The 11th Asian International Conference on Fluid Machinery and Paper number: 112

The 3rd Fluid Power Technology Exhibition November 21-23, 2011, IIT Madras, Chennai, India

Original Paper

Application of Gurney Flaps on a Centrifugal Fan Impeller

T. Manoj Kumar Dundi N. Sitaram M. Suresh
thisismanoj.t@gmail.com nsitaram@iitm.ac.in suremreddy@gmail.com
Thermal Turbomachines Laboratory, Department of Mechanical Engineering,
Indian Institute of Technology Madras, Chennai – 600036, India
Abstract
The objective of the present investigation is to explore the possibility of improving the performance
of a centrifugal fan at low Reynolds numbers using a simple passive means, namely Gurney flap
(GF). GFs of 1/8th inch brass angle (3.175 mm) corresponding to 15.9% of blade exit height or 5.1%
of blade spacing at the impeller tip are attached to the impeller blade tip on the pressure surface.
Performance tests are carried out on the centrifugal fan with vaneless diffuser at five Reynolds
numbers (viz., 0.30, 0.41, 0.55, 0.69, 0.82x105, i.e., at five speeds respectively at 1,100, 1,500, 2,000,
2,500 and 3,000 rpm) without and with GF Static pressures on the vaneless diffuser hub and shroud
are also measured for each speed at flow coefficients [φ=0.23 (below design flow coefficient),
φ=0.34 (design flow coefficient), φ=0.45 (above design flow coefficient), φ=0.60 (above design flow
coefficient)] with and without GF. From the performance curves it is found that the performance of
the fan improves considerably with GFs at lower Reynolds numbers and improves marginally at
higher Reynolds number. Similar improvements are observed for the static pressures on the diffuser
hub and shroud. The effect of Reynolds number on the performance and static pressures is
considerable. However the effect is reduced with GFs.

Keywords: Centrifugal fan, Gurney flap, Experimental investigation, Performance, Static pressure

1 Introduction and Motivation

Industrial fans are widely used for cooling, ventilation, vacuuming and dust removal, inflating, etc., and
account for a large fraction of the worldwide industrial energy demand. Centrifugal fans are widely used for
these applications, because of low cost, ease of fabrication, robustness, higher pressure ratio and reasonable
design and off-design efficiency. The range of application of centrifugal fans has been extended to cool
portable electronic devices where they have to be run at very low Reynolds numbers. Other examples
include computer cooling fans, refrigeration fans, air conditioning fans, and automotive cooling fans as they
are suitable when space is limited and when there is a high pressure drop environment. Because of their
relatively low rotational speeds in these applications, it results in reduced efficiency.
In recent years, it has been recognized that effective energy conservation can play a major role in reducing
energy consumption and can thus offset growth in energy supply required to keep up with industrial
demand. Given large fraction of the energy consumed by fans, even modest improvements applied at large
scales could result in significant energy savings. Hence there is a need to understand and improve the
performance of centrifugal fans at low Reynolds numbers. A simple and inexpensive passive means that had
shown improving the aerodynamic performance of airfoils and wings is Gurney flap.
A Gurney flap is not a sophisticated device. It is a length of metal or plastic right-angle rigidly bolted,
riveted or glued to the trailing edge of a wing. Dan Gurney’s (All American Racing) team used this type of
flap in 1971 while testing the team’s new USAC car at Phoenix, prior to the season's first race. The team
used this type of flap to increase the ‘‘down force’’ and thus the traction generated by the inverted wings on
his race cars. Numerous wind tunnel tests on Gurney flaps attached to airfoils have been performed.
Received 15 January 20xx; revised 20 March 20xx; accepted for publication 20 June 20xx: Review conducted by Prof. Soon-Wook Kim. (Paper
number O08xxx) (This information will be given when accepted.)
Corresponding author: N. Sitaram, Professor, nsitaram@iitm.ac.in

1
These investigations have shown that the Gurney flap increases the effective camber of the airfoil. There is a
significant increase in lift with only a small increase in drag as long as the flap height scales with the local
boundary layer thickness. Despite their widespread use in aeronautics (Lee [1], Lee and Su [2] and Wang et
al. [3]), very little research work has been done on turbomachinery (Byerley et al. [4], Chen et al. [5],
Greenblatt [6], Janus [7] and Myose et al. [8]). An effort has been made here to find the effectiveness of
Gurney flaps on the performance of centrifugal fan.

2 Experimental Facility, Instrumentation, Program and Procedure

The experiments are carried out on a low specific speed centrifugal fan available at the Thermal
Turbomachines Laboratory, Department of Mechanical Engineering, IIT Madras as shown in Fig. 1. The test
rig is a single stage centrifugal fan driven by a 10 HP AC motor, whose speed is controlled by a variable
speed drive. The major design details of the fan are given in Table 1. A digital micro manometer (Model
FCO012, range: + 200 mm of water gage) and a 20 way scanning box (Model FCO091) manufactured by
M/s. Furness Controls, Bexhill, U.K. are used for pressure measurements. For measurements at lower
Reynolds numbers, 10% range of the manometer is used.
P

P
6
7

1 2 3 4

INLET
5

1 Inlet nozzle 2 Inlet duct 3 Centrifugal fan 4 Drive motor 5 VSD

6 Delivery duct 7 Throttle SS Suction pressure PP Delivery pressure
Fig. 1 Schematic layout of the experimental facility

2
 Table 1 Design Details of the Centrifugal Fan
Pressure ratio, P02/P01 1.08 Mass flow rate, m 0.56 kg/s
Speed, N 3,000 rpm Shape number, Nsh 0.076
Inducer hub diameter, D1h 110 mm Inducer tip diameter, D1t 225 mm
Blade angle at inducer hub, β1h 45 Deg. Blade angle at inducer tip, β1t 29 Deg.
Impeller exit diameter, D2 393 mm Number of impeller blades 20
Blade angle at exit, β2 90 Deg. Blade height, h2 20 mm
Exit diameter of vaneless diffuser, D3 600 mm All angles are w.r.t. tangential direction

Performance tests are carried out on the centrifugal fan with vaneless diffuser at five Reynolds numbers viz.,
0.82, 0.69, 0.55, 0.41, 0.30x105 i.e., at five speeds respectively at 3,000, 2,500, 2,000, 1,500 and 1,100 rpm
without Gurney flap to investigate the effect of Reynolds number. Static pressures on the vaneless diffuser
hub and shroud are also measured for each speed at the following flow coefficients.
φ=0.23 (below design flow coefficient), φ=0.34 (design flow coefficient), φ=0.45 (above design flow
coefficient), φ=0.60 (above design flow coefficient).
A 30 cm long brass angle of 1/8th inch side (3.175 mm) manufactured by M/s Special Shapes and is supplied
by M/s Small Parts is cut into pieces of 20 mm length. One surface of the brass pieces is roughened using
emery paper. Similarly the pressure surface of the impeller blade near its tip is roughened using emery
paper. These brass pieces are attached to the impeller blade pressure surface near tip using instant glue
Anabond. As both the surfaces of the brass pieces and impeller pressure surface near the blade tip are
roughened, the brass pieces are rigidly fixed to the impeller pressure surface. The brass pieces stayed in
place up to an impeller speed of 3,000 rpm. These brass pieces are Gurney flaps as shown in Fig. 2. The
height of the Gurney flaps corresponds to 15.9% of impeller blade height at exit or 5.1% of blade spacing at
the impeller tip. Then the performance tests carried for baseline configuration without Gurney flap are
repeated with Gurney flap to determine the effect of Gurney flap on the performance of centrifugal fan.
Static pressures on the vaneless diffuser hub and shroud are also measured for Reynolds numbers 0.69, 0.55,
0.41, 0.30x105 (2,500, 2000, 1500, 1100 rpm) at flow coefficients φ=0.23 (below design flow coefficient),
φ=0.34 (design flow coefficient), φ=0.45 (above design flow coefficient) and φ=0.60 (above design flow
coefficient) for configurations without and with Gurney flap.

Fig. 2 Impeller blade tip with Gurney flap

3
 3 Results and Discussion
3.1 Effect of Reynolds number on Performance
Performance characteristics without Gurney flap: The performance characteristics for baseline
configuration without Gurney flap in terms of non-dimensional parameters, energy coefficient (ψ) vs. flow
coefficient (φ) are plotted at Reynolds numbers 0.82, 0.69, 0.55, 0.41, 0.30x105 are shown in Fig. 3. There is
a considerable effect of Reynolds number where the fan exhibits lower operating range w.r.t. φinstability point at
lower Reynolds number flows as shown in Table 2. The trends for all Reynolds numbers are similar in nature
but higher differences in energy coefficient values can be observed at mid peak regions of the curves.
Performance characteristics with Gurney flap: The trends of performance parameters are similar in nature
with Gurney flap as shown in Fig. 4. But the values of energy coefficient exhibited by the higher Reynolds
number flows 0.82, 0.69x105 (3,000 and 2,500 rpm) are surprisingly lower when compared to the lower
Reynolds numbers of 0.55, 0.41, 0.30x105 (2,000, 1,500 and 1100 rpm respectively). It is also seen that the
operating range exhibited at different Reynolds numbers varies significantly and does not follow any specific
trend as shown in Table 2. However the fan exhibits highest operating range at the lowest Reynolds number of
0.30x105 (1,100 rpm) and lowest operating range at the highest Reynolds number of 0.82x105 (3,000 rpm).
1.4 1.4
Energy Coefficient, ψ

Energy Coefficient, ψ

1.2 1.2

5 5
Speed (rpm) Reb2x10 Speed (rpm) Reb2x10
1.0 1.0 1100 0.30
1100 0.30
1500 0.41 1500 0.41
2000 0.55 2000 0.55
2500 0.69 2500 0.69
3000 0.82 3000 0.82
0.8 0.8
0.0 0.2 0.4 0.6 0.8 0.0 0.2 0.4 0.6 0.8
Flow Coefficient, φ Flow Coefficient, φ
Fig. 3 Effect of Reynolds number on energy Fig. 4 Effect of Reynolds number on energy
coefficient of centrifugal fan without GF coefficient of centrifugal fan with GF
Table 2 Comparison of Important Performance Parameters without and with GF
N
Reb2
φinstability point φop
(rpm) w/o GF with GF w/o GF with GF
5
1,100 0.30x10 0.311 0.252 0.521 0.612
1,500 0.41x105 0.291 0.295 0.563 0.584
2,000 0.55x105 0.311 0.312 0.550 0.578
2,500 0.69x105 0.281 0.282 0.584 0.600
3,000 0.82x105 0.262 0.290 0.588 0.574
φop : Operating range = φmax-φinstability point
φinstability point : Flow coefficient where there is a sudden drop in energy coefficient
4
 3.2 Effect of GF on Performance
Performance characteristics without and with Gurney flap for Reynolds numbers 0.30, 0.41, 0.55, 0.69,
0.82x105 (1,100, 1,500, 2,000, 2,500, 3,000 rpm respectively) are compared as shown in the Fig. 5 below.
A striking feature of the graph is the qualitative difference in performance curve shapes between the baseline
and Gurney flap data. There is clearly a considerable increase in energy coefficient for almost complete flow
coefficient range and the increase being dominating in low Reynolds number, whereas only small increase is
found in case of high Reynolds number 3,000 rpm i.e., at 0.82x105 and 2,500 rpm i.e., at 0.69x105.

1.4
Energy coefficient, ψ

1.2

N (rpm) N (rpm)
1100 1500
2000 2500
3000
1.0

Solid line+solid symbol: w/o GF Solid line+solid symbol: w/o GF

Dashed line+open symbol: with GF Dashed line+open symbol: with GF
0.8
0.0 0.2 0.4 0.6 0.8 0.0 0.2 0.4 0.6 0.8
Flow coefficient, φ
Fig. 5 Comparison of energy coefficient of the centrifugal fan without and with GF

3.3 Effect of Reynolds number on diffuser hub and shroud static pressures
Static pressure is measured on the diffuser hub and shroud at one circumferential location and seventeen
radial locations at each speed for four flow coefficients, viz., φ=0.23 (below design flow coefficient), φ=0.34
(design flow coefficient), φ=0.45 (above design flow coefficient) and φ=0.60 (above design flow coefficient),
without and with Gurney flaps. There is considerable effect of Reynolds number on static pressures on
diffuser hub and shroud. For the sake of brevity, only representative results are presented. All the static
pressure data are presented in Manoj Kumar Dundi [10]. More or less similar nature is exhibited for various
Reynolds numbers for corresponding flow coefficients as shown in Fig. 6. Static pressure coefficient
increases with radius for all flow coefficients at different speeds as flow diffusion occurs due to area
increase. The values of static pressure coefficients on diffuser hub for corresponding flow coefficients for
various speeds are higher than those on the diffuser shroud. However the difference is small at low flow
coefficients. Static pressure increases continuously without any sudden change upto a radius ratio of about
1.25 on the diffuser hub and shroud. Then there is a sudden change in the slope of static pressure curve.
Static pressure coefficient is generally higher as the Reynolds number increases. Trends for static pressure
with GF configuration are similar as shown in Fig. 7.

5
 Static pressure coefficient on diffuser hub and shroud
1.2 φ=0.23 1.2 φ=0.45

1.0 1.0

5
Speed (rpm) Reb2x10
0.8 1100 0.30 Diffuser hub 0.8
1500 0.41 Diffuser hub
2000 0.55 Diffuser hub
2500 0.69 Diffuser hub
1100 0.30 Diffuser shroud
1500 0.69 Diffuser shroud
0.6 2000 0.69 Diffuser shroud 0.6
2500 0.69 Diffuser shroud
1.0 1.2 1.4 1.0 1.2 1.4
Radius Ratio, R Radius Ratio, R
Fig. 6 Effect of Reynolds numbers on static pressure on diffuser hub and shroud without GF
Static pressure coefficient on diffuser hub and shroud

1.2 φ=0.34 1.0 φ=0.60

1.0 0.8

5
Speed (rpm) Reb2x10
0.8 0.6
1100 0.30 Diffuser hub
1500 0.41 Diffuser hub
2000 0.55 Diffuser hub
2500 0.69 Diffuser hub
1100 0.30 Diffuser shroud
1500 0.69 Diffuser shroud
0.6 2000 0.69 Diffuser shroud
0.4
2500 0.69 Diffuser shroud
1.0 1.2 1.4 1.0 1.2 1.4
Radius ratio, R Radius ratio, R
Fig. 7 Effect of Reynolds numbers on static pressure on diffuser hub and shroud with GF

3.4 Effect of Gurney flap on diffuser hub and shroud static pressures
The radial variation of static pressure coefficient on the diffuser hub and shroud is compared without and
with Gurney flaps for representative flow coefficients and speeds in Fig. 8. In general static pressure on the
diffuser hub and shroud is higher with Gurney flaps compared to the basic configuration of without Gurney
flaps. However the difference is reduced as the Reynolds number increases with almost negligible difference at
the speed of 2,500 rpm corresponding to a Reynolds number of 0.69x105. The results of static pressure
correspond well with those of performance characteristics.

6
 φ=0.23, 1100 rpm φ=0.34, 1500 rpm
Static pressure coefficient on diffuser hub and shroud
1.2 1.2

1.0 1.0

0.8 without GF (Diffuser hub) 0.8

without GF (Diffuser shroud)
with GF (Diffuser hub)
with GF (Diffuser shroud)
1.2
φ=0.45, 2000 rpm 1.0
φ=0.60, 2500 rpm

1.0 0.8

0.8 0.6

0.6 0.4

1.0 1.2 1.4 1.0 1.2 1.4

Radius ratio, R
Fig. 8 Effect of Gurney flap on static pressure on diffuser hub and shroud

7
4 Conclusions
1. Performance tests on the centrifugal fan without and with Gurney flaps on the impeller blade tip have
shown that the fan performance improves with Gurney flaps. In addition the maximum volume flow
across the fan increases slightly with Gurney flaps. However the performance without and with Gurney
flaps remains almost same at Reynolds number of 0.69x105 and higher Reynolds numbers.
2. The effect of Reynolds number on the performance parameters is as follows: At lower Reynolds
numbers the impeller without Gurney flap shows lower operating range compared to the impeller
with Gurney flaps. This difference decrease as the Reynolds number increases, with almost
negligible difference at Reynolds number of 0.69x105 and higher Reynolds numbers.
3. Gurney flaps of 1/8th inch brass angle (3.175 mm), 15.9% of blade exit height are successful in
producing higher values of energy coefficient for complete flow range for low Reynolds numbers. At
the higher Reynolds number of 0.69x105 (2,500 rpm), the difference in static pressure without and
with Gurney flaps is negligible.
Nomenclature

b2 diffuser width (m) W specific work (m2/s2)

D2 impeller tip diameter (m) α flow angle in degrees
N speed in rpm φ flow coefficient
r2 impeller tip radius (m) ρ density of air (kg/m3)
r3 diffuser outlet radius (m) ψ energy coefficient
R radius ratio (r3/r2) ψws static pressure coefficient
Reb2 Reynolds number based on impeller on diffuser hub and shroud
blade height at exit (U2b2/ν) Subscripts
V volume flow (m3/s) 1 impeller inlet
2 impeller exit
References
[1] Lee, T., 2011, “PIV study of near-field tip vortex behind perforated Gurney flaps”, Experiments in
Fluids, Vol. 50, pp. 351–361.
[2] Lee, T. and Su, Y. Y., 2011, “Lift enhancement and flow structure of airfoil with joint trailing-edge
flap and Gurney flap”, Experiments in Fluids, Vol. 50, pp. 351–361.
[3] Wang, J. J., Li, Y. C., and Choi, K.-S., 2008, “Gurney Flap: Lift Enhancement, Mechanisms and
Applications”, Prog. Aerosp. Sci., 44, pp. 22–47.
[4] Byerley, A. R., Störmer, O., Baughn, J. W., Simon, T. W., Van Treuren, K. W., and List, J., 2003, “Using
Gurney Flaps to Control Laminar Separation on Linear Cascade Blades”, ASME Journal of
Turbomachinery, Vol. 125, No. 1, pp. 114–120.
[5] Chen, P. H., Qiao,W-Y. and Luo, H-L., 2010, “Investigation of low solidity LP turbine cascade with
flow control: Part 2–Passive flow control using Gurney flap”, ASME Paper GT2010-22330.
[6] Greenblatt, D., 2011, “Application of Large Gurney Flaps on Low Reynolds Number Fan Blades”,
ASME Journal of Fluids Engineering, 133, pp. 021102-1 to 021102-7.
[7] Janus, J. M., 2000, “Analysis of Industrial Fan Designs with Gurney Flaps”, AIAA Paper No. 2000-983.
[8] Myose, R. Y., Lietsche, J. C., Scholz, D., Zinge, H., Hayashibara, S. and Heron, I., 1996, “Flow
Visualization Study on the Effect of a Gurney Flap in a Low Reynolds Number Compressor Cascade”,
AIAA Paper 2006-7809.
[9] Casey, M., 1985,“The Effects of Reynolds Number on the Efficiency of Centrifugal Compressor Stages”,
ASME Journal of Engineering for Gas Turbines and Power, Vol. 107, No. 2, pp. 541-548.
[10] Manoj Kumar Dundi, T., 2011, “Effect of Gurney flaps on the performance and diffuser wall static pressures
of a centrifugal fan”, M. Tech. Project Report, Department of Mechanical Engineering, IIT Madras.