46
NASA Technical Memoran_ 4435 i : -' °"--Tff .=; ] - ;y-_j.' J / J l :" ' " ,'" / , J- ":: V_ J Hypersonic Lateral a_d Directional Stability Characte cs of ;f Aeroassist Flight EXperiment Configuration in Air and CF 4 ZZ:ZI 7 = 5 " = John R. Micol and William L. Wells:;:?ii_2_ q_2 ' =:7::_Z Y:: : -7 :-_-T , JUNE 1993 (', ,'.:.-]" -_q35) HYPEv. S,._'IIC LAT:.kAL _".._ .;I_'_CTICI_L STA31LITY C_.ARACT_RISTICS C}F AEROASSIST r-LICHT ±X_'FR[MC_,T CO_FIGU_,ATION IN AIk ANI_ C_q (NASA) 42 p ';_3-:2 _166 Uncl as HI/02 0174946 https://ntrs.nasa.gov/search.jsp?R=19930019977 2018-06-18T06:56:39+00:00Z

Hypersonic Lateral a d Directional Stability Characte cs ... · Aeroassist Flight EXperiment Configuration in Air ... be used to perform the first phase ... model length ill symmetry

Embed Size (px)

Citation preview

NASA Technical Memoran_ 4435i

: -' °"--Tff

.=; ] •

- ;y-_j.'

J

/ •

J l :" ' " ,'"

/ ,

J- ":: V_J

Hypersonic Lateral a_d Directional

Stability Characte cs of;f

Aeroassist Flight EXperiment

Configuration in Air and CF 4

ZZ:ZI

7 =5 "

=

John R. Micol and William L. Wells:;:?ii_2_

q_2 '

=:7::_ZY ::: -7:-_-T,

JUNE 1993

(', ,'.:.-]" -_q35) HYPEv. S,._'IIC LAT:.kAL

_".._ .;I_'_CTICI_L STA31LITY

C_.ARACT_RISTICS C}F AEROASSIST

r-LICHT ±X_'FR[MC_,T CO_FIGU_,ATION IN

AIk ANI_ C_q (NASA) 42 p

';_3-:2 _166

Uncl as

HI/02 0174946

https://ntrs.nasa.gov/search.jsp?R=19930019977 2018-06-18T06:56:39+00:00Z

NASA Technical Memorandum 4435

Hypersonic Lateral and Directional

Stability Characteristics of

Aeroassist Flight Experiment

Configuration in Air and CF 4

John R. Micol and William L. Wells

Langley Research Center

Hampton, Virginia

National Aeronautics andSpace Administration

Office of Management

Scientific and TechnicalInformation Program

1993

Summary

Tile proposed Aeroassist Flight Experiment

(AFE) utilized a 14-ft-diameter raked and bhmted

elliptical cone to demonstrate the flight character-istics of space transfer vehicles (STV's). The AFE

was to be carried to orbit by and launched from

the Space Shuttle orbiter, where instrumentation for

10 on-board experiments would have obtained aero-dynamic and aerothermodynamic data for velocities

near 32000 ft/sec at altitudes above 245000 ft. A

preflight ground-based test program was initiated

to assess the aerodynamic arid aerothermodynamic

characteristics of the baseline concept and to pro-vide benchmark data for calibration of computational

fluid dynamics codes to be used in flight predictions.

The data reported herein are results from one phase

of this ground-based study. Static lateral and di-

rectional stability characteristics were obtained fortile AFE configuration at angles of attack from -10 °to 10 °. Tests were conducted in air at Mach num-

bers of 6 and l0 and in tctrafluoromethane (CF4)

at Mach 6 to examine tile effects of Mach number,Reynolds number, and normal-shock density ratio.

Changes in Mach number from 6 to 10 in air or

in Reynolds number by a factor of 4 at Mach 6 had

a negligible effect on the lateral and directional sta-

bility characteristics of the baseline AFE configura-tion. Variations in density ratio across the normal

portion of the bow shock from approximately 5 (air)

to 12 (CF4) had a measurable effect on lateral and di-

rectional aerodynamic coefficients, but no significant

effect on lateral and directional stability character-istics. The tests in air and CF4 indicated that the

configuration was laterally and directionally stable

through the test range of angle of attack.

Unfortunately, the AFE program was cancelledin late 1991. The realization of an AFE flight in the

future is possible but uncertain. Thus, this paper

documents the lateral and directional aerodynamiccharacteristics of the baseline AFE vehicle for use in

the design of fllture aeroassist space transfer vehicles.

Introduction

Among the space transportation systems pro-

posed for the future are space transfer vehicles

(STV's), which are designed to ferry cargo between

higher Earth orbits (for example, geosynchronous

and lunar orbits) and lower Earth orbit where theSpace Shuttle and Space Station Freedom will op-

erate. (This class of vehicle was formerly referredto as orbital transfer vehicles or OTV's.) Upon re-

turn of the vehicle from high Earth orbit, its velocity

must be greatly reduced to attain a nearly circular

low Earth orbit. This decrease in velocity can beachieved either by using retrorockets or by guiding

the vehicle through a portion of the atmosphere and

allowing aerodynamic drag forces to slow the vehi-

cle. Studies have shown that lower propellant loads

would be required for the aeroassist method (rcf. 1);thus, payloads could be increased.

Future STV's that will be designed to use Earth

atmosphere for deceleration are generally referred toms aeroassisted space transfer vehicles or ASTV's

(formerly AOTV's). These vehicles will have highdrag and a relatively low lift-to-drag ratio and will

fly at very high altitudes and velocities throughout

the atmospheric portion of the trajectory. Before theactual flight vehicle can be designed with optimal

aerodynamic and acrothcrmodynamic characteris-

tics, additional information about very high-altitude,

high-velocity flight is required. To obtain such in-

formation, a subscale flight was proposed wherebya 14-ft-diametcr ASTV configuration with 10 on-

board experiments would be launched from the Space

Shuttle and accelerated back into the atmosphere

with a rocket. This Aeroassist Flight Experiment

(AFE) would make a sweep through the atmosphereto an altitude of about 245 000 ft with a velocity of

nearly 32000 ft/sec to gain aerodynamic and aero-thermal information and return to low Earth orbit

for retrieval by the Space Shuttle. The on-board in-strumentation would measure and record the aero-

dynamic characteristics and aerothermodynamie en-

vironment of this entry trajectory, and the data

would be used to validate computational fluid dy-

namics (CFD) computer codes and ground-to-flightextrapolation of experimental data for use in future

ASTV designs. This flight experiment was proposed

because the high-velocity, low-density flow environ-

ment cannot be duplicated or simulated in presenttest facilities, nor can it be predicted with certainty

by existing techniques.

Naturally, the AFE wouht require an extensiveaerodynamic and aerothermodynamic experimental

and computational data base for its design and sue-

cessful flight. Present test facilities, in conjunction

with the best CFD codes, would provide this infor-

mation. For this reason, a preflight test programin ground-based hypersonic facilities (ref. 2) was

initiated to develop the required aerodynamic and

aerothermodynamic data base. This data base will

be used to perform the first phase of CFD computercode calibration. The experimental results presented

herein are part of an extensive ground-based test

program performed at the Langley Research Center.

Previous results are presented in references 3 6. The

details of the rationale for the flight experiment are

outlinedin reference7, and the set of experiments to

be performed is described in reference 8.

A primary concern for the AFE vehicle is the

aerothermal heating oil the fore- and aftbody thermal

protection system (TPS). Because of these aerother-mal concerns, low values of sideslip angles are desir-

able to minimize heating to the aftbody or payload

and to prevent large thermal fluctuations on the heatshield. Thus, an accurate knowledge of the lateral

and directional stability characteristics of the AFE is

required. (Lateral and directional stability require-ments for a low lift-to-drag aeromaneuvering vehicle

are discussed in ref. 9.)

CFD codes are not generally used to provide aero-

dynamic information for vehicles at sideslip angles.Computed lateral and directional stability charac-

teristics for the AFE would require calculations of

the entire body at various sideslip angles, thus in-

creasing computational time, complexity, and cost.Hence, determination of these stability characteris-

tics for the flight vehicle must rely on experimental

data obtained in ground-based facilities.

This paper addresses the effects of Mach number,

Reynolds nmnber, and normal-shock density ratio (a"real gas" simulation parameter) on lateral and direc-

tional aerodynamic characteristics measured on thebaseline AFE configuration. Tests were conductedat Mach 6 and 10 in air and at Mach 6 in tetra-

fluoromethane (CF4) through a range of angle of at-

tack and sideslip.

During the continuum-flow portion of the flight,the AFE vehicle is expected to undergo normal-shock

density ratios of about 18, whereas conventional hy-

personic wind tunnels that use air or nitrogen as the

test gas only produce ratios of 5 to 7. In flight, this

large density ratio results from dissociation of air asit passes into the high-temperature shock layer. This

real-gas effect may have a significant impact on shock

detachment distance, distributions of heating and

pressure, and aerodynamic characteristics (ref. 10).

For blunt bodies at hypersonic speeds, the pri-

mary factor that governs the shock stand-off distanceand inviscid forebody flow is the normal-shock den-

sity ratio. (See ref. 10.) Certain aspects of a real

gas can be simulated by the selection of a test gasthat has a low ratio of specific heats and provides

large values of density ratio. These conditions can

be obtained in the Langley Hypersonic CF4 Tun-nel, which provides a simulation of this phenomenon

by producing a density ratio of about 12 across the

shock. This tunnel, in conjunction with tile Lang-

ley 20-Inch Mach 6 Thnnel, provides the capability

to test a given model at the same free-stream Maeh

number and Reynolds number, but at two values of

density ratio (5.25 in air and 12.0 in CF4). Thus,

data for code calibration are provided that includethe effects of normal-shock density ratio. Tests were

performed in air at Mach 10 and through a range of

Reynolds numbers at Mach 6 to verify that aerody-

namic characteristics were independent of significantchanges in Math numbers and Reynolds numbers for

the blunt AFE configuration in hypersonic contin-UUln flOW.

However, the AFE program cancellation ended

the research efforts on this configuration. Thus,

this paper documents the lateral and directionalcharacteristics of the baseline AFE vehicle for use in

the design of future aeroassist space transfer vehicles.

Symbols

Cl

Cl 3

Cn f_

Cy

Cy_

d

M

P

q

Re2,d

S

T

U

X

X, y, z

rolling-moment coefficient,

Rolling momentqocdS

= ACI/A[3, per deg

yawing-moment coefficient,

Yawing momentqecdS

= ACn/A/3, per deg

side-force coefficient, Side forceqocS

= ACv/A[], per deg

model length ill symmetry plane,in.

Mach number

pressure_ psia

dynamic pressure, psia

unit free-stream Reynolds

number, ft- 1

postshock Reynolds immberbased on d

reference area, model base area,

in 2 (10.604 in 2 when d = 3.67 in.and 4.936 in 2 when d = 2.50 in.)

temperature, °R

velocity, ft/sec

moment transfer distance in axial

direction (fig. 4), in. (1.673 ill.when d = 3.67 in. and 1.559 in.

when d = 2.50 in.)

axial, lateral, and vertical coordi-

nates for AFE (fig. 4)

O_

/

P

monlent transfer distance in

normal direction (fig. 4), in.

(0.129 in. when d = 3.67 ill. and

0.0979 in. when d = 2.50 in.)

angle of attack, (leg

angle of sideslip, deg

ratio of specific heats of tile test

gas

density of tile test gas, lbm/in :_

Subscripts:

t total conditions

oc free-stream conditions

2 conditions behind the normalshock

AFE Configuration

The AFE flight vehicle wouM consist of a 14-ft-

diameter drag |)rake, an instrument carrier at thebase, a solid-rocket propulsion motor, and smallcontrol motors. A sketch of the vehicle is shown

in figure 1. The drag brake (fig. 2), which is the

forebody configuration, is derived from a bhmted60 ° half-angle elliptical cone that is raked at 73 °

to the cone centerline to produce a circular raked

plane. A skirt, with an arc radius equal to one-

tenth the rake-plane diameter and with an arc length

correspondiilg to 60 ° has I)('en attached to the rake

plane to reduce aerodynamic heating around the I)auset)eril)hery. The t)hmt nose is an ellipsoid with an

ellipticity equal to 2.0 in the symmetry plane. The

ellipsoid nose and the skirt are at a tangent at their

rest)ective intersections to the elliptical cone surface.A detailed description of the forebody analytical

shat)e is presented in reference 11.

Apparatus and Tests

Facilities

Langley 31-Inch Maeh 10 Tunnel. The

Langley 31-Inch Math 10 _iSnmel (formerly the Lang-ley Continuous Flow Hypersonic Tunnel) expands

dry air through a three-(limensional coIltoured nozzle

to a 31-in-square test secti(m to achieve a nonfinal

Math number of 10. The air is heated to approxi-mately 1850°R by an electrical resistance heater, and

the maximum reservoir pressure is approximately

1500 psia. The tunnel operates in the blowdown

mode with run times of approximately 60 sec. Forceand moment data can be obtained through a range

of angle of attack or sideslip during one run |)y uti-

lization of the t)itch-pause cat)al)ility of the model

support system. This tunnel is described in moredetail in reference 12.

Langley 20-Inch Mach 6 Tunnel. The 20-Inch Mach 6 _Dmnel is a blowdown win(t tunnel that

uses dry air as the test gas. The air may be heated to

a maxinmm temperature of approximately 1100°R byan electrical resistance heater: the maximum reser-

voir pressure is 525 psia. A tixed-geometry, two-

dimensional, contoured nozzle with parallel side walls

expands the flow to a Mach number of 6 at the 20-in-

square test. section. The Ino(tel injection mechanism

allows changes in angle of attack and sideslip during

a run. Run durations are usually 60 to 120 sec, al-though longer times can I)e attained by connection

to auxiliary vacuum storage. A description of this

facility and the calibration results are t)resented inreference 13.

Langley 20-Inch Maeh 6 CF, I Tunnel. The

20-Inch Mach 6 CF.1 Tunnel is a blowdown wind

tunnel that uses CF4 a,s the test gas. The CF4

can 1)e heated to a maxinmm temperature of 153(}°F1

by two molten lead t)ath heat exchangers comm('tedin t)aralM. The maxinmm pressure in the tunnel

reservoir is 2600 psia. Flow is exl)ande(t through an

axisymmetric, contoured nozzh' designed to generatea Math mmfl)er of 6 at the 20-in-diameter exit. This

facility has an open-jet test section. Run duration

can be as long as 30 sec, but 10 sec is sufficient

for most te.sts because the model injection system

is not presently cat)able of changing angle of attackor sideslip (luring a run. A detaih'(t description

of the 20-Inch Mach (i CFI tunnel is presented inreference 14.

Just before the present test series, the tunnel wa,s

modified extensiw_ly. Included in those modificationswere a new nozzle, a new test section and model in-

je(:tion system, a new diffuser, an(t iInprovements in

wiring of the controls and of the data acquisition

system. The new nozzle was designed to improve

flow quality along th(' centerline and to more ch)selymatch the Math number in the Math 6 air tunnel

that is often used to I)roduce data for comparison

with the CF4 data. Calibration results (ref. 15) thatwere obtained after the new nozzle was installed indi-

(:ate greatly improved flow unifornfity near the nozzlec(;nterline. For tim present test series, the model was

tested on the tunnel centerline. Previously, modelswere tested off centerline to avoid flow disturbances.

(See ref. 14.)

3

Models

Two aerodynamicmodelswerefabricatedandtested.Themodelswereidenticalexceptforsize;thebaseheights(d in fig.2)at thesymmetryplanewere3.67in. (2.2percentscale)asshownin figure 3(a) and

2.50 in. (1.5 percent scale) as shown in figure 3(b).

Tile 3.67-in-diameter model is made in three parts

a stainless steel forebody (aerobrake), an ahmfinum

aftbo(ly (instrument carrier and propulsion motor),and a stainless steel balance holder. The 2.50-

in-diameter model, shown mounted in tile Langley

20-Inch Ma(:h 6 CF 4 Tunnel in figure 3(c), is fabri-cated of aluminum and does not include the circu-

lar or hexagonally shaped aftbody and tile sinmlated

propulsion motor of previous models that were tested

(ref. 16). A cylinder protrudes from tile base to ac-cept the balance. The acute angle between the bal-

ance and cylinder axis and the base in the symmetry

plane is 73 °. Tile 2.50-in-diameter model was fabri-cated to provide an air gap between the end of tile

balance and the end of the cavity in the forebody;

its purpose was to reduce conductive heating. Forboth Inodels, shrouds were built to shield tile bal-ance from base-flow closure. The shrouds attach to

tile sting, and clearance was provided to avoid in-

terference with the balance during model movementwhen forces and moments were applied. The fore-

bodies were machined to the design size an(I shape

within a tolerance of +0.003 in. Angle of attack (see

fig. 2) and sideslip (see fig. 4) in this paper are refer-enced to the axis of the original elliptical cone.

Instrumentation

Aerodynamic force and moment data were mea-

sured with sting-supported, six-component, water-

cooled, internal strain gauge balances. Two ther-

mocouplcs were installed in the water jacket thatsurrounds tile measuring elements to monitor inter-

nal balance temperatures. The load rating for each

component of tile two balanccs (one for each model

size) is presented in table I. The calibration accuracyis 0.5 percent of the maximuin load rating for each

component.

Test Conditions

The tests were conducted at nominal free-stream

Mach numbers of 6 and 10 in air and at Mach 6

in CF4. (Nominal test conditions are presented in

table II.) The angles of attack for Mach 6 in air were

0° and -1-5° with nominal sideslip anglcs of 0°, -2 °,and -4 °. Tests at Mach 6 in CF4 were at angles of

attack of 0 ° , +5 ° , and +10 ° with nominal sideslip

angles of 0 °, +2.5 ° , and ±5°; at Mach 10 (exceptfor a = -2.5 °, where only a negative /_ sweep was

4

performed), the angles of attack were 00, ±2.5 °, ±5 °,and ± 10 ° with nominal sideslip angles of 0 °, ±2 °, and±4 ° .

Test Procedures

Bhmt models are conducive to heat conduction

through the forcbody face during a run, which gener-

ally produces a gradual increase in temperature gra-dients along the balance even though the balance is

water cooled. Because temperature gradicnts wcrc

not accounted for in the laboratory calibration of thebalance, efforts were made to minimize these gradi-

ents by limiting the test times. In the 20-Inch Mach 6CF4 Tunnel, the model was mounted at the desired

angle of attack and sideslip before the run. After thetest-stream flow was cstablishcd, the model was in-

jected to the test-stream ccntcrline. Data wcrc gath-crcd for approximately 5 sec, then the model was re-

tracted. In the air tunnels, the model was mounted

at (_ = fl = 0° before the run. After test-streamflow was established, the model was injected to the

stream ccnterlinc, then pitched to the next angle of

attack (or sideslip angle) by the pitch-pause mech-anism. Data wcrc taken while the model was sta-

tionary at each position. The balance thcrmocouplcswcrc monitored during each run to assure that the

temperature gradient within the balance remained

within an acceptable limit. Typical run timcs for a

set of _ and fl sweeps in the air facilities wcrc about15 sec.

Data Reduction and Uncertainty

Each of the thrcc test facilitics has a dedicated

stand-alone data system. Output signals from thebalances were sampled and digitized by an analog-

to-digital converter, then stored and processed by

a computer. The analog signals wcrc sampled at

a rate of 50 per second in the Mach 6 CF4 andMach 10 air tunnels and at 20 per second in the

Mach 6 air tunnel. A single value of data reported

herein represents an average of values measured for2 scc in the Mach 6 CF4 and Mach 6 air tunnels andfor 0.5 sec in the Mach 10 air tunnel. Corrections

were made for model tare weights at each angle ofattack and for interactions between different elements

of the balances. Corrections were not made for base

pressures.

Balance-related calculated uncertainties in the

measured static aerodynamic coefficients are given intable III. Thcsc uncertainties are based on balance

output signals related to forces and moments by a

laboratory calibration that is accurate to ±0.5 per-

cent of the rated load for each component. (S_e ta-

ble I.) For the AFE, the moment reference center is

locatedat thecenterof therakeplane.(Seefig. 4.)Thus,momentsreducedaboutthemodelrake-planecenterandreportedhereinhavegreateruncertainties

o

than those measured at the balance moment center.

The yawing and rolling moments at the balance have

an uncertainty of only +0.5 percent of the rated load,

whereas the moment at the rake-plane center also in-dudes uncertainties associated with the forces in the

transfer equation. The transfer equation is

Yawing momentR/, = Yawing monmntl/ - (X) (Side fi)rce)

and

Roiling moment/t/, = Rolling monmnt/_ - (Z) (Side force)

where the subscripts RP and B denote tile rake-plane

center and the balance moment center, respectively.

The transfer distances X and Z are defined in fig-

ure 4. In coefficient form, the uncertainty A relatedto the balance calibration for the side force is

=t=(0.005) (Force rating)

A C y = q:x:S

Tire uncertainty for the yawing moment is

=t=(0.005) (Moment rating)

q_dS

and an identical equation applies for the rolling mo-ment. These balance uncertainties are sufficient for

measurements at the balance moment center. How-

ever, at the rake-plane center, the yawing-momentmmertainty is

0.5

AC,,.x_p = :t: (AC,.t_) _ + AQ_ _-

and the rolling-moment mwcrtainty is

(}.5

[ /AG._p = :k (AG._) 2 + AC:j-_

Note that all the terms inclu<te the free-stream dy-

namic pressure in the denonfinator so that the m_-certainties are less at test conditions where qzc is

large that is, at a higher Reynolds number rather

than at a lower Reynohts mmfl)er. The uncertainty

in dynamic pressure is +3 percent. The flow condi-

tions for which the present uncertainties have beencalculated are presente(t in table II.

Results and Discussions

The aerodynamic data from the Mach 10 air testsare tabulated in table IV. The Mach 6 results arc

presented in tables V and VI for air and in table VIIfor CF4. The test Reynolds number and modeldiameter are indicated in each table title.

The aerodynamic coefficients Cy, Cn, and C l are

plotted for an angle-of-sideslip range at various an-gles of attack in each facility and presented in fig-

ures 5 -7 for Math 10 in air, Mach 6 in air, and Mach 6

in CF4, respectively. Data obtained at Maeh 6 in

air (fig. 6) indicated no effect of Reynolds numberon measured lateral and directional coefficients for

a factor-of-4 increase in postshock Reynolds num-

ber. (Similar trends with respect to Reynolds num-

ber were also observed for AFE longitudinal acro-

dynamic characteristics presented in ref. 16 in which

a negligible effect of Reynohts number was notedfor Maeh 6 and 10 in air and at Mach 6 in CF4.)

Therefore, the assumption is made that the effect

of Reynolds number on measured lateral and direc-tional data at Mach 10 in air and Math 6 in CF4

is also negligible. The data are amenable to linear

curve fit.s as shown in figures 5 7, for which the ordi-

nate scale is quite sensitive. These curves wouht be

expected to go through the origin because the mode/

wa,s symmetrical about the pitch plane. However, asobserved in figures 5 7, an offset exists. This offset

may be attributed to model misalignment or t.o any

small stray signal in the data system that could cause

a constant data offset because of the very small val-ues being measured relative to the load range of thebalance.

For example, if a slight misalignment of the model

in tire roll direction were introduced during modelsetup or if the balance location within the model were

slightly misaligned, thereby producing a small offset

in the center of gravity location (that is, within a

few thousandths of an inch) in the side plane (y di-rection in fig. 4), then the effect of the large axial-

force component on this small moment arm may pro-

duce a continuous bias in the measured quantities,For instance, from reference 16 at (_ = 3 = 0 °,

Re:,c = 0.46 x 106/ft, and Mach 6 in CF4, the axial-

force coefficient is 1.382. The yawing-moment coeffi-cient, from table VII for similar conditions, is 0.004.

In much the same way as the change in the cen-

ter of pressure in longitudinal aerodynamics is lo-cated, forming the ratio of yawing-moment coeffi-

cient to axial-force coefficient yields the moment arm

in the y direction, which for this case is approxi-

mately 0.003 in. and thus within acceptat)le fabri-cation tolerances. A second linear curve, paralM 1o

the data-faired curve, is drawn through the origin in

each part of figures 5 7. Values from measurements

and the curve through the origin of figures 5 7 are

presented in tables IV VII. Use of the slopes of these

parallel curves through the origin to represent the

lateral and directional stability derivatives should be

valid because the data curves are linear through the

test sideslip range.

The lateral and directional stability derivatives

are presented in figure 8 and table VIII through the

range of angle of attack for which tests were per-

formed in each facility. For all test conditions, the

configuration was laterally and directionatly stable,

as indicated by the positive values of Cn;_ and nega-

tive values of Cl/_. A comparison of lateral and direc-

tional stability derivatives obtained at Mach num-

bers of 6 and 10 in air illustrates no significant

effect of Mach number on stability characteristics; a

comparison of these stability derivatives with those

obtained at Mach 6 in CF4 indicates a small but

measurable effect of normal-shock density ratio on

lateral and directional stability characteristics. Al-

though the numerical values for air and CF4 are not

greatly different, the data trends in air and CF4 ap-

pear to be opposite. (Similar trends were observed

in the longitudinal aerodynamic characteristics dis-

cussed in ref. 16.) This trend is most obvious for

Ci,3, wherein the small numerical values require an

expanded scale on the graph. The wind tunnel re-

sults in CF4 are believed to be a better sinmlation

of flight data than those in air because the shock de-

tachment distance for CF4 is closer to the distance

predicted tot the actual flight case. (For example, see

refs. 6 att(t 16.)

Concluding Remarks

Static lateral and directional stability character-

istics were obtained for the Aeroassist Flight Exper-

iment (AFE) configuration through a range of angle

of attack from -10 ° to 10 ° . Tests were conducted

on two different-sized models at Mach numbers of 6

and 10 in air and at a Mach number of 6 in tetra-

fluoromethane (CF4). The effects of Maeh number,

Reynolds number, and normal-shock density ratio on

lateral and directional stability characteristics were

examined.

Changes in Mach number from 6 to 10 in air or

in Reynolds number by a factor of 4 at Maeh 6 had

a negligible effect on the lateral and directional sta-

bility characteristics of the baseline AFE configura-

tion. Variations in density ratio across the normal

portion of thc bow shock from approximately 5 (air)

to 12 (CF4) had a measurable effect on lateral and

directional aerodynamic coefficients, but no signifi-

cant effect on lateral and directional stability char-

acteristics. The tests in air and CF4 indicated that

the configuration is laterally and directionally stable

through the test range of angle of attack as indicated

by the positive values of Cnz and negative values of

C/:_ (positive effective dihedral).

In late 1991, the AFE program was cancelled and

thus ended research efforts on this configuration. The

realization of an AFE flight in the future is possible

but uncertain. Hence, this paper documents the

lateral and directional aerodynamic characteristics

of the baseline AFE vehicle for use in the design of

future aeroassist space transfer vehicles.

NASA Langley Research Center

Hampton, VA 23681-0001

March 25, 1993

References

1. Walberg, Gerald D.: A Review of Aeroassisted Orbit

Transfer. AIAA-82-1378, Aug. 1982.

2. Wells, William L.: Wind-T_mnel Preflight Test Program

for Aeroassist Flight Experiment. Technical Papers

AIAA Atmospheric Flight Mechanics Conference,

Aug. 1987, pp. 151_ 163. (Available as AIAA-87-2367.)

3. Wells, William L.: Free-Shear-Layer Turning Angle in

Wake o| Aeroassist Flight Experiment (AFE) Vehicle at

Incidence in M = 10 Air and M = 6 CF4. NASA TM-

100479, 1988.

4. Micol, John R.: Experimental and Predicted Pressure, and

Heating Distributions for Aeroa,ssist Flight Experiment

Vehicle. ,l. TheTwmphys. _ Heat Transl., July Sept. 1991,

pp. 301 307.

5. Wells, William L.: Surface Flow and Heating Distributions

on a Cylinder in Near Wake of Aeroassist Flight Experi-

ment (AFE) Configuration at Incidence in Mach 10 Air.

NASA TP-2954, 1990.

6. Micol, John R.: Simulation of Real-Gas Effects on Pres-

sure Distributions for Aeroassist Flight Experiment Vehi-

cle and Comparison With Prediction. NASA TP-3157,

1992.

7. Jones, Jim J.: The Rationale for an Aeroassist Flight

Experiment. AIAA-87-1508, June 1987.

8. Walberg, G. D.; Siemers, P. M., III; Calloway, R. L.;

and Jones, J. J,: The Aeroassist Flight Experiment. IAF

Paper 87-197, Oct. 1987.

9. Gamble, Joe D.; Spratlin, Kenneth M.; and Skalecki, Lisa

M.: Lateral Directional Requirements for a Low L/D

Aeromaneuvering Orbital Transfer Vehicle. A Collection

of Technical Papers AIAA Atmospheric Flight Mechan-

ics Conference, Aug. 1984, pp. 402 413. (Availabh. as

AIAA-84-2123.)

10.Jones,RobertA., andHunt,JamesL. (appendixAbyJamesL. Hunt,KathrynA. Smith,andRobertB.ReynoldsandappendixBby,JamesL. HuntandLillianR.Boney):L,_eof Tetrafluoromethane To Si_rtulate Real-

Gas Effects on the Hypersonic Aerodynamics of Blunt

Vehicles. NASA TR R-312, 1969.

11. Cheatwood, F. McNeil; DeJarnette, Fred R.; and ttamil-

ton, H. Harris, II: Geometrical Description for a Pro-

posed Aeroassisted Fli,qht Experinzent l,'_h_ele. NASA TM-

877'14, 1986.

12. Miller, C. G.: Langley Hypersonic

Aerothermodynamic Testing Capabilities

Future. AIAA-90-1376, June 1990.

Aerodynamic/

Present and

13. Miller, Charles G., Ill; an(t Gnoffo. Peter A.: Pressure

Distributions and Shock" Shapes for 12.84°/7 ° On-Axis

and Bent-Nose Bieonies in Air at Maeh 6. NASA TM-

83222, 1981.

14. Midden, Raymond E.', and Miller, Charles G., IIl: De-

scription and Calibratior_ of the Langley Hypersome CF4

Tunnel A Facility for Simulating Low "_ Flow as Oeeutw

for a Real Gas. NASA TP-2381, 1985.

15. Micol. John R.; Midden, Raymond E.; and Miller, Charles

G., llI: Langh,y 20-In('h Hypersonic CF4 Tmmel: A Facil-

ity for Simulating Real-Gas Effects. AIAA-92-3939. July

1992.

16. Wells, William L.: Measured a_d Predicted Aerody_mmie

CoeJJicien_s and Shock Shapes for Aeroassist Fhght Exper-

iment (AFE) ConfiguratioTt. NASA TP-2956, 1990.

Table I. Balance Load Ratings Used in Tests

Load rating

Model size, in. Side, lbf Roll, in-lbf3

15

2.50

3.67 Yaw, in-lbf2 3

10 15

ft- l

1.09 x 106 [

0.63 x 10 (;

2.21

Table II. Nominal Test Conditions

ll_] I T_' [- P_' T_ U_¢, q_ ------_in2 I R llbf/in2 R Moc I ft/see lbf/in 2 P2/P_c I

Re2,ft-1

70o118,00.017519,.719.9014651I 1.20I 6.oLangley 20-Inch Mach 6 Tunnel; air as test

30 845

126 910

Langley 31-Inch Maeh 10 Tunnel; air as test gas

I0.87 x 105 ]

gas

0.023 108.3 5.84 2975 0.54 5.2 0.96 x 105

.084 112.5 5.94 3095 2.10 5.2 3.50

Langley 20-Inch Maeh 6 Tunnel; CF4 as test gas

0.46x106 1515 1280 0.063 1386 5.87 3000 I 1.29 11.8 1.77

2.23 11810 1.34

1.00 845 1.40

3.86 910 1.40

_ 2.54 1271 1.10

Table III. Balance-Related Uncertainties in Experimental Lateral

and Directional Aerodynamic Coefficients

Re2, d I d, in. I ACl, i ACn, -t- ACy, _

2$toc = 10 air as test gas

18042 l 2.50 l 0.0007 i 0.0019 l 0.0025

M_c = 6; air as test gas

29400 t 3.67 I 0.0024 0.0069 0.0131107000 3.67 .0006 .0018 .0034

Moc = 6; CF4 as test gas

37000 L 2.50 _ 0.0006 I 0.0017 l 0.0023

8

TableIV. MeasuredLateralandDirectionalAerodynanlieCharacteristicsin Airfor Aide,= 9.90,/Re_= 1.09x 106/[t,Re'2.d = 18042, and d = 2.50 in.

[Numbers with asterisk are derived from curve through 0,(/in fig. 5]

(a) c_= 0°

_,deg

-0.019

.098

-.024

-.019-.015

-.054

-.031

.052-2.113

-2.157

-2.063

-2.085

2.242

2.0992.225

2.075

-4.296

-4.3024.364

4.350

O*-2.0*

2.0*

-4.0*4.0*

c_

0.00096

.00101

.00126

.00111

.00138

.00103.00155

.00171

.00281

.00309

.00251.00308

-.00041

-.00021

-.00034-.00039

.00426

.00463

-.00173-.00171

O*

.00145*

-.00145*.00290*

-.00290*

C71

0.00161.00328

.00197

.00181.00259

.00252

.00306

.00375

-.00248

-.00152-.00169

-.00103

.00589

.00597.00676

.00667

-.00666

-.00520.01030

.01058

O*-.00377"

.00377*

-.00755"

.00755*

Cy

-- 0.00826.00273

.00562

.00775

.00314

.00650

.00327

.00386

.01521

.01305

.01238.01323

-.00079

.00098

-.00464-.00221

.02418

.02051

-.00987

-.01067

O*.00740*

-.00740"

.01480*-.01480*

L

9

TableIV. Continued

(b) a = 2.5°

_, deg G C. Cy-0.011

-2.114

-4.316

-2.090

.065-.017

2.233

4.410

2.130-.003O*

-2.0*

2.0*

-4.0*4.0*

0.00104

.00282

.00438

.00253.00101

.00110

-.00032

-.00228-.00071

.00126O*

.00152'-.00152'

.00304*

-.00304*

0.OO188

.00222

.00699

.00221

.00241

.00192

.00632

.01057

.00645

.00282O*

-.00402*.00402*

-.00804*.00804*

0.00692

.01328

.02341

.01468

.00591

.00554

-.00226

-.00934-.00284

.00379O*

.00757*-.00757*

.01514*

-.01514"

(c) a = -2.5 °

_, deg

-0.022

-2.169

-4.280-2.083

.0550*

-2.0*

-4.0*

0.00126

.00240.00425

.00262

.00111O*

.00140'

.00280"

CTt

0.00219-.00261

-.00626

-.00168

.00265O*

-.00408"

-.00816*

0.00426

.01508

.02186

.01270

.00359O*

.00850*

.01700"

10

_, deg-0.006

2.2454.3802.076-.037-.013

-2.113-4.252-1.998

.078O*

-2.0*2.0"

-4.0*4.0*

TableIV. Continued

(d) a = 5.0°

CI

---_0.00116

-.00057

-.00225

-.00048

.00113

.00136

.00307

.00479

.00293

.00122

O*.00164*

-.00164*

.00328*

-.00328*

C71

0.00181.00682

.01168

.00674

.00227

.00206

-.00308

-.00775

-.00254

.00251

O*-.00452*

.00452*

-.00903*

.00903*

Cy

0.00645

-.00293

-.01189-.00302

.00667

.00528

.01548

.02400.01485

.00536

O*.00841*

-.00841*

.01681*

-.01681*

(e) a= -5.0 °

2, deg CI C,, C_0.00182 0.00544

-0.027

2.1914.379

2.050

-.065

-.018

-2.155-4.319

-2.057

.055

O*-2.0*

2.0*

-4.0*

4.0*

0.00127-.00045

-.00191

-.00036.00120

.00128

.00266

.00664

.01057.00660

.00249

.00202

-.00198

-.00389

-.01165-.00391

.00389

.00500

.01200

.00394

.00245.00092

O*

.00136*-.00136*

.00273*

-.00273*

-.00607

-.00172

.00225

0*-.00388*

.00388*

-.00777*

.00777*

.02079

.01178

.00498

O*

.00746*

-.00746*.01492*

-.01492*

11

TableIV. Concluded

(f) c_ = 10.0 °

/3, deg

-0.0152.237

4.334

2.096

.004

-.015-2.068

-4.203

-2.012

.1140*

-2.0*2.0*

-4.0*

4.0*

G0.00071

-.00148-.00346

-.00152

.00065

.00040

.00260

.00417

.00227

.00011O*

.00182*-.00182*

.00363*

-.00363*

0.00167

.00798.01335

.00780

.00264

.00224-.00342

-.00868

-.00260

.00331O*

-.00517*

.00517*

-.01034*

.01034*

0.00508

-.00585

-.01427

-.00523

.00383

.00421

.01295

.02153.01033

.00200O*

.00833*

-.00833*

.01665*

-.01665*

(g) (_= -10.0 °

3 deg

-0.017

2.209

4.2892.104

.011

-.015-2.141

-4.201

-1.999

.098O*

-2.0*

2.0*

-4.0*4.0*

0.00145

-.00054

-.00228-.00053

.00119

.00123

.00311

.0O472

.00303

,00129O*

.00167*

-.O0167*

.00334*

-.00334*

0.00066

.00590

.01080

.00669

.00212

.00056-.00414

-.00836

-.00340

.001920*

-.00460*.00460*

-.00920*

.00920*

0.00584

-.00347

-.01404

-.00536.00217

.00612

.01723.02457

.01536.00446

O*

.00933*

-.00933*.01865*

-.01865*

12

TableV. MeasuredLateralandDirectionalAerodynamicCharacteristics in Air for

Aim = 5.84, Rem= 0.63 × 106/ft Re2,d = 29 400, and d = 3.67 in.

[Numbers with asterisk are derived from curve through 0,0 in fig. 6]

(a) _ = 0°

(3, deg CI Cn Cy

0.0

-2.0

-4.00*

-2.0*

-4.0*

-0.0004.0010

.0026

O*.0015"

.0030*

0,0003-.0036

-.0078

O*-.0043*

-.0085*

0.0018.0093

.0176

O*.0080"

.0159"

(b) a = 5 °

i3, deg CI C,_ C u

0.0

-2.0

-4.00*

-2.0*

-4.0*

-0.0003

.0013

.00310*

.0017"

.0034*

0.0006-.0040

-.0086

O*-.0046*

-.0092*

0.0022.0102

.0183

O*.0079*

.0158"

_,deg C1

0.0

-2.0

-4.0

0*

-2.0*-4.0*

-0.0004

.0012

.0027

O*

.0015'

.0030*

(c) a = -5 °

C_

0.0007

-.0036-.0078

O*

-.0041"

-.0081"

Cy

0.0018

.0101

.01840*

.0080*

.0160'

13

TableVI. MeasuredLateralandDirectionalAerodynamicCharacteristicsin Air forM_ = 5.94, Re_c = 2.21 x 106/ft, Re2, d = 107000, and d = 3.67 in.

[Numbers with asterisk are derived from curve through 0,0 in fig. 61

,5', deg

0.0

-2.0

-4.0

O*

-2.0*

-4.0*

(a) a = 0°

@ c_-0.0002 0.0004

.0014 -.0039

.0030 -.00820* 0*

.0015" -.0043*

.0030" -.0085*

c_

.0097

.01780*

.0080*

.0159"

(b) a=5 °

_, deg

0.0

-2.0

-4.0O*

-2.0*

-4.0*

@-0.0003_

.0014

.0031O*

.0017".0034"

c_

0.0006_

-.0040-.00850*

-.0046*

-.0092*

c_0.0020

.0097

.0175O*

.0079*

.0158"

8, deg

0.0

-2.0-4.0

O*

-2.0*-4.0*

(c) a = _5 °

.0013

.00280*

.0015"

.0030*

c_

o.ooo-$_-.0035

-.0076O*

-.0041"

-.0081"

c_0.0019

.0095

.01740*

.0080*

.0160"

14

TableVII. MeasuredLateralandDirectionalAerodynamicCharacteristicsin CF4forMx_ = 5.87, Revc = 0.46 x 106/ft, Re2,d = 37000, and d = 2.50 in.

[Numbers with asterisk are derived from curve through 0,0 in fig. 71

(a) c_= 0°

_, deg

0

0

2.502.50

5.0

5.0

-2.50-2.50

-5.0

-5.0

0*-2.5*

2.5*

-5.0*

5.0*

CI Cn

0.0000

.0008

-.0016

-.0015-.0035

-.0035

.0032

.0033

0.0037

.0036

.0094

.0094

.0144

.0148

-.0028-.0029

.0048

.0053

O*

.0024*-.0024*

.0047*

-.0047*

-.0084

-.0085

0*

-.0064*-.0064*

-.00128*

.00128*

Cy

0.0017

.0021

-.0084-.0086

-.0179

-.0187

.0136.0136

.0234

.0237

O*.0108"

-.0108"

.0215"-.0215"

(b) a = 5°

_,deg Cl Cn Cy0.0026

0.0

2.50

5.0-2.50

-5.0

O*-2.5*

2.5

-5.0*

5.0*

0.0007

-.0012-.0030

.0027

.0046

O*.0020*

-.0020*

.0039*-.0039*

0.0035

.0084.0133

-.0018

-.0066

O*-.0050*

-.0050*

-.0099*.0099*

-.0061-.0152

.0121

.0212

O*.0093*

-.0093*

.0185"-.0185"

15

TableVII. Continued

(c) _ = -5 °

fl, deg Cl Cn Cy0.0

2.505.0

-2.50

-5.0O*

-2.5*

2.5*

-5.0*5.0*

0.0008

-.0014

-.0037

.0033

.0053O*

.0023*

-.0023*.0045*

-.0045"

0.0035

.0091

.0159-.0028

-.0083O*

-.0060"-.0060*

-.00120"

.00120*

0.0018

-.0094-.0216

.0138

.0245O*

.0115"

-.0115"

.0230"-.0230*

(d) a = 10 °

fl, deg Cl Cn Cy0.0

2.50

5.0

-2.50

-5.0O*

-2.5*2.5*

-5.0*

5.0*

0.0007-.0011

-.0030

.0026

.0045O*

.0019"

-.0019"

.0037*

-.0037*

0.0034

.0084

.0137

-.0018

-.0070O*

-.0050*

.0050*

-.00100'

.00100'

0.0027

-.0060

-.0152.0123

.0214O*

.0093*

-.0093"

.0185"

-.0185"

16

TableVII, Concluded

(e)a = -10°

i3, deg CI Cn C,_

0.0

2.505.0

-2.50

-5.0O*

-2.5*

2.5*

-5.0*

5.0*

0.0011

-.0011-.0029

.0031

.00490*

.0020'

-.0020*

.0040*

-.0040*

0.0035

.0086

.0138

-.0041

-.0078O*

-.0053*

.0053*

-.00106*.00106*

0.0010-.0101

-.0213

.0131

.0248O*

.0118'

-.0118"

.0235*-.0235*

17

TableVIII. LateralandDirectionalStabilityCharacteristics

(a) Mc¢ = 9.90; Reoc = 1.09 × 106/ft; Re2, d --- 18 042; d = 2.50 in.; air as test gas

c_, Cyst, Cn_, Ct_s,

deg deg- 1 deg- 1 deg- 1

-9.80

-4.77-2.31

.21

2.73

5.26

10.35

-0.004663

-.003730

-.004249-.003700

-.003785

-.004203

-.004163

0.002299

.001942

.002040

.001887

.002010

.002258

.OO2585

-0.000834

-.000682

-.000699-.000725

-.000760

-.000821

-.000908

(b) M_c = 5.84; Re_c = 0.63 × 106/ft; Re2, d = 29400; d = 3.67 in.; air as test gas

deg-5.0

0

5.0

cuz,deg- 1 deg - 1 deg- 1

-0.0040

-.0040

-.0040

0.0020

.0021

.0023

-0.00075-.00075

-.00084

(c) M_ = 5.94; Re_ = 2.21 × 106/ft; Re2, d = 107000; d = 3.67 in.; air as test gas

deg deg -1 deg -1 deg -1-5.0 -0.0040 0.0020 -0.00075

0 -.0040 .0021 -.00075

5.0 -.0040 .0023 -.00084

(d) M_ = 5.87; Rec¢ -- 0.46 × 106/ft; Re2, d = 37000; d = 2.50 in.; CF4 as test gas

deg deg -1 deg -1 deg -1-10.0

-5.00

5.0

10.0

-0.0047

-.0046

-.0043-.0037

-.0037

0.00212

.00240

.00256

.00198.00200

-O.OOO80

-.00090

-.00094

-.00078-.00074

18

Top view14 ft

Front view

b---t--

%

- Forebody(aerobrake)

C

[

[

tU_

I

\

Instrumentcarrier

_ Rocket

motor

Side view

Figure 1. AFE flight vehicle configuration.

19

I

!

Flow

Rake plane

Ellipsoid nose-skirt

junction --_

Ellipsoid region

\

Ellipsoidjunction

Cone re.(

/

d

Base plane

Rake plane

Cone-skirt junction

Figure 2. AFE vehicle forebody development from elliptical cone.

2O

BLACK AND WHITE PHOfOGRAFH

(a) The 3.67-in-diameter model, aftbody and balance holder.

Balance shroud

Figure 3.

=,i_:{ !:.._ -=

L-90-11872

(b) Tit{' 2.50-in-diameter model and balance shroud.

The AFE models used in lateral and directional aerodynamics tests.

21

#Jf-_Ac;_,. A,.,,_: +_w*_!TE. P_"IOTQ(_f'7A'Fh

(c) Tile 2.50-in-dianmter model mounted ill Langley 20-hmh Mach (i CF4 Tunnel.

Figure 3. Concluded.

L-90-14502

22

Side force

Y

Rolling moment

Flow_______----_p-

(o_= o°)

Referencemomentcenter

(rake-plane center)

+0_

f

Rake plane

/--- Balance/ moment

_ , _ _enter

_ Yawing moment

Z

Figure 4. System of axes with positive direction of forces, moments, velocities, and angles indicated.

23

Cy

.025

.020

.015

.010

.005

0

-,005

-.010

-.015

-.020

-.025

\

I I I I

-5.0 -4.0 -3.0 -2.0 -1.0 0.0

9, deg

\

}

\\\

I I i I I

1.0 2,0 3.0 4.0 5.0

13,deg

o 0 to 4.289

[] 0 to -4.201

Measurement, linear fitParallel line through origin

Measurement uncertaintyfor all data sets as indicated

O n

.014

.012

.010

.008

.006

.004

.002

0

-.002

-.004

-.006

-.008

-.010

-.012

-.014

tl

I i I I

J

I I , i

-5.0 -4.0 -3.0 -2.0 -1.0 0.0 1.0 2,0 3.0

i I

4.0 5.0

_,deg

CI

.OO5

.004

.003

.002

.001

0

-.001

-.002

-.003

-.004\

-.005 , , , , i i _ _ J-5,0-4.0-3.0-2.0-1.0 0.0 1.0 2.0 3.0 4.0 5.0

_, deg

(a) c_ =-10.0 °.

Figure 5. Variation of lateral and directional aerodynamic characteristics with angle of sideslip in air atM_o = 9.90 and Re_ = 1.09 x 106/ft.

24

Cy

.025

.020

.015

.010

.005

0

-.005

-.010

-.015

-.020

-.025 I I I I

}

½

I I I I I

-5.0 -4.0 -3.0-2.0 -1.0 0.0 1.0 2.0 3.0

13,deg

4.0 5.0

I], deg

o 0 to 4.379

o 0 to -4.319

Measurement, linear fit

Parallel line through origin

Measurement uncertaintyfor all data sets as indicated

C n

.014

.012

.010

.008

.006

.004

.002

0

-.002

-.004

-.006

-.008

-.010

-.012

-.014

}/

/

I I I I I I I

-5.0 -4.0 -3.0-2.0 -1.0 0.0 1.0 2.0 3.0

/

I I

4.0 5.0

13,deg

.005

.004

.003

.002

.001

C / 0

-.001

-.002

-.00:

-.00,

-.005

"_ \i

I I I I I I

-5.0 -4.0 -3.0 -2.0 -1.0 0.0 1.0 2.0

[3, deg

(b) a =-5.0 °.

Figure 5. Continued.

• I I

3.0 4.0 5.0

25

.025

.02C

.01

.01

.00

Cy 13

-.005

-.0113

-.015

-.020

-.025

\

I I i i

-5.0 -4.0 -3.0-2,0 -1.0 0.0

_,deg

I i I I i

1.0 2.0 3.0 4.0 5.0

[3, deg

0 to -4.280

Measurement, linear fitParallel line through origin

Measurement uncertaintyfor all data sets as indicated

C n

.o141r

.012[

.010_

.0081"

.0061"

.oo41-

.002

0

-.002

-.004

-.006

-.008

-.010

-.012

-.0141

{

,.i

/

I I I I I I I

-5.0-4.0-3.0-2.0-1.0 0.0 1.0 2.0 3.0

_,deg

I I

4.0 5.0

.005i

.0041

.003!i

.oo2

.001

C / 0

-.001

-.002

-.003

-.004

-.005

B

%

%

m I i I

-5.0 -4.0 -3.0 -2.0 -1.0 0.0 1.0 2.0 3.0 4.0

J3,deg

(c) a =-2.5 °.

Figure 5. Continued.

I

50

26

Cy

.025 r A

o2o1:-o'__.o,C . {.o,o!.005 \

o \T_%

-oo5 F''_

-.010 / _ _ "_

-.015 \

-.020

-.025 , J , a , , I i ,-5.( -4.0-3.0-2.0-1.0 0.0 1.0 2.0 3.0 4.0 5.0

[_, deg

13, deg

o 0 to -4.302

o 0 to 4.35

0 to 4.364

A 0 to 4296

Measurement, linear fitParallel line through origin

Measurement uncertaintyfor all data sets as indicated

.014

.012

.010

.008

.006

.004

.002

Cn 0

-.002

-.O04

-.006

-.008

-.010

-.012

-.014

q

/,,

.f

I I I I

-5.0 -4.0 -3.0-2.0 -1.0 0.0

13, deg

I I I I I

1.0 2.0 3.0 4.0 5.0

.005

.004 I

.003

.002

.001

C I C

-.001

-.002

-.003 \

-,004

-.00 i , , i , ,-5.0-4.0-3.0-2.0-1.0 0.0 1.0 2.0

]3, deg

I I I

3.0 4.0 5.0

(d) (x = 0 °.

Figure 5. Continued.

2?

Cy

.025

.020

.0151

.01(

.00!

-.00

-.01{

-01. _

-.02G

-.02! i i i ,-5.0 -4.0 -3.0-2.0 -1.0 0.0

\

I_, deg

I I I I I

1.0 2.0 3.0 4.0 5.0

0rl

{

,B, deg

0 to 4.410

0 to -4.316

Measurement, linear fit

Parallel line through origin

Measurement uncertaintyfor all data sets as indicated

C n

.014

.012

.010

.008

.006

.004

.002

0

-.002

-.004

-.006

-.008

-.010

-.012

-.014

r

/

f J /

/

I i I I I I I I I

-5.0 -4.0 -3.0 -2.0 -1.0 0.0 1.0 2.0 3.0 4.0 5.0

_, deg

.OO5

.004

.003

.002

.001

C1 0

-.001

-.002

-.003!

-.004

-.005

=

I I I I

%

I I I I

-5.0-4.0-3.0-2.0-1.0 0.0 1.0 2.0 3.0 4.0

_,deg

(e) _=2.5 ° .

Figure 5. Continued.

I

5.0

28

Cy

.025

.020

.015

.010

.005

0

-.005

-.010

-.015

-.020

-.025

\.%

I I I I-5.0 -4.0 -3.0 -2.0 -1.0 0.0

I_, deg

\\

\

\

I I I I I1.0 2.0 3.0 4.0 5.0

0

[]

_, deg

0 to 4.380

0 to -4.252

Measurement, linear fitParallel line through origin

Measurement uncertaintyfor all data sets as indicated

Cn

.014

.012

.010

.008

.006

.004

.002,0

-.002

-.004

-.006

-.008

-.010

-.012

-.014 I I I I I I I I I

-5.0 -4.0 -30 -2.0 -1.0 0.0 1.0 2.0 3.0 4.0 5.0

#,deg

.OO5

.004

.003

.002

.001

C! 0

-.001

-.002

-.003

-.004

-.005 I I i I

\

-5.0 -4.0 -3,0 -2.0 -1.0 0.0 1.0

_,deg

(f) _ = 5.0°.

Figure 5. Continued.

\

I I I I

2.0 3.0 4.0 5.0

29

•0251.020.015!

.010

.005Cy 0!

-.005

-.010=

-.015

-.020

-.025

\

I I I I

t

\

I I I I i

-5.0 -4.0 -3.0 -2.0 -1.0 0.0 1.0 2.0 3.0 4.0 5.0

J3,deg

o

C]

t

9, deg

0 to 4.334

0 to -4.203

Measurement, linear fitParallel line through origin

Measurement uncertaintyfor all data sets as indicated

O n

.014

.012 _

.010

.008

.006

.004

.002

0

-.002

-.004

-.006

-.008

-.010

-.012 /

-.014 I I I I-5.0 -4.0 -3.0-2.0 -1.0 0.0

/ /

/

I I I I I

1.0 2.0 3.0 4.0 5.0

_,deg

.OO5

.004

.003

.002

.001

CI 0

-.001

-.002

-. 003

-.004

-.005

\ \

i i i i

k\

I I l l I3.0 4.0 5.0-5.0 -4.0 -3.0 -2.0 -1.0 0.0 1.0 2.0

_,deg

(g) a= 10.0 °.

Figure 5. Concluded.

3O

Cy

.025

.020

.015

.010

.005

0

-.005

Uncertainty

I I I 1

©[]

Re=, ft -1 13,deg

0.63 x 106 0 to -4

2.2 x 106 0 to -4

Measurement, linear fit

Parallel line through origin

Cn

.004

.002

0

-.002

-.004

-.006

-.008

-.010

-I-UncertaintyA 1

I I I J

C / .004

.002

0

.006 -

Uncertainty9"

-.002 I I J l J-5.0 -4.0 -3.0 -2.0 -1.0 0 1.0

13,deg

(a) _ =-5%

Figure 6. Variation of lateral and directional aerodynamic characteristics with angle of sideslip in air at 2rice = 6for two values of Re_c.

31

.O25

.020 -

,015 -

Cy .010 -(

.005

0

-.005

.O04

,002

0

-.O02

C n -.004 -

-.006 -

-.008

-.010

.006

C I .004

.002

0

-.002-5.0

-.Uncertainty

-I

1

I I I I

0[]

L.

_-Uncertainty_i t

I

-C _ _- _'1''_'_1"" ._._ /t.. •_'t"_'''1t

i

I I I I

Uncertainty

-

I I I 1

-4.0 -3.0 -2.0 -1.0

13,deg

(b) e = 0 °.

Figure 6. Continued.

Reoo, ft -1 13,deg

0.63 x 106 0 to -4

2.2 x 106 0 to -4

Measurement, linear fit

Parallel line through origin

I

1.0

32

.025

.020

.015

Cy .010

.005 I0

I

-.005

.004

.002 -

Ol

-.002

C n -.004 -

-.006 -

-.008 -

-.010

.006 -

C/ .004

.002

0

Uncertainty

0

0

-

I 1 I 1

- Uncertainty

Reoo, ft-1 13,deg

0.63 x 106 0 to -4

2.2 × 106 0 to -4

Measurement, linear fit

Parallel line through origin

-(

t I I I

Uncertainty

-.002 J J-5.0 -4.0 -3.0

I I

-2.0 -1.0

13,deg

(c) _ = s°.

Figure 6. Concluded.

I

0 1.0

33

Cy

"°30C_

-.0100 t •

0

I I 1 I I

0

[]

13,deg

Oto 5

0 to -5

Measurement, linear fit

Parallel line through origin

Measurement uncertaintyfor all data sets as indicated

O n

.020

.010

0

-.010 [

-.020

- 0

I I I I I I I I I

C!

oo6 .004 I---...

.-..-._

.002 - " -., ,_

--.,0

-.002 -

-.004 -

-.006 I I t [-5.0 -4.0 -3.0 -2.0 -1.0

I I I I I

0 1.0 2.0 3.0 4.0 5.0

13,deg

(a) a =-10 °.

Figure 7. Variation of lateral and directional aerodynamic characteristics with angle of sideslip in CI:4 at

Moo = 6 and Reoo = 0.46 x 106/ft.

34

Cy

,030[_

-.010 f-.020

-.030 I I I I

O

I I I I I

.O20

.010

Cn 0

-.01C_

-.020 I I I I I I I I I

CI

.006I

.004

.0O2

0

-.0O2

-.004 -

-.006-5.0

I I l I

-4.0 -3.0 -2.0 -1.0 0

13,deg

I I I I I

1.0 2.0 3.0 4.0 5.0

(b) c_=-5 °,

Figure 7. Continued.

0

[]

tS,deg

Oto 5

0 to -5

Measurement, linear fit

Parallel line through origin

Measurement uncertaintyfor all data sets as indicated

35

Cy

Ca

CI

.030

.020 [

.010

0

-.010

-.020

-.030 I ! I I

0

I, I I 1 I

.020

.010

0

-.010_

-.020

m

_ _ _ _ _ _ _ S _

.006 -

E.004

.002

0

-.002

-.004

-.006-5.0

,I I 1 I

-4.0 -3.0 -2.0 -1.0I I I [ I

1.0 2.0 3.0 4.0 5.0

(c) ,_ = 0°.

Figure 7. Continued.

0

[]

13,deg

Oto 5

0 to -5

Measurement, linear fit

Parallel line through origin

Measurement uncertaintyfor all data sets as indicated

36

.030

.020[

.010

Cy 0

-.010 -

-.020

-.030

.020 -

.010

Cn 0

-.010 [

-.020

.006

.004 [

.002

C / 0

-.0O2

-.004 -

-.006-5.0

I I I I

0

I I I I I

I I I I I I I I I

B

I I I I , I I I I I

-4.0 -3.0 -2,0 -1.0 0 1.0 2.0 3.0 4.0 5.0

J], deg

(d) r_ = 5°.

Figure 7. Continued.

©

[]

13,deg

Oto 5

0 to -5

Measurement, linear fit

Parallel line through origin

Measurement uncertaintyfor all data sets as indicated

37

Cy

On

C/

.030

.020[

.010

0

-.010

-.020

-.030I I I I

0

0

[]

.020

.010

0

..010 [

-.020

I I I I I

oI I I I I I I I

I I I I I

1.0 2.0 3.0 4.0 5.0

(e) c_ = 10 °.

Figure 7. Concluded.

.006[_

.oo4I'--

.002 " - ..

0

-.002 -

-.004 -

-,006 I I I J-5.0 -4.0 -3.0 -2.0 -1.0

13,deg

Oto 5

0 to -5

Measurement, linear fit

Parallel line through origin

Measurement uncertaintyfor all data sets as indicated

38

-.006 -

_oo,___Cyl3,

deg-1 -.002

Moo Test Gas

O 6 CF 4

[] 6 Air

/_ 10 Air

_ A _

I I I I

_A_ 4_

I I I I I

.oo_Cnp' .00

deg -1 02__ A t_ A _ &

I I I I I I I I I

-.0010 -

-.ooo_

C/p,-.0006

deg-l_.o004 _

-.0002 -

0-10.0

A

A _ "0

I I I I I I I I I

-8.0 -6.0 -4.0 -2.0 0 2.0 4.0 6.0 8.0 10.0

_, deg

Figure 8. Lateral and directional stability characteristics in air and CF4. Note sign change in top and bottomfigures.

*U.S. GOVERNMENT PRINTING OFFICE: 1993-728-150/60053

39

Form Approved

REPORT DOCUMENTATION PAGE OMB No. 070#-0188

Public reporting burden for this collection of information is estimated to average I hour per response, including the time for reviewing instructions, searching existing data sources.gathering and maintaining the data needed, and completing and reviewing the collection of information Send comments regarding this burden estimate or any other aspect of thiscollection of reformation, including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 JeffersonDavis Highway. Suite 1204, Arlington, VA 22202-4302, and to the Office of Management and Budget. Paperwork Reduction Project (0704-0188), Washington, DC 20503

1. AGENCY USE ONLY(Leaveb/ank) 2. REPORT DATE 3. REPORT TYPE AND DATES COVERED

June 1993 Technical Memorandum

4. TITLE AND SUBTITLE 5. FUNDING NUMBERS

Hypersonic Lateral and Directional Stability Characteristics of

Aeroassist Flight Experinlent Configuration in Air and CF4 WU 506-40-41-01

6. AUTHOR(S)

John R. Micol and William L. Wells

7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES)

NASA Langley Research Center

Hampton, VA 23681-0001

9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES)

National Aeronautics and Space Administration

Vv_hington, DC 20546-0001

8. PERFORMING ORGANIZATION

REPORT NUMBER

L-17154

10, SPONSORING/MONITORING

AGENCY REPORT NUMBER

NASA TM-4435

11. SUPPLEMENTARY NOTES

12a. DISTRIBUTION/AVAILABILITY STATEMENT

Unclassified Uidimited

Subject Category 02

12b. DISTRIBUTION CODE

13. ABSTRACT (Maximum 200 words)

Hypersonic lateral and directional stability characteristics measured on a 60 ° half-angle elliptical cone, which

was raked at an angle of 73 ° from the cone centerline and with an ellipsoid nose (elliptieity equal to 2.0 in thesymmetry plane), are presented for angles of attack from - 10 ° to 10 °. The high normal-shock density ratio of

a real gas was simulated by tests at a Mach number of 6 in air and CF4 (density ratio equal to 5.25 and 12.0,

respectively). Tests were conducted in air at Maeh 6 and 10 and in CF 4 at Mach 6 to examine the effects of

Math number, Reynolds number, and normal-shock density ratio. Changes in Maeh number from 6 to 10 in

air or in Reynolds number by a factor of 4 at Mach 6 had a negligible effect on lateral and directional stabilitycharacteristics. Variations in nornml-shock density ratio had a mea,surable effect on lateral and directional

aerodynanlic coefficients, but no significant effect on lateral and directional stability characteristics. Tests in

air and CF4 indicated that the configuration was laterally and directionally stable through the test range olangle of attack.

i

14. SUBJECT TERMS

AFE; Aerodynamics; Real-gas simulation; Hypersonic; Blunt body

17. SECURITY CLASSIFICATION

OF REPORT

Unclassified

_ISN 7540-01-280-5500

18. SECURITY CLASSIFICATION

OF THIS PAGE

Unclassified

19. SECURITY CLASSIFICATION

OF ABSTRACT

15. NUMBER OF PAGES

40

16. PRICE CODE

A03

20. LIMITATION

OF ABSTRACT

iStandard Form 298(Rev, 2-89)Prescribed by ANSI Std Z39 18298- I02

NASA-Langley, 1993