AN INVESTIGATION ON HVLS FAN PERFORMANCE WITH DIFFERENT BLADE CONFIGURATIONS

Abstract

High-volume low-speed (HVLS) fans are one category of ceiling fan installed in large enclosings such as warehouses, large barns and health clubs in order to generate comfortable air circulation. As a rotary blade, aerodynamic performance of a HVLS fan is predominantly related to its airfoil(s), and the pitch and twist angles. This paper first, investigates the effects of airfoil on the performances of three different HVLS fans with NACA 5414, 6413 and 7415 airfoils. The fans have six untwisted blades with the diameter of 6 m and rotate at 60 RPM. The blades pitch angels are 12^{\circ}, 12^{\circ} and 13^{\circ}, respectively. The results are presented in the form of the aerodynamic forces and moments, volumetric flow rate and streamlines. Regarding the volumetric flow of air, the results show that the model with NACA 7415 has the best performance. Hence, two other HVLS fans with the same airfoil but, with four and five blades are studied in order to investigate the effects of number of blades. From the point of view of air circulation still the six-bladed fan is the best one; however, the five-bladed fan is more efficient in power consumption.

Full text

80 / J. Comput. Fluids Eng. Vol.19, No.4, pp.80-85, 2014. 12
A
N
I
NVESTIGATION ON
HVLS
F
AN
P
ERFORMANCE
WIT H
D
IFFERENT
B
LADE
C
ONFIGURATIONS
Mohammad Moshfeghi,
1
Nahmkeon Hur,
*1,2
Young Joo Kim
3
and Hyun Wook Kang
4
1
Multi-Phenomena CFD Engineering Research Center, Sogang University
2
Department of Mechanical Engineering, Sogang University
3
Green Transport & Logistic Institute, Korea Railroad Research Institute
4
Technical Research Lab, Building and Tunnel Technologies Inc.
날개 형상에 따른
HVLS
의 성능에 관한 연구
Mohammad Moshfeghi,
1
허 남 건,
*1,2
김 영 주,
3
강 현 욱
4
1
서강대학교 다중현상 CFD 연구센터
2
서강대학교 기계공학과
3
한국철도기술연구원 녹색교통물류시스템공학연구소
4
(주)비엔텍아이엔씨 기술연구소 연구개발팀
High-volume low-speed (HVLS) fans are one category of ceiling fan installed in large enclosings such as
warehouses, large barns and health clubs in order to generate comfortable air circulation. As a rotary blade,
aerodynamic performance of a HVLS fan is predominantly related to its airfoil(s), and the pitch and twist angles.
This paper first, investigates the effects of airfoil on the performances of three different HVLS fans with NACA
5414, 6413 and 7415 airfoils. The fans have six untwisted blades with the diameter of 6 m and rotate at 60
R
PM.
The blades pitch angels are 12
o
, 12
o
and 13
o
, respectively. The results are presented in the form of the
aerodynamic forces and moments, volumetric flow rate and streamlines. Regarding the volumetric flow of air, the
results show that the model with NACA 7415 has the best performance. Hence, two other HVLS fans with the
s
ame
airfoil but, with four and five blades are studied in order to investigate the effects of number of blades. From the
p
oint of view of air circulation still the six-bladed fan is the best one; however, the five-bladed fan is more efficien
t
in power consumption.
Key Words :
High-volume low-speed fan, Pitch angle, Volumetric air flow, k-
ε
turbulence model
Received: December 2, 2014, Revised: December 22, 2014,
Accepted: December 22, 2014.
* Corresponding author, E-mail: nhur@sogang.ac.kr
DOI http://dx.doi.org/10.6112/kscfe.2014.19.4.080
Ⓒ
KSCFE 2014
1. Introduction
High-volume low-speed fans (HVLS) are rather new
solutions for generating a silent, comfortable gentle air
circulation in large enclosing such as warehouses, large
barns and health clubs. Unlike residential ceiling fans that
are typically 0.9 m to 1.3 m in diameter, the diameter of
HVLS fans are between 2.5 ~ 7.5 m. Also, in spite of
normal ceiling fans, HVLS fans rotate very slowly in the
range of 45 ~ 80 RPM. Technically speaking, HVLS fans
must satisfy the two requirements as[1]:
(a) The volumetric air flow through the blade disc during
a single revolution must be no less than 14 cubic meters;
(b) The blade tip speed of a HVLS fan must not be
greater than 27 m/s.
The main benefits of HVLS fans are their capability to
be used in both summer and winter together with the
heating, ventilation and air conditioning; their low noise
and gentle air flow and also the average energy savings
of 49 percent plus consequent reductions in the generation
of CO
2
and carbon[2-3].
Since parametric investigation of a HVLS fan is highly
expensive and to some extent impossible, the CFD has
employed as a reliable functional gadget. Hence, many

A
N
I
NVESTIGATION ON
HVLS
F
AN
P
ERFORMANCE WITH
D
IFFERENT…
Vol.19, No.4, 2014. 12 / 81
(
a
)
NACA 5414
(
b
)
NACA 6413
(
c
)
NACA 7415
Fig. 1 Geometry of the airfoils
studies have been conducted on the inquiry of HVLS
fans’ performance inside enclosing. Momoi et al.[4]
examine the utilization a ceiling fan for airflow control in
a large air-conditioned room. The research presents the
measured airflow pattern around a ceiling fan and
compares it with the CFD simulation results and concludes
that the CFD result are in good agreement with the
measurement result concerning the average of air velocity.
In another study, Forrest and Owen[5] have investigated a
product MegaFan company and also NACA 0012 in order
to model the airflow and heat transfer in a heated
warehouse building to investigate the levels of
de-stratification offered by HVLS fans. Despite the
majority of the research works which are based on CFD,
some filed measurements are also available[6-7].
The research presented here is one part of a
comprehensive industrial HVLS fan project proposed by
Korea Railroad Research Industry (KRRI), which is carried
out at CFD-ERC at Sogang University. The final goal for
this literature is to investigate the best case among three
models with the NACA 5414, 6413 and 7415 airfoils with
the pitch angles of 12
o
, 12
o
and 13
o
, respectively. Also it
studies the effects of different number of blades. All of
the CFD simulations have been carried out using
steady-state assumption with multiple reference frame
(MRF) technique using Realizable k-
ε
turbulence model in
STAR-CCM+ Version 8.04.
2. Turbulence model
For an acceptable CFD simulation, a proper turbulence
model plays an important role. A good choice of
turbulence model can result in an agreeable simulation,
while an inappropriate one misleads to waste of time,
energy and wrong results. If the blades were highly
twisted with different local pitch angles, the SST-k-
ω
(
a
)
(
b
)
(
c
)
(
d
)
Fig. 2 Domain shape and the location of the fan
would be required[8]. However, because of the simplicity
of the present blades with untwisted configurations and as
it is recommended in many references[5-6,9], a k-
ε
[10] (or
a k-
ε
-based) turbulence model can be considered as an
appropriate model for the present HVLS fan simulations.
The Realizable k-
ε
model[11] which is adopted in the
present paper is a relatively recent development and differs
from the standard model in two important ways: (a) The
model contains a new formulation for the turbulent
viscosity; (b) A new transport equation for the dissipation
rate,
ε
, has been derived from an exact equation for the
transport of the mean-square vorticity fluctuation. For the
wall treatment in the Realizable k-
ε
model there is an
options which is called “all-wall treatment”. This option is
a hybrid treatment that is suitable for both coarse and fine
meshes[12]. In present research this option is used for the
simulations.
3. Geometrical of the model
The NACA 5414, 6413 and 7415 airfoils used in the
present study are shown in Fig. 1(a)-(c). The HVLS fans’
diameters are
Φ
= 6 m, with a constant chord length of
20 cm. In addition, the size of the enclosing is assumed
to be 24 m x 24 m with the height of 18 m. In order to
use the MRF technique, an inner axi-symetric domain with
section is created. The details of the domain
dimensions are shown in Fig. 2.

82 / J. Comput. Fluids Eng.
M. Moshfeghi
․
N. Hur
․
Y.J. Kim
․
H.W. Kang
(
a
)
Mesh distribution at domain mid-sectio
n
(
b
)
Blade ti
p(
left
)
(
c
)
Around the airfoil
(
Ri
g
ht
)
(
d
)
Near the
b
lade root and hub
Fig. 3 Mesh system generated by the setting listed in Table 1, 2
4. Mesh resolution and CFD settings
The two-sub-domain configuration (Fig. 2(a)) also can
help us to put more cells inside the inner domain. Table
1, 2 present the details of mesh resolution for the inner
and outer domains, respectively. The mesh is polyhedral
mesh with 8 prism layers at the blade surface. Fig. 3
Model
No. of prism
layer/Factor
/Total thickness
(
mm
)
Min
size
(
mm
)
Max
size
(
mm
)
Surface
grow th
rat
e
Blade
8/1.3/ 30 5 10 1.3
Hub
8/1.3/30 8 15 1.3
Ti
p
8/1.3/30 0.5 1.5 1.3
Shaft
- 20 100 1.3
Top surface
(at ceiling)
- 40 120 1.3
Table 1 Details of mesh information for inner domain
Model
No. of prism
layer/Factor
/Total thickness
(
mm
)
Min
size
(
mm
)
Max
size
(
mm
)
Surface
grow th
rat
e
Floor
- 60 250 1.3
Side walls
- 60 250 1.3
Ceiling
- 100 500 1.3
Interface
around rotor
- 30 100 1.3
Interface
around shaf
t
-20801.3
Table 2 Details of mesh information for outer domain
Fig. 4 Velocity magnitude and streamlines for the case with
NACA 7415_AOA 13
o
shows the mesh arrangement at different locations. By
applying the mesh setting of Table 1, 2, total number of
cells in the inner and outer domains are 4.1 and 1.1
million, respectively.
Since the Realizable k-
ε
model is adopted, the y+
value could be in the order of 100 and a coarse mesh
also can result in an acceptable value. As mentioned
earlier, the “all-wall treatment” option has been chosen,
which ensures us about the accuracy of the results.
Generally, because the system is rotary blades, numerical
convergence is slow and it may need small value of
under-relaxation factor or small timescale[8]. To assure the
accuracy of the results, in addition to the residual values,
the convergence of lift value over the blades has been
also monitored as an additional convergence criterion.
5. Re sults and conclusions
Generally, there are several industrial parameters for the
evaluation of, or comparison between, HVLS fans. From
the point of view of the air circulation, the volumetric
flow rate in a single blade revolution and the air speed
value at some specific locations are usually considered as
the industrially accepted factors for the HVLS fans
performance. In addition, the aerodynamic loads are
commonly studied as crucial factors for mechanical
designing. In this section, these parameters are presented
and compared for the different cases which are studied in
the present research.
5.1 Comparison between cases with different airfoils
In this part, the results of the three models with
NACA 5414, NACA 6413, NACA 7415 are listed in

A
N
I
NVESTIGATION ON
HVLS
F
AN
P
ERFORMANCE WITH
D
IFFERENT…
Vol.19, No.4, 2014. 12 / 83
(
a
)
(
b
)
(
c
)
Fig. 5 Downward velocity and the magnitude for the fan with four, five and six blades
Table 3. As the results of the volumetric flow rate (Q)
show, the fan with the NACA 7415 generates the highest
volumetric flow rate. However, from the aerodynamic
point of view the blades of this HVLS fan undergo higher
lift and torque values, and, hence, the HVLS fan with
NACA 7415 needs more electrical power to be operated.
In addition, the streamlines and the velocity contour of
this case are presented in Fig. 4, which show that the
area below the rotor and also the area close to the wall
(~ 3 m) have high air recirculation.
Another criterion for the performance evaluation of
HVLS fans is the effective area served by the fan in an
empty enclosing (unobstructed space). As the protocols of
HVLS fan suggest[12], the velocity of the generated air
flow has to be measured at the heights of 100 mm, 1100
mm and 1700 mm above the floor surface. Table 4
presents the average of air velocity at the forementioned
heights for all cases, in which the superiority of the case
with NACA 7415 can be observed.
Industrially, it is important to design a product with an
acceptable performance. Hence, a comparison between the
Airfoil_Pitch angle Lift( N) Torque(N·m) Q(m
3
/s)
NACA 5414_12
o
163.7 80.0 149.2
NACA 6413_12
o
174.5 90.2 164.2
NACA 7415_13
o
200.6 104.8 174.3
Table 3 Comparison between performances of the HVLS fans with
different airfoils
Airfoil_Pitch angle 100 mm 1100 mm 1700 mm
NACA 5414_ 12
o
1.49(m/s) 1.00(m/s) 0.78(m/s)
NACA 6413_ 12
o
1.55(m/s) 1.06(m/s) 0.80(m/s)
NACA 7415_ 13
o
1.75(m/s) 1.15(m/s) 0.86(m/s)
Table 4 Air velocities generated by different HVLS fans at
different heights above the floor
nominal power required for NACA 7415_13
o
with three
industrial fans is presented in Table 5, which shows the
acceptable performance of the present model.
5.2 Effects of number of blades
As shown in the previous section, the HVLS fan based
on NACA 7415 airfoil generates the highest volumetric
flow rate and also the fastest air circulation. To investigate
the effects of number of blades, here, three models of this
HVLS with four, five and six blades are compared.
First, the areas with downward flow are shown in Fig.
5(a)-(c). The figures clearly show that with an increase in
number of the blades the downward area becomes wider.
This results in a higher Q for the six-bladed model as
presented in Table 6. However, it is notable that Table 6
also shows that from the point of view of energy, the
five-bladed HVLS fan requires the minimum aerodynamic
torque for generating one unit of volumetric flow rate,
HVLS fan Diameter
(m)
Speed
(RPM)
Q
(m
3
/s
)
Power
(kW)
ProTav_5500
5.5 50-60 135.0 2.2
ProTav_6500
6.5 50-60 89.7 2.2
GFAN6 1
6.1 67 137.8 1.5
NACA 7415_13
o
6.0 60 174.3 1.3
Table 5 Comparison between the generated volumetric flow rate
and the required power for different HVLS fans
Number of blades Lift
(N)
Torque
(
N
∙
m
)
Q
(m
3
/s
)
Q/Torque
4
155 88 153.4 1.74
5
174.5 89 167.7 1.88
6
200.6 104.8 174.3 1.66
Table 6 Comparison between performances of the HVLS fans with
NACA 7415_AOA 13
o
with different number of blades

84 / J. Comput. Fluids Eng.
M. Moshfeghi
․
N. Hur
․
Y.J. Kim
․
H.W. Kang
(a) Four blades
(b) Five blades
Fig. 6 Velocity magnitude and streamlines for the HVLS fans with
different number of blades
which is a factor for optimum energy consumption.
In addition, it is noteworthy to mention another
importance of the blade pitch angle and airfoil by
comparing the result of five-bladed model of Table 6, 7
with the NACA 6413_12
o
model in Table 3, 4. As the
results demonstrate, the two HVLS fans, one with five
and the other with six blades, behave very similar to each
other. Consequently, this means that with an optimum
design, one can considerably lighten the total weight of a
HVLS fan. Furthermore, the streamlines and velocity
magnitudes are presented in Fig. 6(a),(b). Similar to the
Fig. 4, here also the recirculation areas exist below the
fan disc and near the side walls.
Number of Blades 100 mm 1100 mm 1700 mm
4
1.49(m/s) 1.00(m/s) 0.74(m/s)
5
1.58(m/s) 1.07(m/s) 0.83(m/s)
6
1.75(m/s) 1.15(m/s) 0.86(m/s)
Table 7 Air velocities generated by NACA 7415_AOA 13
o
with
different number of blades at different heights above the
floor surface
6. Concluding remarks
This paper investigates three different high-volume
low-speed (HVLS) fans with NACA 5414, 6413 and 7415
airfoils. All of the fans have six blades with a diameter
of
Φ
= 6 m and rotate at 60 RPM. The blades have
untwisted rectangular shapes with a chord of 20 cm. The
results show that although the HVLS fan based on the
NACA 7415 generates the highest volumetric flow rate
and fastest air circulation, it needs more torque.
Furthermore, the compression between the four, five and
six-bladed versions show that the six-bladed version again
generates the highest volumetric flow rate and fastest air
circulation.
However, from the point of view required torque per
volumetric flow rate, the case with five blades is the most
efficient. The strong influence of the airfoil and pitch
angle is also noted by observing the similarity between
the results of the NACA 7415_13
o
(five-bladed) and
NACA6413_12
o
(six-bladed).
Acknowledgement
This research was supported by the National Research
Foundation of Korea (NRF) grant funded by the Korea
Government (MSIP): Multi-phenomena CFD Research
Center (ERC) in Sogang University (No. 2009-0083510);
and the grant from “Transportation & Logistics Research
Program” funded by Ministry of Land, Infrastructure and
Transport(MOLIT) of Korean government."
Ref erences
[1] http://en.wikipedia.org/wiki/high-volume_low-speed_fan.
[2] http://continuingeducation.construction.com/
article.php?L=193&C=635&P=8.
[3] 2001, Dougan, D. and Damiano, L.A., "ASHRAE.
Standard 62. Comprehensive summary," (http://www.
automatedbuildings.com/news/jan03/articles/ebtron/
ebt.htm).
[4] 2004, Momoi, Y., Sagara, K., Yamanaka, J. and
Kotani, H., "Modeling of ceiling fan based on
velocity measurement for CFD simulation of airflow
in large room,"
In: Proceedings of the 9th
International Conference on Air Distribution in Rooms
,
Coimbra, Portugal, pp.145-150.
[5] 2010, Forrest, J. and Owen, I., "MegaFan Warehouse

A
N
I
NVESTIGATION ON
HVLS
F
AN
P
ERFORMANCE WITH
D
IFFERENT…
Vol.19, No.4, 2014. 12 / 85
Case Study:Final Report".
[6] 2003, Kammel, D.W., Raabe, M.E. and Kappelman,
J.J., "Design of high volume low speed fan
supplemental cooling system in dairy free stall barns,"
Proc., 5th Dairy Housing Conf. Publ 701P0203
,
pp.243-254.
[7] 2003, Aynsley, R. and Ali, M., "Optimizing Ceiling
Fan Locations with CFD,"
Architectural Engineering
2003
, pp.1-4.
[8] 2012, Moshfeghi, M., Song, Y.J. and Xie, Y.H.,
"Effects of near-wall grid spacing on SST-k-
ω
model
using NREL Phase VI horizontal axis wind turbine,"
J.
of Wind Eng. and Ind. Aerodyn
, Vol.107, pp.94-105.
[9] 2011, Bassiouny, R. and Koran, N.S., "Studying the
features of air flow induced by a room ceiling-fan,"
J.
of Energy and Buildings
, Vol.43(8), pp.1913-1918.
[10] 1997, Bardina, J.E., Huang, P.G. and Coakley, T.J.,
"Turbulence Modeling Validation, Testing, and
Development,"
NASA Technical Memorandum 110446
.
[11] 1995, Shih, T.H., Liou, W.W., Shabbir, Z. and Zhu,
J., A New-k-
ε
Eddy-Viscosity Model for High
Reynolds Number Turbulent Flows-Model Development
and Validation Computers Fluids, 24(3), pp.227-238.
[12] 2013, "STAR-CCM Ver. 8.04 help manual".
[13] 1985, Rohles, F., "The development of a protocol for
measuring the air velocities from the industrial ceiling
fan,"
Institute for Environmental Research Report
No.85-1
, Kansas State University.

ASCHT 2015
The Asian Symposium on Computational Heat Transfer and Fluid Flow
June 21 – 24 2015
Busan, Korea
AN INVESTIGATION ON EFFECTS OF INSTALATION HEIGHT
AND LOCATION ON HVLS FAN PERFORMANCE
Mohammad Mosfeghia, Nahmkeon Hura,b*, Young Joo Kimc, Hyun Wook Knagd
a Multi-Phenomena CFD Engineering Research Center, Seoul, Korea
b Department of Mechanical Engineering, Sogang University, Seoul, Korea
cSenior Researcher, Korea Railroad Research institute, Korea
d Building and Tunnel Technologies Inc., Korea
E-mail: nhur@sogang.ac.kr
ABSTRACT
HVLS fans are type of ceiling fans used in large enclosings such as warehouses, large barns halls and salons and
shopping centers. Due to the economic benefits and environmental comfort, the high-volume low-speed fans
(HVLS) have taken a special position in buildings ventilation. The range of blade diameter of the HVLS fans is
between D=2.5~7.5 m. And in spite of normal fan, a HVLS fan rotates very slowly, and gently streams the air inside
the enclosing. To have the best and uniform air circulation the fan should be installed in the center of the enclosing
at a specific height. However, it is evident that due to some likely limitation for the installation, a HVLS fan may be
mounted at different height from the optimum one or in an off-centric position. This paper aims to numerically
investigate several cases in which a HVLS fan is installed in an off-center position. In addition it investigates the
effects of different installation heights. Hence, first a survey is carried out for selecting an appropriate turbulence
model. Having chosen k-ε, the CFD simulations are performed using MRF techniques. The results show the effects
of different installation location in terms of the volume flow passing through the fan disc CFM (cubic feet per
minute) and the streamlines.
Keyword : High-volume low-speed fan, Off-centric, NACA 7415, MRF, k-ε.
NOMENCLATURE
H
[m]
Height form the room floor
L
[N]
Lift on the HVLS fan blade
Q
[m3/s]
Volumetric flow rate generated by HVLS fan
T
[N·m]
Torque on the HVLS fan blade
h
[m]
Height form the plane of the rotation of the HVLS fan
x
[m]
Cartesian axis direction
y
[m]
Cartesian axis direction
z
[m]
Cartesian axis direction
INTRODUCTION
Recently, the high-volume low-speed fans (HVLS) have
become a desirable solutions for a silent, comfortable and
gentle air circulation in large halls and salons such as large
barns, warehouses, factories, and health clubs. This types of
fans are different from their previous generations (residential
ceiling fans) in many aspects:
-A conventional ceiling fan diameter is typically 0.9~1.3 m,
while the diameter of a HVLS fan is between 2.5~7.5 m.
-A conventional ceiling fan usually rotates at a rotational speed
of 150~500, while a HVLS fan rotate very slowly in the range
of 45~80 RPM.
From the point of view of industry, the term “HVLS” can be
added to a fan if it satisfies two operational conditions [1]:
- During a single revolution, the volumetric air flow through the
blade disc must be equal or more than 14 cubic meters;
- The blade tip speed of a HVLS fan must smaller than 27 m/s.
In fact there are several environmental and economic reasons
that the HVLS fans have become popular in the market. The
most important advantages of a well-designed HVLS fan are
[2-3]:

- The capability to be used in both summer and winter together
with the heating, ventilation and air conditioning;
- The low level of generated noise;
- Comfortable and gentle air flow;
- Average energy savings, (e.g. 49% as mentioned in [2])
- Reductions in the generation of CO2 and carbon.
Technically speaking, the parametric experiments are very
rare [4,5] and for a HVLS can be highly expensive. Also it is
also impossible to measure and record all of the field
parameters. However, the CFD simulations provide a very
suitable tool as a replacement of the tests, and, hence, many
studies have been conducted on the inquiry of HVLS fans.
Likewise other fields of research, the CFD study of HVLS fans
can be carried out from different purposes such as: (a)
Evaluation and classification of different CFD parameters in
the simulation of HVLS fans [6,7] (b) study of the airflow and
circulation patterns generated inside a large room [8]; (c)
effects of the HVLS fan on the temperature distribution in a
large enclosing [9].
The research presented in this paper is a part of an industrial
HVLS fan project proposed by Korea Railroad Research
Industry (KRRI). The CAD modeling and the numerical
simulations have been carried out at the CFD-ERC at Sogang
University, Seoul Korea. The final goal for the current article is
to investigate the effects of off-centric location and also
installation heights on the air circulation in a large hall. All of
the CFD simulations have been carried out using as steady-state
assumption with multiple reference frame (MRF) technique
using Realizable k-ε turbulence model in STAR-CCM+
Version 8.04.
TURBULENCE MODEL
A proper turbulence model is a key factor for the CFD
simulation and plays an important role in the accuracy of the
results. An appropriate turbulence model can result in an
agreeable simulation, while an irrelevant one misleads to wrong
results. Very similar to a small horizontal axis wind turbine, for
the simulation of a HVLS fan with highly twisted blades, the
SST-k-ω would be required [10]. However, because of the
untwisted configuration in the model and as it is recommended
in many references [4, 9, 11], a k-ε (or a k-ε-based) turbulence
model [12] is appropriate for the present HVLS fan simulations.
To further increase the reliability of the results, the
Realizable k-ε option [11] which is adopted. This option differs
from the standard model in two important ways:
(a) Contains a new formulation for the turbulent viscosity;
(b) A new transport equation for the dissipation rate, ε, has
been derived from an exact equation for the transport of the
mean-square vorticity fluctuation.
In addition, the wall treatment in the Realizable k-ε model
can be activated using “all-wall treatment” option, which
provides a hybrid treatment suitable for both coarse meshes and
fine meshes and works more accurately [12].
GEOMETRY OF MODELS
All of the HVLS fans which are modeled in the present
study have NACA 7415 airfoil (Figure 1) with the constant
chord length of 20 cm, the pitch angle of 13o and the diameter
of 6 m. In addition, the room size is assumed to be 24 m24 m
with the height of 18 m. In order to use the MRF technique, an
inner axi-symetric domain with ╩ section is created. The fan is
assumed to rotate at RPM=60. The details of domain
dimensions for the baseline case are shown in Figure. 2.
Figure 1 Shape of the NACA 7415 airfoil
(a)
(b)
(c)
(d)
Figure 2 Domains shape and location of fan for baseline
MESH RESOLUTION AND CFD SETTINGS

The two-sub-domain configuration (Fig. 2(a)) lets the
models to have finer mesh inside the inner domain. The mesh is
polyhedral mesh with 8 prism layers around the blade and hub
surfaces. The minimum and maximum mesh size in the inner
domain are 5 and 120 mm, respectively. Because the mesh
resolution in the outer domain is not very sensitive issue, these
values for the outer domain are 60 and 500 mm, respectively.
The total number of cells in the inner and outer domains are 4.1
and 1.1 million. Figure 3 shows the mesh arrangement at
different parts of the domain.
(a) Mesh distribution at domain mid-section
(b) Near the blade root and hub
(c) Around the airfoil
Figure 3 Mesh arrangement in different locations
RESULTS AND CONCLUSIONS
A) Effects of installation height
In the first group, the HVLS fan in considered to be
mounted at the center point of the room (Figure 1(c)) at
different height. The values of Δh=0 m is associated to the
baseline case and three other cases are simulated.
To evaluate these cases, the volumetric flow rate (Q) in a
single blade revolution and the aerodynamic lift (L) and
torque (T) are listed in table 1. In addition, the air speed at
some specific locations from the floor (H=0.1, 1.1., 1.7 m) are
compared. These parameters are usually considered as the
industrially accepted factors for the HVLS fans performances.
Tables 1-2 presents the results of comparison for the effects of
installation height. In addition, the streamlines at the center
plane are shown in Figure 4.
Table 1 Comparison between the performances the HVLS fan
at different heights
L (N)
T (N·m)
Q (m^3/s)
Δh=+2 m
216.2
108.1
138.8
Δh=0 (Baseline)
200.6
105
174.3
Δh=-1 m
197.5
105.5
168.1
Δh=-2 m
194.2
106.1
140.3
Table 2 Comparison between the air velocities generated by the
HVLS fan at different heights.
H=0.1 m
H=1.1 m
H=1.7mm
Δh=+2 m
1.7 (m/s)
1.04(m/s)
0.79(m/s)
Δh=0 (Baseline)
1.75 (m/s)
1.15 (m/s)
0.86 (m/s)
Δh=-1 m
1.83 (m/s)
1.06 (m/s)
0.82 (m/s)
Δh=-2 m
1.60 (m/s)
1.13 (m/s)
0.87 (m/s)
B) Effects of eccentricity
In the second set of comparisons, the fan is assumed to be at
the same height of the baseline case (see Figure 1), but at
different horizontal locations (Δx and Δy) in the room. Again,
the results are compared and presented in terms of the
aerodynamic performance in Tables 3-4. In addition,
distribution of downward flow rotor plane is presented in
Figure 6. As the results show, when the location of fan is close
to the sidewalls, the aerodynamic loads on the fan increase,
while the volumetric flow rate decreases. Likewise, the velocity
magnitudes diminish for the cases near the side walls.
Table 3 Comparison between the performances the HVLS fan
at different horizontal locations
L (N)
T (N·m)
Q (m^3/s)
Δx=Δy=4 m
218.7
111.5
159.4
Δx=Δy=2 m
199.2
106.5
182.9
Baseline
200.6
105
174.3
Table 4 Comparison between the air velocities generated by the
HVLS fan at different heights
H=0.1 m
H=1.1 m
H=1.7mm
Δx=Δy=4 m
1.32 (m/s)
0.9 (m/s)
0.66 (m/s)
Δx=Δy=2 m
1.39 (m/s)
1.05 (m/s)
0.81 (m/s)
Baseline
1.75 (m/s)
1.15 (m/s)
0.86 (m/s)

(a) Δh=+2 m
(b) Δh=0 m (baseline)
(c) Δh=-1 m
(d) Δh=-2 m
Figure 4 Velocity magnitude and streamlines for the NACA 7415_AOA 13
Δx=Δy=4 m
Δx=Δy=2 m
Δx=Δy=0 m (Baseline)
Figure 5 Downward velocity and the magnitude for the fan with four, five and six blades
CONCLUDING REMARKS
This paper investigates effects of different installation
location on the aerodynamic performance of a high volume low
speed fan (HVLS). The fan’s blade are rectangular untwisted
with the NACA 7415 airfoils and a pitch angle of 13o. All of
the fans have six blades with a diameter of 6 m and rotate at 60
rpm. The results show that the installation height of the fan has
noticeable influence on the angle of streamlines exactly
underneath of the rotor and near the floor. The higher the fan
location, the smaller the area of this recirculation underneath of
the rotor. However, when the fan location is close to the ceiling,
the lack of space and air behind the fan decreases the HVLS fan
performance and increases the aerodynamic loads on the blades.
Also it is seen that for the eccentric cases the downward
airstreams are not uniformly distributed. It is also shown that
the aerodynamic loads on the blades increase for the eccentric
installations.
ACKNOWLEDGEMENT
This research was supported by the National Research
Foundation of Korea (NRF) grant funded by the Korea

Government (MSIP): Multi-phenomena CFD Research Center
(ERC) in Sogang University (No. 2009-0083510); and
the grant from “Transportation & Logistics Research Program”
funded by Ministry of Land, Infrastructure and Transport
(MOLIT) of Korean government."
REFERENCES
[1] http://en.wikipedia.org/wiki/High-volume_low-speed_fan
[2]http://continuingeducation.construction.com/article.php?L=193&C
=635&P=8
[3] Dougan, D., Damiano L.A., ASHRAE. Standard 62.
Comprehensive summary, ISBN-13: 978-9992418659, 2001,
(http://www. automatedbuildings.com/news/jan03/articles/ebtronebt.
htm)
[4] Kammel, D.W., Raabe , M.E. and Kappelman, J.J., Design of high
volume low speed fan supplemental cooling system in dairy free
stall barns, Proc., 5th Dairy Housing Conf.. Publ 701P0203, 2003,
pp. 243-254.
[5] Aynsley, R. and Ali, M., Optimizing Ceiling Fan Locations with
CFD”, Architectural Engineering, 2003: pp. 1-4.
[6] Zhai, Z. et al., Evaluation of various turbulence models in
predicting airflow and turbulence in enclosed environments by CFD:
Part-1: summary of prevent turbulence models, HVAC&R Research,
13(6), 2007.
[7] Zhang, Z., et al., Evaluation of various turbulence models in
predicting airflow and turbulence in enclosed environments by CFD:
Part-2: comparison with experimental data from literature,
HVAC&R Research, 13(6), 2007.
[8] Momoi, Y., et al., Modeling of ceiling fan based on velocity
measurement for CFD simulation of airflow in large room, Proc. of
the 9th International Conference on Air Distribution in Rooms,
Coimbra, Portugal, 2004, pp.145-150.
[9] Forrest, J., Owen, I., MegaFan Warehouse Case Study:Final Report
(http://www.megafans.co.uk/resources/MegaFan_Final_Report_UoL
.pdf), 2010.
[10] Moshfeghi, M., et al. Effects of near-wall grid spacing on SST-K-
ω model using NREL Phase VI horizontal axis wind turbine, J. of
Wind Eng. and Ind. Aerodyn, Vol. 107,2012, pp. 94–105.
[11] Bassiouny, R. and Koran, N.S.,Studying the features of air flow
induced by a room ceiling-fan, J. of Energy and Buildings, Vol.43
(8), 2011, pp. 1913–1918.
[12] Bardina, J.E., Huang, P.G., Coakley, T.J., Turbulence Modeling
Validation, Testing, and Development, NASA Technical
Memorandum 110446, 1997.
[13] Shih T.H., et al. A New- k-ε Eddy-Viscosity Model for High
Reynolds Number Turbulent Flows - Model Development and
Validation Computers Fluids, 24(3), 1995, pp. 227-238.
[14] STAR-CCM Ver. 8.04 help manual, CD-adapco, London, UK,
2013,.
[15] Rohles, F., The development of a protocol for measuring the air
velocities from the industrial ceiling fan, Institute for Environmental
Research Report No. 85-1, Kansas State University, 1985.