250
Z. ZARlC
Kim ri a/., 1968), as well as by Corino and Brodkey (1969), Grass (1971).
and Shlanchiauskas (1972) have been most helpful in revealing extremely
interesting structural features in the wall layers. The “bursts” of Kline,
which were found to be responsible for the largest part of turbulent energy
generation, were later interpreted by other investigators as containing not
only ejections of low momentum fluid from the layers closest to the wall.
but also inrushes of the high momentum fluid from the outer layers all the
way into the viscous sublayer. These two highly intermittent phases have
been found to play a leading role in the turbulence production mechanism.
Similar conclusions could be drawn from the studies of fluctuating turbulence
parameters at the wall, such as velocity gradient (Popovich, 1%9), shear
stress (Duhamel and Py, 1972), temperature (Meek and Baer, 1970). and
heat flux (Armistead and Keyes, 1968).
lntermittency is a phenomenon known to provoke departures from Gaussianity. It is also a phenomenon beginning to grow in importance in turbulence research since the time of the studies of Townsend (1949) and
Corrsin and Kistler (195S), who investigated turbulent-nonturbulent interfaces at the edge of a wake, a round jet, and a boundary layer. Kovasnay et
a/. (1970) made an important contribution to the study of intermittency of
the free boundaries by introducing conditional sampling and the averaging
technique which later was employed by many other investigators. This
technique enabled the detection of the apparent interaction between interface intermittency and large-scale fluctuations inside the inner layers, and
again this interaction is found to play a dominant role in turbulent energy
production. Quite recently, Blackwelder and Kovasznay (1972) found nonzero correlations between the interface intermittency and the turbulent
“bursts *’ in the buffer layer.
Conditional averaging has proven to be a very efficient technique in
analyzing intermittent, non-Gaussian statistical phenomena. It is therefore
natural to also employ it in the analysis of the innct layer intermittency. In
fact, Grass (1971) has employed conditional averaging in revealing the importance of the inrush phases in the wall vicinity. Wallace et al. (1972) and
Willmartli and Lu (1972) have used the technquc in determining the principal contributions of the intermittent phases to the Reynolds stresses. Gupta
et ul. (1971) and Blackwelder and Kaplan (1972) have combined the
technique with the use of digital analysis in investigating “bursting” phenomena in the wall layers.
In a previous study we also employed conditional sampling techniques in
analyzing wall turbulence phenomena (Zaric, 1972b,d). This paper presents
the results of further development of the conditional sampling and averaging
analysis as applied to the wall layers, including the viscous sublayer, in an
attempt to separate intermittent phases from the overall signal.
Z. ZARlC
Kim ri a/., 1968), as well as by Corino and Brodkey (1969), Grass (1971).
and Shlanchiauskas (1972) have been most helpful in revealing extremely
interesting structural features in the wall layers. The “bursts” of Kline,
which were found to be responsible for the largest part of turbulent energy
generation, were later interpreted by other investigators as containing not
only ejections of low momentum fluid from the layers closest to the wall.
but also inrushes of the high momentum fluid from the outer layers all the
way into the viscous sublayer. These two highly intermittent phases have
been found to play a leading role in the turbulence production mechanism.
Similar conclusions could be drawn from the studies of fluctuating turbulence
parameters at the wall, such as velocity gradient (Popovich, 1%9), shear
stress (Duhamel and Py, 1972), temperature (Meek and Baer, 1970). and
heat flux (Armistead and Keyes, 1968).
lntermittency is a phenomenon known to provoke departures from Gaussianity. It is also a phenomenon beginning to grow in importance in turbulence research since the time of the studies of Townsend (1949) and
Corrsin and Kistler (195S), who investigated turbulent-nonturbulent interfaces at the edge of a wake, a round jet, and a boundary layer. Kovasnay et
a/. (1970) made an important contribution to the study of intermittency of
the free boundaries by introducing conditional sampling and the averaging
technique which later was employed by many other investigators. This
technique enabled the detection of the apparent interaction between interface intermittency and large-scale fluctuations inside the inner layers, and
again this interaction is found to play a dominant role in turbulent energy
production. Quite recently, Blackwelder and Kovasznay (1972) found nonzero correlations between the interface intermittency and the turbulent
“bursts *’ in the buffer layer.
Conditional averaging has proven to be a very efficient technique in
analyzing intermittent, non-Gaussian statistical phenomena. It is therefore
natural to also employ it in the analysis of the innct layer intermittency. In
fact, Grass (1971) has employed conditional averaging in revealing the importance of the inrush phases in the wall vicinity. Wallace et al. (1972) and
Willmartli and Lu (1972) have used the technquc in determining the principal contributions of the intermittent phases to the Reynolds stresses. Gupta
et ul. (1971) and Blackwelder and Kaplan (1972) have combined the
technique with the use of digital analysis in investigating “bursting” phenomena in the wall layers.
In a previous study we also employed conditional sampling techniques in
analyzing wall turbulence phenomena (Zaric, 1972b,d). This paper presents
the results of further development of the conditional sampling and averaging
analysis as applied to the wall layers, including the viscous sublayer, in an
attempt to separate intermittent phases from the overall signal.