Civil Engineering Reference
In-Depth Information
[MAR 10] use two hot wires, one of which is fixed at
1/ 2
and the other has a variable position within a
zone commencing near to the wall, and running to
+
y
=
3.9
Re
τ
M
y
+
at
.
The
correlation
between
Re
τ
=
7,300
(
)
(
)
+
+
+
+
+
and
the
very
large-scale
ut euyt e
τ
;
≡
;
,
OM
L
M
τ
(
)
component of the local signal
+++
is maximal when we
have a shift corresponding to the angle of tilt of the
structures [BRO 77]. The coefficient
L
uyt
;
()
+
corresponds to
α
y
()
the maximum value of the correlation between
OM
ut
+
+
and
(
)
()
L
uyt
+++
. The optimum parameter
;
β
y
+
is determined
(
)
such that the universal component
+++
, calculated on
the basis of equation [6.4], contains no amplitude
modulation. The set of these quantities arising from the
procedure that we have just described is used to predict the
statistics of the fluctuations
uyt
*
,
(
)
uy
++
at any Reynolds
number, by using equation [6.4]. Given that these
parameters are supposed to be universal, the only “input”
necessary to determine
,
(
)
y
+ +
uy
++
in the range
0
,
<<
at a
M
(
)
given
Re
τ
+ +
. Figure 6.23
compares the predicted distributions of the turbulent
intensity
uu
+
and the skewness factor
is the temporal series
O
ut
τ
;
()
3/2
3 2
u
Suu
with the
=
measurements in the range
2,800
. The
accordance is more than satisfactory. The profiles predicted
at
Re
<<
19, 000
are also shown in Figure 6.23, but are not
compared against the measurements (for which no data are
available) in the inner layer at this very high Reynolds
number.
6
Re
τ
=×
1.4
10
Search WWH ::
Custom Search