International Conference on Mechanical, Industrial and Energy Engineering 2018

23-24 December, 2018, Khulna, BANGLADESH

ICMIEE18-104
The Effect of Swirl on Array of Turbulent Impinging Jets
Sudipta Debnath*, Md Habib Ullah Khan, Zahir U. Ahmed, Md. Mahmodul Alam
Department of Mechanical Engineering, Khulna University of Engineering & Technology, Khulna-9203, BANGLADESH

ABSTRACT
Impinging jets are widely used for their effective heat and mass transfer for several decades. Arrays of jet impingement have
also been studied before due to its practical relevance to electronics cooling. A number of jet variations and jet-to-jet
orientations have previously been studied, mainly to further improve the magnitude and uniformity of heat transfer. In recent
years, swirling jets has also gained interest in heat transfer application due to their inherent mixing and spreading
characteristics, which is believed to be an improvement on overall heat and mass transfer. As such, this paper numerically
investigates an array of circular jets with and without swirl that impinges vertically onto a flat surface located at a fixed vertical
distance H = 2D and at Reynolds number equals to 11,600, where D is the nozzle diameter. As the entire was symmetric, only
quarter of the model was constructed for numerical analysis to save computational cost. In this case, numerical calculations
were done via commercial software package ANSYS Fluent using SST k-ω turbulence model. Inlet conditions were taken from
experimental data. The jet flows were in downward direction and gravity was not considered. This paper also compares
numerical predictions with previously published literature for non-swirling and swirling jets.

Keywords: Swirl jet, Impingement, Turbulence, Heat transfer, CFD.

1. Introduction operations to reduce fracture and to improve grain

Jet impingement is one of the most effective media for growth for higher strength [4].
removing heat from heated surfaces with high heat flux. In order to experiment that whether increasing the
The high heat removal rate of jet impingement had number of jets increases the heat transfer or not jet
gained a prior position of research topic among the arrays were studied by several researchers during the
researchers for several decades. Impinging jets are past decades. Jet interference in impingement arrays is
widely used in various engineering and industrial the main striking difference compared to a single jet. It
applications, such as cooling of turbine blades and is reported that in case of array impinging jet the
micro electric components, quenching and annealing of nozzles spacing of from 4-6 diameters results best heat
non-ferrous sheet metals, tempering of glass, freezing of transfer results [5]. Results of many investigations on
tissues in cryosurgery [1]. circular non-swirling jets are in a good agreement that
Jet is a stream of fluid that is projected into a single jet yields better heat transfer than that of array
surrounding medium, usually from some kind of a impinging jets [6]. The central jet has the shortest core
nozzle, aperture or orifice. Jets can travel long distances and the highest kinetic energy due to higher number of
without dissipating. Swirl means to move with an neighboring jets and the reverse is true for peripheral
eddying or whirling motion. jets [7]. For larger jet-to-jet distance regardless of jet-to-
In majority of the previous research, orthogonal jet plate distances the cross flow doesn’t disturb the fluid
impingements onto flat surfaces or plates from which motion of the neighboring jets. It had also been reported
heat transfer occurs are primarily focused. The jet flows that the jet-to-jet distance is the major factor whereas
emanating from a nozzle may either be non-swirling or jet-to-plate is the minor [8]. The above results are valid
swirling, with their own pros and cons. Jets can also be only for non-swirling jets. As such, it would be
categorized into circular jet, slot jet, inclined jet etc interesting to examine swirling flows in arrays, as single
depending on the orifice opening or geometric swirling jets were found to have rather different
arrangements. Their effects on fluid flow and heat behaviors.
transfer behavior are also different. For example, heat In this paper, an array of circular air jets is studied to
transfer by convection from a hot gas jet to a plane determine the effect of swirl on heat transfer during
surface was observed to increase as twice as the initial cooling by an array of swirl jets. The effect of swirl jets
value with the change of axis-symmetric angle of the jet that impinged vertically onto a flat surface located at a
from 15°-90° [2]. Again, for the same flow rate the fixed vertical distance H/D = 2 and at Reynolds number
circular jet yielded 8% higher heat transfer than the slot equals to 11,600 were studied. The widely used k-ω
jet [3]. In comparison to non-swirling jet, swirling jets model does not properly represent the flow features and
were examined by many researchers. Research out highly over predicts the rate of heat transfer and yields
comes suggest that swirling jets are beneficial over non physically unrealistic behavior [9]. SST k-ω turbulence
swirling jets. It had been found that swirling jet has model is a combination of the k-epsilon in the free
higher Nusselt number and better uniformity in heat- stream and the k-omega models near the walls. It does
transfer which is the prior criteria in many cooling not use wall functions and tends to be most accurate

* Corresponding author. Tel.: +88-01759999904

E-mail addresses: ssudipta69@gmail.com
when solving the flow near the wall. Hence this model conditions in all nozzles. The fluid inlet temperature is
had been used to calculate the results in this analysis. set to ambient (300K). Outlet boundary was set as
pressure outlet with backflow turbulent intensity 5%
and backflow turbulent viscosity ratio 10 similar to the
2. Numerical Methodology numerical simulation of Ahmed et al. [12]. Symmetry
An array of 25 circular nozzles were considered for boundary condition was applied in two side surfaces for
developing the swirling and non-swirling impinging jets. non-swirling jets and periodic boundary conditions for
The nozzles are arranged in inline and staggered swirling jets. In this regard, corresponding surfaces
arrangements. All of them were axi-symmetric, equally were match controlled during meshing in order to
spaced in 3 circles around a central nozzle in which the impose periodic boundary. The pressure velocity
diameter of first circle is 5D, the second circle 9D and coupling was solved using the coupled solver with
the third circle 13 D. The diameter of each nozzle is 40 Green-Gauss Cell special discretization for gradients,
mm, the distance between the centers of two PRESTO for pressure and second order upwind was
corresponding nozzle is 80 (2D) mm while the distance used for momentum, turbulent kinetic energy, specific
between them is 40 (D) mm. As the entire two models dissipation rate and energy. All the residuals were set to
were symmetric, only quarters of the models were 10-05 for accuracy except energy to 10-06. Mesh
constructed for numerical analysis to save consisting of 898k nodes was used to predict the final
computational cost. Jet-to-plate distance was H=2D and result. Inflation was applied with 15 layers and growth
the radial extent of circular plate is considered 16 D. rate of 1.2 near the plate region and the mesh was
generated from fine mesh near the axis to coarser mesh
in the radially outward direction.

3. Data Validation

Fig.1 Top view of the array of nozzles

Fig.3 Data validation of Cp vs R/D chart

Fig. 3 presents a comparison of the effect of swirl on the

radial distributions of (impingement surface) static
pressure for S=0 and 0.74 and H=2D at Re=11600
between the current simulation data and experimental
Fig.2 Front view of the array of nozzles data derived from the literature. The current numerical
data is found to be in good agreement with the literature
For numerical simulation, SST k-ω viscous model is except for swirl number 0.74 Cp decreases (with radial
used for solving the problem. Air was taken as fluid distance) from a maximum at the stagnation point and
with the following properties: Density-1.225 kg/m3, shows a Gaussian-like distribution, with maximum Cp
Specific Heat-1006.43 J/kg-K, Thermal Conductivity- at the center line similar to non-swirling flows. The
0.0242 W/m-K and Viscosity-1.78e-05 kg/m-s. Inlet radial position of the maximum Cp moves outward due
conditions of the swirling impinging jets were taken to the higher centrifugal effects as swirl number
from Ahmed et al. [11] in which swirling jet was increases. The reduction is largely attributed to the axial
produced with an aerodynamic swirl generator. The deceleration of the swirl flow.
profile of the data was set as the inlet boundary

ICMIEE18-104- 2
4. Results and Discussion
Fig. 4 shows the Cp variation for non-swirl and swirl Fig. 5 represents the velocity streamline of array
inline jets at impinging plane along the radial line in impinging jet for non-swirl and swirl flow for both
horizontal direction where r/D=0 indicates the origin inline and staggered arrangements of jets. The origin
and r/D=16 indicates the end of the fluid body. It is (0,0) corresponds to the center and the flow is
observed from the figure that for non-swirl, reductions distributed radially between 0 to 16D.
of Cp occur particularly around the impingement
regions between the jets and gradually decreases in
between 6D to 16D.

Fig.5 Velocity streamlines for S= 0 and 0.74 at inline

arrangement of nozzles

It is seen that a recirculation region appears between

two jets for both non-swirling and swirling flows. For
non-swirling jets, the jets create a fountain, with an
upward movement of the flow, whereas in swirling jets
such upward fountain is not evident. Rather, the flow
passes between recirculating zones, which occurs
between the first two jets from the center and between
the jets near the wall. It is also clear from the above
streamlines that for non-swirl flow the jet strikes the
heated plate whereas swirl jets deviates from well
before the impingement surface resulting a moderate
recirculation around the stagnation zone. It is clear that
from 4.5 D to 16 D both the streamline of swirl flow
and non-swirl flow are closely attached to the heated
plate. Depending on the distance or proximity of the
neighboring jet the flow field may vary with significant
changes which may affect the heat transfer distribution
over the impingement surface. The results for velocity
streamlines ensure that the strength of recirculation for
swirl flow is greater than that of non-swirl flow and the
distance between the neighboring jets plays a great role
Fig.4 Cp for S=0 and S=0.74 at inline arrangement of in this recirculation effect.
nozzles

For swirl it is observed that gradual increasing and Fig. 6 shows contour map of convective heat transfer
reduction of Cp happened in between impinging region coefficient (h) over the range S=0 and S=0.74 in near-
(r/D <8). The radial position of the maximum Cp moves field (H=2D) impingement for the case of inline
outward due to the higher centrifugal effects as swirl arrangement.
number increases. The results show that pressure is
found to be the maximum in the stagnation regions
where the flow strikes the plate and then it reduces fast
both radially and circumferentially. For swirl flow the
pressure is relatively more uniformly distributed over
the impingement plate than non-swirl flow.

ICMIEE18-104- 3
 5. Conclusion
This paper describes the behaviors of an array of
incompressible turbulent impinging air jets for both
non-swirl and swirl flows. The governing equations are
solved using a commercial software package Ansys
Fluent v16.2.by using a turbulence model SST k- ω. The
study is performed for two conditions: non-swirling (S =
0) and highly swirling (S = 0.74) for a nozzle-to-plate
distance of 2D at Re=11,600. Velocity streamline,
pressure distribution, heat transfer distribution are
investigated for both of these arrangements separately.
The results for velocity streamlines ensure that the
strength of recirculation for swirl flow is greater than
that of non-swirl flow. The distance between the
neighboring jets plays a great role in this recirculation
(a) Inline non-swirl effect. The pressure distribution indicated a more
uniform distribution for swirl impinging arrays than its
non-swirling counterpart. This distribution also plays a
great role for achieving a better result in the goal of
impingement cooling. Convective heat transfer
coefficient distribution showed the larger heat transfer
zones in the stagnation zone for non-swirling jets and
between the neighboring jets for swirling jets.

NOMENCLATURE
Cp : Coefficient of pressure
P : Pressure, Pa
S : Swirl Number
h : Convective heat transfer coefficient
R : Radial distance, m
T : Temperature, K
(b) Inline swirl
D : Diameter of nozzle, m
H : Distance between nozzle tip and plate, m
Fig.6 Convective heat transfer coefficient for S= 0 and
0.74 at inline arrangement of nozzles Re : Reynolds Number

For S=0, a higher convective heat transfer coefficient

(h) band appears outside the jet center with a lower h REFERENCES
zone occurring immediately at the jet center. Outside
the periphery of this higher h zone, another (outer) low [1] Behnia, M., Parneix, S., and Durbin, P.,
h zone occurs, before heat transfer peaks up again at Accurate modeling of impinging jet heat
even further radial distances. For the swirling jets, a transfer, Center for Turbulence research,
strikingly difference heat transfer behavior is predicted. Annual Research Briefs, pp. 149-164, 1997
In this regard, maximum heat transfer occurs between
the neighboring nozzles, not below the nozzle as in the [2] Womac, D. J., Ramadhyani, S., and Incropera,
case of non-swirling jet arrays. Additionally, the lowest F. P., Correlating equations for impingment
heat transfer occurs under the central jet, which perhaps cooling of small heat sources with single
due to the recirculation seen in the velocity streamline. circular jets, Journal of Heat Transfer, vol.
The occurrence of relatively two maxima in convective 115(1), pp. 106-155, 1993
heat transfer coefficient (inner and outer) have been [3] Fitzgerald, J. A., and Garimella, S. V., Flow
attributed to the rapid change of radial velocity in the field effects on heat transfer in confined jet
streamline deflection region and to destruction of the impingement, Oceanographic Literature
thermal boundary layer by the large-scale eddies which Review, V. 1(45), pp. 199, 1998
strikes the surface. The minima in h around the jet
center may also be attributed to the weak penetration of [4] Cooper, D., Jackson, D.C., Launder, B.E., Liao,
shear layer induced turbulence, which is particularly g.x. Impinging jet studies for turbulence model
true at H = 2D due to this relatively small impingement assessment. Part I: Flow-field experiments, Int.
distance. J. Heat Mass Transfer, Vol. 36, pp. 2675-2684,
1993

ICMIEE18-104- 4
[5] Brevet, P., Dejeu, C., Dorignac, E., Jolly, M.,
and Vullierme, J., J., Heat transfer to a row of
impinging jets in consideration of optimization,
International Journal of Heat and Mass
Transfer, vol. 45(20), pp. 4191-4200, 2002
[6] Viskanta, R., Heat transfer to impinging
isothermal gas and flame jets, Experimental
Thermal and Fluid Science, vol. 6(2), pp. 111-
134, 1993
[7] Geers, L. F., Tummers, M. J., and Hanjalic´, K.,
Experimental investigation of impinging jet
arrays, Experiments in Fluids, vol. 36(6), pp.
946-958, 2004
[8] Wienand, J., Riedelsheimer, A., and Weigand,
B., Numerical study of a turbulent impinging
jet for different jet-to-plate distances using
two-equation turbulence models, European
Journal of MechanicsB/Fluids, vol. 61, pp.
210-217, 2017
[9] Behnia, M., Parneix, S., and Durbin, P. A.,
Prediction of heat transfer in an axisymmetric
turbulent jet impinging on a flat plate,
International Journal of Heat and Mass
Transfer, vol. 41(2), pp. 1845-1855, 1998
[10] Geers, L. F., Multiple impinging jet arrays: an
experimental study on flow and heat transfer,
PhD Thesis, Technical University, Delft , The
Netherlands, 2004
[11] Ahmed, Z. U., Al-Abdeli, Y. M., and Guzzomi,
F. G., Impingement pressure characteristics of
swirling and non-swirling turbulent jets, vol.
68, Experimental Thermal and Fluid Science,
pp.722–732, 2015
[12] Ahmed, Z. U., Al-Abdeli Y. M., and
Matthews, M. T., The effect of inflow
conditions on the development of non-swirling
versus swirling impinging turbulent jets, vol.
118, Computers & Fluids, pp. 255-273, 2015

ICMIEE18-104- 5