Investigation into the use of freon 12 as a working medium in a high-speed wind-tunnel

CoA Note No.72

Kariaali

LrJ -«Jj^ _{• iu«iu}

THE COLLEGE OF AERONAUTICS

CRANFIELD

INVESTIGATION INTO THE USE OF FREON 12 AS A

WORKING MEDIUM IN A HIGH-SPEED WIND-TUNNEL

by

TECHNISCHE HOGESCHOOL VL!EGTUIG!iOL;WKU;^-D(E KênaaUtraet 10 - DcLFT NOTE NO, 72 NOVI'SffiER 1 9 5 7 .

T H E C O L L E G E O P A E R O N A U T I C S

C R A N F I E L D

"INVESTIQiVriON INTO THE USE OP PEEON 1 2 AS A WOEKING lIEDIUiVI I N A KIGH-SEEED WIND-TUNNEL".

t y

2

-SUMMARY

The Preon 12 tunnel project at the College of Aeronautics was started to giin experience in the design of such tunnels and

subsequently to investigate their use for aerodynamic measurements and the correlation of these with similar measurements in air.

Reports dealing with the use of heavy gases are listed and

the factors affecting the choice of any particular gas are discussed. The most suitable gases are the flutro-carbons and amongst them

C P n-perfluoirobutane shows most promise. It can be shown that : 4 10

^ Power ^ Size ' B . P t .

r e l a t i v e t o a i r a t same R ^

Mol.wt. *T

P-12 (CCOigPj 121 3«5 33 -30

C P 238 1.8 23 - 2

4 10

The application of transonic similarity laws to 2-dimensional flow over wings of infinite aspect ratio to convert pressure

measurements obtained in heavy gases to equivalent air values is suggested.

Pressure and wake drag measurement on a IC^S thick 2in. chord aerofoil RAG.104 over a Mach number range of 0,4 to 0.85, at a Reynolds number of 1 . 7 x 1 0 ^ at M = 0.85 an<3. incidences of 0 and 4 , have been corrected to equivalent air values by means of transonic and area similarity rioles, and have been compared with measurements in air on similar aerofoil sections at the National Physical Laboratory. The validity of these rules could not be

completely proved, though the small differences v/ere most likely due to the tests not being performed on identical models and in identical tunnels.

The application of transonic similarity to the conversion of wave drag in Preon-12 to equivalent air results gave very good agreement with air measurements. The magnitude of the effects of relaxation, times in Preon-12 in the flow about small diameter total head-tubes is shown to be negligible at atmosphex-ic pressure and at a stagnation temperature of 300 K. The increase of drag about a 2" chord model at simalar stagnation conditions is about 0,3% and since smaller models are unlikely to be used this effect is quite negligible.

SUMMARY Cont'd.

-%

The advantages to be gained in flutter testing by using a heavy gas are compared for both the compressible and the

incompressible cases. In general the use .of a heavy vapour tunnel results in the reduction of flutter velocity, frequency, model stresses and also in the pressure ratio ±n comparison to a compressed air tixnnel,

LIST OP CONTENTS.

Summary, 2 List of contents, . 4

Notation, 7 Introduction, "li

Choice of working substance• 12 The effect of Y i n Subsonic and Transonic Plov/, 14

5.1. The effect of the ratio of specific heats on

isentropic flow relations. 14 5.2. Cocrpressibility effects on pressiore and lift. lé

5.3. Transonic similarity laws. 1? 5.4. Application of transonic similarity laws to

2-D.flow. 18 5.5. Application of transnoic similarity laws to finite

aspect ratio wings. 19 5.6. Review of experimental work by Von Doenhoff and

Braslow (ll). 21 The College of Aeronautics Preon-12 Y^jund-tunnel 23

6.1. General design considerations. 23

6.2. Purification. 25

G,3. The 5.25 i n . x 7.5 i n . annular r e t u r n Preon-12

tunnel a t the College of Aeronautics. 25

6 . 4 . Preon C i r c u i t l a y o u t . 27

6 . 5 . C i r c u l a t i o n . 2? 6 . 6 . D e t a i l e d d e s c r i p t i o n of components. .28

CEilibration of Preon-12 Tunnel 30

7 . 1 . I n t r o d u c t i o n . 30 7 . 2 . Tunnel and apparatus. 31

7.3- Accuracy. 31 7 . 4 . Experimental measiirements, 34

7 . 5 . Resiilts from t o t a l head measvirements. 34 7 . 6 . Transverse v a r i a t i o n of s t a t i c p r e s s u r e i n t h e

5

-L i s t of Contents Cont^d,

Page No.

7 . 7 . V a r i a t i o n of s t a t i c p r e s s u r e along the working

s e c t i o n . 36

7 . 8 . Working section Mach No, 37

7.9. P u r i t y of Preon. 37

8. Two-Dimensional Aerofoil T e s t s i n the Preon-12 Tunnel. 38

8 . 1 . Model. 38

8.2. Model mounting and pressure meastirements. . 38

8.3. Pitot traverse gear.

39

8.4. Calculation of results. - 39

8.5. Pressure distributions in Air and Preon-12. a = 0 41

8.6. Drag measiirement a = 0 42

8.7. Pressure distributions in Air and Preon-12, a = 4 44

8.8. Drag measurements a = 4 . 44

9. Conclusions, 45

Acknowledgements, 46

References, 47

Appendices, 51

1. Derivation of Transonic Similarity rules,

2. Application of Transonic and area similarity rules to

two-dimensional flow in Hreon-1

2

and air. 56

3. . Application of Transonic Similarity laws to finite aspect

ratio wings in Preon-12 and Air, 58

4. Application of Area Similarity laws to convert the pressure

distribution on wings in Preon-12 to Air resiilts , 59

5. Power Economy, reduction in size and model stresses in

heavy gas tunnels. 60

6. Mach No. limit for condensation in heavy gases, 62

7. Reynolds No. in gas mixture^ 63

8. Reynolds No. in Bi:'e.on-12 and Preon-12. Air. Mixtures.

66

9. Variation of skin friction between tests in air and Preon-1^

6S

6

-P a g e N». JJXSX or uonxemjs u o m i ' a ,

11, Determination of gas mixture proportion by means of a

sound measToring gauge. 78 12, Coniparison of Flutter model testing in couipressed

NOTATION a = speed of sound

A = area of stream tube A,B,C = parameters

iE = aspect ratio

b = span of aerofoil; tunnel b r e a d t h o = chord of aerofoil

C ' = force and mom.ent coefficient ratio Cj, = drag coefficient

C„ = skin friction coefficient

C. =j xd.bration heat capacity

Cj^ = lift coefficient

%i A moment coefficient

C„ = normal force coefficient

C = pressure coefficient (^ ~ S° ) ; specific heat at

P ip TS'^

constant pressure 2 oo eo

C = heat capacity of active degrees of freedom C = specific heat at constant volume

d = impact tube diameter

D ( K , H ) = drag function

f = fiaiction; frequency P = integrating factor

g = wing thickness distribution

H = similarity parameter jR (y+1 )M^ T ; total head; Entha:lpy

8

-K = similarity parameter -:: 2

_Vi;

relaxation

parameter"-1 = model reference length

L ( H , K )

= lift function

m = molecular weight

M = Mach nos.

M ( H , K )

-

moment function

n

-

no, of atoms in molecule, viscosity scaling factor

p = pressure at any of the wall tappings 1, 2, 3, 4.

p = pressure

p' = static pressure at traversing gauge

p. = atmospheric pressure

p. = pressure in settling chamber

P = power

P ( H , K ;

—, ^-, r-) = pressure function

P ( K ; X', Z') = pressure funetion

q = dynamic pressure

q. = local velocity'- aroiond aerofoil

Q,J,G,E = f\anGtions

r = density or pressure ratio

R or R = Reynolds numbers

R = universal gas constant

P U X 00 00 _ —p—

R

X

S,S.,S' = entropy

= thickness of aerofoil

e = block age correction

e = solid blockage correctior

-9-t

c ~ ^ = aerofoil thickness chord ratio

T = temperature

u, V, w = velocity components in x, y, z directions

U = free stream velocity

X = mole-fraction of air in gas mixture

(x, y, z) = car-tesian coordinates with x in mainstream

direction and j in plane of wing perpendicular

to mainstream direction

Z = ordinate of aerofoil

a

= incidence

C

y = ratio of specific heat

r^

V

AC_^ = increment in C_^

e = stiffness

e = solid blockage correction factor

s

°

e

~

wake blockage correction factor

Tj =

power factor; slope

\ = linear scaling factor

^

= viscosity

p = density

c = Prandtl number; v/ing density

r = relaxation tirr.e; aerofoil thickness chord ratio

T

=

-wall

shear stress

w

0 = perturbation velocity potential; function of

Mach number

$ = v e l o c i t y p o t e n t i a l

0) = exponent i n viscositjT^teraperature r e l a t i o n ;

wave number

13 / l - " " Mc

1 C -SUB SCRIPTS A i r c o n r e c t e d ; c o m p r e s s i b l e v a l u e c r i t i c a l c o n d i t i o n , Mach n o , = 1 P r e o n - 12 incompressible value stagnation

settling chamber pressure uncorrected

working section

differentiation with respect to x direction; mixtiare properties

differentiation with respect to y direction differentiation with respect to z direction free stream condition

^ 1 1

-3. Introduction

Heavy vapours have been sugpested as possible media for wind-tunnels mainly because of the large power savings that can be

obtained in comparison to air (appendix 5 ) .

Other means of obtaining power reduction are by

(a) Pressurisation; this has the disadvantage of increasing m.odel-st i-e s 3 e s

(b) Re.frigerat ion; this is generally difficult within a structure like a wind-tunne.l

Probably the earli^t published report in vrfiich attention is drawn to the advantages of using a medium other than air in a Vidnd-tunnel is by Margoulis, 1920, (ref,l). He considers carbon dioxide and shows that in the cases in vdiich (a) Reynolds Number,

(b) Mach Numbers(c) both Reynolds number and Mach number are simulated, considerable saving of power is achieved. In a report published at about the same time, M.Mxmk, 1921 (ref.2), pointed out that a reduction in size and power is avhieved by using compressed air, and also mentions the possibility of using carbon dioxide for the same purpose but dismisses it as impractical because of the additional complications involved.

Rubinski 1939 (ref.5) considers a wide range of heavy gases for use in supersonic tijnnels and concludes that among other, the Preons, Carbon Tetracl-iloride and Sulphur-Hexaflucride merit special attention. He also considers the design of a wind-tunnel using CCl to run at a Mach number of 1,3. Smelt, 1945 (ref,4), discusses the relative advantages that result from the use of heavy gases, compressed air and also through refrigeration and lists a large number of possible compounds together with the reduction in power and tionnel size obtainable in each case. The use of heavy vapours were the subject of papers by McKinnon Wood,

1945, (ref,5) who also considers the use of sulphur hexafluoride for powered model tests, and by Bottle, 1948, (ref,6).

The effect of y, the ratio of specific heats, v/hich for poly atomic gases differs from the value 1,4 for air and other diatonic gases, on supersonic flow was considered by Relf (ref,7), 1952.

The thennodynamic properties and the time lag of vibrational heat-capacity of Preon-12 have been examined by Huber, 1946 (ref.8) to determine its suitability for aerodynamic testing,

The use of a mixture of a monatomic gas, Y = 1,67, with a heavy polyatomic gas with a low value of y, such that the resultant y is equal to 1,4 was suggested by Relf 1952 (ref,9) and Chapman, 1954,

(refolO). The latter report contains a very extensive list of suitable heavy gases.

Von Doenhoff and Braslov/, 1953, (ref,1l) used Preon-12 in a wind-tunnel, and have evolved a method of correcting the measured force coefficients for the different value of Y to equivalent air values up to the transonic range

In addition gases other than air h.ave been used for ballistic reseEirch by Buel, 1948, (ref,12) and Donaldson and Sabol

1951, (ref,13) mainly to achieve very high Mach numbers,

Theodorsen and Regier, 1944 (ref,l4) used Preon-12 and Preon-IO to measure the drag of revolving discs, cylinders and streamline rods at high Revnolds and Mach numbers; and Kantrowitz 1953, (ref,15} used Preon 12 to test supersonic compressors,

Gases other than air have been used by Duff 1951 (ref,l6) jai a shook-tube and Helium was employed by Bogdonoff and Hammitt, '1954 (ref.l7) in a hypersonic tunnel as it can be expanded to higii Mach numbers vd.thoufc liquification.

4» Choice of working substance

The main requirements for an advantageous mediimi is a. low speed of sound (see aj^pendix 5) together with a low boiling point. It should also be stable and inert chemically, cheap and easily available.

For perfect gases at the sane temperatures

V"

m

where a = speed of sound m = molecular weight

and hence the interest lies mainly with heavy vapours, but unfortvinately the boiling point of vapours generally increases with increasing

molecular weight.

It appears that most of the suitable gases are compounds of flucjrine, for a given molecular weight. They have the lowest boiling pojnts a.nd are mostly extremely stable. /imong these the Preons

requ:i.re special mention since they are readily available in large quantities and they have already been used experimentally, see ref.

(11 and 12). There are however, many compotmds which are more stable and efficient than the Preons and yet have a reasonably low boiling point, e,g, C. P - one of the series of fluoro-carbons. This can be

-13-seén from the following tables

%

power

%

Size B,Pt,

relative to air at same R,M.

Mol,wt,.

^

Preon-12 (CClv, P2) 121

b*5 33

-30

C4 P^o 238 1.8 23 - 2

Many gases of this tjrpe have been developed recently and it is believed

that their usefulness will largely depend on their cost and availability.

In addition to the above considerations several other factors

must be taken into account. These are thermal and caloric imperfections,

relaxation time effects and the value y . Generally y = 2n+^

2nfi

where n is the number of atoms in the molecule and hence y for most

of the heavy vapours with complex molecules is smaller than y ~ 1.4,

the value for air and most diatomic gases.

Deviations from perfect gas laws under normal v/orking

conditions are found to be usually small and caloric imperfections

become considerable only at high temperatures.

Suggestions were made by Relf (ref,9) and independently by

Chapman (ref,10) that it would be possible to obtain a mixture of a

suitable monatomic gas, e,g. either Argon, Krypton or Xenon, y = 5

3

and a heavy polyatomic gas with low y such that the value of y for

the mixture \vas 1,4 and yet obtain considerable power economy, though

not as great as using just a heavy gas by itself. Chapman found that

on the whole the values were of the follov/ing order to obtain the same

Mach and Rejmolds nimiber

Argon and heavy vapcior

Mixtures Krypton "

of Xenon "

II II

CCl^ P^

C4 P10 % Pov;er

relative to

30

20

10

b.b

1.8

air

% Size

relative to air

70

60

50

33

23

but of the monatomic gases Argon i s the only one a v a i l a b l e i n l a r g e

q u a n t i t i e s . Chapman also observes t h a t i t i s p o s s i b l e t o obtain

mixtures which under low temperature wind tunnel conditions a r e

14

-that is dimensionless macroscopic parameters, for instance those involving relaxation time phenomena and temperature variation of specific heat, arc equal, whereas air under low temperature wind tunnel condition v/ould behave dissimilarly.

Some of the disadvantages in using a heavy gas are a different value of y , structural complication because of the necessity of isolating the medium from air and the necessity of additional equipment for purification purposes. The requirements for such a tiinnel are somewliat similar to those of a pressurised tunnel, but it is more important to avoid leaks. Additional instrumentation is required to measure the purity of the medi-um.

5. The effect of y in subsond.c and transonic flow.

5.1, The effect of ratio of specific heat on the isentropic flow relations.

In general y the ratio of specific heats can be written as 2n + 3

y = o^^—7 5.1.1

' 2n + 1

5 •vrfiere n is the number of atoms in a molecule, y thus varies from -r:

for monatomic gases (e.g. Helium, i'lrgon) to a limiting value approaching unity for heavy gases with coinplex molecules.

The isentropic flow relations (with suffixes st and or denoting stagnation and critical values respectively)

(a) Pressure ratio ' 5.1.2

(b)

- y

Lt ^ ^ - = e 2

y . 1 ''^^

Density r a t i o

S u

y-' ^st "

- > - i

5.1.3

-15-(c) Area ratio A_ i" -•''br M

f2 + (y~i) M^

y + 1

\ 2 ( y - i )

) 5.1.4, V-I A^^

(d) Ratio dynamic pressure to stagnation pressure

y

q

Pst

Lt

y - i

=

1

Pst

2 =

M^

2

0

0 y -+ 2 -M^ 2 5.1.5.

(e) Temperature ratio

St \ '^ / 5.1.6, Lt '.p_

y-^ '^st ^

have been plotted in figs. 1, 2, 3, 4, 5, for a Mach number range 0-3 for values of y equal to 5/-z> 7/r, 9/o and the limiting value of y=1.

An examination of these graphs shows that the variation with y of the area ratio up to M = 1,4 is small. The variation

of •^— , when Y changes from Y = 7/^ to 9/^ is about 1C^, but Pst

when y ^ of 1+0^0.

when y varies from y = 5/7; to y = 1,0 the variation is of the order

T

The temperature ratio -— and t.he ratio of dynamic pressure st

to stagnation pressure are much more sensitive to a change in Y , 7

The value of Y = ~ corresponds to the value for air and 9

^^16-Pigo5 shov/s c l e a r l y t h e much la:!rg-3r a r e a I ' a t i o r e q u i r e d f o r supersonn.G dwcts i f a p-p-s «it.h a low va.lue of y i s u s e d a s . oomT)ared w i t h t i i o s s urijr.g a mona.tomic gas w i t h Y = 5/'-/

in ref.7

The effect of Y on some -supersonic phenomena is discussed

5»'<•-•> Co.r.ipra[.'i".>il)ility effects on prensui^e and lift.,

'Vhe ra-l.io of specific heats does not appear in the P.randtl-Glauert rale (.ref ,19) derived from the .linearised coitgpressible flow equations, Th:!.a rale is l:.mdted to small perturbations and hence very thin profiles at .••xTiiül angles of attack at speeds v/ell below the speed of nouiid, Hance under these conditions t h e effect of different y can b e expected to be negligible.

lFvjr*.heT evidence f o r the small influence of y on compressible flow is inQ.icated b y the similarity of results übt8.ined in a hjrdraulic analogy taxix, where the effective y = 2, and compressible air flow (ref,20) and. also from the widely used Karman-Tsien re.lations (ref,22, 23) which can b e derived from the compressible flow equations b y putting y = - 1 .

The Prandtl-Glauert rule Vvas modified b y Laitone (ref,24) wiio obta?jned

% c

1

vi'j f r> s / y — 1 2 \ - . .(i~M" + M (l^. -—- M ) 0 .

Aj u o o o ' ' ' 2 o o ' pj_

where C , G . are the compressible and incompressible values

pc p i ^ -^

of the pressiore coefficient respectively. T h e lüitone r^.^Le gives better experimental agreement and is applicable at somewhat hig.aer

subsox-iic !Vi£!.ch numbers and press-are coefficients tb-an the Prandtl-(xla,uert rule. This rule is note-wcrthy for tb.a fact that it includes

the factor V.

Kaplan (ref,25) obtained a n e:cpression relating t h e compressible and incompressible coefficients on a n elliptical, cylinder and exbended it to an e,rbit.rary syimetricaJ. profile,

C~" = ^ "^ "ïTf f/^'Cp-'i) +-^ (''•"••)

(iT-^y^

5.2,2

-17-•I

where P - —• and T ~ thiclaiess ratio.

This relation is plotted in fig.7 for T =, 0,05, 0.10, 0.15, 0.20 and Y = 1,4.

The lift correction, for the same values of rand M , ' 00 ' A C ( C , ) i C

^c ^ "'c/A - \ ^ c ' ' P

\

^L

^ c y p

1 ^ ( y ^ > - Y ^ ) C ^ H f

''^J^ p ( / ^ - i ) + l (yp+i)(/?^-i)n 5.2.3.

was calcxilated in t h e case of Preon and a i r ( Y, = 1,4 an<3.^-ri =

A r 1.125) and is presented in fig,6 for the cases of 7- = 0.20, 0,15,

0.10, 0.05. It is seen that the effect of the difference in

specific heats is quite small, eog. at M^ = 0.8, A C_ varies from

^ to ^o5% for thickness ratios of 3% to 2 0 ^

\

5.3. Transonic sim:'.larity laws,

The transonic similarity laws indicate useful generalisation concerning the floiï behaviour at transonic speeds. Their applicability to relating flows in media with different values of the raiio of specific heats, y, will be examined.

The transonic similarity laws are derived on the assumption of small perturbation velocities and velocity gradients about either the critical s^nic velocity (ref.24, 25) or under the less restrictive conditions of small perturbations about the free-sbream velocity (ref,26). A derivation of the transonic rules for two and three dimensional wings

is given in appendix . I.

The transonic similarity laws are based on potential flow thus excluding from consideration regions having vorticity. They can be applied successfully to subsonic flows with shock waves provj.ded these are weak so that the entropy change across them is small and provided they are only slightly curved so that the variation of entropy from one streamline to another is small and hence the vorticity induced by the shocks can be neglected.

1 8

-5.4» Application of t r a n s o n i c s i m i l a r i t y r u l e s t o tv/o-dj-mensional flov/ v/ith d i f f e r e n t value of the r a t i o of s p e c i f i c h e a t s ,

Per tv/o-dimensional wings t h e p r o f i l e o r d i n a t e s a r e given by

Z ^ / x

o "= ^ ^ \c J 5.4.1.

an.d the relevant similarity parameter is K = 1 - M ~

£(yvOlC^

y^Z

5.4.2

The transonic similarity rules for pressure and force coefficients (eqn, 21, 22, 23, 24, 25 appendix I.) can then be ^Tx-itten

C = _ l f : j _ w , P (K; x', z')

r(y^^)M: 1 ^

1 / •ïshere x' - — and z' = (1-M^ ) — o ^ 00 ' o

^''°L= - - 1 - ^ _ 1 A L(K)

T 2/

°M =

'

2_ , M (K) . , ,

p (y+1) Mi J / 3

r 5/.

°D = ^™,/ D(K)

-(y+1) Mr-)V3

In the case of x-vings of identical profiles and at the saro.e incidence in tvra flews with different value of the ratio of specific heats y.,, y„, (v/here suffixes i, 2, are used to designate tfie tvro flows) but at the same value of the similarity parameter Kj, vre have from eqni5<.4,2

(l - Ivf ^ ) "• " ^C 2

and at points, given by the same dimensionless parameters x' = x' , 2' = z' .

1 9 -¥ e have from e q u a t i o n s 5 , 4 . 3 « P. p K,x , 3 2 1 1 (y +1) ï f

niA

(y +1) J^ 1 oo< - > and C' = C.

= V;

/°M

I D 1 5 . 4 . 5 . L ' K ^M K S K 2 2 (y +1) M^ ₀₀₂

iA

_ ( y , + i ) M % ^ ^ Thus t h e r a t i o of t h e f o r c e c o e f f i c i e n t s f o r s i m i l a r t w o - d i m e n s i o n a l flow about i d e n t i c a l p r o f i l e s accord:ijig t o t h e t r a n s o n i c s i m i l a r i t y law i s g i v e n b y e q u a t i o n s 5 . 4 . 5 . and 5 . 4 . 6 . 5 . 5 . A p p l i c a t i o n of t r a n s o n i c s i m i l a r i t y .laws t o f i n i t e a s p e c t r a t i o w i n g s , I n t h e c a s e of t h r e e - d i m e n s i o n a l f l o w , tv/o s i m i l a r i t y param.eters H,K, o c c u r i n t h e t r a n s o n i c s i m i l a r i t y r u l e s g i v e n b y e q u a t i o n s 2 1 , 2 3 , 24, 25 a p p e n d i x 1. These s i m i l a r i t y r u l e s a r e where 2 / ,

[(YH-I)

M^Jj^6

[(y+1) Mi]^/3

2 / . M

°D =

[(y+1) My ^ 6

5 / . K = H

P ( H , K . j f , f , f )

L (H,K) M ( H , K ) D ( H , K )

[(y-^i) M^] ^ 3

1 - M ^

[(Y+I)M;^J^'^3

= ^ 1^ (y+l) 1^ rl'^^3

5 . 5 . 1 . 5 . 5 . 2 . 5 . 5 . 3 . 5 . 5 . 4 . 5.5.5. 5 . 5 . 6 ,

-20-To obtain similar flows in fluids of different values of Y, the ratios

of the specific heats, both H and K must be kept constant, and two

special cases

arise:-(a)

M = &

^ ' 1 2 M

= M 5.5.7

oo^ 0O2

and (y +l)r = (y 4.1 )r

1 1 2 2

These satisfy the requirement of H and K being constant,

The similarity rules then give

Jv.) (%.)

V /M„ , iR, (Y+1) r ; I , J , I = ^ C % ^ , ^ , (y+l)r

V «=' ' ^ ' ' c ' b ' b y

'2 '2

^T/[ V

y / M „ , iR, (y+1) r r y +1 5.5.8

2 2 1 . .

and / C„ \ / V o

ö^) =(L)=(v^T

\ ^

M^, JR, (y+l)r \ r^ / V y^+1 / 5.5.9

Thus similarity exists for flovifs v/ith different y about

three-dimensional wings of the same aspect ratio and at the same Mach

number, but for related profiles of different thickness parameter,

(b)

r^

=

r^

Per H and K to be constant

1 -

MS>

J». ;L2/, = const,

[(Y+I)M: ] 5

5.5.10

r •

"1-and JE j 1 - J^ J ^ = const.

or

m

Ry+l) M ^ l ""^3 = const. 5.5.11

In this case to preserve similarity, the thickness parameter

remains the same, but both the corresponding Mach numbers and

asTject ratios are different.

-21-/

this case.

The similarity laws give for the force coefficients in

P,,

.-

K

"'^^'^' c » b * b 2E Z 1 H,K, r

M

V,

\ . _y.^ H,K, r

D

y .

c.

_'D

')k

1 - M^ - M. _{CO.J __} 5.5.12

5.6. Re^/iew of experimental v/ork by von Doenlioff and BrasloT/ (ref.1l) Von Doenhoff and Braslow (ref.1l) used Preon-12 (Y= 1.13)

as a testing medium in the Langley Low T^jrbulence Tujmel. Measurements of the pressure distributie as v/ere made in the Mach nimiber range of

0.4 - 1.2 and at Reynolds numbers (at M:il) of about 9,5x10*^ per foot chord on ST/ept and unsvvept wings of aspect ratios of 4,0 to 9.0. The measurements were then compared with those made in air and methods of converting Preon data to air data T/ere developed based on the sti-eam-line similarity concept using the critical area ratio A

cr where A _ = Area at sonic speed

c r

A

any other speed

Stream-line s i m i l a r i t y i s shov^n t o be consistent with transonic

s i m i l a r i t y lav/s i f the thickness r a t i o of the body in Preon i s

_i_

+ Yr

1 + y..

or 1.050 as great as in air. In practice this difference

in thickness was considered to be negligible and bodies of the same thickness were tested in air and in Preon, The results showed that no basic differences in flo\7 phenomena exist at subsonic speeds and

on the whole the corrections obtained on the assumption of area similarity seem to deal adequately vvith the small differences that do exist betv/een Preon and air flow. An. uncertainty exists when streamline similarity is applied to swept wings. Por wings of small aspect ratio it is sufficient to apply it to the flov/ in the free-stream direction v/hereas for infinite M it should be applied to the flow normal to the leading edge. Pressure distances meas^ared in Preon-12 a<.t various free-stream Mach numbers for a nxjmber of aerofoils were converted,by assuming streamline similarity, to equivalent pressure

— ?p™

distributions in air. The conversion factors for norTra.1 forces and pitching and hinge moment coefficiente ¥/ere plotted against the

freQ-stream Ma,ch nu:r!ber,^\Y-on Doenhoff conciud.es that the overall conversion ^ coefficient can be considered to depend only on the free-stream llach

number and to be independent of the raagnitiado of the pressure distri-bution. irhilst the resultii quoted for conversion factors of the normal force coefficients seem to justify such a simplication, in the case of pitching moment and hin.ge moment conversion coefficients this seems doubtful as the scatter of the results is appreciable.

The conversion factors based on streaanline similarity are different for the induced part of the drag and the wake drag and both depend main2.y on the free-stream Mach number . To convert

subsonic Preon drag rcsiilts an estimate of the ijiduced drag part is made by ca.lGulatian and the rest is assumed to be walce di-ag and the separate conversion factors applied. At supt^rsonic speeds the drag is resolved into that due to skin friction, which is asstmied to be independer.t of angle of attack and that due to presyure wliich for the normal tliin supersonic aerofoils with low leading edge suction can be assum^cd to be ~ C„ «

~ N

where C„ = norm.al force coefficient

o. - angle of attack.

Hence to convert drag at supersonic speeds the drag is

measured vvhen C„ = 0 and the wake drag conversion factor is applied to this pai't of the drag and at a:iy other incidence the normal-force conversion coei'ficient is applied to the remainder of the drag.

Si.nce the spamvise normal-force coefficients a.re all multi.plied. ''oy the sam.o factor, the normal force conversion factor also applieiö to the rolling moment.

To convert side-force and yawing moments the nature of the forces involved has to be decided. Por instance in the case of the yaw:j.ig moment due to non-Tnifonn skin-friction the Vv-ake drag conversion coefficient is applied btjt for the yawing moment caused by the d.eflection

of the rudder t.he application of the normal force conversion coefficient is necessary.

*" The gi'eatest d.ifference between Preen and air results was considered to be most liiely near the trailing edge. In order to obtain a critical check on the method of convertrjig freon data

« generally and moments in particular in this region tiie hinge moments on an elevator on a fnxLl-span unswept tail p].ane were measvxed in Preon-'! 2 tmd in air. The converted Preon data agreed well with the air data to at least the a.ccuracy of the air data itself.

-23-Critical discussion of von Doenhoff's restüts and conclusions dra7,n fi.-'om it.

Von Doenhoff's results indicate that no basic differences exist between air (y= 1.4) and Preon-12 (y= 1,13) flows at subsonic and low supersonic Mach numbers. This is so even in the presence of strong shocks and vri.th a large area of the v/ing stalled. Stream line similarity v/as successfully employed to transform pressure

distributions and overall forces, but the following should be noted:-(l) ¥£Ü:e-drag. No actual measurements of the loss of total head

in the v/ake in Freon-12 were made by von Doenhoff. Experimental justification of the applicability of stream-line similarity

would be desirable.

(2) Swept Wings. An ambiguity exists in the application of streamline similarity to sv/ept vings.

(3) Moments. The accioracy obtained iS the conversion factor is assumed to be dependent only on the free-stream Mach number may not be sufficient. Tnis is indicated by the scatter of the conversion factor when plotted against free-stream mach number.

The difference in area ratio at the same Mach number between air and Preon become appreciable at supersonic Mach numbers (e.g. M> 1.2) and hence it is doubtful whether stream-line similarity exists at

supersonic Mach numbers much in excess of about 1,2. To investigate the usefulness of Preon-12 at Mach numbers greater than 1,2 would

require a supersonic tunnel equipped v/ith Schlieren Apparatus, Por economic reasons this tunnel coiiLd be an intermittent type, Shock-wave boundary layer interaction could be studied in such a tunnel.

Measurements in Preon-12 of wake drag and pressure distributions on sY/ept wing models are necessary in order to obtain the appropriate conversion factors. The first of these has been partially explored in the present series of tests.

6. The College of Aeronautics Preon-12 Wind-tunnel

The general design of viiiid-tunnels using a medium other than air is discussed, follov/ed by a detailed description of the College of Aeronautics Preon-12 wind timnel.

6.1, General design considerations

The simplest approach to the use of a medium other than air for aerodynamic testing is a blow-dovm tunnel in ¥."hich the gas is allowed to escape directly to atmosphere (ref.17) The savings obtained by dispensing with any additional equipment are offset by

-24-the loss of -24-the gas after each run.

In general the design is not unlike that of a coraprss&ed

air tunnel except that it is more important to make the timnel

leak-proof to avoid either large losses or the contamination of the gas by air. In addition provision has to be made

for:-(a) Pilling the tunnel with the gas (b) remov-lng the gas

(c) pxirilfying the gas

(d) s t o r a g e .

Additional problems may arise because of the special properties of the mediijm, though in general the gases that are considered are substantially non-corrosive, non-toxic, non-inflammable and stable (see section 6,3)

To fill the tunnel with the gas the air can be evacuated, as in the College of Aeronautics tunnel, or if a heavy vapour is used it is possible to introduce it slowly at a low point in the t\innel and to remove the air from the highest point. This latter method is

especially usefijil if the tunnel shell cannot withstand pressure differences as is often the case.

Recovery

If it is desirable to avoid the loss of the gas each time access to the tunnel is required, provision has to be made to remove most of it from the tunnel, to a storing unit. Pressure bulkheads

arou;id the working section v/ill be a means 'of considerable time saving v/hen access only to the model is required. The capacity of the

pun^iing equipment will depend on the volijme of the tunnel system and

the time interval in which it is required to reach a specified lev/

pressure. It shouild be pointed out that as the system is inva.riably never completely leak proof the lowest pressure that can be reached will depend on the capacity of the pimip and the rate of leakage.

The storage containers are much more bulky if the medium is stored in the gaseous state rather than as a liquid. Liquid storage is only practical if the critical temperature of the medium is v/ell above the room temperature encountered, otherwise constant refrigeration v/ould be necessary.

2 5

-6.2. Fi.irificat.lon

Some mixing of the gas v/ith air and moistuj-^e is unavoidable,

If the gas is to be used repeatedly some means of purifying it will be

necessary. Moisture can be removed by pass.ing the gas over a suitable

drier containing for example silica gel or activated alumina. Both

these desiccants are liable to disintegrate into small particles which

are then carried away by the gas. The risk of disintegration increases

at highenr gas velocities. The drying of the medium in the liquid phase

when the velocities are much lower and the time of contact correspondingly

longer, might prove advantageous.

T/hen the vapour is liquified, non-condensible gas ijiipurities,

such as air v/ill collect above the liquid and the pressure v/ill rise

as liquification proceeds. The non-condensible vapours v/i.ll have to

be purged when the maximian delivery pressure of the condensing unit

is reached. Ideally the proportion of the vapour in the p\irged ^as

will be equal to the ratio of the saturation va.pour pressmre p at the

temperatiire of the liquid to the pressure in the liquid container, p.

i.e. to TD

P

The saturation pressure decreases v/ith the liquid temperature and

hence for minimum vapour losses it is desirable to have low liquid

temperat'ure combined with high pressure in the output stage of the

condenser. In practice the proportion of vapcui' lost vail probably

be greater than p , as additional vapoixr tends to remain suspended

p ~

in the form of fine mist iji the purged mixture,

6.3.

The 5.25in. x7,5i'nAnnular return P-12 Turjiel at the College of

Aeronautics

A programme of research into the use of media other than

air in vdnd-tunnels v/as initiated at the College of Aeronautics in

1951 and Precn-12 vras chosen as a v/orking medium,

Properties of Freon-12 (Dichlorodi-flucro-methane, C01_ P )

The physical and thermodynamic properties are listed in

ref.35(at low pressures ) and ref,33,

The critical temperature is 111,5 0 and critical pressure

39.6 atm. The vapour pressure at 60*^ is 72,4 Ib/sq, in. Preon-12

in the liquid state is colourless and has a faint ethereal odour,

It is stable at ordinary temperaturesbut dissociates when in contact

with a red-hot surface or flame. It is also non toxic, non-irritant,

non-inflammable and non-explosive.

It does not corrode ordinary metals and alloys used industrially excejjt for magnesium, and its alloys.

Most lubricants used for re.frigeration puiTposes are mixi.ble without reaction, though separation can occur at lov/ temperatiu'es.

It-.is seen from this description that no special difficulties a.rise in the handling of Freon-12, except that special bearing design is required because of the de-oiling properties.

Tunnel

I t v/as decided t o u t i l i s e a l / l 6 t h model of t h e Ri'iE /jm'ular RetiArn High Speed Tunnel ( r e f , 2 9 ) . A c r o s s - s e c t i o n a l dravóng apioears i n f i g . 8 . The v/orking s e c t i o n i s 7.5inx: 5.25ineand lO,25iaT,long. Access t o t h e workijig s e c t i o n i s ••ria. removable covers i n t h e o u t e r - s h e l l and the annulus. The annulus i s made of gun metal and supported by stream-l i n e s t r u t s , Press'ore connections t o outside manometers e n t e r the annulus i n s i d e t h e s e s t r u t s . The o u t e r s h e l l i s of c a s t i r o n and can wi.thstand vacuum and two atmospheres a b s o l u t e . The c a s t i r o n s h e l l v/as found t o be extremely porous and had t o be p a i n t e d w i t h s p e c i a l r e s i n s t o make i t leakproof a t low p r e s s u r e s . The i n t e r n a l volijime of t h e t u n n e l i s about 18 c u . f t ,

Pan & Motor

To seal the fan-shaft effectively which has to revolve at high speed, is very difficult and hence the whole motor was enclosed

in a pressure shell which was ventilated to the rest of the tunnel, Originally the fan was mounted on a fan-shaft v/hich was supported on two special impregnated cage bearings. These were tried because of the de-lubricating properties of Preon-12, The motor drove the fan-sha.ft tlii-ough a flexible coupling. Serious

venti-lation trouble was experienced which was mainly due to insufficiently rigid mounting of the motor, the tunnel housing itself a.cting as a •

diaphragm. The original arrangement v/as abandoned in favoior of the

present one which has the fan mounted directly on the motor sha.ft. The motor is now fixed to the timnel end cover so as to provide least overhang. The motor is a s;>'nchronous induction motor giving 100 H.P. at 10000 RPM. A sketch of the associated control equipment is shov/n in fig, 9.

The 13 blade fan is of one stage with fixed guide vanes upstream and dovrnstream.

^Zj.

Preon- 12Circuit Lay-out

A diagrammatic sketch of the Preon-1 2 t\annel system is l^resented in fig, 9, A general description of the system emd the manner of operation is given below, followed by a detailed description

of the various \Jnits,

The gas exit pipes leave the top of the tunnel at three points and there is also one connection to the top of the gas storage tank, Similar connections exist to the bottom of the tunnel and storage tank for the entry pipes. This ensures that any residual air left over in the tunnel is efficiently displaced by the heavy Preon gas. The exit pipe leads to a Geryk vacuum pump capable of exhausting the whole system to 0.5 nun Hg abs. pressure in about -g- hr.

The Preon-12, which is stored as a liquid in a special container at up to 180 Ib/sq, in, abs pressure, is then allowed to evaporate through an orifice plate. The orifice plate prevents a too rapid evaporation and consequent large drop in temperature v/hich

would cause any moisture present to freeze and probably block the passages. Prom there the Preon passes through a "stillite" oil filter which

serves to remove any oil mist carried over by the vapour. This may be caused by oil dissolved in the liquid Preon phase. Most of the oil is picked up in either the condenser pumps or the circulation puinp, both of which are of the reciprocating type.

The Preon-12 vapour then passes through an activated alumina bed and then a silica gel drier into either the gas storage tank or through a heat exchanger into the tunnel. The activated alumina v/ill also remove any oil mist that may have remained in the Preon, The heat exchanger utilises the hot air from a 1 kv/ heater viiich is provided in the silica . gel unit for reactivation. In practice it is not used, as the gas after evaporation very quickly reaches room temperature,

6,5» Circulation

Constant stagnation pressiiro is maintained in the tunnel by means of an automatic valve, controlling the rate of flow through the circulation pimp and the suction of the condensing unit. Fresh gas is allowed to enter the tunnel from the gas storage tank. By continually circulating the gas in the tunnel its purity can be maintained over longer periods. It also serves to maintain lower

running temperatures especially i^f the condensing unit is used to circulate the gas. The gas storage tank is used for the dual piirpose of increasing the effective volume of the tunnel and also to permit the removal of the gas from the tunnel vdthout reliquifying it.

-28-The condenser can v/ork xjcp to a delivery pressure of 165 Ib/sq, in, This is just over double the vapour pressure of Preon-12 at room

temperature and hence the coiqjosition of the mixture purged dviring liquification will be a.lmost of equal amoimt of Preon~i2 ar'-'' non-condensible gases. T?ius efficient recovery of the Preon-12 is not possible for mixtures containing relatively large proportion of air.

The minimum inlet pressure to the condensing unit is about •5 atm, and hence for efficient recovery of Preon-i2 from the system it would be necessary to use an auxiliary pump.

The capacity of the gas storage tank is 70 cu.ft. and that of the whole system approximately 100 cu.ft.

The total charge of Freon required at 1 atm. pressure and 60 P is about 36 lb.

•^.Tiile the tunnel is running a small amount of Preon controlled by a valve is allowed to pass through the sound measuring gauge, (see

section 6,6, ) so as to keep a constant check on the purity of the Freon -1'2.

Care v,as taken in the design of the system to prevent any liquid Preon being sealed in the pipe circuits as this might cause rupture should the temperature increase,

The tunnel pressure v/hen evacuated can be measured on a vacustat,

Valves are provided in the circuits so that in addition to operational requirements it is possible to seal off each \mit from the tunnel to allow quick repairs. It was found necessary at all times to ensure that the valve packing glands v/ere adequately tightened to prevent leakage.

6.6. Detailed Description of Components

Liquid Receiver; The capacity is apiaroximately 36OO cu, in. for a maximum charge of 152 lb of Preon-12. Provision is made for the transfer of liquid Preon-12 directly from the manufacturers containers. The v/hole receiver • is suspended from a balance to enable the mass of gas to be determined.

Freon Condensing Unit

The 'ï'iulsometer" compressor is a twin cylinder reciprocating type delivering 385 cu,ft. per hr., at 7OO r.p.m. It is capable of delivering approximately 0.75 cu.ft,/min, at atmospheric inlet pressure

-29-and an exit pressui--e of about 120 Ib/sq,in. The water circulation through the cooler is about 150 g,p,h, at 60 P.

S^t_illite filter

Stillite is a. form of slag wool of which the fibres are of very small diameter, the general theory of fibre filters has been the

subject of papers by Stairmand (ref,56).

The required packing density iU 17 lb/ft , at a face velocity of 1 ft/sec. This gives a pressure drop of 0,3 Ib/sq:,in/foot depth of packing. The filters operate in the Stokes' flov/ regime and hence the pressure drop is proportional to the face velocity,

The filter on the frecn-tunnel consists of a horizontal disc of 14" diameter and 3" thick. The gas flo/ra upv/ards through the disc and at face velocities less than 0.5 ft/sec. the oil should be able to drain against the flov/ of the gas.

Activated Alumina Drier

The drier consists of a cylinder of 8" diam.eter find 3'6" long filled with granules of activated alumina. During operation cold Preon-12 gas from the liquid condenser is passed through coils

embedded in the activated alumina to remove the heat liberated by the absorption of moisture from the vapour that is passed through the alum.ina itself.

Silica Gel Drier

The plant is designed to dry 2 c.f.m of Preon-12 at pressiores ranging from 14,7 Ib/sq. in to 30 Ib.sq.in. v/ith inlet temxse.ratures of about 60 P down to a moistixre content of less than 0,0002 lbs per lb.

of dry Preon. The plant can iDorform this duty for a period of about 8 hrs. It has then to be regenerated by heating and drying by means of the fan and heater unit supplied with it,

Autoniatic stagnation pressiire control valve

Tliis is a Short & Mason pneumatically operated control valve operating over the pressure range zero to two atmospheres using

atmospheric pressure as the reference pressure, Fulnometei" type 'Geryk' Rotary Vacuimi I'ump

The sv/ept volume of the pxanp is 20 cu,ft/min at 36O r.p,m, and it is capable of achieving 0.005 ram,. Hg. The lowest pressure

-30-Circulating Fvxnp (polsometer)

This is a reciprocating type of pimip having a swept volume of 7 c,f,m. The minimum inlet pressure v/as 1 in.Hg. absolute vrfien the delivery pressure v/as about 30 Ib/sq, in. Some difficulties were experienced vdth the use of this pump due to the gas picking up oil in its passage through the pump.

Description of the sound measuring gauge

A special gauge to measure the speed of sound in Freon-i2 was designed by the Electrical Department of the College of Aeronautics. A schematic diagram of the gauge appears in fig,10.

The principle employed in this gauge is to measure the time required for a signal to travel a given distance.

A high intensity sharp fronted signal is fed into one of the transdxAcers and is also used to open a gate allowing pulses from a quartz oscillator to pass to a microsecond counter. The arrival of the

signal at the receiving transducer is used to shut the gate. In order to obtain the speed of sound from the measurements the temperature of the gas has to be known and is measured by a thermometer in the outlet pipe of the tube. The sampling rate of the gauge can be adjusted to

give 1 sec, 2 sec, 5 sec, and 15 sec, sampling intervals.

The transducers consist of thin aluminiijm foil stretched betv/een a magnet. This gives maximum responsiveness to sharp edged signals and also acts as a good positional location,

The distance between the two transducers is three feet and the diameter of the pipe is 4in.

The proportion of air in the freon is found by the method discussed in appendix 11.

7. Calibration of Preon Tunnel 7,1, Introduction

The l/l6th scale model of the R,A.E. High Speed Tunnel was calibrated with Preon-12 over the entire speed range. The \jniformity of flow in the v/orking section was determined by measuring the pressure at wall tappings and by means of a traversing pitot-static tube,

"31-7,2. T\Jnnel and Apparatus

The tunnel was calibrated in ter:ns of the pressure differences between the settling chamber and the first tapping which is 33" ahead

of the v/orking section. The pressure at this tapping shoxild not be affected by the presence of the model. This pressure difference was measured on a "master manometer", -v/hich is a null reading instrument, and can be read to v/ithin 0.002" of mercury. A diagram of the arrange-ment appears in fig.11. The pressures at the four tappings in the working section and at the tapping ahead of it are denoted by p , Pp, p,, p, , Pc respectively and the pressure recorded by the static tube

is denoted by p' . The difference betv/een these pressures was measured on a bank of mananeters filled with Butyl Phthlate. In addition the stagnation p.ressiare, and the total head and static pressure of the traversing pitot tube v/ere recorded against an atmospheric datum, A thermo-couple mounted on a st.rut projecting into the centre of the settling chamber alD.owed the stagnation teirperature to be read,

Figures 12 and 13 show the pitot-static tube and its traversing mechsmism. The gauge is derigned to fit into the limited space in the annulus and is driven by a remotely controlled selsyn motor through an arrangement of gears and a screw jack,

7,3..Accuracy

Notation p = atmosphejpic pressure

p = pressure at wall tapping in settling chamber

p = pressirre at wal.l tapping 1,7 in, downstream of beginning of v/orking section p = pressure at v/all tapping 4 in, " " " " " " p_ = pressure at wall tapping 6,3 in. " " " " " "

3

p. = pressure at wall tapping 8,45 iJ^. " " " " " " Pp. = pressure at v/all tapping 38 in, " " " " " " p^ - pressure at any of the wall tappings 1, 2, 3, 4,

H = total head pressure in working section

s

(a) Instruments

The manometers can be read t o within 0,05in.with the exception

of the master manometer and the barometer used to measure atmospheric

p r e s s u r e . These can be read t o within 0,002in.Hg, The specific g r a v i t y

-32-With p . = 30inHg and p, = 400inPutyl P h t h l a t e

Ap^ = ± 0.001 -hl. Hg.

. ' . A(p - V^) = - O.linButyl P h t h l a t e

Thus Ap_^ = Ap + A(p - p ) ^ ± 0,1 i n . Butyl P h t h l a t e

and the accuracy of t h e s e t t l i n g chamber pressure i s

— - = - 0.00G3

Pt

The accuracy of t h e difference betv/een any of t h e p r e s s u r e s

Vy Pg. Vy P^, P5. P' ^^- - 0.2inButyl P h t h l a t e

or Ap

—s. .., ± 0,0005

Pt

(b) Accuracy of Measurements

The pressure difference p. - p^ measured on t h e master

manometer was found t o be s i i f f i c i e n t l y steady a t low speed, M^ = 0,4,

t o be read witliin - O.OIOinHg and a t M^ = 0,8 t o only about - O.linJIg,

At M^ = 0 , 8 , t h e difference between the p r e s s u r e s p . , Pg.'.Pc

could be read vd.th an estimated accuracy of - 0,2in.Butyl P h t h l a t e ,

Hence (at atmospheric stagnation pressure) and M^ = 0,4,

^ . ± % 2 1 « ±0.0003 . ^ '

v^ 30 ^

Since Zi(p^ - P^) = ^ O.OijnJIg.

we have A(p^ - P3) ^ + ^^^^3 i f p^ _ 3 ^^.^g^

Pt - P 5

Hence ^ { H_:^p ) ~ - O.OO3

H_

H

-H

3 3 -( ^ ^ - ) and w i t h ( ^~r^- j = 0 , 1 M = 0 . 4

Y—H^).

b e h a v e 1 ^."^ ) + c , 0 0 0 3 T h e r e f o r e AM = i 0.001 and ^oo = i 0 . 2 ^ M At M = 0 , 8 °° 00 S i n c e A (p. - p ) = ± O . l i i J I g , a n d ( p . - p - ) = 10 i n . H g . t 5 "c 5

v/e have A(p - p )

i ^ = t 0.01 P t - P 5

T h e r e f o r e A^ H - p J = 0,01 x 0 , 3 = t 0,003

AM

and A M^ = t 0.OO5 o r -£• = - Q),5%.

Aocurac.y of t h e Mach number v a r i a t i o n a.long t h e v/orking s e c t i o n

At M = 0 . 4 00 A(Pc - P^) ~ Ap = ± 0 . l i n B u t y l P h t h l a t e g i v i n g Au O..J_ ~ + p 400 u.uuyj). -, AM ^ ^ - ~ ^ = ± 0.002 or AM = ± 0,001 M 00 CO At M = 0 . 8 CO '^(Pc - P^) - -^p = i 0 , 4 B u t y l P h t h l a t e g i v i n g Ap _CV4 p - 300 = u , u u i , ^ ^ ^ = i 0,001 or AM^ = ± 0,001 M —

-34-7,4. Experimental Measurem.ents

All vhe mancmeters, shov/n in fig, 11 and the stagnation temperature were read v/hile the fan speed v/as kept constant. The operator controlling the motor speed v/as unable to observe the ma.ster manometer and hence no attempt was made to run the tunnel at any

specific Mach number or to keep it accurately constant during a. run, In the liigher speed range the readings v/ere talcen before the tunnel reached its terminal temperat-ure v«hich continued to increase over a considerable period in spite of continuous cooling of the tunnel, However the change in temperature vxas small during the intemral required to take readings. Sets of readings were taken v;hile the fan i'p,m, was increased in steps of 300 from about 36OO to 7200 rpm

equivalent to a Mach number range of 0.4 to 0.9.

Readings were taken with the total head tube at 0.75", 1.25", 1,75", 2.25'; 2.75", 3.25", 3.75", 4.25" and 4,5" from the wall from v/hich the gauge v/as mounted,

Readings v/ere aloO taken at 1,7" (opposite pressure hole no,l) 5,4" and 8,45" (opposite pressure hole no.4) downstream of the beginning

of the working section along the centre-line of the tunnel.

7.5 Results from total head measurements

The error in assuming that the pitot tube pressure is the actijal total head pressure should be negligible even if the pitot tube is yawed slightly, whi.ch might be the case at high speeds.

Pt

v/as plotted over the v/hole speed range and the

v-arious transverse positions for each of the three downstream positions (see fig. 14, 15, 16), be' denotes the dovrnstream distance of the plane of measurement and 'y' the distance of the total head tube from the v/ell on v/hich it is mounted.

The variation of H - p, for the varioxis transverse positions of the gauge

Pt was t h e n found. Prom t h e above f i g u r e s AH _ AH - P Tj ~ 1 Z ±0.»001 abM = 0 . 4 P t g i v i n g AM t 0,005 M and AM = - 0 . 0 0 2 .

-35-At M = 0 , 8 oo

% - i ^ ^ l l l i . ) - i 0.003.

givlag AM

co X 00

or AM = i 0,003

oo

The average values of H - p. a t the various t r a n s v e r s e

Pt

p o s i t i o n s , are p l o t t e d in f i g . 1 7 for t h e t h r e e dovrnstream p o s i t i o n s .

They agree vvith one another to v/ithin 0,0005 and also are close t o

the tunnel c e n t r e - l i n e values.

"FT •— T'ï

They also agree well v/ith t h e c a l c u l a t e d value of " • t based

P t

on the Area r a t i o

A

7^ = 6,55

where A = settling chamber area and A = v/orking section area

w

7.6. Transverse Variation of static pressiAre in the working section The variation of static pressure across the working section is more difficult to measure, as corrections due to vajriable blockage and yaw which might arise through bending of the tube under load may be large.

P. - P' P, - P'

and -^ were p l o t t e d

Pt Pt

i n f i g , l 8 and 19 i n t h e case where the plane of measurement coincided

v/ith p r e s s u r e holes 1 and 4 r e s p e c t i v e l y .

P' " Pi P' " P4 ~ +

1 and = ~ i 0,0005

Pt Pt

TECHNISCHE HOGESCHOOL Vi.iEGTUIGÖOüVVK'.JND':: Karadstraal -! O - D'^LrT -3b-^^^ ~ '^ r.r.-' AM + „ _,^. g i v i n g _ °° ~ 0,003 o r QQ = - 0.,001

i^L

a t M = 0 , 4 oo

At higiaer sx^eeds the scatter of results becomes greater but is I)robably not representative of the actual static pressure variation

in the absence of the pitot -tube.

Prom the variation of static and total head pressure in the v/orking section the flo-v/ is estimated to be uniform at least to within AM^'^i 0,005

7.7? Vari-ation of static pressure along the workixig section

Meas-orement of the x^ressure at the four wall tappings in the working section (p , p^, p.^, p, ) and at the tapping ahead of the working section (p^.) were ifiade'^witft an empty working section,

The values cf p^ - p, 5 1

P4.

P2 ~ Pj , P3 -Pi ^ " P -( _{with the tunnel empty agreed with} Pt Pt ' Pt

those made when the pitot static tube v/as mounted opposite press-ure hole 4 (e.g. 8,45i'n.downstream of beginning of w,s,)

They are plotted in figs. 20, 21, 22, 23 respectively as a function of the tunnel pressure ratio p - p^.

P t P5 - Pl

-^ i s p l o t t e d i n f i g . 23 u s i n g meas-urements made v / i t h t h e

P t

t-unnel c o m p l e t e l y empty.

The v a r i a t i o n of Mach raraber from t h e v a l u e a t s t a t i o n 2 was c a l c u l a t e d from

AM = 9M ' t^fv] 9 M , ^ P . J . ^ „ .

.ot-F(i;^) \EJ = ^m H ^'^^^ ^ = ^'^25

assuming no variation of total head.

Using the values of P3 - P^ ^ Pi - Pg P^ " Pi obtained Pt Pt Pt from figs. 20 - 23 as A2 in the above equation the value of the

-3/-Mach number variation a-long the working section v;as calculated and is plotted in fig.24 for M^ = 0.4, 0,5, 0,6, 0,-7, 0,8, 0.85.

The maximum deviation of Mach nxjmber in the working section from the value at station 2 on the basis of these measurements is

AM = 0.003 at Mg = 0.4 and AM = 0.005 at M^ = 0.85.

7.8. Working section Mach number

The v/orking section Mach number was based on the static pressiore measured at No.2 taj-ping in the v/orking section. Thus, if

Pt and p^ - Pg

~ ~ P T ~

= h then H = p^ (h +1) = JT or p^ = p. (l - ^) givung p^ ^ ^ _ ^ H 1 + h

h is found from fig.15 and w from figs. 13, 20,

The Mach number was then calciilated from the isentropic flow relation

t = [ i . ^ M-J-y^^

with- y = 1,125

The accuracy of the calibration is estimated to be A M = i 0,005,

7.9, Purity of Preon

The amount of air present in the Preon v/as usually from 2^

to 6% (by mole fraction) this was frequently checked by moans of the sound gauge meas-urements,

-3o-8. Two-dimensional aerofoil tests in the Freon»-i2 tunnel 8.1. Model

A 2in;3hord model of an Ri'JS 104 IC^S Symmetrical aerofoil section v/as obtained through the courtesy of the N.P.L. The ordinate of the RAS 104 section are given in (ref,3l) and the position of the pressure holes on the model are shown in fig.25.

A large n-umber of tests have been made at the N.P.L. on the two-dimensio-nal models of PAE 104 synmetrical aerofoils including some on a 53i%chord model in the 18210c 1ipii.tunnel,

Comparison of measurements were made between those on the

2in.chord model in the Freon-12 timnel at atmospheric stagnation pressure, and similar measurem.ents in the N.P.L. l8±xx I4iaair tunnel on the

5 in. chord m.odel at atmospheric stagnation pressure, In this coniparison we have

(a) The Reynolds number is almost identical

(b) The ratio of tunnel height to tunnel width in the two cases is very similar,

e,g;:-i = (J|) = 1.3 »d(e,g;:-i%) =1.4.

Air Freon-12 and

(c) The ratio of model maximum thickness to t-unnel height in the case of air and Freon-12 tunnel is

•%! = 0,028 and-§4 = 0,027 18 7.5

respectively. Hence the magnitude of the solid blockage correction vn.ll be of the same order. This allows for a comparison on the basis of results uncorrected for blockage.

8.2. Model McuT-ting and pressure measurem^ents

Nev/ side v/alls v/ith turn-tables to take the 2in.chord model were made. The turn-tables were equipped v/ith a vernier allov/ing the model incidence to be positioned with an accuracy of approximately - 5 minutes.

The position of the model in the v/orking section is shown in fig.26.

-39-Press'jre MeasiJrements

The pressure difference between the v/orking section reference pressure and the settling chamber i^ressure was measured on a master manometer. The pressures on the wing v/ere measured on a butyl phthlate manometer at low speeds and on a mercury manometer at high speeds. The difference between atmospheric pressure and the settling chamber pressure v/as also recorded. The total head presstire can then be obtained from fig. 17 or directly from readings of a 1 mm outside diam.. total head tube mounted 1 in.downstream of the trailing edge and 1,5in.from the v/alls and

floor (see fig,26). 8,3» Pitot traverse gear

A remotely controlled traverse gauge was designed for pitot traverse measurement from which the wake drag v/as determined, As there was not sufficient room in the annular space adjacent to the

walls parallel with the wing, the gauge had to be moxonted of the same walls as the v/ing itse.lf (fig, 12). A nearly vertical traverse of the pitot tube v/as made by swinging it about a hinge mounted in the turjiel wall. The radius of the pitot head arm was 3.3in,so that the traverse of the v/ake was not at mid-span. The head v/as driven by a. selsjm motor through a reducing gear box. The total head tube was mounted about a hinge in the working section v/all and was c^riven by a wheel working on a quadrant affixed to it,

The v/hole driving mechanism was contained in a gas proof box to prevent leaks into the working section. One turn of the motor moved the head through 0.001 in,

The motor was driven electrically and the number of turns . with respect to the central position were noted. At the start of

these tests the pitot head consisted of a -fin. length of 0,02Cin. outside diam. tubing telescoi^^ed into a tube of larger diameter, but this was found

unsatisfactory due to a too large time lagi.. Finally a head of 0.040inputside diam.. tubing -was used. This head proved satisfactory except that it read

lov/ when the tunnel warmed up after running at high speeds for prolonged periods. This v/as eventually traced to some plastic tubes v/hich

developed leaks at the higher temperature encountered iri the turjiel. Modification to the tube connections provided adequate sealing,

8.4. Calc\ilation of Results

The imcorrected Mach number M v/as obtained from the calibration chart. The pressure coefficient C was calculated from

°P = % " \% L, , =V^-^ u^ i P u 2 / =

•f -t-^ \ - ^ / b l o c k a g e «^ oo . «^ oo "-^

Pn-Pco PQ-POC

( P n ' ' P 5 ) + ( p ^ - P g ) * ( P 2 - P c ^ IP IT2

3 o

-8 . Two-dimensional a e r o f o i l t e s t s i n t h e P r e o n - i 2 t u n n e l

8 . 1 , Model

A 2inphord model of an R/iE 104 10^ Symmetrical aerofoil section was obtained through the courtesy of the N.P.L. The ordinate of the RAS 104 section are given in (ref,31) and the position of the pressure holes on the model are shown in fig.25.

A large n\amber of tests have been made at the N.P.L. on the two-dimensional models of RAE 104 symmetrical aerofoils including some on a 5itT.chord model in the I8it%.x 14in,tunnel,

Con^arison of measurements v/ere made between those on the

2in.chord model in the Preon-12 tunnel at atmospheric stagnation pressiire, and similar measurem.ents in the N.P.L. l8dtn,x I4in.air tunnel on the

5 in. chord m.odel at atmospheric stagnation pressure. In this comparison we have

(a) The Reynolds number is almost identical

(b) The ratio of tiinnel heigjit to tunnel -vvidth in the two cases is very similar,

e.g:-Air Preon-12 and

(c) The ratio of model maximum thickness to tunnel height in the case of air and Freon-12 tunnel is

% | = 0.028 and-14 = 0.027

18 7.5

respectively. Hence the magnitude of the solid blockage correction vd.ll be of the same order. This allows for a ccmparison on the ba.sis of results xmcorrected for blockage.

8.2. Model Mcur.ting and pressure measuremients

New side walls v/ith turn-tables to take the 2ün,chord model were made. The turn-tables were equipped v/ith a vernier allov/ing the model incidence to be positioned with an accuracy of approximately - 5 minutes.

The position of the model in the working section is shown in fig,26.

-39-Press^ore Measiorements

The pressure difference between the working section reference pressure and the settling chamber i3ressure was measured on a master manometer. The pressures on the wing v/ere measured on a butyl phthlate manometer at low speeds and on a mercury manometer at high speeds. The difference between atmospheric pressure and the settling chamber pressure wfas also recorded. The total head pressure can then be obtained from fig. 17 or directly from readings of a 1 mm outside'diam.. total head tube mounted 1 in.downstream of the trailing edge and 1,5in.fr'om the v/alls and

floor (see fig,26), 8.3. Pitot traverse gear

A remotely controlled traverse gauge was designed for pitot traverse measurement from which the wake drag was determined. As there -.vas not sufficient room in the a.nnular space adjacent to the

walls parallel -with the wing, the gauge had to be motmted of the same walls as the v/ing itse.lf (fig. 12). A nearly vertical traverse of the X)itot tube v/as made by swinging it about a hinge mounted in the turJiel wall. The radius of the pitot head arm was 3,3in,so that the traverse of the wake v/as not at mid-span. The head v/as driven by a selsjm motor through a reducing gear box. The total head tube was mounted about a hinge in the working section v/all and was driven by a wheel working on a quadrant affixed to it,

The v/hole driving mechanism v/as contained in a gas proof box to prevent leaks into the working section. One turn of the motor moved the head through 0,001 in,

The motor was driven electrically and the number of turns . v/ith respect to the central position were noted. At the start of

these tests the pitot head consisted of a -fin. length of 0,02Qin. outside diam. tubing telescoped into a tube of larger diameter, but this v/as found

unsatisfactory due to a too large time lagi,. Finally a head of 0,040input side diam,, tubing was used. Th.is head proved satisfactory except that it read

lov/ when the tunnel warmed up after running at high speeds for prolonged periods. This v/as eventtially traced to some plastic tubes v/hich

developed leaks at the higher temperatiore encoimtered ixi the -tunnel, Modification to the tube connections pro-vided adequate sealing,

8.4. Calculation of Results

The uncorrected Mach number M v/as obtained from the

c a l i b r a t i o n chaEt. The pressure coefficient C v/as c a l c u l a t e d from

P

Pjj*"Poo PQ"*POO