Airfoil flow instabilities induced by background flow oscillations
W. C. Selerowicz, A. P. Szumowski
Abstract The effect of background ﬂow oscillations on
transonic airfoil (NACA 0012) ﬂow was investigated ex-
perimentally. The oscillations were generated by means of
a rotating plate placed downstream of the airfoil. Owing to
oscillating chocking of the ﬂow caused by the plate, the
airfoil ﬂow periodically accelerated and decelerated. This
led to strong variations in the surface pressure and the
airfoil loading. The results are presented for two angles of
attack, a ¼ 4° and a ¼ 8.5°, which correspond to the
attached and separated steady airfoil ﬂows, respectively.
Self-excited instabilities of airfoil ﬂow frequently appear in
the transonic ﬂow regime owing to shock waves and ﬂow
separations. This phenomenon leads to unsteady airfoil
loading and intense noise generation.
In the majority of papers, the airfoil ﬂow oscillations
were investigated for steady background ﬂows (Mabey
1989; Geissler 1993). However, several cases can be noted
where the airfoil ﬂow instabilities are induced by external
excitations, e.g. the ﬂow around the helicopter rotor blade,
the ﬂow in the turbine or compressor blade rows, etc. In
the former case, the instabilities appear owing to period-
ical changes of the relative ﬂow velocity ahead of the blade
as well as periodical variations in the blade angle of attack.
In the latter case, the blade ﬂow is disturbed by vortices
shed from upstream blades.
The present work concerns the airfoil ﬂow instabilities
generated by oscillatory varying background ﬂow velocity.
It is obvious that the strongest effect of external excitation
on the airfoil ﬂow can be expected in the airfoil trailing
section. This occurs owing to the positive pressure gradi-
ent existing downstream of the airfoil maximum thickness.
When the pressure gradient is strong enough, it induces
It is known that the separation bubble is very sensitive
to external excitation. Acceleration of excited ﬂow can
reduce or even cancel the separation, whereas deceleration
strengthens it. These effects are visible in ﬂow
photographs presented by Szumowski and Meier (1996).
Following this reference, the present work aims to inves-
tigate the effect of background ﬂow oscillations on the
airfoil loading resulting from unsteady airfoil surface
pressure distributions. Both the attached and separated
airfoil ﬂows are taken into account.
The transonic wind tunnel analogous to the one described
by Szumowski and Meier (1996) was used in the present
experiments (Fig. 1). The tunnel operated on a down-
stream vacuum maintained in a tank of 150 m
. Air was
sucked from a container made of impregnated fabric, in
which the air humidity was constant and equal to 12%.
The duration of the tunnel operation cycle was 0.5 s, and
for 0.3 s the ﬂow in the test section was stable. This time
was sufﬁcient to measure the ﬂow properties under
observation and store the results in a computer.
The NACA 0012 airfoil with achord length of 120 mm
was used in the experiments. Thirteen Kulite XCS 093
pressure transducers were placed inside the airfoil along
its side (Fig. 2) in the symmetry plane of the tunnel. Two
additional transducers were located at the side wall of the
test section: one ahead of the airfoil and the other at its
trailing edge. Owing to the symmetry of the airfoil, the
pressure distributions along its suction and the pressure
surfaces were measured with the transducers in their
constant positions. To this end, the angle of attack of the
airfoil was changed to the opposite value after the pressure
distribution had been measured. Then the pressure
distribution was measured again.
The background oscillations were generated by a plate
40 · 100 mm placed downstream of the airfoil. The plate
was rotated in the frequency range from 8 to 80 Hz by
means of an electric motor.
Preliminary investigation was performed to check
whether the ﬂow Mach number ahead of the airfoil de-
pends on the rotation of the plate. The results for two
angles of attack, a ¼ 4° and a ¼ 8.5°, obtained at a con-
stant opening of the adjusting valve of the tunnel are
displayed in Fig. 3. The empty and ﬁlled points in this
ﬁgure show the time mean values of maximal and minimal
ﬂow Mach numbers (M
) for several oscilla-
tion circles, respectively. Taking these points into account,
we obtained the averaged constant ﬂow Mach number and
its amplitude M ¼ 0.7 ± 0.05 for the considered airfoil
angles of attack and the rotation of the plate. The averaged
Reynolds number of the airfoil ﬂow was 2.2 · 10
Experiments in Fluids 32 (2002) 441–446 Ó Springer-Verlag 2002
Received: 6 June 2000 / Accepted: 18 October 2001
W. C. Selerowicz, A. P. Szumowski (&)
Warsaw University of Technology
ul. Nowowiejska 24, 00-665 Warszawa, Poland
This work was supported by State Committee for Scientiﬁc
Research in Poland (Project No 7 T07A 026 16).