EXPERIMENTAL AND NUMERICAL INVESTIGATIONS ON
PASSIVE DEVICES FOR TIP-CLEARANCE INDUCED NOISE
REDUCTION IN AXIAL FLOW FANS
A. Corsini – B. Perugini – F. Rispoli
Department of Mechanics and Aeronautics, University of Rome “La Sapienza”, Rome, Italy
alessandro.corsini@uniroma1.it, bruno.perugini@uniroma1.it, franco.rispoli@uniroma1.it
A. G. Sheard – I. R. Kinghorn
Fläkt Woods Ltd, Colchester, Essex, UK
geoff.sheard@flaktwoods.com, iain.kinghorn@flaktwoods.com
ABSTRACT
This paper presents experimental and computational investigations on a family of axial fans,
in fully-ducted configuration, to investigate an improved blade tip concept designed to recuce
the tip-clearance induced noise. The nature of the flow mechanisms in the fan tip region is
correlated to the specific blade design features that promote a marked reduction of the fan
aero-acoustic signature in both tonal and broad-band noise components. The tip vortical flow
structures are characterised, and their role in creation of overall stage acoustic emissions
clarified. It was found that the tip geometrical modification markedly affects the multiple
vortex behaviour of leakage flow, by altering the near-wall fluid flow paths on both blade
surfaces. The rotor loss behaviour, in the blade tip region, was also discussed in order to assess
the effect of blade tip passive device onto the rotor efficiency.
NOMENCLATURE
k
turbulent kinetic energy
blade chord

l.e.
leading edge
PS
pressure side
p
static pressure
r
radius
SS
suction side
t
pitch
t.e.
trailing edge
Uc
casing relative peripheral velocity
v, w absolute and relative velocities
x, y, z Cartesian coordinates
Greek letters

turbulent dissipation rate

total loss coefficient, p0in  p0

efficiency

hub-to-casing diameter ratio
t
turbulent viscosity
2
0.5  win
1
i
s





absolute vorticity vector
absolute streamwise vorticity, s  (i  wi ) /(2 w ) 
blade solidity
global flow coefficient (annulus area-averaged axial velocity normalised by Uc)
rotor tipclearance
pressure rise coefficient (p/(0.5 U c2 ))
rotor angular velocity
Subscripts and superscripts
a, p, r axial, peripheral and radial
c
casing wall
h
hub wall
i
Cartesian component index
in
inlet section
s
streamwise component
pitch-averaged value
INTRODUCTION
Often in axial flow fans the design specifications demand large tip gap according to the
requirements of operating with variable stagger or pitch angles, e.g. cooling fans, or in some cases
for emergency operation at up to 400°C for two hours to extract smoke in the event of a fire, e.g.
ventilating fans. As well known, the tip clearance plays a detrimental role affecting the rotor aerodynamics [1-3], significantly contributing to the aero-acoustic signature of impeller in low speed
ventilating equipments. In this pictures the tip clearance flow is recognized to inflence the rotor
noise spectra by discrete frequency noise, due to periodic velocity fluctuation, and a broadband or
high-frequency noise, due to velocity fluctuation in the blade passage [4-6]. To this end there is a
strong motivation to look for deliberate aerodynamic design in order to minimize the negative
effects of tip gap and to manage the fan or compressor tip clearance flow to minimize its impact on
performance. Thus techniques and concepts that help to reduce tip clearance noise without
sacrificing aerodynamic efficiency are highly desired and needed.
By surveying the techniques for noise control in fans and compressors, it was found that the
solutions proposed could be grouped into active and passive noise control techniques, conceptually
designed to accomplish this goal by reducing the leakage flow rate or by enhancing the primarysecondary flow momentum transfer.
In the ambit of active control techniques for fans and compressors, recently a number of
experimental studies reconsider the control of tip clearance flow by means of fluid injection on the
casing wall in axial compressor [7], and low-speed axial flow fan [8].
As far as the passive control techniques are concerned, the literature review puts in evidence the
role of three approaches respectively focused on three-dimensional blade design and on geometrical
modifications of the equipments in the gap region. The first concept makes use of sweep technique
in blade design, recognized as a remedial strategy to improve the aerodynamic limits in compressor
and low-speed axial fan rotors owing to the capability of affecting the rotor stall margin by reducing
secondary flows effects and the flow leakage by unloading the blade tip [9-11].
A second family of control technique is based on gap geometrical manipulation and makes use of
casing treatments in the shroud portion, over the blade tip, which is reported since the early 70s to
improve the stable flow range by weakening the tip leakage vortex. Noticeable contributions deal
with the use of grooves and slots [12-13], or stepped tip gaps [14]. Furthermore, in the ambit of fan
technologies recirculating vanes, and annular rings have been proposed as anti-stall devices [15].
2
A final technique, appeared during the last decade, proposed the blade tip modifications by
means of anti-vortex appendices such as the end-plates investigated by Quinlan and Bent [6], or the
solutions recently proposed by industrial patents for ventilating fans [16-19].
In this ambit, the present paper aims to investigate on the use of profiled end-plates at the blade
tip [20]. The study focuses on a family of commercially available fans and compares the
aerodynamic and aeroacoustic performance of the datum blade against two improved tip geometries,
respectively with constant or variable thickness end-plate [21]. The paper reports on the
experimental and numerical assessment of the pay-off derived from the blade tip concept developed
at Fläkt Woods Ltd with respect to the aerodynamic performance of a class of low noise level
industrial fans.
The single rotor investigations are carried out at design operating condition for three
configurations of the six-blade axial flow fan under investigation, namely: the datum fan, coded
AC90/6; two fans modified by the adoption of tip features, respectively coded AC90/6/TF and
AC90/6/TFvte. The studies have been carried in ducted configuration, adopting a high tip pitch
angle configuration, i.e. 28 degrees, where the fan provides the higher static pressure and flow rate
of its operational range.
The comparative aerodynamic performance experiments have been carried out according to ISO
5801 for type D fully ducted configuration set-up. The noise performance test have been carried out
in accordance with the British Standard BS484, Part 2 for outlet noise hemispeherical measurement.
The fans have been tested employing a type A configuration, in the Fläkt Woods Ltd anechoic
chamber at the design operating conditions.
The tip flow characteristics are analysed by using a three-dimensional (3D) steady ReynoldsAveraged Navier-Stokes (RANS) formulation, with use of first order non-isotropic turbulence
closure successfully validated for fan rotor flows [22-23]. Despite the steady-state condition, the
RANS is considered an effective investigation tool for vortical structure detection [24]. The authors
adopt a parallel multi-grid (MG) scheme developed for the in-house finite element method (FEM)
code [25]. The FEM formulation is based on a highly accurate stabilized Petrov-Galerkin (PG)
scheme, modified for application to 3D with equal-order spaces of approximation.
By means of such a numerical investigation, the tip leakage flow structures of the fans are
analysed in terms of vortical structures detection, leakage flow energy and loss behaviours.
Emphasis is laid on the assessment of the benefits related to the improved tip geometry in terms of
efficiency and operating margin gains. The overall objective is to investigate, via steady
computational simulations, the technical merits of a passive control strategy for controlling the
leakage flow and reducing tip clearance vortex/stator interaction noise and rotor-tip self noise.
TEST AXIAL FLOW FANS
The present study was performed on a family of commercially available highly efficient cooling
fans. The in-service experiences indicated that this family of fans gives good acoustic performance
with respect to the state-of-the-art configurations. The investigated fans have six-blade unswept
rotor, with blade profiles of modified ARA-D geometry type originally designed for propeller
applications. The blade profiles geometry is given in Table 1, for the datum fan AC90/6 at the hub,
and tip sections respectively.
Table 1 AC90/6 fan family specifications. Blade profile geometry and rotor specifications.
3
AC90/6 fans
blade geometry
hub
tip
/ t
1.32
0.31
pitch angle (deg)
36
28
camber angle (deg)
46
41
fan rotor
blade number
6
blade tip pitch angle (deg)
28
hub-to-

tip diameter (mm)
0.22
900.0
 (% span)
rated rotational frequency (rpm)
1.0
900 – 935
The studied
blade
configurations, for datum and modified rotors, feature a high tip stagger angle, i.e. 28 degrees,
measured, as is customary in industrial fan practice, from the peripheral direction. This rotor angular
setting has been chosen in order to exploit operating points where the vortical flow near the rotor tip
dramatically affects the aerodynamic performance and noise characteristics of the fans under
investigation.
Test fan tip end-plates
The fan blades are drawn in Figure 1, together with a detailed view of blade tip for the datum
rotor, and the improved rotors developed for low noise emission labeled: AC90/6/TF and
AC90/6/TFvte. Fig. 1 compares, in a qualitative view not to scale, the thickness distributions of the
developed improved tip concepts against the datum base-line.
The improved blade tip geometry, for AC90/6/TF fans, was originally inspired by the technique
developed for tip vortex control and induced drag reduction by preventing 3D flows in aircraft
wings, also used as anti-vortex devices for catamaran hulls.
AC/90/6/TFvte rotor
AC/90/6/TF rotor
datum rotor
Fig. 1 Test fan rotor blades and tip end-plates (not to scale)
The tip blade section was modified by adding an end-plate along the blade pressure surface that
ends on the blade trailing edge with a square tail. By means of the introduction of the end-plate, the
blade section is locally thickened of a factor 3:1 with respect to the maximum thickness at the tip of
datum blade. According to the theory behind the end-plate design, this dimension was chosen as the
4
reference radial dimension of leakage vortex to be controlled that could be estimated in the range
0.2  0.1 blade span, as shown by former studies on rotors of axial compressor [26] and fan [11]. A
recent investigation, carried out by Corsini and co-workers [20], assessed the aerodynamics and
aeroacoustics gains of rotor AC90/6/TF with respect to the datum one. The numerical simulation
founded the evidences of a tip leakage vortex breakdown that affect the rotor AC90/6/TF at the
design operating condition. In order to correct this negative feature, responsible for a loss in fan
efficiency, Corsini in [21] recently proposed a blade tip geometry that exploits a variable thickness
distribution of the end-plate according to safe rotation number chord-wise gradient concept, here
labeled AC90/6/TFvte.
INVESTIGATION TECHNIQUES
Few words of introduction including the operating points (vedi TE06 paper)
Experimental procedure
???? from the introduction ???
The aerodynamic and noise performance tests were carried out at Fläkt Woods Ltd laboratory in
Colchester.
Numerical procedure
The Reynolds-averaged Navier-Stokes equations are solved by an original parallel Multi-Grid
Finite Element flow solver [25]. The physics involved in the fluid dynamics of incompressible 3D
turbulent flows in rotating frame of reference, was modelled with a non-linear k- model [27], here
used in its topology-free low-Reynolds variant. This turbulence closure has been successfully
validated on transitional compressor cascade flows, as well as high-pressure industrial fan rotors
[10, 23].
The numerical integration of PDEs is based on a consistent stabilised Petrov-Galerkin
formulation developed and applied to control the instability origins that affect the advectivediffusive incompressible flow limits, and the reaction of momentum and turbulent scale determining
equations. The latter ones, respectively, related to the Coriolis acceleration or to the
dissipation/destruction terms in the turbulent scale determining equations [22]. Equal-order linear
interpolation spaces (Q1-Q1) are used for primary-turbulent and constrained variables, implicitly
eliminating the undesirable pressure-checkerboarding effects. Concerning the solution strategy, a
hybrid full linear MG accelerator was built-in the in-house made overlapping parallel solver. The
Krylov iterations in the smoothing/solving MG phases are parallelized using an original additive
domain decomposition algorithm. The message passing operations were managed using the MPI
libraries. By that way, the fully coupled solution of sub-domain problem involves an efficient nonconventional use of Krylov sub-space methods. The preconditioned GMRes(5) and GMRes(50)
algorithms were respectively used as smoother and core solver.
Axial fan modeling and boundary conditions
The mesh has been built according to a non-orthogonal body fitted coordinate system, by
merging two structured H-type grid systems. The mesh in the main flow region, surrounding the
blade, and an embedded mesh in the tip gap region. The mesh has 1546858 nodes, respectively in
the axial, pitch, and span wise directions. In the axial direction the node distribution consists of
20%, 50% and 30% of nodes respectively upstream the leading edge, in the blade passage and
downstream of it. Moreover, there are 14 grid nodes to model the tip-clearance along the span. The
computational grid is illustrated, in Figure 2, providing detailed view at the tip of the mesh in
meridional and blade-to-blade surfaces.
The mesh has an adequate stretch toward solid boundaries, with the ratio of minimum grid
spacing on solid walls to mid-span blade chord set as 210-3 on the blade tip, casing wall, and blade
5
surfaces. The adopted grid refinement towards the solid surfaces controls the dimensionless distance
+ value about 1 on the first nodes row.
Fig. 2 Computational grid of fan rotor, mesh details in the tip gap region
Standard boundary condition set has been adopted, already used in recent numerical studies on
high performance fans [10-11].
The Dirichlet conditions for the relative velocity components are imposed at the inflow section
half a mid-span chord far upstream the leading edge. The velocity profile has been obtained from
flow simulation in an annular passage of identical hub-to-casing diameter ratio that includes an
upstream spinner cone. The inlet distribution of the turbulent kinetic energy k is obtained from axisymmetric turbulence intensity (TI) profile derived on the basis of former studies on ducted
industrial fans [10]. The TI profile features a nearly uniform value in the core region (about 6
percent) and it grows markedly approaching the endwalls (about 10 percent). The inlet profile of
turbulence energy dissipation rate is basedon the characteristic length scale l set to 0.01 of rotor
pitch at mid-span. Flow periodicity upstream and downstream the blading, and Neumann outflow
conditions (homogeneous for k and  and non-homogeneous for the static pressure) complete the set
of boundary data.
The numerical investigation compares fan leakage flow patterns for datum, AC90/6/TF and
AC90/6/TFvte fan rotors operated in near-design condition (D) with volume flow rate 7 m3/s and
global flow coefficient = 0.278. The Reynolds number based on tip diameter and rotor tip speed
is 8.3 105, for normal air condition.
PERFORMANCE EXPERIMENTS
Aerodynamic tests
The aerodynamic tests were conducted according to ISO 5801 set up, for fully ducted
configuration and installation type D. This installation features ducted inlet and outlet regions and
the fan is supplied with a properly-shaped inlet bell mouth. The primary performance parameters
measured were the fan static pressure and the efficiency. Figure 3 compares the static pressure and
efficiency characteristic curves for datum, AC90/6/TF and AC90/6/TFvte rotors. The analysis of
static pressure curves of Fig. 3, gives the evidence of a small performance reduction in rotor
AC90/6/TF with improved tip concept, e.g. about 2% at 6m3/s, according to the interaction between
the tip clearance flow and the suction side near surface fluid [28]. This performance penalty is, on
the contrary, partially receovered by rotor AC90/6/TFvte whose static pressure rise increases
remarkably by throttling the rotor toward the peak pressure.
Concerning the efficiency curves it is evident that both the modified rotors feature an efficiency
improvement in the range of volume flow rate higher than the design one. Moreover the efficiency
6
curve comparison shows that the adoption of the tip end-plates results always in the appearance of a
clear efficiency plateau, that shift the peak  volume flow rate towards the rotor stall margin.
250
100
p(Pa)
95
)
datum
AC90/6/TF
AC/90/6/TFvte
200 90
85
80
150
250
75
70
100
200 65
60
55
150
50
80 50

45
100
0 40
704
5
6
7
8
35
50 30
102
103
104
60
0
4
50
5
6
7
volume flow (m3/s)
8
Fig. 3 Static pressure and efficiency characteristic curves (dashed lines: datum fan; solid lines:
AC90/6/TF fan; line-symbols: AC90/6/TFvte fan)
40
The rotor performance were assessed along the operating line. The predicted overall
performance for 900 rpm rotational frequency are compared in Table 2 to the experimental data
measured at Fläkt Woods
Ltd measurements according to the fan performance test standards
30
4
6
7
8
ISO5801:1997 for installations type5 D with inlet
bell-mouth.
Rotor
efficiency  is computed in
terms of static pressure rise. The comparison confirms the validity of the predicted performance at
the chosen setting angle, where the blades of this fan are the most loaded and more readily prone to
flow separation.
It is also worth noting that the prediction of performance parameters have been referred to axial
sections respectively located at the inlet of the domain, and 1.2 midspan chord downstream the
blade trailing edge. The comparison confirms the validity of the predicted performance.
Table 2 Predicted and measured fan overall performance @ 7 m3/s
Measurements
datum
AC90/6/TF
AC90/6/TFvte
Predictions
pstat (Pa)

pstat (Pa)

134.8
126.2
129.0
0.490
0.510
0.519
133.3
126.1
128.2
0.510
0.504
0.516
Noise tests
The noise performance test have been carried out in accordance with the British Standard
BS484, employing a type A testing configuration. In this method the fan is placed downstream of a
plenum chamber with a free outlet, in an arrangement similar to that used for compact cooling fans.
7
Figure 4 compares the measured power spectra in one-third-octave band. Fig. 4.a and Fig. 4.b
respectively show the measured sound power level and the A-weighted sound power level spectra
for the frequency-dependent human audition. The noise tests have been done in order to compare
the rotors aeroacoustic signature for identical static pressure rise, e.g. 190 Pa close to the peak
pressure operation. As shown in Fig. 4, the effectiveness of the improved tip concepts, with constant
and variable thickness end-plates, is demonstrated by the reduction of the rotor aeroacoustic
signature both in terms of tonal noise and broad-band noise. These noise components are related to
the main recognized tip noise generation mechanisms in axial decelerating turbomachinery. The
convection of the primary tip vortex and its interaction with the statoric structures produces mainly
tonal noise, while the oscillating tip vortex behaviour could be linked to the production of
broadband self-generated noise [29].
100
dB
100
dB(A)
datum
AC90/6/TF
AC/90/6/TFvte
95
90
95
datum
AC90/6/TF
AC/90/6/TFvte
90
85
80
85
75
80
70
75
65
60
70
55
65
50
45
60
40
55
35
50
10
2
10
3
frequency (Hz)
10
30
4
10
2
10
3
frequency (Hz)
10
4
Fig. 4 Sound power level spectra in one-third-octave band. a) un-weighted spectra and b) Aweighted spectra (dashed lines: datum fan; solid lines: AC90/6/TF fan; line-symbols:
AC90/6/TFvte fan)
Finally, Table 3 compares the overall sound power levels for the family of fans under
investigation.
Table 3 Overall sound power levels @ 190 Pa
datum
AC90/6/TF
AC90/6/TFvte
un-weighted SWL
dB
94.6
93.0
91.5
A-weighted SWL
dB(A)
88.7
86.3
88.1
RESULTS ANALYSIS
The improved blade tip concepts, developed with constant thickness and variable thickness endplates, has been shown to control the aeroacoustic signature of the family of fans under
investigations and this pay-off is correlated to a large efficiency improvement in the peak pressure
operating range. These experimental evidences prompted a comparative investigation, for the datum
fan and the fan with blade tip features, onto the inner workings of the passive control device used to
influence the structure of the tip leakage vortex and other systems of secondary vorticities. The
effectiveness of the passive tip devices is analysed by comparing the normalized streamwise
vorticity map evolution along the chord, and the tip vortex core paths. These analyses are
complemented with the evaluation of tip leakage flow energy contents, affecting the rotor
aeroacoustic signatures, and with the presentation of loss coefficient map evolution within the blade
8
passage to assess the influence of the geometrical modification at the blade tip and the rotor
aerodynamic efficiency along in operating range of the fans.
Tip leakage flow survey
The tip leakage vortical structures are now investigated by using the normalized helicity Hn
based on the absolute vorticity [3, 24] as the detection tool. Hn is defined and normalized as:
H n  i  wi  /   w  with i = 1, 3, where i and wi are the Cartesian components of the absolute
vorticity and relative velocity vectors,  and w their norms. Figure 5 shows the normalized helicity
distribution in the blade tip region by comparing the contours on cross flow planes in near-design
operating condition. The probing planes are located, respectively, at 0.25, 0.43, 0.65, 0.89 and 1.2
blade chord  from the tip section leading edge. The normalized helicity distribution is plotted with
the vortex cores colored by the its local magnitude.
For the investigated fan rotors, a clear vortex cores identification is only observed for the
leakage flow structures emerging in the front portion of the tip blade sections. In the multiple vortex
behaviour of datum fan rotor, Fig. 5.a, the helicity field shows that a main clock-wise vortical
structure (TLV1) develops through the passage with high skewing angle with respect to the blade
surface.
Fig. 5 Helicity contours
Moreover Fig. 5.a shows the existence of a weak tip secondary vortex, co-rotating with TLV1, in
the vicinity of the suction surface, as clearly shown on 0.65  plane by the streamwise vorticity
distribution. In the front portion of the blade, 0.25  downstream the leading edge, the helicity map
shows also the presence of a third vortical structure spreading from the leading edge of the blade. In
the blade aft region, the leakage flow is mainly characterized by the merging of tip separation vortex
TLV2 and leading edge vortex with TLV1 one, resulting in a unique clock-wise vortical structure
able to affect a large share of the blade pitch on the casing annulus.
As far as the AC90/6/TF rotor is concerned, Fig. 5.b gives evidence of a clear modification in the
tip leakage phenomenon. First of all, Fig. 5.b shows the existence of a vortex limiting the mass
leakage along the blade pressure surface, and recognized as the evidence of the pressure side leg of
a horse-shoe like structure. In the first quarter of the blade chord, it is evident the trace of an
additional leading edge vortex co-rotating with TLV1, in a similar way to datum rotor.
Both the TLV1 and the leading edge vortices feature, since their appearance on 0.25  plane, a
smaller in-passage extension coupled with a reduced helicity magnitude when compared to the
datum rotor field. Furthermore, the leading edge vortex never collapses into the TLV1, but it decays
as indicated on 0.43  plane. Moving downstream, about mid-chord, the TLV1 features a gradual Hn
reduction owing to the weakening of flow vorticity and to the deflection of vortex core.
This circumstance agrees with the hypothesis of mass leaking reduction along the chord, that
gives rise to a leakage flow structures nearly adjacent to the blade suction surface (i.e. TLV2
structure). In the aft portion of the blade, due to the leakage flow un-feeding the TLV1 collapses,
producing a bubble-type separation recognized as the evidence of a vortex breakdown by Corsini
and co-workers [20].
The separated flow turns into a counter-clock wise vortex under the influence of trailing edge
leakage flow streams, rapidly washing-out the vortex behind the rotor so that on 1.2  plane no
coherent vortical structure is evident.
Finally, Fig. 5.c shows that the use of variable thickness concept markedly influences the tip flow
features. The origin position of the main vortical structure TLV1 moves downstream with respect to
datum and AC90/6/TF blades, e.g. about 0.3  , and, as shown by the vortex core path, it develops
closer to the blade suction surface. At mid-chord the weakened TLV1 vortex interacts with highly
energetic leakage jet that causes the inversion of vortex rotation. In the aft chord fraction, the tip
9
vortex finally merges with a rotating cell and exits the blade passage appearing as a coherent clockwise vortical structure.
The helicity maps confirms, for this end-plate configuration, that a driving role is played by the
pressure side vortical structures as given by the horse-shoe vortex leg and by the corner vortex
developing under the end-plate.
Fig. 6 Tip leakage vortex cores and streamtraces
Fig. 7 Tip leakage flow a) velocity and b) skewing angle distributions
Fan tip losses
Rotor loss behaviors at the blade tip is discussed with reference to the local total loss coefficient
defined . The loss behaviour is, finally, investigated with reference to the local total loss
2
2
coefficient defined as:  = p0in  p0 0.5  win , where p0 is the local total pressure, p0in and 0.5  win are
respectively the reference pitch-averaged relative total and dynamic pressures computed at the inlet
mid-span plane.
Figure 9 shows the total loss coefficient distribution within the blade passage, probing the flow
fields in the vicinity of the blade leading edge, about mid-chord and in the region behind the blade
(i.e. respectively about 0.25 and 0.65 and 1.2 chord from the leading edge). In the design operating
condition, Fig.s 9.a, 9.b and 9.c, the predicted loss evolutions agree with the evidences found in
literature for low-speed rotors. At the rotor inlet, all rotor distributions feature loss cores mainly
concentrated on the hub annulus walls. Moving toward the blade aft, the loss maps are characterized
by loss core directly related to the development of primary tip vortices crossing the blade vane. As
shown in Fig. 9.b, the improved tip concept rotor AC90/6/TF, owing to the vortex breakdown, is
affected by a larger peak loss core. Nonetheless, by comparing the loss map on 1.2  plane with the
datum rotor one, it is clearly evident that the improved tip rotor presents a spanwise beneficial loss
distribution, outperforming the datum one within the wake and on the hub end-wall on pressure and
suction side corners.
Fig. ???????????
As far as the AC90/6/TFvte fan rotor is concerned, Fig. 9.c gives the evidence that the end-plate
design concept [21], by controlling the appearance of leakage vortex breakdown is able to reduce
the high loss core at the tip, behind the rotor and along the blade suction side within the interaction
region between leakage and near surface flows. This comparative behaviour, in agreement with the
aerodynamic test results, could be considered as a consequence of the reduced 3-D flow rearrangement consequent to the reduction in the mass leaking through the tip gap. The limited radial
migration of near-wall surface fluid, induces smaller hub loss core and the contraction of
suction/corner stall (shown by the hub corner stall core in Fig. 9.b and Fig. 9.c).
CONCLUSIONS
We have carried out a study on the structure of tip leakage flow and its influence on the fluid
dynamical behaviour in a family of axial flow fans. The aim of the paper has been to speculate on
the effectiveness of improved blade tip concepts to control the leakage flow phenomena and rotor
aeroacoustic signatures. Two end-plate geometries have been developed the first one with constant
thickness, and the second one with a variable thickness distribution according to safe rotation
number chord-wise gradient concept.
10
The aerodynamic tests have shown that the improved tip concepts are affected by a small
performance de-rating, but the efficiency curves give evidence of an improvement with increased
peak performance and wider high efficiency plateau towards the rotor stall margin. The noise test,
furthermore, demonstrated a reduction of the rotor aeroacoustic signature both in terms of tonal
noise and broad-band noise.
The investigation was based on an in-house developed parallel finite element Navier-Stokes
solver. The physical interpretation of the detailed 3D flow field predictions were discussed by means
of streamlines, streamwise vorticity, or leakage flow kinetic energy and loss maps. The following
conclusions are drawn on the basis of the found fluid dynamical behaviours.
The comparison of the detected leakage vortical structure evolutions showed that the datum rotor
features a multiple vortex behaviour characterized by a dominant leading edge vortical structure,
highly skewed with respect to the local relative flow direction, and a weak tip secondary vortex in
the vicinity of the suction surface. The presence of the end-plates influence the leakage flows
structure at the leading edge by changing its orientation with respect to the local relative streamlines
that govern the secondary flow advection in the rotor frame. Both the rotors designed with the
improved tip geometries appear to be characterized by two vortical structures at the tip, respectively
the trace of the pressure side leg of an incoming horse-shoe vortex like structure and the suction side
trace of the main leakage flow. In this viewpoint the existence of this pressure side vortex, peculiar
of the improved tip concepts rotor, is recognized as one of the factor contributing to the control of
the leakage phenomenon promoting a vena contracta effect. By that way the mass leaking is un-fed
close to the leading edge and the gained control on leakage flow onset turns into potentially
improved aeroacoustic performance.
The analysis of the leakage flow energy contents, indirectly related to the recognized tip leakage
noise mechanisms, has given the evidence of a reduction in the rotational kinetic energy and
turbulence level on the casing owing to the presence of the anti-vortex device.
The loss coefficient distributions confirm that the highest loss regions were always observed in
coincidence with the leakage vortex core with a nearly constant peak loss value. The comparative
analysis of mechanical energy loss within the gap showed that the presence of the anti-vortex device
at the tip leads to a reduction of mechanical energy loss within the gap, suggesting that the loss level
within the tip gap is mainly controlled by the mass leaking.
ACKNOWLEDGEMENTS
The present research was done in the context of the contract FW-DMA03, between Fläkt
Woods Ltd and Dipartimento di Meccanica e Aeronautica, University of Rome “La Sapienza”.
REFERENCES
See TE06 ???
11