visualisation of relaminar flows

7/24/2019 Visualisation of Relaminar flows

1/9

R. VISWANATH*, R. NARASIMHA AND A. PRABHU

1977; Revised on January 21, 1978

e phenomenon of releminariza.

y a variety of different mechanism s.

transition, Flow

fluid motionso much

continue to be) greeted with varying degrees of dis-

is even contended that such reversionor relaminarization, as it

order and so must be thermo-

The thermodynamic objection is of course not valid, as the systems claimed to be

if crystallisation is possible, so is reversion. Further-

such reverting flows have been studied in some detail in recent

e.g. a

recent survey by Narasimha'). Nevertheless, there appears to be a linger-

that whatever effects are observed cannot be considered to have resulted

the log law in wall flows) are often only asymptotically valid, and in particular may

It is the purpose of the present paper to describe some simple flow visualization experi-

159


2/9

160

. It . VISWA NA TH

et al.

of the three classes of reversion that have been distinguished by Narasimhail i

first of these, turbulence energy is dissipated by the action of a molecular

tr a

parameter, like viscosity or conductivity. For example, flow through a pip

e

enlarged from one diameter to another in a short divergence may revert to th

e

i

state downstream if the Reynolds number there drops to a sufficiently low vah

the second class, an external agency (like gravity) may work against turbulence e

and so destroy it ; an example is the suppression of turbulence in a stably

st

flow, such as an inversion in the atmosphere. This class may be called the Ri

c h

type, after the author of the first studies of inversions.

2

There is a third category of reversions in which the absolute value of the

t u r

stresses is not necessarily reduced everywhere but the mean momentum balance is

mined largely by the presence of an additional force which overwhelms the Rey

stress. This is what happens in a turbulent boundary layer subjected to severe acc

ration. The reversion observed here is only in part due to the two mechanis

me ntioned a bove ; the ma jor factor i s the dominat ion of pres sure forces over the s low

responding Reynolds stresses in an originally turbulent flow, accompanied by the g

ration of a new laminar sub-boundary layer stabilized by the favourable pre

gradient .

The experiments to be described below employ flow visualization by dye

i n

and colour schlieren photography in air. The

dyn mics

of the observed

p h e n

will not be discussed at length here, as this has been done elsewhere.

2 . Diss ipa t ive revers ion

This can be demonstrated most easily in a bifurcating pipesay across a

T

-

t u b e , a r r

as shown in

Fig.

1 a.

The division of the flow between the downstream

b r

of the T can be controlled by valves. The flow is kept turbulent in the

u p

branch by increasing the Reynolds number well above the critical ; in the presentexp

meats this Reynolds number was about 3,000 (based on average velocity and

p i

meter). By operating the valves, the downstream Reynolds number in

t h e s t

section of the T is reduced to about 500.

Fig . 1

b

is a photograph of the flow visualized using dye. The purple dye injec

upstream diffuses rapidly, showing turbulence in the flow; the green dye injected

30 diameters downstream of the bifurcation remains a clean, thin filament, sh

that the flow has reverted to the laminar state. (To avoid mixing of colours,

the p

graph was taken just before the time that the purple dye-injected upstream rea

the downstream injection station.)


3/9

VISUALIZATION OF RELAW NARIZING FLOWS

61

300

I

a.

Apparatus to demonstrate dissipative revezsion in bifuicating pipe flow.

rrangement consisting of a square tank with two

jet of dye issues from a vertical

he tank is filled with cold

ith a coil rated at 2 kW the

temperature rise near the free surface

In the absence of heating, one finds that the jet of dye first undergoes a transition

flow, and then mixes rapidly with the water in the tank

b),

all of which gets coloured in less than a minute. If the heating is switched

min noted above, the jet is also turned on (the velo-

by a pinch cock so that turbulence in the jet is seen clearly),

cloud ' layer near the top

here is very little activity in the cloud as the turbulence is suppressed ;

transition in the jet followed by reversion

c.

The phenomenon observed in this experiment has similarities with atmospheric inver-

row layer in which the temperature locally increases with

ng in


4/9

162

. R. VISWANATH

et at

-f---

h....,

P u r p l e e ty

2 k W

He ating coi l

Dye

cloud

Turbulel

et

Laminar

8 mm

1 4 ,

W att r

00 x 200x250 Imrn Tank

1 mm id.

tube

FIG.

2

a.

Experimental set up to demonstrate reversion in a stably stratified fluid.

4

Reversion by domination

. Here we demonstrate reversion in a supersonic turbulent boundary layer subjected

sudden acceleration at an expansion corner. A sharply boat-tailed backward facin

s t ep see

Fig. 3 a)

was chosen so that, in addition to reversion in the boundary lay

other interesting flow features may be seen.

This experiment was performed in a 7 in. x 5 in. (17 .5 cm x 12 .5 cm) superso

wind tunnel. The experimental conditions are summarised below :

Mach number

.5

Stagnation pressure

00 kPa (abs)

Boundary layer upstream of corner

urbulent

thickness

mm

Boat-tail (or expansion) angle

5 deg.

The expansion angle of 15 deg. was chosen

to

satisfy the criterion for reversion fou

by Narasimha and Viswartath.

3

A conventional two-mirror schlieren technique (d

cribed,

e.g.,

in ref. 4) was used for flow visualization, with colour filters in place of

knife edge. The filters were Kodak gelatin type, and were horizontal strips of r


5/9

;c

. 1

1 k

1 1 9

.'

I

-44

crin

t

4 .

.

1 % .

1 1 , 4

tar

:

. -

c

:P T

Aillg011r.

vtaikaa,4;

ar;C::

7 ...:3

ir . tin SY .

.

lis n

FIG.

lb

. Relaminarization in bifurcating pipe flow.

FIG.

2c. Reversion, with heating coil switched

no. 2b. Flow without heating.

n.


6/9

AG. 3b. Reiaminarization at an expansion corner in supersonic flow.

Fla. 4b. Reversion in a coiled tube.


7/9

VISUALIZATION OF RELAMINARIZING FLOWS

6 3

a p i a

0

blue

=

green

> 0, moderate

red

> 0, strong

white.

ep/ey

cause correspondingly large deflections of the light

Fig. 3

b

shows a schlieren photograph taken on Agfa colour sheet film of size

ith a mercury vapour lamp of 250 watts (although a white

see

also key in Fig. 3

a) :

P

ay

Turbulent

boundary lay

Blue CO

Green

Red

0, moderate

White >0

1

strong

Viscous

sub

layer

ip separation

shocic

Outer inviscid

rotational

layer

(Nevi) laminar sub-

boundary layer

Shear toyer

Recompression

waves

at reattachmeni

3 a.

Supersonic flow past a boat-tailed step, also sei ving as key to Fig.

3b.

growing

laminar (sub-) boundary layer (another White streak) downstream,

bove this, a thicker, weakly sheared layer (pale red) that is a

a s

the


8/9

300

nwn

' o

164

. R. VISWANATH

et

al.

5 .

Reversion in coiled tubes

This is a most interesting and spectacular phenomenon. It is easily demonstrated

a tube of some flexible, transparent material like tygon. The present experim ent

an 8 mm i.d. tygon tube, which has an initial straight section about 20 cm in l

e

is then curled around a cylinder of 1] cm diameter and has a final straight se

of 15 cm length

see

Fig. 4

a).

The fluid velocity is so adjusted that the Reyn

number in the initial section is about 3,000 and flow is turbulent. As Fig. 4

s h

F I G .

4a.

Apparatus to show reversion in a coiled tube.

the purple dye injected in this part of the tube mixes rapidly, indicating turbulen

However, the green dye injected at the fourth coil does not diffuse at all, showing th

the flow is laminar there. (To avoid mixing of colours, the photograph was again tak

just before the time that the purple dye injected upstream reached the fourth coi

Further dow nstream along the straight section, the green dye diffuses and the flow b ecom

turbulent again, but this occurs outside the field covered by the photograph.

Although this kind of effect was observed by Taylor

6

in 1 92 9, no detailed investigati

of reversion in coiled tubes appear to have been made yet. It is therefore not possib

yet to identify the mechanism unambiguously ; indeed it is likely that several medi

rusms are operating simultaneously in this case. First of all turbulence on a conv


9/9

VISUALIZATION OF RELAMINARIZING FLOWS

65

C onclusion

reversion from turbulent to laminar

Acknowledgement

e thank M r. N. S. A swathanarayana, Mr. H. Gopalakrishna and Mr. D. J. Thvaffarai

stance in c onducting the experimental w ork.

W e are grateful to Professor

assistance in providing colour

tpport of the work reported

.

NARA SIMHA, R.

.

RICHARDSON, L. F.

The three archetypes of relaminarization.

BSc Aero Report 77

FM

7, 1977;

also Proceedings 6th Canadian Congress of Applied

Mechanics,

1977.

The supply of energy from and to atmospheric eddies.

Proc.

Roy. Soc.,

1920, A97, 354-73.

. TAYLOR, G. I.

o BRADSHA W, P.

Reverse transition at an expansion corner in supersonic flow.

AIAA 1

1975, 13, 693

5 .

Elements of Gasdynamics,

1957, Ch. 6, 153-62, John Wiley and

Sons, Inc.

Effect of boat-tailing on two-dimensional tui buient base flews.

M.E. Dissertation Dept. of Aero. Engg. IlSc.

1969 .

The criterion for turbulence in curved pipes.

Proc. Roy. Soc.

1929, Al24, 243-49.

Effect of streamline curvature on turbulent flow.

Agardograph,

169. 1973.

.

N A RA S DA HA , R . A N D

VISWANATH, P. R.

.

LIEPMANN, H. W. AND

ROSHKO, A.

.

VISWANATH, P. R.

visualisation of relaminar flows

Documents