CP. No. 308 (18.7112)

A.R.C. Technical Report

C.P. No. 308 (18,712)

A.R.C. Technical Report

MINISTRY OF SUPPLY

AERONAUTICAL RESEARCH COUNCIL

CURRENT PAPERS

The Development of an Improved Diffuser for a 3 ft. x 3 ft. Wind Tunnel

J. B. Scott-Wikon, and D. I. T. P. Llewelyn-Davies

LONDON: HER MAJESTY’S STATIONERY Of:FICE

1956

TWO SHILLINGS NET

U.D.C. No. 533.6.07Ll I 533,6.0%5

Te&nical. Not e No. Aero 24i1.6

July, 1956

The Development of zn Improved Diffuser for a 3f%x3f%Tunzlcl

bY

J. B. Scott-Tifson and

D, I, T. P, Llewelyn-Davies

of Royal Aircraf't Establishment, Bedford

A aunmary is given of the significant results of diffiscr developnerd tests made in a &' x 4” supersonic tunnel at Mach numbers fk+oih I.4 to 2.0. The application of these results to the aesign of an mr0vca first diffuser stage for the 3 ft x 3 ft supersonic tunnel, is described. The design incorporates along single wedge centrebody, &dch ditides the diffbser into two appzvximatcly oonstant area passages. Using this improved difisor, the pressure ratio needed to run the tunnel at its m&mumM~h number of 2.0, is r&&cd fkom 2.04 to I .76.

1 Mmduction

2 Description of the 3 I? %unnel

3 Diffbser Tests in the 4” x 4” Tunnel

4 'Phe Design of the Balancs Section with Quadrant

5 Tunnel. Parfozxwce Measuremexts

6 Cbnclusions

Reference s

SE

3

3

3

b

5

5

5

General armngment of tunml cimuit and auxiliaries - 3 ft x 3 ft sqymion5.c windtml

Diagmmmatic ~LYOU~ of I;a.lance section -xith CalibratioiI gear

Details of w-dxebdies used for tests in the 4” x 4” tunnel

Effect of centrebodies on diffuser pwfomame 4" x 4" tunnel

Effect of ecpivdent conical semkm$Le on diffiser performance 4” x 4” tunnel

Diagrammatic layout of balance section with quadrant

Tunnel performance me asurelnents 3 ft x 3 f% t-l#Llel

1

2

3

4

I Introduction

The 3 f% x 3 ft transonic and-supersonic tunnel at R.A,E, is a closed circuit pressurized tunnel capable of operation at Mach mbcrs from about 0.7 to 2.0. A feature of the design of this tunnel is the use of two easily removable sections, the "balame section", containing the suppcrt gear for the model sting, and the "mobile dif'fL~ser", inmediately downstream of the working section. The rigging of models is greatly simplified by this arranguncnt, and alternative balance sections canbe

US& The avtilable length of the movable se&ions however imposes a limitation on the design of a diffuser for supersonic: operation; it is not possible to install a simple adjustable flap system, for example. Ixrring the first four years that the tunnel was in servioe only one balance section, the one designed to hold the calibration gozr for the tunnel', was available, and a temporary model support system was installed in this section. Recently a second balance section, -&th a c~adrant type of mod01 support system has been completed. The interior geometry of this section was designed as a result of modclteats in a 4" x 4” tunnel, to give impmved ef'f'iciemy as a diff'user at supersomc speeds, In this note the significant results of the diffiser develoment tests in the 4” x 4” tunnel are presented, and the perf'ormarrce of the 3 f-t x 3 ft tunnelwiththe second balanoo section is compared with its earlier pcrf~nnance to show the gain in efficiency achieved.

2 Description of the 3 f% x 3 ft Tunnel

Fig.1 shows the general arrangement of tho tunnel, the driving plant and the tili;iry machinery. For supersonic operation the working section is 3 ft square, and the Mach number range is from I.:5 to 2.0. Single-sided fixed block nozzles are used to generate the flow; nozzles arc available to give working section Mach numbers of 1.32, I,&?, ? .61, 1.82 and 2-00, The pressure in the tunnel circuit is variable from '/2Oth atm, to 2 atm. absolute. The ‘air is circulated by ttm doubk entry single stage centri- figal compressors, which are set in series with an inte~ooler, Vith the Mach number 2,0 nozzle the pressure ratio available at the maxim per- missible speed of the driving plant is just over 2.0, An aftoroooler, located in the settling chpmbcr is used to control the stagnation tempera- twe,

The bal,ance section with the calibrating gc,ar that vas used for all the early running of the tunnel, is shown diegr,~tWLLy in Fi.g.2. The entry and exit areas of thi s section EIJ% fixed by the tunnel design, and give an overall expansion rote e@vLLont to that of a cone of semi-angle 2O22', The interior liners are, however, &signed to mainkin a const,ark area passage up to the end of the calibration saddle,, and then to cxp3nd with an cqyiwdcnt conicLit s~mi-~mgle of 3’13’ to the exit section.

The design of the balance section with the quadr,ant is described in Section 4, after a summary of the diffuser tests in the 4” x 4” tunnel has been given.

3 Diffuser Tests in the 4” x 4" Yunnel

A brief sua[p11~xy is given here of the significant results of the diffuser development tests, which provided the basis of the dc:sign of the internal layout of the second balance section for the 3 f% tunnel. These tests were made in the 4” x 4” tunnel, which is one of t~ro small supersonic tunnels powered by the 3 ft tunnel CLUXU~~~ plant, The settling chzonber, nozzle and working so&ion of this tunnel arc l/pth scale nc)dcls of the larger tunnel. Tho first diffiscr stage, which is removable, and to which modi- ficat ions wore made during these tests, is not to scale, and pmvid2s a

-3-

rather more rapid exp~~i~n, the e~@.mlent conical semi-e&e being 2'&2*, than the 3 ft tunnel balance section. This difYuser stage has a leiigth of 1 S", akl basicalLy four straight diverging walls. The modifications mac?e, for the results given here, were as follo-kvs:

(FA) the fitting of two-dkensional centrebodies,

(b) the reduction of the angle between the VELUS for the first 44” of the diffuser.

In the first series of tests four centrebodies, vrith dimensions as shown in Fig.3, were fitted, each vertically spanning the diffiser. All the centre- bodies v;ere set Wth their leading edges 2.5" from the end of the works section. The wedge angle of 72.12~ was chosen by designing centrebo6y C so that the diffuser free area at its trailing edge eras the s3me as that at its leading edge. Centrebody D was the longest centrebody that ccjuld be satis- factorily supported in the diffiser. The pSes=re ratios meamred vlith the four centrebodies are shortn in Fig.4. These measurements rlere made at a Z?eynoEis rwnber, based on tunnel height of zbout 2.0 x '106. These results sho~thd large gains in performaze can be achieved by using a long single wedge centrebody forming a roughly constant area passage in the diffiser, The addition of a tail. wedge appears to have no large effect on the performance,

For the second series of tests the centrebody vias removed and the sngle between the walls was reduced over the first 14” of the ciiff'dser. Dil'%se:rs were made with equivCalent conical semi-angles of this entry length of O”, 0028', 0~56’ ad l”37f. Xt the end of the entry length dl the diffisers expanded to the same exit cross -section area as the bask dif?%ser. A fkixed convergirq entry was not considered practical for a difi‘user requires to

operate at low supersorLic speeds because of the risk of choking. The effect of these modifications on the pressure ratio is shown in Fig mea&surements vjere &so made at a ricynolds number of 2.0 x IO 6

5. These . It appezs

from these results that decreasing the expanzion rate aver the entry length improves ti?e efficiency at Mach numbers of '1.6 and above, the m,urimum efficiency being achieved vLth a constant ,area entry. At lower Mach numbers there is an optimum expansion rate, dependent on Jk.ch number, for minimum pres.sdre ratio,

Comparing the results o f these two series of tests, the long single vzdge centrebody modification gives a slightly better perform,ance at the higher Mach numbers than the best modification to the diffkser malls only, and also avoids the penalty of a loss in efficiency at low Mach mbers.

4 The Design of the "nalance Section with &adrmt

The basic steel shell, and the model support quadrant for the second. balance section had been designed end inrere being manufactured at the time when the 4” x 4" tunnel diffiser tests were ma&e, but the design was such that ply-v~ood walls were to be fitted to the shell to form the interior faces of the diff'user. As a resd.t of the 4' x 4” tunnel tests it was decide5 that this section should be fitted v&th a long single wedge centrebofiy -idth the q.mZmnt forming its leading edge. To reduce the risk of a loss in efficiency at lam Mach numbers due to choking, it was decided to pmvic2e a slightly expanding passage up to the quadrant. A di3grcjmmatic layout of this balance section is sho?;n in Fig .6. The top a.nd bottom walls are straight, but the sidewalls are kinked out to make the ratio of the diftiser free area at the shoulder of the qudrant to the working section arc3a 1.05, the equivalent conical semi-angle being O"43' for this part of the dif- fuser. The wedge faking provides the same free ra.rea Et the do~instre~ end of the section <as at the leading edge of the g~~.dr&, EviCiing the di.ff?user into two appz0ximatel.y constant ;31'ea passages.

-49

5 Tunnel Performance IIeamrements

The tunnel. performance nas been measured with both balance sections, snd the results are shorn in Fig.7 as the variation of pressure ratio with Mwh mmber. The measurexnsnts were made at a Reynolds number, based on tmel height of about I2 x 106, The pressure ratio shown is the minimum pressure ratio required to achieve or msints3.n fully supersonic flow 3.n the working sectzion. The minimum pressure ratio has been determined both with in~m~ting and d&a&g pressure ratio, and no evidence of hysteresis found, The pressure ratios shown for the bGLa3ce sectionwith quadrant were obtained with a conical. f&ring as shown in Fig.6, in place of amodel and sting. A brief check has shown that with a model and sting at zero incidence, the mininxm~ pressure ratios are reduced by about 0.01.

The rescllts in Fig.7 show that a reduction in pI%ssure ratio of 0.28 has been achieved at aMach number of 2.00 by the USC: of the balance section with quadrant, At the lowest Mach number, l&-2, at which both sections ware tested, the balame section with quadrant required a pressure ratio O,C& higher than the calibration section. .

Some idea of the effeotive blockge oaused by the quadrant and wedge was obtained by measurhg the choking Mach rarmber when the tunnel was YWI empty tith the flat subsonic nozzle. Urrier these conditions choking occurred not in the diffuser, but upstream of the working section, pre- sumably due to ovemorrectior of the nozzle blocks for boundary layer growth, and it was found possible to obtain a working section Mach number Of 1.07.

6 Conclusions

From the results of tests in a 4” x 4” supersorxio tunnel it has been found that the efficiency of a straight w&ed divergent diftiser, operating at Mach numbers in the range 'I.4 to 2.0 can be great:Ly improved by fitting a long single wedge centrebody to form a whly constant area passage along the first diffuser stage, Slightly snaller gains in efficiency were achieved by reducing the angle betvmen the walls at the diffuser entry.

A "bslence section" (first diffuser stage) fitted with a quadrant type of model support system, and a long single vJedge centrebody, has been designed for the 3 f% x 3 f% supersonic tunnel. Using this improved diffuser the pressure ratio needed tc run-the tunnel at aMach number of 2.0 is reduced from Z,O.&, with the earlier balance section with csJ.ibraticn gear, to 1.76.

& Author

I Morris, D, E.

Title, eta.

CXLibration of the flow in the wtzking section of the 3 ft x 3 f't tunnel. National Aezonautioal Establishment, C,P, 261, September lYf9.t..

UT.2078.CP.308.K3 - Printed in Great Britain.

GROUND FLOOR PLAN

LONGITUDINAL SECTION ON LINE A-A 0 10 .o 60 w PO s 00

CROSS SECTION ON B-B

VIEW IN DIRECTION OF C-C

I i I I I I I SCALE I* FEET FIG. 1. GA. OF TUNNEL CIRCUIT & AUXILIARIES -3’x3’ SUPERSONIC WIND TUNNEL.

X BALANCE SECTION WITH CALIBRATION GEAR

@ BALANCE SECTION WITH QUADRANT

I I I

1*0 I-2 l-4 I-6 1-B 2-o

MACH NUMBER

FIG.7. TUNNEL PERFORMANCE MEASUREMENTS. 3FT. X 3 FT. TUNNEL.

t

144”

QBL U//////,/,/y/ / ////////,,, ////// //// /////

I

‘00, rrl +

I I II

I \I I /Af/////’ ///// ///

/w//y”/“/ f7/

CALIBRATION GEAR /

BALANCE SECTtON SIDE ELEVATION

AND SADDLE

3FFLlSER FREE i APPROX. CONS1 c ,,, ////

\ WORKING SECTION END ELEVAT IO N

SCALE

SECTION VIEW ON A A

FIG.2. DIAGRAMMATIC LAYOUT OF BALANCE SECTION WITH CALI BRATION GEAR.

t- 4.67” / 4=67’ / 4~67’ -1

I C ’ 1’42’ 7’42’ ’ I

LEADING EDGE

CENTRE BODY A

CENTRE BODY 6

- I 7”42’ LEADING EDGE

CWTRE. BODY C

I L

I 17”42’ 8”s’ ( _I

LEADING EDGE

CENTREBODY D

FIG.3. DETAILS OF CENTRE BODIES USED FOR TESTS IN 4” X 4” TUNNEL.

PLAN VIEW OF DIFFUSER WlTH CENTREBODY C

0 BASIC DIFFUSER

V CENTRE00DY A (SEE F4.3)

h CENTREBODY B (SEE F4.3)

0 CENTRE BODY C (SEE FI q.3)

X CENTREl3ODY 0 (SElE F&3) I

I*0 I-2 I*4 l-6 I.8 210

MACH NUMBER

FIG.4. EFFECT OF CENTREBODIES ON DI FFUSER PERFORMANCE 4” x 4” TUNNEL.

M= 1699

M=l*79

I*4

I*2

\-0

M= I.39

--e CORRESPONDIN(i VALUES FOR BASIC I DIFFUSER WlTH CENTRE BODY ‘C’ L’w

I I z1 mu

O0 I' 2’ 3* EQUIVALENT CONICAL SEMI- ANGLE OF ENTRY LENGTH

PLAN VIEW OF DIFFUSER WITH lo 37’ ENTRYAN@E

FIG.5 EFFECT OF EQUIVALENT CONICAL SEMI-ANGLE ON DIFFUSER PERFORMANCE

4”x 4’ TUNNEL.

t- i4V

NOBILE ‘DlFFU5ER

\

SIDE EtEVATION

f /////,

///////////,,/,/,,, ////////,/,, ,,,,,

I ///////l/l

= P

>I I WEDGE’ FAIRING QUADRhT 1 \I I A

///// /m \\y////////“““““‘~ /////////////“““~”

SECTION VIEW ON AA

WORKING SECTION

/

/

7-l O &I I- l

%

1 \\\\\\\

END ELEVATION

SCALE

FiG.6. DIAGRAMMATIC LAYOUT OF BALANCE SECTION WITH QUADRANT.

I I x BALANCE SECTION WITH CALl6RATIW qEAR

IT! @ BALANCE SECTION WITH QUADRANT

I - I

-

I.2 ID4 I.6 I*8 2-O

MACH NUMBER

FI G.7. TUNNEL PERFORMANCE MEASUREMENTS. 3FT’a X 3 FT. TUNNEL.

A.R.C.Technrcal Report

PRINTED IN GREAT BRITAIN

S.Q. Code No. 23-9010-08