UPkJASH FLOWJ FIELDS..(U) GRUMMAN AEROSPACE CORP BETHPAGENY RESEARCH AND DEVELOPMENT C.. B GILBERT MAV 83

UNCLASSIFIED RE-667 RFOSR-TR-84 -88? F49608-82-C-0025 F/G 20/4 NLEEEEEEEEEI EIIEEEElhEEEllEEIElEEEEEEEE

'L1.

11111 1.0 2L3 63

L 36 III

MICROCOPY RESOLUJTION TEST CHARTNATIONAL BUREAU OF STANDARDS-1I963-A

n.

%.

AIFOSR-TR- 8 4 - 0 7

Grumman Research and Development Center Report RE- 667

AN INVESTIGATION OF TURBULENCE MECHANISMS IN V/STOL UPWASH FLOW FIELDS

-' by

Barry 1. Gilbert* Aerosciences

Research and Development CenterGrummnan Aerospace Corporation

Bethpage, New York 11714

* ym4submitted to

* lt'.Dr. James D. WilsonExternal Aerodynamics and Fluid Mechanics

'.C Office of Scientific ResearchU. S. Air Force

First Annual Report on Contract F49620-82-C-0025

May 1983

DTICfELECTEFEB 13 1984j

B

LLJ~ for p.blie rel88sk;

jfISTBUTION -STATEMENT AP. C.* 2AppovedforPublic f916""e

F Distribution Unlimited

84 02 10 A0O

.-SECURITY CLfML~b M AGE ("n Dato.Entered)-.RREPORT DOCUMENTATION PAGE READINSTRUCTIONS

REPOT DOUIMNTA ION AGEBEFORE: COMPLETING FORM4

I. REPORT NUMBER 2. GOVT ACCESSION NO 3. RECIPIENT'S CATALOG NUMBER

I .X)OSR.1T{. 84-0087 P,. ",'4. TITLE (and Subtitle) S. TYPE OF REPORT & PERIOD COVERED

An Investigation of Turbulence Mechanisms in AnnualV/STOL Upwash Flow Fields March 1982 - March 1983

6. PERFORMING OIG. REPORT NUMBER

7. AUTHOR(s) S. CONTRACT OR GRANT NUMBER(s)

" Barry Gilbert F49620-82-C-0025

9. PERFORMING ORGANIZATION NAME AND ADDRESS 10. PROGRAM ELEMENT. PROJECT. TASK

AREA & WORK UNIT NUMBERS

1 G10-1 FSouth Oyster Bay Road 'A 3o") 1 1<

,Bethpage, New York 11714I. CONTROLLING OFFICE NAME AND ADDRESS 12. REPORT DATE

USAF, AFSC May 1983Air Force Office of Scientific Research/A/' 13. NUMBER OF PAGES

Building 410, Bolling AFB. DC 20332 3514. MONITORING AGENCY NAME & ADDRESS(I different from Controllin# Office; IS. SECURITY CLASS. (of this report)

Unclassified

IS.. DECLASSI FICATION/DOWNGRADING

SCHEDULE

16. DISTRIBUTION STATEMENT (of this Report)

,., . ... -Li --or a1LOae;I DISTRIBUTION STATEMENT Adistribution uniuilt ed Approved for public ieIe l 4 ,

Distribution Unlimited

17. DISTRIBUTION STATEMENT (of the ebetract entered in Block 20, it different from Report)

%C

19. SUPPLEMENTARY NOTES

IS. KEY WORDS (Continue an reverse side it necessary and Identify by block number)

Turbulence, V/STOL Upwash, Fluid Dynamics

'.

20. ABSTRACT (Continue an reverse side if neceseea end identify by block number)

-The results of the first year of an experimental investigation of theabnormally high turbulence level and mixing layer growth rate characteristicsfound in the upwash regions of V/STOL flows in ground effect are presented.The overall objectives of this program were to examine the origin of theincreased fluctuations, to systematically characterize the development andstructure of the upwash, and to determine the parameters that influence thesecharacterisicts. The approach adopted was to investigate the fundamentalturbulent V/STOL upwash mechanisms in increasinqly more comDlex flow

DD J 1473 EDITION OF I NOV 6S IS OBSOLETE

SECURITY CLASSIFICATION OF THIS PAGE (When Date Entered)

.* . . .

Q , -, ,% - , . % * . 1 , - . - * • , "- " ." . ° " ,

SECURITY CLASSIFICATION OF THIS PAGE(Whai Data Entered)

configurations.. In the first year'.s effort, a two-dimensional upwash wasformed by the collision of opposed two-dimensional wall jets. Extensivemeasurements were made in the two-dimensional wall jet to establish thestarting conditions of the upwash. Evaluation of these measurements haveshown classical wall jet behavior, and by the time the wall jet reaches thecollision zone, both the mean and turbulence profiles are fully developed.

."A unique set of velocity profiles were obtained at six locations in the

upwash. f Two components of the velocity were found simultaneously. using anX -probe-"nmometer. This baseline set of two component velocity profileshas never been reported before. While the turbulence levels and mixinglayer growth rates were larger than those found in a free two-dimensionaljet, these values were less than those reported by previous investigators.

- .

-

Accession ForNTIS GRA&I

DTIC TAB

Unannounced 1JJustificatio

ByDist ribution/

Availability Codes

lAvail and/or'Dist Special

-I I'' UIi

..- n j.55:rU L 14

SECURITY CLASSIFICATION OF T, PAGE(WWhen Daa Ener.d)

%

* ,_ . . _.-. r

Gruman Research and Development Center Report RE-667

AN INVESTIGATION OF TURBULENCE MECHANISMS IN V/STOL UPWASH FLOW FIELDS

by

Barry L. GilbertAerosciences

Research and Development CenterruGruan Aerospace CorporationBethpage, New York 11714

submitted to

Dr. James 0. WilsonExternal Aerodynamics and Fluid Mechanics

Office of Scientific ResearchU. S. Air Force

First Annual Report on Contract F49620-82-C-0025

May 1983

Approved by:Wichard A( Sce n

P-- 0rector, &Cne

-?SC)AIR ?',fOTTCI " - I is

Thi.

8P• , -i-

%..,,. . -

,:% T..-a '• " oa. + ,nror emt-ir wivis4- •

ABSTRACT

The results of the first year of an experimental investigation of the

abnormally high turbulence level and mixing la) r growth rate characteristics

found in the upwash regions of V/STOL flows in ground effect are presented.

The overall objectives of this program were to examine the origin of the

increased fluctuations, to systematically characterize the development and

structure of the upwash, and to determine the parameters that influence these

characteristics. The approach adopted was to investigate the fundamental

turbulent V/STOL upwash mechanisms in increasingly more complex flow

configurations. In the first year's effort, a two-dimensional upwash was

formed by the collision of opposed two-dimensional wall jets. Extensive

measurements were made in the two-dimensional wall jet to establish thestarting conditions of the upwash. Evaluation of these measurements have

shown classical wall jet behavior, and by the time the wall jet reaches the

collision zone, both the mean and turbulence profiles are fully developed.

A unique set of velocity profiles were obtained at six locations in theupwash. Two components of the velocity were found simultaneously using an X-

probe anemometer. This baseline set of two component velocity profiles hasnever been reported before. While the turbulence levels and mixing layer

growth rates were larger than those found in a free two-dimensional jet, these

values were less than those reported by previous investigators.

RESEARCH OBJECTIVES

The development of aircraft with vertical/short take-off and landing

(V/STOL) capability has led to a requirement for understanding the unique

turbulence phenomena encountered in the interaction of lift jets with the

" ground. When a V/STOL aircraft is in ground effect (IGE), the exhaust from* the aircraft lift jets interacts with the ground, producing an upwash flow

directed towards the underside of the aircraft. Two characteristics of the.

- upwash flow that make its behavior very difficult to analyze are an abnormally

*: high turbulence level and a much greater mixing layer (fan width) growth rate

compared to other types of turbulent flows (Ref. 1,2).

Although a number of investigations of overall flow in ground effect

'1

have been carried out, measurements in these highly unsteady flows are very

difficult, and interpretations of these measurements vary widely (Ref. 3-7).

The problem is made computationally difficult, by the intrinsic three-

dimensionality of the upwash. However, even when these difficulties are

overcome, the numerical codes require better definition of the turbulent

structure in order to make reliable predictions of the fountain flow andlater, the fountain/ aircraft interaction.

*. . Previous investigations have attempted to study the full V/STOL flow

field with its full geometric complexity. Some of these have even made

measurements with an aircraft planform. These are configuration specific

studies that necessarily miss the fundamental flow characteristics. Our

approach was to employ a simple two-dimensional flow configuration. In thisconfiguration, the complex V/STOL upwash flow geometry was simplified. The

lifting jet impingement region with the ground has been eliminated. The

radially spreading wall jets were replaced by the much simpler two-dimensional

wall jets. This part of the study had the goal of producing a baseline data

set showing the high turbulence level and the increased mixing rate. In

4addition, during this part of the study, the upwash turbulence structure was

examined in finer detail than ever reported before.

The first major research objective for the first year's effort was the

." design and construction of the experimental apparatus used to produce the two-

*dimensional upwash. After the facility was running and sufficient

measurements were obtained to assure two dimensionality and uniformity of the

* exit profiles, detailed measurements of the wall jet profiles were to be

obtained. These measurements are very important since these two-dimensional

wall jets represent the initial flow conditions into the formation region of

4the upwash. Finally, a single wire anemometer was to be used to measure one

component mean and turbulence profiles at six locations in the upwash. These

would form a comparison set of data to the relatively small sample of upwash

measurements that exist in the current literature. This data set would then

be repeated using an X-probe anemometer to measure two component velocities,

resultinq in new measurements never before appearing in the literature.

V.

,.-.'.

SIGNIFICANT ACCOMPLISHMENTS

In this section the significant accomplishments and progress made during

the first year's research effort will be described. In summary, the test

facility is functioning well and a basic set of calibration exit profile data

have been taken. Wall jet profiles were obtained at 20 locations from the jet

exit to the facility centerline. These surveys showed the rapid development

of the mean and turbulence similarity profiles. They also exhibited the wellestablished mixing layer growth rate and mean velocity decay rate thatcharacterize wall jets. Careful measurements were made at six heights in the

upwash using both a single hot film probe and an X-hot film probe The

expected abnormally high mixing layer growth rate was found in the two-dimensional upwash flow. However, the turbulent intensity was of the sameorder as is found in ordinary two-dimensional free jets. There is basic

agreement between our data and prior studies. Although there is some

disagreement in specific details such as spread rate, our values are in the

range of those of others. Our set of carefully generated data from a well

defined two-dimensional source shows symmetry of the turbulence energy

profiles In the upwash, data not given by others. This baseline upwash data

show mean velocity decay and spread rate trends required by conservation

considerations. These trends have eluded some earlier investigators. Data

interpretation difficulties due to directional ambiguity of the hot film

sensor were somewhat resolved when the profile measurements were repeated

using an X-probe. These X-probe measurements have shown, for the first time,

the cross component mean velocity in the upwash. In addition, the turbulence

profiles for both components were obtained. Finally, one component of the

Reynolds stress was measured as a sign of the accuracy of the measurement

technique, these cross component data show remarkable symmetry. In the

remainder of this section, these accomplishments will be detailed.

The wind tunnel facility designed and constructed for the first year's

effort is diagrammed in Fig. 1, and the test section is shown in Fig. 2. Each

of two independent fans drives the flow through identical, 90 cm long, 26.50half angle, subdivided diffusers. The plenum chamber has honeycomb, three

sets of 16 mesh/inch screens and a gentle balsa contraction. The nozzle

section employs a symmetric ASME long nozzle contraction to 10 cm height

followed by an asymmetric lemniscate curve contraction to a 1 cm high exit.

.* .* * * * * 4 . . .*J.-.*.*...*.

LL

,I d

'LCa

* 0- I

4

U

NA.No

,,:... .8 2

• -- 2

8- W2,2 .2

, I

I.-

,0 , - . .;. , ,, . ,, , ., . ,, . , ,, -,. . ,. .,.,. ... . .. , -. -.: . .. . .. . .. . , . ,. -.-.

iz

0-~

a. 06

7 fir.

t;

1. u

00

CC

IC-

co0)I x2

p..

This nozzle geometry was carefully chosen to minimize the interference to the

entrainment flow over the top of the nozzle. The success of this design can

be seen in the nozzle exit profiles later. The exit aspect ratio is 50:1.

The test section has plexiglass side walls to aid in maintaining the two

dimensionality of the flow. The ground plane is instrumented with static

pressure taps connected to pressure transducers capable of high frequency

response. To facilitate the ease in understanding and in comparing the data,

a coordinate system was chosen that allows the X direction to be the mean

direction of the largest velocity component. That is, X tracks some

centerline streamline and Y is always perpendicular to it. This results in a

900 rotation of X from the wall jet to the upwash as shown in Fig. 3.

Therefore, for clarity, wall jet parameters are subscribed with w. U and u'

are the mean and fluctuation components in the X direction. V and v' are the

components in the Y direction.

The facility has the capability of producing two independently controlled

wall jets with flow rates that may be balanced up to an exit velocity of 67

m/s. Figure 4 shows hot film anemometry measurements of a typical exit plane

velocity profile taken vertically across the nozzle exit and includes the

entrained flow velocity over the top of the nozzle. There was negligible

variation in these exit profiles taken at various locations across both

nozzles. This flow is uniform to 0.5% across the 50 cm long dimension

excluding the regions near the ends of the nozzle. Disregarding the boundary

layer, the mean velocity is uniform to 0.75% with turbulent intensities u'/U

of about 0.6%. The single jet external entrainment velocity increases from

about 6.6% of the mean exit velocity to 9.7% when both wall jets are used to

form an upwash. The instrumentation plate is 84 cm long (nozzle to nozzle).

The first curve in Fig. 5 shows the mean velocity decay from a single

nozzle to a position past the plate centerline. This trace was taken at a

single height (Yw/exit height D - 0.5), and so although it is not a local

maximum decay profile, it is similar to it, and certainly the qualitative

information is correct. The second curve shows the mean velocity at the same

height using both jets. There is a definite problem with data interpretation

of velocity direction in the interference zone. Since these data were

obtained with a single wire constant temperature anemometer, with the sensor

perpendicular to the page, it measures the velocity component normal to its

6

- , ,, * %".l- o.

* ,° ,' .% , ,'°°,, -,,- . . , - . .,- . .- * .* , . S. .-.. . .- ,

Aor-

> 0CJ J D

ix I

2U

I

92.5

.1~- 2.2Sa..., .SINGLE JET FLOW

* . PLENUM VELOCITY

1.75S 65.0 m/s

* Ito 1.50

1.25

Lu 1.00

0.7

0.50

9, 0.25

00 0.12 0.24 0.36 0.48 0.60 0.72 0.84 0.96 1.08 1.20

MEAN VELOCITY, U/U.jet

2.501

2.25 . SINGLE JET FLOW

2.00 4.PLENUM VELOCITY

1.75 66 s.0 rn/i

.1.50

X z 1 .2 5

X 1.00-

* 0.7 f

21 j__0 0.03 0.06 0.09 0.12 0.15 0.18 0.21 0.24 0.27 0.30

TURBULENCE INTENSITY. u'/I 0c@I

R834g2-004PP

Fig. 4 Typical Jet Exit Velocity Profiles

-r-1;

L

-' .'%.

j-o t.t0 -.- rr- --

0.880 SINGLE WALLJET DECAY

. 0.770-- -

0-

•0.440, Z TWO WALL JET

1. / INTERFERENCE EFFECT

o.o °4

,[, • 0.330

i 0.110

. ' -,o .i o o , ,__ _ _ _ _ _ _ _ _ -

-12 6 0 6 12 1s 24 30 36 42

DISTANCE FROM CENTER LINE, Xw/OwR$3-0892-05PP: "'2365"12(T)

Fig. 5 Mean Velocity Decay Trace at Yw/Dw, 0.5 with a Single Wall Jet and with Colliding Wall Jets

F

S[4 4.' I

-.°

axis, that is, the component in the plane of the page composed of the vector

sum of U + V. There are two features of these curves that are important.

First, the effect of the second jet is confined f- the collision zone as shown

by the unaffected decay tract from the jet exit to the zone. Second, as will

be shown in the next set of plots, the wall jet layer would be of the order of

- four nozzle heights at the centerline if the second jet was not running. One

would expect the collision zone to be twice this wide (consistent with Kind &

Suthanthiran (Ref. 3)). Figure 5 shows the collison zone to be of that order.

Wall jet mean and turbulence profiles were taken at 20 locations from the

jet exit nozzle to the instrumentation plate centerline. These profiles were

made at equal distances along the plate in increments of approximately two

nozzle heights. Each profile contains 24 data points. The measured variation

in probe position above the plate from the first to the last profile position

was less than 0.005 in. The data acquisition and positioning of the single

element hot film probe were accomplished under the total control of the

automatic digital data system.

Figure 6 shows only the 10 alternate mean velocity profiles. The data*1.m are normalized by the source jet velocity and the initial nozzle height Ow.

From these data, using the maximum velocity, the wall jet "half velocity

heights," Bw, were obtained by linear interpolation. The half velocity height

is the usual length scale used to characterize wall jets. It is the height

(above the maximum velocity point) where the mean velocity is half the maximum

velocity.

- -'.--A plot of the wall jet growth rate as characterized by the half velocity

height vs the distance downstream is given in Fig. 7a. A linear least squares

curve fit of the data from station 6 through 20 (Xw/Dw > 10) gives a growth

rate of 0.0728. This is exactly the growth rate established as the "correct"

value for self-preserved two-dimensional wall jets on plane surfaces at the

1980-81 AFOSR-HTTM Stanford conference (Ref. 8) of 0.073 t .002. The first

five stations were eliminated from the curve fit because they appeared to be

in the development region. Figure 7b shows the linear decay of the maximum

velocity squared vs distance. This relationship is required by conservation

of momentum considerations. Normalizing the data on Fig. 6 by the

@9 characteristic half height dimension and replotting this as Fig. 8 shows that

the mean velocity similarity exists as early as Xw/Dw = 10, much sooner than

Z 10

" " I' . . . " "- . .. . . . . . . ...-

6= % % ., ' (. ' ." % " . " - . . "- . . "w - . - -o - . . ...

-4 . ,4. . , ., -, . ., .: , , ., , . - , ..., ., , . -, --, , --., -. .. . . . . . . . - .., -

8.00

X 5.52

6.00 -- 9.72

18.16Qa6.00

LU4.00

C. 13~.00 - --

1.00 -

0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 1.10

0.00 MEAN VELOCITY, U1UJET

8.00

7.00 xwO

UMU ~22.25

GD -- 30.61

.... 38.92

4.00

4.00

S2.00

1.00

1.000.0

0.00W0.0 .0 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.0 10 10

0.00 0.10MEAN VELOCITY, U1UE

R 53 - 692 - OGPPFig. S W 811 J t M e n V elocity Profiles

0 **. . . .

4.00

3a) ' GROWTH RATE PROFILE

3.5

'~3.00

°" .

-.. -.

zo

IG CURVE FIT

-4-2.00 y 0.651 + 0.0728x

-J1.50

3 0 0

1.00

0.00 4.00 8.00 12.00 16.00 20.00 24.00 28.00 32.00 36.00 40.00

DISTANCE DOWNSTREAM OF JET EXIT, XwIDvv

3.25

3.00 b) VELOCITY DECAY PROFILEci4

S2.75

2.25

"'2.00

1.73 CURVE FIT.75.35+0.065x

0 1.50

z 1.25

2 1.00 o * a

0.75

0.500.00 4.00 8.00 12.00 16.00 20.00 24.00 28.00 32.00 36.00 40.00

r DISTANCE DOWNSTREAM OF JET EXIT, Xw/DwRaa-011,2.oo7PP

Fig. 7 Two-dimensional Wall Jet Characteristics

12-'.-- .- ".-

5.500-

4.500-

4.0

3.500

4~~ 3.00

2.000.

1.000.

0.000.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00

MEAN VELOCITY, U/UMax

Fig. 8 WWIt Jet Mean Velocity Profiles in Similarity Form (See Fig. 6 For Symbol Key)

4.13

.. J%

the 50 slot heights quoted at Stanford.

The one-component turbulent intensity data are shown on Fig. 9 at the

same 10 locations and normalized the same way as in Fig. 6. Figure 10 shows

. the same turbulence profiles normalized by the half velocity width. These

show similarity at Xw/Dw about 20.

Using the results shown on Fig. 6 through 10, the wall jet

characteristics at the centerline of the collision zone may be determined.

From Fig. 5 it was noted that the collision process is a very local

phenomenon. The wall jet parameters when no collision occurs should be used

to normalize upwash data in a manner similar to using the wall jet nozzle

height as an initial characteristic dimension. At the centerline, the wallEl. jet half height is 8/0 w a 3.702 - D for the upwash. Further, Umax/U jet =

0.571 at the centerline.

The next set of figures (Fig. 11) shows velocity profiles in the upwash

at six heights from 1.35 0 to 8.10 D in 1.35 0 increments. These data were

4-< obtained by positioning the single element hot film probe at a point and

waiting five integration periods before taking the measurement. Each

measurement is the average of 10 statistically independent *samples, each

having been analog filtered with a time constant sufficient to assure

reproducibility. Each profile is composed of 150 positions with a maximum

spacing of 5 mm in the tails and a minimum of about 1 mm near the center of

the upwash. The full trace is about 50 cm long. The profiles have been

shifted somewhat arbitrarily to an estimated symmetry point, and profile half

widths were determined crudely by finding the positions of half maximum

velocity.

There is a great deal of information that can be obtained from these

data. The residual velocities In the tails re similar to other studies (Refs

* 3-7). This flow in the tails is the entrainment flow towards the center,

perpendicular to the upwash direction. This has been verified by smoke flow

visualization studies. Of course, a single wire perpendicular to the page

cannot distinguish these velocity components. The mean velocity profiles are

symmetric, and beyond H/D = 2.70, the turbulence profiles have symmetric

peaks. Kotansky does not give these turbulent profile data (Ref. 5); Witze

(Ref. 4) and Foley (Ref. 6) show only one-sided turbulence measurements, that

is, they do not show the symmetric data; only Kind and Suthanthiran (Ref. 3)

14

,-* .... ** . . . - * . - *

400 Ir -F 4- w%.

7.50

7.006.50

6.005.50

450

4.00 I'*.3.50

2 .00 -

- W 2~~.50 - ---

1.00-

0.50 .-

0.00 _

0.50 .- - - --

0.00

8.00 U

5.507.00

6.50

6.00 -

5.50 -

4.00

3.503.00

0.00

0.0 .0 02 .03 .04 .05 .06 .07 .08 .09 .10 .11 .12 .13 .14 .15 .16 .17 .18 .19 .20

Fig 9 allJetTurbulence Intensity Profiles (Se. Fig. 6 for Symbol Key)

15

-N e. e

6.00.

V..5

10

9 0.o . 2 0 0 . 60

1 . 80 2

3: 3.00 4

T F U 6 CE~ T N 4T .u U

2 O-Wtfut

r fl' nsm lrF r SeFi.6 F rS m E lK y

.J.1

0.0 16o K V

of ...

r ..

e Fig. 6 Fo.S

.4- je

.171

0.50 0.70

0.45 X/D - 1.35 0.63

0.40 0 - MEAN VELOCITY 0.56PROFILE "

0.35 _ o 0.49- TURBULENT C a

0.30 INTENSITY t i0.42PROFILE a a

U/U.- 0.25 a 035u /U

0.15 a 0.21

0.10 d \ k 0.140.05

uot a' Pl lti pa a o ac oo nm ft ""p!e -'- o~i am u n~ l.' 0.07

----------- i-0. 20 -15 -10 -5 0 5 10 15 20 25

DISTANCE FROM CENTER LINE, YID

0.5 X/ -2.70 0.63

0.40 MEAN VELOCITY0.56PROFaIaLEa a e 0.07

0.25 -315.10 .5 0i t 50.4192 2

0.45 TURBULENT . 0.UuPROFILE a, *

0.30 INTENSITY " " a " m 0.42U/U.e PROFILE . * 9 ' u7U .4

025 a a .. 0.35

0.20 0.28

0.15 0.21

0.10 ** ": 0.14

0.05 0%~ 0.07

-25 .20 15 .10 .5 0 5 10 15 20 25

DISTANCE FROM CENTER LINE, Y/D4o2.06,o11(213)PP

Fig. 11 Mean and Turbulent Intensity Profiles in the Upwash (Uo - 67.0 mis) (Sheet I of 3)

17

0.50 0.70

0.45 X/O - 4.05 0 63

0.40 (3 - MEAN VELOCITY

PR O FILE a0.49

TURBULENT 940.30 INTENSITY 0.420.30 PROFILE

U/U a u'IU0.25 0 a o3 0.35

0.20 a. 0.28

& .0.15 " *a 0.21

0.10 aaa IaatQuItp~a~~aa 0.140.05 'Ift /" Wt-b 0.07

-25 .20 -I -10 .5 0 10 1 20 25

DISTANCE FROM CENTER LINE, Y/D

0.50 , 0.70

0.45 X/O - 5.40 0.63

0.40 0 - MEAN VELOCITY 0.56PROFILE

0.35 <> - TURBULENT 0.49

INTENSITY "'4 0.420.30 PROFILE do 0.42

" UI 0.25- 0.35 U'/U

a la

0.20 0'' J ao 0.28

. ' qba, .,...:

0.10 .. % 0.14O.S ataoo a o oo Do 00° 0'

0.05 ,aaa' , 0.07

4° 0.25 -20 -15 .10 .5 0 5 10 15 20 25

DISTANCE FROM CENTER LINE, YIDM83-0892-011(2/3)PP

Fig. 11 Min and Turbulent Intensity Profiles in the Upwush (U t - 67.0 mls) (Sheet 2 of 3)

18

.' .

°o. .

0.50 0.70

0.45 X/D *6.75 0.63

0.40 [3 MEAN VELOCITY 0.56PROFILE

0.35 0- TURBULENT 0.49

0.301INTENSITY0.0PROFILE 0.42

UIU . u... iU. ..

jot 0 . 2 0.35

ogo

0.20 .... 0.28

* e.

0.15 P 0.21.. a 0.20. ." 0.07 , .

0.10 - . . 0.140.05 -- am ". a,, Dat...f

* 0.06

0-25 -20 -15 -10 -5 0 5 10 is 20 25

DISTANCE.FROM CENTER LINE. Y/O

0.50 0.70

0.45 XJO- 8.10 0.63

0.40 0 - MEAN VELOCITY 0.56PROFILE

*0.35 0.49- TURBULENT

INTENSITY0.30 PROFILE 0.42

U/U. a /U0.25

0.35

0 0 .3

0.20 to .W0 -0.2

0.15 - . 0.21

0.10) lj 0 0.140.05 a. 030 a'aso- W,,su .07

0-25 .20 .i .10 .5 0 5 10 15 20 25

DISTANCE FROM CENTER LINE, Y/O

A63.0892O11I(3/3)PP

Fig. II Mean and Turbulent Intensity Profiles in the Upwash (Uj - 67.0 m/s) (Sheet 3 of 3)

";....,-19

% %a'..

show the complete profiles and their data are not symmetric.

Figure 12 shows the spread rate and maximum velocity decay. The rate is

about 0.20, which doesn't agree with previously reported results. However, a

closer look at these other data show inconsistency and, in some cases, plotted

data disagree with written statements. We believe our data are correct. This

- is further supported by X-probe measurements shown next. The proper mean

velocity decay characteristic is shown for X/D greater than 2.70. This is the

* form for the mean velocity decay required by conservation of axial momentum in

the upwash, a characteristic not usually found by others. Between X/D = 1.35

*and 2.70, the mean velocity actually increases and the mixing width decreases

correspondingly. This is a strong indication that the extent of the collision

zone Is between 1.35 D and 2.70 0 consistent with decay trace shown in Fig. 5.

The upwash profiles were repeated using an X-probe hot film anemometer.

- This measurement technique is able, with proper electronic manipulation, to

sense two perpendicular instantaneous velocity components. Using this

* technique, two component mean velocity profiles at the same six heights as

-previously described were obtained by an automatic data acquisition process.

* These were mean velocity profiles in the direction of the upwash (X) and also

the mean component into the upwash (Y) as entrainment or as spreading. These

. data were obtained by digitizing 4000 data pairs in 1.5 sec from each wire of

the X-probe. In addition to the mean (average) value from the new time

series, the turbulent intensity (deviation) in each direction and one

component of the Reynolds stress (cross product) were also obtained.

The mean velocity components in the upwash direction are shown in Fig.

13. As was shown by the single element probe data, at the first height, X/D =

1.35, the data do not follow the trend of the data at the other five

*locations. 1his trend will be very obvious in successive plots.

Fiqure 14 shows the cross-stream mean velocity component. Looking at the

' data at higher stations shows the mean speading velocity to the right on the

riqht hand side and the syimmetric mean spreading velocity (negative values) to

the left, left of the center. These profiles are symmetric about the upwash

, velocity maximum. At X/D = 1.35, these trends are reversed. That is, for

* example, there is an inflow from the right of center. Remember that X= D

would be by definition the position where the wall jet velocity is half the

local wall jet maximum. For X slightly larger than 0, the wall jet still has

20

'_' -' *,-, ....--" *>-T - .*; *', :- . -° ,- . ." ***-'*s* .-.- * , *. ".' .- ...- .. . "- .-

10.000

9-000 a) SPREAD RATE PROFILEI

S8.000

,-7.000

.~,. 6.000CURVE FIT V -2.219+ 0.202 X

3.000

*2.00 0-3.7 O

0.000

0 3.5 7.0 10.5 14.0 17.5 21.0 24.5 28.0 31.5 35.0UPWASH HEIGHT. XID.

4.0

3.6 - b) MEAN VELOCITY DECAY PROFILE

2.6

2.4 UCURVE FIT Y I AM .0433 X

~2.0

S1.6

1.2

0.8 0 - 3.7D

0.4

00 3.5 7.0 10.5 14.0 17.5 21.0 24.5 28.0 31.5 35.0

UPWASH HEIGHT. X/D~

A4082-012PP

Fig. 12 Two-Dimensional Upwash Characteristics

21

0.55 ,

I. 0.50 0

"-"0.4S

A XI0.40 1 .35 U

.&. A 2.70 A>- 0.35 o 4 4.05 0

0.30

> 0.25

0.15 0

0.10 A.•

• oo"° <,,<,A% 00

0.0A

0.00,,-6.00 -6.00 -4.00 -3.00 -2.00 -1.00 0.00 1.00 2.00 3.00 4.00 5.00 6.00

CROSS - STREAM DISTANCE, Y/D

0.55

0.50

0.45 x X/

.0.40 .Ir 0Q( 6.7 X

>: ,* 0 6.7500

L 0.35 X4 00 8.10 0

800 +*.C

> 0.30oIV- 0.25 0

~0.20 ,* + 4.'00 t x 0A a

0.15 ## X O.

0.10 z. 0 0 XX XX 00.

0.05 O0 _ XX x

0.0 ,-4.00 -6.00 -4.00 -3.00 -2.00 -1.00 0.00 1.00 2.00 3.00 4.00 5.00 6.00

39-1P .CROSS - STREAM DISTANCE, Y/Dma134oao24131P

Fig. 13 Upwash Mean Velocity Profiles

S.22 ) o

-. C * * . . . .--.

0.10

) 0.06

- 0.06

w0.04

C0.02

-.?0.00

0.0lu

oj -0.02 a

o-0.041

~-0.10a

0.a.

0.08Zx a

-0.102

-x+' +

. 0 ~.0%U4 xa 0.01 +

*1W00

x U '~it 0.000

go +xX

M -0.01 A0 0 0 04 '

N *

2.00 3.0 '.00 5.00 6.00o 7.00 8.0 90 00 10 2.00 13.0 1400 15.00 16.00

CROSS-STREAM DISTANCE, V/D

lt$3.0692O14PP

Fig. 14 Upwash Cross-Stream Mean Velocity Profiles (See Fig. 13 for Symbol Key)

23

a velocity component towards the centerline, which appears as a cross upwash

- .direction component. This is a direct indication that X/D - 1.35 is still in

the interference collision zone.

Figures 15 and 16 show the component turbulent energy in the upwash and

relative turbulence energy in the cross-stream direction repectively. The

forms of these profiles are similar to those expected for two-dimensional free

p jet flows.

Preliminary evaluation of the turbulence levels found in the upwash shows

these values to be the same as those found in ordinary two-dimensional free

jet flows. This is contrary to statements made by Foley (Ref. 6) and Witze

(Ref. 4) that the turbulence intensity is a factor of three greater than the

free jet case. However, examination of their data indicates ordinary

levels. Only Kind and Suthanthiran (Ref. 3) show factors of three. Kotansky

(Ref. 5) shows no turbulence data at all. Local values of turbulence

mean, are on the order of 60%. This is well in excess of the range of

application of the small perturbation approximation used to evaluation

turbulence properties in thermo-anemometry. In addition, 4000 data pairs are

probably not a long enough sample for fully converged quantities. These

experiments will be repeated after some digital analysis shows a proper sample

length. With this in mind, these data may still be used.

Figure 17 shows one component of the Reynolds stress, uv. Again looking

to the higher stations first, across the center region the Reynolds stress

profiles are anti-symmetric about the centerline and the same magnitude on

either side. Since Reynolds stress measurements are particularly sensitive to

measurement techniques, these plots are also a good indication of the

precision of the entire experiment. Plotted as a correlation coefficient

normalized by the rms values, the scatter in the tails is due to normalizing

by successively smaller values. Again, at X/D = 1.35, the entire profile is

reversed.

The mean velocity profiles in the upwash direction were curve fit with aleast square curve of the form U - A + C exp (-(y - Yo )

2 )" This curve fit

gives the symmetry coordinate yo, the maximum velocity (A+C), and the standard

deviation S. Using the generally accepted definition of half velocity width,

B(U Umax/2) -1.177 S. It should be emphasized that this technique is far

24

. . .. . . . . . . . . ... .

0.09 *'

40V 0.08 *

w 0

Vz 0.07 0 A

a

I A ~, 0 A

~0.03l a cc ° °°q° oo

0.0)2 /ao

0.01 0 coo ac o

*G 0~m

,-2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.D0 11.00 12.00 13.00 .14.00 15.00 16.00

k "i ,; "CROSS -STREAM DISTANCE. Y/D

0.08

0.07 0. +00 ' + 0- 0

" '00 00.06 Oo 0XX x 0

>:a Na~.

0.05 . 0

O.O O0 x "''us +

U 0.04 0 * x

¢z 0 0 0W xx Xx 00.03 0 o

1- 0.02 0 0*x 0, . x X 0 0

0.01 t0 0

0.00 L nZOO 3.00 4.00 5.00 8.00 7.00 8.00 9.0O 10.00 11.0 12.00 13.00 14.00 15.00 16.00

CROSS-STREAM DISTANCE, Y/D

483-0892-OISPP

Figl. 15 Up was hk n Directon Turblenc Energy Pofiles (Soo Fig. 13 For Symbol Key)

90

0A eL

00", 0i00 0 00.0 .

I'.

4.00

3.50

3.00aa

2.50

~2.00A

11.50 * AA

IA * I. g AGi' _ " . .. *o . ."

A.," •0

1.00 oat

0 a A

U- A"

-.50 , , A A,

S0.00

2.00 3.00 4.00 S.00 6.00 7.00 8.00 9.0 10.00 11.00 12.00 13.00 14.00 15.00 16.00

CROSS-STREAM DISTANCE. Y/O

,4. 2.0 -

1.90

1.60 K x

1.40-

x X

1.20 - X

xX x X *.90 0 4 00 , 0 ,. 0 x . x

0 0 ,0 0 0.0 o 00 0 0 A N 0 " oGo 0 0x 1 0 0X

I6 00x X 0 ~Xa~ . *0 0

00 *x 460..: OOII'+OX'-"+X +.^)X00 0

.20

0.00I ,.

2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 11.00 12.00 13.00 14.00 15.00 16.00

CROSS-STREAM DISTANCE. Y/DA83-06C2-14IP

Fig. 16 Upwash Relative Turbulence Energy Profiles (See Fig. 13 for Symbol Key)

26

4..

.4.. . '- .].'-" '..,.- ... "-. -' . ' . " ' - -. -,, - - -" +

, , -- ... . . . . . . . ."

-4. ,

,. 0.60 - < ..

AL0.40 caBk 0* o

0.0aa 4 :0 0 a

0 0 o a~a a o " a oI0 II I

3 0.0 i, -tO

-0.20 a5 *, • * 0 B a

&0 .0 OL o noSA* -A.-o.-.40 a -

-.60 a I

2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 11.00 12.00 13.00 14.00 15.00 16.00

CROSS-STREAM DISTANCE, Y/O

| • .0.80 x

.060 x 0x 0 0

0.40 O 0 0

0.20 o #0

. i 0.0 Oo

-0.20 0 O6 ###x r_ .is" AulxoO Ooxx a0 0 0 o %4. t

-. 40 * w .+.O

++,X

-. 80 JI

2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 11.00 12.00 13.00 14.00 15.00 16.00

CROSS-STREAM DISTANCE, Y/D1483.0492.017

Fig. 17 Upwash Reynolds Stress Profiles (See Fig. 13 for Symbol Key)

27

%m°. .. . . . . . ..

/ * . . . . . . . * * . , .". .+ ' . . " " 5 " _ - ", ". • . . 1

2.40

2.20 Q) SPREAD RATE PROFILE

50

'. -. Y 0.583 +0.212 X

IA. 1.40

1.20

1.000.00 1.00 2.00 3.00) 4.00 5.00 6.00 7.00 5.00 9.00

UPWASH HEIGHT, XID

7.00

* 6.50 bC MEAN VELOCITY DECAY PROFILE

4 .00

Y-191..79

3.50

3.001.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 7.50 8.00 8.50

UPWASH HEIGHT. X/D

R63.0692041S

Fio. 18 Two.dimnsional Upwash Characteristics

28

'";/- ... '..............................,............. .............. . , . . ..,• .- " .

superior to the usual determination of half width. That procedure usually

entails finding Umax and interpolating between data points to determine B.

The method suffers severely from scatter in the data at both Uma x and

particularly at the half velocity point. Also, it rarely gives symmetric half

velocity positions. A least squares curve fit avoids these problems.

The results of the half velocity growth rate so derived are shown on Fig.

* 18a. As with the earlier one element probe data (Fig. 12), the upwash growth

rate is about 0.21 compared to values of about 0.37 reported by Witze, for

example. However, this value is still more than twice the free jet values.

Note, that the X/D - 1.35 data were not used in the curve fit. Finally, Fig.

16b shows the mean velocity decay relationship required by conservation

considerations.

One hypothesized mechanism to explain the large mixing layer growth rate

found in an upwash, is that a simple free jet is waving (Ref. 9). Our

turbulence measurements would tend to support this mechanism. A remaining

task is to try to detect this waving phenomenon by carefully tracking the

upwash boundary. This will be accomplished using two, two-component probes at

various locations in the upwash and measuring correlations. Probes on the

same side of the upwash should show consistent in-phase correlation, and

probes on either side should show consistent out-6f-phase correlations.

In the event that the waving structure does not exist, the increased

mixing and turbulence must be due to a structure formed by the collision

process (Ref. 10). This process will be carefully investigated by controlling

the flow regime of the wall jets forming the upwash. In a laminar flow, there

is no turbulent structure, therefore any structure found in the upwash must

originate in the collision process itself.

Another task will be an investigation of the effects of the wall jet

initial conditions, particularly along the stagnation line, on the upwash

F turbulence characteristics (Refs 6, 11, 12). The location of a small object

on the stagnation line may serve to stabilize the upwash waving if that is a

mechanism. Alternatively, the object will isolate the turbulent structures in

each wall jet from each other during the critical turning phase. These

experiments will employ a single wall jet flowing into "symmetry" planes of

various heights. After a baseline set of data is obtained, a second wall jet

will be added and the height of the dividing "symmetry" plane will be

successively shortened until the real two jet upwash jet flow is simulated.

29

,- - , , ,, *.. ,, . . . . . . . . * ' " . ... .

ft ., . _- . . . ., . , .,. • . - ' -. . .. -. .'. . - - - . o " '. - ". -

REFERENCES

i '1 Rajaratnam, N., Turbulent Jets, Developments in Water Science, 5. Elsevier

Scientific Publishing Co., Amsterdam, 1976.

2 Harsha, P.T. "Free Turbulent Mixing: A Critical Evaluating Theory and

Experiment," AEDC-TR-71-36, Feb 1971.

3 Kind, R.J., and Suthanthiran, K., "The Interaction of Two Opposing PlaneTurbulent Wall Jets," AIAA Paper No. 72-211, AIAA 10th Aerospace Sciences

Mtg, San Diego, CA, 17-19 Jan 1972.

4 Witze, P.O. and Dwyer, H.A., "Impinging Axisymmetric Turbulent Flows: The

Wall Jet, The Radial Jet and Opposing Free Jets," Symp on Turbulent Shear

Flows, Vol 1, 23-30 Apr 1977, Univ Park, PA.

5 Kotansky, D. R. and Glaze, L. W, "The Effects of Ground Wall-Jet

Characteristics on Fountain Upwash Flow Formation and Development," AIAA

Paper #81-1294. AIAA 14th Fluid & Plasma Dyn Conf 23-25 June, 1981, Palo

Alto, CA.

6 Foley, W.H. and Finley, 0. B., "Fountain Jet Turbulence," AIAA Paper #81-

1293, AIAA 14th Fluid & Plasma Dyn Conf 23-25 June 1981, Palo Alto, CA.

*'7 Jenkins, R. A. and Hill, W. G.,Jr., "Investigation of VTOL Upwash FlowsFormed by Two Impinging Jets," Grumman Research Department Report RE-548,

Nov 1977.

8 Kline, S.J. et al., ed, Proc of the 1980-81 AFOSR-HTTM-Stanford Conf on

Complex Turbulent Flows, Published by Stanford University, 1981.

9 deGortari, J.C. and Goldschmidt, V.W., "The Apparent Flapping Motion of a

Turbulent Plane Jet," J of Fluids Engrg, Vol 103, p 119-126, March 1981.

10 Gilbert, B.L., "Diffusion Mixing in Grid Turbulence without Mean Shear,"

,- JFM Vol 100, Part 2, 1980, pp 349-365.

11 Hill, W.G.,Jr. and Jenkins, R.C., "Effects of Nozzle Spacing on Ground4 Interference Forces for a Two-Jet V/STOL Aircraft," AIAA paper 79-1856,

AIA Aircraft Systems and Technology Conf, N.Y., NY, 20-22 Aug 1979.

*Also J of Aircraft, Vol 17, Sept 1980, pp 684-689.

ft.

..

• ° . .,.. . . . ..

*12 Hill, W. G., Jr., Jenkins, R. C. and Gilbert, B. L., "Effects of the

Initial Boundary Layer State on Turbulent Jet Mixing," AIAA J., Vol 14,No. 11, Nov 1976, pp 1513-1514.

S. .

.'.

k31

S . . . . . ...~

L ...

Publications from this contract

"Detailed Turbulence Measurements in a Two Dimensional Upwash," to be

,.-. presented at the AIAA 16th Fluid & Plasma Dynamics Conf, Danvers, MA, 12-14

"" July, 1983.

PERSONNEL

The research conducted during the first year has been carried out by the

staff of the Grumman Aerospace Corporation R&D Center. Dr. Barry Gilbert has

been the Principal Investigator, reporting to Mr. Vincent Calia, Head of the

Experimental Fluid Dynamics Laboratory, and Dr. Richard Oman, Director of

Aerosciences. Dr. Gilbert devoted approximately 80% of his time to this

contract. Some additional research support was provided by Mr. Richard C.

Jenkins, Senior Research Scientist.

* .3

, 32

:! ' ." .- **~*%~ * * *

. * * 5. * '* *** "* **°-. .

P T V ~ 7rrrrr rr r AT

610

f"A

I8

D6