c a ia a - nasa · modifications shown (rev 3.1, ypa10, and rev 8.3) are indicative of changes made...
TRANSCRIPT
C A_IA AY •
AIAA 2000-2685X-38 Experimental Aerother_ao(lynamics
Thomas J. Horvath,
Scott A. Berry, and N. Ronald Merski
NASA Langley Research Center
Hampton, Virginia
Steve M. Fitzgerald
NASA Johnson Spaceflight Center
Houston, Texas34th AIAA Thermophysics
ConferenceJune 19-22, 2000 / Denver, CO
For permission to copy or republish, contact the American Institute of Aeronautics and Astronautics1801 Alexander Bell Drive, Suite 500, Reston, VA 22091
https://ntrs.nasa.gov/search.jsp?R=20000063489 2020-01-29T17:05:01+00:00Z
X-38 EXPERIMENTAL AEROTHERMODYNAMICS
Thomas J. Horvath*, Scott A. Berry', N. Ronald Merski**
NASA Langley Research Center
and
Steve M. Fitzgeraldt*'
NASA Johnson Space Center
Abstract
The X-38 program seeks to denumstrate an autonomously returned orbital test flight vehicle to ,support the
development of an operational Crew Return Vehicle for the International Space Station. The test flight, anticipated
in 2(X)2, is intended to demonstrate the entire mission profile of returning Space Station crew members safely back
to earth in the event of medical or mechanical emergenc\v. Integral to the formulation of the X-38 flight data book
and the design of the thermal protection system, the aerothermodynamic environment is being defined through a
synergistic combination of ground based testing and computational fluid dynamics. This report provides an oveta,iew
of the hypersonic aerothermodvnamic wind tunnel program conducted at the NASA Langley Research Center in
support of the X-38 development. Gh}bal and discrete sur[ace heal transfer, force and moment, surface streamline
patterns, and shock shapes were measured on scaled models of the proposed X-38 configuration in different test gases
at Mach 6. 10 and 20. The test parametrics include angle of attack.#om 0 to 50 degs. unit Reynolds numbers from
0.3x1(_ to 16xl(l_/fi, rudder deflections of O, 2, and 5 deg. and body flap deflections from 0 to 30 deg. Results from
hypersonic aerodynamic screening studies that were conducted as the configuration evoh, ed to the present shape are
presented. Heat T gas simtdation tests have indicated that the primary real gas effects on X-38 aerodynamics at trim
conditiott_" are expected to favorably influence flap efffectiveness. Comparisons of the experimental heating and fi)rce
and moment data to prediction and the current aerodynamic data book are highlighted. The effects of dL_'rete
roughness elements on boundary laver transition were investigated at Mach 6 and the development of a transition
correlation for the X-38 vehicle is described. Extrapolation of ground based heating measurements to flight radialion
equilibrium wall temperatures at Math 6 and 10 were made and generally compared to within 50 deg F of flight
predicti{m .
Nomenclature
b_0_ reference span (in)
h heat transfer coeff. (Ibm, ft--sec), q/(Ha, ,- H,,)
where I-t_ : H, 2
H enthalpy (BTU/Ibm)
L_t reference length (in)
M Mach number
P pressure, psia
q heat transfer rate (BTU/ft-_-sec)
q dynamic pressure (psi)
R radius (in.)
t time (see)
Re unit Reynolds number (lift)
S_f reference area (in'-)
T temperature (°F)
X axial distance from origin (in.)
Y lateral distance from origin (in.)
* Aerospace Technologisl, Aerothermod_namics Branch, N\SAl,angle 3 Research Center, [Iampton, VA
t Acrospace Technologist, \croscienccs Branch. NASA Jt_hnsonSpace (2enter, l[ovston, TX
Member AIAA
(?op) right '_2000 b) the Amencan lnstilute ol Aeronantics andAstronatttics, Inc No cop} righl is asserled ill the United Slalesunder Title 17, [?S (?ode. The 1?.S. Government has a royalt 3 -free license to exercise all rights under the copyright claimedherein tot government purposes All other righls are reserved b)the cop wighl owner
ct angle of attack (deg)
6 control surface deflection (deg)
p density (Ibm/in _)
¥ ratio of specific heats
CN normal-force coefficient (N/qS_¢r)
CA axial-force coefficient (A/qS_¢f)
Cm pitching-moment coefficient (m/qL_,fS_0
C_ rolling-moment derivative (l/deg)
C._ yawing-moment derivative (l/deg)
Cv_ side-force derivative (l/deg)
L/D lift to drag ratio
Subscripts
aw adiabatic wall
BF body flap_c free-stream conditions
n model nose
t, I reservoir conditions
2 stagnation conditions behind normal shock
w wall
Introduction
The Crew Return VehiclC (CRV) envisioned by
NASA is intended to provide emergency return-to-
earth capability from the International Space Station
(ISS) in the event of medical or mechanical problems
and Shuttle non-availability. Presently scheduled to
be operational at the ISS in 2005, several CRV's are
to be constructed by industry and delivered to the
to be constructed by industry and delivered to thestation by the Shuttle. The X-38 program _ led byNASA Johnson Space Center (JSC) seeks to fly afull-scale technology demonstrator to validate keydesign and operational aspects for the CRV.
Conceived to demonstrate CRV technologies, thepresent X-38 design is considered flexible enough toevolve to a Crew Transfer Vehicle (CTV). The
potential for CRV/CTV dual use has led to acooperative NASA/European effort of the X-38design 3'4. The CTV would be integrated to anexpendable booster, such as the French Ariane 5,permitting personnel to be ferried to and from thestation.
In the near term, the X-38 CRV technologydemonstrator mission planned for 2002 calls for a28.5 ft long vehicle (designated as V201) to bereleased from a Shuttle that is positioned in a highinclination ISS orbit. Following the jettison of adeorbit engine module, the X-38 will returnunpowered (similar to the Space Shuttle) and thenuse a steerable parafoil 5, a technology first developedby the Army, for its final descent. Landing will beaccomplished on skids rather than wheels.
Consistent with the X-38 program's goal to takeadvantage of available equipment and technology toreduce vehicle development costs by an order ofmagnitude 67. the shape of the X-38 draws upon a
synthesis of work performed by the U.S. governmentand industry over the last few decades 89. The initial
X-38 shape (sketched in Fig.l at 0.0295 scale)proposed by NASA JSC was based upon a liftingbody concept originally developed and flown duringthe U.S. Air Force's PRIME (X-23/SV-5D) j°'tl andPILOT (X-24A) _zprojects in the mid-1960s and early
70's. Referred to as Rev 3.1, this lifting bodyconfiguration was initially selected by JSC for theCRV mission due to it's relatively high hypersonicL/D (higher L/D translates to larger cross rangecapability and shorter loiter times in orbit) andvolumetric efficiency (room for all station crew ifnecessary). The current shape (Rev 8.3) departs fromthe X-23/X-24A and the initial Rev 3.1 in that it
reflects changes to the vehicle upper surface toaccommodate a vehicle capable of a CTV missionshould the agency and its international partners decideto pursue this option. High approach speeds and longrollout distances associated with the low subsonic
L/D from this lifting body requires that the landing beaugmented with a steerable parafoil _. Critical forinjured or incapacitated crew, this method permits theCRV to land within close proximity of medicalfacilities with minimal g-loads.
Under the NASA/European partnership,Daussault Aviation serves as prime contractor for thedevelopment of an X-38 Aerodynamic andAerothermodynamic databases 14'_5. The role of theNASA Langley Research Center (LaRC)
Aerothermodynamics Branch (AB) has been to providehypersonic laminar and turbulent global surfaceheating and force and moment (F&M) data for CFDvalidation and as a cross check of X-38 data obtained
in European facilities. Results from early LaRC
wind tunnel heating tests on the initial Rev 3.1. .. 16 17
compared favorably to CFD computanons ' .Transition data was obtained 18 which could be
compared to similar measurements made on the
Shuttle _9 to support the use of a Re0/M e criteria t5 for
assessment of manufacturing (step) tolerances of theThermal Protection System (TPS) tiles. Hypersonicaerodynamic screening studies on Rev 3.1 wereconducted at LaRC to assess the potential for real gaseffects 2°. Since the time of these publications,
additional aerodynamic and aeroheating tests havebeen completed which include design changes to thevehicle outer mold lines. For complex thermaldesign environments on the X-38 (e.g. region behindsplit deflected body flaps), the original database hasbeen supplemented. In this area, LaRC wind tunneldata in the form of discrete measurements has served
as the primary source of heating information.
The purpose of this paper is to present anoverview of the LaRC AB experimental program tocharacterize the X-38 hypersonic aerothermodynamicenvironment. As discussed in Ref. 21, the term"aerothermodynamics" is taken to encompassaerodynamics, aeroheating, and fluid dynamics. The
experimental results were obtained in the LaRCAerothermodynamic Facilities Complex (AFC) 22.Over 1400 tunnel runs from 16 different entries in 4
facilities have been completed since May 1996. Tablel lists of all the LaRC wind tunnel tests to date in
support of X-38 aerothermodynamics. In terms ofMach and Reynolds number simulation, the mostcurrent X-38 flight trajectory (designated as Cycle 8)considered by NASA JSC (Fig. 2) has indicated thatthe 28.5-ft long flight vehicle would experiencelength Reynolds numbers (ReL) of approximately0.7x106, 3.6xl06, and 6xlO 6 at freestream Mach
numbers of 20, 10, and 6 respectively. Fig. 2 alsoindicates the corresponding range of length Reynoldsnumber that can be produced in the AFC with anappropriately sized models for each facility.Prediction along the current flight trajectory wouldplace peak heating to the stagnation point of areference sphere near Mach 23 at ct=40 deg.
Test techniques that were utilized during thesetests include thermographic phosphors and thin film
thermometry, which provide global and discretesurface heating respectively; oil-flow, which providessurface streamline information; schlieren, which
provides shock details; and a six component straingage balance to provide aerodynamic force andmoment loads. Parametrics from the tests conducted
in 4 facilities were Mach numbers of 6,10, and 20,
normalshockdensityratiosof 4 to 12producedinthreetestgases;arangeof angleof attack,0 to50deg.;unitReynoldsnumberof 0.3to 16million/ft);andbodyflapdeflectionsfrom0to 30degin2.5degincrements.
Experimental Methods
Models
The evolution of the X-38 Outer Mold I_ines
(OML) to the present shape (Rev 8.3) is highlightedin Fig. 3, a photograph of the various 0.0175 modelscale force and moment models. All proposedvehicle configurations have incorporated symmetricfins with deflectable rudders and two body flaps thatdeflect away from the windward fuselage foraerodynamic control. The progression of OMLmodifications shown (Rev 3.1, YPA10, and Rev 8.3)are indicative of changes made to the upper surface (orleeside); the windward surface has essentially retainedthe same basic shape as the SV-5D lifting bodiesflown in the 1960's. Two basic model lengths (6 andlO-inch), determined by facility test core size, yieldedmodel scale factors of 0.0175 and 0.0295.
As the LaRC test results were to providebenchmark data, model mold line accuracy relative to
the CFD surface grid definition was important. Thistype of quality control was also necessary due to thecomplimentary tests conducted in Europe on separatemodels. To assure precision, numerically controlledmilling machines were utilized in the fabrication ofthe X-38 family of force & moment wind tunnel
models utilizing CAD geometry sul_l_lied by JSC.. . • .-t.
Rapid prototypmg/castmg techmques- - were used
for construction of the heat transfer models to providean early assessment of the heating environment.
Forces and moments: Initial aerodynamicscreening studies on X-38 conducted in the unheated22-in Mach 20 Helium tunnel utilized resin SLA
models for testing as temperature was not an issue.Two 6-in. (0.0175 scale) resin models were fitted
internally with a steel sleeve in order to accept abalance. The models were designed to accept bodyflaps of various deflections.
Later studies on Rev 3. I and Rev 8.3 conducted
in the heated tunnels of the Aerothemodynamic
Facilities Complex required the use of metallicmodels. The bodies of these two 6-in. models (Rev3.1 and Rev 8.3 ) were fabricated from aluminumwhile the nose, fins, and control surfaces were 15-5stainless steel. As with the earlier resin models, the
metallic models incorporated multiple control surfacesettings, and internally fitted with interchangeablesteel sleeves to accept several different balances.
Heat transfer: Two types of heat transfermodels were fabricated: (1) a series of cast silica
ceramic models for global information and (2) twothin film models providing discrete measurements.At LaRC, the global phosphor technique has largelyreplaced the thin film technique due to the dramaticreduction in model fabrication time andinstrumentation costs. While these benefits are
revolutionary, the technique can have limitationswhere fast (sensor) response times are desired or the
optical view of the model surface is limited. In thepresent X-38 heating studies, the required optical viewbehind the deflected body flaps was limited. Thinfilm instrumentation was used to compliment the
global technique during this phase of testing.
Global phosphor thermography: Over 40
cast ceramic models were fabricated in support of theLaRC X-38 aerothermodynamic program, all ofwhich share a common construction technique. Arapid prototyping technique was first used to build aresin stereolithography (SLA) model with various,detachable body flaps on both the port and s_region of the base of the vehicle. The SLA resinmodel was then assembled with the desired control
surface settings and served as a pattern to constructmolds from which the cast ceramic model
configurations were made. A magnesia ceramic wasused to backfill the ceramic shells, thus providingstrength and support to the sting support structure.
A photograph of six ceramic 0.0295 scale ( 10-in.) Rev 3.1 model configurations with various bodyflap deflections are shown in Fig.4. Typically, twocasts of each configuration were made: the primarybeing immediately prepared for testing and the back-up shell held in reserve, in case of problems with theprimary. In order to obtain accurate heat transfer datawith the phosphor technique, the models are cast witha material with low thermal diffusivity and welldefined, uniform, isotropic thermal properties. Thephosphor coatings typically do not requirerefurbishment between runs in the wind tunnel ;_1
have been measured to be approximately 0.001 inchesthick. Details concerning the model fabricationtechnique and phosphor coating can be found inRefs. 18 and 24. Fiducial marks were placed on themodel surface to assist in determining spatiallocations accurately.
Once the phosphor testing was completed, theuntested backup models were prepared (spray-coatedand kiln fired with a thin black glazing) for use as oil-flow and schlieren models.
Discrete thin fibn: Two metallic models wereconstructed and fitted with a machinable ceramic
insert instrumented on the surface with smallresistance thermometers. The first model, cast withaluminum, was created from a mold that used a resin
SLA model as a pattern. This SLA resin model wasidentical to the pattern used to construct the castceramic phosphor heating models. The second model
wasuniquein thesensethatit was the first rapidprototype, selective laser sintered metallic modelsuccessfully tested in a I.aRC wind tunnel. Incontrast to the more conventional rapid prototypingtechniques that utilize resins or wax to create a patternfrom which the actual tunnel model is cast, this
technique omits the intermediate step by "building"the model, layer by layer, using a metallic powder.Originally, this model was intended as a backup tothe cast aluminum model. Surface verificationmeasurements indicated that the laser-sintered model
was superior to the cast aluminum model in terms oflinear shrinkage. Based on these results it was decidedto instrument and test this model as well.
A photograph of the 0.0295 scale (10-in.) Rev8.3 thin film model installed in the Mach 6 air tunnel
is shown in Fig.5. The Cavity region behind thedeflected body flaps was instrumented (see Fig.5ainset) with thin film gages to characterize the localheating in a region with restricted optical access.Sixty-eight sensors were placed at predeterminedlocations on the Macor (trademark of Coming GlassWorks) substrate, Fig.5b.
Facility Descriptions
The Aerothermodynamic Facilities Complex(AFC) consists of five hypersonic wind tunnels thatrepresent two-thirds of the nation's conventionalaerothermodynamic test capability. Collectively,they provide a wide range of Mach number, unitReynolds number, and normal shock density ratio) 22
This range of hypersonic simulation parameters isdue, in part, to the use of three different test gases(air, helium, and tetraflouromethane), thereby makingseveral of the facilities unique national assets. TheAFC facilities are relatively small and economical tooperate, hence ideally suited for fast-pacedaerodynamic performance and aeroheating studiesaimed at screening, assessing, optimizing, and bench-marking (when combined with computational fluiddynamics) advanced aerospace vehicle concepts andbasic fundamental flow physics research.
20-1nch Math 6 Air Tunnel: Heated, dried,
and filtered air is used as the test gas. Typicaloperating conditions for the tunnel are: stagnationpressures ranging from 30 to 500 psia; stagnationtemperatures from 760-deg to 1000-degR; ardfreestream unit Reynolds numbers from 0.5 to 8million per foot. A two-dimensional, contourednozzle is used to provide nominal freestream Machnumbers from 5.8 to 6.1. The test section is 20.5 by20 inches; the nozzle throat is 0.399 by 20.5-inch. Abottom-mounted model injection system can insertmodels from a sheltered position to the tunnelcenterline in less than 0.5-sec. Run times up to 15minutes are possible with this facility for F&M andpressure measurements. For the heat transfer and
flow visualization tests, the model residence time in
the flow is only a few seconds.
20-1nch Mach 6 CF4 Tunnel: Heated,
dried, and filtered tetrafluoromethane (CF4) is used asthe test gas. Typical operating conditions for thetunnel are: stagnation pressures ranging from 85 to2000 psia, stagnation temperatures up to 1300R, andfreestream unit Reynolds numbers from 0.01 to 0.3million per foot. A contoured axisymmetric nozzle isused to provide a nominal freestream Mach numbersfrom 5.9 to 6.01. The nozzle exit diameter is 20
inches with the flow exhausting into an open jet testsection; the nozzle throat diameter is 0.466-inch. A
bottom-mounted model injection system can injectmodels from a sheltered position to the tunnelcenterline in less than 0.5-sec. Nominal run time for
F&M testing is approximately 20 seconds in thisfacility.
31-1nch Mach 10 Air Tunnel: Heated,
dried, and filtered air is used as the test gas. Typical
operating conditions for the tunnel are: stagnationpressures ranging from 150 to 1350 psia; stagnationtemperatures from 1750-deg to 1850-degR; andfreestream unit Reynolds numbers from 0.25 to 2million per foot. A three-dimensional, contourednozzle is used to provide nominal freestream Machnumber of 10. The test section is 31 by 31 inches;the nozzle throat is 1.07 by 1.07-inch. A side-mounted model injection system can insert modelsfrom a sheltered position to the tunnel centerline inless than 0.5-sec. Run times up to 1.5 minutes arepossible with this facility, although for heat transferand flow visualization tests, the model residence timerequired in the flow is only a few seconds.
22-hwh Math 20 Helium Tunnel: Heated
or unheated, dried, purified, and filtered helium is usedas the test gas. Typical operating conditions for the
tunnel are: stagnation pressures ranging from 300 to3300 psia; stagnation temperatures from 490-deg to900-degR; and freestream unit Reynolds numbersfrom 2.5 to 22 million per foot. A contouredaxisymmetric nozzle is used to provide a nominalfreestream Mach numbers from 18.1 to 22.3. Thenozzle exit diameter is 22 inches with the flow
exhausting into an open jet test section; the nozzlethroat diameter is 0.622-inch. A bottom-mounted
model injection system can insert models from asheltered position to the tunnel centerline in less than0.5-sec. Run times up to 30 seconds are possiblewith this facility.
Test Conditions and Setup
Nominal reservoir and corresponding freestream flow conditions for the four tunnels are
presented in Table 2. The freestream properties weredetermined from the measured reservoir pressure mdtemperature and the measured pitot pressure at the testsection (or inferred from previous calibrations). Test
4
section wall static and pitot pressures were monitored
if possible and compared to tunnel empty conditionsto assess if model blockage effects existed. No
significant differences in pitot pressure were measuredand it was concluded that significant blockage did notexist. The ratio of projected model frontal area(or=40 deg) to tunnel cross sectional area for the0.0175 scale model was less than 0.1.
All heating and force and moment models weresupported by a base mounted cylindrical sting withthe exception of the two thin film, heat transfermodels which were blade supported from the leeside.This was done to minimize possible supportinterference associated with the flap cavitymeasurements. Details of the X-38 ceramic heattransfer model installation in the NASA LaRC 20-Inch Mach 6 Air Tunnel can be found in ref 18.
Test Techniques
Forces and Moments: Aerodynamic force
and moment loads were measured throughout the testprogram using several sting-supported, six-component, water-cooled, internal strain gagebalances. Balance temperature was monitored usingintegrated water jacket thermocouples to ensureexcessive thermal gradients did not develop during therun. Typically, the aerodynamic models were testedin the inviscid test core flow on the tunnel centerline.
In the CF 4 tunnel, the model was locatedapproximately 1.O-inches downstream of the nozzleexit and laterally displaced 4-inches from the tunnelcenterline to avoid small disturbances that are
characteristic in axisymmetric nozzles. Limited testsmade with the model on tunnel centerline did not
indicate any measurable effect on the aerodynamiccharacteristics of the present configuration over the
range of angle of attack tested.
Phosphor Thermography: Advances in
image processing technology which have occurred inrecent years have made digital optical measurementtechniques practical in the wind tunnel. One suchoptical acquisition method is two-color relative-intensity phosphor thermography -'5-'7which has beenutilized in several aeroheating tests conducted in thehypersonic wind tunnels of NASA Langley ResearchCenter. 19'zs'29 With this technique, ceramic windtunnel models are fabricated and coated with
phosphors that fluoresce in two regions of the visiblespectrum when illuminated with ultraviolet light.The fluorescence intensity is dependent upon theamount of incident ultraviolet light and the localsurface temperature of the phosphors. By acquiringfluorescence intensity images with a color videocamera of an illuminated phosphor model exposed toflow in a wind tunnel, surface temperature mappingscan be calculated on the portions of the model that arein the field of view of the camera. A temperaturecalibration of the system conducted prior to the study
provides the look-up tables that are used to convertthe ratio of the green and red intensity images toglobal temperature mappings. With temperatureimages acquired at different times in a wind tunnelrun, global heat transfer images are computedassuming one-dimensional semi-infinite heatconduction. The primary advantage of the phosphortechnique is the global resolution of the quantitativeheat transfer data. Such data can be used to identifythe heating footprint of complex, three-dimensionalflow phenomena (e.g., transition fronts, turbulent
wedges, boundary layer vortices, etc.) that areextremely difficult to resolve by discrete measurementtechniques. Because models are fabricated andinstrumented more rapidly and economically, globalphosphor thermography has largely replaced discreteheating instrumentation in Langley's AFC.
Thin Fihn: Thin film resistance gage were
used to infer convective heating in the region behindthe deflected body flaps. Standard mechanicaldeposition techniques 3° developed at LaRC were usedto fabricate the 0.030-in. by 0.040-in. platinumsensing elements. Surface temperatures wereintegrated over time to determine the local heattransfer rate using the IDHEAT code developed byHoilis 31. Both analytical 3-'-_3 and numerical finite-volume heat transfer models are incorporated into thiscode. The analytical solutions are derived from one-dimensional, semi-infinite solid heat conduction
theory with the assumption of constant substrate(model) thermal properties. When using theanalytical option the inferred heating rates areempirically corrected for the effects of variable modelthermal properties. For the present study, theuncertainty associated with variable wall thermal
properties is believed to be minimal, particularly inthe flap cavity region where surface temperatureincreases of 5 deg F or less were measured.
Flow Visualization: Flow visualization in
the form of schlieren and oil-flow techniques, wasused to complement the surface heating and force &moment tests. The LaRC 20-Inch Mach 6 air and
CF 4 Tunnels are equipped with a pulsed white-light,Z-pattern, single-pass schlieren system with a field ofview encompassing the entire test core. Images wererecorded on a high-resolution digital camera, enhancedwith commercial software and electronically placed
into this report. The 31-Inch Mach 10 air and 22-Inch Mach 20 He Tunnels do not have schlieren
systems.
Surface streamline patterns were obtainedusing the oil-flow technique. Backup ceramic modelsor the metallic force models were spray-painted blackto enhance contrast with the white pigmented oilsused to trace streamline movement. A thin basecoat
of clear silicon oil was first applied to the surface, and
5
then a mist of pinhead-sized pigmented-oil drops wasapplied onto the surface. After the model surface wasprepared, the model was injected into the airstreamand the development of the surface streamlines wasrecorded with a conventional video camera. The
model was retracted immediately following flowestablishment and formation of streamline patterns,and post-run digital photographs were taken.
Data Reduction and Uncertainty
A 16-bit analog-to-digital facility acquisitionsystem acquired flow condition data. Measured values
of Pt, l and Tt, l are believed to be accurate to within
±2 percent.
Heating rates were calculated from the globalsurface temperature measurements using one-dimensional semi-infinite solid heat-conduction
equations, as discussed in detail in Ref. 27. As
discussed in Ref. 27, the accuracy of the phosphorsystem measurement is dependent on the temperaturerise on the surface of the model. For the windward
side heating measurements, the phosphor systemmeasurement accuracy is believed to be better than--8%, and the overall experimental uncertainty of thephosphor heating data due to all factors is estimatedto be -+15%. In areas on the model where the surface
temperature rise is only a few degrees (i.e. leeside orflap cavity), the estimated overall uncertainty is onthe order of --.25%. Repeatability for the normalizedwindward centerline (laminar) heat transfermeasurements was found to be generally better than±4%.
Based on the analysis of Refs. 27 and 34, thediscrete thin film heat transfer measurements are
believed to be accurate to within ±8 percent.Repeatability for the cavity heat transfermeasurements was found to be generally better than±2 percent.
In general, aerodynamic data was obtained in adescending alpha sweep during each run to minimizeerrors associated with balance heating at the morerelevant hypersonic entry angles-of-attack. In CF,,two separate runs were required to complete an angle-of-attack sweep due to the short run time. The datawas collected by a analog-to-digital data acquisitionsystem and averaged over a one second interval foreach angle of attack (model held at fixed angle ofattack for approximately 5 sec). The raw data wastransferred to a Hewlett-Packard 9000 computer for
data reduction and storage. During data reduction,corrections for weight tares, sting deflections, andbalance interactions were made.
The force and moment data measured at thebalance electrical center has been transferred to a
moment reference center located at 57 % along themodel x-axis. The model outer mold lines werechecked and transfer distances were inferred from
measurement by the surface verification laboratory.Table 3 lists the reference area and lengths used tocalculate the aerodynamic coefficients.
Base pressure measurements were made but wereonly used to provide information to help assesspotential interference effects that may be present dueto the sting/support system. All axial forcecoefficients, CA, are reported as uncorrected for basepressure.
The estimated uncertainty in the reportedaerodynamic coefficients was determined and reported -'°using the small sample method presented by Kiineand McClintok _5. Where appropriate, estimated errorsin the aerodynamic coefficients are indicated on thefigure legend symbols.
Prediction Mqthgd
X-38 heating computations for selected angles-of-attack and test conditions were performed byseveral organizations within the Europeancomputational community. An overview of thecomputational methodology for X-38 developmenthas been provided in Refs.14 andl5. Because of thebroad range of CFD codes presently being used toprovide wind tunnel predictions, it was consideredimpractical to present details associated with surface
and volume grid topology, grid sensitivity studies,and turbulence models in this experimental overviewpaper. All comparisons to CFD predictions wereobtained from the X-38 aerothermodynamic datat_semanaged by NASA JSC. Every attempt was made toprovide an appropriate reference to the original sourceof each numerical solution if the work was publishedin the open literature. In some cases, a referencabiedocument was not available, and these cases are listed
as unpublished.
Results and Discussion
Preface
As an overview paper of NASA LaRC ABcontributions to X-38 aerothermodynamics, thissection will highlight some of the more relevantobservations to date. Details of how the presentresults are integrated into the JSC/European X-38design methodology can be found in Refs. 14 and 15.First, the early aerodynamic screening exercises arereviewed and the data compared to the SV-5Dpreflight datatxase. The influence of Mach number,Reynolds number, and gamma (real gas effect) on X-38 hypersonic aerodynamics is discussed. Pitchingmoment data has been presented around a momentreference center located at 57% of the body (reference)length. The aerodynamic results are followed by asynopsis of both global and localized heatingmeasurements including some boundary layer
6
transition results, and an extrapolation of wind tunnelmeasurement to flight. Flow visualization in theform of surface oil flows is presented when necessaryto assist in the interpretation of both force anclmoment and surface heating data. Heating mappingsand distributions are presented in normalized form.
For the global images, a constant color bar maximumvalue was selected for data presentation (except wherenoted) to maintain consistency when viewing orcomparing the images. On the contour scale, thecolors tending towards red indicate areas of higherheating (temperatures) while the colors towards bluerepresent areas of lower heating. In areas where thelocal heating exceeded the selected maximum colorbar value, such as the deflected body flaps or finleading edges, a gray "overscale" will be evident.
Aerodynamics
The initial hypersonic screening conducted inthe Mach 20 Helium tunnel focused on the rapid
aerodynamic assessment of a modified SV-5D vehicleshape (Rev 3.1) using stereolithography (SLA)models. Parametrics were confined to modifications
to the vehicle leeside and fin leading edges and asexpected, produced no measurable effects on
longitudinal aerodynamics at entry angles of attack(not shown). The Helium tests were then expanded inscope to evaluate control surface effectiveness over arange of body flap deflections and to permitcomparison to the PRIME SV-5D preflight windtunnel database3X(comparison of the Mach 20 Heliumaerodynamic coefficients with computationalprediction is presented in Ref. 17). While real gaseffects on aerodynamics were not simulated in thepresent laminar Helium test results at Mach 20, thevehicle trim characteristics and L/D are in close
agreement with values incorporated into the 1960'sdatabase as shown, Fig. 6a-b. This agreement withthe SV-5D database was not surprising as real gaseffects were never quantified and incorporated into theoriginal SV-5D preflight aerodynamic database.From a facility perspective, the quantification of realgas effects on aerodynamics was in its infancy, and
predictive tools were not yet available during theoriginal SV-5D development.
The role of LaRC in X-38 aerodynamic testingwas broadened to include tests in other LaRC
facilities (e.g. Unitary Plan Wind Tunnel, 16-footTransonic Tunnel, and the Low Speed Spin Tunnel)
to compliment tests conducted in Europe. In thehypersonic regime, the focus of this overview paper,the Rev 3.1 configuration was found to belongitudinally stable; trimming hypersonically at_t _- 40 deg with approximately 15 deg body flaps anda corresponding L/D =0.9. This configuration alsoexhibited positive directional control across themeasured angle-of-attack range ( 10 <_t < 50 deg.).The effect of the primary simulation parameters onlongitudinal aerodynamics (CA , Cx , and C0, ) was
performed by testing of the samemodel/balance/support system combination in theLaRC AFC. Viscous effects (Reynolds number)associated with a laminar boundary layer weredetermined at Mach 6. Compressibility effects (Machnumber) were determined with comparisons of Mach6 and 10 air data. Similarly, comparison ofaerodynamic measurements between Mach 6 air andCF4, provided an indication of the significance of realgas effects for X-38 through the variation in normal
shock density ratio produced with the two testmedium.
The effects of Reynolds number on Rev 3.1
longitudinal aerodynamics at Mach 6 for 25 deg bodyflap deflection is presented in Figs 7. Normal force,Fig. 7a, coefficient remained essentially unchangedover the Reynolds number range while pitchingmoment, Fig. 7c, indicated a nose-down incrementwith increasing Reynolds number which suggestedenhanced flap effectiveness. As Reynolds numberwas increased, it was found that the extent of
separated flow on the control surface (as inferred fromsurface oil flow visualization-not shown) decreased as
was indicated by the forward movement of thereattachment line on the flap. The presence of flowseparation in the vicinity of the deflected controlsurfaces at all Reynolds numbers suggested aturbulent flow was not obtained at these tunnel
conditions and a limiting case for flap effectivenesswas not achieved (balance load limits were exceeded athigher Reynolds numbers). For Mach 6 flight it isanticipated that boundary layer upstream of thecontrol surfaces will be turbulent which may furtherimprove flap effectiveness relative to the laminarwind tunnel results. The decrease in axial coefficient,
Fig. 7b, can be attributed to the expected decrease inskin friction with increasing Reynolds number. Atincidence angles more typical of hypersonic entry,crossrange performance in terms of L/D (not shown)was essentially unaffected over this same laminarReynolds number range.
Compressibility effects on Rev 3.1 laminarlongitudinal aerodynamics, particularly at hypersonicentry angles-of-attack, were generally within theexperimental uncertainty of the measurements (Fig.8) with the exception of the Mach 20 axial coefficient(Fig. 8a)obtained in the Helium Tunnel. The lowMach 20 values of CA relative to Mach 6 and 10 aremore likely due to the factor of 6 difference inReynolds number between the Mach 6 and 10 air dataand the Mach 20 Helium tunnel than a
compressibility effect. In terms of pitching moment,Fig. 8c, it is interesting to note a slight crossover inthe coefficient between Mach 6 and 10 near
¢t = 35 deg. This same trend has been observed bothexperimentally _9and computationally _ within the X-33 program. While a variety of suggestions" havebeen offered to explain this trend of decreasing
stabilitywithdecreasingMachnumber,a consistentexplanationhasnotbeendetermined.It shouldbenotedthatthesedifferencesinX-38C,,atMach6 and10arewithintheuncertaintyof themeasurementandmorethansufficientcontrolauthorityis producedbythedeflectedflapstotrimthevehicle.
It was recognizedearly in the X-38developmentphasethat,realgaseffectswereneverquantifiedandincorporatedinto theoriginal SV-5D
preflight aerodynamic database. That is, hightemperature effects due to dissociation could not beaccurately determined (at that time) for the SV-5Dbasic body pitching moment and body flapeffectiveness. Two decades later, the Shuttle Orbiter
would experience a significant nose-up pitchingmoment increment relative to pre-flight predictionsresulting in body flap deflections of twice the amountnecessary to achieve trimmed flight. Thisphenomenon was later accurately simulated in theLangley CF 4 Tunnel and was coupled withcomputational methods to provide a high degree ofconfidence in estimating hypersonic entryaerodynamics 42. It is commonly recognized today that
the primary effect of a real gas on aerodynamics is tolower the specific heat ratio (,/) within the shock
layer which in turn will produce a greater degree offlow compression and expansion relative to a perfectgas. Thus, expansion surfaces will have acorrespondingly lower surface pressure. Because ofthe Shuttle experience and the presence of a windwardexpansion surface (boattail) on the aft end of the Rev3.1 (with the flaps stowed at 0 deg deflection) it wassuggested that aerodynamic real gas simulationtesting (similar to that conducted on the Orbiter post-flight) be performed. The resulting aerodynamicmeasurements obtained in air and CF4 at identicalMach and laminar Reynolds number, Fig.9, indicatethat testing in a heavy gas (CF4) resulted in smalldecreases in normal and axial force coefficients,
Fig.9a-b, and a corresponding nose-up pitchincrement for body flap deflection 0 (leg, Fig.9c.This trend in the basic body pitching moment(neutral control surface deflection) was also noted intests conducted on the Orbiter 42.
The similarity in real gas trends between theX-38 and the Shuttle ended when the flaps weredeployed. With the Shuttle, the relative nose-upincrement between air and CF4 persisted with thebody flap deflected for trim 42. This was not evident
from the present X-38 Rev 3. l test series which haveindicated that for deflections larger than 15 degrees arelative nose-down increment in pitching momentwas present, Fig. 9c. This is best explained whengeometry differences of the respective windwardsurfaces are recognized. The Orbiter has a windwardexpansion surface that begins and coincides with thelargest planform area; deflecting the body flap doesnot alter the expansion surface as the flap hangs off
the Orbiter base. In contrast, the X-38 flap hinge lineis located farther forward on the body with the controlsurfaces deflecting away from the fuselage. Unlike theOrbiter, the windward expansion surface (boat-tail)found on the X-38 is effectively eliminated for bodyflap deflections greater than 15 deg. Thus, theprimary real gas effects on X-38 aerodynamics at trim
conditions are expected to influence flap effectiveness.Relative to laminar perfect gas results the
heavy gas simulation tests revealed an increase inbody flap effectiveness across the angle-of-attackrange, Fig.10. Real gas flight computations fromRef. 17 (not shown) for 25 deg body flap indicatedhigher pressure coefficients at the nose and body flaprelative to perfect gas calculations. The pressureincrease was more substantial at the flap and effected alarger area, which resulted in a net nose downincrement. It is of interest to note that Ref. 10 has
indicated that actual flight pressures on the SV-5Dboattail expansion surface during hypersonic entrywere lower than that obtained in wind tunnel tests,
and that flight trim flap deflections were less thanpredicted by the preflight data base. The lower flightpressure on the boattail surface is consistent with areal gas effect and the increased body flap effectivenessin flight are consistent with trends observed from thepresent heavy gas wind tunnel tests and
computational prediction. In general, the incrementsand trends provided by real gas simulation tests in airand CF4 are applicable to flight provided that (1) thevehicle aerodynamics are dominated by the windwardsurface, (2) ¥ within the flight windward shock layer
does not significantly vary spatially, and (3) ¥within the flight windward shock layer is close inmagnitude to that produced in CF4 (¥=l.1).
Changes to the vehicle outer mold linescontinued as the shape was optimized for performanceacross the speed range. These modifications havebeen confined to the leeside and it is assumed that
these changes will not result in a significant departurefrom the real gas effects previously discussed. LaRCfacilities have been utilized to assist in the evaluation
of these changes on aerodynamics from transonic tohypersonic speeds. Tests at Mach 6 conducted on aninterim OML designated YPAIO (between Rev3.1 and8.3) revealed marginal lateral stability at lowerincidence angles Fig.l la-c. In-situ modelmodifications resulted in a laterally stable vehicleacross the angle-of-attack range. As expected, theaddition of an ISS docking mechanism to the Vehicle(Rev 8.3) leeside did not produce any measurableeffects in the longitudinal or lateral/directionalaerodynamics (not shown) at Mach 6 for15 >ct > 50 deg.
The program has frozen the OML lines(Rev8.3) and is presently focusing efforts ondeveloping and refining an aerodynamic design databook for flight. The most current X-38 Aerodynamic
DataBook43(ADB-RevG)providesa singlesourcereferenceforall X-38vehicledataappropriateto fullscaleflightperformance.Aerodynamics in the databook up to Mach 10 are based on wind tunnel tests of
various scale models tested in Europe. Included in thebook are aerodynamic body flap control effectivenessdata across the Mach range, which are presented interms of increments from an undeflected state.
Laminar body flap pitching moment increments.(ACm), from LaRC tests at Mach 6 and 10 at a lengthReynolds number of 1 x 10° are presented, Fig. 12a-b, along with the current ADB (Rev G) values.Differences between the LaRC results and the ADB
(Rev G) are small and are believed to be within theuncertainties that will be allowed for by the X-38
flight control system.
Windward Surface Heatina
Flight surface heating data from the SV-5DPrecision Recovery Including Maneuvering Entry(PRIME) project of sufficient quality and quantity isnot available for X-38 TPS design. The primaryobjective of the PRIME program was to demonstrate,through flight testing, lifting body aerodynamicperformance during hypersonic entry at crossranges upto 700 miles. With the emphasis placed onaerodynamics, the flight program did not attempt toproduce a large flight heating data base 44 (heating overmost of the body was inferred from flight pressuremeasurements). Ground based tests conducted at the
time proved to be sufficiently accurate for aconservative heatshield design _t. Lightweightablative materials were used over the entire surface as
anticipated heating conditions were considered to bemore extreme than those expected for largeroperational vehicles (such as the Shuttle flownalmost 20 years later).
In terms of experimental and predictivemethods, the aerothermodynamic community hasprogressed considerably since the development of theSV-5D vehicles. Up until the mid 1990's, however,aeroheating information for configuration assessmenthad continued to lag behind aerodynamic informationdue primarily to model and instrumentationcomplexities associated with aerothermodynamictesting. The X-38 program was able to takeadvantage of recent developments in the two-colorglobal phosphor thermography technique, providingan opportunity to conduct an aerothermodynamicscreening/trade study concurrent with aerodynamictests. The aeroheating measurements from LaRCwere primarily used in the continued development andvalidation of computational tools used to predict theX-38 aeroheating environment. Similar to theaerodynamic methodology of the program, the LaRCresults were also intended to duplicate or complimenttest results obtained in European facilities.
Fig. 13 illustrates a typical global comparisonbetween experimental and numerically predictedlaminar heating j7 at wind tunnel conditions. Thecomparisons suggest a high level of confidence in thelaminar numerical simulation over the acreage of thewindward surface where the flow remains attached.
Differences in the magnitude of the deflected flapsurface heating and the apparent size of the separatedflow upstream are indicative of a flow complexitythat is more challenging to simulate numerically.
Extracted heating distributions for this condition nearthe windward centerline and two axial stations are
presented, Fig.14a-c, and are compared to laminarGASP prediction _7. Measured centerline SV-5D
heating data _5 from wind tunnel tests conducted atAEDC Tunnel C at Mach 10 and the Martin Hot
Shot tunnel at Mach 20 over 35 years ago are alsoshown with the present LaRC Mach 10 results andprediction in Fig. 14a. Differences between the SV-5D data sets from Ref. 45 were noted at that time butwere never resolved (while state of the art at the time,
large uncertainties in the SV-5D data are presumeddue to the nature of the thin skin calorimetry testtechnique). The SV-5D data sets bracket the presentMach 10 experimental and computational results. Inretrospect, the SV-5D heating environmentdetermined experimentally was adequate for aconservative TPS design. Modern quantitative globalcapabilities (experimental and computational)available today provide orders of magnitude moreinformation in a fraction of the time permitting a lessconservative design.
While primarily intended to provide data forcomputational validation, the LaRC tests were alsointended to quantitatively assess the effects ofReynolds number, angle-of-attack, boundary layertransition, and configuration changes on heating.Subsonic/transonic aerodynamic optimization did leadto leeside OML changes and as a result some effortwas devoted to the assessment of these changes on theheating environment. No significant issues regardingleeside heating (i.e. canopy, docking ring, or aftflare) were identified experimentally.
On the windward surface, the influences of
Reynolds number were most pronounced in thevicinity of the deflected flaps. Rev 3. I windwardheating at M_=6, tx=40 deg and body flapdeflection=25degare presented in Fig. 15 for a rangeof Reynolds number. The extracted longitudinaldistributions are taken just off centerline so as tocapture flapheating trends (heating to the cavity floorbetween the flap-split gap will be discussed at a laterpoint). The collapse of the heating distributions withReynolds number upstream of flap flow separationindicated the approaching flow was laminar. Thecorresponding windward global heating images andsurface streamline patterns are shown, Fig. 16a-cFig. 17a-c, respectively. Consistent with conclusions
9
inferred from the aerodynamic results, the extent offlow separation diminished with increasing Reynoldsnumber. The flow reattachment downstream on the
flap was observed to be in close proximity and nearlyparallel to the 10 deg inboard swept flap hingeline.The flap hingeline gap is presently designed to besealed to prevent circulation of this high-energy flow
into the cavity. Tests are being planned to providemore detailed heating in this area. Near reattachmentthe streamlines are highly three-dimensional; inflowtowards the flap split gap and expansion over theoutboard flap edge was evident over the range ofReynolds numbers and at all angles-of-attack. The
variation of heating levels on the flap with Reynoldsnumber suggested a transitional/turbulent flowreattachment process. Flap heating will be discussedin more detail in a subsequent section.
Fi n/Rudder Heati ng
The circulation of separated flow upstream of
reattachment appeared to result in outboard flowspillage onto the nearby fin. Rev 8,3 heating images(note scale change), Fig. 18a-c, obtained at Mach 6on the fuselage side (c_--40 deg, body flap deflection =20 deg) indicated possible boundary layer transition ofthis entrained flow up onto the rudder surface. Acomparison of extracted data along the fin chord in thevicinity of the rudder from the present tests with datafrom Ref. 45 is shown in Fig. 19. Boundary layertransition in the present tests has been inferred fromthe heating increase observed on the fin withincreasing Reynolds number. Transitional/turbulentflow on the fin was not reported from the Mach 20ground based wind tunnel test supporting the SV-5Ddevelopment _5 and was likely due to boundary layerstabilization at high Mach number and low Reynoldsnumber. Tripping of the flow with discreteroughness elements suppressed the measured finheating, Fig. 20, from transitional to turbulentlevels. When present, transitional heating on the finrudder is 4 or more times the laminar value. A
comprehensive numerical analysis of the heating inthe rudder/fin gap "_ has indicated that a seal (or
alternatively, an ablator material) may be required. InRef. 46 it was shown that high heating on the rudderoccurred at Mach 17.5 and 11 which conservativelyassumed laminar and turbulent flow over the fin,respectively. The implications of transitional orturbulent flow at the higher Mach number have notbeen assessed. Additional LaRC tests at Mach 10 are
planned to determine the susceptibility of the fin toboundary layer transition at Mach 10.
Windward Flap Heating
The thermal environment associated with the
X-38 body flaps is considered to be a challenge from adesign perspective due to the complex three-dimensional flowfield and resulting high surfacetemperatures anticipated in flight. The windward flap
temperatures in flight will be driven by severalfactors; three- dimensional flow separations, shearlayer transition, multiple shock processing of theflow (bow, separation, reattachment), and flowexpansion and acceleration over the flap edges and
through the split gap. The X-38 flaps are designed asa hot structure 47 and will be manufactured from
C/SiC, a ceramic matrix composite _ (CMC). Earlyestimates of flap thermal loads suggested that CMCtechnology could provide an adeq_te thermal margin.Vehicle weight growth and trajectory refinementshave significantly reduced this margin. Acomprehensive computational and experimental efforthas been initiated to more accurately predict theheating environment associated with the windwardsurface of the deflected body flaps and to insure thismargin is not exceeded.
The peak heating to the deflected body flaps islargely determined by the state of the sepm-ated flowas it reattaches on the control surface. Three
situations may arise: (1) laminar separation withlaminar reattachment, (2) laminar separation withtransitional or turbulent reattachment, and (3)
turbulent separation with turbulent reattachment.Limited testing in Mach 10 with smaller scale Rev
3.1 heating models has suggested that laminarconditions prevailed at and downstream of flapreattachment. Comparison of the present LaRCMach 10 heating extracted along the flap chord nearthe centerline (ct=40 deg, flap deflection of 20 deg)with laminar SV-5D data from Ref. 45 and
unpublished laminar Navier-Stokes prediction fromthe CEVCATS code z9 is presented in Fig.21. Theagreement with the experimental data from Ref. 45 isnoteworthy considering the state of the art ininstrumentation, signal conditioning, and datareduction in the 1960's. The laminar Navier-Stokes
solution underpredicted the measured flap heatingwhich has suggested that either the numericalsimulation of the laminar flow separation andreattachment process was inadequately modeled, orthat nonlaminar conditions prevailed in the Mach 10ground based tests.
Laminar flap heating was not identified atMach 6 for all Reynolds number conditions and it isbelieved that the flap heating was indicative oftransitional or turbulent flow reattachment. TheLaRC X-38 Rev 3.1 and 8.3 data was instrumental in
developing flap heating augmentation and flightscaling factors for separating transitional or turbulentflows. The effects of flap deflection on Rev 3.1global heating images obtained at Mach 6, (t=40 deg,and Re_t, = 2x106 are shown, Fig.22a-c. Extractedflap span heating distributions (X/L=0.98) arepresented, Fig.23. The heating distributions alongthe body fuselage and flap chord (y/b=0.2) are shown.Fig.24, and correspond to the deflections shown inFig. 23. The images at these conditions (as well as
10
all others) did not reveal the presence of Gortlervortices as is sometimes evident downstream of flow
reattachment. This type of flow instability can
produce heating augmentations of 30-50 percentabove turbulent values and was a concern in the
initial thermal design specifications of the flap.Extracted heating distributions along the flap spannear the trailing edge (Fig. 23) were constant with theexception of the area near the flap split gap. Anapproximate increase of 15 percent in heating to theflap edge was measured and is the result of the inflowtowards the gap (and inferred acceleration over theedge) observed in the streamline patterns presentedearlier (Fig. 17). Increasing the body flap deflection
angle from 20 to 30 deg resulted in a 40 percentincrease in overall heating levels on the flapdownstream of reattachment. Closer to the point
flow reattachment (X/L=0.9), flap deflection had amore pronounced effect on heating as shown in Fig.24. The "overshoot" in heating at reattachment for
body flap deflections of 25 and 30 deg is characteristicof transitional flow. As observed with the fin heating,
forced turbulence with discrete roughness elementsplaced upstream of separation suppressed the measuredtransitional heating on the flap, Fig. 25. Thesuppression of reattachment heating on the flaprelative to transitional levels were also consistent
with results obtained from heating studies conductedat AEDC on the SV-5D _° (not shown).
Numerical prediction of transitional mdturbulent interactions such as that which occurs with
the X-38 deflected flap remains challenging.Turbulence models play a crucial role in thesimulation of complex flows where separation,shock/boundary layer interaction, and flowreattachment are present. The correct prediction ofsurface heating depends to a large degree on theturbulence model. Predicted (published andunpublished) laminar and turbulent body and flapheating distributions from the CFD codes detailed inRefs. 49, 51, and 52 were compared to the measuredLaRC Mach 6 heating data to develop a higher degreeof confidence in predictive techniques utilized for X-38 flap design. The experimental heatingdistributions presented in Fig. 26a-b (_=40 deg, bodyflap =20 deg, and Re_ L = 4x10 _') correspond tolaminar and turbulent flow upstream of the deflectedflap. The predicted laminar heating distributions fromthe two Navier-Stokes solvers z's_ and a two-layer
method t2 agreed with measured values and indicatedthe boundary layer upstream of flow separation waslaminar, Fig. 26a. On the deflected flap, themeasured heating was a factor of three higher thanlaminar predictions, which suggested non laminarflow reattachment (the two-layer method was not usedto predict flap heating). Experimentally, theboundary layer was forced turbulent via discreteroughness and the resulting heating distribution
compared to turbulent prediction 49's_52 Fig. 26b. As
expected, the algebraic Baldwin-Lomax turbulencemodel 52 did not perform well in the vicinity of theflap where an adverse pressure gradient and flowseparation exist. A modified two-equation (k-0_)turbulence model s_ more accurately predicted theheating magnitude on the flap. The Shear StressTransport (SST) turbulence model most faithfullyreproduced the measured heating distribution on thedeflected flap. While this comparison does not implythe turbulence model has been validated for flight, itdoes suggest that of the three numerical modelsinvestigated the SST may be the most suitable for
application to X-38 flap design. Ref. 51 provides adetailed discussion of the numerical turbulence
models, and comparisons to additional LaRC X-38data and other benchmark experiments.
Flap Cavity HeatingThe aerothermal environment of the cavity
located behind the deflected flaps represents anextreme challenge from an experimental andnumerical modeling perspective. In flight, forcedconvection through the flap gap, radiative heatingbetween the flap leeside and aft cavity surfaces, flowseparation, and three-dimensionality are all present.The presence of critical component hardware such asthe flap actuator rod, Fig. 27, requires an accurateprediction of the environment to insure properperformance and adequate thermal protection.Experimentally, the cavity flowfield behind the flapswas dominated by the jet-like impingement of theflow through the flap split gap onto the cavity floor.
A photograph of the surface streamline patterns fromflow impingement at Mach 6, cz=40 deg, body flap=20 deg is shown, Fig. 28. The impingementproduced longitudinal and spanwise variations insurface shear on the cavity floor.
To characterize the heating from the jet-likeimpingement, NASA LaRC provided the first detailedconvective heating measurements made on the X-38cavity surface. A comparison of Navier--Stokesunpublished laminar and turbulent prediction from theCFD code detailed in Ref. 51 with measured cavity
floor heating at wind tunnel conditions (tz=40 deg,body flap =20 deg, and Re_, = 4xlO 6) is shown,Fig.29. For presentation purposes, the numericalsolution on the left of the symmetry plane
corresponds to turbulent flow (k-0_) and on the right,laminar flow. The non-dimensional heating
magnitudes from the discrete measurements (indicatedwithin symbol) have been assigned color contourlevels corresponding to that used for prediction.Similar to the phosphor results, the wanner colors(yellow, red etc) correspond to areas of higher heating.With the exception of the location of the cavityheating peak between the flap gap, the predictionscaptured the two-dimensional surface heatingcharacteristics measured on the cavity floor. At this
II
Reynolds number condition it was determinedexperimentally that the heating peak to the cavitysurface was located on centerline near the flaphingeline. In contrast, the computationally predictedpeak was located near the vehicle trailing edge.
Additional experimental tests revealed that theheating peak to the cavity floor exhibited a strongspatial sensitivity to Reynolds number not predictedcomputationally. To visually capture global heatingcharacteristics of the cavity floor from the discretethin film measurements, the data obtained at Mach 6,
6t=40 deg, body flap =25 deg are interpolated andpresented in the form of a color contour plot, Fig.30a-e. The forward movement of the heating peaktoward the flap hingeline with increasing Reynoldsnumber was observed. Secondary heating peaks were
measured outboard of the centerline near the cavityvertical sidewall and corresponded to vortical flowinferred from the increase in shear in the surface
streamline pattern, Fig. 28. The same data isreplotted in a more conventional format, wherebycenterline cavity normalized heating distributions areplotted vs. vehicle length (X/L), Fig. 31. The rangeof Reynolds numbers was sufficient to producelaminar and turbulent flow on the flap windwardsurface. The increase in magnitude and forwardmovement of the heating maximum on the cavityfloor with Reynolds number coincided with theforward movement of flow reattachment (decreasingseparation) on the windward flap surface. Themagnitude increase and shift of the heating maximumwith Reynolds number, were consistent with Mach10 trends obtained from heating studies conducted atAEDC on the SV-5D _ (this trend was observed with
extreme flap deflections of 40 deg and consequently,not shown). A direct comparison of the SV-5Dcavity centedine Mach 10 data of Ref. 50 with thecorresponding LaRC Mach 6 X-38 heatingdistribution was possible for a flap deflection of 20deg and is shown, Fig. 32. While the data from Ref.50 was limited spatially (two thermocouples on thecavity floor), the heating maximum would appear tohave been located near the aft end of the vehicle. This
would be consistent with the present Mach 6 trendsobserved at low Reynolds number. The opening of ahinge line flap seal on the wind tunnel model of Ref.50 appem_ to have shifted the cavity-heatingmaximum forward towards the flap/cavity interface.Relief of the separated flow on the flap windwardsurface through the gap would appear to haveproduced a smaller recirculation region, emulating thehigh Reynolds number Mach 6 trends from thepresent test. Additional tests at higher Mach number(M>6) are requited to determine if differences in theheating magnitude from the present test and Ref. 50are due to compressibility effects. At equilibriumconditions during hypersonic entry, the cavity heatingsensitivity to Reynolds number may not be as strong
as inferred from the perfect gas wind tunnelenvironment. The mechanism driving the cavityheating, flow separation on the flap windward surfacewould be less extensive in flight.
The effect of angle--of-atack and body flapdeflection on cavity heating is presented, Figs. 33 and
34. Turbulent conditions on the flap windwardsurface prevail at this Reynolds number in the windtunnel (Re_L= 8XI0 6) and are anticipated in flight atMach 6. In contrast to the decreased heating on thewindward flap surface, the lower flap deflectionsproduced a more severe thermal environment on the
cavity floor Fig 33. The effect of angle-of-attack oncavity heating at Re_ L = 8x106 for a fixed flap
deflection of 25 degrees is shown in Fig. 34. Peakheating on the cavity behind the flaps approached 30percent of reference stagnation values at 45 degreesangle-of-attack (Fig.34). Incidence angles of greaterthan 40 deg are presently being considered forhypersonic entry to moderate heating due to vehicleweight growth. It is reasonable to assume that theflap cavity interface may see significant heating if rollcontrol authority requires flap deflections between 10and 15 degrees at these higher entry angles-of-attack.
Currently, the design environment for this areahas been compiled completely from LaRCexperimental wind tunnel data. The measured heating
distributions were invaluable in developing a thermaldesign model and flight scaling factors applicable tothis localized region. Future computational work andexperimental tests are anticipated to refine this modeland reduce uncertainties.
Boundary_ Layer Transition
The proposed TPS of the X-38 windwardsurface consists of Shuttle-like ceramic tiles and
similar to Orbiter flight experience, boundary layertransition is expected to be roughness dominated.Surface roughness may arise from inherent TPS tilemismatch due to manufacturing tolerances or mayresult from protruding gap filler material. Anexperimental effort 18was made to determine if X-38boundary layer transition could be forced from discreteroughness. The analysis of data from Ref. 18 has
been used to quantify the effects of isolated roughnessalong the centerline of the windward surface and to
develop a transition correlation for the X-38 vehicle.Information from such a correlation has been used to
provide manufacturing guidelines and constraints forthe (step and gap) tolerances of the TPS tiles. Withsuch step tolerances defined, an estimate of whenboundary layer transition should occur in flight maybe made.
The experimentally determined transitioncorrelation was developed using the same
methodology reported in Refs. 19 and 53. Phosphorheating images were used to identify the transitionfootprint located downstream of systematically ptaced
roughness elements. The size and height of the
12
discrete tripping devices were methodically varied aswas freestream unit Reynolds number in order toproduce transitional and fully turbulent flow.Laminar boundary layer edge conditions at the triplocation were computed by a boundary layer code _(LATCH) for a range of Reynolds numbers. Tocorrelate the data, the experimental transition resultswere compared using the transition parameter ofmomentum thickness Reynolds number over edge
Mach number (Re0/M_) and the disturbance parameterof roughness height over boundary layer thickness(k/6). Figure 35 provides the results of thiscorrelation for all the discrete trip results along the X-38 centerline for Mach 6 at an angle-of-attack of 40degrees. Curve fits representing transition onset, andfully turbulent flow have been experimentallydetermined for X-335_ and are superimposed on the X-38 data set for comparative purposes. The correlatedX-38 data are consistent with that determined for X-
33. It is recognized that the determination oftransition onset from discrete roughness can beinfluenced by tunnel noise 55 and that the incipientcurve defined in Fig.35 may be conservative. Thecurrent X-38 TPS manufacturing guidelines specify
step tolerances no larger than 0.08-in near the nosecapon the windward surface.
]_xtrapolation to flight
A feature of the phosphor thermographyanalysis package 26 (IHEAT) is the ability toextrapolate ground based heating measurements toflight radiation equilibrium wall temperatures. Thesuccessful application of this technique to predictMach 6 flight surface temperatures for both laminarand turbulent conditions was demonstrated in the X-
3426'27and X-33 '_ programs. Based on the successfulMach 6 extrapolation of X-33 and X-34 wind tunneldata and the good agreement between the X-38measurement and prediction presented in this report,phosphor data were extrapolated to flight surfacetemperatures at Mach 6 and 10. Comparison ofextrapolated data to turbulent Mach 6 and laminarequilibrium Mach 10 flight prediction are made, Figs.36a-c and 37a-c. The Mach 6 tunnel data was
obtained on Rev 8.3 at ez = 40 deg, flap deflection of20 deg, and Re_l = 4 x 106 where turbulence wasforced with discrete roughness. Mach 6 flightconditions at (_ = 40 deg correspond to an altitude of127,000 ft., velocity of 6226 ft/s, and a length
Reynolds number of 5 x 10_'. The Mach 10 tunneldata was obtained on Rev 3.1 at _ = 40 deg, flapdeflection of 25 deg, and Re:_i = I x 106. Mach 10flight conditions at (z = 40 deg correspond to analtitude of 157,000 ft., velocity of 11,361 ft/s, and a
length Reynolds number of 2.3 x 10_'. No significantreal-gas aeroheating effects were anticipated at theMach 6 and 10 X-38 flight conditions.
The extrapolated phosphor images weremapped to the three-dimensional vehicle surface
geometry with the IHEAT code Map3D tool. I'hemapping technique permits a more accurate spatial
representation of the global data particularly whenextraction and comparison with numerical predictionare desired. The turbulent extrapolated surface
temperatures at Mach 6 agreed quite well relative topredicted flight temperatures, Fig. 36a, with theexception of the body flap region. Consistent withthe poor agreement found in the wind tunnelcomparisons (Fig. 26b), the Baldwin-Lomax algebraicturbulence model used for the flight computationunder-predicts the extrapolated (experimental)temperatures on the deflected body flap, as shown inFig. 36b. Comparison of extrapolated temperaturewith turbulent prediction along an axial station well
upstream of the expansion surface and deflected flaps,Fig. 36c, were in much better agreement.
The excellent comparison of laminar
extrapolated temperature at Mach 10 to flightprediction, Fig. 37. illustrates the versatility of theextrapolation theory and has extended the
demonstrated range of applicability to Mach 10.Similar to the Mach 6 data, the surface temperatures
compared well over the entire image with theexception of the body flap. This again was notsurprising as the extent of laminar separation was notcaptured computationally at wind tunnel conditions(see Fig. 13). Upstream of the flap interaction, theextrapolated wind tunnel data at Mach 6 and 10generally compared to within 50 deg F (or better) offlight prediction. The extrapolation methodology hasthe potential to provide detailed and timely designinformation early in a design cycle, when a largenumber of vehicle parametrics are being considered.This type of experimentally derived globalinformation provided to the designer early in the TPSevaluation process would be invaluable for materialselection and sizing requirements.
Concluding Remarks
The X-38 program plans to demonstrate anautonomously returned orbital test flight vehicle to
support the development of an operational CrewReturn Vehicle for the International Space Stationthat will return crew members safely back to earth inthe event of medical or mechanical emergency. This
report provides an overview of the hypersonicaerothermodynamic wind tunnel program conducted atthe NASA Langley Research Center to date by theAerothermodynamics Branch in support of the X-38vehicle design.
The X-38 program was able to take advantageof recent developments in a two-color global
phosphor thermography technique, providing anopportunity to conduct heating screening/trade studyconcurrent with aerodynamic tests. The LaRC groundbased tests contributed significantly to the
development and validation of the flight data book for
13
longitudinal and lateral aerodynamic characteristics aswell as control surface effectiveness. Comparison ofaerodynamic measurements between Mach 6 air and
CF4, provided an indication of the significance of realgas effects for X-38. Global and discrete surface heat
transfer measurements were primarily used in thecontinued development and validation ofcomputational tools used to predict the X-38aeroheating environment. Under the presentNASA/European partnership, the aerodynamic mdheating measurements provided by LaRC wereutilized to augment and compliment test resultsobtained in European facilities. The synergismbetween the experimental and computational workperformed within the X-38 program has led to animproved understanding of complex flows.
The hypersonic aerodynamic wind tunnel testsindicated that the X-38 has more than sufficient
control authority for pitch control. Pitching momentincrements from the LaRC Mach 6 and 10 tests
compared favorably with the data book values derivedfrom European aerodynamic tests. The heavy gassimulation tests have indicated that the real gas effectson X-38 aerodynamics at trim conditions are expectedto primarily influence flap effectiveness. Relative tolaminar perfect gas results the heavy gas simulation
tests revealed an increase in body flap effectivenessacross the angle-of-attack range.
Global heating measurements for attachedlaminar flows were in good agreement withpredictions from CFD codes used to define the flightaeroheating environment. Experimental heatingmeasurements in the vicinity of control surfaces(body flaps and rudder) were made to provide initialdesign information from which thermal marginassessments were made. Predicted deflected flapheating distributions were compared to measuredheating data from the LaRC tests in an effort todevelop a higher degree of confidence in predictivetechniques utilized for separating reattaching flows.Transitional flow reattachment represented a challengefrom a numerical modeling perspective. In areaswhere predictive tools could not provide accurateinformation, such as the cavity behind the deflectedflaps, the design environment has been compiledcompletely from LaRC experimental wind tunneldata. The heating distributions on the cavity floorwere invaluable in developing a thermal design modeland flight scaling factors applicable to this localizedregion. Future computational work and experimentaltests are anticipated to refine this model and reduceuncertainties. A more comprehensive experimentaland computational effort has been initiated to moreaccurately predict the flight-heating environmentassociated with the windward surface of the deflected
body flaps.
The global aeroheating results obtained atMach 6 have been used to quantify the effects of
isolated roughness along the centerline of thewindward surface and to develop a transitioncorrelation for the X-38 vehicle. Information from
the correlation was used to provide manufacturingguidelines for the (step and gap) tolerances of the TPStiles. With such step tolerances defined, estimates of
when boundary layer transition would occur in flightcan be made.
Extrapolation of ground _sed heatingmeasurements to flight radiation equilibrium walltemperatures were made at Mach 6 and 10. The
extrapolated wind tunnel data generally compares towithin 50 deg F (or better) of flight prediction. Thistype of information provided to the designer early inthe TPS evaluation process would be invaluable and
could potentially result in significant savings ofcomputational time required for flight predictions
Acknowledgments
Without the assistance of the followingindividuals this work would not have been possible:Mark Cagle, Joe Powers, Mike Powers, MarkGriffith, Ed Covington and Tom Burns for modeldesign, fabrication, instrumentation, and surfaceinspection support; John Ellis, Rhonda Manis, GraceGleason, Melanie Lawhorne, Harry Stotler, Steve
Jones, and Jeff Werner for wind tunnel support;Sheila Wright and Bert Senter for data acquisitionassistance; Tuan Trong, Leo Perez, Chuck Campbell,and Steve Labbe for analysis support; The GermanAerospace Center (DLR) and the partners of theGerman national program on re-entry technologies,(TETRA) for CFD support; Dassault Aviation for the2-Layer laminar and turbulent wind tunnel heatingpredictions; and Richard Wheless for documentation
assistance. The authors gratefully acknowledge theircontributions and behind-the- scenes work.
References
l.Brown, D., "Manning the Lifeboats on theInternational Space Station," Launch.space, April-May 1998, pp. 20-22.
2.Asker, J. R., "For Myriad o1 Woes, NASA Calls X-38
to the Rescue," Aviation Week & Space Technology,Nov. 11, 1996, pp. 70-71.
3.de Selding., P. B., "ESA, United States AdvanceToward Common Crew Craft," Space News, Nov. 18-24, 1996, p.3.
4.de Selding., P. B., "Germany, NASA Pen Deal lot CrewLifeboat Work," Space News, Nov. 24-30, 1997.
5.Smith, B. A., "Weather May Hamper X-38 DropTests," Aviation Week & Space Technology, Dec. 8,1997, p. 33.
14
6.Kandcbo, S. W., "Vista F-16 Tests Previe_ X-38
Flight Controls" Aviation Week & Space
I'echnolo_y, Oct. 19, 1998, p. 4[.
7.Covault, C., "Second X-38 Set l_r Flight" Aviation
Week & Space Technology, Aug. 31, 1998, p. 58.
greed, R. D. "Wingless Flight The Lifting BOd,_ Stor} ,"
NASA SP-4220, 1997.
9.Barret, C., "Lifting Bod 5 Stabilit,,, and Control,"
NASA TM -1999-209255, March 1999.
IO."SV-5D Prime Final Flight Test Summary," Martin
Marietta Engineering Report ER 14465, Sept.,
1967.
l l."Prime Follo_'-on Program: Addition Aerodynamic
and Aerothermodynamic Data Analysis," Martin
Marietta Engineering Report ER 14686, Jan., 1968.
12.Hallion, R.P., "'The Hypersonic Revolution: Eight
Case Studies in the History of H}personic
Technology," Vol. 2, 1987, pp. 893-923
13.Dornhiem, M. A., "X-38 Transitions from [+ilting
Bod 5 to Parafoil" Aviation Week & Space
Technology, March, 23, 1998, p. 92
14.Labbe, S. G., Perez, I+.F.+ Fitzgerald, S ML, I+ongo,
J. M. A.+ and Rapuc, M., +'X-38 NASA/DLR/ESA-
Dassault Aviation Integrated Aerodynamic and
Aerothermodynamics Activities," Atmospheric
Reentry Vehicle and Systems Symposium, Arachon,
France, March 16-18, 1999.
15.Tribot, J. P., Tran, P., Pallegoix, J. F., Orlo_ski,
M., Bruck, S., Andres, O. P., and Fitzgerald, S.M.
"X-38 Aerothermod}namics,'" Atmospheric Reentry
Vehicle and Systems Syml:x+sium, Arachon, France,
March 16-18, [999.
16.Campbell, C+ H., Caram, J M., Bert3, S A.,
Hor_ath, T.J. Merski, N. R., L¢_mis, M. P., and
Venkatapathy, E., "An Overvie_ ot X-38 Hypersonic
Aerothermodynamic Wind Tunnel Data and
Comparison with Numerical Rest, Its," AIAA Paper
97-2475, 1997.
17.L¢×_mis, M. P., Venkatapath), E., Papadopoulos, P.,
Davies, C. B., Berry, S. A., Horvath, T. J., and
Campbell, C. H., "Aeroheating and Aer_l)namicCFD Validation and Prediction Ibr the
Computational/ Expenmental Aeroheating
Predictions for the X-38 Program," AIAA Paper 97-
2478, June, 1997.
18.BerL',, S. A., Hor'_ath, Y. J+, Roback, V. E., and
Williams. G. B., "Results of Aerothermodynamic and
BoundaD-La_er Transition Testing of 0.0362-Scale
X-38 tRey. 3.1) Vehicle in NASA [.angle,,' 20-Inch
Math 6 Tunnel," NASA TM-112857, September1997.
19.Berr 3, S A., Bouslog, S. A., Brauckmann, G. J., and
Caram, J. M., "Shuttle Orbiter Experimental
BoundaD-La}er Transition Results with Isolated
Roughness," Journal of Spacecraft and Rocket._, Vt_l
35, No. 3, 1998, pp+ 241-248.
20+Campbell, C. H., Caram, J. M+, Berrs, S. A.,
Horvath, T. J., and DiFulvio, M., "'Overvie,,_ of X-38
Hypersonic Wind Tunnel Data and Comparison with
Numerical Results," AIAA Paper 97-0567, January1997.
21,Miller, C. G., "Development of X-33/X-34
Aerothermodxnamic Data Bases: Lessons Learned
and Future Enhancements," NATO Research and Tech.
Org., Applied Vehicle Technology Symposium
Aerodynamic Design ans Optimization o1" Flight
Vehicles in a Concurrent Multi-Disciplinary
Environment, paper no. 32, Oct 1990.
22.Miller, C. G., "Langle3 Hypersonic
Aerodynamic/Aerothermodynamic Testing
Capabilities - Present and Future," A[AA Paper 90-
1376, June 1990.
23.Buck, G. M., "Rapid Model Fabrication and Testing
for Aerospace Vehicles," AIAA Paper 2()0<_-0826,
January 2000.
24.Buck, G. M. and Vasquez, P., "An Insestment
Ceramic Slip-Casting Technique for Net-Form,
Precision, Detailed Casting of Ceramic Models," t_
S. Patent 5,266,252, November 30, 1993
25.Buck, G. M., "Automated Thermal Mapping
Techniques Using Chromatic Image Analysis."
NASA TM 101554, April 1989.
26.Merski, N. R., "Reduction and Analysis of Phosphor
Thermograph} Data With the IHEAT Software
Package," AIAA Paper 98-0712, Jan. 1998.
27.Merski, N.R., "Global Aeroheating Wind-Tunnel
Measurements Using Improved Ts_o-Color Phosphor
Thermographx Method" Journal of Spacecraft and
Rockets, Vol. 36, No. 2, 1998, pp. 160-170.
28.Berry, S. A., Horvath, T. J., DiFulvio, M., Glass, C ,
and Merski, N. R., +'X-34 Expenmental Aeroheating
at Mach 6 and tO," Journal of Spacecraft and
Rockets, Vol. 36, No. 2, 1998, pp. 171-178.
29. Horvath, T J., Bert 3, S. A., Hollis, B.R., Liecht 3,
D.S., and Merski, N. R., "X-33 Experimental
Aeroheating at Mach 6 Using Phosphor
Thcrmography," AIAA Paper 99-3558, June, 1999.
30Miller, Charles G. Ill, "Comparison of Thin-Film
Resistance Heat-Transfer Gages With Thin-Skin
Transient Calorimeter Gages in Convcntional
Hypersonic Wind Tunnels," NASA
TM 83197.
31.Hollis, Brian R., "User's Manual for the One-
Dimensional H3personic Experimcntal Acro-
Thermodynamic (IDHEAT) Data Reduction Code,"
NASA CR 4691, (Grant Nos. NAGl-16133,
NAGW- 1331).
32.C_)k, William J. "Unstcad3 Heat Transfer
Measurements to a Semi-lnfinitc Solid With
15
Arbitrary Surtace 'Femperature Histor) and Variable
Thermal Properties," Iowa State Univ. Tech Report
ISU-ERI-AMES-67_)0.
33.Kendall, D. N.; Dixon, W. P.; and Schulte,
E. H. " Semiconductor Surface
Thermocouples for Determining Heat-Transfer
Rates," IEEE Transactions on Aerospace and
Electronic Systems,
Vol AES-3, No. 4.
34.Hollis, Brian R., " Experimental and Computational
Aerothermodynamics of a Mars Entry Vehicle," Phd
Dissertation, North Carolina State University.
(Grant Nos. NAGI-1663, NAGW-1331), 1996.
35.Kline, S., and McClintok, F., "Describing
Uncertainties in Simple Experiments," Mechanical
Engineering, pp. 3-8, 1953.
36.AeroSoft, "'GASP Version 3, The General
Aer,,x:lynamic Simulation Program, Computational
Flow Analysis Software lk)r the Scientist and
Engineer, User's Manual," AeroSoft, Inc.,
Blacksburg, VA., May 1996.
37.Olynick, D. R., and Tam, T., "Trajectory Based
Validation of the Shuttle Heating Environment,"
Jo,rnal of Spacecraft and Rockets, Vol. 34, No. 2,
1997, pp. 172-181.
38.F.C. Michel., "Aerodynamic Data Bt_k- SV-5D
Aerodynamic Characteristics. 680A Program," CR-
113 E, 5 'h Revision, Oct., 1, t965.
39.Murphy, K., Nowak, R., Thompson, R. A., and
Hollis, B.R., "X-33 Hypersonic Aertxtynamic
Characteristics," AIAA Paper 99-4162, August 1999.
40.Hollis, B.R., Thompson, R. A., Murphy, K.J.,
Nowak, R.J., Riley, C.J., W(×xt, W.A., Alter, S.J.,
and Prabhu, R.K. "X-33 Aerodynamic and
Aeroheating Computations for Wind Tunnel and
Flight Conditions," AIAA Paper 99-4163, August1999.
41.Thompson, R.A. "Revie_x of X-33 Hypersonic
Aerodt'namic and Aerothermodynamic
Development," ICA 0323, August 2000.
42.Brauckmann. G. J., Paulson, J. W., and
Weitmuenster, J. K., "Experimental and
Computational Analysis of the Space Shuttle Orbiter
Hypersonic Pitch-Up Anomaly," AIAA Paper 94-
0632, January 1994.
43.Perez, L., F., "X-38 Aercxlynamic Design Data
Book," JSC-28854, Rev. G, Feb., 2000.
44.Heppenheimer, T. A., "The Space Shuttle Decision.
NASA,s Search for a Reusable Space Vehicle," NASA
SP-422 t, 1999.
45.Screen, E.N., and Nickel, H.S., "Htpersonic Heat
Transfer Test Results From a 9.688(_ M(x.tel SV-5D
Vehicle For Program 680A In the Martin Hypersonic
Tunnel-Second Serics," ER-13686, Feb., 1965.
46.Wong, H., Kremer, F., "Numerical Assessment on the
Heating of the Rudder/Fin Gap in the X-38 Space
Vehicle," Atmospheric Reentry Vehicle and Systems
Symposium, Arachon, France, March 16-18, 1999.
47.Muhlrazer, A., Frohlich, A., and Wildenrotter, K.,
"Design, Material and Manufacturing Aspects of the
X-38 Body Flaps," Atmospheric Reentr t Vehicle and
Systems Symposium, Arachon, France, March 16-
18, 1999.
48.Trabandt U., Schmid, T., Reinkober, H., and Ritter,
H., "Light-Weight CMC TPS for Future RLV Applied
On X-38 Nose Skirt," Atmospheric Reentry Vehicle
and Systems Symposium, Arachon, France, March
16-18, 1999.
49.Bruek, S., Radespiel, R., and Longo, J.,
"Comparison of Nonequilibrium Flows Past a
Simplified Space-Shuttle Configuration," AIAA
Paoer 97-0275, Jan., 1997.
50Rhudy, R., and Merz, G., "Heat Transfer and PressureDistribution on a Scale Model SV-5D at Math 10,"
AEDC-TR- 194, September, 1965.
51.Weber,, C., Behr, R.., and Weiland, C.,
"Investigation of Hypersonic Turbulent Flow Over
the X-38 Cre_ Return Vehicle," AIAA Paper 2000-
2601, June 2000.
52.De[attre, N., et al., "Numerical Prediction of
Hypersonic Reacting Flo_vs with the Finite Volume
Solver FLU3NEQV," AIAA Paper 97-0304, Sept.,
1997.
53.Bert t, S. A.,, Horvath, T. J., Hollis, B.R.,
Thompson. R. A., Hamilton, H. H., "X-33
Hypersonic Boundary Layer Transition," AIAA Paper
99-3560, June 1999.
54.Hamilton, H., Green, F., and DeJarnette, F.,
"Approximate Method lk_r Calculating Heating Rates
on Three-Dimensional Vehicles," Journal of
Spacecraft and Rockets, Vol. 31, No. 3, 1994.
55.Schneider, S. P., "Effects of High-Speed Tunnel Noise
on Laminar-Turbulent Transition," AIAA Paper
2(X)0-2205, June, 2000.
16
Table 1:X-38 Aero/Aeroheating Tests in NASA LaRC AB Tunnels
Year Tunnel
1996 22-In Mach20 He
1996 3 l-In Machl0 Air
1996 22-In Mach20 He
1996 20-In Mach6 CF4
1996 20-In Mach6 Air
1996 20-1n Mach6 Air
1996 20-1n Mach6 Air
1997 31-1n Machl0 Air
1998 20-In Mach6 CF4
1998 20-In Mach6 Air
1998 31-In Machl0 Air
1998 20-In Mach6 Air
1998 20-In Mach6 CF4
1998 20-1n Mach6 CF4
1999 20-In Mach6 Air
20OO
Test Configuration(s)556 Rev 3.0/Rev 3.1
322 Rev 3.0
560 YPAIO
113 Rev 3. I/YPAIO
6722 Rev 3.1
6733 Rev 3. I/YPAIO
6735
335 Rev 3.1/YPAIO
120 Rev 8.3
6765 Rev 8.3
345 Rev 8.3
6774
123
Runs
35
42
20
56
6
90
174 Rev 3.1
197
47
50
122
138
50
Objective
F&M Initial screenin£
Heatin s Initial screenin 8F&M Revised OML
F&M Gamma effects
F&M Schlieren
F&M Gamma effects
Heatin 8 TransitionF&M Mach effects
F&M Screening new OML
F&M Screenin_ new OMLF&M Mach effects
Rev 8.3 F&M Gamma effects
Rev 8.3 F&M Gamma effects
Rev 8.3 F&M Gamma effects125 143
6782 216 Heating Global/cavity
31-In Machl0 Air 362 TBD Rev 8.3 Heatin_ Global/cavity
Table 2: Nominal Flow Conditions in NASA LaRC AFC Tunnels
Facility q_(psi) Pt.t(psi) Tta(*F) ReJft (xl0 6)
Rev 8.3
M_
22-In Mach20 He 17.4
31-In Mach 10 Air 9.7
9.83
9.95
20-In Mach6 Air 5.91
5.90
5.94
5.98
6.02
20-In Mach6 CF4* 5.98
1.25 500 80
0.66 350 1350
1.25 720 1350
2.41 1450 1350
0.51 30 410
1.04 60 430
2.10 125 450
4.07 250 450
7.52 475 475
0.80 950 850
P2/P_
4.0 3.99
5.96 0.53
5.98 1.01
5.98 2.00
5.23 0.54
5.23 1.04
5.27 2.08
5.28 4.06
5.29 7.28
11.68 0.35
*Mach 15-20 simulation due to high normal shock
within shock layer.
Table 3:
Dimension Full scale (V201) O.OI75-scale
S,_f 233.28 ft-" 10.288 in 2
Lr_f 27.6 ft 5.796 in
bre f (=Lref) 27.6 ft 5.796 in
Moment reference center 15.732 ft 3.304 in
0.57 x L,.f
density ratio (pjp_) and/or low values of specific heat ratio (¥)
X-38 Reference Dimensions
17
FS0 00
4 8688
_I-- 0_8t ," 100_PIVOTAXI$ ( a_L_00 L
. 130" _14 _ ¢ |
_0 431 _ -CLO 00-
0216R ' 15"9° '"
0771_" -FSO 00- 3 184
Figure 1. Dimensions (in inches) for 0.0295-scale X-38Rev-3.1.
Turbulent Laminar5O
10 7'
3O
105
10
:|i104 -10
0 5 10 15 20 25 30
M=
Figure 2. X-38 cycle 8 trajectory.
Figure 4. O.0295-scale ceramic heat transfer models.
Rev 8.3 YPAIO Rev 3.1
Figure 5a. Thin-film heat transfer model installed in the NASALaRC 20-Inch Mach 6 Air Tunnel.
Figure 3. O.OI75-scale metallic force & moment models.
(b) Flap cavity thin film heat transfer sensors.
18
C m
LID
0.04 -
002
p
-Q02
-004
-0.06
-0.080
$ 6 bf = 0 deg_b
t_bf = 25 deg A
Open X-38 Rev 31 Re= L = 3.2x106
Closed SV-5D Re=, L = 19x106 (ref 38)
I_ Estimated uncertainty, present results
6 bf = 20 deg
I JJ I]lttJllttll .... I .... I
10 20 30 40 50
Angle-of-attack
(a) Pitching moment coefficient, M_ --20 Helium.Figure 6. Comparison of X-38 Rev 3. I aerodynamics to
SV-5D preflight database (ref. 38).
1.25_-
0.75
0.5!
F0.25_
[]
• t]
Open X-38 Rev 3.1 Rex,L = 32x106
Closed SV-5D Re L = 19x106(ref. 38)
0_l J t _ I , , _ , J J J J J I _ J _ , [ J _ J J I
0 10 20 30 40 50
Angle-of-attack
(b) Lift-to-drag ratio, M_ = 20 Helium, 6,_-- 20 deg.
CA
0.2 -
0.15 -
0.1 -
Rein L (106)
0.3
05lO
Estimateduncertainty-7indicatedon symbol /
O.O5 I I I I I I I I10 15 20 25 30 35 40 45 50
Angle of Attack, deg
(a) Axial force coefficient.
Figure 7. Reynolds number effects on X-38 Rev 3. Ilongitudinal aerodynamics M_ = 6 air, bm= 25 deg.
1.2 -
1 --
0.8-
0.6-CN
0.4-
- O0.2
0
I-0.2
10 15
0.04 -
0.02 -
C m 0-
-0.02 -
-0.04
-006
-0.081'0
Re_. L (106)
0.3
q] 0.5
? 1.0
1.5
E Estimateduncertainty--Iindicatedon symbol1I I I 1 1 I I20 25 30 35 40 45 50Angleof Attack, deg
(b) Normal force coefficient.
Re=, L (10 6) I_
_p 0.3m 0.5
1.0
15
I I 1 I I I15 20 25 30 35 40
Angle of Attack, deg
(c) Pitching moment coefficient.
1 145 50
19
0.2 - 0.2
CA
0.15
0.1 --
10 1.4
m 20 1.67 (Re=, L =3.2 x 106)
E Estimated uncertainty "]intimated on symbol
o _ S a _ 8o o[] o
o
0.05 1 I I I t I I I10 15 20 25 30 35 40 45 50
Angle of Attack, deg
(a) Axial force coefficient.
Fig. 8. Mach number effects on X-38 Rev 3. l longitudinal
aerodynamics Re=, L = 0.5 x l06, 6RF = 25 deg.
1.2 -- M Y=® 6 1.4
1 - [] 10 1.4
20 1.67 (Re= L = 32x106)
0.8 - [-Estimated uncertainly-]L indicated on symbol _1
0.6-
CN #0.4-
0.2-
0-
-0.210
I I I I I I I15 20 25 30 35 40 45
Angle of Attack, deg
(b) Normal force coefficient.
0
0
0
I5O
0.04 --
0.02
Cm 0-
-0.02 -
-004-
-0.06 -
-0.0810
%
M= Y_
d) 6 1.4[] 10 1.4
¢, 20 1.67 (Re=, L = 3"2x106)
E Estimated uncertainty -]indicated on symbol
I I I I I I15 20 25 20 35 40
Angle of Attack, deg
(c) Pitching moment coefficient.
o
o
I I45 50
0.15
CA
0.1 _-
0.0510
Test gas y = O_o/P2
Air 1.4 5
0004 1.2 12
Estimated uncertainty']indicated on symbol /
o o o o _3 t_[] [] [] []
0
I I I I I I I I15 20 25 30 35 40 45 50
Angle of Attack, deg
(a) Axial force coefficient.
Fig. 9. Gamma (y) effects on X-38 Rev 3. l longitudinal
aerodynamics M= = 6 Re= L = 0.25 x l0 ", 6RF = 15 deg.
12-
C m
1 --
0.8-
0.6-CN
0.4-
0.2- m
o
-o.2 I10 15
0.04
0.02 -
0 -
-0.02 -
-0.04 -
-0.06 -
-0.0810
m
m
Test gas ¥ = P=/P2
Air 1.4 5
0004 1.2 12
E Estimated uncertainty]indicated on symbol JI I I I I I I
20 25 30 35 40 45 50
Angle of Attack, deg
(b) Normal force coefficient.
O8 O o oo[3 0 []
L3 0n 15°£ o
£ o
Test gas Y _ Poc/P2 _]
Air 1.4 5 _ 25°
m (3=4 1.2 12 o
I Estimated uncertainty]
indicated _n syml_°l J I 1
15 20 25 20 35 4OAngle of Attack, deg
(c) Pitching moment coefficient.
I I45 50
20
2 -
__. 15(
--o--- <_
"<>"
o_-o- -o..-..K.3 __.o
gN
0.5- --e.-- 15
-_-- 20
-*- 25
o I I I I I I II0 15 2O 25 3O 35 40 45
Angle Of Attack, deg
Fig. 10. Effect of gamma (y) on X-38 Rev 3.1 body flap
effectiveness Rex L = 0.25 x 106.
0.008 -
0006
0 004
0002_
Cnl_ 0'
-0.002 [
-0004
-0.006
-0008
o o
o o
[] []
00
00
Stable
0 0 0 0
0 0 0 0
[] [] []
_ o Rev 3 1
YPAIO
- o YPAIO with LaRC modifications
I I I I I I I I0 15 20 25 30 35 40 45 50
Angle of Attack, deg
(a) Yawing moment derivative.
Fig. 11. Configuration effects on X-38 lateral/directional
aerodynamics M_ = 6air Re_ L = 0.25 x 10", tS_F = 0 deg
- 0,_+ 2 deg.
0 006 -
o
0.004 - []
o0 002 -
0
-0002
-0004
-0006
cq_
Rev 3.1
YPAIO
YPAIO with LaRC modifications
_- °g
Stable-0.008 -
-0 01 I I I I I I I I
10 15 20 25 30 35 40 45 50
Angle of Attack, deg
(b) Rolling moment derivative.
cyl_
0 03 -
00.02 -
13
o0.01 -
0-
-0 01 -
-0 02
-O03 -
-0 04 -
-005 I I I I I I 1 I10 15 20 25 30 35 40 45 50
Angle of Attack, deg
(c) Side force derivative.
Rev 3 1
YPAIOYPAIO with LaRC modifications
Body flap=0 deg
IL == - - - - -
-0,02 - _lap-10 deg
-004 - _lap=-20 degAC m _l_,,_f
_.06 - _ |-0.08 - Body flap--30 deg
-0.1 _ADB revG M._=5.5 (ref. 43)
• LaRC M_=6 Re=eL = lx106
-0.12 I I I I 1 I I I15 20 25 30 35 40 45 50 55
Angle of attack, deg
(a) M= = 6 air.
Fig. 12. Comparison of X-38 Rev 8.3 pitching momentincrements with X-38 preflight aerodynamic data book, Rev G
(ref. 43).
0 --
-0.02 -
-0.04-
AC m
-0.06
-0.08
-0.1
Body flap--O deg
- - - - - flap=10 deg
- Body flap<30 deg
_ADB revG M._=10 (ref. 43)
• LaRC M==10 Re= L = lx106
I , , , , _ I I-01215 20 25 30 35 40 50 55
Angle of attack, deg
(b) M= = 10 air.
21
Ex perim enta I
Predictio n
0.6
'Iz
0
Fig 13. Comparison of measured X-38 Rev 3.1 global
windward heating with laminar prediction (ref. 17)
M_ = 10, ct -- 40 deg, 6BF = 25 deg, Re_, L = 0.5 x 106.
1- Configuration ""-- -- ---._ _"
Pr_liction X-38 Rev 3.1 Laminar GASP M = 10 (ref 17)0.8
" _ Exp X-38 Rev 3 1 LaRC M== 10 Air
..... Exp SV-5D AEDC-C M_= 10 Air (ref. 45)
_. _ " " Exp SV-5D Martin HS M_= 20 N 2 (ref. 45)0.6
._
0.4o
Z
0.2
0 02 0.4 0.6 0.8 1
X/L
(a) Windward centerline.
Fig. 14. Comparison of measured X-38 Rev 3.1 and SV-5D
(ref. 45) heating distributions with laminar prediction (ref. 17)
M_ = I0, (_ = 40 deg, 8BF = 25 deg, Re=, L = 0.5 x l0 _
0.5
0.4
0.3
021
0.1
0.0 .... I , , i _ I a _ i _ I i , i i I0 01 02 03 04
2yfo
(b) Axial station, X/L = 0.25 M= = 10 Air.
0.5
0.4
0.3
oz
0.1
I
j.i f"
I
/k Exp X-38Rev31[][]
Prediction X-38 Rev 3 1[]
(laminar GASP) (ref 17)
, I , I , I , I , I , I , I ,
0'00 0.2 0.4 0.6
2y/b
(c) Axial station, X/L = 0.?3, M= = 10 Air.
I
0.8
00 0.2 0.4 0.6 0.8 1
r/l_
Fig. 15. Effect of Reynolds number on Rev 3.1 windward
centerline heating distribution
M_ = 6 Air, (_ = 40 deg, t5t_ = 25 deg.
22
(a) Re. L = 1 X 10 6 (a) Re. L = 1 X 10 6
(b) Re.,,L = 2 X 10 6 (b) Re... L • 2 X 10 6
(c) Re., L = 4 X 10 6
Normalized heating
0 1
(c) Re,o L= 4X 10 6
Fig. 16. Effect of Reynolds number on Rev 3. I global
windward heating.
M_ = 6 Air, _t = 40 deg, b_r = 25 deg.
Fig. 17. Effect of Reynolds number on Rev 3. I windward
surface streamlines.
M_ = 6 Air, ct = 40 deg, bBv = 25 deg.
23
(a) Re.o,L = 1X10 6
(b)Re=o,L = 2X106
(c) Re.o,L = 4X10 6
Normalized heating
0 0.3
Fig. ]8 Effect of Reynolds number on Rev 8.3 fin/rudderglobal heating M= = 6 Air, (_ = 40 deg, _3BF= 20 deg.
0.3
025
._=
,_ 0.2
_N 0.15
_ 0.1
120.05
-O- X-38 Rev 8.3 LaRC M = 6 Air
• SV-5DMartinHSM =20N 2(ref. 45)
Fin/rudder interface
4 x 106
2x 106
1 x 106
• 0,4 x 106
00 0.5 1
Nondimensional fin chord length
Fig. 19. Effect of Reynolds number on X-38 Rev 8.3 fin/rudder
heating distribution and comparison with SV-5D (ref. 45)M= = 6 Air, a = 40 deg, _3aF.= 25 deg.
C
|e-
-g.N
E
Z
0.3
0.25
0.2
0.15
0.1
0.05
0
- Fin#udder
_ interface --_
_ Untripped(Transitional)
,__ _" _ f _:"'_:_ Tripped(Turbulent)k = O.O05-in
0 0.5 1
Nondimenaional fin chord length
Fig. 20. Effect of Reynolds number on X-38 Rev 8.3 fin/rudder
chordwise heating distribution
M= = 6 Air, (3t= 40 deg, Re:_ L = 4 x 106, 6Br. = 25 deg.
e-
Q
m
.EO
z
0.8
0.6
"-[B-I_aRC X-38 Rev31 Reoc ' L =0.5X10 6
• AEDC C SV-5D Re=, L = 1X106 (ref. 45)
m Laminar prediction Re = L = 0.5X10 6 (unpub.)
0.4
0.2
O_0 0.2 0.4 0.6 0.8 1
Nondimensional flap chord length
Fig. 2 I. Comparison of measured X-38 Rev 3.1 and SV-5D
(ref. 45) body flap heating distribution with laminar
prediction (unpublished)
M® = 10 Air, c_ = 40 deg, 8BL:= 20 deg.
24
(a) 5bf = 20 deg
(b) 5bf = 25 deg
(c) ,Sbf = 30 deg
Normalized heating
0 1
Fig. 22. Effect on body flap deflection on Rev 3.1 global
windward heating
M_ = 6 Air, ct = 40 deg, Re_ L = 2.0 x 10 +'.
r'-
N
OZ
1
0.8
0.6
0.4
0.2
bbf'o_x302520deg _ -_---.
C'''W--0 .... i .... i .... i .... i .... + .... i .... i .... J .... i
0.2 0.3 0.4 05 0.6 07 0.8 09 1X/L
Fig. 24. Effect on body flap deflection on Rev 3. I longitudinal
heating distribution at y/b = 0.2
M_=6 Air, ct =40deg, Re= L= 2.0 x I0 +'
Tripped (k = 0.005-in)
m 0.8c
Untripped
0.6
z 0.2Iz x
0- I I I I- I0.2 0.4 0.6 0.8 1
X/L
tSbf'o20deg0.9 _ 25
x 30 0.808 r,_,
0.5 0.4
0.3 0.2
0.2 000.1
-0.6 -0.4 -0.2 0 0.2 0.4 0.6
y¢o
Fig. 23. Effect on body flap deflection on Rev 3.1 flap spanheating distribution at X/L = 0.98
M= = 6 Air, ct = 40 deg, Re=. L = 2.0 x 106.
Fig. 25. Effect of boundary layer trip on X-38 Rev 3.1
longitudinal body flap heating distribution
M= = 6 Air, ct = 40 deg, Re= L = 4 x 106, bm= 25 deg.
1 _ Exp LaRC-- Lam. CFD DLR Navier-Stokes (unpub.)...... Lam. CFD DASA Navier-Stokes (unpub.)-- " -- Larn CFD DASSAULT 2-Layer (unpub.)
=,,
I I
X/L
(a) Laminar.
Fig. 26. Comparison of measured X-38 Rev 8.3 body and flaplongitudinal heating distribution with prediction (ref. 51)
M= = 6 Air, ct = 40 deg, Re= L = 4 x 10 _' bm -- 20 deg.
25
0.8
t-k_
8 06J¢
0.40
Z0.2
Acuator
Fig. 27
---O---- Exp LaRC (Forced transition)....... Turb CFD DLR Baldwln-Lomax (NS) (unpub)..... Turb CFD DASA k-_,_ (NS) (ref. 51)
Turb CFD DASA SST (NS) (ref. 51)-- - -- Turb CFD DASSAULT Baldwin-Lomax
(2L) (unpub)
_¢:k " _'"C_, u_ ,. °..'_]_"• .°°--."[]
i %r_ Trip location I,! I0.2 0.4 0.6 0.8 1
X/1.
(b) Turbulent.
floor
Body flapleewardsurface
X-38 Rev 8.3 flap cavity and actuator arm location
(second flap omitted for clarity).
Turbulent
prediction (k-o_)
Re=, L = 33X106
Laminar
prediction
Re= L = 3'3X106
Experimental
iNormalized heating
railll0 0.1 0.2 0.3
Fig.29 Comparison of discrete thin film experimental cavity
floor heating with laminar and turbulent prediction (unpub.)
M_ = 6 Air, Re_ L = 4 x 106, 6Rv = 20 deg.
(a) Re=, L = 0.5X10 6
(b) Re,. L = 1.0X10 6
(c) Re,o L =20x106
Secondery
reattachment
Impengement
thru flap gap
Windward
flap separation
Fig. 28 X-38 rev 8.3 surface streamlines on flap cavity floor
M_ = 6 Air, ot = 40 deg, Re0¢ ' L = 2.0 x 106, ilBr = 20 deg.
(d)Re** _ = 4.0X106
(e) Re=,, L = 8.0X106
Increasing heating
Fig 30. Effect of Reynolds number on X-38 Rev 8.3 measured
cavity floor heating
M_ = 6 Air, (z = 40 deg, i3r_v= 25 deg.
26
0.3 Re=, L (x106) 0.3 ct, deg
025 -{3- 1 0.25 "-E}'- 25
-'O- 2
0.1 _ 01
z0.05 0.05
0 I 0 I I I I I0.9 0.92 0.94 0.96 0.98 1
X/t.
Fig 3 I. Effect of Reynolds number on X-38 Rev 8.3 centerlinecavity floor heating distribution
M_ = 6 Air, ¢z = 40 deg, 6w_ = 25 deg.
Reoc (xl0 6) Flap hinge0.500 , L seal
05 LaRCM =6 Closed1 LaRC M== 6 Closed
0.400 _ 2 LaRC M_c=6 Closed_ 4 LaRC M._= 6 Closed
0.9 0.92 0.94 096 0.98 1X/L
Fig 34. Effect of angle-of-attack on X-38 Rev 8.3 centerline
cavity floor heating distribution
M_ = 6 Air, Rex, L = 8 x 10 _', 6_ = 25 deg..
25O
20O
'i --v-0 LaRcM =0 C,osed
0.300 • 1 AEDC-C M== 10 Closed (ref. 50)
• 1 AEDC-C M_c=10 (3.n,-n (r_,f 5n) 150
OlOO° ° " " too500,00
09 0.92 0.94 0.96 0.98 1X/L
Fig 32. Comparison of measured X-38 Rev 8.3 and SV-5D(ref.50) centerline cavity floor heating distribution
M_ = 6 Air, (t = 40 deg, 6r_ = 20 deg.
0,3 m
0.25
O2
i 0.150.1
m
-13-- 15--0-2o
- -_- 25--V-- 3o
I I I I
0.05
0.9 0.92 0.94 0.96 0.98 1X/L
Fig 33. Effect of body flap deflection on X-38 Rev 8.3
centerline cavity floor heating distribution
M,=6Air,(_t=40deg,,Re=L=SX 10 ".
Re0/Me. C(k/t_)-l.0
X-33 Incipient C = 45X-33 Effective C = 60
O X-38 Laminar[] X-38 Transitional• X-38 Turbulent
00 02 04 0.6 08 1 12
146
Fig. 35. Experimental transition correlation of X-38 Rev 3. Iwindward centerline discrete roughness data and comparison
with X-33 results (ref. 53)
M= = 6 Air, ct = 40 deg
27
PHOSPHOR(Forced Turbulent)
Boundary LayerTrips
CFD(Turbulent - BaldwinlLomax)
TemP.(°_: 0 IgO _ STO _ 048 113(I 1329
(a) Global surface temperature mapping
Fig 36. Comparison of extrapolated turbulent experimental data
with turbulent flight prediction (unpublished)
M_ = 6, ot = 40 deg, 6RF = 20 deg, Re_, L = 5 x 106.
PHOSPHOR(laminar)
CFD(laminar)
mTemp ('F) 0 202 404 S0i I0| 1010 t212 1414 1115 1816
(a) Global surface temperature mapping
Fig 37. Comparison of extrapolated laminar experimental data
with laminar flight prediction (unpublished)
M_= 10, ot=40deg, 6R_=25deg, Re_¢ L=5X 100 .
Phosphor Turbulent Extrapolation
CFD Turbulent DLR Baldwin-Lomax (unpub)
i 1500_ ,u. 2000,,q
_" Boundary layer trip location
0 I I I I I 00 1
CFD Laminar GASP (Fief 17)
0.2 0.4 0.6 0.8X/L
(b) Longitudinal station, Y/L = 0.06.
I I I I I0 0.2 0.4 0.6 0.8 1
X/L
(b) Longitudinal station, Y/L = 0.06.
2000--
,u, 1500.=
" 500tn
0
2500 -
Phosphor Turbulent Extrapolation
CFD Turbulent DLIq-Baldwin-Lomax (unpub)
1000
, o5 5oo-
I I I I0.05 0.1 0.15 0.2
Y/L
(c) Axial station, X/L = 0.58.
Phosphor Laminar Extrapola_on
CFD Laminar GASP (Ref 17)
I I I0.05 0.1 0.15
Y/L
(c) Axial station, X/L = 0.58.
I0.2
28