Experimental study of thermal mixing layer using variable
temperature hot-wire anemometry
Received: 14 March 2013 / Revised: 5 August 2013 / Accepted: 4 September 2013 / Published online: 25 September 2013
Ó Springer-Verlag Berlin Heidelberg 2013
Abstract The buoyancy effects on the development of
the thermal mixing layer downstream from a horizontal
separating plate were studied by comparing stable and
unstable counter-gradient conﬁgurations. In this study, the
novel experimental technique called parameterizable con-
stant temperature anemometer, proposed by Ndoye et al.
(Meas Sci Technol 21(7):075401, 2010), was improved to
make possible the simultaneous measurement of tempera-
ture and two velocity components with an x-wire probe.
The buoyancy effects on the ﬂow are discussed through the
transport equations of turbulent kinetic energy and tem-
perature variance. In view of the low Richardson numbers
at stake (Ri
\ 0.03), the buoyancy forces appeared logi-
cally to be quantitatively negligible compared to the main
driving forces, but such a low-energy forcing mechanism
was in fact sufﬁcient in unstable conﬁgurations to increase
the shear stress and the expansion rate of the mixing layer
signiﬁcantly, both phenomena being associated with
enhanced production of turbulence.
The thermal plane turbulent mixing layer is a free shear
ﬂow induced by two parallel incident streams with velocity
and temperature differences. This ﬂow exhibits certain
features, which make it an excellent representative model
of the complex ﬂows often encountered in various
industrial applications, especially when the velocity and
temperature proﬁles are in counter-gradient conﬁgurations.
Understanding of the inherent coupling between velocity
and temperature ﬁelds and the transfer mechanisms of heat
and momentum has been the focus of extensive investi-
gations. Improving understanding requires simultaneous
measurement of velocity and temperature with good spatial
and temporal resolution. Among the wide range of mea-
surement techniques used in the literature to obtain these
quantities, the most accurate are based on HWA
Corrsin (1947) and Perry (1982) developed a CTA
method based on the repetition of experiments with a hot-
wire probe operating at different overheat ratios. This
method allows the assessment of mean statistical quantities
such as second-order moment of temperature and velocity,
but not the instantaneous distribution of temperature and
velocity. This method was used by Fulachier and Dumas
(1976) in a turbulent boundary layer and by Farcy (1981)in
a slightly heated subsonic jet.
Sakao (1973), Chevray and Tutu (1972, 1978) and Blair
and Bennett (1987) proposed the use of one or more hot-
wire probes operating in CTA mode combined with a cold
wire operating in CCA
mode or the use of two parallel
wire probes with two different overheat ratios placed close
to each other. Different geometrical conﬁgurations of the
sensors have been proposed and used (Antonia and Bilger
1976; Beguier et al. 1978; Antonia et al. 1984). However,
some critical limitations of the performance of these
methods were explained by Lienhard and Helland (1989).
Indeed, this method lacks spatial resolution, and the wire
interference is not easy to ﬁlter out.
K. Sodjavi Á J. Carlier (&)
IRSTEA, UR TERE, 35044 Rennes, France
K. Sodjavi Á J. Carlier
enne de Bretagne, 35000 Rennes, France
Constant temperature anemometer.
Constant current anemometer.
Exp Fluids (2013) 54:1599