* '
258
%. ZARIC
0.4+--'
- -
T--'
1
1
I
I
1
- 3
- 2
- I
0
1
2
3
4
5
( U- Ui)/k
Pic;. 6. Probubility distributions of the whole signal (I), the inrush phase (2). the ejection
phare (3) and the nonconfamina1ed part of the signal (4). at Ys = 13.85.
earlier, the increased importance of the ejection phase leads to negative
skewness further from the wall.
'The intermittmcy factors
Y =
corresponding to the ejections (ye), the inrush phases (yl) and the aoncontaminated signal (1 - ye - y,) are presented in Fig. 7 as a function of Y +. The
variance ( K 2 ) of the whole signal could be given by
K Z = ( e k h + (%)I + (edo
where ( E ~ ) ~
= ys(K2)B for ejections; ( e r ) [ = y,(K2), for the inrush phases,
and (rr>O = (1 - yB - Y , ) ( K ~ ) ~
for the noncontaminated signal. The ( e k )
values corresponding to different phases are presented in Fig. 8 as a function
of Y
+
as an indication of the relative contribution of dflerent phases to the
turbulent energy at a g i n n Y+. It is seen that the contribution crf the
noncontaminated signal is approximately to y ' throughout the inner layers.
Very close to the wall, at Y+ = 2, the contribution of the inrush phases is
about 70 "/6 while that of the ejection phase is but a few perant. At about
Y + = 12, the contributions of both intermittent phases are equal. Further
on, the contribution of the inrush phases continues to diminish while that of
the ejections continues to rise.