The Ranque-Hilsch Effect: CFD Modeling

Vlad Bezprozvannykh
1*)
, Hank Mottl
1*)

DYCOR Technologies, 17944 106A Avenue, Edmonton, AB, Canada T5S1V3
1
Email: vlad.bez@dycor.com, hank.mottl@dycor.com


*)
This work was supported under DYCOR, Canada Internal Development Project #00023. Authors
gratefully acknowledge the support.
ABSTRACT
The phenomenon of noticeable temperature
distribution in confined steady rotating gas flows
is referred as Ranque-Hilsch effect. The simple
counter-flow Ranque-Hilsch tube consists of a
long hollow cylinder with tangential nozzles at
one end for injecting compressed gas. Rotating
gas escapes the tube through two outlets a
central orifice diaphragm placed near the
injection nozzle plane (cold stream) and a ring-
shaped peripheral outlet placed at the opposite
end of the tube (hot stream). The flow is
essentially three-dimensional, turbulent,
compressible, and spinning such that any
theoretical simplifications are questionable, if at
all possible.
Fluent suite of software was applied at Dycor
Technologies, Canada to the task of numerical
simulation of the Ranque-Hilsch effect that is
part of Dycor's program of fundamental and
applied research. The behavior of two types of
fluids in the Ranque-Hilsch tube was
investigated air and water. Three-dimensional
continuity, momentum, and energy equations
were solved for incompressible and compressible
flows for water and air cases correspondingly.
The Reynolds stress turbulence model was
originally applied to close the Reynolds-
averaged Navier-Stokes equations.
The visualization of the velocity and temperature
fields inside the vortex tube helped to understand
the details of fluid flow. It was shown that CFD
approach is applicable for simulating Ranque-
Hilsch effect. It was found that various levels of
complexity in turbulence modeling are suitable
for vortex tube analysis. No vortex effect was
observed for incompressible flow. The

The area of practical vortex effect applications is
wide ranging from the local cooling industrial
devices to mixture separation equipment. As an
energy re-distribution tool without any moving
parts, the vortex tube is an attractive instrument
for inventors and designers in spite of its low
efficiency

dependence of vortex tube cooling ability on
initial gas pressure was investigated. Numerical
simulation data are consistent with available
experimental results.

1. RANQUE-HILSCH (VORTEX) EFFECT
Ranque [1] reported the phenomenon of
noticeable temperature distribution in confined
steady rotating gas flows in 1933. Hirsch [2]
performed the detailed examination of the effect.
He also suggested the working vortex tube
design. The simple counter-flow vortex tube
consists of a long hollow cylinder with tangential
nozzles at one end for injecting compressed gas.
Rotating gas escapes the tube through two
outlets a central orifice diaphragm placed near
the injection nozzle plane (cold stream) and a
ring-shaped peripheral outlet placed at the
opposite end of the tube (hot stream).

[3]. As to the theoretical description of
processes governing the energy separation in the
vortex tube, the number of suggested theories is
comparable with the number of Ranque-Hilsch
effect researchers. Even the basic physics of the
phenomenon is still under question, as a vortex
tube can be considered a practical embodiment
of Maxwell's hypothetical demon, detaching the
molecules with high kinetic energy from
homogeneous gaseous medium [3]. Basically,
there are two schools of researchers explaining
the Ranque-Hilsch effect. The first school, the
majority, considers the interaction of
compressibility, shear stresses and turbulent
pulsations as the basis for energy separation [4].
The second school believes that the acoustic
effects cause that separation [5].


2. THE MODELING APPROACH AND
VORTEX TUBE PARAMETERS
Modern computational fluid dynamics provides
researchers with an excellent means for
analyzing complex fluid flows. CFD tools are
especially helpful in cases like flow in a vortex
tube when capabilities for modeling several
simultaneous physical processes are desirable.
The Ranque-Hilsch effect itself, on the other
hand, represents the physical situation that can
help to verify the ability of CFD software to
simulate the real world processes. The flow is
essentially three-dimensional, turbulent,
compressible, and spinning such that any
theoretical simplifications are questionable, if at
all possible.

Figure 1. Simple vortex tube with two tangential
inlet nozzles


The behavior of two types of fluids in the
Ranque-Hilsch tube was investigated air and
water. Water was modeled as incompressible
fluid with constant density. Air was considered
as a compressible fluid with the ideal gas law for
governing density-pressure-temperature
relationship. Three-dimensional continuity,
momentum, and energy equations were solved
for incompressible and compressible flows for
water and air cases correspondingly. The
Reynolds stress turbulence model was applied to
close the Reynolds-averaged Navier-Stokes
equations. The concept of Reynolds' analogy of
turbulent heat transport to turbulent momentum
transport was applied with the Reynolds stress
model for convective heat transfer modeling. It
was expected that this model of turbulence
would produce more accurate results compared
to one- or two-equation models because it does
not use the isotropic eddy-viscosity hypothesis
and thus takes into account the effects of swirl
and rapid changes in strain rate.
As to the authors' knowledge, Fluent Inc. at
present time provides one of the most advanced
universal fluid mechanics software tools able to
handle complex multi-physics of fluid and gas
flows. The Fluent suite of software was applied
at Dycor Technologies, Canada to the task of
numerical simulation of the Ranque-Hilsch effect
that is part of Dycor's program of fundamental
and applied research.
The modeled vortex tube's (Figure 1) relative
geometrical parameters are considered to be
optimal [6]: L/D = 5.5, where L is the tube
length, and D is the diameter; two inlet tubular
nozzles with their total relative area S
n
/(D
2
/4) =
0.14; the cold air relative outlet diameter d
c
/D =
0.37; the relative hot air outlet ring-shaped area
S
h
/(D
2
/4) = 0.052. Other researchers [7] believe
that different sets of geometrical parameters are
optimal.

3. THE RESULTS OF NUMERICAL
SIMULATIONS
The visualization of the velocity and temperature
fields inside the vortex tube helped to understand
the details of fluid flow. The following two
figures illustrate the distribution of air velocities
and temperatures in the interior of the Ranque-
Hilsch tube for the case of 0.5 MPa inlet
pressure.



Figure 2. Air velocity contours inside the vortex
tube
Figure 4. Distribution of the stagnation
temperature




Various levels of accuracy in turbulence
modeling for simulating Ranque-Hilsch effect
were tested. It was found that k- two-equation
turbulence model and Spallart-Allmaras
one-equation model (known as giving good
results for boundary layers subjected to adverse
pressure gradients) provide the velocity and
temperature distributions similar to those built
with Reynolds stress model of turbulence. As to
the attempts to get the vortex effect for
incompressible fluid, they are failed. For water,
no measurable temperature distribution in the
tube inlet cross-section was found.





Figure 3. Temperature distribution inside the
vortex tube


The contours of the air stagnation temperature
are shown at Figure 4. The area of minimal
energy is around the tube axis near the inlet
nozzles. The area of maximal gas energy is near
the hot outlet ring. The radial differential of the
stagnation temperature peaks at the inlet nozzles
section. This physical picture is confirmed by the
measurements of the stagnation temperature in
counter flow vortex tubes (see [7]).

Figure 5. Static pressure distribution


According to the results of numerical simulations
the following qualitative explanation of the
Ranque-Hilsch effect seems to be reasonable.
Expanding from inlet nozzle stream of
compressed air transforms into highly intensive
swirl flow with significant radial pressure
gradient (data on pressure distribution are shown
at Figure 5). The flow is turbulent, and turbulent
eddies can travel in the tube cross-section to the
center of tube as well as to the peripheral layers.
Micro-volumes of fluid traveling from the
central core of the vortex to its peripheral area
with relatively higher pressure are compressed
with corresponding heating. Fluid micro-
volumes moving to the center of the tube are
expanding with cooling. The higher values of
initial gas pressure will intensify all the
processes responsible for energy exchange and
increase the cooling ability of the vortex tube.
Figure 6 represents numerical data on the level
of increasing vortex tube cooling ability with
rising pressure at the inlet nozzle. Numerical
analysis results are consistent with original
experimental data of R.Hilsch for similar
geometry vortex tube.





Figure 6. Dependence of vortex tube cooling
ability on inlet nozzle pressure (line numerical
simulation results, points original R.Hilsch's
data


In conclusion, it is shown that CFD approach is
applicable for simulating Ranque-Hilsch effect.
The details of flow velocity and temperature
distribution are presented. It is found that various
levels of complexity in turbulence modeling are
suitable for vortex tube analysis. No vortex
effect is observed for incompressible flow. The
dependence of vortex tube cooling ability on
initial gas pressure is investigated. Numerical
simulation data are consistent with available
experimental results.


REFERENCES

1. Ranque G. Expriences sur la dtente giratoire
avec productions simultanes d'un nchappement
d'air chaud et d'un nchappement d'air froid. J.
de Physique el de la Radium 1933; 4. P.1125.
2. Hilsch R. The use of the expansion of gases in
a centrifugal field as a cooling process. The
Review of Scientific Instruments. 1947
February; 18(2). p.108.
3. Coccerill T. Thermodynamics and fluid
mechanics of a Ranque-Hilsch vortex tube.
Cambridge: University of Cambridge; 1998.
4. Deissler R, Permutter M. Analysis of the flow
and energy separation in a turbulent vortex.
International Journal of Heat and Mass Transfer.
1960; 1. p.173.
0
10
20
30
40
50
0 2 4 6 8
Pressureat inlet nozzle(atm)
M
a
x
i
m
u
m

t
e
m
p
e
r
a
t
u
r
e

d
r
o
p

(
K
)

5. Kurosaka M. Acoustic streaming of swirling
flow. Journal of Fluid Mechanics. 1982
November; 124. p.139.


6. Gupta K, Lilly D, Syrod N. Swirl flows.
England: Abacus Press; 1984.

7. Gutsol A. The Ranque effect. Physics
Uspekhi. 1977; 40(6). p.639.





An Lxperimental Perlormance Lvaluation ol Vortex Jube
P K SingL, Ncn-member
R C JatLgir, Ncn-member
D CangacLaryulu, Ncn-member
C S Crewal, Ncn-member
Tbe .crtex tube ts a stmp/e de.tce, ba.tng nc mc.tng parts, ubtcb prcduces bct and cc/d atr streams stmu/tanecus/y at
tts tuc ends /rcm a scurce c/ ccmpressed atr. Ltterature re.teu re.ea/s tbat tbere ts nc tbecry sc per/ect, ubtcb gt.es tbe
satts/actcry exp/anattcn c/ tbe .crtex tube pbencmencn as exp/atned by .artcus researcbers. Tbere/cre, tt uas tbcugbt tc
carrycut expertmenta/ tn.esttgattcns tc understand tbe beat trans/er cbaractertsttcs tn a .crtex tube uttb respect tc
.artcus parameters /tke mass //cu rates c/ cc/d and bct atr, nczz/e area c/ tn/et ccmpressed atr, cc/d crt/tce area, bct end
area c/ tbe tube, and LD
T
rattc. Tbe expertmenta/ tn.esttgattcns uere carrted cut based upcn tuc destgns, te,
maxtmum temperature drcp tube destgn and maxtmum ccc/tng e//ect tube destgn. It ts cbser.ed tbat tbe e//ect c/ nczz/e
destgn ts mcre tmpcrtant tban tbe cc/d crt/tce destgn tn getttng btgber temperature drcps. Cc/d /racttcn as ue// as
adtabattc e//tctency are mcre tn//uenced by tbe stze c/ tbe cc/d crt/tce ratber tban tbe stze c/ tbe nczz/e. Htgber
temperature drcps are cbtatned tn .crtex tube made c/ maxtmum temperature drcp tube destgn, ubereas, mcre cc/d
/racttcn and btgber adtabattc e//tctency are cbtatned uttb maxtmum ccc/tng e//ect tube destgn. Lengtb c/ tbe tube bas
nc e//ect cn tbe per/crmance c/ tbe .crtex tube tn tbe range c/ 45 tc 55 (LD
T
rattc.
Keyuords: Vortex tube, Heat transler cLaracteristics, ozzle diameter, Cold orilice diameter, Cold and Lot air temperatures
P K Singh is with Department ot Mechanical Engineering, Government
Polytechnic GTB Garh, Moga, Pun|ab, K G Tathgir and D Gangacharyulu
are with Mechanical and Chemical Engineering Departments, respectively,
Thapar Institute ot Engineering and Technology, Patiala 147 001, while
G S Grewal was with Thapar Centre tor Industrial Kesearch and
Development, Patiala 147 001.
JLis paper was received on August 19, 2002. Written discussion on tLe paper
will be entertained till April 30, 200+.
Vc/ 84, january 2004 1+9
OTATIO
A
C
, A
H
: areas ol tLe cold orilice and at tLe Lot end,
respectively, mm
2
A
N
, A
T
: areas ol tLe nozzle and tLe tube,
respectively, mm
2
D
C
, D
N
, D
T
: diameters ol tLe cold orilice, tLe nozzle and
tLe tube, respectively, mm
L : lengtL ol tLe tube, mm
m
C
, m
H
: mass llow rates ol cold and Lot air,
respectively, kg/s
m
t
: mass llow rate ol air at inlet nozzle, kg/s
T
a
: ambient temperature,

C
T
C
, T
H
: temperature ol air at cold and Lot ends,
respectively,

C
T
c
: temperature ol air inside tLe compressor
tank,

C
.
C
, .
H
: velocity ol tLe air at cold orilice and at Lot
end, respectively, m/s
T
C
: temperature drop in vortex tube,

C
T
C
: temperature drop due to adiabatic
expansion,

C

C H
, : density ol air at cold orilice and at Lot end,
respectively, kg/m
3

1
: A
N
/A
T

2
: A
C
/A
T

3
: L/D
T

+
: Lot end area/tube area, A
H
/A
T

1
: m
C
/m
t

2
:
1
T T
C C
/

: specilic Leat ratio

ITKODLCTIO
JLe vortex tube also known as HilscL or Ranque tube
1,2,3,9,11,13
is a simple device Laving no moving parts, wLicL produces Lot
and cold air streams simultaneously at its two ends lrom a
source ol compressed air. It consists ol a long tube Laving a
tangential nozzle near one end and a conical valve at tLe otLer
end, as sLown scLematically in Iigure 1.
A diapLragm called cold orilice, witL a suitable sized Lole in
its centre is placed immediately to tLe lelt ol tLe tangential
150 IL (I jcurna/-/C
inlet nozzle. JLe compressed air is tLen introduced into tLe
tube tLrougL tLis nozzle. JLe tangential llow imparts a
wLirling or vortex motion to tLe inlet air, wLicL subsequently
spirals down tLe tube to tLe rigLt ol tLe inlet nozzle. Conical
valve at rigLt end ol tLe tube conlines tLe exiting air to regions
near tLe outer wall and restricts it to tLe central portion ol tLe
tube lrom making a direct exit. JLe central part ol tLe air
llows in reverse direction and makes exit lrom tLe lelt end ol
tLe tube witL sizeable temperature drop, tLus creating a cold
stream. JLe outer part ol tLe air near tLe wall ol tLe tube
escapes tLrougL tLe rigLt end ol tLe tube and is lound to Lave
temperature LigLer tLan tLat ol inlet air.
Literature review
1-15
reveals tLat tLere is no tLeory so perlect,
wLicL gives tLe satislactory explanation ol tLe vortex tube
pLenomenon as explained by various researcLers. JLerelore,
it was tLougLt to carryout experimental investigations lor
understanding tLe Leat transler cLaracteristics in a vortex
tube. An attempt Las been made to study tLe ellect ol various
parameters like mass llow rates ol cold and Lot air, nozzle area
ol inlet compressed air, cold orilice area and Lot end area ol
tLe tube. JLe study also investigates tLe ellect ol L/D
T
ratio
on tLe perlormance ol tLe vortex tube.
DESIG AD COSTKLCTIOAL IEATLKES
In general, tLere are two design leatures associated witL a
vortex tube, namely, (a) maximum temperature drop vortex
tube design lor producing small quantity ol air witL very low
temperatures, and (b) maximum cooling ellect vortex tube
design lor producing large quantity ol air witL moderate
temperatures. JLese two design considerations Lave been used
in tLe present study. JLe parameters investigated in tLe study,
to understand tLeir inter-relationsLips and tLeir ellect on tLe
perlormance ol tLe vortex tube, are:

nozzle diameter,

cold orilice diameter,

mass llow rates ol cold and Lot air,

lengtL ol tLe tube, and

area at tLe Lot end.

In tLe present investigation, a nozzle area to tube area ratio ol
0.11 _ 0.01 lor maximum temperature drop and a ratio ol
0.08+ _ 0.001 lor acLieving maximum elliciency Las been
considered, as suggested by Soni and JLomson
9
. IurtLer, tLey
suggested tLat tLe ratio ol cold orilice area to tube area
sLould be 0.080 _ 0.001 lor acLieving maximum temperature
drop and it will be 0.1+5 _ 0.035 lor attaining tLe maximum
elliciency. Also, tLe same researcLers suggested tLat tLe lengtL
ol tLe vortex tube sLould be greater tLan +5 times tLe tube
diameter but no upper limit was specilied.
Keeping tLese suggestions in view, a vortex tube ol size 15.30 mm
diameter ol PVC was selected. JLe casing ol tLe vortex tube
was a 6+ mm long CI cylinder. Jwo sets ol nozzles were made
and eacL set consisted ol two nozzles. JLe lirst set ol nozzles
Lad a 3.1 mm diameter Lole and was +0 mm long, wLereas tLe
second set nozzles Lad a 3.5 mm diameter Lole and was 38 mm
long. Similarly, two cold orilices were used Laving Lole
diameters ol +.5 mm and 5.8 mm, respectively. A conical valve
made ol mild steel was provided on tLe rigLt Land side ol tLe
tube to regulate tLe llow. Also, a mulller was provided to
reduce tLe noise levels at tLe Lot end.
EXPEKIMETAL SETLP
Jo study tLe ellect ol various parameters sucL as mass llow
rates ol cold and Lot air, nozzle area, cold orilice area, Lot end
area and L/D
T
ratio on tLe perlormance ol tLe vortex tube, an
experimental set up was prepared as per design described
earlier. JLe vortex tube components were made in a manner
sucL tLat tLe geometry ol tLe tube could be cLanged lrom
maximum temperature drop tube design to maximum cooling
ellect tube design. JLe line diagram ol tLe experimental setup
is sLown in Iigure 2. JLe temperature ol air at cold and Lot
ends was measured witL digital tLermometers witL an
accuracy ol
_
0.1

C . JLe air velocities were measured made

by an anemometer. JLe pressures were measured by tLe
Bourden tube pressure gauges. All tLe instruments were
calibrated belore tLe measurements were actually observed/
recorded.
JLrougLout tLe experiment, tLe inlet compressed air pressure
was maintained at +.1 bar. JLe operating parameters noted
during tLe experiment lor eacL vortex tube design were:

air pressure in compressor receiver, bar,

air pressure near inlet to vortex tube, bar,

compressed air temperature,

C ,

cold air temperature,

C ,

Lot air temperature,

C ,

cold air velocity, m/s,

Lot air velocity, m/s,

Lot end area, mm

2
, and

cold orilice area, mm

2
.
TEST PKOCEDLKE
JLe various parameters, wLicL allect tLe perlormance ol tLe
vortex tube, were divided into lollowing input and output
variables:
Input variables: Ratio ol nozzle area to tube area (
1
), ratio
ol cold orilice area to tube area (
2
), L/D
T
ratio ol tLe
tube (
3
) and ratio ol Lot end area to tLe tube area
(
+
).
Iigure 1 Schematic diagram ot a vortex tube
Compressed air
Cold air
Vortex Tube Hot air
Hot air
Throttle
valve
Regulator
Vc/ 84, january 2004 151
Cutput variables: Jemperature ol cold air (T
C
), cold
lraction (
1
), and adiabatic elliciency (
2
).
JLe compressor was initially run lor about 30 min to get a
stable compressor air tank pressure ol 5.1 bar and a line
pressure ol +.1 bar. Iirstly, two sets ol nozzle diameters were
cLosen. Iirst value represented tLe nozzle size ol tLe
'maximum temperature drop tube design` wLereas tLe second
value represented tLe nozzle ol tLe 'maximum cooling ellect
tube design`. JLen lor eacL set ol nozzles, two values ol cold
orilices were cLosen as given by tLe two design criteria
described earlier. ow lor eacL cold orilice area, tLree lengtLs
ol tLe tubes were selected. IurtLer, lor eacL lengtL ol tube lour
Lot end openings were cLosen. Ior eacL reading, tLe value ol
one input variable was cLanged, keeping tLe otLer input
variables constant. Jwo readings ol eacL output variable were
taken lor eacL run ol tLe experiment. JLe values ol density ol
air at botL ends were calculated assuming tLe llow to be
adiabatic.
KESLLTS AD DISCLSSIO
Since, tLe study was conducted based upon two tube-designs,
tLe ellect on output variables needed to be concluded. JLree
output variables, namely, temperature ol cold air T
C
, cold
lraction (
1
) and adiabatic elliciency (
2
) were evaluated and
plotted lor dillerent tube designs against tLe Lot end area.
Also, tLe ellects ol parameters sucL as nozzle area and cold
orilice area on tLe output variables were studied by mixing tLe
two tube designs.
Iigure 3 sLows tLe variation ol temperature ol cold air (T
C
),
witL respect to cLange in tLe Lot end area lor lour design
conditions. It sLows tLe variation ol (T
C
) lor a L/D
T
ratio ol
+5 and it is observed tLat tLe temperature ol cold air in all tLe
cases decreases as tLe Lot end is opened, wLicL means as tLe
amount ol cold air is reduced its temperature gets lowered. It is
also observed tLat temperatures ol cold air in Case 1 were
LigLest wLereas in Case 2 it was minimum. It clearly indicates
tLat lor obtaining low temperatures, nozzle as well as cold
orilice sLould be ol 'maximum temperature drop tube design`.
JLe temperatures ol cold air observed in Case 3 (wLere nozzle
is ol 'maximum temperature drop tube design`) were LigLer as
compared to tLat ol Case + (wLere nozzle is ol 'maximum
cooling ellect tube design`) lor all tLe values ol Lot end area. It
indicates tLat tLe ellect ol nozzle area is more prominent in
getting LigLer temperature drops. Similar, ellects were
observed lor vortex tube Laving L/D
T
ol 50 and 55.
Iigure + sLows tLe variation ol cold lraction (
1
) witL respect
to cLange in Lot end area lor lour design conditions as
explained earlier. JLe ligure Las been plotted lor L/D
T
- +5.
It is observed lrom tLe grapL tLat tLe cold lraction is LigLest in
tLe Case 2 (wLere nozzle and cold orilice were ol 'maximum
cooling ellect tube design`). Also, tLe cold lractions observed
in Case 3 were LigLer as compared to tLat ol tLe Case +. It
Case 1 . nozzle ot 'maximum temperature drop tube design' , cold oritice ot
'maximum temperature drop tube design', Case 2 . nozzle ot 'maximum
cooling ettect tube design' , cold oritice ot 'maximum cooling ettect tube
design', Case 3 . nozzle ot 'maximum temperature drop tube design' , cold
oritice ot 'maximum cooling ettect tube design', and Case 4 . nozzle ot
'maximum cooling ettect tube design' , cold oritice ot 'maximum
temperature drop tube design'
Iigure 3 Variation ot temperature ot cold air with respect to change in
hot end area
Performance of Vortex Tube
Case 1 . nozzle ot 'maximum temperature drop tube design' , cold oritice ot
'maximum temperature drop tube design', Case 2 . nozzle ot 'maximum
cooling ettect tube design' , cold oritice ot 'maximum cooling ettect tube
design', Case 3 . nozzle ot 'maximum temperature drop tube design' , cold
oritice ot 'maximum cooling ettect tube design', and Case 4 . nozzle ot
'maximum cooling ettect tube design' , cold oritice ot 'maximum
temperature drop tube design'
Iigure 4 Variation ot cold traction with respect to change in hot end area
.
Performance of Vortex Tube
(Hot end area/Tube area)
P
1
. air pressure in compressor receiver, kg/cm
2
, P
2
. air pressure near inlet
to vortex tube, kg/cm
2
, T
a
. compressed air temperature,

C
, T
C
.
cold air temperature,

C
, T
H
. hot air temperature,

C
, u
C
. cold
air velocity, m/s, u
H
. hot air velocity, m/s
Iigure 2 Line diagram ot the experimental setup
P
1
T
a
T
H
V
H
P
2
T
C
V
C
Vortex tube
(Hot end area/Tube area)
T
C
C
o
l
d

a
i
r

t
e
m
p
e
r
a
t
u
r
e
,

C

C
o
l
d

f
r
a
c
t
i
o
n

1
152 IL (I jcurna/-/C
indicates tLat tLe cold lraction is more inlluenced by tLe size
ol tLe cold orilice ratLer tLan tLe size ol tLe nozzle. Similar,
ellects Lave been observed in case ol vortex tube Laving L/D
T
ol 50 and 55.
Iigure 5 sLows tLe variation ol adiabatic elliciency (
2
) witL
respect to cLange in Lot end area lor lour design conditions.
JLe ligure sLows similar pattern as observed in tLe case ol cold
lraction. It is observed lrom tLe ligure tLat adiabatic elliciency
is LigLest in tLe Case 2 (wLere nozzle and cold orilice were ol
'maximum cooling ellect tube design`). Also, tLe adiabatic elli-
ciencies observed in Case 3 were LigLer as compared to tLose
observed in Case +. It indicates tLat adiabatic elliciency is also
inlluenced by tLe size ol tLe cold orilice ratLer tLan tLe size ol
tLe nozzle. Similar, ellects Lave been observed lor vortex tube
Laving L/D
T
ol 50 and 55.
Iigure 6 sLows tLe variation ol temperature ol cold air (T
C
)
witL respect to cLange in tLe Lot end area lor tLe tLree cases.
Case 1 is lor L - +5 D
T
, Case 2 is lor L - 50 D
T
, and Case 3 is
lor L - 55 D
T
. It is observed tLat all tLe cases sLow very
similar trends, wLicL indicate tLat tLe lengtL ol tLe tube Las
no ellect on tLe perlormance ol tLe tube wLen tLe lengtL is
increased beyond +5 D
T
up to 55 D
T
. Similar, results Lave been
observed lor cold lraction (
1
) (Iigure /) and lor adiabatic
elliciency (
2
) (Iigure 8).
All tLe grapLs ol Case 2 and Case 3 sLow tLe start ol reversed
trend wLen tLe Lot end opening is increased beyond 20% ol
tLe tube cross sectional area.
COCLLSIOS
1. JLe ellect ol nozzle design is more important tLan tLe cold
orilice design in getting LigLer temperature drops.
2. Cold lraction as well as adiabatic elliciency is more inllu-
enced by tLe size ol tLe cold orilice ratLer tLan tLe size ol tLe
nozzle.
3. HigLer temperature drops are obtained in vortex tube made
ol maximum temperature drop tube design, wLereas, more
cold lraction and LigLer adiabatic elliciency are obtained witL
maximum cooling ellect tube design.
Case 1 . nozzle ot 'maximum temperature drop tube design' , cold oritice ot
'maximum temperature drop tube design', Case 2 . nozzle ot 'maximum
cooling ettect tube design' , cold oritice ot 'maximum cooling ettect tube
design', Case 3 . nozzle ot 'maximum temperature drop tube design' , cold
oritice ot 'maximum cooling ettect tube design', and Case 4 . nozzle ot
'maximum cooling ettect tube design' , cold oritice ot 'maximum
temperature drop tube design'
Iigure 5 Variation ot adiabatic etticiency with respect to change in hot
end area
nozzle ot 'maximum temperature drop tube design' , cold oritice ot
'maximum temperature drop tube design'
Case 1 . L/D
T
= 45 , Case 2 . L/D
T
= 50, and Case 3 . L/D
T
= 55
Iigure 6 Variation ot cold air temperature with respect to change in hot
end area
nozzle ot 'maximum cooling ettect tube design' , cold oritice ot
'maximum cooling ettect drop tube design'
Case 1 . L/D
T
= 45, Case 2 . L/D
T
= 50, and Case 3 . L/D
T
= 55
Iigure 7 Variation ot cold traction with respect to change in hot end area
Performance of Vortex Tube
Performance of Vortex Tube
nozzle ot 'maximum cooling ettect tube design' , cold oritice ot
'maximum cooling ettect drop tube design'
Case 1 . L/D
T
= 45, Case 2 . L/D
T
= 50, and Case 3 . L/D
T
= 55
Iigure 8 Variation ot adiabatic etticiency with respect to change in hot
end area
(Hot end area/Tube area)
A
d
i
a
b
a
t
i
c

e
f
f
i
c
i
e
n
c
y
(Hot end area/Tube area)

C
o
l
d

f
r
a
c
t
i
o
n
A
d
i
a
b
a
t
i
c

e
f
f
i
c
i
e
n
c
y
(Hot end area/Tube area)

2
Performance of Vortex Tube
(Hot end area/Tube area)
T
C
C
o
l
d

a
i
r

t
e
m
p
e
r
a
t
u
r
e
,

+
Vc/ 84, january 2004 153
+. LengtL ol tLe tube Las no ellect on tLe perlormance ol tLe
tube wLen it is increased beyond +5 D
T
up to 55 D
T
.
5. Jwo design conditions (Case 1 and Case 2) sLow a start ol
reverse trend wLen tLe Lot end area is increased beyond 20%
ol tLe tube area.
ACKOWLEDGEMETS
JLe autLors express tLeir gratitude to Dr S C Saxena,
Director, JLapar Institute ol Lngineering and JecLnology
and Dr M P Kapoor, Director, JLapar Centre lor Industrial
R&D, Patiala lor providing necessary lacilities and
encouragement. JLe autLors also express tLeir sincere tLanks
to Mr SoLan Lal, Laboratory Superintendent and Ms Anu
ParasLar, Cllice-cum-JecLnical Assistant lor tLeir continuous
Lelp in experimentation and preparation ol tLis manuscript,
respectively.
KEIEKECES
1. C ] Ranque. 'Lxperiments on Lxpansion in a Vortex witL Simultaneous
LxLaust ol Hot and Cold Air.` Le jcurna/ De Pbystque, et /e Radtum (Parts,
vol +, ]une 1933, pp 1125-1130.
2. R HilscL. 'JLe use ol tLe Lxpansion ol Aires in a Centrilugal Iield as a
Cooling Process.` Re.teu c/ Sctentt/tc Instruments, vol 13, Iebruary 19+/,
pp 108-113.
3. C D Iulton. 'Ranque Jube.` jcurna/ c/ tbe ASRL, Re/rtgerattcn Lngtneertng,
vol 58, May 1950, pp +/3-+/9.
+. C W ScLeper ( ]r). 'JLe Vortex Jube Internal Ilow Data and a Heat Jransler
JLeory.` jcurna/ c/ tbe ASRL, Re/rtgerattcn Lngtneertng, vol 59, Cctober 1951,
pp 985-989.
5. ] P Hartnett and L R C Lckert. 'Lxperimental Study ol tLe Velocity and
Jemperature Distribution in a HigL Velocity Vortex Jube Ilow.` Transacttcns
c/ AS/L, vol /9, May 195/, pp /51-/58.
6. B B Parulekar. 'JLe SLort Vortex Jube.` jcurna/ c/ Re/rtgerattcn, vol +, 1961,
pp /+-80.
/. R B Aronson. 'JLe Vortex Jube: Cooling witL Compressed Air.` jcurna/ c/
/acbtne Destgn, December 19/6, pp 1+0-1+3.
8. ] L Lay. 'An Lxperimental and Analytical Study ol Vortex Ilow and
Jemperature Separation by Superposition ol Spiral and Axial Ilow, Part 1 and
Part 2.` AS/L jcurna/ c/ Heat Trans/er, vol 81, no 3, August 1959, pp 316-31/.
9. Y Soni and W ] JLomson. 'Cptimal Design ol Ranque - HilscL Vortex
Jube.` AS/L jcurna/ c/ Heat Trans/er, vol 9+, no 2, May 19/5, pp 316-31/.
10. K Landecker. 'A Jwo Stage Relrigeration and Power Producing
Arrangement Consisting ol a Vortex Cooling Jube and a JLermoelectric
Stage.` Lnergy Ccn.erstcn, vol //, no 3, 19//, pp 119-122.
11. R L Collins and R B Lovelace. 'Lxperimental Study ol Jwo PLase Propane
Lxpanded tLrougL tLe Ranque - HilscL Jube.` AS/L jcurna/ c/ Heat Trans/er,
vol 101, no 2, May 19/9, pp 300-305.
12. H JakaLama and H Yokosawa. 'Lnergy Separation in Vortex witL a
Divergent CLamber.` Transacttcns c/ AS/L, vol 103, May 1981, pp 196-203.
13. M Kurosaka. 'Acoustic Streaming in Swirling Ilow and tLe Ranque -
HilscL Lllect.` jcurna/ c/ I/utd /ecbantcs, 1982, p 139.
1+. Y D Raiskii and L L Jankel. 'Inlluence ol Vortex Jube Conliguration and
LengtL on tLe Process ol Lnergetic Air Separation.` jcurna/ c/ Lngtneertng
Pbystcs, vol 2/, no 6, December 19/+, pp 15/8-1581.
15. V I Metenin. 'Investigations ol Vortex Jube Jype Compressed Air
Separators.` Sc.tet Pbystcs - Tecbntca/ Pbystcs, vol 5, 1961, pp 1025-1032.