Energy and Power Engineering, 2010, 2, 313-319
doi:10.4236/epe.2010.24044 Published Online November 2010 (http://www.SciRP.org/journal/epe)
Experimental Investigation of Boundary Layer
Characteristics on Blade Surface under Different Inlet
Flow Conditions
Xiangfeng Bo, Bo Liu, Pengcheng Zhao, Zhiyuan Cao
National Aerodynamics Lab. of Airfoil and Cascade, School of Engine and Energy,
Northwestern Polytechnical University, Xi’an, China
Received July 18, 2010; revised September 4, 2010; accepted October 24, 2010
Abstract
In this paper, an experimental study is conducted on cascade boundary layer under different inlet conditions.
New method is used to measure the total pressure in blade surface boundary layer directly using total pres-
sure probe. Total pressure in both suction and pressure surfaces are acquired at different inlet conditions by
changing incidence angle and inlet Mach number. In addition, a series of parameters related to boundary
layer characteristics are calculated. The objective of the experiment is to investigate the influence of inlet
flow conditions on them. The results indicate that influence of incidence angle is significant when other con-
ditions are the same. Displacement thickness, momentum thickness as well as other parameters display some
disciplines for variation. In contrast, inlet Mach number has only a small influence in that boundary layer
becomes a litter thinner with increasing Mach number. Comparisons of experimental results with theoretical
expectations demonstrate that the method in this experiment is effective and reliable.
Keywords: Boundary Layer, Cascade, Inlet Flow Conditions, Total Pressure Probe
1. Introduction
Modern developing trend for aero-engine is toward high
performance, wide range of work, high efficiency and
high thrust-to-weight ratio [1-3]. Compressor is a key
component of aero-engine which holds a large share in
size and weight. Therefore, improving the performance
of fan and compressor plays a vital role [4].
End-wall boundary layer, blade surface boundary layer
and their interaction lead to great flow loss. Inlet flow
conditions have an important impact for boundary layer
on its development, separation and transition [5-6].
Along history, wind tunnel experiment is always an im-
portant method for research. In early times only inlet or
outlet flow parameters are directly measured. This
method does no help to understand the specific loss
mechanism in cascade. With the variation of research
emphasis, the whole flow conditions in cascade need to
be investigated. Many scholars have already conducted
such experiments and provided some valuable results
[7-9].
In this paper, an experimental research on both suction
and pressure surface boundary layer is conducted on high
subsonic cascade wind tunnel. The impacts of different
inlet flow conditions on boundary layer characteristics
are discussed in-depth. Experimental investigation is
conducted to get detailed information about the velocity,
displacement thickness and momentum thickness of
boundary layer.
2. Design of Experiment and Experimental
Devices
2.1. Design of Experiment
Total pressure in boundary layer on both suction and
pressure surface is measured using total pressure probe.
The cascade wind tunnel is composed of gas source,
regulator section, convergence section, experimental
section, end wall boundary layer suction devices and
control system. The wind tunnel structure is showed in
Figure 1. Parameters of the tested cascade are in Table
1.
The total pressure probe can only measure one posi-

X. F. BO ET AL. 314
Table 1. Parameters of tested cascade.
Chord length 65 mm
Cascade pitch 43.92 mm
Inlet geometry angle 42º
Outlet geometry angle 78º
Stagger angle 27º
Figure 1. Sketch of cascade wind-tunnel.
tion each time and a large number of measuring points
must be selected. Measuring points are selected accord-
ing to practical requirements and experimental condi-
tions [10-11]. In order to understand the characteristics
of boundary layer along the entire blade surface, meas-
uring points are selected from leading edge to trailing
edge at a distance of 10% of chord length, which is
shown in Figure 2. Limited by the size of experimental
device, finally five measuring points are selected on suc-
tion surface from 50% to 90% of the chord length, while
7 points are chosen on pressure side from 10% to 70%
chord length. The cascade is linear so the air flow is
similar along the blade height. In this experiment meas-
uring points are chosen at 40% blade height. In order to
measure the thickness of boundary layer, pressure probe
has to move along the blade surface vertically. The
moving step is 0.1 mm each time.
2.2. Design of Total Pressure Probe
As shown in Figure 3, total pressure probe is installed in
position 1. Component 4 can revolve on component 6 so
that component 2 can move along the vertical direction
on blade surface. 3 is a rivet which can fix component 2.
Bolt 5 is supposed to fix 4. The whole device is installed
in the groove of cascade. During the experiment total
pressure probe is adjusted along the vertical direction of
blade surface. The first measuring position is determined
where probe touches the blade surface but does not have
elastic deformation. In order to ensure the accuracy of
probe moving, micrometer is used to determine the exact
location.
The total pressure probes are shown in Figure 4. Dif-
ferent probes are used to measure the boundary layer on
suction and pressure surfaces respectively. Please notice
that there is a turn of 3° at position 3 because the probe
needs some bending when it is close to blade surface.
Total pressure probes are calibrated in calibration wind
tunnel of Northwestern Polytechnical University with an
accuracy of 0.1 to ensure the accuracy of experiment.
3. Results and Discussion
Flow parameters in boundary layer are measured under
x(mm)
y(mm)
010 20 30 40 50 60 70
-10
-5
0
5
10
15
10%
20%
30% 40% 50% 60%
70%
90%
80%
Figure 2. Sketch of measuring positions on blade surface.
1
2
3
4
5
6
Figure 3. Pressure probe installation device.

X. F. BO ET AL. 315
1
2
3
Figure 4. Pressure probe. 1-pressure surface; 2-suction sur-
face.
different inlet conditions. Inlet Mach numbers are 0.60,
0.66 and 0.71 while incidence angles are 0º, 4º and −4º
respectively. Displacement thickness, shape factor and
momentum thickness were calculated in order to further
research the characteristics of boundary layer. Related
equations are as follows:
101
e
udz
U







01
ee
uu
dz
UU







1
H



3.1. Effects of Flow Conditions on Pressure Side
Boundary Layer
Figure 5 shows the displacement thickness, shape factor
and momentum thickness distribution on pressure side
under different inlet conditions. It can be seen from Fig-
ure 5(a) that the displacement thickness changes obvi-
ously when the incidence changes. Under 0º incidence
the displacement thickness gradually increases when
measuring point moves toward trailing edge. Under 4º
and −4º incidence, the displacement thickness shows a
downward trend after the first rise but finally rises again.
It can also be found that there is a slight decline in dis-
placement thickness when inlet Mach number rises.
Figure 5(b) displays the momentum thickness and shape
factor on pressure side. Although momentum thickness
has some variations, shape factor is small and almost
maintains the same value along the chord under the three
different inlet conditions. Thus a conclusion can be
drawn that there is no obvious boundary layer separation
on pressure surface under the inlet conditions that are
tested.
3.2. Effects of Inlet Flow Conditions on Suction
Side Boundary Layer
Figure 6 shows the boundary layer displacement thick-
ness and shape factor on suction side under different inlet
flow conditions. It can be seen from Figure 6(a) that the
influence of incidence on displacement thickness of
boundary layer is significant. Generally the displacement
thickness has a trend of growth when measuring point
moves toward trailing edge. On the other hand, it can be
seen that the change of displacement thickness under
different Mach number is much smaller. However, the
figure does suggest that there is a slight decline in dis-
placement thickness when inlet Mach number is increas-
ing.
Figure 6(b) is the shape factor and momentum thick-
ness of suction surface boundary layer under different
incidences. Shape factor is an important parameter which
suggests whether the flow is laminar or turbulence and
whether the boundary layer has separated. Shape factor
of laminar flow is larger than that of turbulence and ac-
cording to engineering experience, laminar flow begins
to separate when shape factor grows to 3.5 or larger
while turbulence boundary layer starts to separate when
it is larger than 2.2. As shown in these figures, the flow
has already turns into turbulence at 50% chord since the
shape factor is around 1.5 at that position. Under
0ºincidence, turbulent flow becomes instable at the posi-
tion about 70% chord length and boundary layer may
start to separate at 85% chord length. Under 4º incidence,
there is a sharp increase in shape factor at about 70%
chord length which may indicate a boundary layer sepa-
ration. When the incidence is −4º, boundary layer sepa-
ration may not happen on suction surface as the shape
factor maintains a low value all the time.
3.3. Effects of incidence angle on flow velocities
in the boundary layer
Since inlet Mach number in the test has only a small in-
fluence on boundary layer characteristics as demon-
strated before, in this part focus is mainly concentrated
on incidence angle. Figures 7 and 8 display the velocity
distribution in boundary layer under different incidences
with an inlet Mach number of 0.71. “h” represents the
vertical distance away from blade surface and “v” means
the velocity. “10%-90%” suggests the position of meas-
uring point from leading edge to trailing edge. Figure 7
is the velocity distribution on suction side. It is clear that

X. F. BO ET AL. 316
(a) (b)
Figure 5. Displacement thickness and shape factor on pressure side. (a) displacement thickness; (b) Shape factor and mo-
mentum thickness.
the boundary layer becomes thicker toward trailing edge
while the velocity in and out of it has a trend of decline.
Under 0º incidence, the boundary layer is 0.5 mm thick
at 50% chord length. It increases along the chord and
reaches 2.5 mm at 90% chord length. At the same time,
the velocity in and out of boundary layer decreases, with
the mainstream velocity dropping from 250 m/s at 50%
chord to 180 m/s at 90% chord. When the incidence is 4º,
the boundary layer evidently becomes thicker along the
whole chord than that under 0º incidence. It becomes
stable after 70% chord with the value about 2 mm, and
the mainstream velocity also turns to be stable with the
value about 200m/s. This phenomenon can be caused by
separation on suction surface at about 70% chord as
demonstrated in part 3.2. Under −4º incidence, boundary
layer thickness has only a slight variation along the chord,
almost all 1.2 mm, but mainstream velocity declines
gradually toward trailing edge. This agrees well with the
conclusion made before that separation may not happen
on suction side at negative incidence.
Figure 8 is the velocity distribution in boundary layer
on pressure surface. As shown in the figure that inci-
dence angle has a significant influence on boundary layer
thickness as well as the flow velocity in it. Unlike the
suction side, pressure surface boundary layer has a trend
of becoming thicker when incidence angle decreases. It

X. F. BO ET AL. 317
(a) (b)
Figure 6. Displacement thickness and shape factor on suction side. (a) displacement thickness; (b) shape factor.
can be seen that when incidence is 4º, the boundary layer
is about 0.75 mm thick, it grows to 1.3 mm under 0º and
1.5 mm under −4º. Another phenomenon can be found
that boundary layer thickness nearly remains the same
along the whole chord when inlet flow conditions are
fixed. This demonstrates a conclusion made before that
no obvious separation occurs on pressure surface under
tested inlet conditions. According to Figure 8(a), under
0º incidence, flow velocity near the wall declines mark-
edly from 205 m/s at 10% chord to 160 m/s at 70% chord
when measuring point moves toward outlet. When inci-
dence is 4º, mainstream velocity varies largely along
pressure surface with a range of 209 m/s to 188m/s. Un-
der −4ºincidence, flow velocity in boundary layer has a
varying range from 200 m/s to 165 m/s. Mainstream ve-
locity outside boundary layer has a trend of decline from
215 m/s to 200 m/s.
4. Conclusions
1) Inlet Mach number has only a small influence on
boundary layer characteristics. At different Mach num-
bers in the experiment, aerodynamic parameters display
almost the same distributions. Nevertheless, displace-
ment thickness does have a little decline when inlet
Mach number increases.
2) Incidence angle has significant influence on bound-
ary layer characteristics compared with Mach number.
At different incidences, displacement thickness, mo-
mentum thickness and shape factor all show diverse dis-

X. F. BO ET AL. 318
***********
*
*
*
*
*
v( m/s)
h(mm)
050100 150 200 250 300
0
0.5
1
1.5
2
2.5
50%
60%
70%
80%
90%
*
(a)
*****
**
*****
**
*
*
*
*
*
*
*
v(m/s)
h( mm)
050100 150200 250 300
0
1
2
3
4
5
50%
60%
70%
80%
90%
*
(b)
*
**
****
*
*
***
*
*
*
*
v(m/s)
h(mm)
50100 150 200 250 300
0
1
2
3
4
5
50%
60%
70%
80%
90%
*
(c)
Figure 7. Velocity distribution in suction side boundary
layer. (a) I = 0º, Ma=0.71; (b) I = 4º, Ma = 0.71; (c) I = −4º,
Ma = 0.71.
*
*
*
*
*
*
*
*
*
*
*
*
*
v( m/ s)
h(m m)
150 160 170 180 190 200 210 220
0
0.5
1
1.5
2
2.5
3
3.5
10%
20%
30%
40%
50%
60%
70%
*
(a)
*
*
*
*
*
*
*
*
*
v(m/s)
h(mm)
150 160170 180190 200210
0
0.25
0.5
0.75
1
1.25
1.5
1.75
2
10%
20%
30%
40%
50%
60%
70%
*
(b)
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
v(m/s)
h( mm )
160 170 180 190 200210 220
0
0.5
1
1.5
2
2.5
3
3.5
10%
20%
30%
40%
50%
60%
70%
*
(c)
Figure 8. Velocity distribution in pressure side boundary
layer. (a) (a) I = 0º, Ma=0.71; (b) I = 4º, Ma = 0.71; (c) I =
−4º, Ma = 0.71.

X. F. BO ET AL. 319
tributions on both suction and pressure surfaces. Under
negative incidence, no separation occurs on suction or
pressure surfaces. With incidence angle rising, separation
begins to take place on suction side, and separation hap-
pens earlier when incidence angle increases. On the other
hand, flow can always maintain on pressure surface
without obvious separation at tested conditions.
3) From the influence of incidence angle on flow ve-
locities in boundary layer, on suction surface, under
positive incidence boundary layer thickness begins to be
stabilized at about 70% chord, which can be caused by
separation starting at this position as mentioned in part
3.2. At negative incidence boundary layer thickness is
almost the same from inlet to outlet, which further
proves that no separation happens; on pressure surface,
boundary layer can almost maintain the same thickness
along the chord length at all the three incidences, which
is also consistent with the shape factor distribution ana-
lysed before.
Not only are these experimental results consistent with
each other, but they also agree well with theoretical pre-
dictions and practical experience thus they can get rea-
sonable explanations. As a result, a conclusion can be
safely made that the experimental method used in this
paper is effective and reliable. In the future, more in-
depth experiments of boundary layer can be conducted
using this method.
5. References
[1] S. C. Smith, “Use of Shere-Sentive Liquid Crytals for
Surface Flow Visualization,” Journal of Aircraft, Vol. 29,
No. 2, 1992, pp. 289-293.
[2] F. Albano, A. Auletta and F. de Gregorio, “An Expe-
rimental Study of Laminar Separation Bubbles on Airfoil
at Low Reynolds, ”AIAA2006-3529.
[3] M. Ali, “Aerodynamics of a Low-Pressure Turbine Air-
foil under Steady and Periodically Unsteady Conditions,”
Ph.D. Dissertation, University of Carleton, Locus, 2003.
[4] B. Liu and Y. G. Wang, “Effects of Inlet Flow Conditions
on Boundary Flow on Compressor Blade Surface,” Jour-
nal of Propulsion Technology, Vol. 20, No. 3, 1999, pp.
64-68.
[5] X. Y. Nan, B. Liu and J. Jin, “A Progressive
Optimization for the Inverse Design of Blade Profile
Based on Sequential Quadratic Programming,” Proceed-
ings of IMechE. Part C: Journal of Mechanical Engineer-
ing Science, Vol. 221, No. 1, 2007, pp. 23-31.
[6] R. D. Clark, “Boundary Layer Research Using the
Millersville University Tethered Balloon Facility,”
America Institute of Aeronautics & Astronautics, AIAA
99-3961.
[7] M. S. Zhao, T. G. Owano and C. S. Kruger, “Boundary
layer Diagnostics of an Atmospheric Pressure Plasma
Jet,” AIAA 2000-0202.
[8] B. Hollon and J. Jacob, “Experimental Investigation of
Separation on Low Pressure Turbine Blades,” AIAA
2001-0447.
[9] H. Bhanderi and H. Babinsky, “Improver Boundary
Layer Quantities in the Shock Wave Boundary Layer
Interaction Region on Bumps,” AIAA 2005-4896.
[10] Philip P. Geoghegan, William J. Growther, “Measure-
ment of Boundary Layer Velocity Profiles by Ultrasonic
Tomography for the Prediction of Flow Separation,”
AIAA 2006-2805.
[11] F. Chen, G.-J. Zhao, Y.-P. Song, et al., “Influence of
Incidences on the Distribution of Static Pressure of
Swept-Curved Compressor Cascades,” Journal of Pro-
pulsion Technology, Vol. 25, No. 6, 2004, pp. 521-525.
Nomenclature
b = chord
Ma = Mach number
u = axial velocity [m/s]
Ue = external axial velocity [m/s]
v = transversal velocity [m/s]
t = cascade pitch
β1k = inlet geometry angle
β2k = outlet geometry angle
βy = stagger angle
x = axial coordinate [mm]
y = transversal coordinate [mm]
z = vertical coordinate [mm]
h = vertical distance from blade surface [mm]
H = shape factor
i = inlet flow incidence [deg]
δ = boundary layer thickness [mm]
δ1 = displacement thickness [mm]
θ = momentum thickness [mm]