Adhikari - 2009 - The Impact of A Vortex Ring On A Porous Screen

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 27

Home Search Collections Journals About Contact us My IOPscience

The impact of a vortex ring on a porous screen

This article has been downloaded from IOPscience. Please scroll down to see the full text article.

2009 Fluid Dyn. Res. 41 051404

(http://iopscience.iop.org/1873-7005/41/5/051404)

View the table of contents for this issue, or go to the journal homepage for more

Download details:
IP Address: 128.59.62.83
The article was downloaded on 09/04/2013 at 06:51

Please note that terms and conditions apply.


IOP PUBLISHING FLUID DYNAMICS RESEARCH
Fluid Dyn. Res. 41 (2009) 051404 (26pp) doi:10.1088/0169-5983/41/5/051404

INVITED PAPER

The impact of a vortex ring on a porous screen


D Adhikari and T T Lim
Department of Mechanical Engineering, National University of Singapore, Singapore 117576,
Singapore
E-mail: mpelimtt@nus.edu.sg

Received 20 December 2007, in final form 22 November 2008


Published 2 September 2009
Online at stacks.iop.org/FDR/41/051404

Communicated by Y Fukumoto

Abstract
In this paper, we re-examine the issue raised by Morton (2000) regarding the
impact of a vortex ring on a porous screen. We show, through a systematic
experiment, that Morton’s observation can only occur at very low Reynolds
number or above a certain critical Reynolds number, which is dependent on
the porosity of the screen. In particular, we found that a primary vortex ring
of moderate Reynolds number impacting a porous screen can also undergo
vortex rebound similar to that observed in vortex ring–solid wall interaction.
At a higher Reynolds number, the self-induced flow of the primary vortex ring
around the axis of symmetry passes through the screen as a jet-like flow, which
later regenerates into a new vortex ring of the same sign but weaker circulation.
At an even higher Reynolds number, the primary vortex ring has sufficient
energy to pass through the screen via a possible mechanism of vortex stretching
and reconnection, and continues as a modified ring in its lee.

1. Introduction

The dynamics of vortex rings have been studied for more than 100 years (see Thomson
1867) and obviously many papers have published on the subject. While many of the earlier
studies have concentrated on the properties of a vortex ring such as its formation and
development, propagation speed, structures, stability and interactions (see, for example, Allen
and Naitoh 2005, Auerbach 1988, Dabiri and Gharib 2004, Didden 1979, Fukumoto and
Hattori 2005, Fukumoto and Moffatt 2001, Gharib et al 1998, Glezer and Coles 1990,
Hattori and Fukumoto 2003, Hicks 1885, Kambe and Oshima 1975, Kambe and Takao
1971, Konstantinov 1997, Lim 1989, 1997a, Lim and Nickels 1992, Lim et al 2008,
Maxworthy 1972, 1977, Mohseni et al 2001, Moore and Saffman 1974, Nitsche and
Krasny 1994, Norbury 1972, Pullin 1979, Riley and Steven 1993, Saffman 1970, 1978,
Virk et al 1994, Widnall and Sullivan 1973, Yamada and Matsui 1979), numerous studies
have also been devoted to practical applications of vortex rings. For example, cavitating vortex

© 2009 The Japan Society of Fluid Mechanics and IOP Publishing Ltd Printed in the UK
0169-5983/09/051404+26$30.00 1
Fluid Dyn. Res. 41 (2009) 051404 D Adhikari and T T Lim

rings produced by exciting a cavitating water jet have been used for underwater cleaning and
rock cutting (Chahine and Genoux 1983). Some researchers have even suggested using vortex
rings to extinguish gas and oil well fires (Akhmetov et al 1980) and to project smoke and other
effluents to high altitudes in the atmosphere without the need to use tall chimneys (Fohl 1967,
Turner 1960). In addition, vortex rings are known to occur in many biological flows (see, for
example, Birch and Dickinson 2001, Kheradvar et al 2007, Triantafyllou et al 2000, van den
Berg and Ellington 1997). The above list is by no means exhaustive and readers should refer
to the review articles by Shariff and Leonard (1992) and Lim and Nickels (1995), and the
books by Saffman (1992) and Lugt (1996).
It is well known that a vortex ring of low Reynolds number incident normally on a plane
wall expands in diameter near the wall until it stops due to vorticity diffusion and cancellation
with secondary vorticity of opposite sign near the wall. On the other hand, a vortex ring of
higher Reynolds number subjected to the same interaction not only expands in diameter but
also experiences the phenomenon of ‘vortex rebound’, which is caused by its mutual interac-
tion with a secondary vortex ring (see Lim et al 1991, Gendrich et al 1997, Walker et al 1987).
In his unpublished work, Morton (2000) has commented that when ‘the plane wall is
replaced by a fine gauze (such as a fly screen), the motion of the ring is little affected at low
Reynolds numbers but at higher Reynolds numbers the ring passes through the screen and
continues as a modified vortex ring in its lee’. The above observation is fascinating, but it
also raises a number of interesting questions. (i) Are there only two possible outcomes when
a vortex ring impacts on a porous screen? (ii) What is the effect of screen porosity on the
interaction? (iii) How different is the rate of vortex stretching (or radial expansion of vortex
ring) on a porous screen compared to that on a plane wall at low Reynolds number? To answer
these questions, we conducted a series of experiments by varying the Reynolds number of a
vortex ring (0/ν) (where 0 is the circulation of the vortex ring and ν is the kinematic viscosity
of water) and the screen porosity (β) (open area ratio) while keeping constant other parameters
such as wire diameter of the screen and vortex ring diameter. Unfortunately, vortex core size
increases with decreasing Reynolds number due to viscous diffusion, and therefore cannot
be maintained constant without affecting the other flow parameters. Nevertheless, we do not
expect this to have a significant effect on the overall results.

2. Experimental apparatus and procedure

The current experiment was performed using an apparatus, which is shown schematically in
figure 1. It is a modified version of the one used by Lim (1997b). It consists of a water tank,
which measures 500 mm (W ) × 500 mm (H ) × 1000 mm (L). The tank is made of Plexiglas,
which allows good visual access for flow visualization and DPIV measurements. The vortex
ring generator consists of a circular tube or nozzle with an internal diameter (DN ) of 26 mm
and a wall thickness of 10.5 mm. One end of the tube protrudes 120 mm (4.61DN ) inside
and 120 mm above the bottom of the tank, and the other end of the tube is connected
to a piston–cylinder arrangement. The piston measures 112 mm in diameter and can slide
smoothly inside a matching Plexiglas cylinder. One end of the piston is connected to a lead
screw, which forms part of the gear-drive mechanism that converts a rotational motion into
a linear motion of the piston, as shown in figure 1. The piston is driven by a stepper motor,
which in turn is controlled by TTL signals generated by a Labview program in a PC with
an Intel Pentium 4 processor. The motor can achieve 10 000 micro-steps per revolution or
0.036◦ per step. The Labview program can control not only the speed and the traversing
distance of the piston, but also its velocity profile. In the present study, a near ‘impulsive’
start velocity profile is achieved through a large area ratio of the piston (Ap ) and the nozzle
2
Fluid Dyn. Res. 41 (2009) 051404 D Adhikari and T T Lim

(a)

PC Motor
controller
Stepper
Porous motor
screen Nozzle
Dp

L2 L1 Bevel Lead
Piston gear screw

Water
tank
Camera

(b)
Pulsed laser

PC Motor
controller

Laser sheet

Figure 1. Schematic diagram of the experimental setup. (a) Top view, Dp = 112 mm,
L 1 = 120 mm, L 2 = 140 mm. (b) Side view. For clarity, the laser sheet is not shown
in (a).

(AN ) (i.e. Ap /AN = 18.55). To produce constant diameter vortex rings at different Reynolds
numbers, L N /DN is maintained at a constant value of 1 throughout the experiment, where
L N is the length of the slug of fluid ejected through the nozzle. The choice of L N /DN = 1 is
dictated by ease of operation of the apparatus; a different value of L N /DN produces a different
diameter vortex ring (see Auerbach 1988, Lim and Nickels 1995).
Although porous screens are available commercially, they do not meet our requirements
because (a) most commercial screens are fabricated with the ‘strings’ intertwined in a manner
similar to the strings on a tennis racket, leading to an uneven plane and (b) the diameter
of the strings often changes with the screen porosity. As these two variables may introduce
uncertainties into the results, we decided to fabricate our own screens using rapid prototyping
technology. The rapid prototyping machine used is Objet Eden 260 using Fullcure 720
transparent material made up of translucent acrylic-based photopolymer. The screens were
first designed using computer-aided drawing software, Solidworks R
. Unlike the commercial
screens, the strings of our screens are ‘fused’ at the junctions (see figure 2) and the string
diameter (dw ) is maintained constant at 0.2 mm regardless of the screen porosity. Although
screens of 90% and 80% porosity (β) were fabricated successfully, we were unable to produce
screens of lower porosities due to limited capability of our rapid prototyping machine. In view
of this, we decided to proceed with the experiment with the two available screens. Note that
the screen and the supporting frame were fabricated as a single piece in order to ensure that
the screen is uniformly mounted along the perimeter of the support. Earlier attempts to stretch
and mount the screen taut on a separate support led to tearing due to non-uniform stretching.
After the screens were painted white for the purpose of flow visualization, measurements
under a microscope showed that the porosities of the screens are reduced to 81% and 62%,
3
Fluid Dyn. Res. 41 (2009) 051404 D Adhikari and T T Lim

(a) (b)

Figure 2. Porous screen used in the current investigation. (a) Schematic diagram showing the
construction of the screen. (b) The 81% porosity screen (fabricated using a rapid prototyping
machine) after it has been painted white. Each subdivision of the ruler is 1 mm.

with a standard deviation of 1%. In all the results reported here, there is no evidence of screen
distortion during the vortex interaction.
To visualize the flow, dye (blue laundry brightener with S.G. ≈ 1 and Schmidt number of
O(102 )) was released slowly around the circumference of the nozzle, and when the piston was
given an ‘impulsive push’ by the stepper motor, a cylindrical vortex sheet was produced, which
then rolled up into a toroidal vortex ring. Although a secondary vortex ring was produced
inside the nozzle after the piston motion had stopped, it had negligible influence on the
primary vortex ring. The formation of the secondary vortex ring is common in most vortex
ring generators, and experimental study by Didden (1979) shows that although it causes an
initial reduction in the diameter of the primary vortex ring, its influence diminishes by the
time the primary vortex ring has traveled approximately three ring diameters from the nozzle.
In the present study, the screen was located at 140 mm (or 5.38DN ) from the exit of the nozzle
with its plane normal to the axis of the nozzle. To capture flow images during the interaction,
a Sony CCD camera was used in conjunction with a video recorder (Model DSR-45P). The
camera was running at 25 frames per second with a shutter speed of 1/250 s.
For most of the Reynolds numbers considered, the circulation of the vortex ring before
its impact on the screen was determined from the velocity field measured using the digital
particle image velocimetry (DPIV) technique. The DPIV setup consists of a Nd-YAG pulsed
laser (New Wave Solo PIV) as the illumination source and a Kodak Megaplus CCD camera
(model ES 4.0, 2048 pixels × 2048 pixels) to capture the image data. Dantec R
FlowMap
adaptive cross correlation DPIV algorithm was used to analyze the captured flow images
using an interrogation area of 32 pixels × 32 pixels and 25% overlap in the final step of
the processing. Subpixel accuracy in the velocity vectors was achieved using a three-point
symmetrical Gaussian curve-fit interpolation scheme, which gives an uncertainty of about
4% of the maximum velocities (see Luff et al 1996). Data validation was carried out using
both ‘peak validation’ and ‘moving average validation’ techniques, and velocity vectors were
rejected unless both the criteria were met.
Attempts were initially made to measure velocity fields at different stages of
the interaction, but unavoidable random reflection of the laser sheet from the screen
increases the uncertainty of the DPIV measurements, which makes the results highly
unreliable. Because of this high experimental uncertainty, the velocity and vorticity fields
are not presented here and the discussions that follow are based solely on the flow
visualization.
4
Fluid Dyn. Res. 41 (2009) 051404 D Adhikari and T T Lim

(a) (b)

Figure 3. (a) A vortex ring generated using L N /DN = 1 and Re0 = 1177. (b) Corresponding
velocity vector field obtained using the DPIV technique. The closed loop indicates the path taken
to calculate the circulation (0).

3. Results and discussion

Figure 3(a) shows a snapshot of a typical vortex ring when a slug of the fluid was ejected
through the nozzle according to L N /DN = 1.0. Note that the contrast of the picture has been
enhanced using Photoshop R
by discarding the color and inverting the original vortex ring
in blue color into a ‘white’ vortex ring. For ease of comparison with previously published
results, the picture is orientated with the vortex ring traveling vertically downward toward
the screen. Since the dye was released at the location where vorticity was generated, the dye
sheet depicted in figure 3(a) also represents the vortex sheet. However, one should be mindful
of the fact that dye diffuses much slower than vorticity, and therefore over a long period of
time, dye may not truly represent the actual boundary of the vortex sheet. However, this is
not an issue of concern here since the interaction took place soon after the vortex ring was
fully formed. Figure 3(b) shows the velocity field of a similar vortex ring obtained using the
DPIV technique, from which the circulation (0) of the vortex ring can be evaluated. Here,
the circulation is determined by taking the line integral of velocity around a closed loop with
one of the sides coinciding with the axis of symmetry of the vortex ring and enclosing the
vortex core. For L N /DN = 1, the diameter (D) of the vortex ring obtained is 28.4 mm (or
D/DN = 1.09) and is relatively independent of the ejection velocity (Ue ) through the nozzle
or Reynolds number. This finding is consistent with the result of Auerbach (1988), which
also shows that the diameter of a vortex ring is a function of L N /DN only. The circulation
of the vortex ring, as is to be expected, increases with the ejection velocity (Ue ) and this can
be clearly seen in figure 4, which displays the relationship between the Reynolds number
based on the circulation of the ring (Re0 = 0/ν) and that based on the ejection velocity and
the nozzle diameter (ReN = Ue DN /ν). A line of best fit through the measured data gives the
relationship between the two Reynolds numbers as

Re0 = 0.8344ReN .
Based on the above expression, it can be shown that 0/ 0slug = 1.66, which is consistent with
the results presented in Lim and Nickels (1995), where 0slug is the circulation based on the
slug model. This expression allows the determination of Re0 at different ejection velocities
(Ue ) without the need to conduct further DPIV measurements.
5
Fluid Dyn. Res. 41 (2009) 051404 D Adhikari and T T Lim

3500

3000

2500

2000
Re Γ

1500

1000

500

0
0 500 1000 1500 2000 2500 3000 3500 4000 4500
Re N

Figure 4. Relationship between Re0 (= 0/ν) and ReN (= Ue DN /ν) for L N /DN = 1.0.

The vortex core diameter (dc ), which is defined as the distance between two neighboring
locations of maximum tangential velocity (see also Saffman 1978), is derived from the DPIV
measurements to be approximately 0.40D at Re0 = 1000, and decreases with increasing
Reynolds number until it asymptotes to about 0.35D beyond Re0 = 2500. While it would
be ideal to maintain a constant core size throughout the experiment, this is impossible to
achieve without affecting the other flow parameters. However, changes in the core size are not
expected to have a significant effect on the overall flow development since dw /dc  1 in all
the cases considered here.
In the results to follow, the case of a vortex ring impacting normally on a solid wall
is presented first, followed by the same interaction with porous screens. Although vortex
ring/solid wall interaction has been studied extensively in the past (see Lim et al 1991, Walker
et al 1987) and the phenomenon is well understood, it is presented here for completeness
since a solid wall can be considered as a porous screen with zero porosity. Moreover, vortex
ring–porous screen interaction shares many of the salient flow features of the vortex ring–solid
wall interaction.

3.1. Vortex ring–solid wall interaction


Figure 5 is a sequence of images showing the interaction of a vortex ring of low Reynolds
number (Re0 = 384) with a solid wall. These images are extracted from the video recording of
the interaction based on the important flow features that they display. For ease of comparison
with all the cases considered here, t ∗ (= U t/D) = 0 corresponds to the instant when the
vortex ring is one ring diameter away from the wall or screen. Here, U is the propagation
velocity of the vortex ring before the interaction, and t is the elapsed time. As is to be expected,
figure 5 clearly shows that a primary vortex ring approaching normally a solid wall expands in
diameter and decreases in dye (or vortex) core size in order to satisfy continuity. The increase
in the primary vortex ring diameter can be explained using the ‘method of images’, which is
based on an inviscid theory. With this method, the solid wall is treated as a plane of symmetry,
which can be simulated by introducing an image vortex ring of opposite circulation on the
other side of the wall. This is clearly illustrated in the authors’ interpretation in figure 5(B)(a).
Using the Biot–Savart law, it can be shown that the induced velocity field of one vortex ring
has the effect of causing the other vortex ring to expand radially outward. In an inviscid
6
Fluid Dyn. Res. 41 (2009) 051404 D Adhikari and T T Lim

Direction of
(a) t* = 1.28 propagation (a)

Direction of –Γ +Γ Primary ring


propagation
Vi Vi
Solid wall
Vp Vp

+Γ –Γ Image ring

Solid wall

(b) t* = 1.65 (b)

Vi Vi
Vp Vp

(c) t* = 3.46 (c)

Secondary V ps V sp V spV ps
Secondary vorticity
vorticity

(d) t* = 3.88 (d)

Secondary
vorticity

(A) (B)

Figure 5. Impact of a vortex ring on a solid wall for Re0 = 384. Left column: captured flow
images. t ∗ = 0.0 corresponds to the instant when the vortex ring is one ring diameter from the wall
(picture not shown to save space). Right column: the authors’ interpretation of the interaction. Vp
is the induced velocity of the primary ring on the image ring, Vi the induced velocity of the image
ring on the primary ring, Vps the induced velocity of the primary ring on the secondary vorticity
and Vsp the induced velocity of the secondary vorticity on the primary ring.

7
Fluid Dyn. Res. 41 (2009) 051404 D Adhikari and T T Lim

(a) (c)

(b) (d)

Figure 6. Trajectories of vortex cores versus non-dimensional time (t ∗ = U t/D) on a solid wall
and porous screens. (a) Re0 = 384; (b) Re0 = 779; (c) Re0 = 1597; (d) Re0 = 2369. ♦ solid
wall; ◦ 62% porosity; M 81% porosity.

fluid, a vortex ring can expand indefinitely, but in a real fluid, the expansion is limited by the
presence of secondary vorticity (or a shear layer) of opposite sign to the primary vortex ring
near the wall. This secondary vorticity is generated by the induced velocity of the primary
vortex ring as a result of the no-slip condition at the wall. For the vortex ring of low Reynolds
number presented in figure 5, although there is no apparent roll-up of secondary vorticity, its
induced velocity field still had the effect of causing noticeable rebound of the primary vortex
ring from the wall, a phenomenon commonly referred to as a ‘vortex rebound’ (compare
figure 5(A)(b) with figures 5(A)(c) and (d)). Since the primary vortex ring was considerably
stronger, it entrains the secondary vorticity into its core, resulting in vorticity cancellation
until the primary vortex ring ceases to expand. For a higher Reynolds number primary vortex
ring, the sequence of events during its impact on the wall is essentially the same as that
described above, except that the secondary shear layer is now stronger and able to roll up
into a concentrated secondary vortex ring (see, for example, Walker et al 1987). Mutual
interaction between the primary and secondary vortex rings leads to a more intense rebound
of the primary vortex ring, which is followed by the ejection of the secondary vortex ring
from the wall. For a much higher Reynolds number, past studies (see, for example, Walker
et al 1987) showed that more than one vortex rebound is possible, and a tertiary vortex ring
of the same sign but weaker circulation than the secondary vortex ring can also be generated,
8
Fluid Dyn. Res. 41 (2009) 051404 D Adhikari and T T Lim

resulting in a significantly more complex interaction. Figures 6(a)–(d) show the trajectories of
vortex cores versus non-dimensional time during vortex ring–solid wall interaction for Re0
of 384, 779, 1597 and 2369, respectively. Also included in the figure are the corresponding
vortex core trajectories for 62% and 81% porosity screens, and these results will be discussed
in the following section.
3.2. Vortex ring–porous screen interaction
When the solid wall is replaced by a porous screen, the final outcome of the interaction is very
much dependent on the Reynolds number and the porosity of the screen, and can be broadly
categorized into four Scenarios.
Figure 7 is a sequence of pictures showing the interaction of a vortex ring of low Reynolds
number (i.e. Re0 = 384) with a 62% porosity screen, and figure 8 shows the same interaction
for a vortex ring of slightly higher Reynolds number (i.e. Re0 = 779). In each case, the vortex
ring not only expands in diameter near the screen until it stops as was rightly described by
Morton (2000), but also undergoes a vortex rebound similar to that in a vortex ring–solid
wall interaction. Apart from the above similarities with the solid wall, there are noticeable
differences between the two interactions. This is clearly illustrated in figures 6(a) and (b),
where the trajectories of vortex cores versus non-dimensional time (t ∗ = U t/D) for a 62%
porosity screen are displayed. It is evidenced in figure 6(a) that at the low Reynolds number of
384, the rate of increase of vortex ring diameter (or vortex stretching) decreases as the porosity
of the screen is increased. To gain an insight into this intriguing behavior, we examined closely
the surrounding flow field during the actual interaction, and found that as the vortex ring
approached the screen, part of the self-induced flow around the axis of symmetry of the ring
passed through the screen to form a jet-like flow, while the primary vortex ring remained on
the same side of the screen. This is clearly illustrated in the authors’ interpretation in figure 9
(herein referred to as Scenario 1). Since the transmitted jet-like flow carries with it part of
the original circulation, the strength of the primary vortex ring is reduced accordingly. Also,
the existence of the transmitted flow implies that the screen cannot be treated as a plane of
symmetry, and hence the method of images strictly cannot be applied in the same manner as
in vortex ring–solid wall interaction. Nevertheless, in the local region between the vortex core
and the screen, the streamline is nearly parallel to the screen, and one can use the method of
images locally to approximately explain the stretching of the vortex ring. But with the loss
of circulation of the primary vortex ring to the transmitted flow, the rate of vortex stretching
is reduced accordingly. This observation is supported by the results with the 81% porosity
screen, where a larger open area means that there is more flow transmitted through the screen
and thus a further reduction in the circulation of the primary vortex ring. Although secondary
vorticity is still generated at the screen surface due to the no-slip condition, the weak rebound
of the primary vortex ring indicates that it is relatively weak. The weaker secondary vorticity
can be attributed partly to the reduced circulation of the primary vortex ring and partly to a
smaller solid surface area. It is interesting to note that for this particular Reynolds number,
despite the differing rate of vortex stretching, the vortex rings expanded to about twice their
original diameters on all the three surfaces before the vorticity was completely dissipated.
However, the same behavior does not follow through to a higher Reynolds number vortex ring
as can be seen from the vortex core trajectories displayed in figure 6(b) for Re0 = 779. It
is clear from the figure that the primary vortex ring expands appreciably less on the 62%
porosity screen than on the solid wall. Moreover, the vortex core trajectory on the 81%
porosity screen follows a completely different trend from the other two surfaces. This distinct
behavior can be attributed to the larger open area of the higher porosity screen, which has
allowed the primary vortex ring to pass through it and travel as a modified ring as can be seen
9
Fluid Dyn. Res. 41 (2009) 051404 D Adhikari and T T Lim

(a) t* = 2.21 (d) t* = 3.86


Direction of
propagation

Porous
screen

(b) t* = 2.72 (e) t* = 6.07

(c) t* = 2.83 (f) t* = 6.75

Figure 7. Impact of a vortex ring on a 62% porosity screen for Re0 = 384. Time increases from
(a) to (f). t ∗ = 0.0 corresponds to the instant when the vortex ring is one ring diameter from the
screen.

10
Fluid Dyn. Res. 41 (2009) 051404 D Adhikari and T T Lim

(a) t* = 1.55 (d) t* = 2.93


Direction of
propagation

(b) t* = 1.97 (e) t* = 4.07

(c) t* = 2.44 (f) t* = 6.59

Figure 8. Impact of a vortex ring on a 62% porosity screen for Re0 = 779. Time increases from
(a) to (f). t ∗ = 0.0 corresponds to the instant when the vortex ring is one ring diameter from the
screen.

11
Fluid Dyn. Res. 41 (2009) 051404 D Adhikari and T T Lim

(a)

Primary
vortex ring

Porous screen

Jet-like flow

(b)

Vp Vs Vs Vp
Secondary
vortex ring

(c)
Secondary
vortex ring

(d)

Figure 9. The authors’ interpretation of the vortex ring/porous screen interaction at Re0 = 384
(Scenario 1). Time increases from (a) to (d). Note that broken lines merely represent conceptual
streamlines depicting the transmission of the induced flow of the toroidal ring through the screen.
The actual streamline pattern is more complex. Single arrows represent the direction of the flow
field.
12
Fluid Dyn. Res. 41 (2009) 051404 D Adhikari and T T Lim

in the flow visualization result presented in figure 10. This is the same scenario described by
Morton (2000) for the higher end of the ‘outermost’ condition, which will be discussed later.
For now, we examine those cases prior to this ‘outermost’ flow condition using the interaction
with the 62% porosity screen for the purpose of the discussion.
Returning to figure 6(b), in order to better understand why a vortex ring at Re0 = 779
expands less on the 62% porosity screen than on the solid wall, we took a closer look at
the interaction during the actual experiment and found that the self-induced flow of the
primary vortex ring, after passing through the screen as a jet like flow, started to roll up
into a ring-like structure. This is in contrast to the lower Reynolds number case depicted
in figure 6(a), where the jet flow did not appear to roll up into a well-defined ring-like
structure. Unfortunately, this ring-like structure does not reproduce clearly in figure 8, due
partly to the lack of dye at strategic locations and partly to rapid diffusion of dye after passing
through the screen. The development of this ring-like structure is clearly illustrated in the
authors’ interpretation in figure 11(d) (herein referred to as Scenario 2). It is worth noting
that the sequence of events from figures 11(a)–(c) is similar to that from figures 9(a)–(c).
The ring-like structure has a circulation of the same sign as the primary vortex ring, but its
influence on the primary vortex ring appears to be small (probably due to the presence of the
screen) compared to the influence of the secondary vortex ring. This can be deduced from
the vortex core trajectory for a 62% porosity screen depicted in figure 6(b), which indicates
that the primary vortex ring started to slow down its radial expansion at U t/D ≈ 2.0, which
is during the formation of the secondary vortex ring, but much earlier than the formation
of a ring-like structure. It can be shown using the Biot–Savart law that the velocity induced
by the secondary vortex ring has the effect of retarding the radial expansion of the primary
vortex ring, which may explain the observed trend in figure 6(b). However, with increasing
Reynolds number of the primary vortex ring, the circulations of the secondary vortex ring and
the ring-like structure also increase. Figures 12 and 13 show the flow visualization results
for Re0 = 1597 and 2369, and figures 6(c) and (d) show the corresponding vortex core
trajectories versus non-dimensional time. The ring-like structure eventually develops into a
new vortex ring (henceforth referred to as a ‘regenerated vortex ring’). Due to the presence
of the screen, the influence of the regenerated vortex ring on the primary vortex ring appears
to be small. However, the close proximity of the secondary vortex ring to the primary vortex
ring leads to mutual interaction. Here, the velocity induced by the now stronger secondary
vortex ring causes the primary vortex ring to rebound and, at the same time, reverses its radial
expansion and shrinks. Concurrently, the velocity induced by the primary vortex ring causes
the secondary vortex ring to orbit around the vortex core and toward the inner region of the
primary vortex ring. But at some stage of the entrainment, the self-induced velocity of the
secondary vortex ring is able to escape the induced velocity field of the primary vortex ring and
travels away from the screen. This process is clearly illustrated in the authors’ interpretation
in figure 14 (henceforth referred to as Scenario 3). But when a higher Reynolds number
primary vortex ring impacts the screen as shown in figure 15, the primary vortex ring, after
transferring some of the circulation to the regenerated vortex ring, is still energetic enough
to pass through the screen via a possible mechanism of vortex stretching and reconnection
(see also Ashurst and Meiron 1987, Chatelain et al 2003, Kida and Takaoka 1994, Kida
et al 1991, Melander and Hussain 1989, Saffman 1990) leaving behind the secondary
vortex ring on the other side of the screen. The modified ‘transmitted’ primary vortex ring
eventually pairs up with the regenerated vortex ring as shown in the authors’ interpretation
in figure 16 (herein referred to as Scenario 4). In between figures 16(c) and (d), the
vortex reconnection process is significantly more complex, and the authors’ interpretation
of this vortex reconnection is shown in figure 17. Our model is consistent with the vortex

13
Fluid Dyn. Res. 41 (2009) 051404 D Adhikari and T T Lim

(a) t* = 1.72 (d) t* = 2.65

Direction of
propagation

(b) t* = 2.05 (e) t* = 3.40

(c) t* = 2.36 (f) t* = 6.36

Figure 10. Impact of a vortex ring on an 81% porosity screen for Re0 = 779. Time increases from
(a) to (f). t ∗ = 0.0 corresponds to the instant when the vortex ring is one ring diameter from the
screen.

14
Fluid Dyn. Res. 41 (2009) 051404 D Adhikari and T T Lim

(a)

Primary
vortex ring

Porous screen

(b)

Secondary
V vortex ring
Vpsps Vsp Vsp Vps

(c)
Secondary
Primary vortex ring
vortex ring

(d)

Ring-like
structure

Figure 11. The authors’ interpretation of the vortex ring–porous screen interaction at Re0 = 779
(Scenario 2). Time increases from (a) to (d). Vps is the induced velocity of the primary vortex
ring on the secondary vortex ring and Vsp the induced velocity of the secondary vortex ring on the
primary vortex ring. Single arrows represent the direction of the flow field.

15
Fluid Dyn. Res. 41 (2009) 051404 D Adhikari and T T Lim

(a) t* = 1.47 (d) t* = 3.65


Direction of
propagation

(b) t* = 1.84 (e) t* = 5.73

(c) t* = 2.45 (f) t* = 9.11

Figure 12. Impact of a vortex ring on a 62% porosity screen for Re0 = 1597. Time increases from
(a) to (f). t ∗ = 0.0 corresponds to the instant when the vortex ring is one ring diameter away from
the screen.

16
Fluid Dyn. Res. 41 (2009) 051404 D Adhikari and T T Lim

(a) t* = 1.19 (d) t* = 2.56


Direction of
propagation

(b) t* = 1.57 (e) t* = 4.16

(c) t* = 2.14 (f) t* = 9.49

Figure 13. Impact of a vortex ring on a 62% porosity screen for Re0 = 2369. Time increases from
(a) to (f). t ∗ = 0.0 corresponds to the instant when the vortex ring is one ring diameter away from
the screen.

17
Fluid Dyn. Res. 41 (2009) 051404 D Adhikari and T T Lim

(a)

Primary
vortex ring

Porous screen

(b)

Vsp
Vps Primary
Secondary vortex ring
vortex ring

(c)

Secondary
vortex ring
Primary
vortex ring

Regenerated
vortex ring

(d)

Figure 14. The authors’ interpretation of figures 12 and 13 (Scenario 3). Time increases from (a)
to (d). Vps is the induced velocity of the primary vortex ring on the secondary vortex ring and Vsp
the induced velocity of the secondary vortex ring on the primary vortex ring. Only the induced
velocity vectors on the left-hand side of the vortices are shown in (b). Single arrows represent the
direction of the flow field.
18
Fluid Dyn. Res. 41 (2009) 051404 D Adhikari and T T Lim

(a) t* = 1.13 (d) t* = 1.77


Direction of
propagation

(b) t* = 1.28 (e) t* = 2.44

(c) t* = 1.45 (f) t* = 3.67

Figure 15. Impact of a vortex ring on a 62% porosity screen for Re0 = 4026. Time increases from
(a) to (f). t ∗ = 0.0 corresponds to the instant when the vortex ring is one ring diameter away from
the screen.

19
Fluid Dyn. Res. 41 (2009) 051404 D Adhikari and T T Lim

(a)

Primary
vortex ring

Porous screen

(b)

(c)
Primary Secondary
vortex ring vortex ring

(d)

Merged primary
and regenerated
vortex rings

Figure 16. The authors’ interpretation of figure 15 (Scenario 4). Time increases from (a) to (d).
The passage of the primary vortex ring through the screen is via a possible mechanism of vortex
stretching and reconnection as detailed in figure 17. Note that the wavy ring in (d) is only a
conceptual representation, and the bridges as indicated in figure 17 are not shown here to avoid
complicating the figure. Single arrows represent the direction of the flow field.

20
Fluid Dyn. Res. 41 (2009) 051404 D Adhikari and T T Lim

Figure 17. The authors’ interpretation of vorticity reconnection through a porous screen. Arrows
indicate the direction of vorticity. The shaded region indicates vorticity cancellation. The model
is consistent with a similar model of vorticity reconnection proposed by Kida et al (1991) (see
also Melander and Hussain 1989, Kida and Takaoka 1994, Chatelian et al 2003). Arrows represent
the direction of vorticity.

reconnection model proposed by Kida et al (1991) to explain the collision of two vortex
rings (see also Kida and Takaoka 1994, Melander and Hussain 1989). For clarity, we consider
parallel rows of wire, and only the vorticity lines (and not material lines) are shown. As
depicted in the figure, when the vortex ring impinges on the wires, the parts of the vortex core
(filament) that are obstructed by the wires are stretched locally (see figure 17(b)). During the
stretching, vorticity lines of opposite sign come into contact with each other (see figure 17(c)),
resulting in vorticity cancellation through viscous diffusion (figure 17(d)). Consequently,
they are connected to their counterpart on the other side of the wire. Because of a lower
angular velocity near the interaction area relative to the rest of the flow, the vorticity lines are
entangled due to the difference in the angular velocity and subsequently formed cross-linking
or bridges as shown in figure 17(d) (see Kida and Takaoka 1994, Kida et al 1991, Melander
and Hussain 1989). As the rest of the vortex filaments continue to convect away from the
wire, the reconnection region is stretched until the vorticity is dissipated through viscous
diffusion, and the flow relaxes to a reasonably stable, although slightly distorted, vortex ring.
With the porous screen (i.e. a network of crossed wires), although the actual interaction
21
Fluid Dyn. Res. 41 (2009) 051404 D Adhikari and T T Lim

(a) t* = 1.49 (d) t* = 2.07

(b) t* = 1.67 (e) t* = 2.28

(c) t* = 1.89 (f) t* = 4.23

Figure 18. Impact of a vortex ring on an 81% porosity screen for Re0 = 1597. Time increases
from (a) to (f). t ∗ = 0.0 corresponds to the instant when the vortex ring is one ring diameter away
from the screen.

22
Fluid Dyn. Res. 41 (2009) 051404 D Adhikari and T T Lim

(a)

Primary
vortex ring

Porous screen

(b)

(c)

Region of vortex stretching,


vorticity intensification and
dissipation through viscous
diffusion

Figure 19. The authors’ conjecture of Scenario 4 if dw /dc  1. Time increases from (a) to (c).
In (c), the large wire diameter prevents vorticity lines of opposite sign from coming into contact
with each other, although the stretched vortex filament intensifies the vorticity, hence accelerating
vorticity cancellation through viscous diffusion. Single arrows represent the direction of the flow
field. Note that the resulting vortex motion after (c) is rather convoluted.

is undoubtedly more complicated, we believe that the salient features of the flow do not
differ considerably.
At an even higher Reynolds number, the sequence of events is pretty much the same,
except that the more energetic primary vortex ring is able to pass directly through the screen
in a much shorter time via the same mechanism of vortex stretching and reconnection and
continues as a modified ring as described by Morton (2000). This is essentially the same as
Scenario 4. A much clearer picture of the flow development for Scenario 4 is better observed
with the 81% porosity screen, which occurs at a much lower Reynolds number (see figure 10
for Re0 = 779 and figure 18 for Re0 = 1597). Although screen porosity does not significantly
affect the various scenarios reported here, it does influence the Reynolds number at which a
particular scenario occurs. Generally, the Reynolds number at which a particular scenario
occurs decreases with an increase in the screen porosity as shown in the above example.

23
Fluid Dyn. Res. 41 (2009) 051404 D Adhikari and T T Lim

As noted earlier, besides screen porosity and Reynolds number, the ratio of the wire
diameter (dw ) and vortex core diameter (dc ) also plays an important role in the interaction.
In the present study, dw /dc  1; however, if dw /dc  1, we believe that the final outcome
for Scenario 4 is likely to be different due to the absence of vortex reconnection during the
interaction, but the final outcomes for Scenarios 1–3 would not be significantly different.
Figure 19 shows the authors’ interpretation for Scenario 4 if dw /dc  1, and as before, only
vorticity lines are shown. Our conjecture model is based partly on a related experiment by
Naitoh et al (1995) on a vortex ring traveling across a thin circular cylinder and partly on
our own investigation with larger diameter cylinders. As depicted in figure 19, as a vortex
ring impinges on a coarse screen, parts of the vortex filament obstructed by the wires are
stretched locally. But due to the larger diameter of the wires relative to the vortex core
size, vorticity lines of opposite sign are unlikely to come into contact with each other, as
shown in figure 19(c). This is in contrast to the dw /dc  1 case illustrated in figure 17(c).
While the ‘unaffected’ parts of the vortex filament continue to move away from the wires,
vortex stretching in the affected region causes the vorticity to intensify until it is dissipated
through viscous diffusion. In this case, the original primary vortex ring is unlikely to relax to
a well-defined vortex ring after passing through the screen. Instead, it would merge with the
regenerated vortex ring in a rather convoluted manner before settling down to what would look
like a ring-like structure further downstream. For simplicity, the generation of the secondary
vorticity at the wire surface due to the no-slip condition is not shown in figure 19 as it would
not significantly affect the final outcome.

4. Conclusions

The impact of a vortex ring on a porous screen has been studied experimentally using
flow visualization and DPIV techniques. The purpose is to re-examine the issue raised by
Morton (2000) that a vortex ring of low Reynolds number approaching normally a porous
screen expands in diameter near the screen until it ceases, whereas a vortex ring of high
Reynolds number passes through the screen and continues as a modified ring in its lee.
However, our results obtained using two specially designed screens of 62% and 81% porosity
show that such a flow behavior can only occur at very low Reynolds number or above a
certain critical Reynolds number, which is dependent on the screen porosity. In between these
two ‘outermost’ Reynolds numbers, vortex ring–porous screen interaction is significantly
more complex, resulting in two additional flow scenarios. Further to the observation by
Morton (2000) for a vortex ring of low Reynolds number, our result shows that the rate of
vortex stretching (or radial expansion of the primary vortex ring) on a porous screen decreases
with increasing screen porosity. As far as we know, this observation has not been reported
previously and can be partly attributed to the primary vortex ring losing part of its circulation
to the self-induced flow that passes through the screen to produce a jet-like flow. At the same
time, the primary vortex ring, through its induced velocity and the no-slip condition at the
screen, generates secondary vorticity (or a shear layer) of opposite sign near the screen. At
this low Reynolds number, the shear layer apparently does not roll up into a concentrated
secondary vortex ring. But at a higher Reynolds number, a secondary vortex ring is produced
and the jet-like flow that passes through the screen regenerates into a new vortex ring of the
same sign but weaker circulation than the primary vortex ring. During the mutual interaction
between the primary and secondary vortex rings, the velocity induced by the primary vortex
ring causes the secondary vortex ring to orbit around the core and toward the inner region of
the primary vortex ring. Similarly, the velocity induced by the secondary vortex ring caused
the primary vortex ring to rebound from the screen in a well-known phenomenon of vortex
24
Fluid Dyn. Res. 41 (2009) 051404 D Adhikari and T T Lim

rebound often encountered in vortex ring–solid wall interaction, at the same time limiting the
radial expansion of the primary vortex ring. As the Reynolds number is increased further,
a regenerated vortex ring is still produced, while the induced velocity of the now stronger
secondary vortex ring is able to reverse the radial expansion of the primary vortex ring, causing
it to shrink. At a much higher Reynolds number, the primary vortex ring has sufficient energy
to pass through the screen via a possible mechanism of vortex stretching and reconnection,
and continues as a modified ring in its lee as described by Morton (2000), leaving behind the
secondary vortex ring, which convects away from the screen through its self-induced velocity.
In general, the critical Reynolds number increases with decreasing screen porosity. Because
the transition process is a gradual one, the critical Reynolds number for each screen is difficult
to determine accurately. Moreover, for a screen of low porosity, the ‘critical Reynolds number’
can be so high that the vortex ring becomes turbulent upon its impact on the screen.

Acknowledgments

We thank the reviewers for their thoughtful comments and valuable suggestions.

References

Akhmetov D G, Lugovtsov B A and Tarasov V F 1980 Extinguishing gas and oil well fires by means of vortex rings
Combust. Explosion Shock Waves 16 490–4
Allen J J and Naitoh T 2005 Experimental study of the production of vortex rings using a variable diameter orifice
Phys. Fluids 17 061701
Ashurst W T and Meiron D I 1987 Numerical study of vortex reconnection Phys. Rev. Lett. 58 1632–5
Auerbach D 1988 Some open questions on the flow of a circular vortex rings Fluid Dyn. Res. 3 209–13
Birch J M and Dickinson M H 2001 Spanwise flow and the attachment of the leading-edge vortex on insect wings
Nature 412 729–33
Chahine G L and Genoux P F 1983 Collapse of a cavitating vortex ring J. Fluids Eng. Trans. ASME 105 400–5
Chatelain P, Kivotides D and Leonard A 2003 Reconnection of colliding vortex rings Phys. Rev. Lett. 90 054501
Dabiri J O and Gharib M 2004 Fluid entrainment by isolated vortex rings J. Fluid Mech. 511 311–33
Didden N 1979 Formation of vortex rings—rolling-up and production of circulation Z. Angnew. Math. Phys. 30
101–16
Fohl T 1967 Optimization of flow for forcing stack wastes to high altitudes J. Air Pollut. Control Assoc. 17 730–3
Fukumoto Y and Hattori Y 2005 Curvature instability of a vortex ring J. Fluid Mech. 526 77–115
Fukumoto Y and Moffatt H K 2001 Motion and expansion of a viscous vortex ring. Part 1. A higher-order asymptotic
formula for the velocity J. Fluid Mech. 417 1–45
Gendrich C P, Koochesfahani M M and Nocera D G 1997 Molecular tagging velocimetry and other novel applications
of a new phosphorescent supramolecule Exp. Fluids 23 261–372
Gharib M, Rambod E and Shariff K 1998 A universal time scale for vortex ring formation J. Fluid Mech. 360 121–40
Glezer A and Coles D 1990 An experimental-study of a turbulent vortex ring J. Fluid Mech. 211 243–83
Hattori Y and Fukumoto Y 2003 Short-wavelength stability analysis of thin vortex rings Phys. Fluids 15 3151–63
Hicks W M 1885 Researches on the theory of vortex rings—Part II Phil. Trans. R. Soc. A 176 725–80
Kambe T and Oshima Y 1975 Generation and decay of viscous vortex rings J. Phys. Soc. Japan 38 271–80
Kambe T and Takao T 1971 Motion of distorted vortex rings J. Phys. Soc. Japan 31 591–9
Kheradvar A, Milano M and Gharib M 2007 Correlation between vortex ring formation and mitral annulus dynamics
during ventricular rapid filling ASAIO J. 53 8–16
Kida S and Takaoka M 1994 Vortex reconnection Ann. Rev. Fluid Mech. 26 169–89
Kida S, Takaoka M and Hussain F 1991 Collision of two vortex rings J. Fluid Mech. 230 583–646
Konstantinov M 1997 Numerical investigation of the interaction of coaxial vortex rings Int. J. Heat Fluid Flow 7
120–40
Lim T T 1989 An experimental study of a vortex ring interacting with an inclined wall Exp. Fluids 7 453–63
Lim T T 1997a A note on the leapfrogging between two coaxial vortex rings at low Reynolds number Phys. Fluids 9
239–41

25
Fluid Dyn. Res. 41 (2009) 051404 D Adhikari and T T Lim

Lim T T 1997b On the role of Kelvin-Helmholtz-like instability in the formation of turbulent vortex rings Fluid Dyn.
Res. 21 47–56
Lim T T, Lua K B and Thet K 2008 Does Kutta lift exist on a vortex ring in a uniform cross flow? Phys. Fluids 20
051701
Lim T T and Nickels T B 1992 Instability and reconnection in the head-on collision of two vortex rings Nature 357
225–7
Lim T T and Nickels T B 1995 Vortex rings Fluid Vortices ed S I Green (Kluwer: Academic) pp 95–147
Lim T T, Nickels T B and Chong M S 1991 A note on the cause of rebound in the head-on collision of a vortex ring
with a wall Exp. Fluids 12 41–8
Luff J D, Drouillard T, Rompage A M, Linne M A and Hertzberg J R 1999 Experimental uncertainties associated
with particle image velocimetry (PIV) based vorticity algorithms Exp. Fluids 26 36–54
Lugt H J 1996 Introduction to Vortex Theory (Potomac, MD: Vortex Flow Press)
Maxworthy T 1972 The structure and stability of vortex rings J. Fluid Mech. 51 15–32
Maxworthy T 1977 Some experimental studies of vortex rings J. Fluid Mech. 81 465–95
Melander M and Hussain F 1989 Cross-linking of two antiparallel vortex tubes Phys. Fluids A 1 633–6
Mohseni K, Ran H Y and Colonius T 2001 Numerical experiments on vortex ring formation J. Fluid Mech. 430
267–82
Moore D W and Saffman P G 1974 A note on the stability of vortex ring of small cross-section Proc. R. Soc. A 338
535–7
Morton B R 2000 Convection with Vorticity http://met.psu.edu/∼evans/0CLASSES/CWV/CWV-Morton.html
Naitoh T, Sun B and Yamada H 1995 A vortex ring traveling across a thin circular cylinder Fluid Dyn. Res. 15 43–56
Nitsche M and Krasny R 1994 A numerical study of vortex ring formation at the edge of a circular tube J. Fluid
Mech. 276 139–61
Norbury J 1972 A family of steady vortex rings J. Fluid Mech. 57 417–31
Pullin D I 1979 Vortex ring formation at tube and orifice openings Phys. Fluids 22 401–3
Riley N and Stevens D P 1993 A note on leapfrogging vortex rings Fluid Dyn. Res. 11 235–44
Saffman P G 1970 The velocity of viscous vortex rings Stud. Appl. Math. 49 371–80
Saffman P G 1978 Number of waves on unstable vortex rings J. Fluid Mech. 84 625–39
Saffman P G 1990 A model of vortex reconnection J. Fluid Mech. 212 395–402
Saffman P G 1992 Vortex Dynamics (Cambridge: Cambridge University Press)
Shariff K and Leonard A 1992 Vortex rings Ann. Rev. Fluid Mech. 24 235–79
Thomson W (Lord Kelvin) 1867 On vortex atoms Phil. Mag. 34 15–24
Triantafyllou M S, Triantafyllou G S and Yue D K P 2000 Hydrodynamics of fishlike swimming Ann. Rev. Fluid
Mech. 32 33–53
Turner J S 1960 Intermittent release of smoke from chimneys Mech. Eng. Sci. 2 97–100
van den Berg C and Ellington C P 1997 The vortex wake of a ‘hovering’ model hawkmoth Phil. Trans. R. Soc. B 352
317–28
Virk D, Melander M V and Hussain F 1994 Dynamics of a polarized vortex ring J. Fluid Mech. 260 23–55
Walker J D A, Smith C R and Cerra A W 1987 The impact of a vortex ring on a wall J. Fluid Mech. 181 99–140
Widnall S E and Sullivan J 1973 On the stability of vortex rings Proc. R. Soc. A 332 335–53
Yamada H and Matsui T 1979 Mutual slip-through of a pair of vortex rings Phys. Fluids 22 1245–9

26

You might also like