Journal of Thermal Science Vol.19, No.

1 (2010) 1−6

DOI: 10.1007/s11630-010-0001-0 Article ID: 1003-2169(2010)01-0001-06

Performance and Flow Condition of Small-Sized Axial Fan and Adoption of

Contra-Rotating Rotors

T. Shigemitsu1, J. Fukutomi1, and Y. Okabe2

1. Institute of Technology and Science, The University of Tokushima, 2-1 Minamijyosanjima-cho, Tokushima 770-8506,
JAPAN
2. Graduate School of Advanced Technology and Science, The University of Tokushima, 2-1 Minamijyosanjima-cho,
Tokushima 770-8506, JAPAN

© Science Press and Institute of Engineering Thermophysics, CAS and Springer-Verlag Berlin Heidelberg 2010

Small-sized axial fans are used as air coolers for electric equipments. But there is a strong demand for higher
power of fans according to the increase of quantity of heat from electric devices. Therefore, higher rotational
speed design is conducted, although it causes the deterioration of the efficiency and the increase of noise. Then
the adoption of contra-rotating rotors for the small-sized axial fan is proposed for the improvement of the per-
formance. In the present paper, the performance and the internal flow condition of the small-sized axial fan are
shown as a first step of the research for the contra-rotating small-sized axial fan and the important points to apply
contra-rotating rotors to the small-sized axial fan are discussed. Furthermore, the numerical flow analysis is con-
ducted to investigate the performance of the contra-rotating small-sized axial fan and internal flow conditions and
pressure distributions are clarified and the effect of contra-rotating rotors is considered.

Keywords: Small-sized axial fan, Contra-rotating rotors, Numerical analysis, Performance, Internal flow

Introduction However, the increase of the power by the increase of the

There is a strong demand for higher efficiency of data
centers and electric devices to take one role for solving
energy problem and global warming because the con-
sumption of the electrical power has been increasing sig-
nificantly according to the establishment of ubiquitous
networking society and increase of quantity of heat from
electric devices. Electrical power used for the cooling of
the IT devices for data centers are huge, the same as that
used for the IT devices itself in data centers. Small-sized
axial fans are used as air coolers for electric equipments
i.e. laptop, desk top computers and servers. There is a
strong demand for higher power of fans according to the
increase of quantity of heat from electric devices [1].
fan diameter is restricted because of the limitation of the
space. Therefore, higher rotational speed design is con-
ducted, although it causes the deterioration of the effi-
ciency and the increase of noise. On the other hand,
lower rotational speed design [2] and advantages on the
performance of the contra-rotating fans and pumps are
verified by experimental results [3,4]. Then the adoption
of contra-rotating rotors for small-sized fans is proposed
for the improvement of the performance. In the case of
contra-rotating rotor, the axial space becomes larger than
conventional small-sized axial fans. However, it is an
adequate choice to apply the contra-rotating rotors for the
small sized-fan because the axial space can be ensured in
electrical devices as compared to that of the radial space.

Received: October 2009

Toru Shigemitsu: Associate Professor
www.springerlink.com
2 J. Therm. Sci., Vol.19, No.1, 2010

Nomenclatures
Dh diameter at the hub (mm) Q flow rate (m3/s)
Dt diameter at the tip (mm) r radius (mm)
L shaft power (W) rc radius at the casing (mm)
Pd dynamic pressure (Pa) Vt circumferential velocity (m/s)
Ps static pressure (Pa) Vz axial velocity (m/s)
Pt total pressure (Pa) Z axial distance (m)
ΔP pressure increase (Pa) Greek letter
ΔPd pressure increase at the designed flow rate (Pa) η efficiency (%)

In the case of contra-rotating rotors, it is necessary to measurement duct and booster fan and blew out in the
design a rear rotor considering the unsteady circumferen- ambient atmosphere. The static pressure increase (ΔP)
tial velocity distributions at the outlet of the front rotor was measured by the pressure difference between static
[5]. And it is important to clarify the influence of the holes downstream of the rotor installed at the chamber
wake from the front rotor to the rear rotor on the per- and ambient air. Further, the rotational speed was con-
formance and pressure interaction between front and rear trolled by the servo motor and the flow rates were meas-
rotors [6]. On the other hand, the conventional design ured by the orifice meter set at the measurement duct.
method and the theory for the turbomachinery can’t be The pressure curves from the cutoff flow rate to the large
used for small-sized axial fans because small-sized axial
fans applied to electrical devices belong to extremely
small size field in turbomachinery. Therefore, there is the
strong demand to establish the design method for
small-sized axial fans. In the present paper, the perform-
ance and the internal flow condition of the small-sized
axial fan with 100mm diameter are shown and the limits
of laws to apply the conventional design method to the
small-sized axial fan is discussed. Furthermore, the nu-
merical flow analysis is conducted to investigate the per-
formance and the internal flow conditions of con-
(a) R type (b) RR type
tra-rotating rotors. Then the advantages of the con-
tra-rotating small-sized axial fan on the performance are Fig.1 Small-sized axial fan
considered.
Table 1 Primary dimensions of R and RR types
Experimental Apparatus and Method Hub Mid Tip
Diameter
45 72 98
The rotor design and the primary dimensions of a Blade Number 4
conventional axial fan(Rtype) and a contra-rotating axial NACA NACA NACA
fan(RRtype) are shown in Fig.1 and Table 1 respectively. Rotor Blade Profile
4412 4412 4412
The hub tip ratio Dh/Dt=45mm/98mm and designed flow (R type)
Solidity 0.908 0.361 0.208
rate was Qd=0.016m3/s for both types and pressure in- Stagger Angle 59.84° 70.35° 74.87°
crease at a design point was ΔPd=13.7Pa for Rtype and
Diameter 45 73 98
ΔPd=14.7Pa for RRtype with the same pressure increase
Blade Number 4
of each front and rear rotors. The rotational speed of Front
NACA NACA NACA
Rtype was N=3000min-1 against Nf=Nr=1780min−1 for Rotor Blade Profile
4406 4406 4406
(RR type)
RRtype. In this research, an aerofoil blade was used be- Solidity 1.245 0.508 0.308
cause there was a report which told an advantage of the Stagger Angle 42.40° 60.11° 67.21°
aerofoil blade for the small-sized axial fan[7], however a
Blade Number 5
circular-arc blade was generally used for the small-sized Rear NACA NACA NACA
axial fan. Figure 2 shows the schematic diagram of the Rotor Blade Profile
4409 4409 4409
experimental apparatus. The experimental apparatus was (RR type)
Solidity 0.910 0.447 0.288
designed based on the Japanese Industrial Standard and
Stagger Angle 55.92° 66.91° 68.87°
the air blown in the test section passed the rotor, chamber,
T. Shigemitsu et al. Performance and Flow Condition of Small-Sized Axial Fan and Adoption of Contra-Rotating Rotors
Fig.2 Experimental apparatus
flow rate were investigated in the experiment. The inter-
nal flow measurements at the inlet and the outlet of the
rotor were conducted with a one-hole cylindrical pitot
tube having an outer diameter of 2.0mm. This tube was
used as the substitution of a three-hole pitot tube by ro-
tating it by ±30°. The measurement positions by the pitot
tube are shown in Fig.3. The measurements in axial di-
rection were conducted from 5mm upstream of the lead-
ing edge of the rotor at the hub to 4 other axial upstream
positions by interval of 10mm at the inlet of the rotor and
at 6mm downstream of the rotor at the hub, from 40mm
downstream of it to 6 other axial downstream positions
by interval of 10mm at the outlet of the rotor. Further, the
measurements in radial direction were conducted from
r=28.5mm to r=49.5mm by interval of 3mm at 8 radial
positions. Therefore, the total of the measurements points
were 104 points.
Fig.3 Flow field measurement points
Numerical analysis conditions
The commercial software ANSYS-Fluent was used to
investigate the flow condition which couldn’t be meas-
ured by the experiment. In the numerical analysis, the
numerical model which was almost the same with the
experimental apparatus was used and three dimensional
steady numerical analysis was conducted. The numerical
grids used for the numerical analysis were shown in Fig.4.
3
The numerical domains were comprised of the inlet, rotor,
chamber and outlet duct regions. The numerical grids
number at each region are 218,039 for the inlet region,
1,521,481 in the case of Rtype and 3,613,381 in the case
of RRtype for the rotor region, 667,135 for the chamber
region and 39,875 for the outlet duct region respectively.
The tip clearance was kept 1mm as the same with the
experimental apparatus in the numerical analysis and
numerical grids over 150,000 were ensured at the tip
clearance. The uniform flow at the inlet region and con-
stant pressure at the outlet region were given as boundary
condition. And standard wall function and k-ω turbulence
model was used. The numerical flow analysis was con-
ducted at the 6 different flow rates from 0.6Qd to 1.2Qd.
Fig.4 Numerical analysis grids
Experimental and Numerical Results
In order to investigate the possibility of the adoption
of contra-rotating rotors to the small-sized axial fan, nu-
merical analysis was conducted. Figure 5 shows the
pressure curves of Rtype and RR type obtained by the
numerical analysis. In Fig.5, the pressure curve of Rtype
obtained by the experiment at N=3000min−1 is also
shown to compare with the numerical results. It could be
found that the pressure increased according to the de-
crease of the flow rates and the pressure curves showed
the stable negative curve from the experimental results of
Rtype. The pressure increase of Rtype was ΔP=7.5 Pa at
the designed flow rate Qd=0.016m3/s, however the de-
signed pressure increase was ΔP=13.7Pa. This could be
caused by the design method of this small-sized fan
based on the conventional method, the semi-open inlet
shape of the small-sized fan, the size effect, the boundary
layer on the wall and the pressure loss in downstream
because the stator was not installed. Therefore, it was
confirmed the limits of the laws to apply the conven-
tional design method to the small-sized fan. Moreover, it
was observed that the pressure increase was ΔP=31.7Pa
at the shutoff flow rate, which was approximately 4 times
of the pressure at the designed flow rate. On the other
hand, it was confirmed from Fig.5 that the numerical
results of Rtype showed the quantitative accordance with
the experimental results and the numerical results repre-
sented appropriate data. Therefore, we investigated the
performance of R and RRtypes from the numerical re-
sults. The pressure increase of RRtype was higher than
4 J. Therm. Sci., Vol.19, No.1, 2010
that of Rtype in all flow rates the numerical analysis was
conducted. But the pressure increase of RRtype was
lower than the designed value ΔP=14.7Pa. As a result, it
was clarified that the conventional design method was
not also appropriate for the contra-rotating small-sized
axial fan. The shaft power and the efficiency of Rtype
and RRtype obtained by the numerical analysis are
shown in Fig.6. The efficiency of RRtype at the designed
flow rate Qd=0.016m3/s was (η=44.4%) higher than that
of Rtype (η=34.8%). Furthermore, the efficiency of
RRtype was higher than that of Rtype in all flow rates,
where numerical analysis was conducted. Therefore, the
advantages of the adoption of contra-rotating rotors for
small-sized fan were confirmed.
Fig.5 Pressure increase curves of R and RRtypes
Fig.6 Shaft power and efficiency curves of R and RR types
The important things that need to be considered for the
adoption of the contra-rotating rotors were investigated
with the experimental results of the flow distributions of
Rtype at the inlet and the outlet of the rotor. Figures 7
and 8 exhibit the velocity distributions in radial direction
at the inlet and the outlet of the rotor respectively. Each
(a) and (b) in Figs.7 and 8 shows the axial component
and the circumferential component of the absolute veloc-
ity. The axial velocity and the circumferential velocity
distributions assumed in this design method are shown as
the solid lines in Fig.7(a) and Fig.8(b) respectively. The
vertical axis is non-dimensional radius divided by the
radius at the casing; r/rc=0.45 and r/rc=1.0 correspond to
the hub and the casing. Further, the circumferential ve-
locity Vt is positive in the direction of the rotor rotation.
Focused on the axial velocity in Fig.7, it was found that
the axial velocity decreased gradually from the hub to the
tip at 5mm upstream of the rotor and the axial velocity at
the tip was extremely low with the effect of the
semi-open inlet shape and the boundary layer on the
casing wall, although the axial velocity distributions were
uniform until 15mm upstream of the rotor. It was ob-
served from Fig.7 that the flow inclined toward the hub
region because the axial velocity was large at the hub and
it was considered that this inclined flow condition influ-
enced the circumferential velocity distributions down-
stream of the rotor; the circumferential velocity was large
at the tip and low at the hub in Fig.8. In this test fan, the
constant axial velocity was assumed in radial direction at
the inlet of the rotor. The axial velocity distributions were
not uniform at the inlet of the rotor and a pressure loss
was caused by the separation on the blade surface.
Therefore, it was considered that the designed fan pres-
sure was not obtained at the designed flow rate and this
(a) Axial velocity distributions
(b) Circumferential velocity distributions
Fig.7 Velocity distributions at the rotor inlet of Rtype
T. Shigemitsu et al. Performance and Flow Condition of Small-Sized Axial Fan and Adoption of Contra-Rotating Rotors 5

was known the dynamic pressure was recovered as the

static pressure downstream of the front rotor in the case
of RRtype It was seen from Fig. 10 the gentle slope of
the pressure increase for RRtype compared to that of
Rtype. Further, it was considered that the advantages on
the performance of RRtype were mainly due to the pres-
sure recovery in downstream of the front rotor by the rear
rotor and the gentle pressure gradient in rotors in the case
of RRtype.

(a) Axial velocity distributions

(b) Circumferential velocity distributions Fig.9 Circumferential velocity distributions at the rotor outlet
Fig.8 Velocity distributions at the rotor outlet of Rtype

non-uniform velocity condition caused the deterioration

of the performance. It was clarified from these results
that it was important to consider this non-uniform flow
condition leaned toward the hub region for the design of
the small-sized contra-rotating axial flow fan. Further-
more, it was confirmed that these flow condition at the
rotor outlet should be considered for the rear rotor design
of the contra-rotating small-sized axial fan.
Figure 9 shows the circumferential velocity distribu-
tions for R and RRtypes at the outlet of the rotor obtained
by the numerical analysis as compared to the experimen- Fig.10 Axial distributions of each pressure increase
tal results of Rtype. The circumferential velocity is posi-
tive in the direction of the front rotor rotation for Rtype Concluding Remarks
and in the direction of the rear rotor rotation for RRtype.
The circumferential velocity distributions based on the The performance and the internal flow condition of the
free vortex design are shown as a solid line in Fig.9. small-sized axial fan were investigated experimentally
Further, the mass flow rate averaged total (Pt), static (Ps) and the advantages of contra-rotating type were investi-
and dynamic pressure (Pd) distributions in axial direction gated by the numerical analysis. As a result, following
of R and RRtypes are exhibited in Fig.10. The horizontal concluding remarks were obtained.
axis of Fig.10 shows the axial distance from the leading 1. The static pressure of Rtype based on the conven-
edge of the front rotor at the hub. It was observed from tional design method was lower than the designed value
Fig.9 that the circumferential velocity over Vt=1m/s re- and the limits of the laws to apply the conventional de-
mained at 10mm downstream from the trailing edge of sign method to the small-sized fan were confirmed.
the rotor for Rtype. On the other hand, the circumferen- 2. The axial velocity inclined against the hub region
tial velocity was almost zero (Vt=0) at 10mm downstream and the axial velocity was low in the tip region at the
from the trailing edge of the rear rotor for RRtype and it rotor outlet of Rtype. On the contrary, the circumferential
6 J. Therm. Sci., Vol.19, No.1, 2010
velocity was small in the hub region and large in the tip
region. It was important to consider these flow conditions
for the design of the contra-rotating small-sized axial fan.
3. It was clarified from the numerical results that the
pressure and the efficiency of RRtype were higher than
that of Rtype in all flow rates, where numerical analyses
were conducted.
Acknowledgement
The authors wish to show our special thanks to the
supports by the project research aid from The University
of Tokushima, Japan Science and Technology Agency
and Komiya research aid.
References
[1] Miyahara, M and Fukano, T.: Fan Cooling Technology
for Small Electronic Device, Turbomachinery(in Japa-
nese), Vol. 34, No3, (2006), 129−134.
[2] Furukawa, A., et al.: Performance Test and Flow Meas-
urement of Contra-Rotating Axial Flow Pump, J. Ther-
mal Science, Vol.16, No.1, (2007), 7−13.
[3] Furukawa, A., et. al.: Experimental Study of Pump Char-
acteristics of Contra-Rotating Axial Flow Pump, Proc. 2nd
Int. Symp. on Fluid Machinery and Fluid Eng., Beijing,
67−657, (2000), 245−252.
[4] Kodama, Y., et al.: Experimental Study on the Character-
istics of Fluid Dynamics and Noise of a Counter-Rotating
Fan(1st Report, Effects of the Supporter Shape of the
Electric Motor and the Distance between Two Rotors on
the Characteristics), Trans. JSME (in Japanese), Vol. 60,
No. 576, (1994), 2764−2771.
[5] Shigemitsu, T., et. al.: Experimental Study on Rear Rotor
Design in Contra-Rotating Axial Flow Pump, Proc. 5th
JSME/KSME Fluids Eng. Conf., Nagoya, (2002),
1453−1548.
[6] A, J, Sanders., et. al.: Multi-Blade Row Interactions in a
Transonic Axial Compressor: Part I- Stator Particle Image
Velocimetry (PIV) Investigation, ASME J. Turbomachin-
ery, Vol.124, (2002), 10−18.
[7] Ito, T., et. al.: Experimental Research for Performance
and Noise of Small Axial Flow Fan (Influence of Pa-
rameter of Blade), Trans. JSME (in Japanese), Vol. 72,
No. 715, (2006), 670−677.