Case Studies in Thermal Engineering 1 (2013) 26–37

Contents lists available at ScienceDirect

Case Studies in Thermal Engineering

journal homepage: www.elsevier.com/locate/csite

Effect of axial gap between inlet nozzle and impeller on efficiency

and flow pattern in centrifugal fans, numerical and
experimental analysis
Mojtaba Gholamian n, Gurram Krishna Mohan Rao, Bhramara Panitapu
Department of Mechanical Engineering, JNTUH College of Engineering, Hyderabad, India

a r t i c l e in f o abstract

Article history: Inlet is one of the basic elements of squirrel cage fan that can have great effect on
Received 31 May 2013 performance and losses, especially between inlet nozzle exit and first section of impeller
Received in revised form width. But, enough research has not been done on its parameters. In this paper the effect
3 August 2013
of axial gap between inlet nozzle and impeller on performance and flow pattern is
Accepted 4 August 2013
considered. Four inlet nozzle sizes with respect to impeller size (two smaller, nearly same
Available online 22 August 2013
and bigger than inner impeller diameter) and three axial gaps within the physical
Keywords: dimensions of the casing and impeller were chosen. Numerical simulations with different
Efficiency turbulence models, special geometry and mesh pattern were performed to find the effect
Turbulence model
of this axial gap on flow pattern, performance and efficiency. For the validation of
Squirrel cage fan
numerical results, some experiments were done and all the performance parameters were
Axial gap
compared with that of numerical simulations. These results show good matching between
experimental and numerical results. From the simulation of each case study, flow pattern
and its mechanism and the causes affecting the efficiency and performance due to axial
gap are analyzed and presented.
& 2013 The Authors. Published by Elsevier Ltd. Open access under CC BY-NC-ND license.

1. Introduction

Forward blade centrifugal fans are used extensively in industries and HVAC applications due to their simplicity, ease of
manufacturing and low costs. This type of turbo machine has small chord line length, high width and blades that are joined
together with shroud and hub.
Experiments and studies have shown that the flow in this equipment is three dimensional, complex and turbulent,
especially in two regions: around one-third of impeller width and near tongue in volute. This turbo machine has numerous
geometrical parameters that can affect its performance. If a change in a parameter can improve the performance (especially
efficiency), it means, it might result into more uniform flow pattern and reduced losses in the impeller, particularly at the
blade entry, between blades or at the exit of blades or volute or both of them.
In centrifugal fans, flow enters from inlet almost uniformly and asymmetric [1] and goes into impeller directly. Flow at
the rotor exit is three dimensional, highly fluctuating with jet and wakes. It is reported that in the area between inlet nozzle
and up to one-third of impeller width there is flow separation, high turbulent flow and reversed flow [2,3].

n
Corresponding author. Tel.: þ91 87 9069 2686.
E-mail addresses: [email protected] (M. Gholamian), [email protected] (G.K.M. Rao), [email protected] (B. Panitapu).

2214-157X & 2013 The Authors. Published by Elsevier Ltd. Open access under CC BY-NC-ND license.
http://dx.doi.org/10.1016/j.csite.2013.08.003
 M. Gholamian et al. / Case Studies in Thermal Engineering 1 (2013) 26–37 27

Nomenclature Y coordinate direction

Z coordinate direction
b width of impeller z number of blades
B width of casing ΔPt total pressure difference between blade entry
C gap between impeller and inlet and blade exit flow
Di inlet diameter α angle of tongue position
D1 inner diameter of impeller αs spread angle of casing
D2 outer diameter of impeller β1 inlet angle of blade
E empirical constant (9.793) β2 outlet angle of blade
Gk generation of turbulence kinetic energy due to ε dissipation rate
the mean velocity gradients ρ density (m3/kg)
k turbulent kinetic energy ψ head coefficient
S mean rate of strain tensor φ flow coefficient
Q volume flow rate (m3/kg) η efficiency
N rotational speed of impeller (rpm) μ viscosity of flow
PIn work done by electric motor μt turbulent viscosity
W work done by impeller per mass flow rate (j) τ stress tensor
X coordinate direction

Volute, impeller and inlet are basic elements that have vital importance in centrifugal fans. For impeller and volute, some
case studies (consideration and modifications) have been done to find out and modify the interaction between impeller and
volute and flow pattern at the exit of impeller and also near tongue [3–6]. The availability of simulation tools has facilitated
the study of flow patterns and corresponding mechanisms and analyze before conducting the experimental studies. Lee [5]
used CRUNCH CFD code with K–ε equation for turbulent modeling and wall function procedure and showed that there is a
good agreement between experimental and numerical method in case of static pressure, but not for power. Khelladi [3]
studied flow in impeller–diffuser interface by using FLUENT code with k–ω Shear Stress Transport (SST) model and showed
that there is a good match between numerical and experimental results for pressure in impeller inlet and return channel
outlet. Kim [7] considered the shape optimization studies for centrifugal fans, Rafael [8] considered a special case in which
only the radial gap and clearance between inlet nozzle and impeller is studied and Velarde for volute pressure fluctuation
and some geometry modification [9], used numerical methods with k–ε model. All these studies reported in the literatures,
showed that the numerical simulations were in agreement with the experimental results for different situations considered
for improvement in performance.
Inlet is an important part of fan that can have great effect on performance and variation of losses especially between inlet
nozzle exit and first section of impeller width. But, enough and complete attention is not paid to this issue. The effect of inlet
size on the performance of fans and corresponding modifications was reported by Suzuki and Ugai [4] for backward airfoil
blades with different inlet diameters. All the sizes of the inlet used by them were less than inner impeller diameter and
there is a large axial overlap between inlet nozzle and impeller. They reported that the fan with the largest inlet nozzle
diameter has the best performance.
Axial gap between inlet nozzle and impeller is one of the basic parameters of inlet that can have a great role among the
other inlet parameters. Most of researchers like Suzuki [4] considered the axial gap as fixed parameter in their studies and
others like Montazerin [10] showed that there is an optimum value for the axial gap between inlet nozzle and impeller, but
his result cannot be used as a general result, because in his case study the inlet size was fixed and variation of axial gap for
different inlet sizes was not considered. Hence, the corresponding results might not represent that the decreasing or
increasing of losses (performance and efficiency) is only due to the axial gap. This motivated the authors to conduct case
studies varying the inlet nozzle diameter and the axial gap between inlet nozzle and impeller, while keeping all other
parameters fixed. By this method the effect of the axial gap on performance and efficiency can be found for each inlet size.
In the present study, three different axial gap sizes were used with four inlet nozzle sizes to perform numerical simulations.
To validate the numerical results, some experiments were conducted with the geometrical parameters that are as same as
those considered in the numerical simulations.

2. Numerical simulation

2.1. Fan geometry and mesh generation

For the present case studies, 12 geometries for four different inlet sizes and three axial gap sizes were created by using
the commercial software GAMBIT. The geometrical details are presented in Table 1. A rotor with outer diameter of 216 mm
and 104 mm width was selected such that for optimum value the ratio of outer diameter to inner diameter of impeller is
0.86. Also the ratio of impeller width to impeller diameter is near 0.5 [11] as mentioned in Table 1. The ratio of blade chord
28 M. Gholamian et al. / Case Studies in Thermal Engineering 1 (2013) 26–37

Table 1
Parameters and dimensions of the fan for simulation and experimentation.

Di 175 mm, 190 mm, 205 mm αs 7.51

Dshroud 218 mm ω 900 rpm
D1 186 mm α 301
D2 216 mm C 8 mm
D1/D2 0.86 Z 44
B 125 mm β1 871
b 104 mm β2 1701
B/b 1.2

Fig. 1. General view of squirrel cage fan.

to impeller diameter and inlet and outlet angles of blades are selected as reported by Yahya, Patrick and Ouyang [12–14].
Similarly, the other parameters viz., casing and gaps which also include ratio of casing width to impeller width and position
of tongue were selected from optimum and common ranges as reported by Neise, Raj, Morinushi and Gui [12,15–17].
As mentioned in Table 1, four sizes of inlet diameter: 16 cm and 17.5 cm (smaller than inner diameter of impeller), 19 cm
(nearly same with inner diameter of impeller) and 20.5 cm (greater than inner diameter of impeller) are selected for
numerical simulations. Also for each inlet nozzle three different axial gaps of 4 mm, 8 mm and 12 mm are selected within
the physical constraints of the experimental setup. The rotational speed is selected as 900 rpm for all the 12 geometrical
sizes (to be independent of Reynolds Number [18]).
To construct these geometries, unstructured tetrahedral cells were used and the finest meshes are used in blade
channels. In the gap area between inlet nozzle exit and impeller and also the area around tongue, mesh size is very smaller
than that of other meshes in volute. To have the realistic velocity profile at the entrance of inlet nozzle and also to obtain
well developed velocity profile at outlet channel, a hemisphere was added to inlet nozzle [19] and the length of outlet
channel was increased by a distance more than 10D2. Fig. 1 shows a sketch of the fan with related mesh pattern.
Since the number of cells and their distribution play a great role in numerical calculations, grid independence study was
conducted and presented in Fig. 2 which shows the variation of efficiency (includes static and dynamic pressure and losses)
in fan with five different grids. The figure shows that with the increase of number of cells, there is a steep increase in the
efficiency up to the grid of 2  106cells, beyond which the increase in efficiency is less significant (less than 0.7%) and hence
the grid of 2  106 cells is selected for further simulations.

2.2. Numerical procedure

The flow velocities considered are very low compared to the sonic velocity and the process is essentially isothermal in
nature, hence the fluid is considered as incompressible and constant fluid temperature. Steady flow governing equations are
considered as the experiments were conducted under steady state conditions and the emphasis essentially is to study the
performance, efficiency and flow pattern under steady state conditions. The mass and momentum conservation equations
for steady, 3D, incompressible and isothermal fluid are as follows.
∂U i =∂X i ; i ¼ 1; 2; 3 ð1Þ
 M. Gholamian et al. / Case Studies in Thermal Engineering 1 (2013) 26–37 29

Fig. 2. Using grid pattern in squirrel cage fan.

Table 2
Comparison of efficiencies using different turbulent models for fan with 19 cm inlet nozzle diameter.

Name of method Iteration number Deviation from experimental result (%)

Standard K–ε method 923 0.16

RNG K–ε method 1656 0.27
Standard K–ω methods 1580 0.21
Shear stress Transport (SST) K–ω methods 1707 0.31

∂=∂X i ðρU i U j Þ ¼ ∂P=∂X i þ∂τi;j =∂X i ; i ¼ 1; 2; 3; j ¼ 1; 2; 3 ð2Þ

The commercial CFD software, FLUENT is used to solve the above equations as the flow along the rotating blade channels
is turbulent and complex. To model the turbulences in core flow (far from the walls), standard and RNG K–ε methods,
standard and Shear stress Transport (SST) K–ω methods were studied for each inlet size at the best performance point. It is
observed that the variation in the results, viz., efficiency and performance were not significant as mentioned in Table 2,
however, the computational time varied significantly for each case. In addition, RNG k–ε method is appropriate for low
Reynolds Number flows with high rotation flows, but the accuracy of the results and computational time was observed to be
insufficient when compared with the experimental results. Based on this study, the standard k–ε model was selected for
modeling turbulence. The turbulent kinetic energy (k) and rate of dissipation (ε), is derived from the following equations:

∂=∂X i ðρkU j Þ ¼ ∂=∂X i ½ðm þ mt =sk Þ∂k=∂X j  þ Gk –ρεþ c1 ð3Þ

∂=∂X i ðρεU j Þ ¼ ∂=∂X i ½ðm þ mt =sε Þ∂k=∂X j  þ 1:44ðε=kÞGk–1:92ρε2 =k þ c2 ð4Þ

Gk ¼ mt S2 ð5Þ

and C1 and C2 are user-defined source terms.

Eqs. (3)–(5) are applied to the core flow, but to consider the effect of viscous flow and no-slip condition at high Reynolds
Numbers near the walls, especially impeller and casing walls, on flow momentum, standard wall function approach was
selected. The following equations are applied for mentioned regions at any point, P near the wall:

U n ¼ ð1=0:4187ÞlnðEyn Þ ð6Þ

1=2
yn ¼ ð0:547ρkp yp Þ=m ð7Þ

In this approach, very fine meshes need not be employed at the wall to resolve the viscous sub layer and transition layer
(Yn o133) and hence significant computational time can be saved.
Also for the pressure-velocity coupling, SIMPLE algorithm is applied. For accuracy of results, second order-upwind
scheme is selected for discretization of related parameters with a convergence criterion of 1  10  3.
Mass flow rate of the fluid is specified at the inlet of the nozzle and pressure outlet boundary condition is selected at the
exit plane. Also Multiple Reference Frame technique is applied to model the rotating elements.
30 M. Gholamian et al. / Case Studies in Thermal Engineering 1 (2013) 26–37

3. Experimental validation

To validate the numerical model and to compare the results, experiments were conducted for the same physical
dimensions as mentioned in Table 1. To drive the fan, an electric motor of 0.6 kW is connected to impeller by rigid coupling
to enhance the mechanical efficiency. The rotational speed in all cases was maintained constant at 900 rpm.
To perform experiments and to obtain performance and efficiency curves the standard of ISO 5801 [20] was followed. The
arrangement of the experimental setup is shown in Fig. 3. The performance parameters are given by flow coefficient, head
coefficient and efficiency as presented in Eqs. (8)–(10).
φ ¼ Q =½ðπD2 NÞðπD2 bÞ ð8Þ

ψ ¼ 2ΔP t =½ρðπD2 NÞ2  ð9Þ

η ¼ ΔP t Q =P In ð10Þ

4. Results

The effect of axial gap size on efficiency with different inlet sizes is shown in Figs. 4–7. The figures show an excellent
matching between numerical and experimental results in terms of the magnitudes as well as trends for all the 12 case
studies considered in the present paper. Hence the numerical model has been validated and further studies were performed
with this model. In this case study, for convenient, four fans with inlet size of 19 cm, 17 cm, 16 cm and 20.5 cm are
designated as Fan1, Fan2, Fan3 and Fan4 respectively.
In Fig. 4, for Fan3, it is apparent that by increasing the axial gap from 4 mm to 12 mm, fan efficiency is improving.
The same trend is observed in Figs. 5 and 6 (Fan2 and Fan1) and the efficiency curves, for all axial gaps, are nearly similar
with the difference that, the larger inlet nozzle diameter has higher efficiency. In these cases, as shown in Figs. 4–6, the
effect of decreasing the axial gap from 8 mm to 4 mm with increasing of inlet diameter on efficiency is observed to be more
significant while by increasing the axial gap from 8 mm to 12 mm with increasing the inlet diameter has comparatively less
effect on efficiency.
For Fan1 with the inlet nozzle diameter of 19 cm, same as that of impeller, it is observed that the efficiency in all axial
gaps is nearly same and its magnitude is highest compared to all other cases considered for the analysis.

Fig. 3. Mesh pattern in impeller.

Fig. 4. Mesh number affect on efficiency of fan with 16 cm inlet diameter and 8 mm gap.
 M. Gholamian et al. / Case Studies in Thermal Engineering 1 (2013) 26–37 31

Fig. 5. Equipment arrangement and fan elements.

Fig. 6. Fan with 16 cm inlet diameter: (a) performance and (b) efficiency.

Fig. 7. Fan with 17.5 cm inlet diameter: (a) performance and (b) efficiency.

For Fan4, in Fig. 7, there is an opposite trend i.e. the influence of decreasing axial gap from 8 mm to 4 mm, increased the
efficiency slightly while increasing of this gap from 8 mm to 12 mm has decreased the efficiency and this variation is more
significant compared to increase of efficiency when the axial gap is decreased from 8 mm to 4 mm.
Finally it can be reached to this point that the nozzle with inlet diameter of 19 cm (near inner impeller diameter) has
special characters. For the inlet diameter less than this magnitude, with the increase of axial gap between inlet nozzle and
impeller, efficiency increases while for inlet diameter more than this magnitude, with the decrease of axial gap, efficiency
increases. This shows that the inlet nozzle can be considered as the basis for comparison with all other cases considered for
the analysis.
To study the distribution of static pressure, axial velocity and radial velocity, some contours at different axial positions
and also two areas as shown in Fig. 3 are selected where they are observed to be captured well. The Area1 represents the
space between 301 plane and 3451 plane which is small space under tongue region. The Area2 represents the space between
301 plane and 901 plane. This area is selected as the highest adverse pressure gradient can be observed which causes
32 M. Gholamian et al. / Case Studies in Thermal Engineering 1 (2013) 26–37

Fig. 8. Fan with 9 cm inlet diameter: (a) performance and (b) efficiency.
 M. Gholamian et al. / Case Studies in Thermal Engineering 1 (2013) 26–37 33

Fig. 9. Fan with 20.5 cm inlet diameter: (a) performance and (b) efficiency.
34 M. Gholamian et al. / Case Studies in Thermal Engineering 1 (2013) 26–37

a serious problem for exit flow of corresponding blades. Also some planes at different radii of 9 cm and 11.3 cm, at angle of
01 were considered to capture the variation of total pressures and energy transferred and the observations are presented as
follows.

5. Analysis

5.1. Flow pattern

In squirrel cage fans, exit flow from inlet nozzle is almost axial with high magnitude of velocity; hence flow directly
enters into the impeller. In impeller as the fluid particles move further, gradually their axial velocity component reduces as
shown in Fig. 8a–c and radial velocity component increases along the axis of impeller as shown in Fig. 8d–f.
In this process, since axial velocity towards the end of impeller (negative Z direction) is reduced, the static pressure in
impeller has a positive gradient from inlet of the impeller towards the end plate. Due to high adverse pressure gradient in
Area2, the static pressure at inside of blade channels and at some parts of impeller is influenced by this pressure and the
pressure at these locations is increased as shown in Fig. 9a–c. This event starts even from inside of the inlet nozzle and will
continue up to a point in the impeller where the static pressure (maximum static pressure) is equal to influencing pressure
as shown in Fig. 9b. After this point, both pressures (influencing pressure and pressure inside the impeller) are equal around
inner impeller radius. With decreasing of Z, pressure inside the impeller will increase and static pressure distribution will be
more uniform and proper as shown in Fig. 9c. The static pressure distribution further affects the radial and axial velocity
distribution as observed from the similar contours in the respective cases as shown in Figs. 8 and 9.

5.2. Losses

Squirrel cage fans with forward blades are reaction type blades, so unlike other turbo machines, the static pressure of
fluid is not increased in the blade passages, but it is increased in volute casing. Under such circumstances, due to high blade
exit velocity, if there be an inappropriate space between the impeller and casing at the exit of impeller or faulty volute
design, pressure recovery cannot be achieved properly and turbulences and losses increase significantly. As it is shown in
Fig. 9, in Area1 and even in the space under that, the conversion of dynamic pressure into static pressure (pressure recovery)
is not complete and normally the pressure at this space is same as the pressure at blade exit. Also in the space between
tongue and blades at 301, due to narrow space, leaking of flow from tongue upper side to Area1 is not easy and there is a low
amount of leakage. But leaking of flow with rotation of corresponding blades and also fluctuating axial velocity at the
impeller inlet in Area1 can be the cause of intensive turbulent and vortex flow creation.
Since particle velocity at the blade exit is too high and there is lower pressure on suction side compared to that of
pressure side, there will be a change in relative velocity angle at exit and related problems like slip, jet and wakes [6]. At first
part of the impeller, axial velocity is high and since, pressure at pressure side is more than that of suction side, so there are
reverse flow due to difference in the pressure of suction side and that of pressure side. By decreasing Z, axial velocity

Fig. 10. Axial and radial velocity contours of Fan3 with axial gap of 8 mm. (a) Axial velocity at Z¼ 5 cm, (b) axial velocity at Z¼8 cm, (c) axial velocity
at Z¼ inlet exit, (d) radial velocity at Z¼5 cm, (e) radial velocity at Z¼8 cm and (f) radial velocity at Z¼ inlet exit.
 M. Gholamian et al. / Case Studies in Thermal Engineering 1 (2013) 26–37 35

magnitude decreases while the pressure difference between the pressure and suction sides of blades is increasing. Under
this condition the angle of relative velocity at entry of blades at all axial positions is changing and there is positive incidence.
In Area2 due to adverse pressure there is reverse flow that is coupled with fluctuating axial velocity distribution and
blade rotation that result into intensive turbulent and vortex flows in this area and some parts in the inside of impeller as
shown in Fig. 10. Along the blades with reducing Z, gradually axial velocity is reduced and radial velocity distribution is more
uniform.

5.3. Fan mechanisms

In the present paragraph, an attempt is made to analyze the resulted flow pattern with respect to the cases considered
for the analysis, focusing on the effect of axial gap. Fan3 is considered for presenting this analysis, as the effect of axial gap
on efficiency is more significant compared to the other three fans. Fan3 has the smallest inlet diameter. So flow in this case is
guided into impeller with the highest axial velocity and lowest static pressure. Also, in this fan, in cases of three axial gap
sizes of 4 mm (near impeller), 8 mm (at the middle of the available gap) and 12 mm (almost far from impeller), all the
patterns of static pressure distribution in different axial positions are similar, with the only difference that the static
pressure pattern in the case of axial gap of 4 mm in comparison with the case of 8 mm is achieved at lower Z, implying that
for an axial gap of 8 mm, the fluid particles took smaller axial distance along impeller for achieving the required static
pressure distribution. Similarly when the axial gap is increased from 8 mm to 12 mm, the static pressure pattern is further
retarded with respect to Z. Since static pressure distribution has a great effect on axial and radial velocity and flow pattern
follows the static pressure distribution, these velocities in the case of axial gap of 12 mm, 8 mm and 4 mm reaches to the
uniform and sufficient distribution at a lower axial position respectively.
Totally, for Fan 3, at the same Z, pressure recovery in Area1 is lowest for an axial gap of 4 mm and highest for an axial gap
of 12 mm. In Area2 that includes some number of blades, some part of volute and impeller inside, the influence of adverse
pressure gradient and reverse flow is highest for an axial gap of 4 mm and lowest for the axial gap of 12 mm. Consequently
due to these losses, performance and efficiency of Fan3 in case of 12 mm axial gap are the highest and that of 4 mm is the
lowest.
The reason for these differences is due to the inlet performance at different axial positions. As shown in Fig. 11a with less
axial gap, it can be observed that at the first part of impeller, fluid particles entering with high axial velocity do not have
enough time to reach and adapt themselves to the blade entry and impeller conditions or change their direction to form
radial velocity. Due to this, in these areas (space between points A and B) blades are faced with stalling, unsteadiness and

Fig. 11. Static pressure contours of Fan3 with axial gap of 8 mm. (a) At Z¼ 5 cm (b) at Z ¼7.7 cm, and (c) at Z¼inlet exit.
36 M. Gholamian et al. / Case Studies in Thermal Engineering 1 (2013) 26–37

Fig. 12. Turbulent kinetic energy contours of Fan3 with axial gap of 8 mm at Z¼ 8 cm.

turbulences which are the causes of losses. But in the case of more axial gap, fluid particles have more time compared to the
case of less axial gap, and they pass less axial distance and reach blade entry in higher Z as shown in Fig. 11b. Hence, the
space between points A and B that is influenced by turbulent and vortex flows, is smaller and the effective blade area
transferring energy to flow is more compared to the case of smaller axial gap. This effect is represented in Fig. 12.
This figure shows that the energy transfer and work done by the impeller, along the Z axis at different radial positions of
9 cm and 11.3 cm at the angle of 0o, is less at the first sections of impeller as observed by the total pressure difference
between the two radial positions considered.
In the cases of Fan2 and Fan1, with the increase of the inlet diameter, the static pressure increases and axial velocity
reduces. Here, since the radial distance between inlet nozzle and inner impeller diameter reduces, so the effect of having
enough time for fluid particles to form radial velocity and adapt themselves to impeller conditions will be less significant.
Hence as it is shown in Figs. 4–6 the effect of axial gap on performance and efficiency reduces from Fan3 to Fan1 gradually.
In Fan4, since inlet nozzle diameter is more than inner impeller diameter, so flow directly reaches the blades axially.
Here, by increasing the axial gap between inlet nozzle and impeller, velocity profile during this gap will be distorted. Hence
turbulences, unsteadiness and losses increase as shown in Fig. 7.

6. Conclusions

Simulations were done on forward squirrel cage fans to find the effect of axial gap size on efficiency with different inlet
nozzle diameters. Results showed a good match between experimental data and numerical simulations. The numerical
simulations showed that Fan2 with 19 cm inlet diameter (nearly equal to inner impeller diameter) has special
characteristics. Fan2 has exhibited highest efficiency of nearly 44% at all axial of gaps, compared to all other cases
considered in the analysis. With the inlet nozzle diameter of 19 mm as the basis, with decreasing the inlet diameter, the
effect of increasing the axial gap up to 12 mm on performance and efficiency is more significant towards improvement of
efficiency, while by increasing inlet nozzle diameter, the effect of increasing the axial gap up to 12 mm on performance and
efficiency is more significant towards decreasing the efficiency.
For inlet nozzle size of 19 cm and smaller, flow patterns for the axial gaps considered in the analysis are similar, but with
the difference that flow patterns are less uniform with decrease of axial distance relative to impeller entry. The reason for
this difference and also for its low efficiency is found to be the high axial velocity of flow at the impeller entry and not
having enough time to convert into radial velocity of fluid particles and adapting them with impeller conditions.

References

[1] Gessner FB. An experimental study of centrifugal fan inlet flow and its influence on fan performance. ASME Paper no. 67; 1967.
[2] Kind RJ. Prediction of flow behavior and performance of squirrel-cage centrifugal fans operating at medium and high flow rates. Journal of Fluids
Engineering 1997;119:639–47.
[3] Khelladi S, Kouidri S, Bakir F, Rey R. Flow study in the impeller–diffuser interface of a Vaned centrifugal fan. ASME Journal of Fluid Engineering
2005;127:495–502.
[4] Suzuki S, Ugai Y. Study on high specific speed airfoil fans. Bulletin of the Japan Society of Mechanical Engineers 1977;20:575–83.
[5] Lee YT. Impact of fan gap flow on the centrifugal impeller aerodynamics. Journal of Fluids Engineering 2010;132:1–9.
[6] Meakhail T, Park SO. A study of impeller–diffuser9volute interaction in a centrifugal fan. ASME Journal of Turbomachinery 2005;127:84–90.
[7] Kim KY, Seo SJ. Shape optimization of forward-curved-blade centrifugal fan with Navier–Stokes analysis. ASME Journal of Fluid Engineering 2004;126:
735–42.
[8] Ballesteros-Tajadura Rafael, Velarde-Suarez Sandra, Hurtado-Cruz Juan Pablo. Noise prediction of a centrifugal fan: numerical results and experimental
validation. Journal of Fluids Engineering 2008;130:1–11.
[9] Velarde-Suarez S, et al. Relationship between volute pressure fluctuation pattern and tonal noise generation in a squirrel-cage fan. Applied Acoustics
2009;70:1384–92.
[10] Montazerin N, Damangir A, Mirzaie H. Inlet induced flow in squirrel-cage fans. Proceedings of the Institution of Mechanical Engineers—Part A: Journal
of Power and Energy 2000;214:243–53.
[11] Morinushi K. The influence of geometric parameters on F.C centrifugal fan noise. Journal of Vibration, Acoustics, Stress and Reliability in Design
1987;109:227–34.
[12] YAHYA. Turbines, compressors and fans. 3rd ed.McGraw Hill, New Delhi; 2009 (seventh reprint).
[13] Ouyang H. Internal flow and noise investigations about the cross-flow fan with different blade angles. Journal of Turbomachinery 2012;134:
051023–32.
 M. Gholamian et al. / Case Studies in Thermal Engineering 1 (2013) 26–37 37

[14] Walsh PA, Waslsh EJ, Grimes R. Viscous scaling phenomena in miniature centrifugal flow cooling fans: theory, experiments and correlation. ASME
Journal of Electronic Packaging 2010;132:1–8.
[15] Neise W. Review of noise reduction methods for centrifugal fans. Journal of Engineering for Industry 1982;104:151–61.
[16] Raj D, Swim WB. Measurements of the mean flow velocity and velocity fluctuation at the exit of an FC centrifugal fan rotor. Journal of Engineering for
Power 1981;103:393–8.
[17] Gui Y, Xi P. Study on the methods of determining main geometric parameters of centrifugal fan impeller. IEEE 2010;4:42–6.
[18] Phelan JJ, Russell SH, Zeluff WC. A study of the influence of Reynolds number on the performance of centrifugal fans. Journal of Engineering for Power
1979;101:670–6.
[19] Bhzadmehr A, Mercadir Y, Galanis N. Sensitivity analysis of entrance design parameters of a backward-inclined centrifugal fan using DOE method and
CFD calculations. Journal of Fluids Engineering 2006;128:446–53.
[20] Method of testing fans for general purpose including mine fans part 1. Performance, code B.S.848, Part1. 2nd edition. British Standards Institution
Incorporated by Royal Charter: London; 1963.