure wilh Angular Position of the Interaction Math 15.6 and 16.3 ...

of Peak I leafing and P,'essure with Angular Position of the Interaction Macll 18.8 ...

Variations of Peak I leating and Pressure I,'alio wilh Angular Position of Turbulent Interaction lor Math 6.4, 8.0, and II.0 ...

Variations o1+ Peak tlcating anti Pressure Ratio with Angular Position of Laminar Interaction for Math 11.6, 15.6, and 18.8 ...

Observed Variali_ms of Peak lhcssurc will_ l:rcestrcam Math Number for an Interaction al 0 _-- 20 ° I.ocalion ...

Observed Variali_ms o1 Peal< l leating with Frcestrcam Math Number for an Interaction at 0 + 20 ° I,oealion ...

Variation of Transition Rcytmlds Number wilh Convective Mach Number ... (k)rrciation of Peak I lcating and Pressure Measurements and Comparison with

Predictions Using Morris and Keyes Comlmlatiorml Model ... Wlriation of I+cal< Pressure with Freestrcam Math Numbers for an Interaction at 0 = 20 ° l,ocalion ...

Varialion oI Peak llealing with Frcestrea|n Math Numbers for an Interaction at 0 = 20 ° I,ocation ...

I leat and Pressure Distributions in Muhil_le-Shock-lnteraction Regions Induced

by a 7.5/5 ° Shock (;cnerator Over a 3-inch-I)iametcr Cylinder at Mach 8 for Run 87 ...

I lear and Prcssm+c Distributions in Multiple-Shock-Interaction Regions Induced by a 7.5/5 ° Shock Generator Over a 3-inch-Diameter Cylinder at Mach 8 for Run 88 ...

I lear and Prcssttrc I)islributions in Multil_lc-Shock-lrlleractior_ Regions Induced

by a 7.5/5 ° Sh,_ck Generator ()veta 3-inch Damccr Cylinder at Mach 8 for Run 85 ... Page 65 66 67 68 69 70 71 72 73 74 76 77 78 79 82 83 84 85 86 xi

last of Figures (cont.) i,'igure 38d 38e 39a 39b 39c 39d 39c 40 41 42 43a 43b 44a 44b 45a 45b 4 (_

Ileal anti Pressure I)istributions in Muhil_le-Shock-lnteraction Regions Induced

by a 7.5/5 ° Shock Generator Over a 3-inch Diameter Cylinder at Mach 8 for

Run 86 ...

l leat anti Pressure Distril_utions in Muhil_le-Shock-lnteraction Regions Induced

by a 7.5/5 ° Shock Generator Over a 3 if_ch-Diameter Cylinder at Mach 8 for

RUn 89 ...

lteat and Pressure I)islributkms in Multiple-Shock-Interaction Regions Induced

by a 7.5/6 ° Shock Generator Over a 3-inch-I)iameter Cylinder at Mach 8 for

Run 93 ...

I lcat and Pressure l)istributions in Muhiple -Shock-lnteraction Regions Induced

by a 7.5/6 ° Shock Generator Over a 3-inch-Diameter Cylinder at Math 8 for

Run q2 ...

I lcat and Pressure Distributions in Multiple-Shock-Interaction Regions Induced

by a 7.5/6 ° Shock Generator Over a 3-inch-i)iamcter Cylinder at Math 8 for

Runs 90 arid tTl ...

I leat and Pressure f)istribuLions in Mllltiplc-Shock-lnteraction Regions Induced

by a 7.5/6 ° Shock Generator Over a 3-inclh-Diameter Cylinder at Math 8 for

Run 95 ...

tleat and Pressure Distributions in Multiple-Shock-Interaction Regions Induced

by a 7.5/6 ° Shock Generator Over a 3 incli-Diameter Cylinder at Math 8 for

Run 94 ...

Typical Schlieren Photograph of 15 ° Swcpl Cylinder At Mach 8 ...

Variations of Transition Reynolds Number al the Onset of Transition with

Edge Math Number and Wall Temlxrralure ...

Schematic I)iagram oI Flowficld .qtructures in Swept-Cylinder Flow Configurations

Pressure and ilcat Transfer to Swept Cylinder, X = 15 ° ...

Pressure and ilcat 'l'ransfer to Swept ('ylindcr, X = 30 ° ...

I ieat and Pressure Distributions in Shock/Shock--Interaction Regions Induced

by a 12.5 ° Shock Generator Swept at 15 ° Over a 3-inch-Diameter Cylinder

at Math 8 for I,hm 69 ...

tlcat and Pressure Distributions in Shock/Shc_ck-lnteraction Regions Induced

by a 12.5 ° Shock (icnerator Swept at 30 _' Over a 3-inch-Diameter Cylinder

at Math 8 for Rtm 98 ...

Peak-Pressure Measuremcnls for Swept-Shock Configuration ...

PcaksIleating Mcasuremcnts for SwepI-Shock C0nriguration ...

Variations of Maximum Peak i lcating and Pressure with Sweep Angle ...

Page 87 88 90 91 92 93 94 95 97 98 100 101 102 103 104 105 106 xii

l,ist of Tables

Tal)le Page

Summary or Sh()ck-lnleiference lnvesligalions ... 9

Summary or Gage Posili(ms (m (;ylinder ... 34' Summary of Test Condilions ... 39

Prediction of Peak llealing Rates And Pressures Using

Morris anti Keyes ('Oml_utation;,I Model ... 80

Section l 1 NTRO1)UCTION

el'he healing rates generalud onl I+lunl bodies hy a sllock incident on tile bow shock in the stagnation

region can lye orders o1 nlagnitilde greater than the st;lt, Jialion value in thia absence of the interaction and,

therulore, arc of considerahh.' interest to designers _>l hypersonic vehicles, hliti,i1 studies by Edney 1 of

[Iowficlds and large heating loads generated in shock/shock interaction regions demonstrate that, for

certain incident-shock/l_ow-shock configurations, the pressure recovered on the body can be orders of

magnilude larger Ihan Ihe u,_distu,hed-flow pilot pressure, which, in turn, causes a corresponding heat

transfer rate increase in the stagnation region. Edney and, later, others 2,a showed that six different flow

conliguralions can be gene,'aHed, depending on the slrength of the incident shock and its point of

intersection wilh the bow sh(_ck. IVil.,,ure I shows II_e wirious flow patterns cleveloped for the

incident-shock/bow-shocl,: conlignraltions as tlcline¢l hy Eduey. Types I, II, and V are interactions where

the shock propagates to the surlalCe of the body, resulting in a shock/hotmdary layer interaction. A type

VI interaction results in an exfmnsion-fan boundary layer interaction, which does not cause significant

aerotl_ermall-load enhancen_ent, ilowever, types Ill and IV interactions (shown in Figures 2 and 3) result

in large healing and pressure loads and are of the t,rcaicst inlerest to researchers.

The peak heating developed In regions of type 111 interaction are generated on the body just

dowllslream of the reallachment point o1 the free slle;ir layer. This reattachment phenomenon is very

similar to the reallachnlent of sel_:llaled shear layers developed over spiked bodies, highly indented nose shapes, and compression ramps. _l'he development o1 the shear layer and the peak heating developed in these flows are strongly dependent upon whclher the shear layer is laminar or turbulent, as demonstrated

in extensive studies of separaled flows hy I lolden 4. Tl_ese studius (and many others) have shown that the

slability of the laminar llmv in both the Iree shear layer and Ilte attached houndary layer increases with

local Mach numher. I lowever, in gencrall, transition of the fruc shear layer occurs at Reynolds numbers an

order of magnitude smaller than for attached Ix)unclary layers. To provide a guide in determining which

correlati_m to employ in the semi-empirical prediction techniques, F-dney _ and, later, Birch and Keyes s

correlated shear-layer Iransilion measurements from separated-flow studies and shock/shock-interaction

measurements. The IransilJon correlation developed by I_hch and Keyes is shown in Figure 4. This

correlation suggesls Ihat, in hypersonic flows, shear layers wi{l_ Reynolds numbers of approximately 5 x

10 4 will he transitional in charalcler. Fanl_h>yinglhe analogy hetween reattachment heating in separated

flows and th,lt induced in regions o1' shock/shock interaction, Keyes and Hains 2 (based on the work of

Bushnell and Weinslein _2) suggested Ihe relalionships shown in Figure 2 for estimating the heating rates

for type ill interactions. Wlfile such correlalion techniques may he considered outmoded by current

compulalional procedures for modeling the Iransitio,ud shear layers and I_oundary layers, the correlation

lechniqnes still represent a llsefnl engineering tool.

A type IV interaction is gcneralud when an ol)lique shock is incident close to the normal region of the

bow shock, :is ilhistrated in I:igurc I. lu high Reynolds numl_er flows, shock/shock interaction produces

an essentially inviscid slipslreanl -which Eduey termc'd :l "jet"--in which the supersonic flow is efficiently

compressed by a sel:ies of c_mlpression and expansion waves. This jet, which is hounded by shear layers, is

terinill;iled by a normal shock jnsl ahead of Ihe sHrlace Io produce a narrow stagnation region. The

TYPE V SI-IOCK IMPINGEMENT JET TYPE Vt EXPANSION WAVE IMPINGEMENT EXPANSION FAN TYPE I SHOCK IMPINGEMENT

lAMINAR -_ 0.19

e,,,_J_(ll,,,,, il,,,f ÷

IIIAN';JII( _NAI _(),{);'1 FlOw Sl I(_CK

L

J'"'-_

J

)

[ lh_'''ll''4}'*'l I I)71rh¢l_' _},_1, f). I,_._ J_,S'l, 11,,,,,_i.,I .j M<I M>I _G _jODY -- /,_,__....___ r_ ,_ / SHEAR LAYE R Q }\\

Figure 2 SCHEMATIC DIAGRAM OF TYPE III INTERFERENCE

M>I \ IMPINGING SHOCK _ SHOCK M<I

Q

M>I

JET BOW SHOCK JET,. M>I M<J SHOCK CYLINDER M<I SHOCK WAVE _- EXPANSION WAVE ---- SHEAR LAYER

Figure 3 SCHEMATIC DIAGRAM OF A TYPE IV INTERFERENCE PATTERN IMPINGING

u) £3 O Z LU rr Z o Z I-I0 _ 10 _ 104 10" for (I_ _ 0 (See Figure 2)

0

0

CHAPMAN EI AL. (1958) NACA 1356 CRAWFORD _ Re/_, cm-1 NASA TN D-118 _ _ 1.57 x 105 (1959) _ 0.790.39_0.23

BIRCH AND j_ [] MACH 6 TUNNEL

KEYES (1972) _L_ O 11" TUNNEL

I

I

1

I

1 2 3 4

i;,

Figure 4 VARIATION OF TRANSITION REYNOLDS NUMBER WITH CONVECTIVE

rcgioti that Clill CallSt. large lil:iiliiil2, I-:ilcs relative to i.hc heating generalcd by uiidisturbed flow. A simple

rclalionshil-_, suggested by Edncy, Ior cstimali_g the ratio o1 lype IV heating to the undisturbed value is

shown in Figure 3. However, with increasing Math number and decreasing Reynolds number, the laminar

shear layers that bound the jcl can itd'hlcncc the characteristics of the jet to the point that the structure of

the st,lgnation regions is significantly nlodificd. IJilclcr such conditions, viscous effects can significantly

reduce the heating levels at the base of the jet, as will bc seen later. Alternatively, if the shear layer

upstream of the jet is transitio_ml, the disturh_,nccs generated in the slagnation regions of the jet may

significantly on!lance the peak heating Icvcls.

in this rcporl, we present an experimental study to investigate the effects of Mach number and

Reynolds number oil the detailed distrihulions of pressure and heat transfer in regions of shock/shock

inter:lction over a Iransvcrse cylit_dcr _t Math 6 to 19. Additional studies of the aerothermal

characteristics o1" regions of nlllltil_lc-shock/shock il_lcraction and a preliminary study of sweep effects on

regions of shock/shock inicraciiofls are presented. Wc first review some of the results of earlier studies of

regions of shock/shock interaction. We then discuss the ohjectivcs and design of the present experimental

program. The models and instrumentation employed it_ Ihis study arc presented, together with the test

procedures employed. The results o1 Ihc cxpcrirnen|al program are then presented and discussed.

Dctnilcd listings _[ the model configurations, tcsl conditions, and measurements of pressure and heat

Section 2

REVIEW OF EAI_,I_IER STUDIES

Studies of the aerothermal loads generated by shock/shock/boundary layer interaction, or

"interfer-ence healing," began shortly after the advent ¢)1"supersonic flight and construction of supersonic wind

tunnels. Mosl of the earlier sllJdics were concerned will, !lie aerodynamic loads generated by shock/shock

interaction, llowever the structural failure of the pylon supporting the dummy ramjet engine in X-15 flight

tests 6 and the burning up of the nosctip model in a sled test conducted hy the Air Force 7 provide graphic demonstrations (Figures 5 and 6) of the searing heating loads that can be generated in the regions of shock/shock interaction. One of the most definitive studies of shock/shock interaction resulted from an

investigation of "anomalous heating rates" by Edney _. Other major studies of interference heating have been formulated to investigate the aerothermal loads from (i) sllock incident on a fin, wing or pylon; (it) nose**p/body shock interaction on indented nosetips and spiked bodies; and (iii) impingement of shocks onto inlet cowl lips and inject_r struts. A chronological summary of various investigations is given in Table

I.

Investigations of "shock impingement" heating began in the early 60's with studies of the oblique shock incident on swept and unswept fins. These early uffl_rts were motivated by observations of unusually high heating rates generated during flow-visualization studies of various uninstrumented supersonic air-craft configurations at NASA/Langley. The first definitive investigations of shock-interaction heating were wind tunnel and free-flight studies conducted in superscmic flows (up to Math 5.5), designed by

New]an-der a and Carter and Carr 9 to measure the heating in regions of shock impingement on unswept cylinders.

A typical model conliguralion for Newlander's investigaiion, ahmg with the associated heating rates, is

shown in Figure 7. Similar hcaliJq:, rates arc shown in Figure g for the free-flight investigation of Carter

and Carr. These studies revealed healing enhancements 5 to I0 times higher than the reference stagnation

heating vahle in Ihe presence ¢_f shock/sl_ocl_ interaclion. Measurements on similar configurations were

conducted in hypersonic flows by Francis m. Beckwilh _1 and Bushnell 12, 13 measured the interference

heat-ing (m a swept cylinder close t_) *Is.juuclion will, a wedge, while Jones 14 studied a fin/plate interference at

Math 6. f_ushnell's t2 earlier W_lk focused oi_ lhe iuterference heating problem caused by the root region

of a wedge-swept cylinder configuration, and, initially, analysis of localized effects of shock/shock

interac-tions was not pursued, l lowever, in a subsequent study, I_ushnel113 isolated both the effects of shock/

shock i,deractions and the cffccls of separated root region. By supporting the cylinder away from the

shock generator, he removed the separated region al the wedge/cylinder junction. A typical result of

peak-healing expectations based o,1 Buslmell's work, and tlm typical model configuration, is given in

Figurcs 9a and ()1_.

Also, during the mid-61Ys, sludics were being conducled for the Air Force Flight Dynamics

Lab-bc_ratory (AFFDi.) in the Arnuld I!ngineering Development (?enter (AEDC) facilities by Slier and

Des-kin I_', (;ulbran _ aJ]., Ic''_z Kn(_x, la al_d Ray and I_all<o. m Again, the emphasis in this work was on

leading, edge shock inlpilq, tmlellI and was a direct res, dt of AFFDI.'s experience with high heat loading

¢_bserved during their supers_mic aircraft lcsting plc_grams. Siler's investigation of leading-edge surface

heat transfer yiekled heating c,lhancements o1 5 limes tl_c values with¢_ut shock impingement. The high

heat rates caused by the slmck/shock interacti¢m are illuslratud in Figure 10. This lest was conducted in

ORIGINAL PAGE

BLACK AND WHITE PHOTOGRAPH

Figure 5a SLED IN MOTION AT 7,000 FPS ON TEST TRACK AT HOLLOMAN AFB

Figure 5b SLED ON TEST TRACK AT HOLLOMAN AFB AND RESULTING

INTERACTION-HEATING DAMAGE

ORIGINAL PAGE iS

,_ _ FUSELAGE

... LEADING-EDGE SHOCK --_

\ _ LEADING EDGE

c

sP,

<...__._OOWL-L,P

_ '_JOOWL

_- RAMJET SPIKE FLARE

Figure 6 .... D_MAGE TO X-15 PYLON RESULTING FROM SHOCK/SHOCK INTERACTIONS

O_iGiNA'L PAG£ i_

OF POOR (_JALITY

_® 8_=.... cE =y=o,CCoo n_ ,_._ _ m c _E __mr: O_ -_ I_ LE _ --_'_ : __= =>10 E II

"_,_

_

__

n_. m m E 3 . ®E 0.0. E .E .C E _ a. _ =-_ j_0o£ _ . _ _ _g'. -_-o_ L

i

"aE oE_,= -_ E ,_+8#--o 6 ° ooF#-°°: _m m i O9 o.E_l _ u_ _-_i _ u_ _o_ .2 ,_ o +9,o _.j _ +_ _ _ I ! o x o ,= ._ 9 o oo 9 It3 Z _o .__ x I X X I X x I _- -- = m ILl > Z LU xx x xx xx o x x x x x X X X X X X _EO -r_ × X X_ X XX IX X X X X X X X 0 Z _ .o I_1 n X X X X X X X X X x x x x x x x x x x u. __

°°

7 I'< m v 0 ____ _ _ 0_{I J_E}o_ _{_ -J}. 0 0 g_ g g,_ o 0 ¢= 0 I O o i_ o u_ c_ o 6 o o o 0 o o _0 0 0 0 ("J - l 0 o o I o o u] to o _1 o o_ o o ._o o o _0 _._ o o o o _ o o o u. 0 c _ = _ g gg ,8 % %% %% ®® :E D (/) Q) I--x 2 x oJ 03 _ O0 X XX XX _ 0 0 _ 0 0 2 2 _ _2 2" x x _ x x x x %× c_c_ × ×Z- _ x U3 0 _ 0 0 0 x _ _ x x x

_

_o

o° o co ,_ o o 8 '_:_ o _ _o o oo _o _D IDp.. _ 0

_

_

_

-

o

_j

,_ _ == __ J b= i _ E (0 _QD OIm v E_ v_ z_.-_ o_ _ ),-,,err "r _m

._ = _ _

_ c _ _. c¢O -"o 'S" _oo) >. _ "_

AIR FLOW Z 2.8 INCHES 'i, " k\b\_ 2.5 INCHES ,,, , 16 1/4 o I "iJ_'l SIDE VIEW 3.82 INCHES

FLAT PLATE SURFACE ____ ild" L M,I, 2.65 2.65 2.69 3-51 3.51 3.91 4.44 4.44 4.1_4 V,,, 3.99 x lo 6 2.(,9 i .(_0 3.95 2.7 b 1.(_ 1, .50 3.2o 2.15 It 0.01661 .01356 •OO9)9 .o1472 .01251 • o0938 .Ol 2_2 .01071 .OO871

t

1.0 Z/L 4 3 2 1 0 /l ho

Figure 7 SHOCK IMPINGEMENT HEATING ON A RIGHT CIRCULAR CYLINDER AT MACH

NUMBERS BETWEEN 2.65 AND 4.44 IN INVESTIGATION OF NEWLANDER (Ref. 8)

240O

23OO

2200

o 15 (TYPICAL) __lr__ 1.755 7O0 6OO (.9 en 2100 2000 1900 Figure 8 1 Moo D I 750 I .88 DIAM. (INCHES)

Shlieren Photograph from' Mach 2,65 Wind Tunnel Test

D : .750 IN

Laminar Theory (Ref 9) Turbulent Theory (Ref 9) O One-Dimensional Cl • Aerodynamic cl Wall Temperature (°R) Approximate Location of Shock Intersection TC - Thermocouple 2 3 x/D 5.50; ReD= 1.87 x 10_

FREE-FLIGHT SHOCK INTERFERENCE MEASUREMENTS OF HEAT TRANSFER

ON UNSWEPT CYLINDER AT MACH NUMBERS UP TO 5.5 (Ref. 9)

II ORIGINAL PAGE IS

HEAT TRANSFER TO LEADING EDGE WITH AND WITHOUT GROOVES (TO TRIP THE BOUNDARY LAYER)

Moo = 8 ; X /D _ 12 I 11 600 -400 THEORY j TURBULENT o GROOVED _]_""

E] SMOOTH ._i,_ I_ LAMINAR

LOW HIGH

R_, D R=_, O

L .J. i .[ I.LJ. LI ___L__i i_l J_i lIJ

104 105 106

R,_, D

TYPICAL MODEL CONFIGURATIONS FOR SWEPT

LEADING EDGE STUDY

M_o

l

Figure 9a EFFECTS OF SHOCK IMPINGEMENT ON PEAK-HEATING INVESTIGATION

OF BUSHNELL (Ref. 12, 13)

FLOW FIELD

ASSOCIATED WITH LEADING-EDGE SHOCK IMPINGEMENT

REGION ---/ IMPINGING___ _,_-. ,__VsOHRTEEX ;M_GE ME N ? SHOCK-J

-M-M_. __. "--i_i__" - VORTEX SHEE T

BOW SHOCK - "

L EADING EDGE_ (.,_,..,_

EFFECT OF SHOCK IMPINGEMENT ON MAXIMUM HEATING

_'_5EPARATION AND

IM-Io _ PINGEMENT _I CLOSE

PROXIMITY

8 Moo, 14 L_SOME ROOT SEPARATION,

IMPINGEMENT IN {)E-VE LOPING BOW- SHOCK

REGION

6 o NO ROOT SEPARATION,

IMPINGEMENT

INDE-.... M_o'2 5TOB_////S-._ d'_ VELOPED BOW-SHOCK

°,.o 15 2,6 _{3,0 4D}-- _5.0

P2/P,

MAXIMUM HEATING AS A FUNCTION OF DISTANCE FROM TIP OF CYLINDER

6 A_O=_ Moo:8 /l_ o o l I 1 I 1-_l 0 .2 .4 ,6 .B 1.0 1.2 A/D

SIIOCK STAND OFF DISTANCE

Figure 9b ILLUSTRATION OF INTERFERENCE PATTERN GENERATED BY A HYPERSONIC

WING/FUSELAGE JUNCTION AND A TYPICAL WIND TUNNEL MODEL USED

(Ref. 12, 13)

ORIGINAL PAGE IS OF POOR QUALITY P/P(I 10.00 ... 5.00 I.(X) ()._0 (I.Ill 0

Qage Location Around The Cylinder

(-i I I I, ( 10.00 5.00 "-'T_ Sliutkllnpiiigelll('N| n 7 ] hoiml Sh,iduwijraph £ ',r ___l_.l ... J I V,in. I.(l() ().5l)

6/60

1).1(} 0 ol p i I i I ! 1 i i i t I !

+VY ,

,

12 12 .A \ \ ,G',-..,n_.&_ ^ . N = Q; V I Shock hiipinqement 71 Iroll) ShddOwcjraph _T__J__ L I 0 3 _ 9 V. iil.

Figure 1 0 _{PRESSURE AND HEAT TRANSFER DISTRIBUTION MEASUREMENTS OF} SILER AND DESKIN (Ref. 15) (using N,_ at Mach 19)

conducted in AI.+.I)C Tunnels I+, a_td (_, again will1 a wedge/cylinder model configuration. These studies were conduclcd in air at Math 6 to 10. The peak heal lrans/cq rates observed here were 3 to 4 times the undisturl_ed rate. It is clear that, iu ilmse studies as well as i,I Ihe earlier work, the density of instrumenta-tion was it_stlfficient to define tl_u" distribution and peak heating in the interaction region.

At NASA/Ames, Ilei,'s and I.(_ubsky 2° :llso Sltldit'd I[le effects of shock impingement on the heat transfer to a cylindrical leadin!, edge in the Ames shock tlmnel at Mach 14. This latter study highlighted a number oi problems assf)cJalcd wJlh measurement _f Ihc large heat transJcr ra_es and gradients generated in the interaction regions. More specifically, in the expcrimenls of lleirs and Loubsky, and in many of the other studies, the concluc(ivitit..s of the model surfaces were such that the measured peak heating and the distribulion of heating rates were significantly reduced by heat conduction ahmg the model surface. Al-though a correction procedure can he used, it can I_e Ifiklfly inaccurate, paJticularly if the actual heat transfer distribution is uns|eady. The hml,+ respol_se lime of the inslrut_qentalion used in the past made it impossible Io resolve unsteady m_wemcn! in the interaction regkm. Even wilh these difficulties in instru-n_entation, the measured peak heat _ates on an tlllSWt..pl cylinder impinged by an externally generated shock wave were estimated to be 1() limes those of Ihe unclisturl)ed values. This study also concluded that increasing the swept angle of Ilte cyliJ_der will alleviate the heat loading. At larger swept angles, the heat tl'at_sler rates could be adequately analyzed through the use of a simple two-dimensional boundary layer solulion wilh swept-cylinder Ival]slorm:_lion, as discussed later A typical prediction for an unswept con-Iiguration as compared wilh an experimental resuh at Math 14 is illustrated in Figure 11.

In one of the mos! sie.nificat_t investigations of shock/sh(_cl<--interaction phenomena, Edney _ identified the basic flowfield structure in rcgi_ms _11 shocl,:/shocl< mleraLClion for a number of shock geometries and '+uggestcd simple prediction1 nlethods t_} estimate the acr, lthcrmal loads generated by them. The work of

Edney provided valuable insij:hls to various types of inviscid and viscous interaction problems and will be discussed further in detail. IqJllowing this major wool< (cited earlier), a scries of studies were initiated by NASA/I.gln_,lcy to develop In_)rc acctll',_le mclh_Ms I,_1 pledicling sh(>ck interaction heating with applica-lion to the design of the Shtlllle project. Tim work ,1t Keyes anti llaills _ and, later, Keyes and Morris 22 extended the experin/cntal work of F.dney and provided some simple prediction techniques with which to cslimale the aerodynamic I,mcls, as well as peak heatini,-rale measurements. The work of Keyes and I lains, in particular, Iucused _m tile various lypes of il_leraclions that could occur on the surface of shuttle/ lank conligurations. '['his invcsligali,,n led to the c(,nclt_si_m thai a Type IV interaction could occur on a shuttle/tank configurali¢m, inct'easin!,, the local heali]l_, tale up to 20 times that of the undisturbed ITeestrcan_ Imaling rate. TI_e tcsl results, model c(mfigtH-ation, and typical locations of the interaction are illustrated in Figures 12a aud 12h. The heat transfer measurements made in the Keyes and Haines work were deduced f,'om phase-chane, e paint Icehniques, which have slow response and are accurate to only +30%. floweret, tim evaluali,m of the flowfield characleristics represenls an imporlant contribution. The c(mll-mter program writlen by Morris and Keyes 2t pr(widcs particularly useful information in this regard. Mcasurcmenls of the shock impint,,en_cnt heal loads ,,n Ihc shuttle were also performed in a number of

olhcr sludies mantle with models of lhe orbiler (Rogers _,_) ;_ud (he orbiter/lank (Lanning _4 and Ginouxa_).

During the mid-6()'s Io early 70's, Ihe Ai_ I:orcc. moliva(ed by their experience with the X-aircraft testing program, SUl_porled _,rl< _m shock Jlnpiltgemcnl ,)1_ hhlnl fins (Kaufman 2s) and inlets (Craig _) as well as (al tile Von Karmas Inslittlte) on conical b{)dics and cylinders. The works of Kaufman a°,_l were directed h_ sludy of Ihc detailed slrucl.ure o1 Ihe sep:lrali{m and interaciion regions of a blunt fin mounted on a flat plate The focus of these works was 1{i study detailed flow struclure by various means of flow

i@

hoc,

15 -- _ (cm) _Oo__" Invisc!d streamlines [ I / }_ Viscous streamlines / I Bow shockwave / ! I |

/

/

/;-51 Separation (oil /accumulation) Yawed local flow " _

Impinging shock

Generatedshockw_..._ _ _;1 \

Separation stlock wave _ Shear layer tl ]

Boundary layer ._ " _ _ ' /g - |

J.._...-..---"-__ _Z _ .... __._/-

--_---/--"_--- - 7/- -_ / Leading edge Separation Shock wave generator

10

0 0

--o-- Experiment

"o Extrapolated data

Calculated heat transfer maximum -- Calculated heat transfer distribution,

v Intersection of generator and leading edge shock waves

Y

t-x

_5

/,

/ _____:_1 -- -o - _ I"-°" iV J I I , ,I .5 1.0 1.5 2.0 2.5 3.0 3.5 y. cm

Figure 11 SPANWISE HEAT TRANSFER DISTRIBUTION ON STAGNATION LINE OF THE

BLUNT LEADING-EDGE INVESTIGATION OF HEIRS AND LOUBSKY (Ref. 20)

__Q_ Qsl=i 1o 6 4 7 0 -I00 S_-_ • R LA_[R 7

JET flOW SHOCK IN REGION Q

o \

L -73 -254

(a) Schlieren photograph.

<\

lest run

1Oo

2;+<-;'l;C]-z!

i-_--,,.+,,+°°,,

I

I

i

$li(J ICilcultrled _lk_ _-_ "116"3(Regi°nlli'-- I_ -I_ -_ -40 -N |1 dL'U

,<, [ - 7:-+.>--

-I _:o

....

-l_

.__

_i 7;_1

Q Qlli t

JET HOW SHOCK IN REGION Q

(c) Interference patterns.

Tesl run lpo =o,N/cm_lo,=o . I Iw. I !(_iligo_lCnll_ (il. Oe9

/

----0--- l.llierlment -- -- -- Nolnterlerence _-- C_ku_ed _ks I] B6 14 ---Re_lon II -10 -- - _ ...

...

_i_

-I

/ i -lO0 -_ -60 -40 2O 40 _0 g0

(b) Heat-transfer distribution. (d) Heat-transfer distribution.

Figure ! 2a TYPE IV INTERACTION(S) AND ASSOCIATED HEAT TRANSFER DISTRIBUTIONS

ON A HEMISPHERE AT MACH 5.04. INVESTIGATION OF KEYES AND HAINS

(Ref. 2)

_ref

No Hho(_k li_[erfert, nce

---_--- = 20 -x

Type lI[ or T]ip(! IV

Type V or Type VI

Type [ or Type [r

Shock [n[errerPnce J_f,;llilil_ duriB|: n!,,.,.ll| ¢)_' ;l In:ltpd _Jllll[|,. c¢)nrJLn.trallon.

;-- Typ_ IV, V or VI

/ dppendir_g on

Sh_Tk ._,rw ,or,/')

,"ev,',!,):,_ I Type r. ,)r rv . /- // /

Sh.ek - ]

[,,wattons of iyp(,s 0r il_terf(,rvn(,e h(,alin); on ma),'d confiF, uratlon at _I_ =20.

Figure 12b TYPES II! AND IV SHOCK INTERFERENCE PATTERNS GENERATED

visualizalitm and to perlorm detailed pressure measurements. Although the abnormal peak heat rates were observed, no heat transfer d_lta were Ieporlud. *l'hu later _.'llorls of llasl_tt and Kaufman focused heavily on heal load measurements, and on complex geometrical configurali_ns. The tested configurations in L eluded various combinations of the orbital model with a I'lat-plate receiver and models of the shuttle fuel tanks (Figure 13). These combinations o[ models, al<mg with surface-mounted heat transfer gages, were used to validate a lypical empirical heal-load amplilic:ttion correlation function, an example of which is given in Figure 14. Tile orhiter/tank configuration was used along with heat-sensitive paint to predict residual heat loading on the shuttle fuel-Lank system during the separation stage of the shuttle flight. In the results illustrated in Figure 13, peak heating is shown to be as much as 30 times the undisturbed values. Tile work of Craig and Ortwerth 3 was directed specifically toward cowl lip heating, which currently is a problem of major concern in the design o1 tile National AeroSpace Plane (NASP). Studies by Gu]bran et al. 16,_z and work by Ginou× and Matthews _2,3a provided further measurements and analysis of shock/ shock-interaction problems (_ccurring on blunt I)oclies.

In a later work, Ginoux a:_ sludied the types III and IV shock/shock interaction produced on a wedge/ cylinder model configuration. I lowever, only small increases in the peak heat rates were observed, be-cause of the weak interaction regions generated by the low free_tream Math number used in this investigation. Typical geometrical configurations and distributions of heat transfer are shown in Figure 15.

Tile heating loads devel_ped in regions of shear-layer in_pingement were measured in a number of studies on spiked bodies (lloldcn:_4), indented nose slulpes (llolden as,aG) and Tension Shells (Jones,

Bushnell, and t luntg7). In these studies, the interactk_n regions were generated by the interaction between the nosetip shock and the body shock, as illustrated by Figures 16, 17, anti IS. In most cases, the basic mechanism for healing c,nhancem_.,nt was Ihe reattachmenl of a free shear layer (axisymmetric), although there were a number of cases where a free jet was formed. Figures 17 and Ig show some of the typical heat transfer distributions develop_'d along the body serrates. 'Fhe investigation of the flowfie]d around a tension sllell hy .Jones, i]uslmell, and limit gz utilized phase-cllanging paint to measure heat transfer rate; tilt, test gas of tetrafluoromelhane was used ill an allclllp[ |() ewfluate the effects of 7 on the interaction. 'File investigations of spiked bodies ,_4 and indented nose lips '9_,a6 used high-frequency thin-fi]rn instru-mentation I_ measure heat transler distribution. A graphic illustration of the damage caused by the enor-mous heating rates in reatlachmenl regions on spiked bodies was provided by sled tests conducted at I lolloman Air Force Base. ,,ks sh(_wn in Figure 6, tile shock-induced heating loads caused significant damage to the sled and its rmmers.

Tile techniques used t_ pl_wide simple predictions of tile' flowfields and peak heating levels in regions of shock/shriek interaction are founded on Ihe Fl_>wfield analysis developed by Edney _. The basic ap-proach empt_yed by Edney was expallded and incorp_>rated inLo a numerical code by Keyes and Hains. 2 Ftlney defined the basic flowfiekl structure_ R)r Ihc varifies incidenl-shock/bow-shock configurations

(Figures I, 2 and 3) and formulated techniques to quantify the flowfield structure and estimate thepres-sure and heating levels at the b_>dy surface in the interaction regions. For a type IV interaction (where the jel-likc flowfield structure was p_stulated t_) I_c inviscid in n:_ture), a simple stagnation-point heating relationship was formulated to eslilnale the peak Itealing in the interaction region. For a type Ill

interac-Lion, tile healing at tile attachment point of |he tree shear la_/er to the surface was estimated from

semi-empirical rel,ltionships based on correlations of reattachment heating rates. The correlation was developed Irom measurements over compression ramps, base-cavity flows, and spiked bodies. Edney justified this ,lppr¢)ach by drawing Ihe analogy bt'lween these lath,r Ih)wfields and those produced in regions of shock/

STING SLEEVE STING ADAPTER

7

_ ray 3 r h/h REF 0.30 010 0.03 001 m ----r- 1 ' 1 _ I ' 1 1 I ' I ' u

-- E]RAY 1 14 ° ,4_" Po" 850PSIA u

_ ARAY2 7.5 ° _ _ * :- 0° --O RAY 3 0 O • RAY 4 4 ° _1_o, t_ A

'_@%%_o

_'_=_

---

o°

_---- 0 0 O0 --0 0 0 O0 0 0 0 0 t_ __1__1 _m__l_=__L__ J__ ___l J l , 1

DIRECTION ALONG ARRAY (in) (see above)

Figure 13 INTERFERENCE HEATING DISTRIBUTION ON ORBITER CAUSED BY BOOSTER (Ref. 31)

I (3(I --h ! X

° 1'

O50 PSIA 22" /_ meso / ._/,s Ps,A_ % t I]S/F T _ 75 8E, O x 1 FT • 850 0 _ 0 2O0 75

1- UPSTREAM OF SHOCK 2--DOWNSTREAM OF REFLECTED SHOCK

Figure 14 SHOCK INTERFERENCE HEATING ON BLUNT CYLINDER AT MACH 3.

_,,..._...______lt.O_4 CK ANGLE

S.E__

_'__M,./

=225

h 1.7 1,6 1.5 1.4 1.3 1.2 1.1 1.0 ,9 .8 .7 .6 .5 .4 .3

/

\

\

{) P'0 80 mm Hg laminar shear layer [3 P'0 - 130 mm Hg transitional shear layer

* P'0 175 mm Hg turbulent shear layer ---- Shock Layer Profile Heat Transfer Distribution

0

90 80 70 60 50 40 30 20 10 0 350 340 330 320 310 300 290 280 270 36O

(DEGREES)

FigUre i5- iNTERFERENCE HEATING ON RIGHT CIRCULAR CYLINDER AT MACH 2.25. INVESTIGATION OF GINOUX (Ref. 32)

ORIGINAL PAGE iS OF POOR QUALITY

Figure 16 TYPICAL SCHLIEREN PHOTOGRAPHS OF INTERFERENCE-HEATING

FLOW PATTERNS GENERATED BY A SPIKED BODY DURING STUDIES

CONDUCTED AT CALSPAN (Ref, 34)

ORIGINAL PAGE IS OF pOOR QUALITY

(a) MACH 11 (b) MACH 13

ZII 2.4 2.O ChxlO 2 1.11 1.2 O.IIt 0.4 [] 13 0 r-I O O r_ [] [] [] ¢30 O° o•O O O• [] 0 0 0 [] [] O0 []

Fol , l,,.,i

[] [] 0 0 [] 0

8

! I ! I I I I 0.2 0.4 0.6 0.11 1.0 1.2 1.4 I/IPl R

Figure 17 SHOCK/SHOCK'INTERACTION HEATING ON INDENTED NOSE TIP (Ref. 36)

ORIGINAL p_{A_E},_o _{_5} QE POOR QUAMTY 28 2.4 h/h o .?_0 1,6 12 .8 .4 0 0 Phase-change temperature © 125°F (51.68°C) [] 150°F (65.65°C) h o - STAG ENTHALPY MODEL WITH DIAMETER = D ! °05 a -_ i . i i .10 .15 .20 ,25 ,30 Skirt Shock ,,_ Nearly Isentroplo ,Jc Compression //_ Region / ./a "_.///""_ Imbedded Forebody J'// Shocks

Flow

,.y_/

\

.35 .40 . 5 .50 .6 .5 .4 ,3 .2 .1 X/D < .2 .1 I 0 0

Figure 18 SHOCK/SHOCK-INTERACTION HEATING ON "TENSION SHELLS." INVESTIGATION

OF JONES, BUSHNELL AND HUNT (Ref. 37)

shock il_tev'acLion, a_ illustrated hy Fi_, re t0. Ih_wcv,.'r, previous expcrhnenlal studies have shc)wn that

q II_ese conelaticms of attacilmcil[ healing are re;lstmabty ac_'Clirale on[y when the flows are either fully laminar or lurbulent. Tin,s, a type ill interaction, _vlfic'h is generated by a transilional shear layer, is intrinsically transiLiorlal in nalure and is extremely difficuh to predict, regardless of the computational technique employed, l lowew'r, in hypersonic flows where the Reynolds numbers are low enough to main-Lain a laminar slipstream (shear layer), shear layers bounding the jet formed in a type IV interaction influence the peak healing, as illustrated by the rcsulls of the present study.

Because tim basic flow_eld structure in regions of shock/shock interaction is controlled by principally inviscid phenomena, it is not surprising that numerical techniques to solve the Navier-Stokes equations thal have good shock-capturing prol_erties along wilh high grid resolution have been used successfully to predict the basic flow structure and surface iSressure distrH_utions of these flows, as illustrated by the computations of Tannellill _.,1 a._[l. (f;il,i "e 20). llowevcr, a subslantial grid refinement is required, as illus-trated by the studies of Klopfer and Yue _9 and Stewa,l cI nl. 4° to predicl tile surface healing. Furthermore, withoul an adequate Iransition ,nodel, i! is difficull [_ accurately preclicl lhe heating levels for a type IV interaction occurring at Reynolds numbers of approxinmlely 104 .

Tl_e major problem wilh 1he earlier experimemal studies of shock/shock interaction was that they lacked the instrumenlation lo provide adequate accuracy and resolution of tile heating and pressure distri-butions in the interaction reeions and were not able Io develop high Mach number, high Reynolds number interactions..Also, in most cases, tile heat transfer inslrumentation employed suffered from problems associated with lateral heat conduction and low frequency response. In the studies presented in this re-port, we employed ins[rumentation with Ihe spatial and temporal resolution necessary to determine the detailed distribt,tion of surface properties. We were able to identify the basic character of the boundary layer on tile body for type [11 and type IV inleraclions because of the intrinsically high frequency response of our thin-film heat transfer ins[rumenlntion. The emphasis in these studies was placed on an investiga-tion of Math number elfccts on the l_ealin,, levels and pressure distribution in regions of types III and IV interactions. The Reynolds number was wlried ill the present studies to investigate the effect of shear-layer transition _m sl,rfacc healing. We also invesiigaled tile effects of mul|iple-shock impingement and the effects of _weep in these flows by using a number of geometrical configurations of the shock/shock-in-teraction model.

(o)

(_)

\Q

,/,'

{b)

_H__

....

//,//

p /._./

.

; ":'" \ _.\ H2>1 Q {d} j...--'//-_

/..iS1-

....

:_

_Q/MR

P - SEPARATION POINT IR REATTACHMENT POINT (e) P

_ __M2

_>_.!__

",,_,.'?_

_M3

<1

)

/.

Figure 19 ANALOGY BETWEEN TYPE III AND REATTACHING SHEAR LAYERS (Ref. 1)

0 NU_ fltK A[ _

Comparison of shock shapes.

= ANGLE MEASURED FROM STAGNATION STREAMLINE (WITHOUT IMPINGEMENT) II.0 10 0 9,0-I).0 4 FT.O " 4+0- 5.0- 4.0- 3.0- 2.0- 1.0-.418 _ --40 -?0 O O O O O O O O O O O J. _L [ 0 0 0000000000 ° J I L 1 J 40 60 O0

Comparison of surface heal Iransfer rlle_.

Figure 20 COMPARISON BETWEEN NUMERICAL PREDICTIONS OF TANNEHILL ET AL.

USING NAVIER-STOKES EQUATIONS AND MEASUREMENTS OF KEYES AND

HAINS (Refs. 26 and 27)

Section 3

I,_XI)ERIMENTAI, PR()GI_AM

3.1 PROGRAM OI|JE(?TIVE AND DESIGN

The ()bjective of the present studies was t() provide delailcd pressure and heat transfer measurements

as well as SchIiercn photogriq)hs to define the structure and properties of regions of shock/shock

interac-tion al Mach numbers from 6 to It). The emphasis in these studies was placed on type 11I and type IV

interactions, Ior these provide Ihe largest aer(_thcrmal loads. Studies were performed over a range of

Rcyt_olds nunahers to explore Ihc effects of transition on the heating r4tes. Measurements were also

per-formed for fully laminar c(mclitiotl.s to provide _l data sol that could be compared with theory without

transition or turbulence modeling problems.

The first group of studies explored the aerothcrnml characteristics of the interaction between a planar

shock and the shock layer ahead of a cylinder, supported perpendicular to the flow. The primary

objec-tive of this sttidy was to investigate Ihc elfeels of Math number and Reynolds number on the magnitude

and distribution of heating caused by various types (ff shc)ck/shock interaction on the cylinder. The Mach

number range from 6 to I c) was ()1 particular interest here, because the type III and type IV interactions .

generated at these Math numhcrs c;luse large acrothermal loads. The condition of the shear layer, either

laminar or turbulent, gcncr;itcd hy a type Ill interaclion, and upstream of or surrounding the jet for a type

IV interaction, is the olhcr major l,_ctnr that controls the peak heating levels generated by the shock/

shock interactions. The Mach nt_mbcr and Reynolds number in the shock layer adjacent to the shear

layer arc believed to bc the most important parameters controlling transition of the shear layer, which, in

f.urn, _s controlled I)y the Math t_umber and ReyncJlds number of the freestream. Measurements were

made at Math 6, 8, and 16 fur Reynolds numbers !arge enough to ensure generation of turbulent shear

layers by the sia()ck/shock inlcractions. Tile majority of the sludies at Math numbers from 11 to 19 were

c()nduclcd I()r Reynolds numbers where the shear layers were determined to be fully laminar, based on

analysis of the measured thin-filna gage characteristics.

TI)c second seric,s ()f sltJdies invesligalcd the ,_erolhcrrnal loads associated with the impingement of

two oblique shocks il_ the sia_,nati(m region of ;_ cylinder model at a Math number of 8. The objective was

to dclcrminc whether {lie hcatirL_,, levels gencraled hy {he irl{cr,lclions of two oblique shocks with the bow

sh()ck wollld generate signific;inlly h)wcr hc;.lliiw, I{);I¢ls than a single shock of the same overall strength.

Thi_ invcsti[,,_lion arc)sc fr()m questi(ms c()nccrning the impingement of multiple shocks from the

compres-sion ramp ol ramjet inlets (m the c()wl lip. In these experiments, the effects of the relative strengths of the

two incident shocks and Iheir positions relative to the bow shock were studied. In particular, attempts

were made to) determine whcll_cr the heating loads would be substantially reduced by preventing the two

ramp slmcks from coalescing before they impinged on the bow sllock. Measurements were made to

deter-mine the rekdive magnitudes of the heating loads developed for a single impinging shock and a pair of

focused sl_()cks with the saint _wcrall turning angle.

The Ihild, lind Ihl,d, invcslil.,,atic_n was a prcfinaitmry study to examine the effects of sweeping the

inlcr;_ctiot_ regions (m the pc:ll,, healing goner;iLeal onx the surl:_ce of tile cylinder, in addition to problems

t)l oblaining quasi-two-dimcnsiunal flows over the nH)dels, questions associated with transition of the

ht)undary layer on Ihc swept cylinder in the al3scncc of the sh()ck/shock interaction are an important

issue. The laller Iwo studies were conducled at a range of Reynolds numbers and at a freestream Mach

number of 8.

A numher of key problems must be solved belo,e a meaningful experimental study of shock/shock

interaction at hypersonic speeds can be conducted. First, a blockage-free flow between the shock genera-tor and the cylinder must be obtained while, at the same lime, preventing expansion at the trailing edge of the shock generator from influencing the shock/shock interacfion. Also, a 3-inch-diameter cylinder is required to accurately define the characteristics of the interaction regions, consistent with.the 0.010-inch

gage spacing employed with our thin-film instrumentation. These constraints required that we use a shock

generator 4 feet in lengtln and 18 inc!_es in width to nblain two-dimensional flow over the centerline of the

model. Large experimental facilities are required for such experimental studies. We designed models with

shock-generator angles between 10 and 15 degrees based on Edney's prediction technique, to give the

maximum interference healing enhancement. In the multiple-shock interaction investigation, the two

turning angles were selected s¢_ Ihal the Iotal turning angle was always between 12.5 and 13.5 degrees.

3.2 EXPERIMENTAL FACII,ITIES, MODEI,S, INSTRUMENTATION, AND FLOW

VlSUAI,IZATION

3.2.1 Experimental Facililies

'Fhe experime,_tal studies were conducted in Calspan 4g-inch and 96-inch shock tunnels at Mach

numhers of (_.5, 8.0, 11.7, 15.6, 16.3, and 18.9. The facilities and their performance are described in

Reference 43. The freestrean_ conditions at which the current experimental program was conducted are

ploltud on the map of Math number versus unit Reynolds number shown in Figure 21a. Test conditions

A, B, D, and E were obtained in the 4g-inch tunnel, and test condition C was obtained in the 96-inch

tunnel. At Mach numbers of 6.5 and g, the Reynolds numbers were sufficiently large that the interactions

generated transitional to turbulent shear layers. We obtained completely laminar interactions at Mach

numbers between I I and 19 and transitional inleraclions at Mach II and 16.5.

The oper:nion of Ihe shock lunnel in Ihe reflected-shock mode is shown with the aid of the wave

diagram ill Figure 21h. The tunnel is started by rnpluring a double diaphragm, permitting high-pressure helium in the driver section Io expand into the driven section. This generates a normal shock, which propagales through Ihe low-pressure air. A region of high-temperature, high-pressure air is produced

between Ihis normal-shock f,ont and the gas interface (often referred to as the contact surface) between the driver and driven gases. When the primary or incidenl shock strikes the end of the driven section, it is reflected, leaving a region uf ahnosl stalionary, high-I_ressure, heated :fir. This air is then expanded through a nozzle Io Ihe desired freestream condilions in the test section.

The duration of the flow in Ihe test section is controlled by the interactions between the reflected shock, the gas :interfilce, and the leading expansion wave generated by the non-stationary expansion process occurring in the driver seclion. We normally control the initial conctitions of the gases in the driver and driven sections so that the gas interface beconles Iransparent to the reflected shock interaction. This

is known as operaling under "tailored interface" c(mclilions. Under these conditions, the test time is

controlled hy the time lake,_ fl_r Ihe driver/driven interface to reach the throat, or for the leading

expan-sion wave to deplete the reservoir t_l pressure behind Ihe reflected shock. The flow duration is, therefore,

either driver-,t{as-limi!ed or cxpansio,v-limited. Figure 2 Ic shows the flow duration in the test section as a

funclion o1 the Math mlmber of the incicleul shock. Ilere, it can be seen that, for operation at low

108 107

k.-8

Lt ,,¢, 106 ¢; iJJ 105 Z ¢,o Q .J O 10 4 Z 10 3 102

L_GE 96-INCH LEG

', D " • _!00,000 " , % .12s.ooo , , _.. \ 150,000 ' _ \ :_._. /.. ;J225ooo • \ _ : :./ _/ ' 25O,0OO i _.i <

i

i :

i

;

_._>-27s.ooo

i ALTITUDE AT WHICH Re/Moo

: COORDINATES EXIST IN FRFE FLIGHT

0 4 8 12 16 20 24

MACH NUMBER

{a) PERFORMANCE MAP

TESTING

LEADING ExPANSION.,,.._..,c..WAVE 5 TIME/__

| _DIAPHRAGM STATION DISPLACEMENT--_.

MA N DIAPHRAGM

) NOZZLE -)

[OHIVEH 5EC[ION i AIR SECTION _{r_._.}

(b) WAVE DIAGRAM FOR TAILOREDINTERFACE SHOCK TUBE

CI Z

8

LM ¢/) .J .J p-IP Z I--.?,

I--\

I..

L'-,TE°

15 /. " _/i _ 20-ft DRIVER l0 _N-WAVE LIMITED

HELIUM _"AIR HELIUM DRIVER DRIVER

I I _I I J

2 3 4 5 6

INCIDENT SHOCK MACH NUMBER MI

I¢) TEST TIME AVA